x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

237 files changed, 35620 insertions, 0 deletions

diff --git a/arch/xtensa/Kbuild b/arch/xtensa/Kbuild
new file mode 100644
index 000000000..015baeb76
--- /dev/null
+++ b/arch/xtensa/Kbuild
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-y += kernel/ mm/ platforms/
diff --git a/arch/xtensa/Kconfig b/arch/xtensa/Kconfig
new file mode 100644
index 000000000..f2f9cd9cd
--- /dev/null
+++ b/arch/xtensa/Kconfig
@@ -0,0 +1,817 @@
+# SPDX-License-Identifier: GPL-2.0
+config XTENSA
+	def_bool y
+	select ARCH_32BIT_OFF_T
+	select ARCH_HAS_CPU_CACHE_ALIASING
+	select ARCH_HAS_BINFMT_FLAT if !MMU
+	select ARCH_HAS_CURRENT_STACK_POINTER
+	select ARCH_HAS_DEBUG_VM_PGTABLE
+	select ARCH_HAS_DMA_PREP_COHERENT if MMU
+	select ARCH_HAS_GCOV_PROFILE_ALL
+	select ARCH_HAS_KCOV
+	select ARCH_HAS_SYNC_DMA_FOR_CPU if MMU
+	select ARCH_HAS_SYNC_DMA_FOR_DEVICE if MMU
+	select ARCH_HAS_DMA_SET_UNCACHED if MMU
+	select ARCH_HAS_STRNCPY_FROM_USER if !KASAN
+	select ARCH_HAS_STRNLEN_USER
+	select ARCH_NEED_CMPXCHG_1_EMU
+	select ARCH_USE_MEMTEST
+	select ARCH_USE_QUEUED_RWLOCKS
+	select ARCH_USE_QUEUED_SPINLOCKS
+	select ARCH_WANT_IPC_PARSE_VERSION
+	select BUILDTIME_TABLE_SORT
+	select GENERIC_BUILTIN_DTB
+	select CLONE_BACKWARDS
+	select COMMON_CLK
+	select DMA_NONCOHERENT_MMAP if MMU
+	select GENERIC_ATOMIC64
+	select GENERIC_IRQ_SHOW
+	select GENERIC_LIB_CMPDI2
+	select GENERIC_LIB_MULDI3
+	select GENERIC_LIB_UCMPDI2
+	select GENERIC_PCI_IOMAP
+	select GENERIC_SCHED_CLOCK
+	select GENERIC_IOREMAP if MMU
+	select HAVE_ARCH_AUDITSYSCALL
+	select HAVE_ARCH_JUMP_LABEL if !XIP_KERNEL
+	select HAVE_ARCH_KASAN if MMU && !XIP_KERNEL
+	select HAVE_ARCH_KCSAN
+	select HAVE_ARCH_SECCOMP_FILTER
+	select HAVE_ARCH_TRACEHOOK
+	select HAVE_ASM_MODVERSIONS
+	select HAVE_CONTEXT_TRACKING_USER
+	select HAVE_DEBUG_KMEMLEAK
+	select HAVE_DMA_CONTIGUOUS
+	select HAVE_EXIT_THREAD
+	select HAVE_FUNCTION_TRACER
+	select HAVE_GCC_PLUGINS if GCC_VERSION >= 120000
+	select HAVE_HW_BREAKPOINT if PERF_EVENTS
+	select HAVE_IRQ_TIME_ACCOUNTING
+	select HAVE_PAGE_SIZE_4KB
+	select HAVE_PCI
+	select HAVE_PERF_EVENTS
+	select HAVE_STACKPROTECTOR
+	select HAVE_SYSCALL_TRACEPOINTS
+	select HAVE_VIRT_CPU_ACCOUNTING_GEN
+	select IRQ_DOMAIN
+	select LOCK_MM_AND_FIND_VMA
+	select MODULES_USE_ELF_RELA
+	select PERF_USE_VMALLOC
+	select TRACE_IRQFLAGS_SUPPORT
+	help
+	  Xtensa processors are 32-bit RISC machines designed by Tensilica
+	  primarily for embedded systems.  These processors are both
+	  configurable and extensible.  The Linux port to the Xtensa
+	  architecture supports all processor configurations and extensions,
+	  with reasonable minimum requirements.  The Xtensa Linux project has
+	  a home page at <http://www.linux-xtensa.org/>.
+
+config GENERIC_HWEIGHT
+	def_bool y
+
+config ARCH_HAS_ILOG2_U32
+	def_bool n
+
+config ARCH_HAS_ILOG2_U64
+	def_bool n
+
+config ARCH_MTD_XIP
+	def_bool y
+
+config NO_IOPORT_MAP
+	def_bool n
+
+config HZ
+	int
+	default 100
+
+config LOCKDEP_SUPPORT
+	def_bool y
+
+config STACKTRACE_SUPPORT
+	def_bool y
+
+config MMU
+	def_bool n
+	select PFAULT
+
+config HAVE_XTENSA_GPIO32
+	def_bool n
+
+config KASAN_SHADOW_OFFSET
+	hex
+	default 0x6e400000
+
+config CPU_BIG_ENDIAN
+	def_bool $(success,test "$(shell,echo __XTENSA_EB__ | $(CC) -E -P -)" = 1)
+
+config CPU_LITTLE_ENDIAN
+	def_bool !CPU_BIG_ENDIAN
+
+config CC_HAVE_CALL0_ABI
+	def_bool $(success,test "$(shell,echo __XTENSA_CALL0_ABI__ | $(CC) -mabi=call0 -E -P - 2>/dev/null)" = 1)
+
+menu "Processor type and features"
+
+choice
+	prompt "Xtensa Processor Configuration"
+	default XTENSA_VARIANT_FSF
+
+config XTENSA_VARIANT_FSF
+	bool "fsf - default (not generic) configuration"
+	select MMU
+
+config XTENSA_VARIANT_DC232B
+	bool "dc232b - Diamond 232L Standard Core Rev.B (LE)"
+	select MMU
+	select HAVE_XTENSA_GPIO32
+	help
+	  This variant refers to Tensilica's Diamond 232L Standard core Rev.B (LE).
+
+config XTENSA_VARIANT_DC233C
+	bool "dc233c - Diamond 233L Standard Core Rev.C (LE)"
+	select MMU
+	select HAVE_XTENSA_GPIO32
+	help
+	  This variant refers to Tensilica's Diamond 233L Standard core Rev.C (LE).
+
+config XTENSA_VARIANT_CUSTOM
+	bool "Custom Xtensa processor configuration"
+	select HAVE_XTENSA_GPIO32
+	help
+	  Select this variant to use a custom Xtensa processor configuration.
+	  You will be prompted for a processor variant CORENAME.
+endchoice
+
+config XTENSA_VARIANT_CUSTOM_NAME
+	string "Xtensa Processor Custom Core Variant Name"
+	depends on XTENSA_VARIANT_CUSTOM
+	help
+	  Provide the name of a custom Xtensa processor variant.
+	  This CORENAME selects arch/xtensa/variants/CORENAME.
+	  Don't forget you have to select MMU if you have one.
+
+config XTENSA_VARIANT_NAME
+	string
+	default "dc232b"			if XTENSA_VARIANT_DC232B
+	default "dc233c"			if XTENSA_VARIANT_DC233C
+	default "fsf"				if XTENSA_VARIANT_FSF
+	default XTENSA_VARIANT_CUSTOM_NAME	if XTENSA_VARIANT_CUSTOM
+
+config XTENSA_VARIANT_MMU
+	bool "Core variant has a Full MMU (TLB, Pages, Protection, etc)"
+	depends on XTENSA_VARIANT_CUSTOM
+	default y
+	select MMU
+	help
+	  Build a Conventional Kernel with full MMU support,
+	  ie: it supports a TLB with auto-loading, page protection.
+
+config XTENSA_VARIANT_HAVE_PERF_EVENTS
+	bool "Core variant has Performance Monitor Module"
+	depends on XTENSA_VARIANT_CUSTOM
+	default n
+	help
+	  Enable if core variant has Performance Monitor Module with
+	  External Registers Interface.
+
+	  If unsure, say N.
+
+config XTENSA_FAKE_NMI
+	bool "Treat PMM IRQ as NMI"
+	depends on XTENSA_VARIANT_HAVE_PERF_EVENTS
+	default n
+	help
+	  If PMM IRQ is the only IRQ at EXCM level it is safe to
+	  treat it as NMI, which improves accuracy of profiling.
+
+	  If there are other interrupts at or above PMM IRQ priority level
+	  but not above the EXCM level, PMM IRQ still may be treated as NMI,
+	  but only if these IRQs are not used. There will be a build warning
+	  saying that this is not safe, and a bugcheck if one of these IRQs
+	  actually fire.
+
+	  If unsure, say N.
+
+config PFAULT
+	bool "Handle protection faults" if EXPERT && !MMU
+	default y
+	help
+	  Handle protection faults. MMU configurations must enable it.
+	  noMMU configurations may disable it if used memory map never
+	  generates protection faults or faults are always fatal.
+
+	  If unsure, say Y.
+
+config XTENSA_UNALIGNED_USER
+	bool "Unaligned memory access in user space"
+	help
+	  The Xtensa architecture currently does not handle unaligned
+	  memory accesses in hardware but through an exception handler.
+	  Per default, unaligned memory accesses are disabled in user space.
+
+	  Say Y here to enable unaligned memory access in user space.
+
+config XTENSA_LOAD_STORE
+	bool "Load/store exception handler for memory only readable with l32"
+	help
+	  The Xtensa architecture only allows reading memory attached to its
+	  instruction bus with l32r and l32i instructions, all other
+	  instructions raise an exception with the LoadStoreErrorCause code.
+	  This makes it hard to use some configurations, e.g. store string
+	  literals in FLASH memory attached to the instruction bus.
+
+	  Say Y here to enable exception handler that allows transparent
+	  byte and 2-byte access to memory attached to instruction bus.
+
+config HAVE_SMP
+	bool "System Supports SMP (MX)"
+	depends on XTENSA_VARIANT_CUSTOM
+	select XTENSA_MX
+	help
+	  This option is used to indicate that the system-on-a-chip (SOC)
+	  supports Multiprocessing. Multiprocessor support implemented above
+	  the CPU core definition and currently needs to be selected manually.
+
+	  Multiprocessor support is implemented with external cache and
+	  interrupt controllers.
+
+	  The MX interrupt distributer adds Interprocessor Interrupts
+	  and causes the IRQ numbers to be increased by 4 for devices
+	  like the open cores ethernet driver and the serial interface.
+
+	  You still have to select "Enable SMP" to enable SMP on this SOC.
+
+config SMP
+	bool "Enable Symmetric multi-processing support"
+	depends on HAVE_SMP
+	select GENERIC_SMP_IDLE_THREAD
+	help
+	  Enabled SMP Software; allows more than one CPU/CORE
+	  to be activated during startup.
+
+config NR_CPUS
+	depends on SMP
+	int "Maximum number of CPUs (2-32)"
+	range 2 32
+	default "4"
+
+config HOTPLUG_CPU
+	bool "Enable CPU hotplug support"
+	depends on SMP
+	help
+	  Say Y here to allow turning CPUs off and on. CPUs can be
+	  controlled through /sys/devices/system/cpu.
+
+	  Say N if you want to disable CPU hotplug.
+
+config SECONDARY_RESET_VECTOR
+	bool "Secondary cores use alternative reset vector"
+	default y
+	depends on HAVE_SMP
+	help
+	  Secondary cores may be configured to use alternative reset vector,
+	  or all cores may use primary reset vector.
+	  Say Y here to supply handler for the alternative reset location.
+
+config FAST_SYSCALL_XTENSA
+	bool "Enable fast atomic syscalls"
+	default n
+	help
+	  fast_syscall_xtensa is a syscall that can make atomic operations
+	  on UP kernel when processor has no s32c1i support.
+
+	  This syscall is deprecated. It may have issues when called with
+	  invalid arguments. It is provided only for backwards compatibility.
+	  Only enable it if your userspace software requires it.
+
+	  If unsure, say N.
+
+config FAST_SYSCALL_SPILL_REGISTERS
+	bool "Enable spill registers syscall"
+	default n
+	help
+	  fast_syscall_spill_registers is a syscall that spills all active
+	  register windows of a calling userspace task onto its stack.
+
+	  This syscall is deprecated. It may have issues when called with
+	  invalid arguments. It is provided only for backwards compatibility.
+	  Only enable it if your userspace software requires it.
+
+	  If unsure, say N.
+
+choice
+	prompt "Kernel ABI"
+	default KERNEL_ABI_DEFAULT
+	help
+	  Select ABI for the kernel code. This ABI is independent of the
+	  supported userspace ABI and any combination of the
+	  kernel/userspace ABI is possible and should work.
+
+	  In case both kernel and userspace support only call0 ABI
+	  all register windows support code will be omitted from the
+	  build.
+
+	  If unsure, choose the default ABI.
+
+config KERNEL_ABI_DEFAULT
+	bool "Default ABI"
+	help
+	  Select this option to compile kernel code with the default ABI
+	  selected for the toolchain.
+	  Normally cores with windowed registers option use windowed ABI and
+	  cores without it use call0 ABI.
+
+config KERNEL_ABI_CALL0
+	bool "Call0 ABI" if CC_HAVE_CALL0_ABI
+	help
+	  Select this option to compile kernel code with call0 ABI even with
+	  toolchain that defaults to windowed ABI.
+	  When this option is not selected the default toolchain ABI will
+	  be used for the kernel code.
+
+endchoice
+
+config USER_ABI_CALL0
+	bool
+
+choice
+	prompt "Userspace ABI"
+	default USER_ABI_DEFAULT
+	help
+	  Select supported userspace ABI.
+
+	  If unsure, choose the default ABI.
+
+config USER_ABI_DEFAULT
+	bool "Default ABI only"
+	help
+	  Assume default userspace ABI. For XEA2 cores it is windowed ABI.
+	  call0 ABI binaries may be run on such kernel, but signal delivery
+	  will not work correctly for them.
+
+config USER_ABI_CALL0_ONLY
+	bool "Call0 ABI only"
+	select USER_ABI_CALL0
+	help
+	  Select this option to support only call0 ABI in userspace.
+	  Windowed ABI binaries will crash with a segfault caused by
+	  an illegal instruction exception on the first 'entry' opcode.
+
+	  Choose this option if you're planning to run only user code
+	  built with call0 ABI.
+
+config USER_ABI_CALL0_PROBE
+	bool "Support both windowed and call0 ABI by probing"
+	select USER_ABI_CALL0
+	help
+	  Select this option to support both windowed and call0 userspace
+	  ABIs. When enabled all processes are started with PS.WOE disabled
+	  and a fast user exception handler for an illegal instruction is
+	  used to turn on PS.WOE bit on the first 'entry' opcode executed by
+	  the userspace.
+
+	  This option should be enabled for the kernel that must support
+	  both call0 and windowed ABIs in userspace at the same time.
+
+	  Note that Xtensa ISA does not guarantee that entry opcode will
+	  raise an illegal instruction exception on cores with XEA2 when
+	  PS.WOE is disabled, check whether the target core supports it.
+
+endchoice
+
+endmenu
+
+config XTENSA_CALIBRATE_CCOUNT
+	def_bool n
+	help
+	  On some platforms (XT2000, for example), the CPU clock rate can
+	  vary.  The frequency can be determined, however, by measuring
+	  against a well known, fixed frequency, such as an UART oscillator.
+
+config SERIAL_CONSOLE
+	def_bool n
+
+config PLATFORM_HAVE_XIP
+	def_bool n
+
+menu "Platform options"
+
+choice
+	prompt "Xtensa System Type"
+	default XTENSA_PLATFORM_ISS
+
+config XTENSA_PLATFORM_ISS
+	bool "ISS"
+	select XTENSA_CALIBRATE_CCOUNT
+	select SERIAL_CONSOLE
+	help
+	  ISS is an acronym for Tensilica's Instruction Set Simulator.
+
+config XTENSA_PLATFORM_XT2000
+	bool "XT2000"
+	help
+	  XT2000 is the name of Tensilica's feature-rich emulation platform.
+	  This hardware is capable of running a full Linux distribution.
+
+config XTENSA_PLATFORM_XTFPGA
+	bool "XTFPGA"
+	select ETHOC if ETHERNET
+	select PLATFORM_WANT_DEFAULT_MEM if !MMU
+	select SERIAL_CONSOLE
+	select XTENSA_CALIBRATE_CCOUNT
+	select PLATFORM_HAVE_XIP
+	help
+	  XTFPGA is the name of Tensilica board family (LX60, LX110, LX200, ML605).
+	  This hardware is capable of running a full Linux distribution.
+
+endchoice
+
+config PLATFORM_NR_IRQS
+	int
+	default 3 if XTENSA_PLATFORM_XT2000
+	default 0
+
+config XTENSA_CPU_CLOCK
+	int "CPU clock rate [MHz]"
+	depends on !XTENSA_CALIBRATE_CCOUNT
+	default 16
+
+config GENERIC_CALIBRATE_DELAY
+	bool "Auto calibration of the BogoMIPS value"
+	help
+	  The BogoMIPS value can easily be derived from the CPU frequency.
+
+config CMDLINE_BOOL
+	bool "Default bootloader kernel arguments"
+
+config CMDLINE
+	string "Initial kernel command string"
+	depends on CMDLINE_BOOL
+	default "console=ttyS0,38400 root=/dev/ram"
+	help
+	  On some architectures (EBSA110 and CATS), there is currently no way
+	  for the boot loader to pass arguments to the kernel. For these
+	  architectures, you should supply some command-line options at build
+	  time by entering them here. As a minimum, you should specify the
+	  memory size and the root device (e.g., mem=64M root=/dev/nfs).
+
+config USE_OF
+	bool "Flattened Device Tree support"
+	select OF
+	select OF_EARLY_FLATTREE
+	help
+	  Include support for flattened device tree machine descriptions.
+
+config BUILTIN_DTB_NAME
+	string "DTB to build into the kernel image"
+	depends on OF
+
+config PARSE_BOOTPARAM
+	bool "Parse bootparam block"
+	default y
+	help
+	  Parse parameters passed to the kernel from the bootloader. It may
+	  be disabled if the kernel is known to run without the bootloader.
+
+	  If unsure, say Y.
+
+choice
+	prompt "Semihosting interface"
+	default XTENSA_SIMCALL_ISS
+	depends on XTENSA_PLATFORM_ISS
+	help
+	  Choose semihosting interface that will be used for serial port,
+	  block device and networking.
+
+config XTENSA_SIMCALL_ISS
+	bool "simcall"
+	help
+	  Use simcall instruction. simcall is only available on simulators,
+	  it does nothing on hardware.
+
+config XTENSA_SIMCALL_GDBIO
+	bool "GDBIO"
+	help
+	  Use break instruction. It is available on real hardware when GDB
+	  is attached to it via JTAG.
+
+endchoice
+
+config BLK_DEV_SIMDISK
+	tristate "Host file-based simulated block device support"
+	default n
+	depends on XTENSA_PLATFORM_ISS && BLOCK
+	help
+	  Create block devices that map to files in the host file system.
+	  Device binding to host file may be changed at runtime via proc
+	  interface provided the device is not in use.
+
+config BLK_DEV_SIMDISK_COUNT
+	int "Number of host file-based simulated block devices"
+	range 1 10
+	depends on BLK_DEV_SIMDISK
+	default 2
+	help
+	  This is the default minimal number of created block devices.
+	  Kernel/module parameter 'simdisk_count' may be used to change this
+	  value at runtime. More file names (but no more than 10) may be
+	  specified as parameters, simdisk_count grows accordingly.
+
+config SIMDISK0_FILENAME
+	string "Host filename for the first simulated device"
+	depends on BLK_DEV_SIMDISK = y
+	default ""
+	help
+	  Attach a first simdisk to a host file. Conventionally, this file
+	  contains a root file system.
+
+config SIMDISK1_FILENAME
+	string "Host filename for the second simulated device"
+	depends on BLK_DEV_SIMDISK = y && BLK_DEV_SIMDISK_COUNT != 1
+	default ""
+	help
+	  Another simulated disk in a host file for a buildroot-independent
+	  storage.
+
+config XTFPGA_LCD
+	bool "Enable XTFPGA LCD driver"
+	depends on XTENSA_PLATFORM_XTFPGA
+	default n
+	help
+	  There's a 2x16 LCD on most of XTFPGA boards, kernel may output
+	  progress messages there during bootup/shutdown. It may be useful
+	  during board bringup.
+
+	  If unsure, say N.
+
+config XTFPGA_LCD_BASE_ADDR
+	hex "XTFPGA LCD base address"
+	depends on XTFPGA_LCD
+	default "0x0d0c0000"
+	help
+	  Base address of the LCD controller inside KIO region.
+	  Different boards from XTFPGA family have LCD controller at different
+	  addresses. Please consult prototyping user guide for your board for
+	  the correct address. Wrong address here may lead to hardware lockup.
+
+config XTFPGA_LCD_8BIT_ACCESS
+	bool "Use 8-bit access to XTFPGA LCD"
+	depends on XTFPGA_LCD
+	default n
+	help
+	  LCD may be connected with 4- or 8-bit interface, 8-bit access may
+	  only be used with 8-bit interface. Please consult prototyping user
+	  guide for your board for the correct interface width.
+
+comment "Kernel memory layout"
+
+config INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	bool "Initialize Xtensa MMU inside the Linux kernel code"
+	depends on !XTENSA_VARIANT_FSF && !XTENSA_VARIANT_DC232B
+	default y if XTENSA_VARIANT_DC233C || XTENSA_VARIANT_CUSTOM
+	help
+	  Earlier version initialized the MMU in the exception vector
+	  before jumping to _startup in head.S and had an advantage that
+	  it was possible to place a software breakpoint at 'reset' and
+	  then enter your normal kernel breakpoints once the MMU was mapped
+	  to the kernel mappings (0XC0000000).
+
+	  This unfortunately won't work for U-Boot and likely also won't
+	  work for using KEXEC to have a hot kernel ready for doing a
+	  KDUMP.
+
+	  So now the MMU is initialized in head.S but it's necessary to
+	  use hardware breakpoints (gdb 'hbreak' cmd) to break at _startup.
+	  xt-gdb can't place a Software Breakpoint in the  0XD region prior
+	  to mapping the MMU and after mapping even if the area of low memory
+	  was mapped gdb wouldn't remove the breakpoint on hitting it as the
+	  PC wouldn't match. Since Hardware Breakpoints are recommended for
+	  Linux configurations it seems reasonable to just assume they exist
+	  and leave this older mechanism for unfortunate souls that choose
+	  not to follow Tensilica's recommendation.
+
+	  Selecting this will cause U-Boot to set the KERNEL Load and Entry
+	  address at 0x00003000 instead of the mapped std of 0xD0003000.
+
+	  If in doubt, say Y.
+
+config XIP_KERNEL
+	bool "Kernel Execute-In-Place from ROM"
+	depends on PLATFORM_HAVE_XIP
+	help
+	  Execute-In-Place allows the kernel to run from non-volatile storage
+	  directly addressable by the CPU, such as NOR flash. This saves RAM
+	  space since the text section of the kernel is not loaded from flash
+	  to RAM. Read-write sections, such as the data section and stack,
+	  are still copied to RAM. The XIP kernel is not compressed since
+	  it has to run directly from flash, so it will take more space to
+	  store it. The flash address used to link the kernel object files,
+	  and for storing it, is configuration dependent. Therefore, if you
+	  say Y here, you must know the proper physical address where to
+	  store the kernel image depending on your own flash memory usage.
+
+	  Also note that the make target becomes "make xipImage" rather than
+	  "make Image" or "make uImage". The final kernel binary to put in
+	  ROM memory will be arch/xtensa/boot/xipImage.
+
+	  If unsure, say N.
+
+config MEMMAP_CACHEATTR
+	hex "Cache attributes for the memory address space"
+	depends on !MMU
+	default 0x22222222
+	help
+	  These cache attributes are set up for noMMU systems. Each hex digit
+	  specifies cache attributes for the corresponding 512MB memory
+	  region: bits 0..3 -- for addresses 0x00000000..0x1fffffff,
+	  bits 4..7 -- for addresses 0x20000000..0x3fffffff, and so on.
+
+	  Cache attribute values are specific for the MMU type.
+	  For region protection MMUs:
+	    1: WT cached,
+	    2: cache bypass,
+	    4: WB cached,
+	    f: illegal.
+	  For full MMU:
+	    bit 0: executable,
+	    bit 1: writable,
+	    bits 2..3:
+	      0: cache bypass,
+	      1: WB cache,
+	      2: WT cache,
+	      3: special (c and e are illegal, f is reserved).
+	  For MPU:
+	    0: illegal,
+	    1: WB cache,
+	    2: WB, no-write-allocate cache,
+	    3: WT cache,
+	    4: cache bypass.
+
+config KSEG_PADDR
+	hex "Physical address of the KSEG mapping"
+	depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX && MMU
+	default 0x00000000
+	help
+	  This is the physical address where KSEG is mapped. Please refer to
+	  the chosen KSEG layout help for the required address alignment.
+	  Unpacked kernel image (including vectors) must be located completely
+	  within KSEG.
+	  Physical memory below this address is not available to linux.
+
+	  If unsure, leave the default value here.
+
+config KERNEL_VIRTUAL_ADDRESS
+	hex "Kernel virtual address"
+	depends on MMU && XIP_KERNEL
+	default 0xd0003000
+	help
+	  This is the virtual address where the XIP kernel is mapped.
+	  XIP kernel may be mapped into KSEG or KIO region, virtual address
+	  provided here must match kernel load address provided in
+	  KERNEL_LOAD_ADDRESS.
+
+config KERNEL_LOAD_ADDRESS
+	hex "Kernel load address"
+	default 0x60003000 if !MMU
+	default 0x00003000 if MMU && INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	default 0xd0003000 if MMU && !INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	help
+	  This is the address where the kernel is loaded.
+	  It is virtual address for MMUv2 configurations and physical address
+	  for all other configurations.
+
+	  If unsure, leave the default value here.
+
+choice
+	prompt "Relocatable vectors location"
+	default XTENSA_VECTORS_IN_TEXT
+	help
+	  Choose whether relocatable vectors are merged into the kernel .text
+	  or placed separately at runtime. This option does not affect
+	  configurations without VECBASE register where vectors are always
+	  placed at their hardware-defined locations.
+
+config XTENSA_VECTORS_IN_TEXT
+	bool "Merge relocatable vectors into kernel text"
+	depends on !MTD_XIP
+	help
+	  This option puts relocatable vectors into the kernel .text section
+	  with proper alignment.
+	  This is a safe choice for most configurations.
+
+config XTENSA_VECTORS_SEPARATE
+	bool "Put relocatable vectors at fixed address"
+	help
+	  This option puts relocatable vectors at specific virtual address.
+	  Vectors are merged with the .init data in the kernel image and
+	  are copied into their designated location during kernel startup.
+	  Use it to put vectors into IRAM or out of FLASH on kernels with
+	  XIP-aware MTD support.
+
+endchoice
+
+config VECTORS_ADDR
+	hex "Kernel vectors virtual address"
+	default 0x00000000
+	depends on XTENSA_VECTORS_SEPARATE
+	help
+	  This is the virtual address of the (relocatable) vectors base.
+	  It must be within KSEG if MMU is used.
+
+config XIP_DATA_ADDR
+	hex "XIP kernel data virtual address"
+	depends on XIP_KERNEL
+	default 0x00000000
+	help
+	  This is the virtual address where XIP kernel data is copied.
+	  It must be within KSEG if MMU is used.
+
+config PLATFORM_WANT_DEFAULT_MEM
+	def_bool n
+
+config DEFAULT_MEM_START
+	hex
+	prompt "PAGE_OFFSET/PHYS_OFFSET" if !MMU && PLATFORM_WANT_DEFAULT_MEM
+	default 0x60000000 if PLATFORM_WANT_DEFAULT_MEM
+	default 0x00000000
+	help
+	  This is the base address used for both PAGE_OFFSET and PHYS_OFFSET
+	  in noMMU configurations.
+
+	  If unsure, leave the default value here.
+
+choice
+	prompt "KSEG layout"
+	depends on MMU
+	default XTENSA_KSEG_MMU_V2
+
+config XTENSA_KSEG_MMU_V2
+	bool "MMUv2: 128MB cached + 128MB uncached"
+	help
+	  MMUv2 compatible kernel memory map: TLB way 5 maps 128MB starting
+	  at KSEG_PADDR to 0xd0000000 with cache and to 0xd8000000
+	  without cache.
+	  KSEG_PADDR must be aligned to 128MB.
+
+config XTENSA_KSEG_256M
+	bool "256MB cached + 256MB uncached"
+	depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	help
+	  TLB way 6 maps 256MB starting at KSEG_PADDR to 0xb0000000
+	  with cache and to 0xc0000000 without cache.
+	  KSEG_PADDR must be aligned to 256MB.
+
+config XTENSA_KSEG_512M
+	bool "512MB cached + 512MB uncached"
+	depends on INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	help
+	  TLB way 6 maps 512MB starting at KSEG_PADDR to 0xa0000000
+	  with cache and to 0xc0000000 without cache.
+	  KSEG_PADDR must be aligned to 256MB.
+
+endchoice
+
+config HIGHMEM
+	bool "High Memory Support"
+	depends on MMU
+	select KMAP_LOCAL
+	help
+	  Linux can use the full amount of RAM in the system by
+	  default. However, the default MMUv2 setup only maps the
+	  lowermost 128 MB of memory linearly to the areas starting
+	  at 0xd0000000 (cached) and 0xd8000000 (uncached).
+	  When there are more than 128 MB memory in the system not
+	  all of it can be "permanently mapped" by the kernel.
+	  The physical memory that's not permanently mapped is called
+	  "high memory".
+
+	  If you are compiling a kernel which will never run on a
+	  machine with more than 128 MB total physical RAM, answer
+	  N here.
+
+	  If unsure, say Y.
+
+config ARCH_FORCE_MAX_ORDER
+	int "Order of maximal physically contiguous allocations"
+	default "10"
+	help
+	  The kernel page allocator limits the size of maximal physically
+	  contiguous allocations. The limit is called MAX_PAGE_ORDER and it
+	  defines the maximal power of two of number of pages that can be
+	  allocated as a single contiguous block. This option allows
+	  overriding the default setting when ability to allocate very
+	  large blocks of physically contiguous memory is required.
+
+	  Don't change if unsure.
+
+endmenu
+
+menu "Power management options"
+
+config ARCH_HIBERNATION_POSSIBLE
+	def_bool y
+
+source "kernel/power/Kconfig"
+
+endmenu
diff --git a/arch/xtensa/Kconfig.debug b/arch/xtensa/Kconfig.debug
new file mode 100644
index 000000000..e84172a77
--- /dev/null
+++ b/arch/xtensa/Kconfig.debug
@@ -0,0 +1,48 @@
+# SPDX-License-Identifier: GPL-2.0
+
+config DEBUG_TLB_SANITY
+	bool "Debug TLB sanity"
+	depends on DEBUG_KERNEL && MMU
+	help
+	  Enable this to turn on TLB sanity check on each entry to userspace.
+	  This check can spot missing TLB invalidation/wrong PTE permissions/
+	  premature page freeing.
+
+	  If unsure, say N.
+
+config LD_NO_RELAX
+	bool "Disable linker relaxation"
+	default y
+	help
+	  Enable this function to disable link-time optimizations.
+	  The default linker behavior is to combine identical literal
+	  values to reduce code size and remove unnecessary overhead from
+	  assembler-generated 'longcall' sequences.
+	  Enabling this option improves the link time but increases the
+	  code size, and possibly execution time.
+
+config S32C1I_SELFTEST
+	bool "Perform S32C1I instruction self-test at boot"
+	default y
+	help
+	  Enable this option to test S32C1I instruction behavior at boot.
+	  Correct operation of this instruction requires some cooperation from hardware
+	  external to the processor (such as bus bridge, bus fabric, or memory controller).
+	  It is easy to make wrong hardware configuration, this test should catch it early.
+
+	  Say 'N' on stable hardware.
+
+config PRINT_STACK_DEPTH
+	int "Stack depth to print" if DEBUG_KERNEL
+	default 64
+	help
+	  This option allows you to set the stack depth that the kernel
+	  prints in stack traces.
+
+config PRINT_USER_CODE_ON_UNHANDLED_EXCEPTION
+	bool "Dump user code around unhandled exception address"
+	help
+	  Enable this option to display user code around PC of the unhandled
+	  exception (starting at address aligned on 16 byte boundary).
+	  This may simplify finding faulting code in the absence of other
+	  debug facilities.
diff --git a/arch/xtensa/Makefile b/arch/xtensa/Makefile
new file mode 100644
index 000000000..4c14a0217
--- /dev/null
+++ b/arch/xtensa/Makefile
@@ -0,0 +1,77 @@
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+# Copyright (C) 2001 - 2005  Tensilica Inc.
+# Copyright (C) 2014 Cadence Design Systems Inc.
+#
+# This file is included by the global makefile so that you can add your own
+# architecture-specific flags and dependencies.
+
+# Core configuration.
+# (Use VAR=<xtensa_config> to use another default compiler.)
+
+variant-y := $(CONFIG_XTENSA_VARIANT_NAME)
+
+VARIANT = $(variant-y)
+
+ifneq ($(VARIANT),)
+  ifdef cross_compiling
+    ifndef CROSS_COMPILE
+      CROSS_COMPILE = xtensa_$(VARIANT)-
+    endif
+  endif
+endif
+
+# Platform configuration
+
+platform-$(CONFIG_XTENSA_PLATFORM_XT2000)	:= xt2000
+platform-$(CONFIG_XTENSA_PLATFORM_ISS)		:= iss
+platform-$(CONFIG_XTENSA_PLATFORM_XTFPGA)	:= xtfpga
+
+# temporarily until string.h is fixed
+KBUILD_CFLAGS += -ffreestanding -D__linux__
+KBUILD_CFLAGS += -pipe -mlongcalls -mtext-section-literals
+KBUILD_CFLAGS += $(call cc-option,-mforce-no-pic,)
+KBUILD_CFLAGS += $(call cc-option,-mno-serialize-volatile,)
+KBUILD_CFLAGS += $(call cc-option,-mno-fdpic,)
+ifneq ($(CONFIG_KERNEL_ABI_CALL0),)
+KBUILD_CFLAGS += -mabi=call0
+KBUILD_AFLAGS += -mabi=call0
+endif
+
+KBUILD_AFLAGS += -mlongcalls -mtext-section-literals
+KBUILD_AFLAGS += $(call cc-option,-mno-fdpic,)
+
+KBUILD_LDFLAGS += -m elf32xtensa
+
+ifneq ($(CONFIG_LD_NO_RELAX),)
+KBUILD_LDFLAGS += --no-relax
+endif
+
+CHECKFLAGS += -D$(if $(CONFIG_CPU_BIG_ENDIAN),__XTENSA_EB__,__XTENSA_EL__)
+
+vardirs := $(patsubst %,arch/xtensa/variants/%/,$(variant-y))
+plfdirs := $(patsubst %,arch/xtensa/platforms/%/,$(platform-y))
+
+KBUILD_CPPFLAGS += $(patsubst %,-I$(srctree)/%include,$(vardirs) $(plfdirs))
+
+KBUILD_DEFCONFIG := iss_defconfig
+
+libs-y		+= arch/xtensa/lib/
+
+boot		:= arch/xtensa/boot
+
+all Image zImage uImage xipImage: vmlinux
+	$(Q)$(MAKE) $(build)=$(boot) $@
+
+archheaders:
+	$(Q)$(MAKE) $(build)=arch/xtensa/kernel/syscalls all
+
+define archhelp
+  @echo '* Image       - Kernel ELF image with reset vector'
+  @echo '* zImage      - Compressed kernel image (arch/xtensa/boot/images/zImage.*)'
+  @echo '* uImage      - U-Boot wrapped image'
+  @echo '  xipImage    - XIP image'
+endef
diff --git a/arch/xtensa/boot/.gitignore b/arch/xtensa/boot/.gitignore
new file mode 100644
index 000000000..615f1f741
--- /dev/null
+++ b/arch/xtensa/boot/.gitignore
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: GPL-2.0-only
+uImage
+zImage.redboot
diff --git a/arch/xtensa/boot/Makefile b/arch/xtensa/boot/Makefile
new file mode 100644
index 000000000..d8b0fadf4
--- /dev/null
+++ b/arch/xtensa/boot/Makefile
@@ -0,0 +1,53 @@
+#
+# arch/xtensa/boot/Makefile
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+#
+
+
+# KBUILD_CFLAGS used when building rest of boot (takes effect recursively)
+KBUILD_CFLAGS	+= -fno-builtin
+
+subdir-y	:= lib
+targets		+= vmlinux.bin vmlinux.bin.gz
+targets		+= uImage xipImage
+
+# Subdirs for the boot loader(s)
+
+boot-$(CONFIG_XTENSA_PLATFORM_ISS)    += Image
+boot-$(CONFIG_XTENSA_PLATFORM_XT2000) += Image zImage uImage
+boot-$(CONFIG_XTENSA_PLATFORM_XTFPGA) += Image zImage uImage
+
+all: $(boot-y)
+Image: boot-elf
+zImage: boot-redboot
+uImage: $(obj)/uImage
+xipImage: $(obj)/xipImage
+
+boot-elf boot-redboot: $(addprefix $(obj)/,$(subdir-y))
+	$(Q)$(MAKE) $(build)=$(obj)/$@ $(MAKECMDGOALS)
+
+OBJCOPYFLAGS = --strip-all -R .comment -R .notes -O binary
+
+$(obj)/vmlinux.bin: vmlinux FORCE
+	$(call if_changed,objcopy)
+
+$(obj)/vmlinux.bin.gz: $(obj)/vmlinux.bin FORCE
+	$(call if_changed,gzip)
+
+boot-elf: $(obj)/vmlinux.bin
+boot-redboot: $(obj)/vmlinux.bin.gz
+
+UIMAGE_LOADADDR = $(CONFIG_KERNEL_LOAD_ADDRESS)
+UIMAGE_COMPRESSION = gzip
+
+$(obj)/uImage: $(obj)/vmlinux.bin.gz FORCE
+	$(call if_changed,uimage)
+	$(Q)$(kecho) '  Kernel: $@ is ready'
+
+$(obj)/xipImage: vmlinux FORCE
+	$(call if_changed,objcopy)
+	$(Q)$(kecho) '  Kernel: $@ is ready'
diff --git a/arch/xtensa/boot/boot-elf/.gitignore b/arch/xtensa/boot/boot-elf/.gitignore
new file mode 100644
index 000000000..747340450
--- /dev/null
+++ b/arch/xtensa/boot/boot-elf/.gitignore
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+boot.lds
diff --git a/arch/xtensa/boot/boot-elf/Makefile b/arch/xtensa/boot/boot-elf/Makefile
new file mode 100644
index 000000000..faec20029
--- /dev/null
+++ b/arch/xtensa/boot/boot-elf/Makefile
@@ -0,0 +1,30 @@
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+
+OBJCOPY_ARGS := -O $(if $(CONFIG_CPU_BIG_ENDIAN),elf32-xtensa-be,elf32-xtensa-le)
+
+CPPFLAGS_boot.lds += -P -C
+KBUILD_AFLAGS += -mtext-section-literals
+
+boot-y		:= bootstrap.o
+targets		+= $(boot-y) boot.lds
+
+OBJS		:= $(addprefix $(obj)/,$(boot-y))
+
+$(obj)/Image.o: $(obj)/../vmlinux.bin $(OBJS)
+	$(Q)$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
+		--add-section image=$< \
+		--set-section-flags image=contents,alloc,load,load,data \
+		$(OBJS) $@
+
+$(obj)/../Image.elf: $(obj)/Image.o $(obj)/boot.lds
+	$(Q)$(LD) $(KBUILD_LDFLAGS) \
+		-T $(obj)/boot.lds \
+		--build-id=none \
+		-o $@ $(obj)/Image.o
+	$(Q)$(kecho) '  Kernel: $@ is ready'
+
+all Image: $(obj)/../Image.elf
diff --git a/arch/xtensa/boot/boot-elf/boot.lds.S b/arch/xtensa/boot/boot-elf/boot.lds.S
new file mode 100644
index 000000000..32a3b7c5b
--- /dev/null
+++ b/arch/xtensa/boot/boot-elf/boot.lds.S
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  linux/arch/xtensa/boot/boot-elf/boot.lds.S
+ *
+ *  Copyright (C) 2008 - 2013 by Tensilica Inc.
+ *
+ *  Chris Zankel <chris@zankel.net>
+ *  Marc Gauthier <marc@tensilica.com
+ *  Pete Delaney <piet@tensilica.com>
+ */
+
+#include <asm/vectors.h>
+OUTPUT_ARCH(xtensa)
+ENTRY(_ResetVector)
+
+SECTIONS
+{
+	.ResetVector.text XCHAL_RESET_VECTOR_VADDR :
+	{
+		*(.ResetVector.text)
+	}
+
+	.image KERNELOFFSET: AT (CONFIG_KERNEL_LOAD_ADDRESS)
+	{
+		_image_start = .;
+		*(image)
+		. = (. + 3) & ~ 3;
+		_image_end = .	;
+	}
+
+	.bss ((LOADADDR(.image) + SIZEOF(.image) + 3) & ~ 3):
+	{
+		__bss_start = .;
+		*(.sbss)
+		*(.scommon)
+		*(.dynbss)
+		*(.bss)
+		__bss_end = .;
+	}
+}
diff --git a/arch/xtensa/boot/boot-elf/bootstrap.S b/arch/xtensa/boot/boot-elf/bootstrap.S
new file mode 100644
index 000000000..2dd28931d
--- /dev/null
+++ b/arch/xtensa/boot/boot-elf/bootstrap.S
@@ -0,0 +1,79 @@
+/*
+ * arch/xtensa/boot/boot-elf/bootstrap.S
+ *
+ * Low-level exception handling
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2004 - 2013 by Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Marc Gauthier <marc@tensilica.com>
+ * Piet Delaney <piet@tensilica.com>
+ */
+
+#include <asm/bootparam.h>
+#include <asm/initialize_mmu.h>
+#include <asm/vectors.h>
+#include <linux/linkage.h>
+
+	.section	.ResetVector.text, "ax"
+	.global         _ResetVector
+	.global         reset
+
+_ResetVector:
+	_j _SetupMMU
+
+	.begin  no-absolute-literals
+	.literal_position
+
+#ifdef CONFIG_PARSE_BOOTPARAM
+	.align 4
+_bootparam:
+	.short	BP_TAG_FIRST
+	.short	4
+	.long	BP_VERSION
+	.short	BP_TAG_LAST
+	.short	0
+	.long	0
+#endif
+
+	.align  4
+_SetupMMU:
+#if XCHAL_HAVE_WINDOWED
+	movi	a0, 0
+	wsr	a0, windowbase
+	rsync
+	movi	a0, 1
+	wsr	a0, windowstart
+	rsync
+#endif
+	movi	a0, 0x1F
+	wsr	a0, ps
+	rsync
+
+#ifndef CONFIG_INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	initialize_mmu
+#endif
+
+	rsil    a0, XCHAL_DEBUGLEVEL-1
+	rsync
+reset:
+#if defined(CONFIG_INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX) && \
+	XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY
+	movi	a0, CONFIG_KERNEL_LOAD_ADDRESS
+#else
+	movi	a0, KERNELOFFSET
+#endif
+#ifdef CONFIG_PARSE_BOOTPARAM
+	movi	a2, _bootparam
+#else
+	movi	a2, 0
+#endif
+	movi	a3, 0
+	movi	a4, 0
+	jx      a0
+
+	.end    no-absolute-literals
diff --git a/arch/xtensa/boot/boot-redboot/Makefile b/arch/xtensa/boot/boot-redboot/Makefile
new file mode 100644
index 000000000..c0eef3f6f
--- /dev/null
+++ b/arch/xtensa/boot/boot-redboot/Makefile
@@ -0,0 +1,31 @@
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+
+OBJCOPY_ARGS := -O $(if $(CONFIG_CPU_BIG_ENDIAN),elf32-xtensa-be,elf32-xtensa-le)
+
+boot-y	:= bootstrap.o
+targets	+= $(boot-y)
+
+OBJS	:= $(addprefix $(obj)/,$(boot-y))
+LIBS	:= arch/xtensa/boot/lib/lib.a arch/xtensa/lib/lib.a
+
+$(obj)/zImage.o: $(obj)/../vmlinux.bin.gz $(OBJS)
+	$(Q)$(OBJCOPY) $(OBJCOPY_ARGS) -R .comment \
+		--add-section image=$< \
+		--set-section-flags image=contents,alloc,load,load,data \
+		$(OBJS) $@
+
+$(obj)/zImage.elf: $(obj)/zImage.o $(LIBS)
+	$(Q)$(LD) $(KBUILD_LDFLAGS) \
+		-T $(srctree)/$(obj)/boot.ld \
+		--build-id=none \
+		-o $@ $^
+
+$(obj)/../zImage.redboot: $(obj)/zImage.elf
+	$(Q)$(OBJCOPY) -S -O binary $< $@
+	$(Q)$(kecho) '  Kernel: $@ is ready'
+
+all zImage: $(obj)/../zImage.redboot
diff --git a/arch/xtensa/boot/boot-redboot/boot.ld b/arch/xtensa/boot/boot-redboot/boot.ld
new file mode 100644
index 000000000..b0b9e95b5
--- /dev/null
+++ b/arch/xtensa/boot/boot-redboot/boot.ld
@@ -0,0 +1,59 @@
+OUTPUT_ARCH(xtensa)
+
+SECTIONS
+{
+	.start 0xD1000000 : { *(.start) }
+
+	.text :
+	{
+		__reloc_start = . ;
+		_text_start = . ;
+		*(.literal .text.literal .text)
+		_text_end = . ;
+	}
+
+	.rodata ALIGN(0x04):
+	{
+		*(.rodata)
+		*(.rodata1)
+	}
+
+	.data ALIGN(0x04):
+	{
+		*(.data)
+		*(.data1)
+		*(.sdata)
+		*(.sdata2)
+		*(.got.plt)
+		*(.got)
+		*(.dynamic)
+	}
+
+	__reloc_end = . ;
+
+	. = ALIGN(0x10);
+	__image_load = . ;
+	.image 0xd0003000: AT(__image_load)
+	{
+		_image_start = .;
+		*(image)
+		. = (. + 3) & ~ 3;
+		_image_end = .	;
+	}
+
+
+	.bss ((LOADADDR(.image) + SIZEOF(.image) + 3) & ~ 3):
+	{
+		__bss_start = .;
+		*(.sbss)
+		*(.scommon)
+		*(.dynbss)
+		*(.bss)
+		__bss_end = .;
+	}
+	_end = .;
+	_param_start = .;
+
+
+	PROVIDE (end = .);
+}
diff --git a/arch/xtensa/boot/boot-redboot/bootstrap.S b/arch/xtensa/boot/boot-redboot/bootstrap.S
new file mode 100644
index 000000000..3ed94ad35
--- /dev/null
+++ b/arch/xtensa/boot/boot-redboot/bootstrap.S
@@ -0,0 +1,255 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#include <asm/core.h>
+#include <asm/regs.h>
+#include <asm/asmmacro.h>
+#include <asm/cacheasm.h>
+#include <asm/processor.h>
+	/*
+	 * RB-Data: RedBoot data/bss
+	 * P:	    Boot-Parameters
+	 * L:	    Kernel-Loader
+	 *
+	 * The Linux-Kernel image including the loader must be loaded
+	 * to a position so that the kernel and the boot parameters
+	 * can fit in the space before the load address.
+	 *  ______________________________________________________
+	 * |_RB-Data_|_P_|__________|_L_|___Linux-Kernel___|______|
+	 *                          ^
+	 *                          ^ Load address
+	 *  ______________________________________________________
+	 * |___Linux-Kernel___|_P_|_L_|___________________________|
+	 *
+	 * The loader copies the parameter to the position that will
+	 * be the end of the kernel and itself to the end of the
+	 * parameter list.
+	 */
+
+/* Make sure we have enough space for the 'uncompressor' */
+
+#define STACK_SIZE 32768
+#define HEAP_SIZE (131072*4)
+
+	# a2: Parameter list
+	# a3: Size of parameter list
+
+	.section .start, "ax"
+
+	.globl __start
+	/* this must be the first byte of the loader! */
+__start:
+	abi_entry(32)		# we do not intend to return
+	_call0	_start
+__start_a0:
+	.align 4
+
+	.section .text, "ax"
+	.literal_position
+	.begin literal_prefix .text
+
+	/* put literals in here! */
+
+	.globl _start
+_start:
+
+	/* 'reset' window registers */
+
+	movi	a4, 1
+	wsr	a4, ps
+	rsync
+#if XCHAL_HAVE_WINDOWED
+	rsr	a5, windowbase
+	ssl	a5
+	sll	a4, a4
+	wsr	a4, windowstart
+	rsync
+#endif
+	movi	a4, KERNEL_PS_WOE_MASK
+	wsr	a4, ps
+	rsync
+
+KABI_C0	mov	abi_saved0, abi_arg0
+
+	/* copy the loader to its address
+	 * Note: The loader itself is a very small piece, so we assume we
+	 *       don't partially overlap. We also assume (even more important)
+	 *	 that the kernel image is out of the way. Usually, when the
+	 *	 load address of this image is not at an arbitrary address,
+	 *	 but aligned to some 10K's we shouldn't overlap.
+	 */
+
+	/* Note: The assembler cannot relax "addi a0, a0, ..." to an
+	   l32r, so we load to a4 first. */
+
+	# addi	a4, a0, __start - __start_a0
+	# mov	a0, a4
+
+	movi	a4, __start
+	movi	a5, __start_a0
+	add	a4, a0, a4
+	sub	a0, a4, a5
+
+	movi	a4, __start
+	movi	a5, __reloc_end
+
+	# a0: address where this code has been loaded
+	# a4: compiled address of __start
+	# a5: compiled end address
+
+	mov.n	a7, a0
+	mov.n	a8, a4
+
+1:
+	l32i	a10, a7, 0
+	l32i	a11, a7, 4
+	s32i	a10, a8, 0
+	s32i	a11, a8, 4
+	l32i	a10, a7, 8
+	l32i	a11, a7, 12
+	s32i	a10, a8, 8
+	s32i	a11, a8, 12
+	addi	a8, a8, 16
+	addi	a7, a7, 16
+	blt	a8, a5, 1b
+
+
+	/* We have to flush and invalidate the caches here before we jump. */
+
+#if XCHAL_DCACHE_IS_WRITEBACK
+
+	___flush_dcache_all a5 a6
+
+#endif
+
+	___invalidate_icache_all a5 a6
+	isync
+
+	movi	a11, _reloc
+	jx	a11
+
+	.globl _reloc
+_reloc:
+
+	/* RedBoot is now at the end of the memory, so we don't have
+	 * to copy the parameter list. Keep the code around; in case
+	 * we need it again. */
+#if 0
+	# a0: load address
+	# a2: start address of parameter list
+	# a3: length of parameter list
+	# a4: __start
+
+	/* copy the parameter list out of the way */
+
+	movi	a6, _param_start
+	add	a3, a2, a3
+2:
+	l32i	a8, a2, 0
+	s32i	a8, a6, 0
+	addi	a2, a2, 4
+	addi	a6, a6, 4
+	blt	a2, a3, 2b
+#endif
+
+	/* clear BSS section */
+	movi	a6, __bss_start
+	movi	a7, __bss_end
+	movi.n	a5, 0
+3:
+	s32i	a5, a6, 0
+	addi	a6, a6, 4
+	blt	a6, a7, 3b
+
+	movi	a5, -16
+	movi	a1, _stack + STACK_SIZE
+	and	a1, a1, a5
+
+	/* Uncompress the kernel */
+
+	# a0: load address
+	# a2: boot parameter
+	# a4: __start
+
+	movi	a3, __image_load
+	sub	a4, a3, a4
+	add	abi_arg2, a0, a4
+
+	# a1  Stack
+	# a8(a4)  Load address of the image
+
+	movi	abi_arg0, _image_start
+	movi	abi_arg4, _image_end
+	movi	abi_arg1, 0x1000000
+	sub	abi_tmp0, abi_arg4, abi_arg0
+	movi	abi_arg3, complen
+	s32i	abi_tmp0, abi_arg3, 0
+
+	movi	a0, 0
+
+	# abi_arg0 destination
+	# abi_arg1 maximum size of destination
+	# abi_arg2 source
+	# abi_arg3 ptr to length
+
+	.extern gunzip
+	movi	abi_tmp0, gunzip
+	beqz	abi_tmp0, 1f
+
+	abi_callx	abi_tmp0
+
+	j	2f
+
+
+	# abi_arg0 destination start
+	# abi_arg1 maximum size of destination
+	# abi_arg2 source start
+	# abi_arg3 ptr to length
+	# abi_arg4 destination end
+
+1:
+        l32i    abi_tmp0, abi_arg2, 0
+        l32i    abi_tmp1, abi_arg2, 4
+        s32i    abi_tmp0, abi_arg0, 0
+        s32i    abi_tmp1, abi_arg0, 4
+        l32i    abi_tmp0, abi_arg2, 8
+        l32i    abi_tmp1, abi_arg2, 12
+        s32i    abi_tmp0, abi_arg0, 8
+        s32i    abi_tmp1, abi_arg0, 12
+        addi    abi_arg0, abi_arg0, 16
+        addi    abi_arg2, abi_arg2, 16
+        blt     abi_arg0, abi_arg4, 1b
+
+
+	/* jump to the kernel */
+2:
+#if XCHAL_DCACHE_IS_WRITEBACK
+
+	___flush_dcache_all a5 a6
+
+#endif
+
+	___invalidate_icache_all a5 a6
+
+	isync
+
+	# a2  Boot parameter list
+
+KABI_C0	mov	abi_arg0, abi_saved0
+	movi	a0, _image_start
+	jx	a0
+
+	.align 16
+	.data
+	.globl avail_ram
+avail_ram:
+	.long	_heap
+	.globl end_avail
+end_avail:
+	.long	_heap + HEAP_SIZE
+
+	.comm _stack, STACK_SIZE
+	.comm _heap, HEAP_SIZE
+
+	.globl end_avail
+	.comm complen, 4
+
+	.end	literal_prefix
diff --git a/arch/xtensa/boot/dts/Makefile b/arch/xtensa/boot/dts/Makefile
new file mode 100644
index 000000000..7271294ce
--- /dev/null
+++ b/arch/xtensa/boot/dts/Makefile
@@ -0,0 +1,13 @@
+#
+# arch/xtensa/boot/dts/Makefile
+#
+# This file is subject to the terms and conditions of the GNU General Public
+# License.  See the file "COPYING" in the main directory of this archive
+# for more details.
+#
+#
+
+dtb-$(CONFIG_OF) += $(addsuffix .dtb, $(CONFIG_BUILTIN_DTB_NAME))
+
+# for CONFIG_OF_ALL_DTBS test
+dtb-	:= $(patsubst $(src)/%.dts,%.dtb, $(wildcard $(src)/*.dts))
diff --git a/arch/xtensa/boot/dts/csp.dts b/arch/xtensa/boot/dts/csp.dts
new file mode 100644
index 000000000..885495460
--- /dev/null
+++ b/arch/xtensa/boot/dts/csp.dts
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+
+/ {
+	compatible = "cdns,xtensa-xtfpga";
+	#address-cells = <1>;
+	#size-cells = <1>;
+	interrupt-parent = <&pic>;
+
+	chosen {
+		bootargs = "earlycon=cdns,0xfd000000,115200 console=tty0 console=ttyPS0,115200 root=/dev/ram0 rw earlyprintk xilinx_uartps.rx_trigger_level=32 loglevel=8 nohz=off ignore_loglevel";
+	};
+
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x40000000>;
+	};
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		cpu@0 {
+			compatible = "cdns,xtensa-cpu";
+			reg = <0>;
+		};
+	};
+
+	pic: pic {
+		compatible = "cdns,xtensa-pic";
+		#interrupt-cells = <2>;
+		interrupt-controller;
+	};
+
+	clocks {
+		osc: main-oscillator {
+			#clock-cells = <0>;
+			compatible = "fixed-clock";
+		};
+	};
+
+	soc {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "simple-bus";
+		ranges = <0x00000000 0xf0000000 0x10000000>;
+
+		uart0: serial@0d000000 {
+			compatible = "xlnx,xuartps", "cdns,uart-r1p8";
+			clocks = <&osc>, <&osc>;
+			clock-names = "uart_clk", "pclk";
+			reg = <0x0d000000 0x1000>;
+			interrupts = <0 1>;
+		};
+	};
+};
diff --git a/arch/xtensa/boot/dts/kc705.dts b/arch/xtensa/boot/dts/kc705.dts
new file mode 100644
index 000000000..6887ff090
--- /dev/null
+++ b/arch/xtensa/boot/dts/kc705.dts
@@ -0,0 +1,31 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-128m.dtsi"
+
+/ {
+	compatible = "cdns,xtensa-kc705";
+	chosen {
+		bootargs = "earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=0x38000000";
+	};
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x38000000>;
+	};
+
+	reserved-memory {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		ranges;
+
+		/* global autoconfigured region for contiguous allocations */
+		linux,cma {
+			compatible = "shared-dma-pool";
+			reusable;
+			size = <0x04000000>;
+			alignment = <0x2000>;
+			alloc-ranges = <0x00000000 0x20000000>;
+			linux,cma-default;
+		};
+	};
+};
diff --git a/arch/xtensa/boot/dts/kc705_nommu.dts b/arch/xtensa/boot/dts/kc705_nommu.dts
new file mode 100644
index 000000000..d8e194a0f
--- /dev/null
+++ b/arch/xtensa/boot/dts/kc705_nommu.dts
@@ -0,0 +1,18 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-128m.dtsi"
+
+/ {
+	compatible = "cdns,xtensa-kc705";
+	chosen {
+		bootargs = "earlycon=uart8250,mmio32,0x9d050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug";
+	};
+	memory@0 {
+		device_type = "memory";
+		reg = <0x60000000 0x10000000>;
+	};
+	soc {
+		ranges = <0x00000000 0x90000000 0x10000000>;
+	};
+};
diff --git a/arch/xtensa/boot/dts/lx200mx.dts b/arch/xtensa/boot/dts/lx200mx.dts
new file mode 100644
index 000000000..974a8d904
--- /dev/null
+++ b/arch/xtensa/boot/dts/lx200mx.dts
@@ -0,0 +1,17 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-16m.dtsi"
+
+/ {
+	compatible = "cdns,xtensa-lx200";
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x06000000>;
+	};
+	pic: pic {
+		compatible = "cdns,xtensa-mx";
+		#interrupt-cells = <2>;
+		interrupt-controller;
+	};
+};
diff --git a/arch/xtensa/boot/dts/lx60.dts b/arch/xtensa/boot/dts/lx60.dts
new file mode 100644
index 000000000..7c203c1c7
--- /dev/null
+++ b/arch/xtensa/boot/dts/lx60.dts
@@ -0,0 +1,12 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-4m.dtsi"
+
+/ {
+	compatible = "cdns,xtensa-lx60";
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x04000000>;
+	};
+};
diff --git a/arch/xtensa/boot/dts/ml605.dts b/arch/xtensa/boot/dts/ml605.dts
new file mode 100644
index 000000000..08e5c8d47
--- /dev/null
+++ b/arch/xtensa/boot/dts/ml605.dts
@@ -0,0 +1,12 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+/include/ "xtfpga.dtsi"
+/include/ "xtfpga-flash-16m.dtsi"
+
+/ {
+	compatible = "cdns,xtensa-ml605";
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x08000000>;
+	};
+};
diff --git a/arch/xtensa/boot/dts/virt.dts b/arch/xtensa/boot/dts/virt.dts
new file mode 100644
index 000000000..611b98a02
--- /dev/null
+++ b/arch/xtensa/boot/dts/virt.dts
@@ -0,0 +1,72 @@
+// SPDX-License-Identifier: GPL-2.0
+/dts-v1/;
+
+/ {
+	compatible = "cdns,xtensa-iss";
+	#address-cells = <1>;
+	#size-cells = <1>;
+	interrupt-parent = <&pic>;
+
+	chosen {
+		bootargs = "console=ttyS0,115200n8 debug";
+	};
+
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x80000000>;
+	};
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		cpu@0 {
+			compatible = "cdns,xtensa-cpu";
+			reg = <0>;
+			clocks = <&osc>;
+		};
+	};
+
+	clocks {
+		osc: osc {
+			#clock-cells = <0>;
+			compatible = "fixed-clock";
+			clock-frequency = <40000000>;
+		};
+	};
+
+	pic: pic {
+		compatible = "cdns,xtensa-pic";
+		/* one cell: internal irq number,
+		 * two cells: second cell == 0: internal irq number
+		 *            second cell == 1: external irq number
+		 */
+		#address-cells = <0>;
+		#interrupt-cells = <2>;
+		interrupt-controller;
+	};
+
+	pci {
+		compatible = "pci-host-ecam-generic";
+		device_type = "pci";
+		#address-cells = <3>;
+		#size-cells = <2>;
+		#interrupt-cells = <0x1>;
+
+		bus-range = <0x0 0x3e>;
+		reg = <0xf0100000 0x03f00000>;
+
+		     // BUS_ADDRESS(3)  CPU_PHYSICAL(1)  SIZE(2)
+		ranges = <0x01000000 0x0 0x00000000  0xf0000000  0x0 0x00010000>,
+			 <0x02000000 0x0 0xf4000000  0xf4000000  0x0 0x08000000>;
+
+		     // PCI_DEVICE(3)  INT#(1)  CONTROLLER(PHANDLE)  CONTROLLER_DATA(2)
+		interrupt-map = <
+			0x0000 0x0 0x0  0x1  &pic  0x0 0x1
+			0x0800 0x0 0x0  0x1  &pic  0x1 0x1
+			0x1000 0x0 0x0  0x1  &pic  0x2 0x1
+			0x1800 0x0 0x0  0x1  &pic  0x3 0x1
+			>;
+
+		interrupt-map-mask = <0x1800 0x0 0x0  0x7>;
+	};
+};
diff --git a/arch/xtensa/boot/dts/xtfpga-flash-128m.dtsi b/arch/xtensa/boot/dts/xtfpga-flash-128m.dtsi
new file mode 100644
index 000000000..c33932568
--- /dev/null
+++ b/arch/xtensa/boot/dts/xtfpga-flash-128m.dtsi
@@ -0,0 +1,29 @@
+// SPDX-License-Identifier: GPL-2.0
+/ {
+	soc {
+		flash: flash@00000000 {
+			#address-cells = <1>;
+			#size-cells = <1>;
+			compatible = "cfi-flash";
+			reg = <0x00000000 0x08000000>;
+			bank-width = <2>;
+			device-width = <2>;
+			partition@0 {
+				label = "data";
+				reg = <0x00000000 0x06000000>;
+			};
+			partition@6000000 {
+				label = "boot loader area";
+				reg = <0x06000000 0x00800000>;
+			};
+			partition@6800000 {
+				label = "kernel image";
+				reg = <0x06800000 0x017e0000>;
+			};
+			partition@7fe0000 {
+				label = "boot environment";
+				reg = <0x07fe0000 0x00020000>;
+			};
+		};
+        };
+};
diff --git a/arch/xtensa/boot/dts/xtfpga-flash-16m.dtsi b/arch/xtensa/boot/dts/xtfpga-flash-16m.dtsi
new file mode 100644
index 000000000..7bde2ab2d
--- /dev/null
+++ b/arch/xtensa/boot/dts/xtfpga-flash-16m.dtsi
@@ -0,0 +1,29 @@
+// SPDX-License-Identifier: GPL-2.0
+/ {
+	soc {
+		flash: flash@08000000 {
+			#address-cells = <1>;
+			#size-cells = <1>;
+			compatible = "cfi-flash";
+			reg = <0x08000000 0x01000000>;
+			bank-width = <2>;
+			device-width = <2>;
+			partition@0 {
+				label = "boot loader area";
+				reg = <0x00000000 0x00400000>;
+			};
+			partition@400000 {
+				label = "kernel image";
+				reg = <0x00400000 0x00600000>;
+			};
+			partition@a00000 {
+				label = "data";
+				reg = <0x00a00000 0x005e0000>;
+			};
+			partition@fe0000 {
+				label = "boot environment";
+				reg = <0x00fe0000 0x00020000>;
+			};
+		};
+	};
+};
diff --git a/arch/xtensa/boot/dts/xtfpga-flash-4m.dtsi b/arch/xtensa/boot/dts/xtfpga-flash-4m.dtsi
new file mode 100644
index 000000000..0655b8687
--- /dev/null
+++ b/arch/xtensa/boot/dts/xtfpga-flash-4m.dtsi
@@ -0,0 +1,21 @@
+// SPDX-License-Identifier: GPL-2.0
+/ {
+	soc {
+		flash: flash@08000000 {
+			#address-cells = <1>;
+			#size-cells = <1>;
+			compatible = "cfi-flash";
+			reg = <0x08000000 0x00400000>;
+			bank-width = <2>;
+			device-width = <2>;
+			partition@0 {
+				label = "boot loader area";
+				reg = <0x00000000 0x003f0000>;
+			};
+			partition@3f0000 {
+				label = "boot environment";
+				reg = <0x003f0000 0x00010000>;
+			};
+		};
+	};
+};
diff --git a/arch/xtensa/boot/dts/xtfpga.dtsi b/arch/xtensa/boot/dts/xtfpga.dtsi
new file mode 100644
index 000000000..e46ae07ba
--- /dev/null
+++ b/arch/xtensa/boot/dts/xtfpga.dtsi
@@ -0,0 +1,137 @@
+// SPDX-License-Identifier: GPL-2.0
+/ {
+	compatible = "cdns,xtensa-xtfpga";
+	#address-cells = <1>;
+	#size-cells = <1>;
+	interrupt-parent = <&pic>;
+
+	chosen {
+		bootargs = "earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug";
+	};
+
+	memory@0 {
+		device_type = "memory";
+		reg = <0x00000000 0x06000000>;
+	};
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+		cpu@0 {
+			compatible = "cdns,xtensa-cpu";
+			reg = <0>;
+			clocks = <&osc>;
+		};
+	};
+
+	pic: pic {
+		compatible = "cdns,xtensa-pic";
+		/* one cell: internal irq number,
+		 * two cells: second cell == 0: internal irq number
+		 *            second cell == 1: external irq number
+		 */
+		#interrupt-cells = <2>;
+		interrupt-controller;
+	};
+
+	clocks {
+		clk54: clk54 {
+			#clock-cells = <0>;
+			compatible = "fixed-clock";
+			clock-frequency = <54000000>;
+		};
+	};
+
+	soc {
+		#address-cells = <1>;
+		#size-cells = <1>;
+		compatible = "simple-bus";
+		ranges = <0x00000000 0xf0000000 0x10000000>;
+
+		osc: main-oscillator {
+			#clock-cells = <0>;
+			compatible = "cdns,xtfpga-clock";
+			reg = <0x0d020004 0x4>;
+		};
+
+		serial0: serial@0d050020 {
+			device_type = "serial";
+			compatible = "ns16550a";
+			no-loopback-test;
+			reg = <0x0d050020 0x20>;
+			reg-shift = <2>;
+			reg-io-width = <4>;
+			native-endian;
+			interrupts = <0 1>; /* external irq 0 */
+			clocks = <&osc>;
+		};
+
+		enet0: ethoc@0d030000 {
+			compatible = "opencores,ethoc";
+			reg = <0x0d030000 0x4000 0x0d800000 0x4000>;
+			native-endian;
+			interrupts = <1 1>; /* external irq 1 */
+			local-mac-address = [00 50 c2 13 6f 00];
+			clocks = <&osc>;
+		};
+
+		i2s0: xtfpga-i2s@0d080000 {
+			#sound-dai-cells = <0>;
+			compatible = "cdns,xtfpga-i2s";
+			reg = <0x0d080000 0x40>;
+			interrupts = <2 1>; /* external irq 2 */
+			clocks = <&cdce706 4>;
+		};
+
+		i2c0: i2c-master@0d090000 {
+			compatible = "opencores,i2c-ocores";
+			#address-cells = <1>;
+			#size-cells = <0>;
+			reg = <0x0d090000 0x20>;
+			reg-shift = <2>;
+			reg-io-width = <4>;
+			native-endian;
+			interrupts = <4 1>;
+			clocks = <&osc>;
+
+			cdce706: clock-synth@69 {
+				compatible = "ti,cdce706";
+				#clock-cells = <1>;
+				reg = <0x69>;
+				clocks = <&clk54>;
+				clock-names = "clk_in0";
+			};
+		};
+
+		spi0: spi@0d0a0000 {
+			compatible = "cdns,xtfpga-spi";
+			#address-cells = <1>;
+			#size-cells = <0>;
+			reg = <0x0d0a0000 0xc>;
+
+			tlv320aic23: sound-codec@0 {
+				#sound-dai-cells = <0>;
+				compatible = "tlv320aic23";
+				reg = <0>;
+				spi-max-frequency = <12500000>;
+			};
+		};
+	};
+
+	sound {
+		compatible = "simple-audio-card";
+		simple-audio-card,format = "i2s";
+		simple-audio-card,mclk-fs = <256>;
+
+		simple-audio-card,cpu {
+			sound-dai = <&i2s0>;
+		};
+
+		simple-audio-card,codec {
+			sound-dai = <&tlv320aic23>;
+			simple-audio-card,bitclock-master = <0>;
+			simple-audio-card,frame-master = <0>;
+			clocks = <&cdce706 4>;
+		};
+	};
+};
diff --git a/arch/xtensa/boot/lib/.gitignore b/arch/xtensa/boot/lib/.gitignore
new file mode 100644
index 000000000..805a82492
--- /dev/null
+++ b/arch/xtensa/boot/lib/.gitignore
@@ -0,0 +1,4 @@
+# SPDX-License-Identifier: GPL-2.0-only
+inffast.c
+inflate.c
+inftrees.c
diff --git a/arch/xtensa/boot/lib/Makefile b/arch/xtensa/boot/lib/Makefile
new file mode 100644
index 000000000..0378a22a0
--- /dev/null
+++ b/arch/xtensa/boot/lib/Makefile
@@ -0,0 +1,34 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for some libs needed by zImage.
+#
+
+zlib	:= inffast.c inflate.c inftrees.c
+
+lib-y	+= $(zlib:.c=.o) zmem.o
+
+ccflags-y	:= -I $(srctree)/lib/zlib_inflate
+ifdef CONFIG_FUNCTION_TRACER
+CFLAGS_REMOVE_inflate.o = -pg
+CFLAGS_REMOVE_zmem.o = -pg
+CFLAGS_REMOVE_inftrees.o = -pg
+CFLAGS_REMOVE_inffast.o = -pg
+endif
+
+KASAN_SANITIZE := n
+KCSAN_SANITIZE := n
+KCOV_INSTRUMENT := n
+GCOV_PROFILE := n
+
+CFLAGS_REMOVE_inflate.o += -fstack-protector -fstack-protector-strong
+CFLAGS_REMOVE_zmem.o += -fstack-protector -fstack-protector-strong
+CFLAGS_REMOVE_inftrees.o += -fstack-protector -fstack-protector-strong
+CFLAGS_REMOVE_inffast.o += -fstack-protector -fstack-protector-strong
+
+quiet_cmd_copy_zlib = COPY    $@
+      cmd_copy_zlib = cat $< > $@
+
+$(addprefix $(obj)/,$(zlib)): $(obj)/%: $(srctree)/lib/zlib_inflate/%
+	$(call cmd,copy_zlib)
+
+clean-files	:= $(zlib)
diff --git a/arch/xtensa/boot/lib/zmem.c b/arch/xtensa/boot/lib/zmem.c
new file mode 100644
index 000000000..b89189355
--- /dev/null
+++ b/arch/xtensa/boot/lib/zmem.c
@@ -0,0 +1,81 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/zlib.h>
+
+/* bits taken from ppc */
+
+extern void *avail_ram, *end_avail;
+void gunzip(void *dst, int dstlen, unsigned char *src, int *lenp);
+
+static void exit(void)
+{
+  for (;;);
+}
+
+static void *zalloc(unsigned int size)
+{
+        void *p = avail_ram;
+
+        size = (size + 7) & -8;
+        avail_ram += size;
+        if (avail_ram > end_avail) {
+                //puts("oops... out of memory\n");
+                //pause();
+                exit ();
+        }
+        return p;
+}
+
+#define HEAD_CRC        2
+#define EXTRA_FIELD     4
+#define ORIG_NAME       8
+#define COMMENT         0x10
+#define RESERVED        0xe0
+
+#define DEFLATED        8
+
+void gunzip (void *dst, int dstlen, unsigned char *src, int *lenp)
+{
+	z_stream s;
+	int r, i, flags;
+
+        /* skip header */
+        i = 10;
+        flags = src[3];
+        if (src[2] != DEFLATED || (flags & RESERVED) != 0) {
+                //puts("bad gzipped data\n");
+                exit();
+        }
+        if ((flags & EXTRA_FIELD) != 0)
+                i = 12 + src[10] + (src[11] << 8);
+        if ((flags & ORIG_NAME) != 0)
+                while (src[i++] != 0)
+                        ;
+        if ((flags & COMMENT) != 0)
+                while (src[i++] != 0)
+                        ;
+        if ((flags & HEAD_CRC) != 0)
+                i += 2;
+        if (i >= *lenp) {
+                //puts("gunzip: ran out of data in header\n");
+                exit();
+        }
+
+	s.workspace = zalloc(zlib_inflate_workspacesize());
+        r = zlib_inflateInit2(&s, -MAX_WBITS);
+        if (r != Z_OK) {
+                //puts("inflateInit2 returned "); puthex(r); puts("\n");
+                exit();
+        }
+        s.next_in = src + i;
+        s.avail_in = *lenp - i;
+        s.next_out = dst;
+        s.avail_out = dstlen;
+        r = zlib_inflate(&s, Z_FINISH);
+        if (r != Z_OK && r != Z_STREAM_END) {
+                //puts("inflate returned "); puthex(r); puts("\n");
+                exit();
+        }
+        *lenp = s.next_out - (unsigned char *) dst;
+        zlib_inflateEnd(&s);
+}
+
diff --git a/arch/xtensa/configs/audio_kc705_defconfig b/arch/xtensa/configs/audio_kc705_defconfig
new file mode 100644
index 000000000..0772b5a24
--- /dev/null
+++ b/arch/xtensa/configs/audio_kc705_defconfig
@@ -0,0 +1,134 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_FHANDLE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CPUSETS=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_MEMCG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PROFILING=y
+CONFIG_MODULES=y
+CONFIG_MODULE_UNLOAD=y
+CONFIG_XTENSA_VARIANT_CUSTOM=y
+CONFIG_XTENSA_VARIANT_CUSTOM_NAME="test_kc705_hifi"
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_PREEMPT=y
+CONFIG_HIGHMEM=y
+# CONFIG_PCI is not set
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=0x38000000@0"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="kc705"
+# CONFIG_COMPACTION is not set
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_PM=y
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_MTD=y
+CONFIG_MTD_CFI=y
+CONFIG_MTD_JEDECPROBE=y
+CONFIG_MTD_CFI_INTELEXT=y
+CONFIG_MTD_CFI_AMDSTD=y
+CONFIG_MTD_CFI_STAA=y
+CONFIG_MTD_PHYSMAP_OF=y
+CONFIG_MTD_UBI=y
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+# CONFIG_NET_VENDOR_ARC is not set
+# CONFIG_NET_VENDOR_BROADCOM is not set
+# CONFIG_NET_VENDOR_INTEL is not set
+# CONFIG_NET_VENDOR_MARVELL is not set
+# CONFIG_NET_VENDOR_MICREL is not set
+# CONFIG_NET_VENDOR_NATSEMI is not set
+# CONFIG_NET_VENDOR_SAMSUNG is not set
+# CONFIG_NET_VENDOR_SEEQ is not set
+# CONFIG_NET_VENDOR_SMSC is not set
+# CONFIG_NET_VENDOR_STMICRO is not set
+# CONFIG_NET_VENDOR_VIA is not set
+# CONFIG_NET_VENDOR_WIZNET is not set
+CONFIG_MARVELL_PHY=y
+# CONFIG_WLAN is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
+CONFIG_HW_RANDOM=y
+CONFIG_I2C=y
+CONFIG_I2C_OCORES=y
+CONFIG_SPI=y
+CONFIG_SPI_XTENSA_XTFPGA=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+CONFIG_SOUND=y
+CONFIG_SND=y
+CONFIG_SND_SOC=y
+CONFIG_SND_SOC_XTFPGA_I2S=y
+CONFIG_SND_SOC_TLV320AIC23_SPI=y
+CONFIG_SND_SIMPLE_CARD=y
+# CONFIG_USB_SUPPORT is not set
+CONFIG_COMMON_CLK_CDCE706=y
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+CONFIG_UBIFS_FS=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_LOCKUP_DETECTOR=y
+# CONFIG_SCHED_DEBUG is not set
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+# CONFIG_S32C1I_SELFTEST is not set
diff --git a/arch/xtensa/configs/cadence_csp_defconfig b/arch/xtensa/configs/cadence_csp_defconfig
new file mode 100644
index 000000000..81a057f25
--- /dev/null
+++ b/arch/xtensa/configs/cadence_csp_defconfig
@@ -0,0 +1,113 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CPUSETS=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_INITRAMFS_SOURCE="$$KERNEL_INITRAMFS_SOURCE"
+# CONFIG_RD_BZIP2 is not set
+# CONFIG_RD_LZMA is not set
+# CONFIG_RD_XZ is not set
+# CONFIG_RD_LZO is not set
+# CONFIG_RD_LZ4 is not set
+CONFIG_CC_OPTIMIZE_FOR_SIZE=y
+CONFIG_EXPERT=y
+CONFIG_PROFILING=y
+CONFIG_MODULES=y
+CONFIG_MODULE_FORCE_LOAD=y
+CONFIG_MODULE_UNLOAD=y
+CONFIG_MODULE_FORCE_UNLOAD=y
+CONFIG_XTENSA_VARIANT_CUSTOM=y
+CONFIG_XTENSA_VARIANT_CUSTOM_NAME="csp"
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_PREEMPT=y
+CONFIG_HIGHMEM=y
+# CONFIG_PCI is not set
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="csp"
+# CONFIG_COMPACTION is not set
+CONFIG_XTFPGA_LCD=y
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_INET_DIAG is not set
+# CONFIG_IPV6 is not set
+# CONFIG_WIRELESS is not set
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+# CONFIG_FW_LOADER is not set
+CONFIG_MTD=y
+CONFIG_MTD_CFI=y
+CONFIG_MTD_JEDECPROBE=y
+CONFIG_MTD_CFI_INTELEXT=y
+CONFIG_MTD_CFI_AMDSTD=y
+CONFIG_MTD_CFI_STAA=y
+CONFIG_MTD_PHYSMAP_OF=y
+CONFIG_MTD_UBI=y
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+CONFIG_LEGACY_PTY_COUNT=16
+CONFIG_SERIAL_XILINX_PS_UART=y
+CONFIG_SERIAL_XILINX_PS_UART_CONSOLE=y
+CONFIG_HW_RANDOM=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_USB_SUPPORT is not set
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+# CONFIG_MISC_FILESYSTEMS is not set
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_LOCKUP_DETECTOR=y
+# CONFIG_SCHED_DEBUG is not set
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_PROVE_LOCKING=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_RCU_TRACE=y
+CONFIG_FUNCTION_TRACER=y
+# CONFIG_S32C1I_SELFTEST is not set
+# CONFIG_CRYPTO_ECHAINIV is not set
+CONFIG_CRYPTO_DEFLATE=y
+CONFIG_CRYPTO_LZO=y
+# CONFIG_CRYPTO_HW is not set
diff --git a/arch/xtensa/configs/common_defconfig b/arch/xtensa/configs/common_defconfig
new file mode 100644
index 000000000..09e4a1d9d
--- /dev/null
+++ b/arch/xtensa/configs/common_defconfig
@@ -0,0 +1,54 @@
+CONFIG_PCI=y
+CONFIG_SYSVIPC=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_LOG_BUF_SHIFT=14
+CONFIG_MODULES=y
+CONFIG_MODVERSIONS=y
+CONFIG_XTENSA_PLATFORM_XT2000=y
+CONFIG_GENERIC_CALIBRATE_DELAY=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="console=ttyS0,38400 ip=bootp root=nfs nfsroot=/opt/montavista/pro/devkit/xtensa/linux_be/target memmap=128M@0"
+CONFIG_BINFMT_MISC=y
+CONFIG_NET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_ADVANCED_ROUTER=y
+CONFIG_IP_MULTIPLE_TABLES=y
+CONFIG_IP_ROUTE_MULTIPATH=y
+CONFIG_IP_ROUTE_VERBOSE=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NET_SCHED=y
+CONFIG_NET_SCH_CBQ=m
+CONFIG_NET_SCH_HTB=m
+CONFIG_NET_SCH_PRIO=m
+CONFIG_NET_SCH_RED=m
+CONFIG_NET_SCH_SFQ=m
+CONFIG_NET_SCH_TEQL=m
+CONFIG_NET_SCH_TBF=m
+CONFIG_NET_SCH_GRED=m
+CONFIG_NET_SCH_DSMARK=m
+CONFIG_NET_CLS_ROUTE4=m
+CONFIG_NET_CLS_FW=m
+CONFIG_NET_CLS_U32=m
+CONFIG_NET_CLS_RSVP=m
+CONFIG_NET_CLS_RSVP6=m
+CONFIG_NETDEVICES=y
+CONFIG_DUMMY=y
+# CONFIG_NET_VENDOR_3COM is not set
+# CONFIG_INPUT_MOUSEDEV_PSAUX is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO_SERPORT is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+# CONFIG_VGA_CONSOLE is not set
+CONFIG_NFS_FS=y
+CONFIG_ROOT_NFS=y
+# CONFIG_FRAME_POINTER is not set
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_DEBUG_KERNEL=y
diff --git a/arch/xtensa/configs/generic_kc705_defconfig b/arch/xtensa/configs/generic_kc705_defconfig
new file mode 100644
index 000000000..8327768bb
--- /dev/null
+++ b/arch/xtensa/configs/generic_kc705_defconfig
@@ -0,0 +1,122 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_FHANDLE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CPUSETS=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_MEMCG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PROFILING=y
+CONFIG_MODULES=y
+CONFIG_MODULE_UNLOAD=y
+CONFIG_XTENSA_VARIANT_DC233C=y
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_PREEMPT=y
+CONFIG_HIGHMEM=y
+# CONFIG_PCI is not set
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=0x38000000@0"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="kc705"
+# CONFIG_COMPACTION is not set
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_MTD=y
+CONFIG_MTD_CFI=y
+CONFIG_MTD_JEDECPROBE=y
+CONFIG_MTD_CFI_INTELEXT=y
+CONFIG_MTD_CFI_AMDSTD=y
+CONFIG_MTD_CFI_STAA=y
+CONFIG_MTD_PHYSMAP_OF=y
+CONFIG_MTD_UBI=y
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+# CONFIG_NET_VENDOR_ARC is not set
+# CONFIG_NET_VENDOR_BROADCOM is not set
+# CONFIG_NET_VENDOR_INTEL is not set
+# CONFIG_NET_VENDOR_MARVELL is not set
+# CONFIG_NET_VENDOR_MICREL is not set
+# CONFIG_NET_VENDOR_NATSEMI is not set
+# CONFIG_NET_VENDOR_SAMSUNG is not set
+# CONFIG_NET_VENDOR_SEEQ is not set
+# CONFIG_NET_VENDOR_SMSC is not set
+# CONFIG_NET_VENDOR_STMICRO is not set
+# CONFIG_NET_VENDOR_VIA is not set
+# CONFIG_NET_VENDOR_WIZNET is not set
+CONFIG_MARVELL_PHY=y
+# CONFIG_WLAN is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
+CONFIG_HW_RANDOM=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_USB_SUPPORT is not set
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+CONFIG_UBIFS_FS=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_LOCKUP_DETECTOR=y
+# CONFIG_SCHED_DEBUG is not set
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+CONFIG_LD_NO_RELAX=y
+# CONFIG_S32C1I_SELFTEST is not set
diff --git a/arch/xtensa/configs/iss_defconfig b/arch/xtensa/configs/iss_defconfig
new file mode 100644
index 000000000..324266824
--- /dev/null
+++ b/arch/xtensa/configs/iss_defconfig
@@ -0,0 +1,30 @@
+CONFIG_SYSVIPC=y
+CONFIG_LOG_BUF_SHIFT=14
+CONFIG_EXPERT=y
+# CONFIG_PCI is not set
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="console=ttyS0,38400 eth0=tuntap,,tap0 ip=192.168.168.5:192.168.168.1 root=nfs nfsroot=192.168.168.1:/opt/montavista/pro/devkit/xtensa/linux_be/target memmap=128M@0"
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+# CONFIG_STANDALONE is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_DNOTIFY is not set
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+# CONFIG_FRAME_POINTER is not set
+CONFIG_DETECT_HUNG_TASK=y
diff --git a/arch/xtensa/configs/nommu_kc705_defconfig b/arch/xtensa/configs/nommu_kc705_defconfig
new file mode 100644
index 000000000..fef9ecc37
--- /dev/null
+++ b/arch/xtensa/configs/nommu_kc705_defconfig
@@ -0,0 +1,123 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_FHANDLE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CPUSETS=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_MEMCG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+# CONFIG_RD_BZIP2 is not set
+# CONFIG_RD_LZMA is not set
+# CONFIG_RD_XZ is not set
+# CONFIG_RD_LZO is not set
+# CONFIG_RD_LZ4 is not set
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PERF_EVENTS=y
+CONFIG_MODULES=y
+CONFIG_MODULE_UNLOAD=y
+CONFIG_XTENSA_VARIANT_CUSTOM=y
+CONFIG_XTENSA_VARIANT_CUSTOM_NAME="de212"
+# CONFIG_XTENSA_VARIANT_MMU is not set
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_PREEMPT=y
+CONFIG_MEMMAP_CACHEATTR=0xfff2442f
+# CONFIG_PCI is not set
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="earlycon=uart8250,mmio32native,0x9d050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=256M@0x60000000"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="kc705_nommu"
+CONFIG_BINFMT_FLAT=y
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_MTD=y
+CONFIG_MTD_CFI=y
+CONFIG_MTD_JEDECPROBE=y
+CONFIG_MTD_CFI_INTELEXT=y
+CONFIG_MTD_CFI_AMDSTD=y
+CONFIG_MTD_CFI_STAA=y
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+# CONFIG_NET_VENDOR_ARC is not set
+# CONFIG_NET_VENDOR_BROADCOM is not set
+# CONFIG_NET_VENDOR_INTEL is not set
+# CONFIG_NET_VENDOR_MARVELL is not set
+# CONFIG_NET_VENDOR_MICREL is not set
+# CONFIG_NET_VENDOR_NATSEMI is not set
+# CONFIG_NET_VENDOR_SEEQ is not set
+# CONFIG_NET_VENDOR_STMICRO is not set
+# CONFIG_NET_VENDOR_VIA is not set
+# CONFIG_NET_VENDOR_WIZNET is not set
+CONFIG_MARVELL_PHY=y
+# CONFIG_WLAN is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
+CONFIG_HW_RANDOM=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_USB_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_JFFS2_FS=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+# CONFIG_FRAME_POINTER is not set
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_DEBUG_VM=y
+CONFIG_DEBUG_NOMMU_REGIONS=y
+CONFIG_DEBUG_SHIRQ=y
+CONFIG_LOCKUP_DETECTOR=y
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+# CONFIG_LD_NO_RELAX is not set
+# CONFIG_CRYPTO_ECHAINIV is not set
diff --git a/arch/xtensa/configs/smp_lx200_defconfig b/arch/xtensa/configs/smp_lx200_defconfig
new file mode 100644
index 000000000..b0a8fe3de
--- /dev/null
+++ b/arch/xtensa/configs/smp_lx200_defconfig
@@ -0,0 +1,126 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_FHANDLE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CPUSETS=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_MEMCG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PROFILING=y
+CONFIG_MODULES=y
+CONFIG_MODULE_UNLOAD=y
+CONFIG_XTENSA_VARIANT_CUSTOM=y
+CONFIG_XTENSA_VARIANT_CUSTOM_NAME="test_mmuhifi_c3"
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_PREEMPT=y
+CONFIG_HAVE_SMP=y
+CONFIG_SMP=y
+CONFIG_HOTPLUG_CPU=y
+# CONFIG_INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX is not set
+# CONFIG_PCI is not set
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=96M@0"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="lx200mx"
+# CONFIG_COMPACTION is not set
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_MTD=y
+CONFIG_MTD_CFI=y
+CONFIG_MTD_JEDECPROBE=y
+CONFIG_MTD_CFI_INTELEXT=y
+CONFIG_MTD_CFI_AMDSTD=y
+CONFIG_MTD_CFI_STAA=y
+CONFIG_MTD_PHYSMAP_OF=y
+CONFIG_MTD_UBI=y
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+# CONFIG_NET_VENDOR_ARC is not set
+# CONFIG_NET_VENDOR_BROADCOM is not set
+# CONFIG_NET_VENDOR_INTEL is not set
+# CONFIG_NET_VENDOR_MARVELL is not set
+# CONFIG_NET_VENDOR_MICREL is not set
+# CONFIG_NET_VENDOR_NATSEMI is not set
+# CONFIG_NET_VENDOR_SAMSUNG is not set
+# CONFIG_NET_VENDOR_SEEQ is not set
+# CONFIG_NET_VENDOR_SMSC is not set
+# CONFIG_NET_VENDOR_STMICRO is not set
+# CONFIG_NET_VENDOR_VIA is not set
+# CONFIG_NET_VENDOR_WIZNET is not set
+CONFIG_MARVELL_PHY=y
+# CONFIG_WLAN is not set
+# CONFIG_INPUT_MOUSEDEV is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
+CONFIG_HW_RANDOM=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_USB_SUPPORT is not set
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+CONFIG_UBIFS_FS=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_DEBUG_VM=y
+CONFIG_LOCKUP_DETECTOR=y
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+CONFIG_LD_NO_RELAX=y
+# CONFIG_S32C1I_SELFTEST is not set
diff --git a/arch/xtensa/configs/virt_defconfig b/arch/xtensa/configs/virt_defconfig
new file mode 100644
index 000000000..982dd67ed
--- /dev/null
+++ b/arch/xtensa/configs/virt_defconfig
@@ -0,0 +1,110 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_PREEMPT=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_MEMCG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PERF_EVENTS=y
+CONFIG_XTENSA_VARIANT_DC233C=y
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_XTENSA_KSEG_512M=y
+CONFIG_HIGHMEM=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=0x80000000@0"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="virt"
+# CONFIG_PARSE_BOOTPARAM is not set
+CONFIG_JUMP_LABEL=y
+CONFIG_MODULES=y
+CONFIG_MODULE_UNLOAD=y
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+# CONFIG_COMPACTION is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_PCI=y
+CONFIG_PCI_HOST_GENERIC=y
+CONFIG_UEVENT_HELPER=y
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_VIRTIO_BLK=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+CONFIG_VIRTIO_NET=y
+# CONFIG_ETHERNET is not set
+# CONFIG_WLAN is not set
+CONFIG_INPUT_MOUSEDEV=y
+CONFIG_INPUT_EVDEV=y
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_LEGACY_PTYS is not set
+CONFIG_HW_RANDOM=y
+CONFIG_HW_RANDOM_VIRTIO=y
+# CONFIG_HWMON is not set
+CONFIG_DRM=y
+CONFIG_DRM_VGEM=y
+CONFIG_DRM_VIRTIO_GPU=y
+CONFIG_FB_MODE_HELPERS=y
+# CONFIG_VGA_CONSOLE is not set
+CONFIG_FRAMEBUFFER_CONSOLE=y
+CONFIG_LOGO=y
+# CONFIG_USB_SUPPORT is not set
+CONFIG_VIRTIO_PCI=y
+CONFIG_VIRTIO_INPUT=y
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_CRYPTO_ECHAINIV=y
+CONFIG_CRYPTO_DEFLATE=y
+CONFIG_CRYPTO_LZO=y
+CONFIG_CRYPTO_DEV_VIRTIO=y
+CONFIG_FONTS=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+# CONFIG_SCHED_DEBUG is not set
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+# CONFIG_S32C1I_SELFTEST is not set
diff --git a/arch/xtensa/configs/xip_kc705_defconfig b/arch/xtensa/configs/xip_kc705_defconfig
new file mode 100644
index 000000000..7c61f7c26
--- /dev/null
+++ b/arch/xtensa/configs/xip_kc705_defconfig
@@ -0,0 +1,114 @@
+CONFIG_SYSVIPC=y
+CONFIG_POSIX_MQUEUE=y
+CONFIG_NO_HZ_IDLE=y
+CONFIG_HIGH_RES_TIMERS=y
+CONFIG_PREEMPT=y
+CONFIG_IRQ_TIME_ACCOUNTING=y
+CONFIG_BSD_PROCESS_ACCT=y
+CONFIG_MEMCG=y
+CONFIG_CGROUP_FREEZER=y
+CONFIG_CGROUP_DEVICE=y
+CONFIG_CGROUP_CPUACCT=y
+CONFIG_CGROUP_DEBUG=y
+CONFIG_NAMESPACES=y
+CONFIG_SCHED_AUTOGROUP=y
+CONFIG_RELAY=y
+CONFIG_BLK_DEV_INITRD=y
+CONFIG_EXPERT=y
+CONFIG_KALLSYMS_ALL=y
+CONFIG_PROFILING=y
+CONFIG_XTENSA_VARIANT_DC233C=y
+CONFIG_XTENSA_UNALIGNED_USER=y
+CONFIG_XIP_KERNEL=y
+CONFIG_XIP_DATA_ADDR=0xd0000000
+CONFIG_KERNEL_VIRTUAL_ADDRESS=0xe6000000
+CONFIG_KERNEL_LOAD_ADDRESS=0xf6000000
+CONFIG_XTENSA_KSEG_512M=y
+CONFIG_HIGHMEM=y
+CONFIG_XTENSA_PLATFORM_XTFPGA=y
+CONFIG_CMDLINE_BOOL=y
+CONFIG_CMDLINE="earlycon=uart8250,mmio32native,0xfd050020,115200n8 console=ttyS0,115200n8 ip=dhcp root=/dev/nfs rw debug memmap=0x38000000@0"
+CONFIG_USE_OF=y
+CONFIG_BUILTIN_DTB_NAME="kc705"
+# CONFIG_PARSE_BOOTPARAM is not set
+# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set
+# CONFIG_COMPACTION is not set
+CONFIG_NET=y
+CONFIG_PACKET=y
+CONFIG_UNIX=y
+CONFIG_INET=y
+CONFIG_IP_MULTICAST=y
+CONFIG_IP_PNP=y
+CONFIG_IP_PNP_DHCP=y
+CONFIG_IP_PNP_BOOTP=y
+CONFIG_IP_PNP_RARP=y
+# CONFIG_IPV6 is not set
+CONFIG_NETFILTER=y
+# CONFIG_WIRELESS is not set
+CONFIG_UEVENT_HELPER=y
+CONFIG_DEVTMPFS=y
+CONFIG_DEVTMPFS_MOUNT=y
+# CONFIG_STANDALONE is not set
+CONFIG_BLK_DEV_LOOP=y
+CONFIG_BLK_DEV_RAM=y
+CONFIG_SCSI=y
+CONFIG_BLK_DEV_SD=y
+CONFIG_NETDEVICES=y
+# CONFIG_NET_VENDOR_ARC is not set
+# CONFIG_NET_VENDOR_BROADCOM is not set
+# CONFIG_NET_VENDOR_INTEL is not set
+# CONFIG_NET_VENDOR_MARVELL is not set
+# CONFIG_NET_VENDOR_MICREL is not set
+# CONFIG_NET_VENDOR_NATSEMI is not set
+# CONFIG_NET_VENDOR_SAMSUNG is not set
+# CONFIG_NET_VENDOR_SEEQ is not set
+# CONFIG_NET_VENDOR_SMSC is not set
+# CONFIG_NET_VENDOR_STMICRO is not set
+# CONFIG_NET_VENDOR_VIA is not set
+# CONFIG_NET_VENDOR_WIZNET is not set
+CONFIG_MARVELL_PHY=y
+# CONFIG_WLAN is not set
+# CONFIG_INPUT_KEYBOARD is not set
+# CONFIG_INPUT_MOUSE is not set
+# CONFIG_SERIO is not set
+CONFIG_SERIAL_8250=y
+CONFIG_SERIAL_8250_CONSOLE=y
+CONFIG_SERIAL_OF_PLATFORM=y
+# CONFIG_HWMON is not set
+CONFIG_WATCHDOG=y
+CONFIG_WATCHDOG_NOWAYOUT=y
+CONFIG_SOFT_WATCHDOG=y
+# CONFIG_VGA_CONSOLE is not set
+# CONFIG_USB_SUPPORT is not set
+# CONFIG_IOMMU_SUPPORT is not set
+CONFIG_EXT4_FS=y
+CONFIG_FANOTIFY=y
+CONFIG_VFAT_FS=y
+CONFIG_PROC_KCORE=y
+CONFIG_TMPFS=y
+CONFIG_TMPFS_POSIX_ACL=y
+CONFIG_NFS_FS=y
+CONFIG_NFS_V4=y
+CONFIG_NFS_SWAP=y
+CONFIG_ROOT_NFS=y
+CONFIG_SUNRPC_DEBUG=y
+CONFIG_NLS_CODEPAGE_437=y
+CONFIG_NLS_ISO8859_1=y
+CONFIG_CRYPTO_ECHAINIV=y
+CONFIG_CRYPTO_DEFLATE=y
+CONFIG_CRYPTO_LZO=y
+CONFIG_PRINTK_TIME=y
+CONFIG_DYNAMIC_DEBUG=y
+CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y
+CONFIG_MAGIC_SYSRQ=y
+CONFIG_DETECT_HUNG_TASK=y
+# CONFIG_SCHED_DEBUG is not set
+CONFIG_SCHEDSTATS=y
+CONFIG_DEBUG_RT_MUTEXES=y
+CONFIG_DEBUG_SPINLOCK=y
+CONFIG_DEBUG_MUTEXES=y
+CONFIG_DEBUG_ATOMIC_SLEEP=y
+CONFIG_STACKTRACE=y
+CONFIG_RCU_TRACE=y
+# CONFIG_FTRACE is not set
+# CONFIG_S32C1I_SELFTEST is not set
diff --git a/arch/xtensa/include/asm/Kbuild b/arch/xtensa/include/asm/Kbuild
new file mode 100644
index 000000000..e57af6192
--- /dev/null
+++ b/arch/xtensa/include/asm/Kbuild
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0
+generated-y += syscall_table.h
+generic-y += extable.h
+generic-y += kvm_para.h
+generic-y += mcs_spinlock.h
+generic-y += parport.h
+generic-y += qrwlock.h
+generic-y += qspinlock.h
+generic-y += ring_buffer.h
+generic-y += user.h
+generic-y += text-patching.h
diff --git a/arch/xtensa/include/asm/asm-offsets.h b/arch/xtensa/include/asm/asm-offsets.h
new file mode 100644
index 000000000..d370ee36a
--- /dev/null
+++ b/arch/xtensa/include/asm/asm-offsets.h
@@ -0,0 +1 @@
+#include <generated/asm-offsets.h>
diff --git a/arch/xtensa/include/asm/asm-prototypes.h b/arch/xtensa/include/asm/asm-prototypes.h
new file mode 100644
index 000000000..b0da61812
--- /dev/null
+++ b/arch/xtensa/include/asm/asm-prototypes.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_PROTOTYPES_H
+#define __ASM_PROTOTYPES_H
+
+#include <asm/cacheflush.h>
+#include <asm/checksum.h>
+#include <asm/ftrace.h>
+#include <asm/page.h>
+#include <asm/string.h>
+#include <asm/uaccess.h>
+
+#include <asm-generic/asm-prototypes.h>
+
+/*
+ * gcc internal math functions
+ */
+long long __ashrdi3(long long, int);
+long long __ashldi3(long long, int);
+long long __bswapdi2(long long);
+int __bswapsi2(int);
+long long __lshrdi3(long long, int);
+int __divsi3(int, int);
+int __modsi3(int, int);
+int __mulsi3(int, int);
+unsigned int __udivsi3(unsigned int, unsigned int);
+unsigned int __umodsi3(unsigned int, unsigned int);
+unsigned long long __umulsidi3(unsigned int, unsigned int);
+
+#endif /* __ASM_PROTOTYPES_H */
diff --git a/arch/xtensa/include/asm/asm-uaccess.h b/arch/xtensa/include/asm/asm-uaccess.h
new file mode 100644
index 000000000..7cec86913
--- /dev/null
+++ b/arch/xtensa/include/asm/asm-uaccess.h
@@ -0,0 +1,84 @@
+/*
+ * include/asm-xtensa/uaccess.h
+ *
+ * User space memory access functions
+ *
+ * These routines provide basic accessing functions to the user memory
+ * space for the kernel. This header file provides functions such as:
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_ASM_UACCESS_H
+#define _XTENSA_ASM_UACCESS_H
+
+#include <linux/errno.h>
+#include <asm/types.h>
+
+#include <asm/current.h>
+#include <asm/asm-offsets.h>
+#include <asm/processor.h>
+
+/*
+ * user_ok determines whether the access to user-space memory is allowed.
+ * See the equivalent C-macro version below for clarity.
+ *
+ * On error, user_ok branches to a label indicated by parameter
+ * <error>.  This implies that the macro falls through to the next
+ * instruction on success.
+ *
+ * Note that while this macro can be used independently, we designed
+ * in for optimal use in the access_ok macro below (i.e., we fall
+ * through on success).
+ *
+ * On Entry:
+ * 	<aa>	register containing memory address
+ * 	<as>	register containing memory size
+ * 	<at>	temp register
+ * 	<error>	label to branch to on error; implies fall-through
+ * 		macro on success
+ * On Exit:
+ * 	<aa>	preserved
+ * 	<as>	preserved
+ * 	<at>	destroyed (actually, (TASK_SIZE + 1 - size))
+ */
+	.macro	user_ok	aa, as, at, error
+	movi	\at, __XTENSA_UL_CONST(TASK_SIZE)
+	bgeu	\as, \at, \error
+	sub	\at, \at, \as
+	bgeu	\aa, \at, \error
+	.endm
+
+/*
+ * access_ok determines whether a memory access is allowed.  See the
+ * equivalent C-macro version below for clarity.
+ *
+ * On error, access_ok branches to a label indicated by parameter
+ * <error>.  This implies that the macro falls through to the next
+ * instruction on success.
+ *
+ * Note that we assume success is the common case, and we optimize the
+ * branch fall-through case on success.
+ *
+ * On Entry:
+ * 	<aa>	register containing memory address
+ * 	<as>	register containing memory size
+ * 	<at>	temp register
+ * 	<sp>
+ * 	<error>	label to branch to on error; implies fall-through
+ * 		macro on success
+ * On Exit:
+ * 	<aa>	preserved
+ * 	<as>	preserved
+ * 	<at>	destroyed
+ */
+	.macro	access_ok  aa, as, at, sp, error
+	user_ok    \aa, \as, \at, \error
+.Laccess_ok_\@:
+	.endm
+
+#endif	/* _XTENSA_ASM_UACCESS_H */
diff --git a/arch/xtensa/include/asm/asmmacro.h b/arch/xtensa/include/asm/asmmacro.h
new file mode 100644
index 000000000..a52d49a16
--- /dev/null
+++ b/arch/xtensa/include/asm/asmmacro.h
@@ -0,0 +1,340 @@
+/*
+ * include/asm-xtensa/asmmacro.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_ASMMACRO_H
+#define _XTENSA_ASMMACRO_H
+
+#include <linux/export.h>
+#include <asm/core.h>
+
+/*
+ * Some little helpers for loops. Use zero-overhead-loops
+ * where applicable and if supported by the processor.
+ *
+ * __loopi ar, at, size, inc
+ *         ar	register initialized with the start address
+ *	   at	scratch register used by macro
+ *	   size	size immediate value
+ *	   inc	increment
+ *
+ * __loops ar, as, at, inc_log2[, mask_log2][, cond][, ncond]
+ *	   ar	register initialized with the start address
+ *	   as	register initialized with the size
+ *	   at	scratch register use by macro
+ *	   inc_log2	increment [in log2]
+ *	   mask_log2	mask [in log2]
+ *	   cond		true condition (used in loop'cond')
+ *	   ncond	false condition (used in b'ncond')
+ *
+ * __loop  as
+ *	   restart loop. 'as' register must not have been modified!
+ *
+ * __endla ar, as, incr
+ *	   ar	start address (modified)
+ *	   as	scratch register used by __loops/__loopi macros or
+ *		end address used by __loopt macro
+ *	   inc	increment
+ */
+
+/*
+ * loop for given size as immediate
+ */
+
+	.macro	__loopi ar, at, size, incr
+
+#if XCHAL_HAVE_LOOPS
+		movi	\at, ((\size + \incr - 1) / (\incr))
+		loop	\at, 99f
+#else
+		addi	\at, \ar, \size
+		98:
+#endif
+
+	.endm
+
+/*
+ * loop for given size in register
+ */
+
+	.macro	__loops	ar, as, at, incr_log2, mask_log2, cond, ncond
+
+#if XCHAL_HAVE_LOOPS
+		.ifgt \incr_log2 - 1
+			addi	\at, \as, (1 << \incr_log2) - 1
+			.ifnc \mask_log2,
+				extui	\at, \at, \incr_log2, \mask_log2
+			.else
+				srli	\at, \at, \incr_log2
+			.endif
+		.endif
+		loop\cond	\at, 99f
+#else
+		.ifnc \mask_log2,
+			extui	\at, \as, \incr_log2, \mask_log2
+		.else
+			.ifnc \ncond,
+				srli	\at, \as, \incr_log2
+			.endif
+		.endif
+		.ifnc \ncond,
+			b\ncond	\at, 99f
+
+		.endif
+		.ifnc \mask_log2,
+			slli	\at, \at, \incr_log2
+			add	\at, \ar, \at
+		.else
+			add	\at, \ar, \as
+		.endif
+#endif
+		98:
+
+	.endm
+
+/*
+ * loop from ar to as
+ */
+
+	.macro	__loopt	ar, as, at, incr_log2
+
+#if XCHAL_HAVE_LOOPS
+		sub	\at, \as, \ar
+		.ifgt	\incr_log2 - 1
+			addi	\at, \at, (1 << \incr_log2) - 1
+			srli	\at, \at, \incr_log2
+		.endif
+		loop	\at, 99f
+#else
+		98:
+#endif
+
+	.endm
+
+/*
+ * restart loop. registers must be unchanged
+ */
+
+	.macro	__loop	as
+
+#if XCHAL_HAVE_LOOPS
+		loop	\as, 99f
+#else
+		98:
+#endif
+
+	.endm
+
+/*
+ * end of loop with no increment of the address.
+ */
+
+	.macro	__endl	ar, as
+#if !XCHAL_HAVE_LOOPS
+		bltu	\ar, \as, 98b
+#endif
+		99:
+	.endm
+
+/*
+ * end of loop with increment of the address.
+ */
+
+	.macro	__endla	ar, as, incr
+		addi	\ar, \ar, \incr
+		__endl	\ar \as
+	.endm
+
+/* Load or store instructions that may cause exceptions use the EX macro. */
+
+#define EX(handler)				\
+	.section __ex_table, "a";		\
+	.word	97f, handler;			\
+	.previous				\
+97:
+
+
+/*
+ * Extract unaligned word that is split between two registers w0 and w1
+ * into r regardless of machine endianness. SAR must be loaded with the
+ * starting bit of the word (see __ssa8).
+ */
+
+	.macro __src_b	r, w0, w1
+#ifdef __XTENSA_EB__
+		src	\r, \w0, \w1
+#else
+		src	\r, \w1, \w0
+#endif
+	.endm
+
+/*
+ * Load 2 lowest address bits of r into SAR for __src_b to extract unaligned
+ * word starting at r from two registers loaded from consecutive aligned
+ * addresses covering r regardless of machine endianness.
+ *
+ *      r   0   1   2   3
+ * LE SAR   0   8  16  24
+ * BE SAR  32  24  16   8
+ */
+
+	.macro __ssa8	r
+#ifdef __XTENSA_EB__
+		ssa8b	\r
+#else
+		ssa8l	\r
+#endif
+	.endm
+
+	.macro	do_nsau cnt, val, tmp, a
+#if XCHAL_HAVE_NSA
+	nsau	\cnt, \val
+#else
+	mov	\a, \val
+	movi	\cnt, 0
+	extui	\tmp, \a, 16, 16
+	bnez	\tmp, 0f
+	movi	\cnt, 16
+	slli	\a, \a, 16
+0:
+	extui	\tmp, \a, 24, 8
+	bnez	\tmp, 1f
+	addi	\cnt, \cnt, 8
+	slli	\a, \a, 8
+1:
+	movi	\tmp, __nsau_data
+	extui	\a, \a, 24, 8
+	add	\tmp, \tmp, \a
+	l8ui	\tmp, \tmp, 0
+	add	\cnt, \cnt, \tmp
+#endif /* !XCHAL_HAVE_NSA */
+	.endm
+
+	.macro	do_abs dst, src, tmp
+#if XCHAL_HAVE_ABS
+	abs	\dst, \src
+#else
+	neg	\tmp, \src
+	movgez	\tmp, \src, \src
+	mov	\dst, \tmp
+#endif
+	.endm
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+
+/* Assembly instructions for windowed kernel ABI. */
+#define KABI_W
+/* Assembly instructions for call0 kernel ABI (will be ignored). */
+#define KABI_C0 #
+
+#define XTENSA_FRAME_SIZE_RESERVE	16
+#define XTENSA_SPILL_STACK_RESERVE	32
+
+#define abi_entry(frame_size) \
+	entry sp, (XTENSA_FRAME_SIZE_RESERVE + \
+		   (((frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \
+		    -XTENSA_STACK_ALIGNMENT))
+#define abi_entry_default abi_entry(0)
+
+#define abi_ret(frame_size) retw
+#define abi_ret_default retw
+
+	/* direct call */
+#define abi_call call4
+	/* indirect call */
+#define abi_callx callx4
+	/* outgoing call argument registers */
+#define abi_arg0 a6
+#define abi_arg1 a7
+#define abi_arg2 a8
+#define abi_arg3 a9
+#define abi_arg4 a10
+#define abi_arg5 a11
+	/* return value */
+#define abi_rv a6
+	/* registers preserved across call */
+#define abi_saved0 a2
+#define abi_saved1 a3
+
+	/* none of the above */
+#define abi_tmp0 a4
+#define abi_tmp1 a5
+
+#elif defined(__XTENSA_CALL0_ABI__)
+
+/* Assembly instructions for windowed kernel ABI (will be ignored). */
+#define KABI_W #
+/* Assembly instructions for call0 kernel ABI. */
+#define KABI_C0
+
+#define XTENSA_SPILL_STACK_RESERVE	0
+
+#define abi_entry(frame_size) __abi_entry (frame_size)
+
+	.macro	__abi_entry frame_size
+	.ifgt \frame_size
+	addi sp, sp, -(((\frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \
+		       -XTENSA_STACK_ALIGNMENT)
+	.endif
+	.endm
+
+#define abi_entry_default
+
+#define abi_ret(frame_size) __abi_ret (frame_size)
+
+	.macro	__abi_ret frame_size
+	.ifgt \frame_size
+	addi sp, sp, (((\frame_size) + XTENSA_STACK_ALIGNMENT - 1) & \
+		      -XTENSA_STACK_ALIGNMENT)
+	.endif
+	ret
+	.endm
+
+#define abi_ret_default ret
+
+	/* direct call */
+#define abi_call call0
+	/* indirect call */
+#define abi_callx callx0
+	/* outgoing call argument registers */
+#define abi_arg0 a2
+#define abi_arg1 a3
+#define abi_arg2 a4
+#define abi_arg3 a5
+#define abi_arg4 a6
+#define abi_arg5 a7
+	/* return value */
+#define abi_rv a2
+	/* registers preserved across call */
+#define abi_saved0 a12
+#define abi_saved1 a13
+
+	/* none of the above */
+#define abi_tmp0 a8
+#define abi_tmp1 a9
+
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+#if defined(USER_SUPPORT_WINDOWED)
+/* Assembly instructions for windowed user ABI. */
+#define UABI_W
+/* Assembly instructions for call0 user ABI (will be ignored). */
+#define UABI_C0 #
+#else
+/* Assembly instructions for windowed user ABI (will be ignored). */
+#define UABI_W #
+/* Assembly instructions for call0 user ABI. */
+#define UABI_C0
+#endif
+
+#define __XTENSA_HANDLER	.section ".exception.text", "ax"
+
+#endif /* _XTENSA_ASMMACRO_H */
diff --git a/arch/xtensa/include/asm/atomic.h b/arch/xtensa/include/asm/atomic.h
new file mode 100644
index 000000000..7308b7f77
--- /dev/null
+++ b/arch/xtensa/include/asm/atomic.h
@@ -0,0 +1,269 @@
+/*
+ * include/asm-xtensa/atomic.h
+ *
+ * Atomic operations that C can't guarantee us.  Useful for resource counting..
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2008 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_ATOMIC_H
+#define _XTENSA_ATOMIC_H
+
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <asm/processor.h>
+#include <asm/cmpxchg.h>
+#include <asm/barrier.h>
+
+/*
+ * This Xtensa implementation assumes that the right mechanism
+ * for exclusion is for locking interrupts to level EXCM_LEVEL.
+ *
+ * Locking interrupts looks like this:
+ *
+ *    rsil a14, TOPLEVEL
+ *    <code>
+ *    wsr  a14, PS
+ *    rsync
+ *
+ * Note that a14 is used here because the register allocation
+ * done by the compiler is not guaranteed and a window overflow
+ * may not occur between the rsil and wsr instructions. By using
+ * a14 in the rsil, the machine is guaranteed to be in a state
+ * where no register reference will cause an overflow.
+ */
+
+/**
+ * atomic_read - read atomic variable
+ * @v: pointer of type atomic_t
+ *
+ * Atomically reads the value of @v.
+ */
+#define arch_atomic_read(v)		READ_ONCE((v)->counter)
+
+/**
+ * atomic_set - set atomic variable
+ * @v: pointer of type atomic_t
+ * @i: required value
+ *
+ * Atomically sets the value of @v to @i.
+ */
+#define arch_atomic_set(v,i)		WRITE_ONCE((v)->counter, (i))
+
+#if XCHAL_HAVE_EXCLUSIVE
+#define ATOMIC_OP(op)							\
+static inline void arch_atomic_##op(int i, atomic_t *v)			\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32ex   %[tmp], %[addr]\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32ex   %[result], %[addr]\n"		\
+			"       getex   %[result]\n"			\
+			"       beqz    %[result], 1b\n"		\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp)	\
+			: [i] "a" (i), [addr] "a" (v)			\
+			: "memory"					\
+			);						\
+}									\
+
+#define ATOMIC_OP_RETURN(op)						\
+static inline int arch_atomic_##op##_return(int i, atomic_t *v)		\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32ex   %[tmp], %[addr]\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32ex   %[result], %[addr]\n"		\
+			"       getex   %[result]\n"			\
+			"       beqz    %[result], 1b\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp)	\
+			: [i] "a" (i), [addr] "a" (v)			\
+			: "memory"					\
+			);						\
+									\
+	return result;							\
+}
+
+#define ATOMIC_FETCH_OP(op)						\
+static inline int arch_atomic_fetch_##op(int i, atomic_t *v)		\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32ex   %[tmp], %[addr]\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32ex   %[result], %[addr]\n"		\
+			"       getex   %[result]\n"			\
+			"       beqz    %[result], 1b\n"		\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp)	\
+			: [i] "a" (i), [addr] "a" (v)			\
+			: "memory"					\
+			);						\
+									\
+	return tmp;							\
+}
+
+#elif XCHAL_HAVE_S32C1I
+#define ATOMIC_OP(op)							\
+static inline void arch_atomic_##op(int i, atomic_t * v)		\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32i    %[tmp], %[mem]\n"		\
+			"       wsr     %[tmp], scompare1\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32c1i  %[result], %[mem]\n"		\
+			"       bne     %[result], %[tmp], 1b\n"	\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp),	\
+			  [mem] "+m" (*v)				\
+			: [i] "a" (i)					\
+			: "memory"					\
+			);						\
+}									\
+
+#define ATOMIC_OP_RETURN(op)						\
+static inline int arch_atomic_##op##_return(int i, atomic_t * v)	\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32i    %[tmp], %[mem]\n"		\
+			"       wsr     %[tmp], scompare1\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32c1i  %[result], %[mem]\n"		\
+			"       bne     %[result], %[tmp], 1b\n"	\
+			"       " #op " %[result], %[result], %[i]\n"	\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp),	\
+			  [mem] "+m" (*v)				\
+			: [i] "a" (i)					\
+			: "memory"					\
+			);						\
+									\
+	return result;							\
+}
+
+#define ATOMIC_FETCH_OP(op)						\
+static inline int arch_atomic_fetch_##op(int i, atomic_t * v)		\
+{									\
+	unsigned long tmp;						\
+	int result;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32i    %[tmp], %[mem]\n"		\
+			"       wsr     %[tmp], scompare1\n"		\
+			"       " #op " %[result], %[tmp], %[i]\n"	\
+			"       s32c1i  %[result], %[mem]\n"		\
+			"       bne     %[result], %[tmp], 1b\n"	\
+			: [result] "=&a" (result), [tmp] "=&a" (tmp),	\
+			  [mem] "+m" (*v)				\
+			: [i] "a" (i)					\
+			: "memory"					\
+			);						\
+									\
+	return result;							\
+}
+
+#else /* XCHAL_HAVE_S32C1I */
+
+#define ATOMIC_OP(op)							\
+static inline void arch_atomic_##op(int i, atomic_t * v)		\
+{									\
+	unsigned int vval;						\
+									\
+	__asm__ __volatile__(						\
+			"       rsil    a14, "__stringify(TOPLEVEL)"\n"	\
+			"       l32i    %[result], %[mem]\n"		\
+			"       " #op " %[result], %[result], %[i]\n"	\
+			"       s32i    %[result], %[mem]\n"		\
+			"       wsr     a14, ps\n"			\
+			"       rsync\n"				\
+			: [result] "=&a" (vval), [mem] "+m" (*v)	\
+			: [i] "a" (i)					\
+			: "a14", "memory"				\
+			);						\
+}									\
+
+#define ATOMIC_OP_RETURN(op)						\
+static inline int arch_atomic_##op##_return(int i, atomic_t * v)	\
+{									\
+	unsigned int vval;						\
+									\
+	__asm__ __volatile__(						\
+			"       rsil    a14,"__stringify(TOPLEVEL)"\n"	\
+			"       l32i    %[result], %[mem]\n"		\
+			"       " #op " %[result], %[result], %[i]\n"	\
+			"       s32i    %[result], %[mem]\n"		\
+			"       wsr     a14, ps\n"			\
+			"       rsync\n"				\
+			: [result] "=&a" (vval), [mem] "+m" (*v)	\
+			: [i] "a" (i)					\
+			: "a14", "memory"				\
+			);						\
+									\
+	return vval;							\
+}
+
+#define ATOMIC_FETCH_OP(op)						\
+static inline int arch_atomic_fetch_##op(int i, atomic_t * v)		\
+{									\
+	unsigned int tmp, vval;						\
+									\
+	__asm__ __volatile__(						\
+			"       rsil    a14,"__stringify(TOPLEVEL)"\n"	\
+			"       l32i    %[result], %[mem]\n"		\
+			"       " #op " %[tmp], %[result], %[i]\n"	\
+			"       s32i    %[tmp], %[mem]\n"		\
+			"       wsr     a14, ps\n"			\
+			"       rsync\n"				\
+			: [result] "=&a" (vval), [tmp] "=&a" (tmp),	\
+			  [mem] "+m" (*v)				\
+			: [i] "a" (i)					\
+			: "a14", "memory"				\
+			);						\
+									\
+	return vval;							\
+}
+
+#endif /* XCHAL_HAVE_S32C1I */
+
+#define ATOMIC_OPS(op) ATOMIC_OP(op) ATOMIC_FETCH_OP(op) ATOMIC_OP_RETURN(op)
+
+ATOMIC_OPS(add)
+ATOMIC_OPS(sub)
+
+#define arch_atomic_add_return			arch_atomic_add_return
+#define arch_atomic_sub_return			arch_atomic_sub_return
+#define arch_atomic_fetch_add			arch_atomic_fetch_add
+#define arch_atomic_fetch_sub			arch_atomic_fetch_sub
+
+#undef ATOMIC_OPS
+#define ATOMIC_OPS(op) ATOMIC_OP(op) ATOMIC_FETCH_OP(op)
+
+ATOMIC_OPS(and)
+ATOMIC_OPS(or)
+ATOMIC_OPS(xor)
+
+#define arch_atomic_fetch_and			arch_atomic_fetch_and
+#define arch_atomic_fetch_or			arch_atomic_fetch_or
+#define arch_atomic_fetch_xor			arch_atomic_fetch_xor
+
+#undef ATOMIC_OPS
+#undef ATOMIC_FETCH_OP
+#undef ATOMIC_OP_RETURN
+#undef ATOMIC_OP
+
+#endif /* _XTENSA_ATOMIC_H */
diff --git a/arch/xtensa/include/asm/barrier.h b/arch/xtensa/include/asm/barrier.h
new file mode 100644
index 000000000..898ea397e
--- /dev/null
+++ b/arch/xtensa/include/asm/barrier.h
@@ -0,0 +1,31 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2012 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SYSTEM_H
+#define _XTENSA_SYSTEM_H
+
+#include <asm/core.h>
+
+#define __mb()  ({ __asm__ __volatile__("memw" : : : "memory"); })
+#define __rmb() barrier()
+#define __wmb() __mb()
+
+#ifdef CONFIG_SMP
+#define __smp_mb() __mb()
+#define __smp_rmb() __rmb()
+#define __smp_wmb() __wmb()
+#endif
+
+#if XCHAL_HAVE_S32C1I
+#define __smp_mb__before_atomic()		barrier()
+#define __smp_mb__after_atomic()		barrier()
+#endif
+
+#include <asm-generic/barrier.h>
+
+#endif /* _XTENSA_SYSTEM_H */
diff --git a/arch/xtensa/include/asm/bitops.h b/arch/xtensa/include/asm/bitops.h
new file mode 100644
index 000000000..f7390b676
--- /dev/null
+++ b/arch/xtensa/include/asm/bitops.h
@@ -0,0 +1,218 @@
+/*
+ * include/asm-xtensa/bitops.h
+ *
+ * Atomic operations that C can't guarantee us.Useful for resource counting etc.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_BITOPS_H
+#define _XTENSA_BITOPS_H
+
+#ifndef _LINUX_BITOPS_H
+#error only <linux/bitops.h> can be included directly
+#endif
+
+#include <asm/processor.h>
+#include <asm/byteorder.h>
+#include <asm/barrier.h>
+
+#include <asm-generic/bitops/non-atomic.h>
+
+#if XCHAL_HAVE_NSA
+
+static inline unsigned long __cntlz (unsigned long x)
+{
+	int lz;
+	asm ("nsau %0, %1" : "=r" (lz) : "r" (x));
+	return lz;
+}
+
+/*
+ * ffz: Find first zero in word. Undefined if no zero exists.
+ * bit 0 is the LSB of addr; bit 32 is the LSB of (addr+1).
+ */
+
+static inline int __attribute_const__ ffz(unsigned long x)
+{
+	return 31 - __cntlz(~x & -~x);
+}
+
+/*
+ * __ffs: Find first bit set in word. Return 0 for bit 0
+ */
+
+static inline __attribute_const__ unsigned long __ffs(unsigned long x)
+{
+	return 31 - __cntlz(x & -x);
+}
+
+/*
+ * ffs: Find first bit set in word. This is defined the same way as
+ * the libc and compiler builtin ffs routines, therefore
+ * differs in spirit from the above ffz (man ffs).
+ */
+
+static inline __attribute_const__ int ffs(unsigned long x)
+{
+	return 32 - __cntlz(x & -x);
+}
+
+/*
+ * fls: Find last (most-significant) bit set in word.
+ * Note fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32.
+ */
+
+static inline __attribute_const__ int fls (unsigned int x)
+{
+	return 32 - __cntlz(x);
+}
+
+/**
+ * __fls - find last (most-significant) set bit in a long word
+ * @word: the word to search
+ *
+ * Undefined if no set bit exists, so code should check against 0 first.
+ */
+static inline __attribute_const__ unsigned long __fls(unsigned long word)
+{
+	return 31 - __cntlz(word);
+}
+#else
+
+/* Use the generic implementation if we don't have the nsa/nsau instructions. */
+
+# include <asm-generic/bitops/ffs.h>
+# include <asm-generic/bitops/__ffs.h>
+# include <asm-generic/bitops/ffz.h>
+# include <asm-generic/bitops/fls.h>
+# include <asm-generic/bitops/__fls.h>
+
+#endif
+
+#include <asm-generic/bitops/fls64.h>
+
+#if XCHAL_HAVE_EXCLUSIVE
+
+#define BIT_OP(op, insn, inv)						\
+static inline void arch_##op##_bit(unsigned int bit, volatile unsigned long *p)\
+{									\
+	unsigned long tmp;						\
+	unsigned long mask = 1UL << (bit & 31);				\
+									\
+	p += bit >> 5;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32ex   %[tmp], %[addr]\n"		\
+			"      "insn"   %[tmp], %[tmp], %[mask]\n"	\
+			"       s32ex   %[tmp], %[addr]\n"		\
+			"       getex   %[tmp]\n"			\
+			"       beqz    %[tmp], 1b\n"			\
+			: [tmp] "=&a" (tmp)				\
+			: [mask] "a" (inv mask), [addr] "a" (p)		\
+			: "memory");					\
+}
+
+#define TEST_AND_BIT_OP(op, insn, inv)					\
+static inline int							\
+arch_test_and_##op##_bit(unsigned int bit, volatile unsigned long *p)	\
+{									\
+	unsigned long tmp, value;					\
+	unsigned long mask = 1UL << (bit & 31);				\
+									\
+	p += bit >> 5;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32ex   %[value], %[addr]\n"		\
+			"      "insn"   %[tmp], %[value], %[mask]\n"	\
+			"       s32ex   %[tmp], %[addr]\n"		\
+			"       getex   %[tmp]\n"			\
+			"       beqz    %[tmp], 1b\n"			\
+			: [tmp] "=&a" (tmp), [value] "=&a" (value)	\
+			: [mask] "a" (inv mask), [addr] "a" (p)		\
+			: "memory");					\
+									\
+	return value & mask;						\
+}
+
+#elif XCHAL_HAVE_S32C1I
+
+#define BIT_OP(op, insn, inv)						\
+static inline void arch_##op##_bit(unsigned int bit, volatile unsigned long *p)\
+{									\
+	unsigned long tmp, value;					\
+	unsigned long mask = 1UL << (bit & 31);				\
+									\
+	p += bit >> 5;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32i    %[value], %[mem]\n"		\
+			"       wsr     %[value], scompare1\n"		\
+			"      "insn"   %[tmp], %[value], %[mask]\n"	\
+			"       s32c1i  %[tmp], %[mem]\n"		\
+			"       bne     %[tmp], %[value], 1b\n"		\
+			: [tmp] "=&a" (tmp), [value] "=&a" (value),	\
+			  [mem] "+m" (*p)				\
+			: [mask] "a" (inv mask)				\
+			: "memory");					\
+}
+
+#define TEST_AND_BIT_OP(op, insn, inv)					\
+static inline int							\
+arch_test_and_##op##_bit(unsigned int bit, volatile unsigned long *p)	\
+{									\
+	unsigned long tmp, value;					\
+	unsigned long mask = 1UL << (bit & 31);				\
+									\
+	p += bit >> 5;							\
+									\
+	__asm__ __volatile__(						\
+			"1:     l32i    %[value], %[mem]\n"		\
+			"       wsr     %[value], scompare1\n"		\
+			"      "insn"   %[tmp], %[value], %[mask]\n"	\
+			"       s32c1i  %[tmp], %[mem]\n"		\
+			"       bne     %[tmp], %[value], 1b\n"		\
+			: [tmp] "=&a" (tmp), [value] "=&a" (value),	\
+			  [mem] "+m" (*p)				\
+			: [mask] "a" (inv mask)				\
+			: "memory");					\
+									\
+	return tmp & mask;						\
+}
+
+#else
+
+#define BIT_OP(op, insn, inv)
+#define TEST_AND_BIT_OP(op, insn, inv)
+
+#include <asm-generic/bitops/atomic.h>
+
+#endif /* XCHAL_HAVE_S32C1I */
+
+#define BIT_OPS(op, insn, inv)		\
+	BIT_OP(op, insn, inv)		\
+	TEST_AND_BIT_OP(op, insn, inv)
+
+BIT_OPS(set, "or", )
+BIT_OPS(clear, "and", ~)
+BIT_OPS(change, "xor", )
+
+#undef BIT_OPS
+#undef BIT_OP
+#undef TEST_AND_BIT_OP
+
+#include <asm-generic/bitops/instrumented-atomic.h>
+
+#include <asm-generic/bitops/le.h>
+
+#include <asm-generic/bitops/ext2-atomic-setbit.h>
+
+#include <asm-generic/bitops/hweight.h>
+#include <asm-generic/bitops/lock.h>
+#include <asm-generic/bitops/sched.h>
+
+#endif	/* _XTENSA_BITOPS_H */
diff --git a/arch/xtensa/include/asm/bootparam.h b/arch/xtensa/include/asm/bootparam.h
new file mode 100644
index 000000000..a459ffbaf
--- /dev/null
+++ b/arch/xtensa/include/asm/bootparam.h
@@ -0,0 +1,50 @@
+/*
+ * include/asm-xtensa/bootparam.h
+ *
+ * Definition of the Linux/Xtensa boot parameter structure
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005  Tensilica Inc.
+ *
+ * (Concept borrowed from the 68K port)
+ */
+
+#ifndef _XTENSA_BOOTPARAM_H
+#define _XTENSA_BOOTPARAM_H
+
+#define BP_VERSION 0x0001
+
+#define BP_TAG_COMMAND_LINE	0x1001	/* command line (0-terminated string)*/
+#define BP_TAG_INITRD		0x1002	/* ramdisk addr and size (bp_meminfo) */
+#define BP_TAG_MEMORY		0x1003	/* memory addr and size (bp_meminfo) */
+#define BP_TAG_SERIAL_BAUDRATE	0x1004	/* baud rate of current console. */
+#define BP_TAG_SERIAL_PORT	0x1005	/* serial device of current console */
+#define BP_TAG_FDT		0x1006	/* flat device tree addr */
+
+#define BP_TAG_FIRST		0x7B0B  /* first tag with a version number */
+#define BP_TAG_LAST 		0x7E0B	/* last tag */
+
+#ifndef __ASSEMBLER__
+
+/* All records are aligned to 4 bytes */
+
+typedef struct bp_tag {
+	unsigned short id;	/* tag id */
+	unsigned short size;	/* size of this record excluding the structure*/
+	unsigned long data[];	/* data */
+} bp_tag_t;
+
+struct bp_meminfo {
+	unsigned long type;
+	unsigned long start;
+	unsigned long end;
+};
+
+#define MEMORY_TYPE_CONVENTIONAL	0x1000
+#define MEMORY_TYPE_NONE		0x2000
+
+#endif
+#endif
diff --git a/arch/xtensa/include/asm/cache.h b/arch/xtensa/include/asm/cache.h
new file mode 100644
index 000000000..54e147ac2
--- /dev/null
+++ b/arch/xtensa/include/asm/cache.h
@@ -0,0 +1,40 @@
+/*
+ * include/asm-xtensa/cache.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CACHE_H
+#define _XTENSA_CACHE_H
+
+#include <asm/core.h>
+
+#define L1_CACHE_SHIFT	XCHAL_DCACHE_LINEWIDTH
+#define L1_CACHE_BYTES	XCHAL_DCACHE_LINESIZE
+#define SMP_CACHE_BYTES	L1_CACHE_BYTES
+
+#define DCACHE_WAY_SIZE	(XCHAL_DCACHE_SIZE/XCHAL_DCACHE_WAYS)
+#define ICACHE_WAY_SIZE	(XCHAL_ICACHE_SIZE/XCHAL_ICACHE_WAYS)
+#define DCACHE_WAY_SHIFT (XCHAL_DCACHE_SETWIDTH + XCHAL_DCACHE_LINEWIDTH)
+#define ICACHE_WAY_SHIFT (XCHAL_ICACHE_SETWIDTH + XCHAL_ICACHE_LINEWIDTH)
+
+/* Maximum cache size per way. */
+#if DCACHE_WAY_SIZE >= ICACHE_WAY_SIZE
+# define CACHE_WAY_SIZE DCACHE_WAY_SIZE
+#else
+# define CACHE_WAY_SIZE ICACHE_WAY_SIZE
+#endif
+
+#define ARCH_DMA_MINALIGN	L1_CACHE_BYTES
+
+/*
+ * R/O after init is actually writable, it cannot go to .rodata
+ * according to vmlinux linker script.
+ */
+#define __ro_after_init __read_mostly
+
+#endif	/* _XTENSA_CACHE_H */
diff --git a/arch/xtensa/include/asm/cacheasm.h b/arch/xtensa/include/asm/cacheasm.h
new file mode 100644
index 000000000..34545ecfd
--- /dev/null
+++ b/arch/xtensa/include/asm/cacheasm.h
@@ -0,0 +1,214 @@
+/*
+ * include/asm-xtensa/cacheasm.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006 Tensilica Inc.
+ */
+
+#include <asm/cache.h>
+#include <asm/asmmacro.h>
+#include <linux/stringify.h>
+
+/*
+ * Define cache functions as macros here so that they can be used
+ * by the kernel and boot loader. We should consider moving them to a
+ * library that can be linked by both.
+ *
+ * Locking
+ *
+ *   ___unlock_dcache_all
+ *   ___unlock_icache_all
+ *
+ * Flush and invaldating
+ *
+ *   ___flush_invalidate_dcache_{all|range|page}
+ *   ___flush_dcache_{all|range|page}
+ *   ___invalidate_dcache_{all|range|page}
+ *   ___invalidate_icache_{all|range|page}
+ *
+ */
+
+
+	.macro	__loop_cache_unroll ar at insn size line_width max_immed
+
+	.if	(1 << (\line_width)) > (\max_immed)
+	.set	_reps, 1
+	.elseif	(2 << (\line_width)) > (\max_immed)
+	.set	_reps, 2
+	.else
+	.set	_reps, 4
+	.endif
+
+	__loopi	\ar, \at, \size, (_reps << (\line_width))
+	.set	_index, 0
+	.rep	_reps
+	\insn	\ar, _index << (\line_width)
+	.set	_index, _index + 1
+	.endr
+	__endla	\ar, \at, _reps << (\line_width)
+
+	.endm
+
+
+	.macro	__loop_cache_all ar at insn size line_width max_immed
+
+	movi	\ar, 0
+	__loop_cache_unroll \ar, \at, \insn, \size, \line_width, \max_immed
+
+	.endm
+
+
+	.macro	__loop_cache_range ar as at insn line_width
+
+	extui	\at, \ar, 0, \line_width
+	add	\as, \as, \at
+
+	__loops	\ar, \as, \at, \line_width
+	\insn	\ar, 0
+	__endla	\ar, \at, (1 << (\line_width))
+
+	.endm
+
+
+	.macro	__loop_cache_page ar at insn line_width max_immed
+
+	__loop_cache_unroll \ar, \at, \insn, PAGE_SIZE, \line_width, \max_immed
+
+	.endm
+
+
+	.macro	___unlock_dcache_all ar at
+
+#if XCHAL_DCACHE_LINE_LOCKABLE && XCHAL_DCACHE_SIZE
+	__loop_cache_all \ar \at diu XCHAL_DCACHE_SIZE \
+		XCHAL_DCACHE_LINEWIDTH 240
+#endif
+
+	.endm
+
+
+	.macro	___unlock_icache_all ar at
+
+#if XCHAL_ICACHE_LINE_LOCKABLE && XCHAL_ICACHE_SIZE
+	__loop_cache_all \ar \at iiu XCHAL_ICACHE_SIZE \
+		XCHAL_ICACHE_LINEWIDTH 240
+#endif
+
+	.endm
+
+
+	.macro	___flush_invalidate_dcache_all ar at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_all \ar \at diwbi XCHAL_DCACHE_SIZE \
+		XCHAL_DCACHE_LINEWIDTH 240
+#endif
+
+	.endm
+
+
+	.macro	___flush_dcache_all ar at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_all \ar \at diwb XCHAL_DCACHE_SIZE \
+		XCHAL_DCACHE_LINEWIDTH 240
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_dcache_all ar at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_all \ar \at dii XCHAL_DCACHE_SIZE \
+			 XCHAL_DCACHE_LINEWIDTH 1020
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_icache_all ar at
+
+#if XCHAL_ICACHE_SIZE
+	__loop_cache_all \ar \at iii XCHAL_ICACHE_SIZE \
+			 XCHAL_ICACHE_LINEWIDTH 1020
+#endif
+
+	.endm
+
+
+
+	.macro	___flush_invalidate_dcache_range ar as at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_range \ar \as \at dhwbi XCHAL_DCACHE_LINEWIDTH
+#endif
+
+	.endm
+
+
+	.macro	___flush_dcache_range ar as at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_range \ar \as \at dhwb XCHAL_DCACHE_LINEWIDTH
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_dcache_range ar as at
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_range \ar \as \at dhi XCHAL_DCACHE_LINEWIDTH
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_icache_range ar as at
+
+#if XCHAL_ICACHE_SIZE
+	__loop_cache_range \ar \as \at ihi XCHAL_ICACHE_LINEWIDTH
+#endif
+
+	.endm
+
+
+
+	.macro	___flush_invalidate_dcache_page ar as
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_page \ar \as dhwbi XCHAL_DCACHE_LINEWIDTH 1020
+#endif
+
+	.endm
+
+
+	.macro ___flush_dcache_page ar as
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_page \ar \as dhwb XCHAL_DCACHE_LINEWIDTH 1020
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_dcache_page ar as
+
+#if XCHAL_DCACHE_SIZE
+	__loop_cache_page \ar \as dhi XCHAL_DCACHE_LINEWIDTH 1020
+#endif
+
+	.endm
+
+
+	.macro	___invalidate_icache_page ar as
+
+#if XCHAL_ICACHE_SIZE
+	__loop_cache_page \ar \as ihi XCHAL_ICACHE_LINEWIDTH 1020
+#endif
+
+	.endm
diff --git a/arch/xtensa/include/asm/cacheflush.h b/arch/xtensa/include/asm/cacheflush.h
new file mode 100644
index 000000000..a2b6bb542
--- /dev/null
+++ b/arch/xtensa/include/asm/cacheflush.h
@@ -0,0 +1,181 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CACHEFLUSH_H
+#define _XTENSA_CACHEFLUSH_H
+
+#include <linux/mm.h>
+#include <asm/processor.h>
+#include <asm/page.h>
+
+/*
+ * Lo-level routines for cache flushing.
+ *
+ * invalidate data or instruction cache:
+ *
+ * __invalidate_icache_all()
+ * __invalidate_icache_page(adr)
+ * __invalidate_dcache_page(adr)
+ * __invalidate_icache_range(from,size)
+ * __invalidate_dcache_range(from,size)
+ *
+ * flush data cache:
+ *
+ * __flush_dcache_page(adr)
+ *
+ * flush and invalidate data cache:
+ *
+ * __flush_invalidate_dcache_all()
+ * __flush_invalidate_dcache_page(adr)
+ * __flush_invalidate_dcache_range(from,size)
+ *
+ * specials for cache aliasing:
+ *
+ * __flush_invalidate_dcache_page_alias(vaddr,paddr)
+ * __invalidate_dcache_page_alias(vaddr,paddr)
+ * __invalidate_icache_page_alias(vaddr,paddr)
+ */
+
+extern void __invalidate_dcache_all(void);
+extern void __invalidate_icache_all(void);
+extern void __invalidate_dcache_page(unsigned long);
+extern void __invalidate_icache_page(unsigned long);
+extern void __invalidate_icache_range(unsigned long, unsigned long);
+extern void __invalidate_dcache_range(unsigned long, unsigned long);
+
+#if XCHAL_DCACHE_IS_WRITEBACK
+extern void __flush_invalidate_dcache_all(void);
+extern void __flush_dcache_page(unsigned long);
+extern void __flush_dcache_range(unsigned long, unsigned long);
+extern void __flush_invalidate_dcache_page(unsigned long);
+extern void __flush_invalidate_dcache_range(unsigned long, unsigned long);
+#else
+static inline void __flush_dcache_page(unsigned long va)
+{
+}
+static inline void __flush_dcache_range(unsigned long va, unsigned long sz)
+{
+}
+# define __flush_invalidate_dcache_all()	__invalidate_dcache_all()
+# define __flush_invalidate_dcache_page(p)	__invalidate_dcache_page(p)
+# define __flush_invalidate_dcache_range(p,s)	__invalidate_dcache_range(p,s)
+#endif
+
+#if defined(CONFIG_MMU) && (DCACHE_WAY_SIZE > PAGE_SIZE)
+extern void __flush_invalidate_dcache_page_alias(unsigned long, unsigned long);
+extern void __invalidate_dcache_page_alias(unsigned long, unsigned long);
+#else
+static inline void __flush_invalidate_dcache_page_alias(unsigned long virt,
+							unsigned long phys) { }
+static inline void __invalidate_dcache_page_alias(unsigned long virt,
+						  unsigned long phys) { }
+#endif
+#if defined(CONFIG_MMU) && (ICACHE_WAY_SIZE > PAGE_SIZE)
+extern void __invalidate_icache_page_alias(unsigned long, unsigned long);
+#else
+static inline void __invalidate_icache_page_alias(unsigned long virt,
+						unsigned long phys) { }
+#endif
+
+/*
+ * We have physically tagged caches - nothing to do here -
+ * unless we have cache aliasing.
+ *
+ * Pages can get remapped. Because this might change the 'color' of that page,
+ * we have to flush the cache before the PTE is changed.
+ * (see also Documentation/core-api/cachetlb.rst)
+ */
+
+#if defined(CONFIG_MMU) && \
+	((DCACHE_WAY_SIZE > PAGE_SIZE) || defined(CONFIG_SMP))
+
+#ifdef CONFIG_SMP
+void flush_cache_all(void);
+void flush_cache_range(struct vm_area_struct*, ulong, ulong);
+void flush_icache_range(unsigned long start, unsigned long end);
+void flush_cache_page(struct vm_area_struct*,
+			     unsigned long, unsigned long);
+#define flush_cache_all flush_cache_all
+#define flush_cache_range flush_cache_range
+#define flush_icache_range flush_icache_range
+#define flush_cache_page flush_cache_page
+#else
+#define flush_cache_all local_flush_cache_all
+#define flush_cache_range local_flush_cache_range
+#define flush_icache_range local_flush_icache_range
+#define flush_cache_page  local_flush_cache_page
+#endif
+
+#define local_flush_cache_all()						\
+	do {								\
+		__flush_invalidate_dcache_all();			\
+		__invalidate_icache_all();				\
+	} while (0)
+
+#define flush_cache_mm(mm)		flush_cache_all()
+#define flush_cache_dup_mm(mm)		flush_cache_mm(mm)
+
+#define flush_cache_vmap(start,end)		flush_cache_all()
+#define flush_cache_vmap_early(start,end)	do { } while (0)
+#define flush_cache_vunmap(start,end)		flush_cache_all()
+
+void flush_dcache_folio(struct folio *folio);
+#define flush_dcache_folio flush_dcache_folio
+
+#define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE 1
+static inline void flush_dcache_page(struct page *page)
+{
+	flush_dcache_folio(page_folio(page));
+}
+
+void local_flush_cache_range(struct vm_area_struct *vma,
+		unsigned long start, unsigned long end);
+void local_flush_cache_page(struct vm_area_struct *vma,
+		unsigned long address, unsigned long pfn);
+
+#else
+
+#define flush_icache_range local_flush_icache_range
+
+#endif
+
+#define flush_icache_user_range flush_icache_range
+
+/* Ensure consistency between data and instruction cache. */
+#define local_flush_icache_range(start, end)				\
+	do {								\
+		__flush_dcache_range(start, (end) - (start));		\
+		__invalidate_icache_range(start,(end) - (start));	\
+	} while (0)
+
+#if defined(CONFIG_MMU) && (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+extern void copy_to_user_page(struct vm_area_struct*, struct page*,
+		unsigned long, void*, const void*, unsigned long);
+extern void copy_from_user_page(struct vm_area_struct*, struct page*,
+		unsigned long, void*, const void*, unsigned long);
+#define copy_to_user_page copy_to_user_page
+#define copy_from_user_page copy_from_user_page
+
+#else
+
+#define copy_to_user_page(vma, page, vaddr, dst, src, len)		\
+	do {								\
+		memcpy(dst, src, len);					\
+		__flush_dcache_range((unsigned long) dst, len);		\
+		__invalidate_icache_range((unsigned long) dst, len);	\
+	} while (0)
+
+#define copy_from_user_page(vma, page, vaddr, dst, src, len) \
+	memcpy(dst, src, len)
+
+#endif
+
+#include <asm-generic/cacheflush.h>
+
+#endif /* _XTENSA_CACHEFLUSH_H */
diff --git a/arch/xtensa/include/asm/cachetype.h b/arch/xtensa/include/asm/cachetype.h
new file mode 100644
index 000000000..51bd49e2a
--- /dev/null
+++ b/arch/xtensa/include/asm/cachetype.h
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_XTENSA_CACHETYPE_H
+#define __ASM_XTENSA_CACHETYPE_H
+
+#include <asm/cache.h>
+#include <asm/page.h>
+
+#define cpu_dcache_is_aliasing()	(DCACHE_WAY_SIZE > PAGE_SIZE)
+
+#endif
diff --git a/arch/xtensa/include/asm/checksum.h b/arch/xtensa/include/asm/checksum.h
new file mode 100644
index 000000000..44ec1d0b2
--- /dev/null
+++ b/arch/xtensa/include/asm/checksum.h
@@ -0,0 +1,247 @@
+/*
+ * include/asm-xtensa/checksum.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CHECKSUM_H
+#define _XTENSA_CHECKSUM_H
+
+#include <linux/in6.h>
+#include <linux/uaccess.h>
+#include <asm/core.h>
+
+/*
+ * computes the checksum of a memory block at buff, length len,
+ * and adds in "sum" (32-bit)
+ *
+ * returns a 32-bit number suitable for feeding into itself
+ * or csum_tcpudp_magic
+ *
+ * this function must be called with even lengths, except
+ * for the last fragment, which may be odd
+ *
+ * it's best to have buff aligned on a 32-bit boundary
+ */
+asmlinkage __wsum csum_partial(const void *buff, int len, __wsum sum);
+
+/*
+ * the same as csum_partial, but copies from src while it
+ * checksums, and handles user-space pointer exceptions correctly, when needed.
+ *
+ * here even more important to align src and dst on a 32-bit (or even
+ * better 64-bit) boundary
+ */
+
+asmlinkage __wsum csum_partial_copy_generic(const void *src, void *dst, int len);
+
+#define _HAVE_ARCH_CSUM_AND_COPY
+/*
+ *	Note: when you get a NULL pointer exception here this means someone
+ *	passed in an incorrect kernel address to one of these functions.
+ */
+static inline
+__wsum csum_partial_copy_nocheck(const void *src, void *dst, int len)
+{
+	return csum_partial_copy_generic(src, dst, len);
+}
+
+#define _HAVE_ARCH_COPY_AND_CSUM_FROM_USER
+static inline
+__wsum csum_and_copy_from_user(const void __user *src, void *dst,
+				   int len)
+{
+	if (!access_ok(src, len))
+		return 0;
+	return csum_partial_copy_generic((__force const void *)src, dst, len);
+}
+
+/*
+ *	Fold a partial checksum
+ */
+
+static __inline__ __sum16 csum_fold(__wsum sum)
+{
+	unsigned int __dummy;
+	__asm__("extui	%1, %0, 16, 16\n\t"
+		"extui	%0 ,%0, 0, 16\n\t"
+		"add	%0, %0, %1\n\t"
+		"slli	%1, %0, 16\n\t"
+		"add	%0, %0, %1\n\t"
+		"extui	%0, %0, 16, 16\n\t"
+		"neg	%0, %0\n\t"
+		"addi	%0, %0, -1\n\t"
+		"extui	%0, %0, 0, 16\n\t"
+		: "=r" (sum), "=&r" (__dummy)
+		: "0" (sum));
+	return (__force __sum16)sum;
+}
+
+/*
+ *	This is a version of ip_compute_csum() optimized for IP headers,
+ *	which always checksum on 4 octet boundaries.
+ */
+static __inline__ __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
+{
+	unsigned int sum, tmp, endaddr;
+
+	__asm__ __volatile__(
+		"sub		%0, %0, %0\n\t"
+#if XCHAL_HAVE_LOOPS
+		"loopgtz	%2, 2f\n\t"
+#else
+		"beqz		%2, 2f\n\t"
+		"slli		%4, %2, 2\n\t"
+		"add		%4, %4, %1\n\t"
+		"0:\t"
+#endif
+		"l32i		%3, %1, 0\n\t"
+		"add		%0, %0, %3\n\t"
+		"bgeu		%0, %3, 1f\n\t"
+		"addi		%0, %0, 1\n\t"
+		"1:\t"
+		"addi		%1, %1, 4\n\t"
+#if !XCHAL_HAVE_LOOPS
+		"blt		%1, %4, 0b\n\t"
+#endif
+		"2:\t"
+	/* Since the input registers which are loaded with iph and ihl
+	   are modified, we must also specify them as outputs, or gcc
+	   will assume they contain their original values. */
+		: "=r" (sum), "=r" (iph), "=r" (ihl), "=&r" (tmp),
+		  "=&r" (endaddr)
+		: "1" (iph), "2" (ihl)
+		: "memory");
+
+	return	csum_fold(sum);
+}
+
+static __inline__ __wsum csum_tcpudp_nofold(__be32 saddr, __be32 daddr,
+					    __u32 len, __u8 proto,
+					    __wsum sum)
+{
+
+#ifdef __XTENSA_EL__
+	unsigned long len_proto = (len + proto) << 8;
+#elif defined(__XTENSA_EB__)
+	unsigned long len_proto = len + proto;
+#else
+# error processor byte order undefined!
+#endif
+	__asm__("add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"add	%0, %0, %2\n\t"
+		"bgeu	%0, %2, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"add	%0, %0, %3\n\t"
+		"bgeu	%0, %3, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		: "=r" (sum), "=r" (len_proto)
+		: "r" (daddr), "r" (saddr), "1" (len_proto), "0" (sum));
+	return sum;
+}
+
+/*
+ * computes the checksum of the TCP/UDP pseudo-header
+ * returns a 16-bit checksum, already complemented
+ */
+static __inline__ __sum16 csum_tcpudp_magic(__be32 saddr, __be32 daddr,
+					    __u32 len, __u8 proto,
+					    __wsum sum)
+{
+	return csum_fold(csum_tcpudp_nofold(saddr,daddr,len,proto,sum));
+}
+
+/*
+ * this routine is used for miscellaneous IP-like checksums, mainly
+ * in icmp.c
+ */
+
+static __inline__ __sum16 ip_compute_csum(const void *buff, int len)
+{
+	return csum_fold (csum_partial(buff, len, 0));
+}
+
+#define _HAVE_ARCH_IPV6_CSUM
+static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr,
+					  const struct in6_addr *daddr,
+					  __u32 len, __u8 proto,
+					  __wsum sum)
+{
+	unsigned int __dummy;
+	__asm__("l32i	%1, %2, 0\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %2, 4\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %2, 8\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %2, 12\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %3, 0\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %3, 4\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %3, 8\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"l32i	%1, %3, 12\n\t"
+		"add	%0, %0, %1\n\t"
+		"bgeu	%0, %1, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"add	%0, %0, %4\n\t"
+		"bgeu	%0, %4, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		"add	%0, %0, %5\n\t"
+		"bgeu	%0, %5, 1f\n\t"
+		"addi	%0, %0, 1\n\t"
+		"1:\t"
+		: "=r" (sum), "=&r" (__dummy)
+		: "r" (saddr), "r" (daddr),
+		  "r" (htonl(len)), "r" (htonl(proto)), "0" (sum)
+		: "memory");
+
+	return csum_fold(sum);
+}
+
+/*
+ *	Copy and checksum to user
+ */
+#define HAVE_CSUM_COPY_USER
+static __inline__ __wsum csum_and_copy_to_user(const void *src,
+					       void __user *dst, int len)
+{
+	if (!access_ok(dst, len))
+		return 0;
+	return csum_partial_copy_generic(src, (__force void *)dst, len);
+}
+#endif
diff --git a/arch/xtensa/include/asm/cmpxchg.h b/arch/xtensa/include/asm/cmpxchg.h
new file mode 100644
index 000000000..b6db4838b
--- /dev/null
+++ b/arch/xtensa/include/asm/cmpxchg.h
@@ -0,0 +1,225 @@
+/*
+ * Atomic xchg and cmpxchg operations.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CMPXCHG_H
+#define _XTENSA_CMPXCHG_H
+
+#ifndef __ASSEMBLER__
+
+#include <linux/bits.h>
+#include <linux/stringify.h>
+#include <linux/cmpxchg-emu.h>
+
+/*
+ * cmpxchg
+ */
+
+static inline unsigned long
+__cmpxchg_u32(volatile int *p, int old, int new)
+{
+#if XCHAL_HAVE_EXCLUSIVE
+	unsigned long tmp, result;
+
+	__asm__ __volatile__(
+			"1:     l32ex   %[result], %[addr]\n"
+			"       bne     %[result], %[cmp], 2f\n"
+			"       mov     %[tmp], %[new]\n"
+			"       s32ex   %[tmp], %[addr]\n"
+			"       getex   %[tmp]\n"
+			"       beqz    %[tmp], 1b\n"
+			"2:\n"
+			: [result] "=&a" (result), [tmp] "=&a" (tmp)
+			: [new] "a" (new), [addr] "a" (p), [cmp] "a" (old)
+			: "memory"
+			);
+
+	return result;
+#elif XCHAL_HAVE_S32C1I
+	__asm__ __volatile__(
+			"       wsr     %[cmp], scompare1\n"
+			"       s32c1i  %[new], %[mem]\n"
+			: [new] "+a" (new), [mem] "+m" (*p)
+			: [cmp] "a" (old)
+			: "memory"
+			);
+
+	return new;
+#else
+	__asm__ __volatile__(
+			"       rsil    a14, "__stringify(TOPLEVEL)"\n"
+			"       l32i    %[old], %[mem]\n"
+			"       bne     %[old], %[cmp], 1f\n"
+			"       s32i    %[new], %[mem]\n"
+			"1:\n"
+			"       wsr     a14, ps\n"
+			"       rsync\n"
+			: [old] "=&a" (old), [mem] "+m" (*p)
+			: [cmp] "a" (old), [new] "r" (new)
+			: "a14", "memory");
+	return old;
+#endif
+}
+/* This function doesn't exist, so you'll get a linker error
+ * if something tries to do an invalid cmpxchg(). */
+
+extern void __cmpxchg_called_with_bad_pointer(void);
+
+static __inline__ unsigned long
+__cmpxchg(volatile void *ptr, unsigned long old, unsigned long new, int size)
+{
+	switch (size) {
+	case 1:  return cmpxchg_emu_u8(ptr, old, new);
+	case 4:  return __cmpxchg_u32(ptr, old, new);
+	default: __cmpxchg_called_with_bad_pointer();
+		 return old;
+	}
+}
+
+#define arch_cmpxchg(ptr,o,n)						      \
+	({ __typeof__(*(ptr)) _o_ = (o);				      \
+	   __typeof__(*(ptr)) _n_ = (n);				      \
+	   (__typeof__(*(ptr))) __cmpxchg((ptr), (unsigned long)_o_,	      \
+	   			        (unsigned long)_n_, sizeof (*(ptr))); \
+	})
+
+#include <asm-generic/cmpxchg-local.h>
+
+static inline unsigned long __cmpxchg_local(volatile void *ptr,
+				      unsigned long old,
+				      unsigned long new, int size)
+{
+	switch (size) {
+	case 4:
+		return __cmpxchg_u32(ptr, old, new);
+	default:
+		return __generic_cmpxchg_local(ptr, old, new, size);
+	}
+
+	return old;
+}
+
+/*
+ * cmpxchg_local and cmpxchg64_local are atomic wrt current CPU. Always make
+ * them available.
+ */
+#define arch_cmpxchg_local(ptr, o, n)				  	       \
+	((__typeof__(*(ptr)))__generic_cmpxchg_local((ptr), (unsigned long)(o),\
+			(unsigned long)(n), sizeof(*(ptr))))
+#define arch_cmpxchg64_local(ptr, o, n) __generic_cmpxchg64_local((ptr), (o), (n))
+#define arch_cmpxchg64(ptr, o, n)    arch_cmpxchg64_local((ptr), (o), (n))
+
+/*
+ * xchg_u32
+ *
+ * Note that a14 is used here because the register allocation
+ * done by the compiler is not guaranteed and a window overflow
+ * may not occur between the rsil and wsr instructions. By using
+ * a14 in the rsil, the machine is guaranteed to be in a state
+ * where no register reference will cause an overflow.
+ */
+
+static inline unsigned long xchg_u32(volatile int * m, unsigned long val)
+{
+#if XCHAL_HAVE_EXCLUSIVE
+	unsigned long tmp, result;
+
+	__asm__ __volatile__(
+			"1:     l32ex   %[result], %[addr]\n"
+			"       mov     %[tmp], %[val]\n"
+			"       s32ex   %[tmp], %[addr]\n"
+			"       getex   %[tmp]\n"
+			"       beqz    %[tmp], 1b\n"
+			: [result] "=&a" (result), [tmp] "=&a" (tmp)
+			: [val] "a" (val), [addr] "a" (m)
+			: "memory"
+			);
+
+	return result;
+#elif XCHAL_HAVE_S32C1I
+	unsigned long tmp, result;
+	__asm__ __volatile__(
+			"1:     l32i    %[tmp], %[mem]\n"
+			"       mov     %[result], %[val]\n"
+			"       wsr     %[tmp], scompare1\n"
+			"       s32c1i  %[result], %[mem]\n"
+			"       bne     %[result], %[tmp], 1b\n"
+			: [result] "=&a" (result), [tmp] "=&a" (tmp),
+			  [mem] "+m" (*m)
+			: [val] "a" (val)
+			: "memory"
+			);
+	return result;
+#else
+	unsigned long tmp;
+	__asm__ __volatile__(
+			"       rsil    a14, "__stringify(TOPLEVEL)"\n"
+			"       l32i    %[tmp], %[mem]\n"
+			"       s32i    %[val], %[mem]\n"
+			"       wsr     a14, ps\n"
+			"       rsync\n"
+			: [tmp] "=&a" (tmp), [mem] "+m" (*m)
+			: [val] "a" (val)
+			: "a14", "memory");
+	return tmp;
+#endif
+}
+
+#define arch_xchg(ptr,x) \
+	((__typeof__(*(ptr)))__arch_xchg((unsigned long)(x),(ptr),sizeof(*(ptr))))
+
+static inline u32 xchg_small(volatile void *ptr, u32 x, int size)
+{
+	int off = (unsigned long)ptr % sizeof(u32);
+	volatile u32 *p = ptr - off;
+#ifdef __BIG_ENDIAN
+	int bitoff = (sizeof(u32) - size - off) * BITS_PER_BYTE;
+#else
+	int bitoff = off * BITS_PER_BYTE;
+#endif
+	u32 bitmask = ((0x1 << size * BITS_PER_BYTE) - 1) << bitoff;
+	u32 oldv, newv;
+	u32 ret;
+
+	do {
+		oldv = READ_ONCE(*p);
+		ret = (oldv & bitmask) >> bitoff;
+		newv = (oldv & ~bitmask) | (x << bitoff);
+	} while (__cmpxchg_u32(p, oldv, newv) != oldv);
+
+	return ret;
+}
+
+/*
+ * This only works if the compiler isn't horribly bad at optimizing.
+ * gcc-2.5.8 reportedly can't handle this, but I define that one to
+ * be dead anyway.
+ */
+
+extern void __xchg_called_with_bad_pointer(void);
+
+static __inline__ unsigned long
+__arch_xchg(unsigned long x, volatile void * ptr, int size)
+{
+	switch (size) {
+	case 1:
+		return xchg_small(ptr, x, 1);
+	case 2:
+		return xchg_small(ptr, x, 2);
+	case 4:
+		return xchg_u32(ptr, x);
+	default:
+		__xchg_called_with_bad_pointer();
+		return x;
+	}
+}
+
+#endif /* __ASSEMBLER__ */
+
+#endif /* _XTENSA_CMPXCHG_H */
diff --git a/arch/xtensa/include/asm/coprocessor.h b/arch/xtensa/include/asm/coprocessor.h
new file mode 100644
index 000000000..e0447bcc5
--- /dev/null
+++ b/arch/xtensa/include/asm/coprocessor.h
@@ -0,0 +1,155 @@
+/*
+ * include/asm-xtensa/coprocessor.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2003 - 2007 Tensilica Inc.
+ */
+
+
+#ifndef _XTENSA_COPROCESSOR_H
+#define _XTENSA_COPROCESSOR_H
+
+#include <variant/tie.h>
+#include <asm/core.h>
+#include <asm/types.h>
+
+#ifdef __ASSEMBLER__
+# include <variant/tie-asm.h>
+
+.macro	xchal_sa_start  a b
+	.set .Lxchal_pofs_, 0
+	.set .Lxchal_ofs_, 0
+.endm
+
+.macro	xchal_sa_align  ptr minofs maxofs ofsalign totalign
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + .Lxchal_pofs_ + \totalign - 1
+	.set	.Lxchal_ofs_, (.Lxchal_ofs_ & -\totalign) - .Lxchal_pofs_
+.endm
+
+#define _SELECT	(  XTHAL_SAS_TIE | XTHAL_SAS_OPT \
+		 | XTHAL_SAS_CC \
+		 | XTHAL_SAS_CALR | XTHAL_SAS_CALE )
+
+.macro save_xtregs_opt ptr clb at1 at2 at3 at4 offset
+	.if XTREGS_OPT_SIZE > 0
+		addi	\clb, \ptr, \offset
+		xchal_ncp_store \clb \at1 \at2 \at3 \at4 select=_SELECT
+	.endif
+.endm
+
+.macro load_xtregs_opt ptr clb at1 at2 at3 at4 offset
+	.if XTREGS_OPT_SIZE > 0
+		addi	\clb, \ptr, \offset
+		xchal_ncp_load \clb \at1 \at2 \at3 \at4 select=_SELECT
+	.endif
+.endm
+#undef _SELECT
+
+#define _SELECT	(  XTHAL_SAS_TIE | XTHAL_SAS_OPT \
+		 | XTHAL_SAS_NOCC \
+		 | XTHAL_SAS_CALR | XTHAL_SAS_CALE | XTHAL_SAS_GLOB )
+
+.macro save_xtregs_user ptr clb at1 at2 at3 at4 offset
+	.if XTREGS_USER_SIZE > 0
+		addi	\clb, \ptr, \offset
+		xchal_ncp_store \clb \at1 \at2 \at3 \at4 select=_SELECT
+	.endif
+.endm
+
+.macro load_xtregs_user ptr clb at1 at2 at3 at4 offset
+	.if XTREGS_USER_SIZE > 0
+		addi	\clb, \ptr, \offset
+		xchal_ncp_load \clb \at1 \at2 \at3 \at4 select=_SELECT
+	.endif
+.endm
+#undef _SELECT
+
+
+
+#endif	/* __ASSEMBLER__ */
+
+/*
+ * XTENSA_HAVE_COPROCESSOR(x) returns 1 if coprocessor x is configured.
+ *
+ * XTENSA_HAVE_IO_PORT(x) returns 1 if io-port x is configured.
+ *
+ */
+
+#define XTENSA_HAVE_COPROCESSOR(x)					\
+	((XCHAL_CP_MASK ^ XCHAL_CP_PORT_MASK) & (1 << (x)))
+#define XTENSA_HAVE_COPROCESSORS					\
+	(XCHAL_CP_MASK ^ XCHAL_CP_PORT_MASK)
+#define XTENSA_HAVE_IO_PORT(x)						\
+	(XCHAL_CP_PORT_MASK & (1 << (x)))
+#define XTENSA_HAVE_IO_PORTS						\
+	XCHAL_CP_PORT_MASK
+
+#ifndef __ASSEMBLER__
+
+/*
+ * Additional registers.
+ * We define three types of additional registers:
+ *  ext: extra registers that are used by the compiler
+ *  cpn: optional registers that can be used by a user application
+ *  cpX: coprocessor registers that can only be used if the corresponding
+ *       CPENABLE bit is set.
+ */
+
+#define XCHAL_SA_REG(list,cc,abi,type,y,name,z,align,size,...)	\
+	__REG ## list (cc, abi, type, name, size, align)
+
+#define __REG0(cc,abi,t,name,s,a)	__REG0_ ## cc (abi,name)
+#define __REG1(cc,abi,t,name,s,a)	__REG1_ ## cc (name)
+#define __REG2(cc,abi,type,...)		__REG2_ ## type (__VA_ARGS__)
+
+#define __REG0_0(abi,name)
+#define __REG0_1(abi,name)		__REG0_1 ## abi (name)
+#define __REG0_10(name)	__u32 name;
+#define __REG0_11(name)	__u32 name;
+#define __REG0_12(name)
+
+#define __REG1_0(name)	__u32 name;
+#define __REG1_1(name)
+
+#define __REG2_0(n,s,a)	__u32 name;
+#define __REG2_1(n,s,a)	unsigned char n[s] __attribute__ ((aligned(a)));
+#define __REG2_2(n,s,a) unsigned char n[s] __attribute__ ((aligned(a)));
+
+typedef struct { XCHAL_NCP_SA_LIST(0) } xtregs_opt_t
+	__attribute__ ((aligned (XCHAL_NCP_SA_ALIGN)));
+typedef struct { XCHAL_NCP_SA_LIST(1) } xtregs_user_t
+	__attribute__ ((aligned (XCHAL_NCP_SA_ALIGN)));
+
+#if XTENSA_HAVE_COPROCESSORS
+
+typedef struct { XCHAL_CP0_SA_LIST(2) } xtregs_cp0_t
+	__attribute__ ((aligned (XCHAL_CP0_SA_ALIGN)));
+typedef struct { XCHAL_CP1_SA_LIST(2) } xtregs_cp1_t
+	__attribute__ ((aligned (XCHAL_CP1_SA_ALIGN)));
+typedef struct { XCHAL_CP2_SA_LIST(2) } xtregs_cp2_t
+	__attribute__ ((aligned (XCHAL_CP2_SA_ALIGN)));
+typedef struct { XCHAL_CP3_SA_LIST(2) } xtregs_cp3_t
+	__attribute__ ((aligned (XCHAL_CP3_SA_ALIGN)));
+typedef struct { XCHAL_CP4_SA_LIST(2) } xtregs_cp4_t
+	__attribute__ ((aligned (XCHAL_CP4_SA_ALIGN)));
+typedef struct { XCHAL_CP5_SA_LIST(2) } xtregs_cp5_t
+	__attribute__ ((aligned (XCHAL_CP5_SA_ALIGN)));
+typedef struct { XCHAL_CP6_SA_LIST(2) } xtregs_cp6_t
+	__attribute__ ((aligned (XCHAL_CP6_SA_ALIGN)));
+typedef struct { XCHAL_CP7_SA_LIST(2) } xtregs_cp7_t
+	__attribute__ ((aligned (XCHAL_CP7_SA_ALIGN)));
+
+struct thread_info;
+void coprocessor_flush(struct thread_info *ti, int cp_index);
+void coprocessor_release_all(struct thread_info *ti);
+void coprocessor_flush_all(struct thread_info *ti);
+void coprocessor_flush_release_all(struct thread_info *ti);
+void local_coprocessors_flush_release_all(void);
+
+#endif	/* XTENSA_HAVE_COPROCESSORS */
+
+#endif	/* !__ASSEMBLER__ */
+#endif	/* _XTENSA_COPROCESSOR_H */
diff --git a/arch/xtensa/include/asm/core.h b/arch/xtensa/include/asm/core.h
new file mode 100644
index 000000000..6f02f6f21
--- /dev/null
+++ b/arch/xtensa/include/asm/core.h
@@ -0,0 +1,68 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2019 Cadence Design Systems Inc. */
+
+#ifndef _ASM_XTENSA_CORE_H
+#define _ASM_XTENSA_CORE_H
+
+#include <variant/core.h>
+
+#ifndef XCHAL_HAVE_DIV32
+#define XCHAL_HAVE_DIV32 0
+#endif
+
+#ifndef XCHAL_HAVE_EXCLUSIVE
+#define XCHAL_HAVE_EXCLUSIVE 0
+#endif
+
+#ifndef XCHAL_HAVE_EXTERN_REGS
+#define XCHAL_HAVE_EXTERN_REGS 0
+#endif
+
+#ifndef XCHAL_HAVE_MPU
+#define XCHAL_HAVE_MPU 0
+#endif
+
+#ifndef XCHAL_HAVE_VECBASE
+#define XCHAL_HAVE_VECBASE 0
+#endif
+
+#ifndef XCHAL_SPANNING_WAY
+#define XCHAL_SPANNING_WAY 0
+#endif
+
+#ifndef XCHAL_HAVE_TRAX
+#define XCHAL_HAVE_TRAX 0
+#endif
+
+#ifndef XCHAL_NUM_PERF_COUNTERS
+#define XCHAL_NUM_PERF_COUNTERS 0
+#endif
+
+#if XCHAL_HAVE_WINDOWED
+#if defined(CONFIG_USER_ABI_DEFAULT) || defined(CONFIG_USER_ABI_CALL0_PROBE)
+/* Whether windowed ABI is supported in userspace. */
+#define USER_SUPPORT_WINDOWED
+#endif
+#if defined(__XTENSA_WINDOWED_ABI__) || defined(USER_SUPPORT_WINDOWED)
+/* Whether windowed ABI is supported either in userspace or in the kernel. */
+#define SUPPORT_WINDOWED
+#endif
+#endif
+
+/* Xtensa ABI requires stack alignment to be at least 16 */
+#if XCHAL_DATA_WIDTH > 16
+#define XTENSA_STACK_ALIGNMENT	XCHAL_DATA_WIDTH
+#else
+#define XTENSA_STACK_ALIGNMENT	16
+#endif
+
+#ifndef XCHAL_HW_MIN_VERSION
+#if defined(XCHAL_HW_MIN_VERSION_MAJOR) && defined(XCHAL_HW_MIN_VERSION_MINOR)
+#define XCHAL_HW_MIN_VERSION (XCHAL_HW_MIN_VERSION_MAJOR * 100 + \
+			      XCHAL_HW_MIN_VERSION_MINOR)
+#else
+#define XCHAL_HW_MIN_VERSION 0
+#endif
+#endif
+
+#endif
diff --git a/arch/xtensa/include/asm/current.h b/arch/xtensa/include/asm/current.h
new file mode 100644
index 000000000..7b483538f
--- /dev/null
+++ b/arch/xtensa/include/asm/current.h
@@ -0,0 +1,40 @@
+/*
+ * include/asm-xtensa/current.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CURRENT_H
+#define _XTENSA_CURRENT_H
+
+#include <asm/thread_info.h>
+
+#ifndef __ASSEMBLER__
+
+#include <linux/thread_info.h>
+
+struct task_struct;
+
+static __always_inline struct task_struct *get_current(void)
+{
+	return current_thread_info()->task;
+}
+
+#define current get_current()
+
+register unsigned long current_stack_pointer __asm__("a1");
+
+#else
+
+#define GET_CURRENT(reg,sp)		\
+	GET_THREAD_INFO(reg,sp);	\
+	l32i reg, reg, TI_TASK		\
+
+#endif
+
+
+#endif /* XTENSA_CURRENT_H */
diff --git a/arch/xtensa/include/asm/delay.h b/arch/xtensa/include/asm/delay.h
new file mode 100644
index 000000000..24304b39a
--- /dev/null
+++ b/arch/xtensa/include/asm/delay.h
@@ -0,0 +1,75 @@
+/*
+ * include/asm-xtensa/delay.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ *
+ */
+
+#ifndef _XTENSA_DELAY_H
+#define _XTENSA_DELAY_H
+
+#include <asm/timex.h>
+#include <asm/param.h>
+
+extern unsigned long loops_per_jiffy;
+
+static inline void __delay(unsigned long loops)
+{
+	if (__builtin_constant_p(loops) && loops < 2)
+		__asm__ __volatile__ ("nop");
+	else if (loops >= 2)
+		/* 2 cycles per loop. */
+		__asm__ __volatile__ ("1: addi %0, %0, -2; bgeui %0, 2, 1b"
+				: "+r" (loops));
+}
+
+/* Undefined function to get compile-time error */
+void __bad_udelay(void);
+void __bad_ndelay(void);
+
+#define __MAX_UDELAY 30000
+#define __MAX_NDELAY 30000
+
+static inline void __udelay(unsigned long usecs)
+{
+	unsigned long start = get_ccount();
+	unsigned long cycles = (usecs * (ccount_freq >> 15)) >> 5;
+
+	/* Note: all variables are unsigned (can wrap around)! */
+	while (((unsigned long)get_ccount()) - start < cycles)
+		cpu_relax();
+}
+
+static inline void udelay(unsigned long usec)
+{
+	if (__builtin_constant_p(usec) && usec >= __MAX_UDELAY)
+		__bad_udelay();
+	else
+		__udelay(usec);
+}
+
+static inline void __ndelay(unsigned long nsec)
+{
+	/*
+	 * Inner shift makes sure multiplication doesn't overflow
+	 * for legitimate nsec values
+	 */
+	unsigned long cycles = (nsec * (ccount_freq >> 15)) >> 15;
+	__delay(cycles);
+}
+
+#define ndelay(n) ndelay(n)
+
+static inline void ndelay(unsigned long nsec)
+{
+	if (__builtin_constant_p(nsec) && nsec >= __MAX_NDELAY)
+		__bad_ndelay();
+	else
+		__ndelay(nsec);
+}
+
+#endif
diff --git a/arch/xtensa/include/asm/dma.h b/arch/xtensa/include/asm/dma.h
new file mode 100644
index 000000000..172644539
--- /dev/null
+++ b/arch/xtensa/include/asm/dma.h
@@ -0,0 +1,55 @@
+/*
+ * include/asm-xtensa/dma.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2003 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_DMA_H
+#define _XTENSA_DMA_H
+
+#include <asm/io.h>		/* need byte IO */
+
+/*
+ * This is only to be defined if we have PC-like DMA.
+ * By default this is not true on an Xtensa processor,
+ * however on boards with a PCI bus, such functionality
+ * might be emulated externally.
+ *
+ * NOTE:  there still exists driver code that assumes
+ * this is defined, eg. drivers/sound/soundcard.c (as of 2.4).
+ */
+#define MAX_DMA_CHANNELS	8
+
+/*
+ * The maximum virtual address to which DMA transfers
+ * can be performed on this platform.
+ *
+ * NOTE: This is board (platform) specific, not processor-specific!
+ *
+ * NOTE: This assumes DMA transfers can only be performed on
+ *	the section of physical memory contiguously mapped in virtual
+ *	space for the kernel.  For the Xtensa architecture, this
+ *	means the maximum possible size of this DMA area is
+ *	the size of the statically mapped kernel segment
+ *	(XCHAL_KSEG_{CACHED,BYPASS}_SIZE), ie. 128 MB.
+ *
+ * NOTE: When the entire KSEG area is DMA capable, we subtract
+ *	one from the max address so that the virt_to_phys() macro
+ *	works correctly on the address (otherwise the address
+ *	enters another area, and virt_to_phys() may not return
+ *	the value desired).
+ */
+
+#ifndef MAX_DMA_ADDRESS
+#define MAX_DMA_ADDRESS		(PAGE_OFFSET + XCHAL_KIO_SIZE - 1)
+#endif
+
+/* Reserve and release a DMA channel */
+extern int request_dma(unsigned int dmanr, const char * device_id);
+extern void free_dma(unsigned int dmanr);
+
+#endif
diff --git a/arch/xtensa/include/asm/elf.h b/arch/xtensa/include/asm/elf.h
new file mode 100644
index 000000000..ffcf1ada1
--- /dev/null
+++ b/arch/xtensa/include/asm/elf.h
@@ -0,0 +1,196 @@
+/*
+ * include/asm-xtensa/elf.h
+ *
+ * ELF register definitions
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_ELF_H
+#define _XTENSA_ELF_H
+
+#include <asm/ptrace.h>
+#include <asm/coprocessor.h>
+#include <linux/elf-em.h>
+
+/* Xtensa processor ELF architecture-magic number */
+
+#define EM_XTENSA_OLD	0xABC7
+
+/* Xtensa relocations defined by the ABIs */
+
+#define R_XTENSA_NONE           0
+#define R_XTENSA_32             1
+#define R_XTENSA_RTLD           2
+#define R_XTENSA_GLOB_DAT       3
+#define R_XTENSA_JMP_SLOT       4
+#define R_XTENSA_RELATIVE       5
+#define R_XTENSA_PLT            6
+#define R_XTENSA_OP0            8
+#define R_XTENSA_OP1            9
+#define R_XTENSA_OP2            10
+#define R_XTENSA_ASM_EXPAND	11
+#define R_XTENSA_ASM_SIMPLIFY	12
+#define R_XTENSA_GNU_VTINHERIT	15
+#define R_XTENSA_GNU_VTENTRY	16
+#define R_XTENSA_DIFF8		17
+#define R_XTENSA_DIFF16		18
+#define R_XTENSA_DIFF32		19
+#define R_XTENSA_SLOT0_OP	20
+#define R_XTENSA_SLOT1_OP	21
+#define R_XTENSA_SLOT2_OP	22
+#define R_XTENSA_SLOT3_OP	23
+#define R_XTENSA_SLOT4_OP	24
+#define R_XTENSA_SLOT5_OP	25
+#define R_XTENSA_SLOT6_OP	26
+#define R_XTENSA_SLOT7_OP	27
+#define R_XTENSA_SLOT8_OP	28
+#define R_XTENSA_SLOT9_OP	29
+#define R_XTENSA_SLOT10_OP	30
+#define R_XTENSA_SLOT11_OP	31
+#define R_XTENSA_SLOT12_OP	32
+#define R_XTENSA_SLOT13_OP	33
+#define R_XTENSA_SLOT14_OP	34
+#define R_XTENSA_SLOT0_ALT	35
+#define R_XTENSA_SLOT1_ALT	36
+#define R_XTENSA_SLOT2_ALT	37
+#define R_XTENSA_SLOT3_ALT	38
+#define R_XTENSA_SLOT4_ALT	39
+#define R_XTENSA_SLOT5_ALT	40
+#define R_XTENSA_SLOT6_ALT	41
+#define R_XTENSA_SLOT7_ALT	42
+#define R_XTENSA_SLOT8_ALT	43
+#define R_XTENSA_SLOT9_ALT	44
+#define R_XTENSA_SLOT10_ALT	45
+#define R_XTENSA_SLOT11_ALT	46
+#define R_XTENSA_SLOT12_ALT	47
+#define R_XTENSA_SLOT13_ALT	48
+#define R_XTENSA_SLOT14_ALT	49
+
+/* ELF register definitions. This is needed for core dump support.  */
+
+typedef unsigned long elf_greg_t;
+
+typedef struct user_pt_regs xtensa_gregset_t;
+
+#define ELF_NGREG	(sizeof(xtensa_gregset_t) / sizeof(elf_greg_t))
+
+typedef elf_greg_t elf_gregset_t[ELF_NGREG];
+
+#define ELF_NFPREG	18
+
+typedef unsigned int elf_fpreg_t;
+typedef elf_fpreg_t elf_fpregset_t[ELF_NFPREG];
+
+/*
+ * This is used to ensure we don't load something for the wrong architecture.
+ */
+
+#define elf_check_arch(x) ( ( (x)->e_machine == EM_XTENSA )  || \
+			    ( (x)->e_machine == EM_XTENSA_OLD ) )
+
+#define ELFOSABI_XTENSA_FDPIC 65
+#define elf_check_fdpic(x) ((x)->e_ident[EI_OSABI] == ELFOSABI_XTENSA_FDPIC)
+#define ELF_FDPIC_CORE_EFLAGS 0
+
+/*
+ * These are used to set parameters in the core dumps.
+ */
+
+#ifdef __XTENSA_EL__
+# define ELF_DATA	ELFDATA2LSB
+#elif defined(__XTENSA_EB__)
+# define ELF_DATA	ELFDATA2MSB
+#else
+# error processor byte order undefined!
+#endif
+
+#define ELF_CLASS	ELFCLASS32
+#define ELF_ARCH	EM_XTENSA
+
+#define ELF_EXEC_PAGESIZE	PAGE_SIZE
+#define CORE_DUMP_USE_REGSET
+
+/*
+ * This is the location that an ET_DYN program is loaded if exec'ed.  Typical
+ * use of this is to invoke "./ld.so someprog" to test out a new version of
+ * the loader.  We need to make sure that it is out of the way of the program
+ * that it will "exec", and that there is sufficient room for the brk.
+ */
+
+#define ELF_ET_DYN_BASE         (2 * TASK_SIZE / 3)
+
+/*
+ * This yields a mask that user programs can use to figure out what
+ * instruction set this CPU supports.  This could be done in user space,
+ * but it's not easy, and we've already done it here.
+ */
+
+#define ELF_HWCAP	(0)
+
+/*
+ * This yields a string that ld.so will use to load implementation
+ * specific libraries for optimization.  This is more specific in
+ * intent than poking at uname or /proc/cpuinfo.
+ * For the moment, we have only optimizations for the Intel generations,
+ * but that could change...
+ */
+
+#define ELF_PLATFORM  (NULL)
+
+/*
+ * The Xtensa processor ABI says that when the program starts, a2
+ * contains a pointer to a function which might be registered using
+ * `atexit'.  This provides a mean for the dynamic linker to call
+ * DT_FINI functions for shared libraries that have been loaded before
+ * the code runs.
+ *
+ * A value of 0 tells we have no such handler.
+ *
+ * We might as well make sure everything else is cleared too (except
+ * for the stack pointer in a1), just to make things more
+ * deterministic.  Also, clearing a0 terminates debugger backtraces.
+ */
+
+#define ELF_PLAT_INIT(_r, load_addr) \
+	do { \
+		(_r)->areg[0]  = 0; /*(_r)->areg[1] = 0;*/ \
+		(_r)->areg[2]  = 0; (_r)->areg[3]  = 0; \
+		(_r)->areg[4]  = 0; (_r)->areg[5]  = 0; \
+		(_r)->areg[6]  = 0; (_r)->areg[7]  = 0; \
+		(_r)->areg[8]  = 0; (_r)->areg[9]  = 0; \
+		(_r)->areg[10] = 0; (_r)->areg[11] = 0; \
+		(_r)->areg[12] = 0; (_r)->areg[13] = 0; \
+		(_r)->areg[14] = 0; (_r)->areg[15] = 0; \
+	} while (0)
+
+#define ELF_FDPIC_PLAT_INIT(_r, _exec_map_addr, _interp_map_addr, dynamic_addr) \
+	do { \
+		(_r)->areg[4] = _exec_map_addr; \
+		(_r)->areg[5] = _interp_map_addr; \
+		(_r)->areg[6] = dynamic_addr; \
+	} while (0)
+
+typedef struct {
+	xtregs_opt_t	opt;
+	xtregs_user_t	user;
+#if XTENSA_HAVE_COPROCESSORS
+	xtregs_cp0_t	cp0;
+	xtregs_cp1_t	cp1;
+	xtregs_cp2_t	cp2;
+	xtregs_cp3_t	cp3;
+	xtregs_cp4_t	cp4;
+	xtregs_cp5_t	cp5;
+	xtregs_cp6_t	cp6;
+	xtregs_cp7_t	cp7;
+#endif
+} elf_xtregs_t;
+
+#define SET_PERSONALITY(ex) \
+	set_personality(PER_LINUX_32BIT | (current->personality & (~PER_MASK)))
+
+#endif	/* _XTENSA_ELF_H */
diff --git a/arch/xtensa/include/asm/fixmap.h b/arch/xtensa/include/asm/fixmap.h
new file mode 100644
index 000000000..1c65dc1d3
--- /dev/null
+++ b/arch/xtensa/include/asm/fixmap.h
@@ -0,0 +1,38 @@
+/*
+ * fixmap.h: compile-time virtual memory allocation
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1998 Ingo Molnar
+ *
+ * Support of BIGMEM added by Gerhard Wichert, Siemens AG, July 1999
+ */
+
+#ifndef _ASM_FIXMAP_H
+#define _ASM_FIXMAP_H
+
+#ifdef CONFIG_HIGHMEM
+#include <linux/threads.h>
+#include <linux/pgtable.h>
+#include <asm/kmap_size.h>
+
+/* The map slots for temporary mappings via kmap_atomic/local(). */
+enum fixed_addresses {
+	FIX_KMAP_BEGIN,
+	FIX_KMAP_END = FIX_KMAP_BEGIN +
+		(KM_MAX_IDX * NR_CPUS * DCACHE_N_COLORS) - 1,
+	__end_of_fixed_addresses
+};
+
+#define FIXADDR_END     (XCHAL_KSEG_CACHED_VADDR - PAGE_SIZE)
+#define FIXADDR_SIZE	(__end_of_fixed_addresses << PAGE_SHIFT)
+/* Enforce that FIXADDR_START is PMD aligned to handle cache aliasing */
+#define FIXADDR_START	((FIXADDR_END - FIXADDR_SIZE) & PMD_MASK)
+#define FIXADDR_TOP	(FIXADDR_START + FIXADDR_SIZE - PAGE_SIZE)
+
+#include <asm-generic/fixmap.h>
+
+#endif /* CONFIG_HIGHMEM */
+#endif
diff --git a/arch/xtensa/include/asm/flat.h b/arch/xtensa/include/asm/flat.h
new file mode 100644
index 000000000..4854419dc
--- /dev/null
+++ b/arch/xtensa/include/asm/flat.h
@@ -0,0 +1,19 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_XTENSA_FLAT_H
+#define __ASM_XTENSA_FLAT_H
+
+#include <linux/unaligned.h>
+
+static inline int flat_get_addr_from_rp(u32 __user *rp, u32 relval, u32 flags,
+					u32 *addr)
+{
+	*addr = get_unaligned((__force u32 *)rp);
+	return 0;
+}
+static inline int flat_put_addr_at_rp(u32 __user *rp, u32 addr, u32 rel)
+{
+	put_unaligned(addr, (__force u32 *)rp);
+	return 0;
+}
+
+#endif /* __ASM_XTENSA_FLAT_H */
diff --git a/arch/xtensa/include/asm/ftrace.h b/arch/xtensa/include/asm/ftrace.h
new file mode 100644
index 000000000..f676d209d
--- /dev/null
+++ b/arch/xtensa/include/asm/ftrace.h
@@ -0,0 +1,31 @@
+/*
+ * arch/xtensa/include/asm/ftrace.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2013 Tensilica Inc.
+ */
+#ifndef _XTENSA_FTRACE_H
+#define _XTENSA_FTRACE_H
+
+#include <asm/processor.h>
+
+#ifndef __ASSEMBLER__
+extern unsigned long return_address(unsigned level);
+#define ftrace_return_address(n) return_address(n)
+#endif /* __ASSEMBLER__ */
+
+#ifdef CONFIG_FUNCTION_TRACER
+
+#define MCOUNT_ADDR ((unsigned long)(_mcount))
+#define MCOUNT_INSN_SIZE 3
+
+#ifndef __ASSEMBLER__
+extern void _mcount(void);
+#define mcount _mcount
+#endif /* __ASSEMBLER__ */
+#endif /* CONFIG_FUNCTION_TRACER */
+
+#endif /* _XTENSA_FTRACE_H */
diff --git a/arch/xtensa/include/asm/futex.h b/arch/xtensa/include/asm/futex.h
new file mode 100644
index 000000000..a6f7d7ab5
--- /dev/null
+++ b/arch/xtensa/include/asm/futex.h
@@ -0,0 +1,167 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Atomic futex routines
+ *
+ * Based on the PowerPC implementataion
+ *
+ * Copyright (C) 2013 TangoTec Ltd.
+ *
+ * Baruch Siach <baruch@tkos.co.il>
+ */
+
+#ifndef _ASM_XTENSA_FUTEX_H
+#define _ASM_XTENSA_FUTEX_H
+
+#include <linux/futex.h>
+#include <linux/uaccess.h>
+#include <linux/errno.h>
+
+#define arch_futex_atomic_op_inuser arch_futex_atomic_op_inuser
+#define futex_atomic_cmpxchg_inatomic futex_atomic_cmpxchg_inatomic
+#include <asm-generic/futex.h>
+
+#if XCHAL_HAVE_EXCLUSIVE
+#define __futex_atomic_op(insn, ret, old, uaddr, arg)	\
+	__asm__ __volatile(				\
+	"1:	l32ex	%[oldval], %[addr]\n"		\
+		insn "\n"				\
+	"2:	s32ex	%[newval], %[addr]\n"		\
+	"	getex	%[newval]\n"			\
+	"	beqz	%[newval], 1b\n"		\
+	"	movi	%[newval], 0\n"			\
+	"3:\n"						\
+	"	.section .fixup,\"ax\"\n"		\
+	"	.align 4\n"				\
+	"	.literal_position\n"			\
+	"5:	movi	%[oldval], 3b\n"		\
+	"	movi	%[newval], %[fault]\n"		\
+	"	jx	%[oldval]\n"			\
+	"	.previous\n"				\
+	"	.section __ex_table,\"a\"\n"		\
+	"	.long 1b, 5b, 2b, 5b\n"			\
+	"	.previous\n"				\
+	: [oldval] "=&r" (old), [newval] "=&r" (ret)	\
+	: [addr] "r" (uaddr), [oparg] "r" (arg),	\
+	  [fault] "I" (-EFAULT)				\
+	: "memory")
+#elif XCHAL_HAVE_S32C1I
+#define __futex_atomic_op(insn, ret, old, uaddr, arg)	\
+	__asm__ __volatile(				\
+	"1:	l32i	%[oldval], %[mem]\n"		\
+		insn "\n"				\
+	"	wsr	%[oldval], scompare1\n"		\
+	"2:	s32c1i	%[newval], %[mem]\n"		\
+	"	bne	%[newval], %[oldval], 1b\n"	\
+	"	movi	%[newval], 0\n"			\
+	"3:\n"						\
+	"	.section .fixup,\"ax\"\n"		\
+	"	.align 4\n"				\
+	"	.literal_position\n"			\
+	"5:	movi	%[oldval], 3b\n"		\
+	"	movi	%[newval], %[fault]\n"		\
+	"	jx	%[oldval]\n"			\
+	"	.previous\n"				\
+	"	.section __ex_table,\"a\"\n"		\
+	"	.long 1b, 5b, 2b, 5b\n"			\
+	"	.previous\n"				\
+	: [oldval] "=&r" (old), [newval] "=&r" (ret),	\
+	  [mem] "+m" (*(uaddr))				\
+	: [oparg] "r" (arg), [fault] "I" (-EFAULT)	\
+	: "memory")
+#endif
+
+static inline int arch_futex_atomic_op_inuser(int op, int oparg, int *oval,
+		u32 __user *uaddr)
+{
+#if XCHAL_HAVE_S32C1I || XCHAL_HAVE_EXCLUSIVE
+	int oldval = 0, ret;
+
+	if (!access_ok(uaddr, sizeof(u32)))
+		return -EFAULT;
+
+	switch (op) {
+	case FUTEX_OP_SET:
+		__futex_atomic_op("mov %[newval], %[oparg]",
+				  ret, oldval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ADD:
+		__futex_atomic_op("add %[newval], %[oldval], %[oparg]",
+				  ret, oldval, uaddr, oparg);
+		break;
+	case FUTEX_OP_OR:
+		__futex_atomic_op("or %[newval], %[oldval], %[oparg]",
+				  ret, oldval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ANDN:
+		__futex_atomic_op("and %[newval], %[oldval], %[oparg]",
+				  ret, oldval, uaddr, ~oparg);
+		break;
+	case FUTEX_OP_XOR:
+		__futex_atomic_op("xor %[newval], %[oldval], %[oparg]",
+				  ret, oldval, uaddr, oparg);
+		break;
+	default:
+		ret = -ENOSYS;
+	}
+
+	if (!ret)
+		*oval = oldval;
+
+	return ret;
+#else
+	return futex_atomic_op_inuser_local(op, oparg, oval, uaddr);
+#endif
+}
+
+static inline int
+futex_atomic_cmpxchg_inatomic(u32 *uval, u32 __user *uaddr,
+			      u32 oldval, u32 newval)
+{
+#if XCHAL_HAVE_S32C1I || XCHAL_HAVE_EXCLUSIVE
+	unsigned long tmp;
+	int ret = 0;
+
+	if (!access_ok(uaddr, sizeof(u32)))
+		return -EFAULT;
+
+	__asm__ __volatile__ (
+	"	# futex_atomic_cmpxchg_inatomic\n"
+#if XCHAL_HAVE_EXCLUSIVE
+	"1:	l32ex	%[tmp], %[addr]\n"
+	"	s32i	%[tmp], %[uval], 0\n"
+	"	bne	%[tmp], %[oldval], 2f\n"
+	"	mov	%[tmp], %[newval]\n"
+	"3:	s32ex	%[tmp], %[addr]\n"
+	"	getex	%[tmp]\n"
+	"	beqz	%[tmp], 1b\n"
+#elif XCHAL_HAVE_S32C1I
+	"	wsr	%[oldval], scompare1\n"
+	"1:	s32c1i	%[newval], %[addr], 0\n"
+	"	s32i	%[newval], %[uval], 0\n"
+#endif
+	"2:\n"
+	"	.section .fixup,\"ax\"\n"
+	"	.align 4\n"
+	"	.literal_position\n"
+	"4:	movi	%[tmp], 2b\n"
+	"	movi	%[ret], %[fault]\n"
+	"	jx	%[tmp]\n"
+	"	.previous\n"
+	"	.section __ex_table,\"a\"\n"
+	"	.long 1b, 4b\n"
+#if XCHAL_HAVE_EXCLUSIVE
+	"	.long 3b, 4b\n"
+#endif
+	"	.previous\n"
+	: [ret] "+r" (ret), [newval] "+r" (newval), [tmp] "=&r" (tmp)
+	: [addr] "r" (uaddr), [oldval] "r" (oldval), [uval] "r" (uval),
+	  [fault] "I" (-EFAULT)
+	: "memory");
+
+	return ret;
+#else
+	return futex_atomic_cmpxchg_inatomic_local(uval, uaddr, oldval, newval);
+#endif
+}
+
+#endif /* _ASM_XTENSA_FUTEX_H */
diff --git a/arch/xtensa/include/asm/highmem.h b/arch/xtensa/include/asm/highmem.h
new file mode 100644
index 000000000..b55235f4a
--- /dev/null
+++ b/arch/xtensa/include/asm/highmem.h
@@ -0,0 +1,84 @@
+/*
+ * include/asm-xtensa/highmem.h
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2003 - 2005 Tensilica Inc.
+ * Copyright (C) 2014 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_HIGHMEM_H
+#define _XTENSA_HIGHMEM_H
+
+#ifdef CONFIG_HIGHMEM
+#include <linux/wait.h>
+#include <linux/pgtable.h>
+#include <asm/cacheflush.h>
+#include <asm/fixmap.h>
+
+#define PKMAP_BASE		((FIXADDR_START -			\
+				  (LAST_PKMAP + 1) * PAGE_SIZE) & PMD_MASK)
+#define LAST_PKMAP		(PTRS_PER_PTE * DCACHE_N_COLORS)
+#define LAST_PKMAP_MASK		(LAST_PKMAP - 1)
+#define PKMAP_NR(virt)		(((virt) - PKMAP_BASE) >> PAGE_SHIFT)
+#define PKMAP_ADDR(nr)		(PKMAP_BASE + ((nr) << PAGE_SHIFT))
+
+#define kmap_prot		PAGE_KERNEL_EXEC
+
+#if DCACHE_WAY_SIZE > PAGE_SIZE
+#define get_pkmap_color get_pkmap_color
+static inline int get_pkmap_color(const struct page *page)
+{
+	return DCACHE_ALIAS(page_to_phys(page));
+}
+
+extern unsigned int last_pkmap_nr_arr[];
+
+static inline unsigned int get_next_pkmap_nr(unsigned int color)
+{
+	last_pkmap_nr_arr[color] =
+		(last_pkmap_nr_arr[color] + DCACHE_N_COLORS) & LAST_PKMAP_MASK;
+	return last_pkmap_nr_arr[color] + color;
+}
+
+static inline int no_more_pkmaps(unsigned int pkmap_nr, unsigned int color)
+{
+	return pkmap_nr < DCACHE_N_COLORS;
+}
+
+static inline int get_pkmap_entries_count(unsigned int color)
+{
+	return LAST_PKMAP / DCACHE_N_COLORS;
+}
+
+extern wait_queue_head_t pkmap_map_wait_arr[];
+
+static inline wait_queue_head_t *get_pkmap_wait_queue_head(unsigned int color)
+{
+	return pkmap_map_wait_arr + color;
+}
+
+enum fixed_addresses kmap_local_map_idx(int type, unsigned long pfn);
+#define arch_kmap_local_map_idx		kmap_local_map_idx
+
+enum fixed_addresses kmap_local_unmap_idx(int type, unsigned long addr);
+#define arch_kmap_local_unmap_idx	kmap_local_unmap_idx
+
+#endif
+
+extern pte_t *pkmap_page_table;
+
+static inline void flush_cache_kmaps(void)
+{
+	flush_cache_all();
+}
+
+#define arch_kmap_local_post_unmap(vaddr)	\
+	local_flush_tlb_kernel_range(vaddr, vaddr + PAGE_SIZE)
+
+void kmap_init(void);
+
+#endif /* CONFIG_HIGHMEM */
+#endif
diff --git a/arch/xtensa/include/asm/hw_breakpoint.h b/arch/xtensa/include/asm/hw_breakpoint.h
new file mode 100644
index 000000000..9ec86f440
--- /dev/null
+++ b/arch/xtensa/include/asm/hw_breakpoint.h
@@ -0,0 +1,62 @@
+/*
+ * Xtensa hardware breakpoints/watchpoints handling functions
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2016 Cadence Design Systems Inc.
+ */
+
+#ifndef __ASM_XTENSA_HW_BREAKPOINT_H
+#define __ASM_XTENSA_HW_BREAKPOINT_H
+
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+
+#include <linux/kdebug.h>
+#include <linux/types.h>
+#include <uapi/linux/hw_breakpoint.h>
+
+/* Breakpoint */
+#define XTENSA_BREAKPOINT_EXECUTE	0
+
+/* Watchpoints */
+#define XTENSA_BREAKPOINT_LOAD		1
+#define XTENSA_BREAKPOINT_STORE		2
+
+struct arch_hw_breakpoint {
+	unsigned long address;
+	u16 len;
+	u16 type;
+};
+
+struct perf_event_attr;
+struct perf_event;
+struct pt_regs;
+struct task_struct;
+
+int hw_breakpoint_slots(int type);
+int arch_check_bp_in_kernelspace(struct arch_hw_breakpoint *hw);
+int hw_breakpoint_arch_parse(struct perf_event *bp,
+			     const struct perf_event_attr *attr,
+			     struct arch_hw_breakpoint *hw);
+int hw_breakpoint_exceptions_notify(struct notifier_block *unused,
+				    unsigned long val, void *data);
+
+int arch_install_hw_breakpoint(struct perf_event *bp);
+void arch_uninstall_hw_breakpoint(struct perf_event *bp);
+void hw_breakpoint_pmu_read(struct perf_event *bp);
+int check_hw_breakpoint(struct pt_regs *regs);
+void clear_ptrace_hw_breakpoint(struct task_struct *tsk);
+void restore_dbreak(void);
+
+#else
+
+struct task_struct;
+
+static inline void clear_ptrace_hw_breakpoint(struct task_struct *tsk)
+{
+}
+
+#endif /* CONFIG_HAVE_HW_BREAKPOINT */
+#endif /* __ASM_XTENSA_HW_BREAKPOINT_H */
diff --git a/arch/xtensa/include/asm/initialize_mmu.h b/arch/xtensa/include/asm/initialize_mmu.h
new file mode 100644
index 000000000..101bcb87e
--- /dev/null
+++ b/arch/xtensa/include/asm/initialize_mmu.h
@@ -0,0 +1,245 @@
+/*
+ * arch/xtensa/include/asm/initialize_mmu.h
+ *
+ * Initializes MMU:
+ *
+ *      For the new V3 MMU we remap the TLB from virtual == physical
+ *      to the standard Linux mapping used in earlier MMU's.
+ *
+ *      For the MMU we also support a new configuration register that
+ *      specifies how the S32C1I instruction operates with the cache
+ *      controller.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2008 - 2012 Tensilica, Inc.
+ *
+ *   Marc Gauthier <marc@tensilica.com>
+ *   Pete Delaney <piet@tensilica.com>
+ */
+
+#ifndef _XTENSA_INITIALIZE_MMU_H
+#define _XTENSA_INITIALIZE_MMU_H
+
+#include <linux/init.h>
+#include <linux/pgtable.h>
+#include <asm/vectors.h>
+
+#if XCHAL_HAVE_PTP_MMU
+#define CA_BYPASS	(_PAGE_CA_BYPASS | _PAGE_HW_WRITE | _PAGE_HW_EXEC)
+#define CA_WRITEBACK	(_PAGE_CA_WB     | _PAGE_HW_WRITE | _PAGE_HW_EXEC)
+#else
+#define CA_WRITEBACK	(0x4)
+#endif
+
+#ifdef __ASSEMBLER__
+
+#define XTENSA_HWVERSION_RC_2009_0 230000
+
+	.macro	initialize_mmu
+
+#if XCHAL_HAVE_S32C1I && (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RC_2009_0)
+/*
+ * We Have Atomic Operation Control (ATOMCTL) Register; Initialize it.
+ * For details see Documentation/arch/xtensa/atomctl.rst
+ */
+#if XCHAL_DCACHE_IS_COHERENT
+	movi	a3, 0x25	/* For SMP/MX -- internal for writeback,
+				 * RCW otherwise
+				 */
+#else
+	movi	a3, 0x29	/* non-MX -- Most cores use Std Memory
+				 * Controlers which usually can't use RCW
+				 */
+#endif
+	wsr	a3, atomctl
+#endif  /* XCHAL_HAVE_S32C1I &&
+	 * (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RC_2009_0)
+	 */
+
+#if defined(CONFIG_MMU) && XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY
+/*
+ * Have MMU v3
+ */
+
+#if !XCHAL_HAVE_VECBASE
+# error "MMU v3 requires reloc vectors"
+#endif
+
+	movi	a1, 0
+	_call0	1f
+	_j	2f
+
+	.align	4
+1:
+
+#if CONFIG_KERNEL_LOAD_ADDRESS < 0x40000000ul
+#define TEMP_MAPPING_VADDR 0x40000000
+#else
+#define TEMP_MAPPING_VADDR 0x00000000
+#endif
+
+	/* Step 1: invalidate mapping at 0x40000000..0x5FFFFFFF. */
+
+	movi	a2, TEMP_MAPPING_VADDR | XCHAL_SPANNING_WAY
+	idtlb	a2
+	iitlb	a2
+	isync
+
+	/* Step 2: map 0x40000000..0x47FFFFFF to paddr containing this code
+	 * and jump to the new mapping.
+	 */
+
+	srli	a3, a0, 27
+	slli	a3, a3, 27
+	addi	a3, a3, CA_BYPASS
+	addi	a7, a2, 5 - XCHAL_SPANNING_WAY
+	wdtlb	a3, a7
+	witlb	a3, a7
+	isync
+
+	slli	a4, a0, 5
+	srli	a4, a4, 5
+	addi	a5, a2, -XCHAL_SPANNING_WAY
+	add	a4, a4, a5
+	jx	a4
+
+	/* Step 3: unmap everything other than current area.
+	 *	   Start at 0x60000000, wrap around, and end with 0x20000000
+	 */
+2:	movi	a4, 0x20000000
+	add	a5, a2, a4
+3:	idtlb	a5
+	iitlb	a5
+	add	a5, a5, a4
+	bne	a5, a2, 3b
+
+	/* Step 4: Setup MMU with the requested static mappings. */
+
+	movi	a6, 0x01000000
+	wsr	a6, ITLBCFG
+	wsr	a6, DTLBCFG
+	isync
+
+	movi	a5, XCHAL_KSEG_CACHED_VADDR + XCHAL_KSEG_TLB_WAY
+	movi	a4, XCHAL_KSEG_PADDR + CA_WRITEBACK
+	wdtlb	a4, a5
+	witlb	a4, a5
+
+	movi	a5, XCHAL_KSEG_BYPASS_VADDR + XCHAL_KSEG_TLB_WAY
+	movi	a4, XCHAL_KSEG_PADDR + CA_BYPASS
+	wdtlb	a4, a5
+	witlb	a4, a5
+
+#ifdef CONFIG_XTENSA_KSEG_512M
+	movi	a5, XCHAL_KSEG_CACHED_VADDR + 0x10000000 + XCHAL_KSEG_TLB_WAY
+	movi	a4, XCHAL_KSEG_PADDR + 0x10000000 + CA_WRITEBACK
+	wdtlb	a4, a5
+	witlb	a4, a5
+
+	movi	a5, XCHAL_KSEG_BYPASS_VADDR + 0x10000000 + XCHAL_KSEG_TLB_WAY
+	movi	a4, XCHAL_KSEG_PADDR + 0x10000000 + CA_BYPASS
+	wdtlb	a4, a5
+	witlb	a4, a5
+#endif
+
+	movi	a5, XCHAL_KIO_CACHED_VADDR + XCHAL_KIO_TLB_WAY
+	movi	a4, XCHAL_KIO_DEFAULT_PADDR + CA_WRITEBACK
+	wdtlb	a4, a5
+	witlb	a4, a5
+
+	movi	a5, XCHAL_KIO_BYPASS_VADDR + XCHAL_KIO_TLB_WAY
+	movi	a4, XCHAL_KIO_DEFAULT_PADDR + CA_BYPASS
+	wdtlb	a4, a5
+	witlb	a4, a5
+
+	isync
+
+	/* Jump to self, using final mappings. */
+	movi	a4, 1f
+	jx	a4
+
+1:
+	/* Step 5: remove temporary mapping. */
+	idtlb	a7
+	iitlb	a7
+	isync
+
+	movi	a0, 0
+	wsr	a0, ptevaddr
+	rsync
+
+#endif /* defined(CONFIG_MMU) && XCHAL_HAVE_PTP_MMU &&
+	  XCHAL_HAVE_SPANNING_WAY */
+
+	.endm
+
+	.macro	initialize_cacheattr
+
+#if !defined(CONFIG_MMU) && (XCHAL_HAVE_TLBS || XCHAL_HAVE_MPU)
+#if CONFIG_MEMMAP_CACHEATTR == 0x22222222 && XCHAL_HAVE_PTP_MMU
+#error Default MEMMAP_CACHEATTR of 0x22222222 does not work with full MMU.
+#endif
+
+#if XCHAL_HAVE_MPU
+	__REFCONST
+	.align	4
+.Lattribute_table:
+	.long 0x000000, 0x1fff00, 0x1ddf00, 0x1eef00
+	.long 0x006600, 0x000000, 0x000000, 0x000000
+	.long 0x000000, 0x000000, 0x000000, 0x000000
+	.long 0x000000, 0x000000, 0x000000, 0x000000
+	.previous
+
+	movi	a3, .Lattribute_table
+	movi	a4, CONFIG_MEMMAP_CACHEATTR
+	movi	a5, 1
+	movi	a6, XCHAL_MPU_ENTRIES
+	movi	a10, 0x20000000
+	movi	a11, -1
+1:
+	sub	a5, a5, a10
+	extui	a8, a4, 28, 4
+	beq	a8, a11, 2f
+	addi	a6, a6, -1
+	mov	a11, a8
+2:
+	addx4	a9, a8, a3
+	l32i	a9, a9, 0
+	or	a9, a9, a6
+	wptlb	a9, a5
+	slli	a4, a4, 4
+	bgeu	a5, a10, 1b
+
+#else
+	movi	a5, XCHAL_SPANNING_WAY
+	movi	a6, ~_PAGE_ATTRIB_MASK
+	movi	a4, CONFIG_MEMMAP_CACHEATTR
+	movi	a8, 0x20000000
+1:
+	rdtlb1	a3, a5
+	xor	a3, a3, a4
+	and	a3, a3, a6
+	xor	a3, a3, a4
+	wdtlb	a3, a5
+	ritlb1	a3, a5
+	xor	a3, a3, a4
+	and	a3, a3, a6
+	xor	a3, a3, a4
+	witlb	a3, a5
+
+	add	a5, a5, a8
+	srli	a4, a4, 4
+	bgeu	a5, a8, 1b
+
+	isync
+#endif
+#endif
+
+	.endm
+
+#endif /*__ASSEMBLER__*/
+
+#endif /* _XTENSA_INITIALIZE_MMU_H */
diff --git a/arch/xtensa/include/asm/io.h b/arch/xtensa/include/asm/io.h
new file mode 100644
index 000000000..7cdcc2dea
--- /dev/null
+++ b/arch/xtensa/include/asm/io.h
@@ -0,0 +1,62 @@
+/*
+ * include/asm-xtensa/io.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_IO_H
+#define _XTENSA_IO_H
+
+#include <asm/byteorder.h>
+#include <asm/page.h>
+#include <asm/vectors.h>
+#include <linux/bug.h>
+#include <linux/kernel.h>
+#include <linux/pgtable.h>
+
+#include <linux/types.h>
+
+#define IOADDR(x)		(XCHAL_KIO_BYPASS_VADDR + (x))
+#define IO_SPACE_LIMIT ~0
+#define PCI_IOBASE		((void __iomem *)XCHAL_KIO_BYPASS_VADDR)
+
+#ifdef CONFIG_MMU
+/*
+ * I/O memory mapping functions.
+ */
+void __iomem *ioremap_prot(phys_addr_t phys_addr, size_t size,
+			   pgprot_t prot);
+#define ioremap_prot ioremap_prot
+#define iounmap iounmap
+
+static inline void __iomem *ioremap(unsigned long offset, unsigned long size)
+{
+	if (offset >= XCHAL_KIO_PADDR
+	    && offset - XCHAL_KIO_PADDR < XCHAL_KIO_SIZE)
+		return (void*)(offset-XCHAL_KIO_PADDR+XCHAL_KIO_BYPASS_VADDR);
+	else
+		return ioremap_prot(offset, size,
+				    pgprot_noncached(PAGE_KERNEL));
+}
+#define ioremap ioremap
+
+static inline void __iomem *ioremap_cache(unsigned long offset,
+		unsigned long size)
+{
+	if (offset >= XCHAL_KIO_PADDR
+	    && offset - XCHAL_KIO_PADDR < XCHAL_KIO_SIZE)
+		return (void*)(offset-XCHAL_KIO_PADDR+XCHAL_KIO_CACHED_VADDR);
+	else
+		return ioremap_prot(offset, size, PAGE_KERNEL);
+
+}
+#define ioremap_cache ioremap_cache
+#endif /* CONFIG_MMU */
+
+#include <asm-generic/io.h>
+
+#endif	/* _XTENSA_IO_H */
diff --git a/arch/xtensa/include/asm/irq.h b/arch/xtensa/include/asm/irq.h
new file mode 100644
index 000000000..0f71a51da
--- /dev/null
+++ b/arch/xtensa/include/asm/irq.h
@@ -0,0 +1,42 @@
+/*
+ * include/asm-xtensa/irq.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_IRQ_H
+#define _XTENSA_IRQ_H
+
+#include <linux/init.h>
+#include <asm/core.h>
+
+#ifdef CONFIG_PLATFORM_NR_IRQS
+# define PLATFORM_NR_IRQS CONFIG_PLATFORM_NR_IRQS
+#else
+# define PLATFORM_NR_IRQS 0
+#endif
+#define XTENSA_NR_IRQS XCHAL_NUM_INTERRUPTS
+#define NR_IRQS (XTENSA_NR_IRQS + PLATFORM_NR_IRQS + 1)
+#define XTENSA_PIC_LINUX_IRQ(hwirq) ((hwirq) + 1)
+
+static __inline__ int irq_canonicalize(int irq)
+{
+	return (irq);
+}
+
+struct irqaction;
+struct irq_domain;
+
+void migrate_irqs(void);
+int xtensa_irq_domain_xlate(const u32 *intspec, unsigned int intsize,
+		unsigned long int_irq, unsigned long ext_irq,
+		unsigned long *out_hwirq, unsigned int *out_type);
+int xtensa_irq_map(struct irq_domain *d, unsigned int irq, irq_hw_number_t hw);
+unsigned xtensa_map_ext_irq(unsigned ext_irq);
+unsigned xtensa_get_ext_irq_no(unsigned irq);
+
+#endif	/* _XTENSA_IRQ_H */
diff --git a/arch/xtensa/include/asm/irqflags.h b/arch/xtensa/include/asm/irqflags.h
new file mode 100644
index 000000000..128906815
--- /dev/null
+++ b/arch/xtensa/include/asm/irqflags.h
@@ -0,0 +1,83 @@
+/*
+ * Xtensa IRQ flags handling functions
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_IRQFLAGS_H
+#define _XTENSA_IRQFLAGS_H
+
+#include <linux/stringify.h>
+#include <linux/types.h>
+#include <asm/processor.h>
+
+static inline unsigned long arch_local_save_flags(void)
+{
+	unsigned long flags;
+	asm volatile("rsr %0, ps" : "=a" (flags));
+	return flags;
+}
+
+static inline unsigned long arch_local_irq_save(void)
+{
+	unsigned long flags;
+#if XTENSA_FAKE_NMI
+#if defined(CONFIG_DEBUG_MISC) && (LOCKLEVEL | TOPLEVEL) >= XCHAL_DEBUGLEVEL
+	unsigned long tmp;
+
+	asm volatile("rsr	%0, ps\t\n"
+		     "extui	%1, %0, 0, 4\t\n"
+		     "bgei	%1, "__stringify(LOCKLEVEL)", 1f\t\n"
+		     "rsil	%0, "__stringify(LOCKLEVEL)"\n"
+		     "1:"
+		     : "=a" (flags), "=a" (tmp) :: "memory");
+#else
+	asm volatile("rsr	%0, ps\t\n"
+		     "or	%0, %0, %1\t\n"
+		     "xsr	%0, ps\t\n"
+		     "rsync"
+		     : "=&a" (flags) : "a" (LOCKLEVEL) : "memory");
+#endif
+#else
+	asm volatile("rsil	%0, "__stringify(LOCKLEVEL)
+		     : "=a" (flags) :: "memory");
+#endif
+	return flags;
+}
+
+static inline void arch_local_irq_disable(void)
+{
+	arch_local_irq_save();
+}
+
+static inline void arch_local_irq_enable(void)
+{
+	unsigned long flags;
+	asm volatile("rsil %0, 0" : "=a" (flags) :: "memory");
+}
+
+static inline void arch_local_irq_restore(unsigned long flags)
+{
+	asm volatile("wsr %0, ps; rsync"
+		     :: "a" (flags) : "memory");
+}
+
+static inline bool arch_irqs_disabled_flags(unsigned long flags)
+{
+#if XCHAL_EXCM_LEVEL < LOCKLEVEL || (1 << PS_EXCM_BIT) < LOCKLEVEL
+#error "XCHAL_EXCM_LEVEL and 1<<PS_EXCM_BIT must be no less than LOCKLEVEL"
+#endif
+	return (flags & (PS_INTLEVEL_MASK | (1 << PS_EXCM_BIT))) >= LOCKLEVEL;
+}
+
+static inline bool arch_irqs_disabled(void)
+{
+	return arch_irqs_disabled_flags(arch_local_save_flags());
+}
+
+#endif /* _XTENSA_IRQFLAGS_H */
diff --git a/arch/xtensa/include/asm/jump_label.h b/arch/xtensa/include/asm/jump_label.h
new file mode 100644
index 000000000..38e3e2a9b
--- /dev/null
+++ b/arch/xtensa/include/asm/jump_label.h
@@ -0,0 +1,65 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2018 Cadence Design Systems Inc. */
+
+#ifndef _ASM_XTENSA_JUMP_LABEL_H
+#define _ASM_XTENSA_JUMP_LABEL_H
+
+#ifndef __ASSEMBLER__
+
+#include <linux/types.h>
+
+#define JUMP_LABEL_NOP_SIZE 3
+
+static __always_inline bool arch_static_branch(struct static_key *key,
+					       bool branch)
+{
+	asm goto("1:\n\t"
+			  "_nop\n\t"
+			  ".pushsection __jump_table,  \"aw\"\n\t"
+			  ".word 1b, %l[l_yes], %c0\n\t"
+			  ".popsection\n\t"
+			  : :  "i" (&((char *)key)[branch]) :  : l_yes);
+
+	return false;
+l_yes:
+	return true;
+}
+
+static __always_inline bool arch_static_branch_jump(struct static_key *key,
+						    bool branch)
+{
+	/*
+	 * Xtensa assembler will mark certain points in the code
+	 * as unreachable, so that later assembler or linker relaxation
+	 * passes could use them. A spot right after the J instruction
+	 * is one such point. Assembler and/or linker may insert padding
+	 * or literals here, breaking code flow in case the J instruction
+	 * is later replaced with NOP. Put a label right after the J to
+	 * make it reachable and wrap both into a no-transform block
+	 * to avoid any assembler interference with this.
+	 */
+	asm goto("1:\n\t"
+			  ".begin no-transform\n\t"
+			  "_j %l[l_yes]\n\t"
+			  "2:\n\t"
+			  ".end no-transform\n\t"
+			  ".pushsection __jump_table,  \"aw\"\n\t"
+			  ".word 1b, %l[l_yes], %c0\n\t"
+			  ".popsection\n\t"
+			  : :  "i" (&((char *)key)[branch]) :  : l_yes);
+
+	return false;
+l_yes:
+	return true;
+}
+
+typedef u32 jump_label_t;
+
+struct jump_entry {
+	jump_label_t code;
+	jump_label_t target;
+	jump_label_t key;
+};
+
+#endif  /* __ASSEMBLER__ */
+#endif
diff --git a/arch/xtensa/include/asm/kasan.h b/arch/xtensa/include/asm/kasan.h
new file mode 100644
index 000000000..0da91b64f
--- /dev/null
+++ b/arch/xtensa/include/asm/kasan.h
@@ -0,0 +1,41 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_KASAN_H
+#define __ASM_KASAN_H
+
+#ifndef __ASSEMBLER__
+
+#ifdef CONFIG_KASAN
+
+#include <linux/kernel.h>
+#include <linux/sizes.h>
+#include <asm/kmem_layout.h>
+
+#define KASAN_SHADOW_SCALE_SHIFT 3
+
+/* Start of area covered by KASAN */
+#define KASAN_START_VADDR __XTENSA_UL_CONST(0x90000000)
+/* Start of the shadow map */
+#define KASAN_SHADOW_START (XCHAL_PAGE_TABLE_VADDR + XCHAL_PAGE_TABLE_SIZE)
+/* Size of the shadow map */
+#define KASAN_SHADOW_SIZE (-KASAN_START_VADDR >> KASAN_SHADOW_SCALE_SHIFT)
+/* End of the shadow map */
+#define KASAN_SHADOW_END (KASAN_SHADOW_START + KASAN_SHADOW_SIZE)
+/* Offset for mem to shadow address transformation */
+#define KASAN_SHADOW_OFFSET __XTENSA_UL_CONST(CONFIG_KASAN_SHADOW_OFFSET)
+
+void __init kasan_early_init(void);
+void __init kasan_init(void);
+
+#else
+
+static inline void kasan_early_init(void)
+{
+}
+
+static inline void kasan_init(void)
+{
+}
+
+#endif
+#endif
+#endif
diff --git a/arch/xtensa/include/asm/kmem_layout.h b/arch/xtensa/include/asm/kmem_layout.h
new file mode 100644
index 000000000..694972462
--- /dev/null
+++ b/arch/xtensa/include/asm/kmem_layout.h
@@ -0,0 +1,104 @@
+/*
+ * Kernel virtual memory layout definitions.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2016 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_KMEM_LAYOUT_H
+#define _XTENSA_KMEM_LAYOUT_H
+
+#include <asm/core.h>
+#include <asm/types.h>
+
+#ifdef CONFIG_MMU
+
+/*
+ * Fixed TLB translations in the processor.
+ */
+
+#define XCHAL_PAGE_TABLE_VADDR	__XTENSA_UL_CONST(0x80000000)
+#define XCHAL_PAGE_TABLE_SIZE	__XTENSA_UL_CONST(0x00400000)
+
+#if defined(CONFIG_XTENSA_KSEG_MMU_V2)
+
+#define XCHAL_KSEG_CACHED_VADDR	__XTENSA_UL_CONST(0xd0000000)
+#define XCHAL_KSEG_BYPASS_VADDR	__XTENSA_UL_CONST(0xd8000000)
+#define XCHAL_KSEG_SIZE		__XTENSA_UL_CONST(0x08000000)
+#define XCHAL_KSEG_ALIGNMENT	__XTENSA_UL_CONST(0x08000000)
+#define XCHAL_KSEG_TLB_WAY	5
+#define XCHAL_KIO_TLB_WAY	6
+
+#elif defined(CONFIG_XTENSA_KSEG_256M)
+
+#define XCHAL_KSEG_CACHED_VADDR	__XTENSA_UL_CONST(0xb0000000)
+#define XCHAL_KSEG_BYPASS_VADDR	__XTENSA_UL_CONST(0xc0000000)
+#define XCHAL_KSEG_SIZE		__XTENSA_UL_CONST(0x10000000)
+#define XCHAL_KSEG_ALIGNMENT	__XTENSA_UL_CONST(0x10000000)
+#define XCHAL_KSEG_TLB_WAY	6
+#define XCHAL_KIO_TLB_WAY	6
+
+#elif defined(CONFIG_XTENSA_KSEG_512M)
+
+#define XCHAL_KSEG_CACHED_VADDR	__XTENSA_UL_CONST(0xa0000000)
+#define XCHAL_KSEG_BYPASS_VADDR	__XTENSA_UL_CONST(0xc0000000)
+#define XCHAL_KSEG_SIZE		__XTENSA_UL_CONST(0x20000000)
+#define XCHAL_KSEG_ALIGNMENT	__XTENSA_UL_CONST(0x10000000)
+#define XCHAL_KSEG_TLB_WAY	6
+#define XCHAL_KIO_TLB_WAY	6
+
+#else
+#error Unsupported KSEG configuration
+#endif
+
+#ifdef CONFIG_KSEG_PADDR
+#define XCHAL_KSEG_PADDR        __XTENSA_UL_CONST(CONFIG_KSEG_PADDR)
+#else
+#define XCHAL_KSEG_PADDR	__XTENSA_UL_CONST(0x00000000)
+#endif
+
+#if XCHAL_KSEG_PADDR & (XCHAL_KSEG_ALIGNMENT - 1)
+#error XCHAL_KSEG_PADDR is not properly aligned to XCHAL_KSEG_ALIGNMENT
+#endif
+
+#endif
+
+/* KIO definition */
+
+#if XCHAL_HAVE_PTP_MMU
+#define XCHAL_KIO_CACHED_VADDR		0xe0000000
+#define XCHAL_KIO_BYPASS_VADDR		0xf0000000
+#define XCHAL_KIO_DEFAULT_PADDR		0xf0000000
+#else
+#define XCHAL_KIO_BYPASS_VADDR		XCHAL_KIO_PADDR
+#define XCHAL_KIO_DEFAULT_PADDR		0x90000000
+#endif
+#define XCHAL_KIO_SIZE			0x10000000
+
+#if (!XCHAL_HAVE_PTP_MMU || XCHAL_HAVE_SPANNING_WAY) && defined(CONFIG_USE_OF)
+#define XCHAL_KIO_PADDR			xtensa_get_kio_paddr()
+#ifndef __ASSEMBLER__
+extern unsigned long xtensa_kio_paddr;
+
+static inline unsigned long xtensa_get_kio_paddr(void)
+{
+	return xtensa_kio_paddr;
+}
+#endif
+#else
+#define XCHAL_KIO_PADDR			XCHAL_KIO_DEFAULT_PADDR
+#endif
+
+/* KERNEL_STACK definition */
+
+#ifndef CONFIG_KASAN
+#define KERNEL_STACK_SHIFT	13
+#else
+#define KERNEL_STACK_SHIFT	15
+#endif
+#define KERNEL_STACK_SIZE	(1 << KERNEL_STACK_SHIFT)
+
+#endif
diff --git a/arch/xtensa/include/asm/linkage.h b/arch/xtensa/include/asm/linkage.h
new file mode 100644
index 000000000..0ba997323
--- /dev/null
+++ b/arch/xtensa/include/asm/linkage.h
@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef __ASM_LINKAGE_H
+#define __ASM_LINKAGE_H
+
+#define __ALIGN		.align 4
+#define __ALIGN_STR	".align 4"
+
+#endif
diff --git a/arch/xtensa/include/asm/mmu.h b/arch/xtensa/include/asm/mmu.h
new file mode 100644
index 000000000..71afe418d
--- /dev/null
+++ b/arch/xtensa/include/asm/mmu.h
@@ -0,0 +1,22 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_MMU_H
+#define _XTENSA_MMU_H
+
+#ifndef CONFIG_MMU
+#include <asm-generic/mmu.h>
+#else
+
+typedef struct {
+	unsigned long asid[NR_CPUS];
+	unsigned int cpu;
+} mm_context_t;
+
+#endif /* CONFIG_MMU */
+#endif	/* _XTENSA_MMU_H */
diff --git a/arch/xtensa/include/asm/mmu_context.h b/arch/xtensa/include/asm/mmu_context.h
new file mode 100644
index 000000000..e337ba968
--- /dev/null
+++ b/arch/xtensa/include/asm/mmu_context.h
@@ -0,0 +1,154 @@
+/*
+ * Switch an MMU context.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_MMU_CONTEXT_H
+#define _XTENSA_MMU_CONTEXT_H
+
+#ifndef CONFIG_MMU
+#include <asm/nommu_context.h>
+#else
+
+#include <linux/stringify.h>
+#include <linux/sched.h>
+#include <linux/mm_types.h>
+#include <linux/pgtable.h>
+
+#include <asm/vectors.h>
+
+#include <asm/cacheflush.h>
+#include <asm/tlbflush.h>
+#include <asm-generic/mm_hooks.h>
+#include <asm-generic/percpu.h>
+
+#if (XCHAL_HAVE_TLBS != 1)
+# error "Linux must have an MMU!"
+#endif
+
+DECLARE_PER_CPU(unsigned long, asid_cache);
+#define cpu_asid_cache(cpu) per_cpu(asid_cache, cpu)
+
+/*
+ * NO_CONTEXT is the invalid ASID value that we don't ever assign to
+ * any user or kernel context.  We use the reserved values in the
+ * ASID_INSERT macro below.
+ *
+ * 0 invalid
+ * 1 kernel
+ * 2 reserved
+ * 3 reserved
+ * 4...255 available
+ */
+
+#define NO_CONTEXT	0
+#define ASID_USER_FIRST	4
+#define ASID_MASK	((1 << XCHAL_MMU_ASID_BITS) - 1)
+#define ASID_INSERT(x)	(0x03020001 | (((x) & ASID_MASK) << 8))
+
+void init_mmu(void);
+void init_kio(void);
+
+static inline void set_rasid_register (unsigned long val)
+{
+	__asm__ __volatile__ (" wsr %0, rasid\n\t"
+			      " isync\n" : : "a" (val));
+}
+
+static inline unsigned long get_rasid_register (void)
+{
+	unsigned long tmp;
+	__asm__ __volatile__ (" rsr %0, rasid\n\t" : "=a" (tmp));
+	return tmp;
+}
+
+static inline void get_new_mmu_context(struct mm_struct *mm, unsigned int cpu)
+{
+	unsigned long asid = cpu_asid_cache(cpu);
+	if ((++asid & ASID_MASK) == 0) {
+		/*
+		 * Start new asid cycle; continue counting with next
+		 * incarnation bits; skipping over 0, 1, 2, 3.
+		 */
+		local_flush_tlb_all();
+		asid += ASID_USER_FIRST;
+	}
+	cpu_asid_cache(cpu) = asid;
+	mm->context.asid[cpu] = asid;
+	mm->context.cpu = cpu;
+}
+
+static inline void get_mmu_context(struct mm_struct *mm, unsigned int cpu)
+{
+	/*
+	 * Check if our ASID is of an older version and thus invalid.
+	 */
+
+	if (mm) {
+		unsigned long asid = mm->context.asid[cpu];
+
+		if (asid == NO_CONTEXT ||
+				((asid ^ cpu_asid_cache(cpu)) & ~ASID_MASK))
+			get_new_mmu_context(mm, cpu);
+	}
+}
+
+static inline void activate_context(struct mm_struct *mm, unsigned int cpu)
+{
+	get_mmu_context(mm, cpu);
+	set_rasid_register(ASID_INSERT(mm->context.asid[cpu]));
+	invalidate_page_directory();
+}
+
+/*
+ * Initialize the context related info for a new mm_struct
+ * instance.  Valid cpu values are 0..(NR_CPUS-1), so initializing
+ * to -1 says the process has never run on any core.
+ */
+
+#define init_new_context init_new_context
+static inline int init_new_context(struct task_struct *tsk,
+		struct mm_struct *mm)
+{
+	int cpu;
+	for_each_possible_cpu(cpu) {
+		mm->context.asid[cpu] = NO_CONTEXT;
+	}
+	mm->context.cpu = -1;
+	return 0;
+}
+
+static inline void switch_mm(struct mm_struct *prev, struct mm_struct *next,
+			     struct task_struct *tsk)
+{
+	unsigned int cpu = smp_processor_id();
+	int migrated = next->context.cpu != cpu;
+	/* Flush the icache if we migrated to a new core. */
+	if (migrated) {
+		__invalidate_icache_all();
+		next->context.cpu = cpu;
+	}
+	if (migrated || prev != next)
+		activate_context(next, cpu);
+}
+
+/*
+ * Destroy context related info for an mm_struct that is about
+ * to be put to rest.
+ */
+#define destroy_context destroy_context
+static inline void destroy_context(struct mm_struct *mm)
+{
+	invalidate_page_directory();
+}
+
+
+#include <asm-generic/mmu_context.h>
+
+#endif /* CONFIG_MMU */
+#endif /* _XTENSA_MMU_CONTEXT_H */
diff --git a/arch/xtensa/include/asm/mtd-xip.h b/arch/xtensa/include/asm/mtd-xip.h
new file mode 100644
index 000000000..514325155
--- /dev/null
+++ b/arch/xtensa/include/asm/mtd-xip.h
@@ -0,0 +1,14 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+
+#ifndef _ASM_MTD_XIP_H
+#define _ASM_MTD_XIP_H
+
+#include <asm/processor.h>
+
+#define xip_irqpending()	(xtensa_get_sr(interrupt) & xtensa_get_sr(intenable))
+#define xip_currtime()		(xtensa_get_sr(ccount))
+#define xip_elapsed_since(x)	((xtensa_get_sr(ccount) - (x)) / 1000) /* should work up to 1GHz */
+#define xip_cpu_idle()		do { asm volatile ("waiti 0"); } while (0)
+
+#endif /* _ASM_MTD_XIP_H */
+
diff --git a/arch/xtensa/include/asm/mxregs.h b/arch/xtensa/include/asm/mxregs.h
new file mode 100644
index 000000000..73dcc5456
--- /dev/null
+++ b/arch/xtensa/include/asm/mxregs.h
@@ -0,0 +1,46 @@
+/*
+ * Xtensa MX interrupt distributor
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2008 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_MXREGS_H
+#define _XTENSA_MXREGS_H
+
+/*
+ * RER/WER at, as	Read/write external register
+ *	at: value
+ *	as: address
+ *
+ * Address	Value
+ * 00nn		0...0p..p	Interrupt Routing, route IRQ n to processor p
+ * 01pp		0...0d..d	16 bits (d) 'ored' as single IPI to processor p
+ * 0180		0...0m..m	Clear enable specified by mask (m)
+ * 0184		0...0m..m	Set enable specified by mask (m)
+ * 0190		0...0x..x	8-bit IPI partition register
+ *				VVVVVVVVPPPPUUUUUUUUUUUUUUUUU
+ *				V (10-bit) Release/Version
+ *				P ( 4-bit) Number of cores - 1
+ *				U (18-bit) ID
+ * 01a0		i.......i	32-bit ConfigID
+ * 0200		0...0m..m	RunStall core 'n'
+ * 0220		c		Cache coherency enabled
+ */
+
+#define MIROUT(irq)	(0x000 + (irq))
+#define MIPICAUSE(cpu)	(0x100 + (cpu))
+#define MIPISET(cause)	(0x140 + (cause))
+#define MIENG		0x180
+#define MIENGSET	0x184
+#define MIASG		0x188	/* Read Global Assert Register */
+#define MIASGSET	0x18c	/* Set Global Addert Regiter */
+#define MIPIPART	0x190
+#define SYSCFGID	0x1a0
+#define MPSCORE		0x200
+#define CCON		0x220
+
+#endif /* _XTENSA_MXREGS_H */
diff --git a/arch/xtensa/include/asm/nommu_context.h b/arch/xtensa/include/asm/nommu_context.h
new file mode 100644
index 000000000..7c9d1918d
--- /dev/null
+++ b/arch/xtensa/include/asm/nommu_context.h
@@ -0,0 +1,10 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+static inline void init_mmu(void)
+{
+}
+
+static inline void init_kio(void)
+{
+}
+
+#include <asm-generic/nommu_context.h>
diff --git a/arch/xtensa/include/asm/page.h b/arch/xtensa/include/asm/page.h
new file mode 100644
index 000000000..059493256
--- /dev/null
+++ b/arch/xtensa/include/asm/page.h
@@ -0,0 +1,177 @@
+/*
+ * include/asm-xtensa/page.h
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version2 as
+ * published by the Free Software Foundation.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PAGE_H
+#define _XTENSA_PAGE_H
+
+#include <linux/const.h>
+
+#include <asm/processor.h>
+#include <asm/types.h>
+#include <asm/cache.h>
+#include <asm/kmem_layout.h>
+
+#include <vdso/page.h>
+
+#ifdef CONFIG_MMU
+#define PAGE_OFFSET	XCHAL_KSEG_CACHED_VADDR
+#define PHYS_OFFSET	XCHAL_KSEG_PADDR
+#define MAX_LOW_PFN	(PHYS_PFN(XCHAL_KSEG_PADDR) + \
+			 PHYS_PFN(XCHAL_KSEG_SIZE))
+#else
+#define PAGE_OFFSET	_AC(CONFIG_DEFAULT_MEM_START, UL)
+#define PHYS_OFFSET	_AC(CONFIG_DEFAULT_MEM_START, UL)
+#define MAX_LOW_PFN	PHYS_PFN(0xfffffffful)
+#endif
+
+/*
+ * Cache aliasing:
+ *
+ * If the cache size for one way is greater than the page size, we have to
+ * deal with cache aliasing. The cache index is wider than the page size:
+ *
+ * |    |cache| cache index
+ * | pfn  |off|	virtual address
+ * |xxxx:X|zzz|
+ * |    : |   |
+ * | \  / |   |
+ * |trans.|   |
+ * | /  \ |   |
+ * |yyyy:Y|zzz|	physical address
+ *
+ * When the page number is translated to the physical page address, the lowest
+ * bit(s) (X) that are part of the cache index are also translated (Y).
+ * If this translation changes bit(s) (X), the cache index is also afected,
+ * thus resulting in a different cache line than before.
+ * The kernel does not provide a mechanism to ensure that the page color
+ * (represented by this bit) remains the same when allocated or when pages
+ * are remapped. When user pages are mapped into kernel space, the color of
+ * the page might also change.
+ *
+ * We use the address space VMALLOC_END ... VMALLOC_END + DCACHE_WAY_SIZE * 2
+ * to temporarily map a patch so we can match the color.
+ */
+
+#if DCACHE_WAY_SIZE > PAGE_SIZE
+# define DCACHE_ALIAS_ORDER	(DCACHE_WAY_SHIFT - PAGE_SHIFT)
+# define DCACHE_ALIAS_MASK	(PAGE_MASK & (DCACHE_WAY_SIZE - 1))
+# define DCACHE_ALIAS(a)	(((a) & DCACHE_ALIAS_MASK) >> PAGE_SHIFT)
+# define DCACHE_ALIAS_EQ(a,b)	((((a) ^ (b)) & DCACHE_ALIAS_MASK) == 0)
+#else
+# define DCACHE_ALIAS_ORDER	0
+# define DCACHE_ALIAS(a)	((void)(a), 0)
+#endif
+#define DCACHE_N_COLORS		(1 << DCACHE_ALIAS_ORDER)
+
+#if ICACHE_WAY_SIZE > PAGE_SIZE
+# define ICACHE_ALIAS_ORDER	(ICACHE_WAY_SHIFT - PAGE_SHIFT)
+# define ICACHE_ALIAS_MASK	(PAGE_MASK & (ICACHE_WAY_SIZE - 1))
+# define ICACHE_ALIAS(a)	(((a) & ICACHE_ALIAS_MASK) >> PAGE_SHIFT)
+# define ICACHE_ALIAS_EQ(a,b)	((((a) ^ (b)) & ICACHE_ALIAS_MASK) == 0)
+#else
+# define ICACHE_ALIAS_ORDER	0
+#endif
+
+
+#ifdef __ASSEMBLER__
+
+#define __pgprot(x)	(x)
+
+#else
+
+/*
+ * These are used to make use of C type-checking..
+ */
+
+typedef struct { unsigned long pte; } pte_t;		/* page table entry */
+typedef struct { unsigned long pgd; } pgd_t;		/* PGD table entry */
+typedef struct { unsigned long pgprot; } pgprot_t;
+typedef struct page *pgtable_t;
+
+#define pte_val(x)	((x).pte)
+#define pgd_val(x)	((x).pgd)
+#define pgprot_val(x)	((x).pgprot)
+
+#define __pte(x)	((pte_t) { (x) } )
+#define __pgd(x)	((pgd_t) { (x) } )
+#define __pgprot(x)	((pgprot_t) { (x) } )
+
+# include <asm-generic/getorder.h>
+
+struct page;
+struct vm_area_struct;
+extern void clear_page(void *page);
+extern void copy_page(void *to, void *from);
+
+/*
+ * If we have cache aliasing and writeback caches, we might have to do
+ * some extra work
+ */
+
+#if defined(CONFIG_MMU) && DCACHE_WAY_SIZE > PAGE_SIZE
+extern void clear_page_alias(void *vaddr, unsigned long paddr);
+extern void copy_page_alias(void *to, void *from,
+			    unsigned long to_paddr, unsigned long from_paddr);
+
+#define clear_user_highpage clear_user_highpage
+void clear_user_highpage(struct page *page, unsigned long vaddr);
+#define __HAVE_ARCH_COPY_USER_HIGHPAGE
+void copy_user_highpage(struct page *to, struct page *from,
+			unsigned long vaddr, struct vm_area_struct *vma);
+#else
+# define copy_user_page(to, from, vaddr, pg)	copy_page(to, from)
+#endif
+
+/*
+ * This handles the memory map.  We handle pages at
+ * XCHAL_KSEG_CACHED_VADDR for kernels with 32 bit address space.
+ * These macros are for conversion of kernel address, not user
+ * addresses.
+ */
+
+#define ARCH_PFN_OFFSET		(PHYS_OFFSET >> PAGE_SHIFT)
+
+#ifdef CONFIG_MMU
+static inline unsigned long ___pa(unsigned long va)
+{
+	unsigned long off = va - PAGE_OFFSET;
+
+	if (off >= XCHAL_KSEG_SIZE)
+		off -= XCHAL_KSEG_SIZE;
+
+#ifndef CONFIG_XIP_KERNEL
+	return off + PHYS_OFFSET;
+#else
+	if (off < XCHAL_KSEG_SIZE)
+		return off + PHYS_OFFSET;
+
+	off -= XCHAL_KSEG_SIZE;
+	if (off >= XCHAL_KIO_SIZE)
+		off -= XCHAL_KIO_SIZE;
+
+	return off + XCHAL_KIO_PADDR;
+#endif
+}
+#define __pa(x)	___pa((unsigned long)(x))
+#else
+#define __pa(x)	\
+	((unsigned long) (x) - PAGE_OFFSET + PHYS_OFFSET)
+#endif
+#define __va(x)	\
+	((void *)((unsigned long) (x) - PHYS_OFFSET + PAGE_OFFSET))
+
+#define virt_to_page(kaddr)	pfn_to_page(__pa(kaddr) >> PAGE_SHIFT)
+#define page_to_virt(page)	__va(page_to_pfn(page) << PAGE_SHIFT)
+#define virt_addr_valid(kaddr)	pfn_valid(__pa(kaddr) >> PAGE_SHIFT)
+
+#endif /* __ASSEMBLER__ */
+
+#include <asm-generic/memory_model.h>
+#endif /* _XTENSA_PAGE_H */
diff --git a/arch/xtensa/include/asm/pci-bridge.h b/arch/xtensa/include/asm/pci-bridge.h
new file mode 100644
index 000000000..e320aa5bb
--- /dev/null
+++ b/arch/xtensa/include/asm/pci-bridge.h
@@ -0,0 +1,76 @@
+/*
+ * include/asm-xtensa/pci-bridge.h
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PCI_BRIDGE_H
+#define _XTENSA_PCI_BRIDGE_H
+
+struct device_node;
+struct pci_controller;
+
+/*
+ * pciauto_bus_scan() enumerates the pci space.
+ */
+
+extern int pciauto_bus_scan(struct pci_controller *, int);
+
+struct pci_space {
+	unsigned long start;
+	unsigned long end;
+	unsigned long base;
+};
+
+/*
+ * Structure of a PCI controller (host bridge)
+ */
+
+struct pci_controller {
+	int index;			/* used for pci_controller_num */
+	struct pci_controller *next;
+	struct pci_bus *bus;
+	void *arch_data;
+
+	int first_busno;
+	int last_busno;
+
+	struct pci_ops *ops;
+	volatile unsigned int *cfg_addr;
+	volatile unsigned char *cfg_data;
+
+	/* Currently, we limit ourselves to 1 IO range and 3 mem
+	 * ranges since the common pci_bus structure can't handle more
+	 */
+	struct resource	io_resource;
+	struct resource mem_resources[3];
+	int mem_resource_count;
+
+	/* Host bridge I/O and Memory space
+	 * Used for BAR placement algorithms
+	 */
+	struct pci_space io_space;
+	struct pci_space mem_space;
+
+	/* Return the interrupt number fo a device. */
+	int (*map_irq)(struct pci_dev*, u8, u8);
+
+};
+
+static inline void pcibios_init_resource(struct resource *res,
+		unsigned long start, unsigned long end, int flags, char *name)
+{
+	res->start = start;
+	res->end = end;
+	res->flags = flags;
+	res->name = name;
+	res->parent = NULL;
+	res->sibling = NULL;
+	res->child = NULL;
+}
+
+#endif	/* _XTENSA_PCI_BRIDGE_H */
diff --git a/arch/xtensa/include/asm/pci.h b/arch/xtensa/include/asm/pci.h
new file mode 100644
index 000000000..b56de9635
--- /dev/null
+++ b/arch/xtensa/include/asm/pci.h
@@ -0,0 +1,46 @@
+/*
+ * linux/include/asm-xtensa/pci.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PCI_H
+#define _XTENSA_PCI_H
+
+/* Can be used to override the logic in pci_scan_bus for skipping
+ * already-configured bus numbers - to be used for buggy BIOSes
+ * or architectures with incomplete PCI setup by the loader
+ */
+
+#define pcibios_assign_all_busses()	0
+
+/* Assume some values. (We should revise them, if necessary) */
+
+#define PCIBIOS_MIN_IO		0x2000
+#define PCIBIOS_MIN_MEM		0x10000000
+
+/* Dynamic DMA mapping stuff.
+ * Xtensa has everything mapped statically like x86.
+ */
+
+#include <linux/types.h>
+#include <linux/slab.h>
+#include <linux/scatterlist.h>
+#include <linux/string.h>
+#include <asm/io.h>
+
+/* The PCI address space does equal the physical memory address space.
+ * The networking and block device layers use this boolean for bounce buffer
+ * decisions.
+ */
+
+/* Tell PCI code what kind of PCI resource mappings we support */
+#define HAVE_PCI_MMAP			1
+#define ARCH_GENERIC_PCI_MMAP_RESOURCE	1
+#define arch_can_pci_mmap_io()		1
+
+#endif	/* _XTENSA_PCI_H */
diff --git a/arch/xtensa/include/asm/perf_event.h b/arch/xtensa/include/asm/perf_event.h
new file mode 100644
index 000000000..5aa4590ac
--- /dev/null
+++ b/arch/xtensa/include/asm/perf_event.h
@@ -0,0 +1,4 @@
+#ifndef __ASM_XTENSA_PERF_EVENT_H
+#define __ASM_XTENSA_PERF_EVENT_H
+
+#endif /* __ASM_XTENSA_PERF_EVENT_H */
diff --git a/arch/xtensa/include/asm/pgalloc.h b/arch/xtensa/include/asm/pgalloc.h
new file mode 100644
index 000000000..1919ee9c3
--- /dev/null
+++ b/arch/xtensa/include/asm/pgalloc.h
@@ -0,0 +1,67 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * include/asm-xtensa/pgalloc.h
+ *
+ * Copyright (C) 2001-2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PGALLOC_H
+#define _XTENSA_PGALLOC_H
+
+#ifdef CONFIG_MMU
+#include <linux/highmem.h>
+#include <linux/slab.h>
+
+#define __HAVE_ARCH_PTE_ALLOC_ONE_KERNEL
+#define __HAVE_ARCH_PTE_ALLOC_ONE
+#include <asm-generic/pgalloc.h>
+
+/*
+ * Allocating and freeing a pmd is trivial: the 1-entry pmd is
+ * inside the pgd, so has no extra memory associated with it.
+ */
+
+#define pmd_populate_kernel(mm, pmdp, ptep)				     \
+	(pmd_val(*(pmdp)) = ((unsigned long)ptep))
+#define pmd_populate(mm, pmdp, page)					     \
+	(pmd_val(*(pmdp)) = ((unsigned long)page_to_virt(page)))
+
+static inline pgd_t*
+pgd_alloc(struct mm_struct *mm)
+{
+	return __pgd_alloc(mm, 0);
+}
+
+static inline void ptes_clear(pte_t *ptep)
+{
+	int i;
+
+	for (i = 0; i < PTRS_PER_PTE; i++)
+		pte_clear(NULL, 0, ptep + i);
+}
+
+static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm)
+{
+	pte_t *ptep;
+
+	ptep = (pte_t *)__pte_alloc_one_kernel(mm);
+	if (!ptep)
+		return NULL;
+	ptes_clear(ptep);
+	return ptep;
+}
+
+static inline pgtable_t pte_alloc_one(struct mm_struct *mm)
+{
+	struct page *page;
+
+	page = __pte_alloc_one(mm, GFP_PGTABLE_USER);
+	if (!page)
+		return NULL;
+	ptes_clear(page_address(page));
+	return page;
+}
+
+#endif /* CONFIG_MMU */
+
+#endif /* _XTENSA_PGALLOC_H */
diff --git a/arch/xtensa/include/asm/pgtable.h b/arch/xtensa/include/asm/pgtable.h
new file mode 100644
index 000000000..f00a879dc
--- /dev/null
+++ b/arch/xtensa/include/asm/pgtable.h
@@ -0,0 +1,417 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * include/asm-xtensa/pgtable.h
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PGTABLE_H
+#define _XTENSA_PGTABLE_H
+
+#include <asm/page.h>
+#include <asm/kmem_layout.h>
+#include <asm-generic/pgtable-nopmd.h>
+
+/*
+ * We only use two ring levels, user and kernel space.
+ */
+
+#ifdef CONFIG_MMU
+#define USER_RING		1	/* user ring level */
+#else
+#define USER_RING		0
+#endif
+#define KERNEL_RING		0	/* kernel ring level */
+
+/*
+ * The Xtensa architecture port of Linux has a two-level page table system,
+ * i.e. the logical three-level Linux page table layout is folded.
+ * Each task has the following memory page tables:
+ *
+ *   PGD table (page directory), ie. 3rd-level page table:
+ *	One page (4 kB) of 1024 (PTRS_PER_PGD) pointers to PTE tables
+ *	(Architectures that don't have the PMD folded point to the PMD tables)
+ *
+ *	The pointer to the PGD table for a given task can be retrieved from
+ *	the task structure (struct task_struct*) t, e.g. current():
+ *	  (t->mm ? t->mm : t->active_mm)->pgd
+ *
+ *   PMD tables (page middle-directory), ie. 2nd-level page tables:
+ *	Absent for the Xtensa architecture (folded, PTRS_PER_PMD == 1).
+ *
+ *   PTE tables (page table entry), ie. 1st-level page tables:
+ *	One page (4 kB) of 1024 (PTRS_PER_PTE) PTEs with a special PTE
+ *	invalid_pte_table for absent mappings.
+ *
+ * The individual pages are 4 kB big with special pages for the empty_zero_page.
+ */
+
+#define PGDIR_SHIFT	22
+#define PGDIR_SIZE	(1UL << PGDIR_SHIFT)
+#define PGDIR_MASK	(~(PGDIR_SIZE-1))
+
+/*
+ * Entries per page directory level: we use two-level, so
+ * we don't really have any PMD directory physically.
+ */
+#define PTRS_PER_PTE		1024
+#define PTRS_PER_PTE_SHIFT	10
+#define PTRS_PER_PGD		1024
+#define USER_PTRS_PER_PGD	(TASK_SIZE/PGDIR_SIZE)
+
+#ifdef CONFIG_MMU
+/*
+ * Virtual memory area. We keep a distance to other memory regions to be
+ * on the safe side. We also use this area for cache aliasing.
+ */
+#define VMALLOC_START		(XCHAL_KSEG_CACHED_VADDR - 0x10000000)
+#define VMALLOC_END		(VMALLOC_START + 0x07FEFFFF)
+#define TLBTEMP_BASE_1		(VMALLOC_START + 0x08000000)
+#define TLBTEMP_BASE_2		(TLBTEMP_BASE_1 + DCACHE_WAY_SIZE)
+#if 2 * DCACHE_WAY_SIZE > ICACHE_WAY_SIZE
+#define TLBTEMP_SIZE		(2 * DCACHE_WAY_SIZE)
+#else
+#define TLBTEMP_SIZE		ICACHE_WAY_SIZE
+#endif
+
+#else
+
+#define VMALLOC_START		__XTENSA_UL_CONST(0)
+#define VMALLOC_END		__XTENSA_UL_CONST(0xffffffff)
+
+#endif
+
+/*
+ * For the Xtensa architecture, the PTE layout is as follows:
+ *
+ *		31------12  11  10-9   8-6  5-4  3-2  1-0
+ *		+-----------------------------------------+
+ *		|           |   Software   |   HARDWARE   |
+ *		|    PPN    |          ADW | RI |Attribute|
+ *		+-----------------------------------------+
+ *   pte_none	|             MBZ          | 01 | 11 | 00 |
+ *		+-----------------------------------------+
+ *   present	|    PPN    | 0 | 00 | ADW | RI | CA | wx |
+ *		+- - - - - - - - - - - - - - - - - - - - -+
+ *   (PAGE_NONE)|    PPN    | 0 | 00 | ADW | 01 | 11 | 11 |
+ *		+-----------------------------------------+
+ *   swap	|     index     |   type   | 01 | 11 | e0 |
+ *		+-----------------------------------------+
+ *
+ * For T1050 hardware and earlier the layout differs for present and (PAGE_NONE)
+ *		+-----------------------------------------+
+ *   present	|    PPN    | 0 | 00 | ADW | RI | CA | w1 |
+ *		+-----------------------------------------+
+ *   (PAGE_NONE)|    PPN    | 0 | 00 | ADW | 01 | 01 | 00 |
+ *		+-----------------------------------------+
+ *
+ *  Legend:
+ *   PPN        Physical Page Number
+ *   ADW	software: accessed (young) / dirty / writable
+ *   RI         ring (0=privileged, 1=user, 2 and 3 are unused)
+ *   CA		cache attribute: 00 bypass, 01 writeback, 10 writethrough
+ *		(11 is invalid and used to mark pages that are not present)
+ *   e		exclusive marker in swap PTEs
+ *   w		page is writable (hw)
+ *   x		page is executable (hw)
+ *   index      swap offset / PAGE_SIZE (bit 11-31: 21 bits -> 8 GB)
+ *		(note that the index is always non-zero)
+ *   type       swap type (5 bits -> 32 types)
+ *
+ *  Notes:
+ *   - (PROT_NONE) is a special case of 'present' but causes an exception for
+ *     any access (read, write, and execute).
+ *   - 'multihit-exception' has the highest priority of all MMU exceptions,
+ *     so the ring must be set to 'RING_USER' even for 'non-present' pages.
+ *   - on older hardware, the exectuable flag was not supported and
+ *     used as a 'valid' flag, so it needs to be always set.
+ *   - we need to keep track of certain flags in software (dirty and young)
+ *     to do this, we use write exceptions and have a separate software w-flag.
+ *   - attribute value 1101 (and 1111 on T1050 and earlier) is reserved
+ */
+
+#define _PAGE_ATTRIB_MASK	0xf
+
+#define _PAGE_HW_EXEC		(1<<0)	/* hardware: page is executable */
+#define _PAGE_HW_WRITE		(1<<1)	/* hardware: page is writable */
+
+#define _PAGE_CA_BYPASS		(0<<2)	/* bypass, non-speculative */
+#define _PAGE_CA_WB		(1<<2)	/* write-back */
+#define _PAGE_CA_WT		(2<<2)	/* write-through */
+#define _PAGE_CA_MASK		(3<<2)
+#define _PAGE_CA_INVALID	(3<<2)
+
+/* We use invalid attribute values to distinguish special pte entries */
+#if XCHAL_HW_VERSION_MAJOR < 2000
+#define _PAGE_HW_VALID		0x01	/* older HW needed this bit set */
+#define _PAGE_NONE		0x04
+#else
+#define _PAGE_HW_VALID		0x00
+#define _PAGE_NONE		0x0f
+#endif
+
+#define _PAGE_USER		(1<<4)	/* user access (ring=1) */
+
+/* Software */
+#define _PAGE_WRITABLE_BIT	6
+#define _PAGE_WRITABLE		(1<<6)	/* software: page writable */
+#define _PAGE_DIRTY		(1<<7)	/* software: page dirty */
+#define _PAGE_ACCESSED		(1<<8)	/* software: page accessed (read) */
+
+/* We borrow bit 1 to store the exclusive marker in swap PTEs. */
+#define _PAGE_SWP_EXCLUSIVE	(1<<1)
+
+#ifdef CONFIG_MMU
+
+#define _PAGE_CHG_MASK	   (PAGE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY)
+#define _PAGE_PRESENT	   (_PAGE_HW_VALID | _PAGE_CA_WB | _PAGE_ACCESSED)
+
+#define PAGE_NONE	   __pgprot(_PAGE_NONE | _PAGE_USER)
+#define PAGE_COPY	   __pgprot(_PAGE_PRESENT | _PAGE_USER)
+#define PAGE_COPY_EXEC	   __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_HW_EXEC)
+#define PAGE_READONLY	   __pgprot(_PAGE_PRESENT | _PAGE_USER)
+#define PAGE_READONLY_EXEC __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_HW_EXEC)
+#define PAGE_SHARED	   __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_WRITABLE)
+#define PAGE_SHARED_EXEC \
+	__pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_WRITABLE | _PAGE_HW_EXEC)
+#define PAGE_KERNEL	   __pgprot(_PAGE_PRESENT | _PAGE_HW_WRITE)
+#define PAGE_KERNEL_RO	   __pgprot(_PAGE_PRESENT)
+#define PAGE_KERNEL_EXEC   __pgprot(_PAGE_PRESENT|_PAGE_HW_WRITE|_PAGE_HW_EXEC)
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+# define _PAGE_DIRECTORY   (_PAGE_HW_VALID | _PAGE_ACCESSED | _PAGE_CA_BYPASS)
+#else
+# define _PAGE_DIRECTORY   (_PAGE_HW_VALID | _PAGE_ACCESSED | _PAGE_CA_WB)
+#endif
+
+#else /* no mmu */
+
+# define _PAGE_CHG_MASK  (PAGE_MASK | _PAGE_ACCESSED | _PAGE_DIRTY)
+# define PAGE_NONE       __pgprot(0)
+# define PAGE_SHARED     __pgprot(0)
+# define PAGE_COPY       __pgprot(0)
+# define PAGE_READONLY   __pgprot(0)
+# define PAGE_KERNEL     __pgprot(0)
+
+#endif
+
+/*
+ * On certain configurations of Xtensa MMUs (eg. the initial Linux config),
+ * the MMU can't do page protection for execute, and considers that the same as
+ * read.  Also, write permissions may imply read permissions.
+ * What follows is the closest we can get by reasonable means..
+ * See linux/mm/mmap.c for protection_map[] array that uses these definitions.
+ */
+#ifndef __ASSEMBLER__
+
+#define pte_ERROR(e) \
+	printk("%s:%d: bad pte %08lx.\n", __FILE__, __LINE__, pte_val(e))
+#define pgd_ERROR(e) \
+	printk("%s:%d: bad pgd entry %08lx.\n", __FILE__, __LINE__, pgd_val(e))
+
+#ifdef CONFIG_MMU
+extern pgd_t swapper_pg_dir[PAGE_SIZE/sizeof(pgd_t)];
+extern void paging_init(void);
+#else
+# define swapper_pg_dir NULL
+static inline void paging_init(void) { }
+#endif
+
+/*
+ * The pmd contains the kernel virtual address of the pte page.
+ */
+#define pmd_page_vaddr(pmd) ((unsigned long)(pmd_val(pmd) & PAGE_MASK))
+#define pmd_pfn(pmd) (__pa(pmd_val(pmd)) >> PAGE_SHIFT)
+#define pmd_page(pmd) virt_to_page(pmd_val(pmd))
+
+/*
+ * pte status.
+ */
+# define pte_none(pte)	 (pte_val(pte) == (_PAGE_CA_INVALID | _PAGE_USER))
+#if XCHAL_HW_VERSION_MAJOR < 2000
+# define pte_present(pte) ((pte_val(pte) & _PAGE_CA_MASK) != _PAGE_CA_INVALID)
+#else
+# define pte_present(pte)						\
+	(((pte_val(pte) & _PAGE_CA_MASK) != _PAGE_CA_INVALID)		\
+	 || ((pte_val(pte) & _PAGE_ATTRIB_MASK) == _PAGE_NONE))
+#endif
+#define pte_clear(mm,addr,ptep)						\
+	do { update_pte(ptep, __pte(_PAGE_CA_INVALID | _PAGE_USER)); } while (0)
+
+#define pmd_none(pmd)	 (!pmd_val(pmd))
+#define pmd_present(pmd) (pmd_val(pmd) & PAGE_MASK)
+#define pmd_bad(pmd)	 (pmd_val(pmd) & ~PAGE_MASK)
+#define pmd_clear(pmdp)	 do { set_pmd(pmdp, __pmd(0)); } while (0)
+
+static inline int pte_write(pte_t pte) { return pte_val(pte) & _PAGE_WRITABLE; }
+static inline int pte_dirty(pte_t pte) { return pte_val(pte) & _PAGE_DIRTY; }
+static inline int pte_young(pte_t pte) { return pte_val(pte) & _PAGE_ACCESSED; }
+
+static inline pte_t pte_wrprotect(pte_t pte)
+	{ pte_val(pte) &= ~(_PAGE_WRITABLE | _PAGE_HW_WRITE); return pte; }
+static inline pte_t pte_mkclean(pte_t pte)
+	{ pte_val(pte) &= ~(_PAGE_DIRTY | _PAGE_HW_WRITE); return pte; }
+static inline pte_t pte_mkold(pte_t pte)
+	{ pte_val(pte) &= ~_PAGE_ACCESSED; return pte; }
+static inline pte_t pte_mkdirty(pte_t pte)
+	{ pte_val(pte) |= _PAGE_DIRTY; return pte; }
+static inline pte_t pte_mkyoung(pte_t pte)
+	{ pte_val(pte) |= _PAGE_ACCESSED; return pte; }
+static inline pte_t pte_mkwrite_novma(pte_t pte)
+	{ pte_val(pte) |= _PAGE_WRITABLE; return pte; }
+
+#define pgprot_noncached(prot) \
+		((__pgprot((pgprot_val(prot) & ~_PAGE_CA_MASK) | \
+			   _PAGE_CA_BYPASS)))
+
+#define PFN_PTE_SHIFT		PAGE_SHIFT
+#define pte_pfn(pte)		(pte_val(pte) >> PAGE_SHIFT)
+#define pte_same(a,b)		(pte_val(a) == pte_val(b))
+#define pte_page(x)		pfn_to_page(pte_pfn(x))
+#define pfn_pte(pfn, prot)	__pte(((pfn) << PAGE_SHIFT) | pgprot_val(prot))
+
+static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
+{
+	return __pte((pte_val(pte) & _PAGE_CHG_MASK) | pgprot_val(newprot));
+}
+
+/*
+ * Certain architectures need to do special things when pte's
+ * within a page table are directly modified.  Thus, the following
+ * hook is made available.
+ */
+static inline void update_pte(pte_t *ptep, pte_t pteval)
+{
+	*ptep = pteval;
+#if (DCACHE_WAY_SIZE > PAGE_SIZE) && XCHAL_DCACHE_IS_WRITEBACK
+	__asm__ __volatile__ ("dhwb %0, 0" :: "a" (ptep));
+#endif
+
+}
+
+struct mm_struct;
+
+static inline void set_pte(pte_t *ptep, pte_t pte)
+{
+	update_pte(ptep, pte);
+}
+
+static inline void
+set_pmd(pmd_t *pmdp, pmd_t pmdval)
+{
+	*pmdp = pmdval;
+}
+
+struct vm_area_struct;
+
+static inline bool ptep_test_and_clear_young(struct vm_area_struct *vma,
+		unsigned long addr, pte_t *ptep)
+{
+	pte_t pte = *ptep;
+	if (!pte_young(pte))
+		return false;
+	update_pte(ptep, pte_mkold(pte));
+	return true;
+}
+
+static inline pte_t
+ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+	pte_t pte = *ptep;
+	pte_clear(mm, addr, ptep);
+	return pte;
+}
+
+static inline void
+ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep)
+{
+	pte_t pte = *ptep;
+	update_pte(ptep, pte_wrprotect(pte));
+}
+
+/*
+ * Encode/decode swap entries and swap PTEs. Swap PTEs are all PTEs that
+ * are !pte_none() && !pte_present().
+ */
+#define MAX_SWAPFILES_CHECK() BUILD_BUG_ON(MAX_SWAPFILES_SHIFT > 5)
+
+#define __swp_type(entry)	(((entry).val >> 6) & 0x1f)
+#define __swp_offset(entry)	((entry).val >> 11)
+#define __swp_entry(type,offs)	\
+	((swp_entry_t){(((type) & 0x1f) << 6) | ((offs) << 11) | \
+	 _PAGE_CA_INVALID | _PAGE_USER})
+#define __pte_to_swp_entry(pte)	((swp_entry_t) { pte_val(pte) })
+#define __swp_entry_to_pte(x)	((pte_t) { (x).val })
+
+static inline bool pte_swp_exclusive(pte_t pte)
+{
+	return pte_val(pte) & _PAGE_SWP_EXCLUSIVE;
+}
+
+static inline pte_t pte_swp_mkexclusive(pte_t pte)
+{
+	pte_val(pte) |= _PAGE_SWP_EXCLUSIVE;
+	return pte;
+}
+
+static inline pte_t pte_swp_clear_exclusive(pte_t pte)
+{
+	pte_val(pte) &= ~_PAGE_SWP_EXCLUSIVE;
+	return pte;
+}
+
+#endif /*  !defined (__ASSEMBLER__) */
+
+
+#ifdef __ASSEMBLER__
+
+/* Assembly macro _PGD_INDEX is the same as C pgd_index(unsigned long),
+ *                _PGD_OFFSET as C pgd_offset(struct mm_struct*, unsigned long),
+ *                _PMD_OFFSET as C pmd_offset(pgd_t*, unsigned long)
+ *                _PTE_OFFSET as C pte_offset(pmd_t*, unsigned long)
+ *
+ * Note: We require an additional temporary register which can be the same as
+ *       the register that holds the address.
+ *
+ * ((pte_t*) ((unsigned long)(pmd_val(*pmd) & PAGE_MASK)) + pte_index(addr))
+ *
+ */
+#define _PGD_INDEX(rt,rs)	extui	rt, rs, PGDIR_SHIFT, 32-PGDIR_SHIFT
+#define _PTE_INDEX(rt,rs)	extui	rt, rs, PAGE_SHIFT, PTRS_PER_PTE_SHIFT
+
+#define _PGD_OFFSET(mm,adr,tmp)		l32i	mm, mm, MM_PGD;		\
+					_PGD_INDEX(tmp, adr);		\
+					addx4	mm, tmp, mm
+
+#define _PTE_OFFSET(pmd,adr,tmp)	_PTE_INDEX(tmp, adr);		\
+					srli	pmd, pmd, PAGE_SHIFT;	\
+					slli	pmd, pmd, PAGE_SHIFT;	\
+					addx4	pmd, tmp, pmd
+
+#else
+
+struct vm_fault;
+void update_mmu_cache_range(struct vm_fault *vmf, struct vm_area_struct *vma,
+		unsigned long address, pte_t *ptep, unsigned int nr);
+#define update_mmu_cache(vma, address, ptep) \
+	update_mmu_cache_range(NULL, vma, address, ptep, 1)
+
+typedef pte_t *pte_addr_t;
+
+void update_mmu_tlb_range(struct vm_area_struct *vma,
+		unsigned long address, pte_t *ptep, unsigned int nr);
+#define update_mmu_tlb_range update_mmu_tlb_range
+
+#endif /* !defined (__ASSEMBLER__) */
+
+#define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG
+#define __HAVE_ARCH_PTEP_GET_AND_CLEAR
+#define __HAVE_ARCH_PTEP_SET_WRPROTECT
+#define __HAVE_ARCH_PTEP_MKDIRTY
+#define __HAVE_ARCH_PTE_SAME
+/* We provide our own get_unmapped_area to cope with
+ * SHM area cache aliasing for userland.
+ */
+#define HAVE_ARCH_UNMAPPED_AREA
+
+#endif /* _XTENSA_PGTABLE_H */
diff --git a/arch/xtensa/include/asm/platform.h b/arch/xtensa/include/asm/platform.h
new file mode 100644
index 000000000..f2e50fc12
--- /dev/null
+++ b/arch/xtensa/include/asm/platform.h
@@ -0,0 +1,46 @@
+/*
+ * Platform specific functions
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_PLATFORM_H
+#define _XTENSA_PLATFORM_H
+
+#include <linux/types.h>
+#include <asm/bootparam.h>
+
+/*
+ * platform_init is called before the mmu is initialized to give the
+ * platform a early hook-up. bp_tag_t is a list of configuration tags
+ * passed from the boot-loader.
+ */
+extern void platform_init(bp_tag_t*);
+
+/*
+ * platform_setup is called from setup_arch with a pointer to the command-line
+ * string.
+ */
+extern void platform_setup (char **);
+
+/*
+ * platform_idle is called from the idle function.
+ */
+extern void platform_idle (void);
+
+/*
+ * platform_calibrate_ccount calibrates cpu clock freq (CONFIG_XTENSA_CALIBRATE_CCOUNT)
+ */
+extern void platform_calibrate_ccount (void);
+
+/*
+ * Flush and reset the mmu, simulate a processor reset, and
+ * jump to the reset vector.
+ */
+void cpu_reset(void) __attribute__((noreturn));
+
+#endif	/* _XTENSA_PLATFORM_H */
diff --git a/arch/xtensa/include/asm/processor.h b/arch/xtensa/include/asm/processor.h
new file mode 100644
index 000000000..60a211356
--- /dev/null
+++ b/arch/xtensa/include/asm/processor.h
@@ -0,0 +1,267 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2008 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_PROCESSOR_H
+#define _XTENSA_PROCESSOR_H
+
+#include <asm/core.h>
+
+#include <linux/compiler.h>
+#include <linux/stringify.h>
+
+#include <asm/bootparam.h>
+#include <asm/ptrace.h>
+#include <asm/types.h>
+#include <asm/regs.h>
+
+#define ARCH_SLAB_MINALIGN XTENSA_STACK_ALIGNMENT
+
+/*
+ * User space process size: 1 GB.
+ * Windowed call ABI requires caller and callee to be located within the same
+ * 1 GB region. The C compiler places trampoline code on the stack for sources
+ * that take the address of a nested C function (a feature used by glibc), so
+ * the 1 GB requirement applies to the stack as well.
+ */
+
+#ifdef CONFIG_MMU
+#define TASK_SIZE	__XTENSA_UL_CONST(0x40000000)
+#else
+#define TASK_SIZE	__XTENSA_UL_CONST(0xffffffff)
+#endif
+
+#define STACK_TOP	TASK_SIZE
+#define STACK_TOP_MAX	STACK_TOP
+
+/*
+ * General exception cause assigned to fake NMI. Fake NMI needs to be handled
+ * differently from other interrupts, but it uses common kernel entry/exit
+ * code.
+ */
+
+#define EXCCAUSE_MAPPED_NMI	62
+
+/*
+ * General exception cause assigned to debug exceptions. Debug exceptions go
+ * to their own vector, rather than the general exception vectors (user,
+ * kernel, double); and their specific causes are reported via DEBUGCAUSE
+ * rather than EXCCAUSE.  However it is sometimes convenient to redirect debug
+ * exceptions to the general exception mechanism.  To do this, an otherwise
+ * unused EXCCAUSE value was assigned to debug exceptions for this purpose.
+ */
+
+#define EXCCAUSE_MAPPED_DEBUG	63
+
+/*
+ * We use DEPC also as a flag to distinguish between double and regular
+ * exceptions. For performance reasons, DEPC might contain the value of
+ * EXCCAUSE for regular exceptions, so we use this definition to mark a
+ * valid double exception address.
+ * (Note: We use it in bgeui, so it should be 64, 128, or 256)
+ */
+
+#define VALID_DOUBLE_EXCEPTION_ADDRESS	64
+
+#define XTENSA_INT_LEVEL(intno) _XTENSA_INT_LEVEL(intno)
+#define _XTENSA_INT_LEVEL(intno) XCHAL_INT##intno##_LEVEL
+
+#define XTENSA_INTLEVEL_MASK(level) _XTENSA_INTLEVEL_MASK(level)
+#define _XTENSA_INTLEVEL_MASK(level) (XCHAL_INTLEVEL##level##_MASK)
+
+#define XTENSA_INTLEVEL_ANDBELOW_MASK(l) _XTENSA_INTLEVEL_ANDBELOW_MASK(l)
+#define _XTENSA_INTLEVEL_ANDBELOW_MASK(l) (XCHAL_INTLEVEL##l##_ANDBELOW_MASK)
+
+#define PROFILING_INTLEVEL XTENSA_INT_LEVEL(XCHAL_PROFILING_INTERRUPT)
+
+/* LOCKLEVEL defines the interrupt level that masks all
+ * general-purpose interrupts.
+ */
+#if defined(CONFIG_XTENSA_FAKE_NMI) && defined(XCHAL_PROFILING_INTERRUPT)
+#define LOCKLEVEL (PROFILING_INTLEVEL - 1)
+#else
+#define LOCKLEVEL XCHAL_EXCM_LEVEL
+#endif
+
+#define TOPLEVEL XCHAL_EXCM_LEVEL
+#define XTENSA_FAKE_NMI (LOCKLEVEL < TOPLEVEL)
+
+/* WSBITS and WBBITS are the width of the WINDOWSTART and WINDOWBASE
+ * registers
+ */
+#define WSBITS  (XCHAL_NUM_AREGS / 4)      /* width of WINDOWSTART in bits */
+#define WBBITS  (XCHAL_NUM_AREGS_LOG2 - 2) /* width of WINDOWBASE in bits */
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+#define KERNEL_PS_WOE_MASK PS_WOE_MASK
+#elif defined(__XTENSA_CALL0_ABI__)
+#define KERNEL_PS_WOE_MASK 0
+#else
+#error Unsupported xtensa ABI
+#endif
+
+#ifndef __ASSEMBLER__
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+
+/* Build a valid return address for the specified call winsize.
+ * winsize must be 1 (call4), 2 (call8), or 3 (call12)
+ */
+#define MAKE_RA_FOR_CALL(ra,ws)   (((ra) & 0x3fffffff) | (ws) << 30)
+
+/* Convert return address to a valid pc
+ * Note: 'text' is the address within the same 1GB range as the ra
+ */
+#define MAKE_PC_FROM_RA(ra, text) (((ra) & 0x3fffffff) | ((unsigned long)(text) & 0xc0000000))
+
+#elif defined(__XTENSA_CALL0_ABI__)
+
+/* Build a valid return address for the specified call winsize.
+ * winsize must be 1 (call4), 2 (call8), or 3 (call12)
+ */
+#define MAKE_RA_FOR_CALL(ra, ws)   (ra)
+
+/* Convert return address to a valid pc
+ * Note: 'text' is not used as 'ra' is always the full address
+ */
+#define MAKE_PC_FROM_RA(ra, text)  (ra)
+
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+/* Spill slot location for the register reg in the spill area under the stack
+ * pointer sp. reg must be in the range [0..4).
+ */
+#define SPILL_SLOT(sp, reg) (*(((unsigned long *)(sp)) - 4 + (reg)))
+
+/* Spill slot location for the register reg in the spill area under the stack
+ * pointer sp for the call8. reg must be in the range [4..8).
+ */
+#define SPILL_SLOT_CALL8(sp, reg) (*(((unsigned long *)(sp)) - 12 + (reg)))
+
+/* Spill slot location for the register reg in the spill area under the stack
+ * pointer sp for the call12. reg must be in the range [4..12).
+ */
+#define SPILL_SLOT_CALL12(sp, reg) (*(((unsigned long *)(sp)) - 16 + (reg)))
+
+struct thread_struct {
+
+	/* kernel's return address and stack pointer for context switching */
+	unsigned long ra; /* kernel's a0: return address and window call size */
+	unsigned long sp; /* kernel's a1: stack pointer */
+
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	struct perf_event *ptrace_bp[XCHAL_NUM_IBREAK];
+	struct perf_event *ptrace_wp[XCHAL_NUM_DBREAK];
+#endif
+} __aligned(16);
+
+/* This decides where the kernel will search for a free chunk of vm
+ * space during mmap's.
+ */
+#define TASK_UNMAPPED_BASE	(TASK_SIZE / 2)
+
+#define INIT_THREAD  \
+{									\
+	ra:		0, 						\
+	sp:		sizeof(init_stack) + (long) &init_stack,	\
+}
+
+
+/*
+ * Do necessary setup to start up a newly executed thread.
+ * Note: When windowed ABI is used for userspace we set-up ps
+ *       as if we did a call4 to the new pc.
+ *       set_thread_state in signal.c depends on it.
+ */
+#if IS_ENABLED(CONFIG_USER_ABI_CALL0)
+#define USER_PS_VALUE ((USER_RING << PS_RING_SHIFT) |			\
+		       (1 << PS_UM_BIT) |				\
+		       (1 << PS_EXCM_BIT))
+#else
+#define USER_PS_VALUE (PS_WOE_MASK |					\
+		       (1 << PS_CALLINC_SHIFT) |			\
+		       (USER_RING << PS_RING_SHIFT) |			\
+		       (1 << PS_UM_BIT) |				\
+		       (1 << PS_EXCM_BIT))
+#endif
+
+/* Clearing a0 terminates the backtrace. */
+#define start_thread(regs, new_pc, new_sp) \
+	do { \
+		unsigned long syscall = (regs)->syscall; \
+		unsigned long current_aregs[16]; \
+		memcpy(current_aregs, (regs)->areg, sizeof(current_aregs)); \
+		memset((regs), 0, sizeof(*(regs))); \
+		(regs)->pc = (new_pc); \
+		(regs)->ps = USER_PS_VALUE; \
+		memcpy((regs)->areg, current_aregs, sizeof(current_aregs)); \
+		(regs)->areg[1] = (new_sp); \
+		(regs)->areg[0] = 0; \
+		(regs)->wmask = 1; \
+		(regs)->depc = 0; \
+		(regs)->windowbase = 0; \
+		(regs)->windowstart = 1; \
+		(regs)->syscall = syscall; \
+	} while (0)
+
+/* Forward declaration */
+struct task_struct;
+struct mm_struct;
+
+extern unsigned long __get_wchan(struct task_struct *p);
+
+void init_arch(bp_tag_t *bp_start);
+void do_notify_resume(struct pt_regs *regs);
+
+#define KSTK_EIP(tsk)		(task_pt_regs(tsk)->pc)
+#define KSTK_ESP(tsk)		(task_pt_regs(tsk)->areg[1])
+
+#define cpu_relax()  barrier()
+
+/* Special register access. */
+
+#define xtensa_set_sr(x, sr) \
+	({ \
+	 __asm__ __volatile__ ("wsr %0, "__stringify(sr) :: \
+			       "a"((unsigned int)(x))); \
+	 })
+
+#define xtensa_get_sr(sr) \
+	({ \
+	 unsigned int v; \
+	 __asm__ __volatile__ ("rsr %0, "__stringify(sr) : "=a"(v)); \
+	 v; \
+	 })
+
+#define xtensa_xsr(x, sr) \
+	({ \
+	 unsigned int __v__ = (unsigned int)(x); \
+	 __asm__ __volatile__ ("xsr %0, " __stringify(sr) : "+a"(__v__)); \
+	 __v__; \
+	 })
+
+#if XCHAL_HAVE_EXTERN_REGS
+
+static inline void set_er(unsigned long value, unsigned long addr)
+{
+	asm volatile ("wer %0, %1" : : "a" (value), "a" (addr) : "memory");
+}
+
+static inline unsigned long get_er(unsigned long addr)
+{
+	register unsigned long value;
+	asm volatile ("rer %0, %1" : "=a" (value) : "a" (addr) : "memory");
+	return value;
+}
+
+#endif /* XCHAL_HAVE_EXTERN_REGS */
+
+#endif	/* __ASSEMBLER__ */
+#endif	/* _XTENSA_PROCESSOR_H */
diff --git a/arch/xtensa/include/asm/ptrace.h b/arch/xtensa/include/asm/ptrace.h
new file mode 100644
index 000000000..d0568ff6d
--- /dev/null
+++ b/arch/xtensa/include/asm/ptrace.h
@@ -0,0 +1,116 @@
+/*
+ * include/asm-xtensa/ptrace.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+#ifndef _XTENSA_PTRACE_H
+#define _XTENSA_PTRACE_H
+
+#include <asm/kmem_layout.h>
+#include <uapi/asm/ptrace.h>
+
+/*
+ * Kernel stack
+ *
+ *		+-----------------------+  -------- STACK_SIZE
+ *		|     register file     |  |
+ *		+-----------------------+  |
+ *		|    struct pt_regs     |  |
+ *		+-----------------------+  | ------ PT_REGS_OFFSET
+ * double	:  16 bytes spill area  :  |  ^
+ * excetion	:- - - - - - - - - - - -:  |  |
+ * frame	:    struct pt_regs     :  |  |
+ *		:- - - - - - - - - - - -:  |  |
+ *		|                       |  |  |
+ *		|     memory stack      |  |  |
+ *		|                       |  |  |
+ *		~                       ~  ~  ~
+ *		~                       ~  ~  ~
+ *		|                       |  |  |
+ *		|                       |  |  |
+ *		+-----------------------+  |  | --- STACK_BIAS
+ *		|  struct task_struct   |  |  |  ^
+ *  current --> +-----------------------+  |  |  |
+ *		|  struct thread_info   |  |  |  |
+ *		+-----------------------+ --------
+ */
+
+#define NO_SYSCALL (-1)
+
+#ifndef __ASSEMBLER__
+
+#include <asm/coprocessor.h>
+#include <asm/core.h>
+
+/*
+ * This struct defines the way the registers are stored on the
+ * kernel stack during a system call or other kernel entry.
+ */
+struct pt_regs {
+	unsigned long pc;		/*   4 */
+	unsigned long ps;		/*   8 */
+	unsigned long depc;		/*  12 */
+	unsigned long exccause;		/*  16 */
+	unsigned long excvaddr;		/*  20 */
+	unsigned long debugcause;	/*  24 */
+	unsigned long wmask;		/*  28 */
+	unsigned long lbeg;		/*  32 */
+	unsigned long lend;		/*  36 */
+	unsigned long lcount;		/*  40 */
+	unsigned long sar;		/*  44 */
+	unsigned long windowbase;	/*  48 */
+	unsigned long windowstart;	/*  52 */
+	unsigned long syscall;		/*  56 */
+	unsigned long icountlevel;	/*  60 */
+	unsigned long scompare1;	/*  64 */
+	unsigned long threadptr;	/*  68 */
+
+	/* Additional configurable registers that are used by the compiler. */
+	xtregs_opt_t xtregs_opt;
+
+	/* current register frame.
+	 * Note: The ESF for kernel exceptions ends after 16 registers!
+	 */
+	unsigned long areg[XCHAL_NUM_AREGS] __aligned(16);
+};
+
+# define arch_has_single_step()	(1)
+# define task_pt_regs(tsk) ((struct pt_regs*) \
+	(task_stack_page(tsk) + KERNEL_STACK_SIZE) - 1)
+# define user_mode(regs) (((regs)->ps & 0x00000020)!=0)
+# define instruction_pointer(regs) ((regs)->pc)
+# define return_pointer(regs) (MAKE_PC_FROM_RA((regs)->areg[0], \
+					       (regs)->pc))
+
+# ifndef CONFIG_SMP
+#  define profile_pc(regs) instruction_pointer(regs)
+# else
+#  define profile_pc(regs)						\
+	({								\
+		in_lock_functions(instruction_pointer(regs)) ?		\
+		return_pointer(regs) : instruction_pointer(regs);	\
+	})
+# endif
+
+#define user_stack_pointer(regs) ((regs)->areg[1])
+
+static inline unsigned long regs_return_value(struct pt_regs *regs)
+{
+	return regs->areg[2];
+}
+
+int do_syscall_trace_enter(struct pt_regs *regs);
+void do_syscall_trace_leave(struct pt_regs *regs);
+
+#else	/* __ASSEMBLER__ */
+
+# include <asm/asm-offsets.h>
+#define PT_REGS_OFFSET	  (KERNEL_STACK_SIZE - PT_USER_SIZE)
+
+#endif	/* !__ASSEMBLER__ */
+
+#endif	/* _XTENSA_PTRACE_H */
diff --git a/arch/xtensa/include/asm/regs.h b/arch/xtensa/include/asm/regs.h
new file mode 100644
index 000000000..ce184e7de
--- /dev/null
+++ b/arch/xtensa/include/asm/regs.h
@@ -0,0 +1,118 @@
+/*
+ * Copyright (c) 2006 Tensilica, Inc.  All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2.1 of the GNU Lesser General Public
+ * License as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * Further, this software is distributed without any warranty that it is
+ * free of the rightful claim of any third person regarding infringement
+ * or the like.  Any license provided herein, whether implied or
+ * otherwise, applies only to this software file.  Patent licenses, if
+ * any, provided herein do not apply to combinations of this program with
+ * other software, or any other product whatsoever.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with this program; if not, write the Free Software
+ * Foundation, Inc., 59 Temple Place - Suite 330, Boston MA 02111-1307,
+ * USA.
+ */
+
+#ifndef _XTENSA_REGS_H
+#define _XTENSA_REGS_H
+
+/*  Special registers.  */
+
+#define SREG_MR			32
+#define SREG_IBREAKENABLE	96
+#define SREG_IBREAKA		128
+#define SREG_DBREAKA		144
+#define SREG_DBREAKC		160
+#define SREG_EPC		176
+#define SREG_EPS		192
+#define SREG_EXCSAVE		208
+#define SREG_CCOMPARE		240
+#define SREG_MISC		244
+
+/*  EXCCAUSE register fields  */
+
+#define EXCCAUSE_EXCCAUSE_SHIFT	0
+#define EXCCAUSE_EXCCAUSE_MASK	0x3F
+
+#define EXCCAUSE_ILLEGAL_INSTRUCTION		0
+#define EXCCAUSE_SYSTEM_CALL			1
+#define EXCCAUSE_INSTRUCTION_FETCH_ERROR	2
+#define EXCCAUSE_LOAD_STORE_ERROR		3
+#define EXCCAUSE_LEVEL1_INTERRUPT		4
+#define EXCCAUSE_ALLOCA				5
+#define EXCCAUSE_INTEGER_DIVIDE_BY_ZERO		6
+#define EXCCAUSE_SPECULATION			7
+#define EXCCAUSE_PRIVILEGED			8
+#define EXCCAUSE_UNALIGNED			9
+#define EXCCAUSE_INSTR_DATA_ERROR		12
+#define EXCCAUSE_LOAD_STORE_DATA_ERROR		13
+#define EXCCAUSE_INSTR_ADDR_ERROR		14
+#define EXCCAUSE_LOAD_STORE_ADDR_ERROR		15
+#define EXCCAUSE_ITLB_MISS			16
+#define EXCCAUSE_ITLB_MULTIHIT			17
+#define EXCCAUSE_ITLB_PRIVILEGE			18
+#define EXCCAUSE_ITLB_SIZE_RESTRICTION		19
+#define EXCCAUSE_FETCH_CACHE_ATTRIBUTE		20
+#define EXCCAUSE_DTLB_MISS			24
+#define EXCCAUSE_DTLB_MULTIHIT			25
+#define EXCCAUSE_DTLB_PRIVILEGE			26
+#define EXCCAUSE_DTLB_SIZE_RESTRICTION		27
+#define EXCCAUSE_LOAD_CACHE_ATTRIBUTE		28
+#define EXCCAUSE_STORE_CACHE_ATTRIBUTE		29
+#define EXCCAUSE_COPROCESSOR0_DISABLED		32
+#define EXCCAUSE_COPROCESSOR1_DISABLED		33
+#define EXCCAUSE_COPROCESSOR2_DISABLED		34
+#define EXCCAUSE_COPROCESSOR3_DISABLED		35
+#define EXCCAUSE_COPROCESSOR4_DISABLED		36
+#define EXCCAUSE_COPROCESSOR5_DISABLED		37
+#define EXCCAUSE_COPROCESSOR6_DISABLED		38
+#define EXCCAUSE_COPROCESSOR7_DISABLED		39
+#define EXCCAUSE_N				64
+
+/*  PS register fields.  */
+
+#define PS_WOE_BIT		18
+#define PS_WOE_MASK		0x00040000
+#define PS_CALLINC_SHIFT	16
+#define PS_CALLINC_MASK		0x00030000
+#define PS_OWB_SHIFT		8
+#define PS_OWB_WIDTH		4
+#define PS_OWB_MASK		0x00000F00
+#define PS_RING_SHIFT		6
+#define PS_RING_MASK		0x000000C0
+#define PS_UM_BIT		5
+#define PS_EXCM_BIT		4
+#define PS_INTLEVEL_SHIFT	0
+#define PS_INTLEVEL_WIDTH	4
+#define PS_INTLEVEL_MASK	0x0000000F
+
+/*  DBREAKCn register fields.  */
+
+#define DBREAKC_MASK_BIT		0
+#define DBREAKC_MASK_MASK		0x0000003F
+#define DBREAKC_LOAD_BIT		30
+#define DBREAKC_LOAD_MASK		0x40000000
+#define DBREAKC_STOR_BIT		31
+#define DBREAKC_STOR_MASK		0x80000000
+
+/*  DEBUGCAUSE register fields.  */
+
+#define DEBUGCAUSE_DBNUM_MASK		0xf00
+#define DEBUGCAUSE_DBNUM_SHIFT		8	/* First bit of DBNUM field */
+#define DEBUGCAUSE_DEBUGINT_BIT		5	/* External debug interrupt */
+#define DEBUGCAUSE_BREAKN_BIT		4	/* BREAK.N instruction */
+#define DEBUGCAUSE_BREAK_BIT		3	/* BREAK instruction */
+#define DEBUGCAUSE_DBREAK_BIT		2	/* DBREAK match */
+#define DEBUGCAUSE_IBREAK_BIT		1	/* IBREAK match */
+#define DEBUGCAUSE_ICOUNT_BIT		0	/* ICOUNT would incr. to zero */
+
+#endif /* _XTENSA_SPECREG_H */
diff --git a/arch/xtensa/include/asm/seccomp.h b/arch/xtensa/include/asm/seccomp.h
new file mode 100644
index 000000000..f1cb6b0a9
--- /dev/null
+++ b/arch/xtensa/include/asm/seccomp.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef _ASM_SECCOMP_H
+#define _ASM_SECCOMP_H
+
+#include <asm-generic/seccomp.h>
+
+#define SECCOMP_ARCH_NATIVE		AUDIT_ARCH_XTENSA
+#define SECCOMP_ARCH_NATIVE_NR		NR_syscalls
+#define SECCOMP_ARCH_NATIVE_NAME	"xtensa"
+
+#endif /* _ASM_SECCOMP_H */
diff --git a/arch/xtensa/include/asm/sections.h b/arch/xtensa/include/asm/sections.h
new file mode 100644
index 000000000..e5da6d709
--- /dev/null
+++ b/arch/xtensa/include/asm/sections.h
@@ -0,0 +1,45 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _XTENSA_SECTIONS_H
+#define _XTENSA_SECTIONS_H
+
+#include <asm-generic/sections.h>
+
+#ifdef CONFIG_VECTORS_ADDR
+extern char _WindowVectors_text_start[];
+extern char _WindowVectors_text_end[];
+extern char _DebugInterruptVector_text_start[];
+extern char _DebugInterruptVector_text_end[];
+extern char _KernelExceptionVector_text_start[];
+extern char _KernelExceptionVector_text_end[];
+extern char _UserExceptionVector_text_start[];
+extern char _UserExceptionVector_text_end[];
+extern char _DoubleExceptionVector_text_start[];
+extern char _DoubleExceptionVector_text_end[];
+extern char _exception_text_start[];
+extern char _exception_text_end[];
+extern char _Level2InterruptVector_text_start[];
+extern char _Level2InterruptVector_text_end[];
+extern char _Level3InterruptVector_text_start[];
+extern char _Level3InterruptVector_text_end[];
+extern char _Level4InterruptVector_text_start[];
+extern char _Level4InterruptVector_text_end[];
+extern char _Level5InterruptVector_text_start[];
+extern char _Level5InterruptVector_text_end[];
+extern char _Level6InterruptVector_text_start[];
+extern char _Level6InterruptVector_text_end[];
+#endif
+#ifdef CONFIG_SECONDARY_RESET_VECTOR
+extern char _SecondaryResetVector_text_start[];
+extern char _SecondaryResetVector_text_end[];
+#endif
+#ifdef CONFIG_XIP_KERNEL
+#ifdef CONFIG_VECTORS_ADDR
+extern char _xip_text_start[];
+extern char _xip_text_end[];
+#endif
+extern char _xip_start[];
+extern char _xip_end[];
+#endif
+
+#endif
diff --git a/arch/xtensa/include/asm/serial.h b/arch/xtensa/include/asm/serial.h
new file mode 100644
index 000000000..a8a249326
--- /dev/null
+++ b/arch/xtensa/include/asm/serial.h
@@ -0,0 +1,18 @@
+/*
+ * include/asm-xtensa/serial.h
+ *
+ * Configuration details for 8250, 16450, 16550, etc. serial ports
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SERIAL_H
+#define _XTENSA_SERIAL_H
+
+#include <platform/serial.h>
+
+#endif	/* _XTENSA_SERIAL_H */
diff --git a/arch/xtensa/include/asm/shmparam.h b/arch/xtensa/include/asm/shmparam.h
new file mode 100644
index 000000000..c8cc16c3d
--- /dev/null
+++ b/arch/xtensa/include/asm/shmparam.h
@@ -0,0 +1,21 @@
+/*
+ * include/asm-xtensa/shmparam.h
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ */
+
+#ifndef _XTENSA_SHMPARAM_H
+#define _XTENSA_SHMPARAM_H
+
+/*
+ * Xtensa can have variable size caches, and if
+ * the size of single way is larger than the page size,
+ * then we have to start worrying about cache aliasing
+ * problems.
+ */
+
+#define SHMLBA	((PAGE_SIZE > DCACHE_WAY_SIZE)? PAGE_SIZE : DCACHE_WAY_SIZE)
+
+#endif /* _XTENSA_SHMPARAM_H */
diff --git a/arch/xtensa/include/asm/signal.h b/arch/xtensa/include/asm/signal.h
new file mode 100644
index 000000000..d301e6857
--- /dev/null
+++ b/arch/xtensa/include/asm/signal.h
@@ -0,0 +1,23 @@
+/*
+ * include/asm-xtensa/signal.h
+ *
+ * Swiped from SH.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+#ifndef _XTENSA_SIGNAL_H
+#define _XTENSA_SIGNAL_H
+
+#include <uapi/asm/signal.h>
+
+#ifndef __ASSEMBLER__
+#define __ARCH_HAS_SA_RESTORER
+
+#include <asm/sigcontext.h>
+
+#endif	/* __ASSEMBLER__ */
+#endif	/* _XTENSA_SIGNAL_H */
diff --git a/arch/xtensa/include/asm/smp.h b/arch/xtensa/include/asm/smp.h
new file mode 100644
index 000000000..e446e6fc4
--- /dev/null
+++ b/arch/xtensa/include/asm/smp.h
@@ -0,0 +1,44 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SMP_H
+#define _XTENSA_SMP_H
+
+#ifdef CONFIG_SMP
+
+#define raw_smp_processor_id()	(current_thread_info()->cpu)
+#define cpu_logical_map(cpu)	(cpu)
+
+struct start_info {
+	unsigned long stack;
+};
+extern struct start_info start_info;
+
+struct cpumask;
+void arch_send_call_function_ipi_mask(const struct cpumask *mask);
+void arch_send_call_function_single_ipi(int cpu);
+
+void secondary_start_kernel(void);
+void smp_init_cpus(void);
+void secondary_init_irq(void);
+void ipi_init(void);
+struct seq_file;
+void show_ipi_list(struct seq_file *p, int prec);
+
+#ifdef CONFIG_HOTPLUG_CPU
+
+void __cpu_die(unsigned int cpu);
+int __cpu_disable(void);
+void __noreturn cpu_die(void);
+void cpu_restart(void);
+
+#endif /* CONFIG_HOTPLUG_CPU */
+
+#endif /* CONFIG_SMP */
+
+#endif	/* _XTENSA_SMP_H */
diff --git a/arch/xtensa/include/asm/spinlock.h b/arch/xtensa/include/asm/spinlock.h
new file mode 100644
index 000000000..41c449ece
--- /dev/null
+++ b/arch/xtensa/include/asm/spinlock.h
@@ -0,0 +1,20 @@
+/*
+ * include/asm-xtensa/spinlock.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SPINLOCK_H
+#define _XTENSA_SPINLOCK_H
+
+#include <asm/barrier.h>
+#include <asm/qspinlock.h>
+#include <asm/qrwlock.h>
+
+#define smp_mb__after_spinlock()	smp_mb()
+
+#endif	/* _XTENSA_SPINLOCK_H */
diff --git a/arch/xtensa/include/asm/spinlock_types.h b/arch/xtensa/include/asm/spinlock_types.h
new file mode 100644
index 000000000..6baaeac62
--- /dev/null
+++ b/arch/xtensa/include/asm/spinlock_types.h
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __ASM_SPINLOCK_TYPES_H
+#define __ASM_SPINLOCK_TYPES_H
+
+#if !defined(__LINUX_SPINLOCK_TYPES_RAW_H) && !defined(__ASM_SPINLOCK_H)
+# error "Please do not include this file directly."
+#endif
+
+#include <asm-generic/qspinlock_types.h>
+#include <asm-generic/qrwlock_types.h>
+
+#endif
diff --git a/arch/xtensa/include/asm/stackprotector.h b/arch/xtensa/include/asm/stackprotector.h
new file mode 100644
index 000000000..dd10279a2
--- /dev/null
+++ b/arch/xtensa/include/asm/stackprotector.h
@@ -0,0 +1,33 @@
+/*
+ * GCC stack protector support.
+ *
+ * (This is directly adopted from the ARM implementation)
+ *
+ * Stack protector works by putting predefined pattern at the start of
+ * the stack frame and verifying that it hasn't been overwritten when
+ * returning from the function.  The pattern is called stack canary
+ * and gcc expects it to be defined by a global variable called
+ * "__stack_chk_guard" on Xtensa.  This unfortunately means that on SMP
+ * we cannot have a different canary value per task.
+ */
+
+#ifndef _ASM_STACKPROTECTOR_H
+#define _ASM_STACKPROTECTOR_H 1
+
+extern unsigned long __stack_chk_guard;
+
+/*
+ * Initialize the stackprotector canary value.
+ *
+ * NOTE: this must only be called from functions that never return,
+ * and it must always be inlined.
+ */
+static __always_inline void boot_init_stack_canary(void)
+{
+	unsigned long canary = get_random_canary();
+
+	current->stack_canary = canary;
+	__stack_chk_guard = current->stack_canary;
+}
+
+#endif	/* _ASM_STACKPROTECTOR_H */
diff --git a/arch/xtensa/include/asm/stacktrace.h b/arch/xtensa/include/asm/stacktrace.h
new file mode 100644
index 000000000..a85e785a6
--- /dev/null
+++ b/arch/xtensa/include/asm/stacktrace.h
@@ -0,0 +1,44 @@
+/*
+ * arch/xtensa/include/asm/stacktrace.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+#ifndef _XTENSA_STACKTRACE_H
+#define _XTENSA_STACKTRACE_H
+
+#include <linux/sched.h>
+
+struct stackframe {
+	unsigned long pc;
+	unsigned long sp;
+};
+
+static __always_inline unsigned long *stack_pointer(struct task_struct *task)
+{
+	unsigned long sp;
+
+	if (!task || task == current)
+		sp = current_stack_pointer;
+	else
+		sp = task->thread.sp;
+
+	return (unsigned long *)sp;
+}
+
+void walk_stackframe(unsigned long *sp,
+		int (*fn)(struct stackframe *frame, void *data),
+		void *data);
+
+void xtensa_backtrace_kernel(struct pt_regs *regs, unsigned int depth,
+			     int (*kfn)(struct stackframe *frame, void *data),
+			     int (*ufn)(struct stackframe *frame, void *data),
+			     void *data);
+void xtensa_backtrace_user(struct pt_regs *regs, unsigned int depth,
+			   int (*ufn)(struct stackframe *frame, void *data),
+			   void *data);
+
+#endif /* _XTENSA_STACKTRACE_H */
diff --git a/arch/xtensa/include/asm/string.h b/arch/xtensa/include/asm/string.h
new file mode 100644
index 000000000..e16cda026
--- /dev/null
+++ b/arch/xtensa/include/asm/string.h
@@ -0,0 +1,112 @@
+/*
+ * include/asm-xtensa/string.h
+ *
+ * These trivial string functions are considered part of the public domain.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+/* We should optimize these. See arch/xtensa/lib/strncpy_user.S */
+
+#ifndef _XTENSA_STRING_H
+#define _XTENSA_STRING_H
+
+#define __HAVE_ARCH_STRCPY
+static inline char *strcpy(char *__dest, const char *__src)
+{
+	register char *__xdest = __dest;
+	unsigned long __dummy;
+
+	__asm__ __volatile__("1:\n\t"
+		"l8ui	%2, %1, 0\n\t"
+		"s8i	%2, %0, 0\n\t"
+		"addi	%1, %1, 1\n\t"
+		"addi	%0, %0, 1\n\t"
+		"bnez	%2, 1b\n\t"
+		: "=r" (__dest), "=r" (__src), "=&r" (__dummy)
+		: "0" (__dest), "1" (__src)
+		: "memory");
+
+	return __xdest;
+}
+
+#define __HAVE_ARCH_STRCMP
+static inline int strcmp(const char *__cs, const char *__ct)
+{
+	register int __res;
+	unsigned long __dummy;
+
+	__asm__ __volatile__(
+		"1:\n\t"
+		"l8ui	%3, %1, 0\n\t"
+		"addi	%1, %1, 1\n\t"
+		"l8ui	%2, %0, 0\n\t"
+		"addi	%0, %0, 1\n\t"
+		"beqz	%2, 2f\n\t"
+		"beq	%2, %3, 1b\n"
+		"2:\n\t"
+		"sub	%2, %2, %3"
+		: "=r" (__cs), "=r" (__ct), "=&r" (__res), "=&r" (__dummy)
+		: "0" (__cs), "1" (__ct));
+
+	return __res;
+}
+
+#define __HAVE_ARCH_STRNCMP
+static inline int strncmp(const char *__cs, const char *__ct, size_t __n)
+{
+	register int __res;
+	unsigned long __dummy;
+
+	__asm__ __volatile__(
+		"mov	%2, %3\n"
+		"1:\n\t"
+		"beq	%0, %6, 2f\n\t"
+		"l8ui	%3, %1, 0\n\t"
+		"addi	%1, %1, 1\n\t"
+		"l8ui	%2, %0, 0\n\t"
+		"addi	%0, %0, 1\n\t"
+		"beqz	%2, 2f\n\t"
+		"beqz	%3, 2f\n\t"
+		"beq	%2, %3, 1b\n"
+		"2:\n\t"
+		"sub	%2, %2, %3"
+		: "=r" (__cs), "=r" (__ct), "=&r" (__res), "=&r" (__dummy)
+		: "0" (__cs), "1" (__ct), "r" ((uintptr_t)__cs+__n));
+
+	return __res;
+}
+
+#define __HAVE_ARCH_MEMSET
+extern void *memset(void *__s, int __c, size_t __count);
+extern void *__memset(void *__s, int __c, size_t __count);
+
+#define __HAVE_ARCH_MEMCPY
+extern void *memcpy(void *__to, __const__ void *__from, size_t __n);
+extern void *__memcpy(void *__to, __const__ void *__from, size_t __n);
+
+#define __HAVE_ARCH_MEMMOVE
+extern void *memmove(void *__dest, __const__ void *__src, size_t __n);
+extern void *__memmove(void *__dest, __const__ void *__src, size_t __n);
+
+#if defined(CONFIG_KASAN) && !defined(__SANITIZE_ADDRESS__)
+
+/*
+ * For files that are not instrumented (e.g. mm/slub.c) we
+ * should use not instrumented version of mem* functions.
+ */
+
+#define memcpy(dst, src, len) __memcpy(dst, src, len)
+#define memmove(dst, src, len) __memmove(dst, src, len)
+#define memset(s, c, n) __memset(s, c, n)
+
+#ifndef __NO_FORTIFY
+#define __NO_FORTIFY /* FORTIFY_SOURCE uses __builtin_memcpy, etc. */
+#endif
+#endif
+
+#endif	/* _XTENSA_STRING_H */
diff --git a/arch/xtensa/include/asm/switch_to.h b/arch/xtensa/include/asm/switch_to.h
new file mode 100644
index 000000000..6b73bf0eb
--- /dev/null
+++ b/arch/xtensa/include/asm/switch_to.h
@@ -0,0 +1,22 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SWITCH_TO_H
+#define _XTENSA_SWITCH_TO_H
+
+/* * switch_to(n) should switch tasks to task nr n, first
+ * checking that n isn't the current task, in which case it does nothing.
+ */
+extern void *_switch_to(void *last, void *next);
+
+#define switch_to(prev,next,last)		\
+do {						\
+	(last) = _switch_to(prev, next);	\
+} while(0)
+
+#endif /* _XTENSA_SWITCH_TO_H */
diff --git a/arch/xtensa/include/asm/syscall.h b/arch/xtensa/include/asm/syscall.h
new file mode 100644
index 000000000..bab7cdd96
--- /dev/null
+++ b/arch/xtensa/include/asm/syscall.h
@@ -0,0 +1,93 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ * Copyright (C) 2018 Cadence Design Systems Inc.
+ */
+
+#ifndef _ASM_SYSCALL_H
+#define _ASM_SYSCALL_H
+
+#include <linux/err.h>
+#include <asm/ptrace.h>
+#include <uapi/linux/audit.h>
+
+static inline int syscall_get_arch(struct task_struct *task)
+{
+	return AUDIT_ARCH_XTENSA;
+}
+
+typedef void (*syscall_t)(void);
+extern syscall_t sys_call_table[];
+
+static inline long syscall_get_nr(struct task_struct *task,
+				  struct pt_regs *regs)
+{
+	return regs->syscall;
+}
+
+static inline void syscall_set_nr(struct task_struct *task,
+				  struct pt_regs *regs,
+				  int nr)
+{
+	regs->syscall = nr;
+}
+
+static inline void syscall_rollback(struct task_struct *task,
+				    struct pt_regs *regs)
+{
+	/* Do nothing. */
+}
+
+static inline long syscall_get_error(struct task_struct *task,
+				     struct pt_regs *regs)
+{
+	/* 0 if syscall succeeded, otherwise -Errorcode */
+	return IS_ERR_VALUE(regs->areg[2]) ? regs->areg[2] : 0;
+}
+
+static inline long syscall_get_return_value(struct task_struct *task,
+					    struct pt_regs *regs)
+{
+	return regs->areg[2];
+}
+
+static inline void syscall_set_return_value(struct task_struct *task,
+					    struct pt_regs *regs,
+					    int error, long val)
+{
+	regs->areg[2] = (long) error ? error : val;
+}
+
+#define XTENSA_SYSCALL_ARGUMENT_REGS {6, 3, 4, 5, 8, 9}
+
+static inline void syscall_get_arguments(struct task_struct *task,
+					 struct pt_regs *regs,
+					 unsigned long *args)
+{
+	static const unsigned int reg[] = XTENSA_SYSCALL_ARGUMENT_REGS;
+	unsigned int i;
+
+	for (i = 0; i < 6; ++i)
+		args[i] = regs->areg[reg[i]];
+}
+
+static inline void syscall_set_arguments(struct task_struct *task,
+					 struct pt_regs *regs,
+					 const unsigned long *args)
+{
+	static const unsigned int reg[] = XTENSA_SYSCALL_ARGUMENT_REGS;
+	unsigned int i;
+
+	for (i = 0; i < 6; ++i)
+		regs->areg[reg[i]] = args[i];
+}
+
+asmlinkage long xtensa_rt_sigreturn(void);
+asmlinkage long xtensa_shmat(int, char __user *, int);
+asmlinkage long xtensa_fadvise64_64(int, int,
+				    unsigned long long, unsigned long long);
+
+#endif
diff --git a/arch/xtensa/include/asm/sysmem.h b/arch/xtensa/include/asm/sysmem.h
new file mode 100644
index 000000000..552cdfd85
--- /dev/null
+++ b/arch/xtensa/include/asm/sysmem.h
@@ -0,0 +1,19 @@
+/*
+ * sysmem-related prototypes.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2014 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_SYSMEM_H
+#define _XTENSA_SYSMEM_H
+
+#include <linux/memblock.h>
+
+void bootmem_init(void);
+void zones_init(void);
+
+#endif /* _XTENSA_SYSMEM_H */
diff --git a/arch/xtensa/include/asm/thread_info.h b/arch/xtensa/include/asm/thread_info.h
new file mode 100644
index 000000000..5b74dfc35
--- /dev/null
+++ b/arch/xtensa/include/asm/thread_info.h
@@ -0,0 +1,145 @@
+/*
+ * include/asm-xtensa/thread_info.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_THREAD_INFO_H
+#define _XTENSA_THREAD_INFO_H
+
+#include <linux/stringify.h>
+#include <asm/kmem_layout.h>
+
+#define CURRENT_SHIFT KERNEL_STACK_SHIFT
+
+#ifndef __ASSEMBLER__
+# include <asm/processor.h>
+#endif
+
+/*
+ * low level task data that entry.S needs immediate access to
+ * - this struct should fit entirely inside of one cache line
+ * - this struct shares the supervisor stack pages
+ * - if the contents of this structure are changed, the assembly constants
+ *   must also be changed
+ */
+
+#ifndef __ASSEMBLER__
+
+#if XTENSA_HAVE_COPROCESSORS
+
+typedef struct xtregs_coprocessor {
+	xtregs_cp0_t cp0;
+	xtregs_cp1_t cp1;
+	xtregs_cp2_t cp2;
+	xtregs_cp3_t cp3;
+	xtregs_cp4_t cp4;
+	xtregs_cp5_t cp5;
+	xtregs_cp6_t cp6;
+	xtregs_cp7_t cp7;
+} xtregs_coprocessor_t;
+
+#endif
+
+struct thread_info {
+	struct task_struct	*task;		/* main task structure */
+	unsigned long		flags;		/* low level flags */
+	unsigned long		status;		/* thread-synchronous flags */
+	__u32			cpu;		/* current CPU */
+	__s32			preempt_count;	/* 0 => preemptable,< 0 => BUG*/
+
+#if XCHAL_HAVE_EXCLUSIVE
+	/* result of the most recent exclusive store */
+	unsigned long		atomctl8;
+#endif
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+	/* Address where PS.WOE was enabled by the ABI probing code */
+	unsigned long		ps_woe_fix_addr;
+#endif
+
+	/*
+	 * If i-th bit is set then coprocessor state is loaded into the
+	 * coprocessor i on CPU cp_owner_cpu.
+	 */
+	unsigned long		cpenable;
+	u32			cp_owner_cpu;
+	/* Allocate storage for extra user states and coprocessor states. */
+#if XTENSA_HAVE_COPROCESSORS
+	xtregs_coprocessor_t	xtregs_cp;
+#endif
+	xtregs_user_t		xtregs_user;
+};
+
+#endif
+
+/*
+ * macros/functions for gaining access to the thread information structure
+ */
+
+#ifndef __ASSEMBLER__
+
+#define INIT_THREAD_INFO(tsk)			\
+{						\
+	.task		= &tsk,			\
+	.flags		= 0,			\
+	.cpu		= 0,			\
+	.preempt_count	= INIT_PREEMPT_COUNT,	\
+}
+
+/* how to get the thread information struct from C */
+static __always_inline struct thread_info *current_thread_info(void)
+{
+	struct thread_info *ti;
+	 __asm__("extui %0, a1, 0, "__stringify(CURRENT_SHIFT)"\n\t"
+	         "xor %0, a1, %0" : "=&r" (ti) : );
+	return ti;
+}
+
+#else /* !__ASSEMBLER__ */
+
+/* how to get the thread information struct from ASM */
+#define GET_THREAD_INFO(reg,sp) \
+	extui reg, sp, 0, CURRENT_SHIFT; \
+	xor   reg, sp, reg
+#endif
+
+
+/*
+ * thread information flags
+ * - these are process state flags that various assembly files may need to access
+ */
+#define TIF_SYSCALL_TRACE	0	/* syscall trace active */
+#define TIF_SIGPENDING		1	/* signal pending */
+#define TIF_NEED_RESCHED	2	/* rescheduling necessary */
+#define TIF_SINGLESTEP		3	/* restore singlestep on return to user mode */
+#define TIF_SYSCALL_TRACEPOINT	4	/* syscall tracepoint instrumentation */
+#define TIF_NOTIFY_SIGNAL	5	/* signal notifications exist */
+#define TIF_RESTORE_SIGMASK	6	/* restore signal mask in do_signal() */
+#define TIF_NOTIFY_RESUME	7	/* callback before returning to user */
+#define TIF_DB_DISABLED		8	/* debug trap disabled for syscall */
+#define TIF_SYSCALL_AUDIT	9	/* syscall auditing active */
+#define TIF_SECCOMP		10	/* secure computing */
+#define TIF_MEMDIE		11	/* is terminating due to OOM killer */
+
+#define _TIF_SYSCALL_TRACE	(1<<TIF_SYSCALL_TRACE)
+#define _TIF_SIGPENDING		(1<<TIF_SIGPENDING)
+#define _TIF_NEED_RESCHED	(1<<TIF_NEED_RESCHED)
+#define _TIF_SINGLESTEP		(1<<TIF_SINGLESTEP)
+#define _TIF_SYSCALL_TRACEPOINT	(1<<TIF_SYSCALL_TRACEPOINT)
+#define _TIF_NOTIFY_SIGNAL	(1<<TIF_NOTIFY_SIGNAL)
+#define _TIF_NOTIFY_RESUME	(1<<TIF_NOTIFY_RESUME)
+#define _TIF_SYSCALL_AUDIT	(1<<TIF_SYSCALL_AUDIT)
+#define _TIF_SECCOMP		(1<<TIF_SECCOMP)
+
+#define _TIF_WORK_MASK		(_TIF_SYSCALL_TRACE | _TIF_SINGLESTEP | \
+				 _TIF_SYSCALL_TRACEPOINT | \
+				 _TIF_SYSCALL_AUDIT | _TIF_SECCOMP)
+
+#define THREAD_SIZE KERNEL_STACK_SIZE
+#define THREAD_SIZE_ORDER (KERNEL_STACK_SHIFT - PAGE_SHIFT)
+
+#endif	/* _XTENSA_THREAD_INFO */
diff --git a/arch/xtensa/include/asm/timex.h b/arch/xtensa/include/asm/timex.h
new file mode 100644
index 000000000..3f2462f2d
--- /dev/null
+++ b/arch/xtensa/include/asm/timex.h
@@ -0,0 +1,60 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_TIMEX_H
+#define _XTENSA_TIMEX_H
+
+#include <asm/processor.h>
+
+#if XCHAL_NUM_TIMERS > 0 && \
+	XTENSA_INT_LEVEL(XCHAL_TIMER0_INTERRUPT) <= XCHAL_EXCM_LEVEL
+# define LINUX_TIMER     0
+# define LINUX_TIMER_INT XCHAL_TIMER0_INTERRUPT
+#elif XCHAL_NUM_TIMERS > 1 && \
+	XTENSA_INT_LEVEL(XCHAL_TIMER1_INTERRUPT) <= XCHAL_EXCM_LEVEL
+# define LINUX_TIMER     1
+# define LINUX_TIMER_INT XCHAL_TIMER1_INTERRUPT
+#elif XCHAL_NUM_TIMERS > 2 && \
+	XTENSA_INT_LEVEL(XCHAL_TIMER2_INTERRUPT) <= XCHAL_EXCM_LEVEL
+# define LINUX_TIMER     2
+# define LINUX_TIMER_INT XCHAL_TIMER2_INTERRUPT
+#else
+# error "Bad timer number for Linux configurations!"
+#endif
+
+extern unsigned long ccount_freq;
+
+void local_timer_setup(unsigned cpu);
+
+/*
+ * Register access.
+ */
+
+static inline unsigned long get_ccount (void)
+{
+	return xtensa_get_sr(ccount);
+}
+
+static inline void set_ccount (unsigned long ccount)
+{
+	xtensa_set_sr(ccount, ccount);
+}
+
+static inline unsigned long get_linux_timer (void)
+{
+	return xtensa_get_sr(SREG_CCOMPARE + LINUX_TIMER);
+}
+
+static inline void set_linux_timer (unsigned long ccompare)
+{
+	xtensa_set_sr(ccompare, SREG_CCOMPARE + LINUX_TIMER);
+}
+
+#include <asm-generic/timex.h>
+
+#endif	/* _XTENSA_TIMEX_H */
diff --git a/arch/xtensa/include/asm/tlb.h b/arch/xtensa/include/asm/tlb.h
new file mode 100644
index 000000000..8c3ceb427
--- /dev/null
+++ b/arch/xtensa/include/asm/tlb.h
@@ -0,0 +1,23 @@
+/*
+ * include/asm-xtensa/tlb.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_TLB_H
+#define _XTENSA_TLB_H
+
+#include <asm/cache.h>
+#include <asm/page.h>
+
+#include <asm-generic/tlb.h>
+
+#define __pte_free_tlb(tlb, pte, address)	pte_free((tlb)->mm, pte)
+
+void check_tlb_sanity(void);
+
+#endif	/* _XTENSA_TLB_H */
diff --git a/arch/xtensa/include/asm/tlbflush.h b/arch/xtensa/include/asm/tlbflush.h
new file mode 100644
index 000000000..3edaebeef
--- /dev/null
+++ b/arch/xtensa/include/asm/tlbflush.h
@@ -0,0 +1,205 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_TLBFLUSH_H
+#define _XTENSA_TLBFLUSH_H
+
+#include <linux/stringify.h>
+#include <asm/processor.h>
+
+#define DTLB_WAY_PGD	7
+
+#define ITLB_ARF_WAYS	4
+#define DTLB_ARF_WAYS	4
+
+#define ITLB_HIT_BIT	3
+#define DTLB_HIT_BIT	4
+
+#ifndef __ASSEMBLER__
+
+/* TLB flushing:
+ *
+ *  - flush_tlb_all() flushes all processes TLB entries
+ *  - flush_tlb_mm(mm) flushes the specified mm context TLB entries
+ *  - flush_tlb_page(vma, page) flushes a single page
+ *  - flush_tlb_range(vma, vmaddr, end) flushes a range of pages
+ */
+
+void local_flush_tlb_all(void);
+void local_flush_tlb_mm(struct mm_struct *mm);
+void local_flush_tlb_page(struct vm_area_struct *vma,
+		unsigned long page);
+void local_flush_tlb_range(struct vm_area_struct *vma,
+		unsigned long start, unsigned long end);
+void local_flush_tlb_kernel_range(unsigned long start, unsigned long end);
+
+#ifdef CONFIG_SMP
+
+void flush_tlb_all(void);
+void flush_tlb_mm(struct mm_struct *);
+void flush_tlb_page(struct vm_area_struct *, unsigned long);
+void flush_tlb_range(struct vm_area_struct *, unsigned long,
+		unsigned long);
+void flush_tlb_kernel_range(unsigned long start, unsigned long end);
+
+#else /* !CONFIG_SMP */
+
+#define flush_tlb_all()			   local_flush_tlb_all()
+#define flush_tlb_mm(mm)		   local_flush_tlb_mm(mm)
+#define flush_tlb_page(vma, page)	   local_flush_tlb_page(vma, page)
+#define flush_tlb_range(vma, vmaddr, end)  local_flush_tlb_range(vma, vmaddr, \
+								 end)
+#define flush_tlb_kernel_range(start, end) local_flush_tlb_kernel_range(start, \
+									end)
+
+#endif /* CONFIG_SMP */
+
+/* TLB operations. */
+
+static inline unsigned long itlb_probe(unsigned long addr)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("pitlb  %0, %1\n\t" : "=a" (tmp) : "a" (addr));
+	return tmp;
+}
+
+static inline unsigned long dtlb_probe(unsigned long addr)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("pdtlb  %0, %1\n\t" : "=a" (tmp) : "a" (addr));
+	return tmp;
+}
+
+static inline void invalidate_itlb_entry (unsigned long probe)
+{
+	__asm__ __volatile__("iitlb  %0; isync\n\t" : : "a" (probe));
+}
+
+static inline void invalidate_dtlb_entry (unsigned long probe)
+{
+	__asm__ __volatile__("idtlb  %0; dsync\n\t" : : "a" (probe));
+}
+
+/* Use the .._no_isync functions with caution.  Generally, these are
+ * handy for bulk invalidates followed by a single 'isync'.  The
+ * caller must follow up with an 'isync', which can be relatively
+ * expensive on some Xtensa implementations.
+ */
+static inline void invalidate_itlb_entry_no_isync (unsigned entry)
+{
+	/* Caller must follow up with 'isync'. */
+	__asm__ __volatile__ ("iitlb  %0\n" : : "a" (entry) );
+}
+
+static inline void invalidate_dtlb_entry_no_isync (unsigned entry)
+{
+	/* Caller must follow up with 'isync'. */
+	__asm__ __volatile__ ("idtlb  %0\n" : : "a" (entry) );
+}
+
+static inline void set_itlbcfg_register (unsigned long val)
+{
+	__asm__ __volatile__("wsr  %0, itlbcfg\n\t" "isync\n\t"
+			     : : "a" (val));
+}
+
+static inline void set_dtlbcfg_register (unsigned long val)
+{
+	__asm__ __volatile__("wsr  %0, dtlbcfg; dsync\n\t"
+	    		     : : "a" (val));
+}
+
+static inline void set_ptevaddr_register (unsigned long val)
+{
+	__asm__ __volatile__(" wsr  %0, ptevaddr; isync\n"
+			     : : "a" (val));
+}
+
+static inline unsigned long read_ptevaddr_register (void)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("rsr  %0, ptevaddr\n\t" : "=a" (tmp));
+	return tmp;
+}
+
+static inline void write_dtlb_entry (pte_t entry, int way)
+{
+	__asm__ __volatile__("wdtlb  %1, %0; dsync\n\t"
+			     : : "r" (way), "r" (entry) );
+}
+
+static inline void write_itlb_entry (pte_t entry, int way)
+{
+	__asm__ __volatile__("witlb  %1, %0; isync\n\t"
+	                     : : "r" (way), "r" (entry) );
+}
+
+static inline void invalidate_page_directory (void)
+{
+	invalidate_dtlb_entry (DTLB_WAY_PGD);
+	invalidate_dtlb_entry (DTLB_WAY_PGD+1);
+	invalidate_dtlb_entry (DTLB_WAY_PGD+2);
+}
+
+static inline void invalidate_itlb_mapping (unsigned address)
+{
+	unsigned long tlb_entry;
+	if (((tlb_entry = itlb_probe(address)) & (1 << ITLB_HIT_BIT)) != 0)
+		invalidate_itlb_entry(tlb_entry);
+}
+
+static inline void invalidate_dtlb_mapping (unsigned address)
+{
+	unsigned long tlb_entry;
+	if (((tlb_entry = dtlb_probe(address)) & (1 << DTLB_HIT_BIT)) != 0)
+		invalidate_dtlb_entry(tlb_entry);
+}
+
+/*
+ * DO NOT USE THESE FUNCTIONS.  These instructions aren't part of the Xtensa
+ * ISA and exist only for test purposes..
+ * You may find it helpful for MMU debugging, however.
+ *
+ * 'at' is the unmodified input register
+ * 'as' is the output register, as follows (specific to the Linux config):
+ *
+ *      as[31..12] contain the virtual address
+ *      as[11..08] are meaningless
+ *      as[07..00] contain the asid
+ */
+
+static inline unsigned long read_dtlb_virtual (int way)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("rdtlb0  %0, %1\n\t" : "=a" (tmp), "+a" (way));
+	return tmp;
+}
+
+static inline unsigned long read_dtlb_translation (int way)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("rdtlb1  %0, %1\n\t" : "=a" (tmp), "+a" (way));
+	return tmp;
+}
+
+static inline unsigned long read_itlb_virtual (int way)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("ritlb0  %0, %1\n\t" : "=a" (tmp), "+a" (way));
+	return tmp;
+}
+
+static inline unsigned long read_itlb_translation (int way)
+{
+	unsigned long tmp;
+	__asm__ __volatile__("ritlb1  %0, %1\n\t" : "=a" (tmp), "+a" (way));
+	return tmp;
+}
+
+#endif	/* __ASSEMBLER__ */
+#endif	/* _XTENSA_TLBFLUSH_H */
diff --git a/arch/xtensa/include/asm/traps.h b/arch/xtensa/include/asm/traps.h
new file mode 100644
index 000000000..212c3b9ff
--- /dev/null
+++ b/arch/xtensa/include/asm/traps.h
@@ -0,0 +1,145 @@
+/*
+ * arch/xtensa/include/asm/traps.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 Tensilica Inc.
+ */
+#ifndef _XTENSA_TRAPS_H
+#define _XTENSA_TRAPS_H
+
+#include <asm/ptrace.h>
+
+typedef void xtensa_exception_handler(struct pt_regs *regs);
+
+/*
+ * Per-CPU exception handling data structure.
+ * EXCSAVE1 points to it.
+ */
+struct exc_table {
+	/* Kernel Stack */
+	void *kstk;
+	/* Double exception save area for a0 */
+	unsigned long double_save;
+	/* Fixup handler */
+	void *fixup;
+	/* For passing a parameter to fixup */
+	void *fixup_param;
+#if XTENSA_HAVE_COPROCESSORS
+	/* Pointers to owner struct thread_info */
+	struct thread_info *coprocessor_owner[XCHAL_CP_MAX];
+#endif
+	/* Fast user exception handlers */
+	void *fast_user_handler[EXCCAUSE_N];
+	/* Fast kernel exception handlers */
+	void *fast_kernel_handler[EXCCAUSE_N];
+	/* Default C-Handlers */
+	xtensa_exception_handler *default_handler[EXCCAUSE_N];
+};
+
+DECLARE_PER_CPU(struct exc_table, exc_table);
+
+xtensa_exception_handler *
+__init trap_set_handler(int cause, xtensa_exception_handler *handler);
+
+asmlinkage void fast_illegal_instruction_user(void);
+asmlinkage void fast_syscall_user(void);
+asmlinkage void fast_alloca(void);
+asmlinkage void fast_load_store(void);
+asmlinkage void fast_unaligned(void);
+asmlinkage void fast_second_level_miss(void);
+asmlinkage void fast_store_prohibited(void);
+asmlinkage void fast_coprocessor(void);
+
+asmlinkage void kernel_exception(void);
+asmlinkage void user_exception(void);
+asmlinkage void system_call(struct pt_regs *regs);
+
+void do_IRQ(int hwirq, struct pt_regs *regs);
+void do_page_fault(struct pt_regs *regs);
+void do_unhandled(struct pt_regs *regs);
+
+/* Initialize minimal exc_table structure sufficient for basic paging */
+static inline void __init early_trap_init(void)
+{
+	static struct exc_table init_exc_table __initdata = {
+#ifdef CONFIG_XTENSA_LOAD_STORE
+		.fast_kernel_handler[EXCCAUSE_LOAD_STORE_ERROR] =
+			fast_load_store,
+#endif
+#ifdef CONFIG_MMU
+		.fast_kernel_handler[EXCCAUSE_DTLB_MISS] =
+			fast_second_level_miss,
+#endif
+	};
+	xtensa_set_sr(&init_exc_table, excsave1);
+}
+
+void secondary_trap_init(void);
+
+static inline void spill_registers(void)
+{
+#if defined(__XTENSA_WINDOWED_ABI__)
+#if XCHAL_NUM_AREGS > 16
+	__asm__ __volatile__ (
+		"	call8	1f\n"
+		"	_j	2f\n"
+		"	retw\n"
+		"	.align	4\n"
+		"1:\n"
+#if XCHAL_NUM_AREGS == 32
+		"	_entry	a1, 32\n"
+		"	addi	a8, a0, 3\n"
+		"	_entry	a1, 16\n"
+		"	mov	a12, a12\n"
+		"	retw\n"
+#else
+		"	_entry	a1, 48\n"
+		"	call12	1f\n"
+		"	retw\n"
+		"	.align	4\n"
+		"1:\n"
+		"	.rept	(" __stringify(XCHAL_NUM_AREGS) " - 16) / 12\n"
+		"	_entry	a1, 48\n"
+		"	mov	a12, a0\n"
+		"	.endr\n"
+		"	_entry	a1, 16\n"
+#if XCHAL_NUM_AREGS % 12 == 0
+		"	mov	a12, a12\n"
+#elif XCHAL_NUM_AREGS % 12 == 4
+		"	mov	a4, a4\n"
+#elif XCHAL_NUM_AREGS % 12 == 8
+		"	mov	a8, a8\n"
+#endif
+		"	retw\n"
+#endif
+		"2:\n"
+		: : : "a8", "a9", "memory");
+#else
+	__asm__ __volatile__ (
+		"	mov	a12, a12\n"
+		: : : "memory");
+#endif
+#endif
+}
+
+struct debug_table {
+	/* Pointer to debug exception handler */
+	void (*debug_exception)(void);
+	/* Temporary register save area */
+	unsigned long debug_save[1];
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	/* Save area for DBREAKC registers */
+	unsigned long dbreakc_save[XCHAL_NUM_DBREAK];
+	/* Saved ICOUNT register */
+	unsigned long icount_save;
+	/* Saved ICOUNTLEVEL register */
+	unsigned long icount_level_save;
+#endif
+};
+
+void debug_exception(void);
+
+#endif /* _XTENSA_TRAPS_H */
diff --git a/arch/xtensa/include/asm/uaccess.h b/arch/xtensa/include/asm/uaccess.h
new file mode 100644
index 000000000..6538a29a2
--- /dev/null
+++ b/arch/xtensa/include/asm/uaccess.h
@@ -0,0 +1,294 @@
+/*
+ * include/asm-xtensa/uaccess.h
+ *
+ * User space memory access functions
+ *
+ * These routines provide basic accessing functions to the user memory
+ * space for the kernel. This header file provides functions such as:
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_UACCESS_H
+#define _XTENSA_UACCESS_H
+
+#include <linux/prefetch.h>
+#include <asm/types.h>
+#include <asm/extable.h>
+#include <asm-generic/access_ok.h>
+
+/*
+ * These are the main single-value transfer routines.  They
+ * automatically use the right size if we just have the right pointer
+ * type.
+ *
+ * This gets kind of ugly. We want to return _two_ values in
+ * "get_user()" and yet we don't want to do any pointers, because that
+ * is too much of a performance impact. Thus we have a few rather ugly
+ * macros here, and hide all the uglyness from the user.
+ *
+ * Careful to not
+ * (a) re-use the arguments for side effects (sizeof is ok)
+ * (b) require any knowledge of processes at this stage
+ */
+#define put_user(x, ptr)	__put_user_check((x), (ptr), sizeof(*(ptr)))
+#define get_user(x, ptr) __get_user_check((x), (ptr), sizeof(*(ptr)))
+
+/*
+ * The "__xxx" versions of the user access functions are versions that
+ * do not verify the address space, that must have been done previously
+ * with a separate "access_ok()" call (this is used when we do multiple
+ * accesses to the same area of user memory).
+ */
+#define __put_user(x, ptr) __put_user_nocheck((x), (ptr), sizeof(*(ptr)))
+#define __get_user(x, ptr) __get_user_nocheck((x), (ptr), sizeof(*(ptr)))
+
+
+extern long __put_user_bad(void);
+
+#define __put_user_nocheck(x, ptr, size)		\
+({							\
+	long __pu_err;					\
+	__put_user_size((x), (ptr), (size), __pu_err);	\
+	__pu_err;					\
+})
+
+#define __put_user_check(x, ptr, size)					\
+({									\
+	long __pu_err = -EFAULT;					\
+	__typeof__(*(ptr)) __user *__pu_addr = (ptr);			\
+	if (access_ok(__pu_addr, size))			\
+		__put_user_size((x), __pu_addr, (size), __pu_err);	\
+	__pu_err;							\
+})
+
+#define __put_user_size(x, ptr, size, retval)				\
+do {									\
+	int __cb;							\
+	retval = 0;							\
+	switch (size) {							\
+	case 1: __put_user_asm(x, ptr, retval, 1, "s8i", __cb);  break;	\
+	case 2: __put_user_asm(x, ptr, retval, 2, "s16i", __cb); break;	\
+	case 4: __put_user_asm(x, ptr, retval, 4, "s32i", __cb); break;	\
+	case 8: {							\
+		     __typeof__(*ptr) __v64 = x;			\
+		     retval = __copy_to_user(ptr, &__v64, 8) ? -EFAULT : 0;	\
+		     break;						\
+	        }							\
+	default: __put_user_bad();					\
+	}								\
+} while (0)
+
+
+/*
+ * Consider a case of a user single load/store would cause both an
+ * unaligned exception and an MMU-related exception (unaligned
+ * exceptions happen first):
+ *
+ * User code passes a bad variable ptr to a system call.
+ * Kernel tries to access the variable.
+ * Unaligned exception occurs.
+ * Unaligned exception handler tries to make aligned accesses.
+ * Double exception occurs for MMU-related cause (e.g., page not mapped).
+ * do_page_fault() thinks the fault address belongs to the kernel, not the
+ * user, and panics.
+ *
+ * The kernel currently prohibits user unaligned accesses.  We use the
+ * __check_align_* macros to check for unaligned addresses before
+ * accessing user space so we don't crash the kernel.  Both
+ * __put_user_asm and __get_user_asm use these alignment macros, so
+ * macro-specific labels such as 0f, 1f, %0, %2, and %3 must stay in
+ * sync.
+ */
+
+#define __check_align_1  ""
+
+#define __check_align_2				\
+	"   _bbci.l %[mem] * 0, 1f	\n"	\
+	"   movi    %[err], %[efault]	\n"	\
+	"   _j      2f			\n"
+
+#define __check_align_4				\
+	"   _bbsi.l %[mem] * 0, 0f	\n"	\
+	"   _bbci.l %[mem] * 0 + 1, 1f	\n"	\
+	"0: movi    %[err], %[efault]	\n"	\
+	"   _j      2f			\n"
+
+
+/*
+ * We don't tell gcc that we are accessing memory, but this is OK
+ * because we do not write to any memory gcc knows about, so there
+ * are no aliasing issues.
+ *
+ * WARNING: If you modify this macro at all, verify that the
+ * __check_align_* macros still work.
+ */
+#define __put_user_asm(x_, addr_, err_, align, insn, cb)\
+__asm__ __volatile__(					\
+	__check_align_##align				\
+	"1: "insn"  %[x], %[mem]	\n"		\
+	"2:				\n"		\
+	"   .section  .fixup,\"ax\"	\n"		\
+	"   .align 4			\n"		\
+	"   .literal_position		\n"		\
+	"5:				\n"		\
+	"   movi   %[tmp], 2b		\n"		\
+	"   movi   %[err], %[efault]	\n"		\
+	"   jx     %[tmp]		\n"		\
+	"   .previous			\n"		\
+	"   .section  __ex_table,\"a\"	\n"		\
+	"   .long	1b, 5b		\n"		\
+	"   .previous"					\
+	:[err] "+r"(err_), [tmp] "=r"(cb), [mem] "=m"(*(addr_))		\
+	:[x] "r"(x_), [efault] "i"(-EFAULT))
+
+#define __get_user_nocheck(x, ptr, size)			\
+({								\
+	long __gu_err;						\
+	__get_user_size((x), (ptr), (size), __gu_err);		\
+	__gu_err;						\
+})
+
+#define __get_user_check(x, ptr, size)					\
+({									\
+	long __gu_err = -EFAULT;					\
+	const __typeof__(*(ptr)) __user *__gu_addr = (ptr);		\
+	if (access_ok(__gu_addr, size))					\
+		__get_user_size((x), __gu_addr, (size), __gu_err);	\
+	else								\
+		(x) = (__typeof__(*(ptr)))0;				\
+	__gu_err;							\
+})
+
+extern long __get_user_bad(void);
+
+#define __get_user_size(x, ptr, size, retval)				\
+do {									\
+	int __cb;							\
+	retval = 0;							\
+	switch (size) {							\
+	case 1: __get_user_asm(x, ptr, retval, 1, "l8ui", __cb);  break;\
+	case 2: __get_user_asm(x, ptr, retval, 2, "l16ui", __cb); break;\
+	case 4: __get_user_asm(x, ptr, retval, 4, "l32i", __cb);  break;\
+	case 8: {							\
+		u64 __x;						\
+		if (unlikely(__copy_from_user(&__x, ptr, 8))) {		\
+			retval = -EFAULT;				\
+			(x) = (__typeof__(*(ptr)))0;			\
+		} else {						\
+			(x) = *(__force __typeof__(*(ptr)) *)&__x;	\
+		}							\
+		break;							\
+	}								\
+	default:							\
+		(x) = (__typeof__(*(ptr)))0;				\
+		__get_user_bad();					\
+	}								\
+} while (0)
+
+
+/*
+ * WARNING: If you modify this macro at all, verify that the
+ * __check_align_* macros still work.
+ */
+#define __get_user_asm(x_, addr_, err_, align, insn, cb) \
+do {							\
+	u32 __x = 0;					\
+	__asm__ __volatile__(				\
+		__check_align_##align			\
+		"1: "insn"  %[x], %[mem]	\n"	\
+		"2:				\n"	\
+		"   .section  .fixup,\"ax\"	\n"	\
+		"   .align 4			\n"	\
+		"   .literal_position		\n"	\
+		"5:				\n"	\
+		"   movi   %[tmp], 2b		\n"	\
+		"   movi   %[err], %[efault]	\n"	\
+		"   jx     %[tmp]		\n"	\
+		"   .previous			\n"	\
+		"   .section  __ex_table,\"a\"	\n"	\
+		"   .long	1b, 5b		\n"	\
+		"   .previous"				\
+		:[err] "+r"(err_), [tmp] "=r"(cb), [x] "+r"(__x) \
+		:[mem] "m"(*(addr_)), [efault] "i"(-EFAULT)); \
+	(x_) = (__force __typeof__(*(addr_)))__x;	\
+} while (0)
+
+
+/*
+ * Copy to/from user space
+ */
+
+extern unsigned __xtensa_copy_user(void *to, const void *from, unsigned n);
+
+static inline unsigned long
+raw_copy_from_user(void *to, const void __user *from, unsigned long n)
+{
+	prefetchw(to);
+	return __xtensa_copy_user(to, (__force const void *)from, n);
+}
+static inline unsigned long
+raw_copy_to_user(void __user *to, const void *from, unsigned long n)
+{
+	prefetch(from);
+	return __xtensa_copy_user((__force void *)to, from, n);
+}
+#define INLINE_COPY_USER
+
+/*
+ * We need to return the number of bytes not cleared.  Our memset()
+ * returns zero if a problem occurs while accessing user-space memory.
+ * In that event, return no memory cleared.  Otherwise, zero for
+ * success.
+ */
+
+static inline unsigned long
+__xtensa_clear_user(void __user *addr, unsigned long size)
+{
+	if (!__memset((void __force *)addr, 0, size))
+		return size;
+	return 0;
+}
+
+static inline unsigned long
+clear_user(void __user *addr, unsigned long size)
+{
+	if (access_ok(addr, size))
+		return __xtensa_clear_user(addr, size);
+	return size ? -EFAULT : 0;
+}
+
+#define __clear_user  __xtensa_clear_user
+
+
+#ifdef CONFIG_ARCH_HAS_STRNCPY_FROM_USER
+extern long __strncpy_user(char *dst, const char __user *src, long count);
+
+static inline long
+strncpy_from_user(char *dst, const char __user *src, long count)
+{
+	if (access_ok(src, 1))
+		return __strncpy_user(dst, src, count);
+	return -EFAULT;
+}
+#else
+long strncpy_from_user(char *dst, const char __user *src, long count);
+#endif
+
+/*
+ * Return the size of a string (including the ending 0!)
+ */
+extern long __strnlen_user(const char __user *str, long len);
+
+static inline long strnlen_user(const char __user *str, long len)
+{
+	if (!access_ok(str, 1))
+		return 0;
+	return __strnlen_user(str, len);
+}
+
+#endif	/* _XTENSA_UACCESS_H */
diff --git a/arch/xtensa/include/asm/unistd.h b/arch/xtensa/include/asm/unistd.h
new file mode 100644
index 000000000..30af4dc3c
--- /dev/null
+++ b/arch/xtensa/include/asm/unistd.h
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _XTENSA_UNISTD_H
+#define _XTENSA_UNISTD_H
+
+#define __ARCH_WANT_SYS_CLONE
+#include <uapi/asm/unistd.h>
+
+#define __ARCH_WANT_NEW_STAT
+#define __ARCH_WANT_STAT64
+#define __ARCH_WANT_SYS_UTIME32
+#define __ARCH_WANT_SYS_GETPGRP
+
+#define NR_syscalls				__NR_syscalls
+
+#endif /* _XTENSA_UNISTD_H */
diff --git a/arch/xtensa/include/asm/vectors.h b/arch/xtensa/include/asm/vectors.h
new file mode 100644
index 000000000..704286c35
--- /dev/null
+++ b/arch/xtensa/include/asm/vectors.h
@@ -0,0 +1,93 @@
+/*
+ * arch/xtensa/include/asm/xchal_vaddr_remap.h
+ *
+ * Xtensa macros for MMU V3 Support. Deals with re-mapping the Virtual
+ * Memory Addresses from "Virtual == Physical" to their prevvious V2 MMU
+ * mappings (KSEG at 0xD0000000 and KIO at 0XF0000000).
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2008 - 2012 Tensilica Inc.
+ *
+ * Pete Delaney <piet@tensilica.com>
+ * Marc Gauthier <marc@tensilica.com
+ */
+
+#ifndef _XTENSA_VECTORS_H
+#define _XTENSA_VECTORS_H
+
+#include <asm/core.h>
+#include <asm/kmem_layout.h>
+
+#if defined(CONFIG_MMU) && XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY
+#ifdef CONFIG_KERNEL_VIRTUAL_ADDRESS
+#define KERNELOFFSET			CONFIG_KERNEL_VIRTUAL_ADDRESS
+#else
+#define KERNELOFFSET			(CONFIG_KERNEL_LOAD_ADDRESS + \
+					 XCHAL_KSEG_CACHED_VADDR - \
+					 XCHAL_KSEG_PADDR)
+#endif
+#else
+#define KERNELOFFSET			CONFIG_KERNEL_LOAD_ADDRESS
+#endif
+
+#define RESET_VECTOR1_VADDR		(XCHAL_RESET_VECTOR1_VADDR)
+#ifdef CONFIG_VECTORS_ADDR
+#define VECBASE_VADDR			(CONFIG_VECTORS_ADDR)
+#else
+#define VECBASE_VADDR			_vecbase
+#endif
+
+#if XCHAL_HAVE_VECBASE
+
+#define VECTOR_VADDR(offset)		(VECBASE_VADDR + offset)
+
+#define USER_VECTOR_VADDR		VECTOR_VADDR(XCHAL_USER_VECOFS)
+#define KERNEL_VECTOR_VADDR		VECTOR_VADDR(XCHAL_KERNEL_VECOFS)
+#define DOUBLEEXC_VECTOR_VADDR		VECTOR_VADDR(XCHAL_DOUBLEEXC_VECOFS)
+#define WINDOW_VECTORS_VADDR		VECTOR_VADDR(XCHAL_WINDOW_OF4_VECOFS)
+#define INTLEVEL2_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL2_VECOFS)
+#define INTLEVEL3_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL3_VECOFS)
+#define INTLEVEL4_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL4_VECOFS)
+#define INTLEVEL5_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL5_VECOFS)
+#define INTLEVEL6_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL6_VECOFS)
+#define INTLEVEL7_VECTOR_VADDR		VECTOR_VADDR(XCHAL_INTLEVEL7_VECOFS)
+#define DEBUG_VECTOR_VADDR		VECTOR_VADDR(XCHAL_DEBUG_VECOFS)
+
+/*
+ * These XCHAL_* #defines from varian/core.h
+ * are not valid to use with V3 MMU. Non-XCHAL
+ * constants are defined above and should be used.
+ */
+#undef  XCHAL_VECBASE_RESET_VADDR
+#undef  XCHAL_USER_VECTOR_VADDR
+#undef  XCHAL_KERNEL_VECTOR_VADDR
+#undef  XCHAL_DOUBLEEXC_VECTOR_VADDR
+#undef  XCHAL_WINDOW_VECTORS_VADDR
+#undef  XCHAL_INTLEVEL2_VECTOR_VADDR
+#undef  XCHAL_INTLEVEL3_VECTOR_VADDR
+#undef  XCHAL_INTLEVEL4_VECTOR_VADDR
+#undef  XCHAL_INTLEVEL5_VECTOR_VADDR
+#undef  XCHAL_INTLEVEL6_VECTOR_VADDR
+#undef  XCHAL_INTLEVEL7_VECTOR_VADDR
+#undef  XCHAL_DEBUG_VECTOR_VADDR
+
+#else
+
+#define USER_VECTOR_VADDR		XCHAL_USER_VECTOR_VADDR
+#define KERNEL_VECTOR_VADDR		XCHAL_KERNEL_VECTOR_VADDR
+#define DOUBLEEXC_VECTOR_VADDR		XCHAL_DOUBLEEXC_VECTOR_VADDR
+#define WINDOW_VECTORS_VADDR		XCHAL_WINDOW_VECTORS_VADDR
+#define INTLEVEL2_VECTOR_VADDR		XCHAL_INTLEVEL2_VECTOR_VADDR
+#define INTLEVEL3_VECTOR_VADDR		XCHAL_INTLEVEL3_VECTOR_VADDR
+#define INTLEVEL4_VECTOR_VADDR		XCHAL_INTLEVEL4_VECTOR_VADDR
+#define INTLEVEL5_VECTOR_VADDR		XCHAL_INTLEVEL5_VECTOR_VADDR
+#define INTLEVEL6_VECTOR_VADDR		XCHAL_INTLEVEL6_VECTOR_VADDR
+#define INTLEVEL7_VECTOR_VADDR		XCHAL_INTLEVEL6_VECTOR_VADDR
+#define DEBUG_VECTOR_VADDR		XCHAL_DEBUG_VECTOR_VADDR
+
+#endif
+
+#endif /* _XTENSA_VECTORS_H */
diff --git a/arch/xtensa/include/asm/vermagic.h b/arch/xtensa/include/asm/vermagic.h
new file mode 100644
index 000000000..6d9c670e4
--- /dev/null
+++ b/arch/xtensa/include/asm/vermagic.h
@@ -0,0 +1,17 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _ASM_VERMAGIC_H
+#define _ASM_VERMAGIC_H
+
+#include <linux/stringify.h>
+#include <variant/core.h>
+
+#define MODULE_ARCH_VERMAGIC "xtensa-" __stringify(XCHAL_CORE_ID) " "
+
+#endif	/* _ASM_VERMAGIC_H */
diff --git a/arch/xtensa/include/asm/vmalloc.h b/arch/xtensa/include/asm/vmalloc.h
new file mode 100644
index 000000000..0eb94b70b
--- /dev/null
+++ b/arch/xtensa/include/asm/vmalloc.h
@@ -0,0 +1,4 @@
+#ifndef _ASM_XTENSA_VMALLOC_H
+#define _ASM_XTENSA_VMALLOC_H
+
+#endif /* _ASM_XTENSA_VMALLOC_H */
diff --git a/arch/xtensa/include/uapi/asm/Kbuild b/arch/xtensa/include/uapi/asm/Kbuild
new file mode 100644
index 000000000..da95c5a03
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/Kbuild
@@ -0,0 +1,3 @@
+# SPDX-License-Identifier: GPL-2.0
+generated-y += unistd_32.h
+generic-y += ucontext.h
diff --git a/arch/xtensa/include/uapi/asm/auxvec.h b/arch/xtensa/include/uapi/asm/auxvec.h
new file mode 100644
index 000000000..257dec75c
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/auxvec.h
@@ -0,0 +1,4 @@
+#ifndef __XTENSA_AUXVEC_H
+#define __XTENSA_AUXVEC_H
+
+#endif
diff --git a/arch/xtensa/include/uapi/asm/byteorder.h b/arch/xtensa/include/uapi/asm/byteorder.h
new file mode 100644
index 000000000..5b9f832c2
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/byteorder.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+#ifndef _XTENSA_BYTEORDER_H
+#define _XTENSA_BYTEORDER_H
+
+#ifdef __XTENSA_EL__
+#include <linux/byteorder/little_endian.h>
+#elif defined(__XTENSA_EB__)
+#include <linux/byteorder/big_endian.h>
+#else
+# error processor byte order undefined!
+#endif
+
+#endif /* _XTENSA_BYTEORDER_H */
diff --git a/arch/xtensa/include/uapi/asm/ioctls.h b/arch/xtensa/include/uapi/asm/ioctls.h
new file mode 100644
index 000000000..6d4a87296
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/ioctls.h
@@ -0,0 +1,130 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/ioctls.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2003 - 2005 Tensilica Inc.
+ *
+ * Derived from "include/asm-i386/ioctls.h"
+ */
+
+#ifndef _XTENSA_IOCTLS_H
+#define _XTENSA_IOCTLS_H
+
+#include <asm/ioctl.h>
+
+#define FIOCLEX		_IO('f', 1)
+#define FIONCLEX	_IO('f', 2)
+#define FIOASYNC	_IOW('f', 125, int)
+#define FIONBIO		_IOW('f', 126, int)
+#define FIONREAD	_IOR('f', 127, int)
+#define TIOCINQ		FIONREAD
+#define FIOQSIZE	_IOR('f', 128, loff_t)
+
+#define TCGETS		0x5401
+#define TCSETS		0x5402
+#define TCSETSW		0x5403
+#define TCSETSF		0x5404
+
+#define TCGETA		0x80127417	/* _IOR('t', 23, struct termio) */
+#define TCSETA		0x40127418	/* _IOW('t', 24, struct termio) */
+#define TCSETAW		0x40127419	/* _IOW('t', 25, struct termio) */
+#define TCSETAF		0x4012741C	/* _IOW('t', 28, struct termio) */
+
+#define TCSBRK		_IO('t', 29)
+#define TCXONC		_IO('t', 30)
+#define TCFLSH		_IO('t', 31)
+
+#define TIOCSWINSZ	0x40087467	/* _IOW('t', 103, struct winsize) */
+#define TIOCGWINSZ	0x80087468	/* _IOR('t', 104, struct winsize) */
+#define	TIOCSTART	_IO('t', 110)		/* start output, like ^Q */
+#define	TIOCSTOP	_IO('t', 111)		/* stop output, like ^S */
+#define TIOCOUTQ        _IOR('t', 115, int)     /* output queue size */
+
+#define TIOCSPGRP	_IOW('t', 118, int)
+#define TIOCGPGRP	_IOR('t', 119, int)
+
+#define TIOCEXCL	_IO('T', 12)
+#define TIOCNXCL	_IO('T', 13)
+#define TIOCSCTTY	_IO('T', 14)
+
+#define TIOCSTI		_IOW('T', 18, char)
+#define TIOCMGET	_IOR('T', 21, unsigned int)
+#define TIOCMBIS	_IOW('T', 22, unsigned int)
+#define TIOCMBIC	_IOW('T', 23, unsigned int)
+#define TIOCMSET	_IOW('T', 24, unsigned int)
+# define TIOCM_LE	0x001
+# define TIOCM_DTR	0x002
+# define TIOCM_RTS	0x004
+# define TIOCM_ST	0x008
+# define TIOCM_SR	0x010
+# define TIOCM_CTS	0x020
+# define TIOCM_CAR	0x040
+# define TIOCM_RNG	0x080
+# define TIOCM_DSR	0x100
+# define TIOCM_CD	TIOCM_CAR
+# define TIOCM_RI	TIOCM_RNG
+
+#define TIOCGSOFTCAR	_IOR('T', 25, unsigned int)
+#define TIOCSSOFTCAR	_IOW('T', 26, unsigned int)
+#define TIOCLINUX	_IOW('T', 28, char)
+#define TIOCCONS	_IO('T', 29)
+#define TIOCGSERIAL	0x803C541E	/*_IOR('T', 30, struct serial_struct)*/
+#define TIOCSSERIAL	0x403C541F	/*_IOW('T', 31, struct serial_struct)*/
+#define TIOCPKT		_IOW('T', 32, int)
+# define TIOCPKT_DATA		 0
+# define TIOCPKT_FLUSHREAD	 1
+# define TIOCPKT_FLUSHWRITE	 2
+# define TIOCPKT_STOP		 4
+# define TIOCPKT_START		 8
+# define TIOCPKT_NOSTOP		16
+# define TIOCPKT_DOSTOP		32
+# define TIOCPKT_IOCTL		64
+
+
+#define TIOCNOTTY	_IO('T', 34)
+#define TIOCSETD	_IOW('T', 35, int)
+#define TIOCGETD	_IOR('T', 36, int)
+#define TCSBRKP		_IOW('T', 37, int)   /* Needed for POSIX tcsendbreak()*/
+#define TIOCSBRK	_IO('T', 39) 	     /* BSD compatibility */
+#define TIOCCBRK	_IO('T', 40)	     /* BSD compatibility */
+#define TIOCGSID	_IOR('T', 41, pid_t) /* Return the session ID of FD*/
+#define TCGETS2		_IOR('T', 42, struct termios2)
+#define TCSETS2		_IOW('T', 43, struct termios2)
+#define TCSETSW2	_IOW('T', 44, struct termios2)
+#define TCSETSF2	_IOW('T', 45, struct termios2)
+#define TIOCGRS485	_IOR('T', 46, struct serial_rs485)
+#define TIOCSRS485	_IOWR('T', 47, struct serial_rs485)
+#define TIOCGPTN	_IOR('T',0x30, unsigned int) /* Get Pty Number (of pty-mux device) */
+#define TIOCSPTLCK	_IOW('T',0x31, int)  /* Lock/unlock Pty */
+#define TIOCGDEV	_IOR('T',0x32, unsigned int) /* Get primary device node of /dev/console */
+#define TIOCSIG		_IOW('T',0x36, int)  /* Generate signal on Pty slave */
+#define TIOCVHANGUP	_IO('T', 0x37)
+#define TIOCGPKT	_IOR('T', 0x38, int) /* Get packet mode state */
+#define TIOCGPTLCK	_IOR('T', 0x39, int) /* Get Pty lock state */
+#define TIOCGEXCL	_IOR('T', 0x40, int) /* Get exclusive mode state */
+#define TIOCGPTPEER	_IO('T', 0x41) /* Safely open the slave */
+#define TIOCGISO7816	_IOR('T', 0x42, struct serial_iso7816)
+#define TIOCSISO7816	_IOWR('T', 0x43, struct serial_iso7816)
+
+#define TIOCSERCONFIG	_IO('T', 83)
+#define TIOCSERGWILD	_IOR('T', 84,  int)
+#define TIOCSERSWILD	_IOW('T', 85,  int)
+#define TIOCGLCKTRMIOS	0x5456
+#define TIOCSLCKTRMIOS	0x5457
+#define TIOCSERGSTRUCT	0x5458		     /* For debugging only */
+#define TIOCSERGETLSR   _IOR('T', 89, unsigned int) /* Get line status reg. */
+  /* ioctl (fd, TIOCSERGETLSR, &result) where result may be as below */
+# define TIOCSER_TEMT    0x01		     /* Transmitter physically empty */
+#define TIOCSERGETMULTI 0x80a8545a /* Get multiport config  */
+			/* _IOR('T', 90, struct serial_multiport_struct) */
+#define TIOCSERSETMULTI 0x40a8545b /* Set multiport config */
+			/* _IOW('T', 91, struct serial_multiport_struct) */
+
+#define TIOCMIWAIT	_IO('T', 92) /* wait for a change on serial input line(s) */
+#define TIOCGICOUNT	0x545D	/* read serial port inline interrupt counts */
+
+#endif /* _XTENSA_IOCTLS_H */
diff --git a/arch/xtensa/include/uapi/asm/ipcbuf.h b/arch/xtensa/include/uapi/asm/ipcbuf.h
new file mode 100644
index 000000000..3bd0642f6
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/ipcbuf.h
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/ipcbuf.h
+ *
+ * The ipc64_perm structure for the Xtensa architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_IPCBUF_H
+#define _XTENSA_IPCBUF_H
+
+#include <linux/posix_types.h>
+
+/*
+ * Pad space is left for:
+ * - 32-bit mode_t and seq
+ * - 2 miscellaneous 32-bit values
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ */
+
+struct ipc64_perm
+{
+	__kernel_key_t		key;
+	__kernel_uid32_t	uid;
+	__kernel_gid32_t	gid;
+	__kernel_uid32_t	cuid;
+	__kernel_gid32_t	cgid;
+	__kernel_mode_t		mode;
+	unsigned long		seq;
+	unsigned long		__unused1;
+	unsigned long		__unused2;
+};
+
+#endif /* _XTENSA_IPCBUF_H */
diff --git a/arch/xtensa/include/uapi/asm/mman.h b/arch/xtensa/include/uapi/asm/mman.h
new file mode 100644
index 000000000..99d4ccee7
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/mman.h
@@ -0,0 +1,127 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/mman.h
+ *
+ * Xtensa Processor memory-manager definitions
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1995 by Ralf Baechle
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_MMAN_H
+#define _XTENSA_MMAN_H
+
+/*
+ * Protections are chosen from these bits, OR'd together.  The
+ * implementation does not necessarily support PROT_EXEC or PROT_WRITE
+ * without PROT_READ.  The only guarantees are that no writing will be
+ * allowed without PROT_WRITE and no access will be allowed for PROT_NONE.
+ */
+
+#define PROT_NONE	0x0		/* page can not be accessed */
+#define PROT_READ	0x1		/* page can be read */
+#define PROT_WRITE	0x2		/* page can be written */
+#define PROT_EXEC	0x4		/* page can be executed */
+
+#define PROT_SEM	0x10		/* page may be used for atomic ops */
+#define PROT_GROWSDOWN	0x01000000	/* mprotect flag: extend change to start of growsdown vma */
+#define PROT_GROWSUP	0x02000000	/* mprotect flag: extend change to end fo growsup vma */
+
+/*
+ * Flags for mmap
+ */
+/* 0x01 - 0x03 are defined in linux/mman.h */
+#define MAP_TYPE	0x00f		/* Mask for type of mapping */
+#define MAP_FIXED	0x010		/* Interpret addr exactly */
+
+/* not used by linux, but here to make sure we don't clash with ABI defines */
+#define MAP_RENAME	0x020		/* Assign page to file */
+#define MAP_AUTOGROW	0x040		/* File may grow by writing */
+#define MAP_LOCAL	0x080		/* Copy on fork/sproc */
+#define MAP_AUTORSRV	0x100		/* Logical swap reserved on demand */
+
+/* These are linux-specific */
+#define MAP_NORESERVE	0x0400		/* don't check for reservations */
+#define MAP_ANONYMOUS	0x0800		/* don't use a file */
+#define MAP_GROWSDOWN	0x1000		/* stack-like segment */
+#define MAP_DENYWRITE	0x2000		/* ETXTBSY */
+#define MAP_EXECUTABLE	0x4000		/* mark it as an executable */
+#define MAP_LOCKED	0x8000		/* pages are locked */
+#define MAP_POPULATE	0x10000		/* populate (prefault) pagetables */
+#define MAP_NONBLOCK	0x20000		/* do not block on IO */
+#define MAP_STACK	0x40000		/* give out an address that is best suited for process/thread stacks */
+#define MAP_HUGETLB	0x80000		/* create a huge page mapping */
+#define MAP_FIXED_NOREPLACE 0x100000	/* MAP_FIXED which doesn't unmap underlying mapping */
+#define MAP_UNINITIALIZED 0x4000000	/* For anonymous mmap, memory could be
+					 * uninitialized */
+
+/*
+ * Flags for msync
+ */
+#define MS_ASYNC	0x0001		/* sync memory asynchronously */
+#define MS_INVALIDATE	0x0002		/* invalidate mappings & caches */
+#define MS_SYNC		0x0004		/* synchronous memory sync */
+
+/*
+ * Flags for mlockall
+ */
+#define MCL_CURRENT	1		/* lock all current mappings */
+#define MCL_FUTURE	2		/* lock all future mappings */
+#define MCL_ONFAULT	4		/* lock all pages that are faulted in */
+
+/*
+ * Flags for mlock
+ */
+#define MLOCK_ONFAULT	0x01		/* Lock pages in range after they are faulted in, do not prefault */
+
+#define MADV_NORMAL	0		/* no further special treatment */
+#define MADV_RANDOM	1		/* expect random page references */
+#define MADV_SEQUENTIAL	2		/* expect sequential page references */
+#define MADV_WILLNEED	3		/* will need these pages */
+#define MADV_DONTNEED	4		/* don't need these pages */
+
+/* common parameters: try to keep these consistent across architectures */
+#define MADV_FREE	8		/* free pages only if memory pressure */
+#define MADV_REMOVE	9		/* remove these pages & resources */
+#define MADV_DONTFORK	10		/* don't inherit across fork */
+#define MADV_DOFORK	11		/* do inherit across fork */
+
+#define MADV_MERGEABLE   12		/* KSM may merge identical pages */
+#define MADV_UNMERGEABLE 13		/* KSM may not merge identical pages */
+
+#define MADV_HUGEPAGE	14		/* Worth backing with hugepages */
+#define MADV_NOHUGEPAGE	15		/* Not worth backing with hugepages */
+
+#define MADV_DONTDUMP   16		/* Explicity exclude from the core dump,
+					   overrides the coredump filter bits */
+#define MADV_DODUMP	17		/* Clear the MADV_NODUMP flag */
+
+#define MADV_WIPEONFORK 18		/* Zero memory on fork, child only */
+#define MADV_KEEPONFORK 19		/* Undo MADV_WIPEONFORK */
+
+#define MADV_COLD	20		/* deactivate these pages */
+#define MADV_PAGEOUT	21		/* reclaim these pages */
+
+#define MADV_POPULATE_READ	22	/* populate (prefault) page tables readable */
+#define MADV_POPULATE_WRITE	23	/* populate (prefault) page tables writable */
+
+#define MADV_DONTNEED_LOCKED	24	/* like DONTNEED, but drop locked pages too */
+
+#define MADV_COLLAPSE	25		/* Synchronous hugepage collapse */
+
+#define MADV_GUARD_INSTALL 102		/* fatal signal on access to range */
+#define MADV_GUARD_REMOVE 103		/* unguard range */
+
+/* compatibility flags */
+#define MAP_FILE	0
+
+#define PKEY_DISABLE_ACCESS	0x1
+#define PKEY_DISABLE_WRITE	0x2
+#define PKEY_ACCESS_MASK	(PKEY_DISABLE_ACCESS |\
+				 PKEY_DISABLE_WRITE)
+
+#endif /* _XTENSA_MMAN_H */
diff --git a/arch/xtensa/include/uapi/asm/msgbuf.h b/arch/xtensa/include/uapi/asm/msgbuf.h
new file mode 100644
index 000000000..1ed2c85b6
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/msgbuf.h
@@ -0,0 +1,50 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/msgbuf.h
+ *
+ * The msqid64_ds structure for the Xtensa architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 2 miscellaneous 32-bit values
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ */
+
+#ifndef _XTENSA_MSGBUF_H
+#define _XTENSA_MSGBUF_H
+
+#include <asm/ipcbuf.h>
+
+struct msqid64_ds {
+	struct ipc64_perm msg_perm;
+#ifdef __XTENSA_EB__
+	unsigned long  msg_stime_high;
+	unsigned long  msg_stime;	/* last msgsnd time */
+	unsigned long  msg_rtime_high;
+	unsigned long  msg_rtime;	/* last msgrcv time */
+	unsigned long  msg_ctime_high;
+	unsigned long  msg_ctime;	/* last change time */
+#elif defined(__XTENSA_EL__)
+	unsigned long  msg_stime;	/* last msgsnd time */
+	unsigned long  msg_stime_high;
+	unsigned long  msg_rtime;	/* last msgrcv time */
+	unsigned long  msg_rtime_high;
+	unsigned long  msg_ctime;	/* last change time */
+	unsigned long  msg_ctime_high;
+#else
+# error processor byte order undefined!
+#endif
+	unsigned long  msg_cbytes;	/* current number of bytes on queue */
+	unsigned long  msg_qnum;	/* number of messages in queue */
+	unsigned long  msg_qbytes;	/* max number of bytes on queue */
+	__kernel_pid_t msg_lspid;	/* pid of last msgsnd */
+	__kernel_pid_t msg_lrpid;	/* last receive pid */
+	unsigned long  __unused4;
+	unsigned long  __unused5;
+};
+
+#endif	/* _XTENSA_MSGBUF_H */
diff --git a/arch/xtensa/include/uapi/asm/poll.h b/arch/xtensa/include/uapi/asm/poll.h
new file mode 100644
index 000000000..4d249040b
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/poll.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/poll.h
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_POLL_H
+#define _XTENSA_POLL_H
+
+#define POLLWRNORM	POLLOUT
+#define POLLWRBAND	0x0100
+#define POLLREMOVE	0x0800
+
+#include <asm-generic/poll.h>
+
+#endif /* _XTENSA_POLL_H */
diff --git a/arch/xtensa/include/uapi/asm/posix_types.h b/arch/xtensa/include/uapi/asm/posix_types.h
new file mode 100644
index 000000000..1dc675928
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/posix_types.h
@@ -0,0 +1,40 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/posix_types.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Largely copied from include/asm-ppc/posix_types.h
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_POSIX_TYPES_H
+#define _XTENSA_POSIX_TYPES_H
+
+/*
+ * This file is generally used by user-level software, so you need to
+ * be a little careful about namespace pollution etc.  Also, we cannot
+ * assume GCC is being used.
+ */
+
+typedef unsigned short	__kernel_ipc_pid_t;
+#define __kernel_ipc_pid_t __kernel_ipc_pid_t
+
+typedef unsigned int	__kernel_size_t;
+typedef int		__kernel_ssize_t;
+typedef long		__kernel_ptrdiff_t;
+#define __kernel_size_t __kernel_size_t
+
+typedef unsigned short	__kernel_old_uid_t;
+typedef unsigned short	__kernel_old_gid_t;
+#define __kernel_old_uid_t __kernel_old_uid_t
+
+typedef unsigned short	__kernel_old_dev_t;
+#define __kernel_old_dev_t __kernel_old_dev_t
+
+#include <asm-generic/posix_types.h>
+
+#endif /* _XTENSA_POSIX_TYPES_H */
diff --git a/arch/xtensa/include/uapi/asm/ptrace.h b/arch/xtensa/include/uapi/asm/ptrace.h
new file mode 100644
index 000000000..6e89ea301
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/ptrace.h
@@ -0,0 +1,63 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/ptrace.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _UAPI_XTENSA_PTRACE_H
+#define _UAPI_XTENSA_PTRACE_H
+
+#include <linux/types.h>
+
+/* Registers used by strace */
+
+#define REG_A_BASE	0x0000
+#define REG_AR_BASE	0x0100
+#define REG_PC		0x0020
+#define REG_PS		0x02e6
+#define REG_WB		0x0248
+#define REG_WS		0x0249
+#define REG_LBEG	0x0200
+#define REG_LEND	0x0201
+#define REG_LCOUNT	0x0202
+#define REG_SAR		0x0203
+
+#define SYSCALL_NR	0x00ff
+
+/* Other PTRACE_ values defined in <linux/ptrace.h> using values 0-9,16,17,24 */
+
+#define PTRACE_GETREGS		12
+#define PTRACE_SETREGS		13
+#define PTRACE_GETXTREGS	18
+#define PTRACE_SETXTREGS	19
+#define PTRACE_GETHBPREGS	20
+#define PTRACE_SETHBPREGS	21
+#define PTRACE_GETFDPIC		22
+
+#define PTRACE_GETFDPIC_EXEC	0
+#define PTRACE_GETFDPIC_INTERP	1
+
+#ifndef __ASSEMBLER__
+
+struct user_pt_regs {
+	__u32 pc;
+	__u32 ps;
+	__u32 lbeg;
+	__u32 lend;
+	__u32 lcount;
+	__u32 sar;
+	__u32 windowstart;
+	__u32 windowbase;
+	__u32 threadptr;
+	__u32 syscall;
+	__u32 reserved[6 + 48];
+	__u32 a[64];
+};
+
+#endif
+#endif /* _UAPI_XTENSA_PTRACE_H */
diff --git a/arch/xtensa/include/uapi/asm/sembuf.h b/arch/xtensa/include/uapi/asm/sembuf.h
new file mode 100644
index 000000000..3b9cdd406
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/sembuf.h
@@ -0,0 +1,45 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/sembuf.h
+ *
+ * The semid64_ds structure for Xtensa architecture.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ *
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space.
+ *
+ * Pad space is left for:
+ * - 2 miscellaneous 32-bit values
+ *
+ */
+
+#ifndef _XTENSA_SEMBUF_H
+#define _XTENSA_SEMBUF_H
+
+#include <asm/byteorder.h>
+#include <asm/ipcbuf.h>
+
+struct semid64_ds {
+	struct ipc64_perm sem_perm;		/* permissions .. see ipc.h */
+#ifdef __XTENSA_EL__
+	unsigned long	sem_otime;		/* last semop time */
+	unsigned long	sem_otime_high;
+	unsigned long	sem_ctime;		/* last change time */
+	unsigned long	sem_ctime_high;
+#else
+	unsigned long	sem_otime_high;
+	unsigned long	sem_otime;		/* last semop time */
+	unsigned long	sem_ctime_high;
+	unsigned long	sem_ctime;		/* last change time */
+#endif
+	unsigned long	sem_nsems;		/* no. of semaphores in array */
+	unsigned long	__unused3;
+	unsigned long	__unused4;
+};
+
+#endif /* __ASM_XTENSA_SEMBUF_H */
diff --git a/arch/xtensa/include/uapi/asm/setup.h b/arch/xtensa/include/uapi/asm/setup.h
new file mode 100644
index 000000000..5356a5fd4
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/setup.h
@@ -0,0 +1,17 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/setup.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SETUP_H
+#define _XTENSA_SETUP_H
+
+#define COMMAND_LINE_SIZE	256
+
+#endif
diff --git a/arch/xtensa/include/uapi/asm/shmbuf.h b/arch/xtensa/include/uapi/asm/shmbuf.h
new file mode 100644
index 000000000..bb8bdddae
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/shmbuf.h
@@ -0,0 +1,54 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/shmbuf.h
+ *
+ * The shmid64_ds structure for Xtensa architecture.
+ * Note extra padding because this structure is passed back and forth
+ * between kernel and user space, but the padding is on the wrong
+ * side for big-endian xtensa, for historic reasons.
+ *
+ * Pad space is left for:
+ * - 2 miscellaneous 32-bit values
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SHMBUF_H
+#define _XTENSA_SHMBUF_H
+
+#include <asm/ipcbuf.h>
+#include <asm/posix_types.h>
+
+struct shmid64_ds {
+	struct ipc64_perm	shm_perm;	/* operation perms */
+	__kernel_size_t		shm_segsz;	/* size of segment (bytes) */
+	unsigned long		shm_atime;	/* last attach time */
+	unsigned long		shm_atime_high;
+	unsigned long		shm_dtime;	/* last detach time */
+	unsigned long		shm_dtime_high;
+	unsigned long		shm_ctime;	/* last change time */
+	unsigned long		shm_ctime_high;
+	__kernel_pid_t		shm_cpid;	/* pid of creator */
+	__kernel_pid_t		shm_lpid;	/* pid of last operator */
+	unsigned long		shm_nattch;	/* no. of current attaches */
+	unsigned long		__unused4;
+	unsigned long		__unused5;
+};
+
+struct shminfo64 {
+	unsigned long	shmmax;
+	unsigned long	shmmin;
+	unsigned long	shmmni;
+	unsigned long	shmseg;
+	unsigned long	shmall;
+	unsigned long	__unused1;
+	unsigned long	__unused2;
+	unsigned long	__unused3;
+	unsigned long	__unused4;
+};
+
+#endif	/* _XTENSA_SHMBUF_H */
diff --git a/arch/xtensa/include/uapi/asm/sigcontext.h b/arch/xtensa/include/uapi/asm/sigcontext.h
new file mode 100644
index 000000000..9cbf39e63
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/sigcontext.h
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/sigcontext.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SIGCONTEXT_H
+#define _XTENSA_SIGCONTEXT_H
+
+
+struct sigcontext {
+	unsigned long sc_pc;
+	unsigned long sc_ps;
+	unsigned long sc_lbeg;
+	unsigned long sc_lend;
+	unsigned long sc_lcount;
+	unsigned long sc_sar;
+	unsigned long sc_acclo;
+	unsigned long sc_acchi;
+	unsigned long sc_a[16];
+	void *sc_xtregs;
+};
+
+#endif /* _XTENSA_SIGCONTEXT_H */
diff --git a/arch/xtensa/include/uapi/asm/signal.h b/arch/xtensa/include/uapi/asm/signal.h
new file mode 100644
index 000000000..8060f1914
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/signal.h
@@ -0,0 +1,110 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/signal.h
+ *
+ * Swiped from SH.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _UAPI_XTENSA_SIGNAL_H
+#define _UAPI_XTENSA_SIGNAL_H
+
+
+#define _NSIG		64
+#define _NSIG_BPW	32
+#define _NSIG_WORDS	(_NSIG / _NSIG_BPW)
+
+#ifndef __ASSEMBLER__
+
+#include <linux/types.h>
+
+/* Avoid too many header ordering problems.  */
+struct siginfo;
+typedef unsigned long old_sigset_t;		/* at least 32 bits */
+typedef struct {
+	unsigned long sig[_NSIG_WORDS];
+} sigset_t;
+
+#endif
+
+#define SIGHUP		 1
+#define SIGINT		 2
+#define SIGQUIT		 3
+#define SIGILL		 4
+#define SIGTRAP		 5
+#define SIGABRT		 6
+#define SIGIOT		 6
+#define SIGBUS		 7
+#define SIGFPE		 8
+#define SIGKILL		 9
+#define SIGUSR1		10
+#define SIGSEGV		11
+#define SIGUSR2		12
+#define SIGPIPE		13
+#define SIGALRM		14
+#define SIGTERM		15
+#define SIGSTKFLT	16
+#define SIGCHLD		17
+#define SIGCONT		18
+#define SIGSTOP		19
+#define SIGTSTP		20
+#define SIGTTIN		21
+#define SIGTTOU		22
+#define SIGURG		23
+#define SIGXCPU		24
+#define SIGXFSZ		25
+#define SIGVTALRM	26
+#define SIGPROF		27
+#define SIGWINCH	28
+#define SIGIO		29
+#define SIGPOLL		SIGIO
+/* #define SIGLOST		29 */
+#define SIGPWR		30
+#define SIGSYS		31
+#define	SIGUNUSED	31
+
+/* These should not be considered constants from userland.  */
+#define SIGRTMIN	32
+#define SIGRTMAX	(_NSIG-1)
+
+#define SA_RESTORER	0x04000000
+
+#define MINSIGSTKSZ	2048
+#define SIGSTKSZ	8192
+
+#ifndef __ASSEMBLER__
+
+#include <asm-generic/signal-defs.h>
+
+#ifndef __KERNEL__
+
+/* Here we must cater to libcs that poke about in kernel headers.  */
+
+struct sigaction {
+	union {
+	  __sighandler_t _sa_handler;
+	  void (*_sa_sigaction)(int, struct siginfo *, void *);
+	} _u;
+	sigset_t sa_mask;
+	unsigned long sa_flags;
+	void (*sa_restorer)(void);
+};
+
+#define sa_handler	_u._sa_handler
+#define sa_sigaction	_u._sa_sigaction
+
+#endif /* __KERNEL__ */
+
+typedef struct sigaltstack {
+	void *ss_sp;
+	int ss_flags;
+	__kernel_size_t ss_size;
+} stack_t;
+
+#endif	/* __ASSEMBLER__ */
+#endif /* _UAPI_XTENSA_SIGNAL_H */
diff --git a/arch/xtensa/include/uapi/asm/sockios.h b/arch/xtensa/include/uapi/asm/sockios.h
new file mode 100644
index 000000000..1a1f58f4b
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/sockios.h
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/sockios.h
+ *
+ * Socket-level I/O control calls.  Copied from MIPS.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1995 by Ralf Baechle
+ * Copyright (C) 2001 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SOCKIOS_H
+#define _XTENSA_SOCKIOS_H
+
+#include <asm/ioctl.h>
+
+/* Socket-level I/O control calls. */
+
+#define FIOGETOWN	_IOR('f', 123, int)
+#define FIOSETOWN 	_IOW('f', 124, int)
+
+#define SIOCATMARK	_IOR('s', 7, int)
+#define SIOCSPGRP	_IOW('s', 8, pid_t)
+#define SIOCGPGRP	_IOR('s', 9, pid_t)
+
+#define SIOCGSTAMP_OLD	0x8906		/* Get stamp (timeval) */
+#define SIOCGSTAMPNS_OLD 0x8907		/* Get stamp (timespec) */
+
+#endif	/* _XTENSA_SOCKIOS_H */
diff --git a/arch/xtensa/include/uapi/asm/stat.h b/arch/xtensa/include/uapi/asm/stat.h
new file mode 100644
index 000000000..7b4d90d6d
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/stat.h
@@ -0,0 +1,60 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/stat.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_STAT_H
+#define _XTENSA_STAT_H
+
+#define STAT_HAVE_NSEC 1
+
+struct stat {
+	unsigned long	st_dev;
+	unsigned long	st_ino;
+	unsigned int	st_mode;
+	unsigned int	st_nlink;
+	unsigned int	st_uid;
+	unsigned int	st_gid;
+	unsigned long	st_rdev;
+	long		st_size;
+	unsigned long	st_blksize;
+	unsigned long	st_blocks;
+	unsigned long	st_atime;
+	unsigned long	st_atime_nsec;
+	unsigned long	st_mtime;
+	unsigned long	st_mtime_nsec;
+	unsigned long	st_ctime;
+	unsigned long	st_ctime_nsec;
+	unsigned long	__unused4;
+	unsigned long	__unused5;
+};
+
+struct stat64  {
+	unsigned long long st_dev;	/* Device */
+	unsigned long long st_ino;	/* File serial number */
+	unsigned int  st_mode;		/* File mode. */
+	unsigned int  st_nlink;		/* Link count. */
+	unsigned int  st_uid;		/* User ID of the file's owner. */
+	unsigned int  st_gid;		/* Group ID of the file's group. */
+	unsigned long long st_rdev;	/* Device number, if device. */
+	long long st_size;		/* Size of file, in bytes. */
+	unsigned long st_blksize;	/* Optimal block size for I/O. */
+	unsigned long __unused2;
+	unsigned long long st_blocks;	/* Number 512-byte blocks allocated. */
+	unsigned long st_atime;		/* Time of last access. */
+	unsigned long st_atime_nsec;
+	unsigned long st_mtime;		/* Time of last modification. */
+	unsigned long st_mtime_nsec;
+	unsigned long st_ctime;		/* Time of last status change. */
+	unsigned long st_ctime_nsec;
+	unsigned long __unused4;
+	unsigned long __unused5;
+};
+
+#endif	/* _XTENSA_STAT_H */
diff --git a/arch/xtensa/include/uapi/asm/swab.h b/arch/xtensa/include/uapi/asm/swab.h
new file mode 100644
index 000000000..e677cf4cc
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/swab.h
@@ -0,0 +1,71 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/swab.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_SWAB_H
+#define _XTENSA_SWAB_H
+
+#include <linux/types.h>
+#include <linux/compiler.h>
+
+#define __SWAB_64_THRU_32__
+
+static inline __attribute_const__ __u32 __arch_swab32(__u32 x)
+{
+    __u32 res;
+    /* instruction sequence from Xtensa ISA release 2/2000 */
+    __asm__("ssai     8           \n\t"
+	    "srli     %0, %1, 16  \n\t"
+	    "src      %0, %0, %1  \n\t"
+	    "src      %0, %0, %0  \n\t"
+	    "src      %0, %1, %0  \n"
+	    : "=&a" (res)
+	    : "a" (x)
+	    );
+    return res;
+}
+#define __arch_swab32 __arch_swab32
+
+static inline __attribute_const__ __u16 __arch_swab16(__u16 x)
+{
+    /* Given that 'short' values are signed (i.e., can be negative),
+     * we cannot assume that the upper 16-bits of the register are
+     * zero.  We are careful to mask values after shifting.
+     */
+
+    /* There exists an anomaly between xt-gcc and xt-xcc.  xt-gcc
+     * inserts an extui instruction after putting this function inline
+     * to ensure that it uses only the least-significant 16 bits of
+     * the result.  xt-xcc doesn't use an extui, but assumes the
+     * __asm__ macro follows convention that the upper 16 bits of an
+     * 'unsigned short' result are still zero.  This macro doesn't
+     * follow convention; indeed, it leaves garbage in the upport 16
+     * bits of the register.
+
+     * Declaring the temporary variables 'res' and 'tmp' to be 32-bit
+     * types while the return type of the function is a 16-bit type
+     * forces both compilers to insert exactly one extui instruction
+     * (or equivalent) to mask off the upper 16 bits. */
+
+    __u32 res;
+    __u32 tmp;
+
+    __asm__("extui    %1, %2, 8, 8\n\t"
+	    "slli     %0, %2, 8   \n\t"
+	    "or       %0, %0, %1  \n"
+	    : "=&a" (res), "=&a" (tmp)
+	    : "a" (x)
+	    );
+
+    return res;
+}
+#define __arch_swab16 __arch_swab16
+
+#endif /* _XTENSA_SWAB_H */
diff --git a/arch/xtensa/include/uapi/asm/types.h b/arch/xtensa/include/uapi/asm/types.h
new file mode 100644
index 000000000..2e9217a06
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/types.h
@@ -0,0 +1,30 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+/*
+ * include/asm-xtensa/types.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ */
+
+#ifndef _UAPI_XTENSA_TYPES_H
+#define _UAPI_XTENSA_TYPES_H
+
+#include <asm-generic/int-ll64.h>
+
+#ifdef __ASSEMBLER__
+# define __XTENSA_UL(x)		(x)
+# define __XTENSA_UL_CONST(x)	x
+#else
+# define __XTENSA_UL(x)		((unsigned long)(x))
+# define ___XTENSA_UL_CONST(x)	x##UL
+# define __XTENSA_UL_CONST(x)	___XTENSA_UL_CONST(x)
+#endif
+
+#ifndef __ASSEMBLER__
+
+#endif
+
+#endif /* _UAPI_XTENSA_TYPES_H */
diff --git a/arch/xtensa/include/uapi/asm/unistd.h b/arch/xtensa/include/uapi/asm/unistd.h
new file mode 100644
index 000000000..e67ccdd11
--- /dev/null
+++ b/arch/xtensa/include/uapi/asm/unistd.h
@@ -0,0 +1,26 @@
+/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */
+#ifndef _UAPI_XTENSA_UNISTD_H
+#define _UAPI_XTENSA_UNISTD_H
+
+#include <asm/unistd_32.h>
+
+#define __ARCH_WANT_SYS_OLDUMOUNT
+
+/*
+ * sysxtensa syscall handler
+ *
+ * int sysxtensa (SYS_XTENSA_ATOMIC_SET,     ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_ADD,     ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_EXG_ADD, ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_CMP_SWP, ptr, oldval, newval);
+ *        a2            a6                   a3    a4      a5
+ */
+
+#define SYS_XTENSA_RESERVED               0     /* don't use this */
+#define SYS_XTENSA_ATOMIC_SET             1     /* set variable */
+#define SYS_XTENSA_ATOMIC_EXG_ADD         2     /* exchange memory and add */
+#define SYS_XTENSA_ATOMIC_ADD             3     /* add to memory */
+#define SYS_XTENSA_ATOMIC_CMP_SWP         4     /* compare and swap */
+#define SYS_XTENSA_COUNT                  5     /* count */
+
+#endif /* _UAPI_XTENSA_UNISTD_H */
diff --git a/arch/xtensa/kernel/.gitignore b/arch/xtensa/kernel/.gitignore
new file mode 100644
index 000000000..bbb90f92d
--- /dev/null
+++ b/arch/xtensa/kernel/.gitignore
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+vmlinux.lds
diff --git a/arch/xtensa/kernel/Makefile b/arch/xtensa/kernel/Makefile
new file mode 100644
index 000000000..d3ef04074
--- /dev/null
+++ b/arch/xtensa/kernel/Makefile
@@ -0,0 +1,46 @@
+# SPDX-License-Identifier: GPL-2.0
+#
+# Makefile for the Linux/Xtensa kernel.
+#
+
+always-$(KBUILD_BUILTIN) := vmlinux.lds
+
+obj-y := head.o align.o coprocessor.o entry.o irq.o platform.o process.o \
+	 ptrace.o setup.o signal.o stacktrace.o syscall.o time.o traps.o \
+	 vectors.o
+
+obj-$(CONFIG_MMU) += pci-dma.o
+obj-$(CONFIG_PCI) += pci.o
+obj-$(CONFIG_MODULES) += xtensa_ksyms.o module.o
+obj-$(CONFIG_FUNCTION_TRACER) += mcount.o
+obj-$(CONFIG_SMP) += smp.o
+obj-$(CONFIG_SECONDARY_RESET_VECTOR) += mxhead.o
+obj-$(CONFIG_XTENSA_VARIANT_HAVE_PERF_EVENTS) += perf_event.o
+obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o
+obj-$(CONFIG_S32C1I_SELFTEST) += s32c1i_selftest.o
+obj-$(CONFIG_JUMP_LABEL) += jump_label.o
+obj-$(CONFIG_HIBERNATION) += hibernate.o
+
+# In the Xtensa architecture, assembly generates literals which must always
+# precede the L32R instruction with a relative offset less than 256 kB.
+# Therefore, the .text and .literal section must be combined in parenthesis
+# in the linker script, such as: *(.literal .text).
+#
+# We need to post-process the generated vmlinux.lds scripts to convert
+# *(xxx.text) to  *(xxx.literal xxx.text) for the following text sections:
+#   .text .ref.text .*init.text .*exit.text .text.*
+#
+# Replicate rules in scripts/Makefile.build
+
+sed-y = -e ':a; s/\*(\([^)]*\)\.text\.unlikely/*(\1.literal.unlikely .{text}.unlikely/; ta; ' \
+	-e ':b; s/\*(\([^)]*\)\.text\(\.[a-z]*\)/*(\1.{text}\2.literal .{text}\2/; tb; ' \
+	-e ':c; s/\*(\([^)]*\)\(\.[a-z]*it\|\.ref\)\.text/*(\1\2.literal \2.{text}/; tc; ' \
+	-e ':d; s/\*(\([^)]\+ \|\)\.text/*(\1.literal .{text}/; td; ' \
+	-e 's/\.{text}/.text/g'
+
+quiet_cmd__cpp_lds_S = LDS     $@
+cmd__cpp_lds_S = $(CPP) $(cpp_flags) -P -C -Uxtensa -D__ASSEMBLY__ \
+		 -DLINKER_SCRIPT $< | sed $(sed-y) >$@
+
+$(obj)/vmlinux.lds: $(src)/vmlinux.lds.S FORCE
+	$(call if_changed_dep,_cpp_lds_S)
diff --git a/arch/xtensa/kernel/align.S b/arch/xtensa/kernel/align.S
new file mode 100644
index 000000000..f2d0d831c
--- /dev/null
+++ b/arch/xtensa/kernel/align.S
@@ -0,0 +1,579 @@
+/*
+ * arch/xtensa/kernel/align.S
+ *
+ * Handle unalignment and load/store exceptions.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica, Inc.
+ * Copyright (C) 2014 Cadence Design Systems Inc.
+ *
+ * Rewritten by Chris Zankel <chris@zankel.net>
+ *
+ * Based on work from Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ * and Marc Gauthier <marc@tensilica.com, marc@alimni.uwaterloo.ca>
+ */
+
+#include <linux/linkage.h>
+#include <asm/current.h>
+#include <asm/asm-offsets.h>
+#include <asm/asmmacro.h>
+#include <asm/processor.h>
+#ifdef CONFIG_MMU
+#include <asm/asm-uaccess.h>
+#endif
+
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || defined CONFIG_XTENSA_LOAD_STORE
+#define LOAD_EXCEPTION_HANDLER
+#endif
+
+#if XCHAL_UNALIGNED_STORE_EXCEPTION || defined CONFIG_XTENSA_LOAD_STORE
+#define STORE_EXCEPTION_HANDLER
+#endif
+
+#if defined LOAD_EXCEPTION_HANDLER || defined STORE_EXCEPTION_HANDLER
+#define ANY_EXCEPTION_HANDLER
+#endif
+
+#if XCHAL_HAVE_WINDOWED && defined CONFIG_MMU
+#define UNALIGNED_USER_EXCEPTION
+#endif
+
+/* Big and little endian 16-bit values are located in
+ * different halves of a register.  HWORD_START helps to
+ * abstract the notion of extracting a 16-bit value from a
+ * register.
+ * We also have to define new shifting instructions because
+ * lsb and msb are on 'opposite' ends in a register for
+ * different endian machines.
+ *
+ * Assume a memory region in ascending address:
+ *   	0 1 2 3|4 5 6 7
+ *
+ * When loading one word into a register, the content of that register is:
+ *  LE	3 2 1 0, 7 6 5 4
+ *  BE  0 1 2 3, 4 5 6 7
+ *
+ * Masking the bits of the higher/lower address means:
+ *  LE  X X 0 0, 0 0 X X
+ *  BE	0 0 X X, X X 0 0
+ *
+ * Shifting to higher/lower addresses, means:
+ *  LE  shift left / shift right
+ *  BE  shift right / shift left
+ *
+ * Extracting 16 bits from a 32 bit reg. value to higher/lower address means:
+ *  LE  mask 0 0 X X / shift left
+ *  BE  shift left / mask 0 0 X X
+ */
+
+#if XCHAL_HAVE_BE
+
+#define HWORD_START	16
+#define	INSN_OP0	28
+#define	INSN_T		24
+#define	INSN_OP1	16
+
+.macro __ssa8r	r;		ssa8l	\r;		.endm
+.macro __sh	r, s;		srl	\r, \s;		.endm
+.macro __sl	r, s;		sll	\r, \s;		.endm
+.macro __exth	r, s;		extui	\r, \s, 0, 16;	.endm
+.macro __extl	r, s;		slli	\r, \s, 16;	.endm
+
+#else
+
+#define HWORD_START	0
+#define	INSN_OP0	0
+#define	INSN_T		4
+#define	INSN_OP1	12
+
+.macro __ssa8r	r;		ssa8b	\r;		.endm
+.macro __sh	r, s;		sll	\r, \s;		.endm
+.macro __sl	r, s;		srl	\r, \s;		.endm
+.macro __exth	r, s;		slli	\r, \s, 16;	.endm
+.macro __extl	r, s;		extui	\r, \s, 0, 16;	.endm
+
+#endif
+
+/*
+ *	xxxx xxxx = imm8 field
+ *	     yyyy = imm4 field
+ *	     ssss = s field
+ *	     tttt = t field
+ *
+ *	       		 16		    0
+ *		          -------------------
+ *	L32I.N		  yyyy ssss tttt 1000
+ *	S32I.N	          yyyy ssss tttt 1001
+ *
+ *	       23			    0
+ *		-----------------------------
+ *	L8UI	xxxx xxxx 0000 ssss tttt 0010
+ *	L16UI	xxxx xxxx 0001 ssss tttt 0010
+ *	L32I	xxxx xxxx 0010 ssss tttt 0010
+ *	XXX	          0011 ssss tttt 0010
+ *	XXX	          0100 ssss tttt 0010
+ *	S16I	xxxx xxxx 0101 ssss tttt 0010
+ *	S32I	xxxx xxxx 0110 ssss tttt 0010
+ *	XXX	          0111 ssss tttt 0010
+ *	XXX	          1000 ssss tttt 0010
+ *	L16SI	xxxx xxxx 1001 ssss tttt 0010
+ *	XXX	          1010           0010
+ *      **L32AI	xxxx xxxx 1011 ssss tttt 0010 unsupported
+ *	XXX	          1100           0010
+ *	XXX	          1101           0010
+ *	XXX	          1110           0010
+ *	**S32RI	xxxx xxxx 1111 ssss tttt 0010 unsupported
+ *		-----------------------------
+ *                           ^         ^    ^
+ *    sub-opcode (NIBBLE_R) -+         |    |
+ *       t field (NIBBLE_T) -----------+    |
+ *  major opcode (NIBBLE_OP0) --------------+
+ */
+
+#define OP0_L32I_N	0x8		/* load immediate narrow */
+#define OP0_S32I_N	0x9		/* store immediate narrow */
+#define OP0_LSAI	0x2		/* load/store */
+#define OP1_SI_MASK	0x4		/* OP1 bit set for stores */
+#define OP1_SI_BIT	2		/* OP1 bit number for stores */
+
+#define OP1_L8UI	0x0
+#define OP1_L32I	0x2
+#define OP1_L16UI	0x1
+#define OP1_L16SI	0x9
+#define OP1_L32AI	0xb
+
+#define OP1_S32I	0x6
+#define OP1_S16I	0x5
+#define OP1_S32RI	0xf
+
+/*
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+	.literal_position
+#ifdef CONFIG_XTENSA_LOAD_STORE
+ENTRY(fast_load_store)
+
+	call0	.Lsave_and_load_instruction
+
+	/* Analyze the instruction (load or store?). */
+
+	extui	a0, a4, INSN_OP0, 4	# get insn.op0 nibble
+
+#if XCHAL_HAVE_DENSITY
+	_beqi	a0, OP0_L32I_N, 1f	# L32I.N, jump
+#endif
+	bnei	a0, OP0_LSAI, .Linvalid_instruction
+	/* 'store indicator bit' set, jump */
+	bbsi.l	a4, OP1_SI_BIT + INSN_OP1, .Linvalid_instruction
+
+1:
+	movi	a3, ~3
+	and	a3, a3, a8		# align memory address
+
+	__ssa8	a8
+
+#ifdef CONFIG_MMU
+	/* l32e can't be used here even when it's available. */
+	rsr	a6, ps
+	bbci.l	a6, PS_UM_BIT, 1f # kernel mode
+	movi	a5, 8
+	access_ok a3, a5, a6, a2, .Linvalid_instruction
+1:
+#endif
+	l32i	a5, a3, 0
+	l32i	a6, a3, 4
+	__src_b	a3, a5, a6		# a3 has the data word
+
+#if XCHAL_HAVE_DENSITY
+	addi	a7, a7, 2		# increment PC (assume 16-bit insn)
+	_beqi	a0, OP0_L32I_N, .Lload_w# l32i.n: jump
+	addi	a7, a7, 1
+#else
+	addi	a7, a7, 3
+#endif
+
+	extui	a5, a4, INSN_OP1, 4
+	_beqi	a5, OP1_L32I, .Lload_w
+	bnei	a5, OP1_L8UI, .Lload16
+	extui	a3, a3, 0, 8
+	j	.Lload_w
+
+ENDPROC(fast_load_store)
+#endif
+
+/*
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+#ifdef ANY_EXCEPTION_HANDLER
+ENTRY(fast_unaligned)
+
+	call0	.Lsave_and_load_instruction
+
+	/* Analyze the instruction (load or store?). */
+
+	extui	a5, a4, INSN_OP0, 4	# get insn.op0 nibble
+
+#if XCHAL_HAVE_DENSITY
+	_beqi	a5, OP0_L32I_N, .Lload	# L32I.N, jump
+	addi	a6, a5, -OP0_S32I_N
+	_beqz	a6, .Lstore		# S32I.N, do a store
+#endif
+	/* 'store indicator bit' not set, jump */
+	_bbci.l	a4, OP1_SI_BIT + INSN_OP1, .Lload
+
+#ifdef STORE_EXCEPTION_HANDLER
+
+	/* Store: Jump to table entry to get the value in the source register.*/
+
+.Lstore:movi	a5, .Lstore_table	# table
+	extui	a6, a4, INSN_T, 4	# get source register
+	addx8	a5, a6, a5
+	jx	a5			# jump into table
+#endif
+#ifdef LOAD_EXCEPTION_HANDLER
+
+	/* Load: Load memory address. */
+
+.Lload: movi	a3, ~3
+	and	a3, a3, a8		# align memory address
+
+	__ssa8	a8
+#ifdef UNALIGNED_USER_EXCEPTION
+	addi	a3, a3, 8
+	l32e	a5, a3, -8
+	l32e	a6, a3, -4
+#else
+	l32i	a5, a3, 0
+	l32i	a6, a3, 4
+#endif
+	__src_b	a3, a5, a6		# a3 has the data word
+
+#if XCHAL_HAVE_DENSITY
+	addi	a7, a7, 2		# increment PC (assume 16-bit insn)
+
+	extui	a5, a4, INSN_OP0, 4
+	_beqi	a5, OP0_L32I_N, .Lload_w# l32i.n: jump
+
+	addi	a7, a7, 1
+#else
+	addi	a7, a7, 3
+#endif
+
+	extui	a5, a4, INSN_OP1, 4
+	_beqi	a5, OP1_L32I, .Lload_w	# l32i: jump
+#endif
+#ifdef LOAD_EXCEPTION_HANDLER
+.Lload16:
+	extui	a3, a3, 0, 16		# extract lower 16 bits
+	_beqi	a5, OP1_L16UI, .Lload_w
+	addi	a5, a5, -OP1_L16SI
+	_bnez	a5, .Linvalid_instruction
+
+	/* sign extend value */
+#if XCHAL_HAVE_SEXT
+	sext	a3, a3, 15
+#else
+	slli	a3, a3, 16
+	srai	a3, a3, 16
+#endif
+
+	/* Set target register. */
+
+.Lload_w:
+	extui	a4, a4, INSN_T, 4	# extract target register
+	movi	a5, .Lload_table
+	addx8	a4, a4, a5
+	jx	a4			# jump to entry for target register
+
+	.align	8
+.Lload_table:
+	s32i	a3, a2, PT_AREG0;	_j .Lexit;	.align 8
+	mov	a1, a3;			_j .Lexit;	.align 8 # fishy??
+	s32i	a3, a2, PT_AREG2;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG3;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG4;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG5;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG6;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG7;	_j .Lexit;	.align 8
+	s32i	a3, a2, PT_AREG8;	_j .Lexit;	.align 8
+	mov	a9, a3		;	_j .Lexit;	.align 8
+	mov	a10, a3		;	_j .Lexit;	.align 8
+	mov	a11, a3		;	_j .Lexit;	.align 8
+	mov	a12, a3		;	_j .Lexit;	.align 8
+	mov	a13, a3		;	_j .Lexit;	.align 8
+	mov	a14, a3		;	_j .Lexit;	.align 8
+	mov	a15, a3		;	_j .Lexit;	.align 8
+#endif
+#ifdef STORE_EXCEPTION_HANDLER
+.Lstore_table:
+	l32i	a3, a2, PT_AREG0;	_j .Lstore_w;	.align 8
+	mov	a3, a1;			_j .Lstore_w;	.align 8	# fishy??
+	l32i	a3, a2, PT_AREG2;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG3;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG4;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG5;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG6;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG7;	_j .Lstore_w;	.align 8
+	l32i	a3, a2, PT_AREG8;	_j .Lstore_w;	.align 8
+	mov	a3, a9		;	_j .Lstore_w;	.align 8
+	mov	a3, a10		;	_j .Lstore_w;	.align 8
+	mov	a3, a11		;	_j .Lstore_w;	.align 8
+	mov	a3, a12		;	_j .Lstore_w;	.align 8
+	mov	a3, a13		;	_j .Lstore_w;	.align 8
+	mov	a3, a14		;	_j .Lstore_w;	.align 8
+	mov	a3, a15		;	_j .Lstore_w;	.align 8
+#endif
+
+	/* We cannot handle this exception. */
+
+	.extern _kernel_exception
+.Linvalid_instruction:
+
+	movi	a4, 0
+	rsr	a3, excsave1
+	s32i	a4, a3, EXC_TABLE_FIXUP
+
+	/* Restore a4...a8 and SAR, set SP, and jump to default exception. */
+
+	l32i	a0, a2, PT_SAR
+	l32i	a8, a2, PT_AREG8
+	l32i	a7, a2, PT_AREG7
+	l32i	a6, a2, PT_AREG6
+	l32i	a5, a2, PT_AREG5
+	l32i	a4, a2, PT_AREG4
+	wsr	a0, sar
+	mov	a1, a2
+
+	rsr	a0, ps
+	bbsi.l  a0, PS_UM_BIT, 2f     # jump if user mode
+
+	movi	a0, _kernel_exception
+	jx	a0
+
+2:	movi	a0, _user_exception
+	jx	a0
+
+#ifdef STORE_EXCEPTION_HANDLER
+
+	# a7: instruction pointer, a4: instruction, a3: value
+.Lstore_w:
+	movi	a6, 0			# mask: ffffffff:00000000
+
+#if XCHAL_HAVE_DENSITY
+	addi	a7, a7, 2		# incr. PC,assume 16-bit instruction
+
+	extui	a5, a4, INSN_OP0, 4	# extract OP0
+	addi	a5, a5, -OP0_S32I_N
+	_beqz	a5, 1f			# s32i.n: jump
+
+	addi	a7, a7, 1		# increment PC, 32-bit instruction
+#else
+	addi	a7, a7, 3		# increment PC, 32-bit instruction
+#endif
+
+	extui	a5, a4, INSN_OP1, 4	# extract OP1
+	_beqi	a5, OP1_S32I, 1f	# jump if 32 bit store
+	_bnei	a5, OP1_S16I, .Linvalid_instruction
+
+	movi	a5, -1
+	__extl	a3, a3			# get 16-bit value
+	__exth	a6, a5			# get 16-bit mask ffffffff:ffff0000
+
+	/* Get memory address */
+
+1:
+	movi	a4, ~3
+	and	a4, a4, a8		# align memory address
+
+	/* Insert value into memory */
+
+	movi	a5, -1			# mask: ffffffff:XXXX0000
+#ifdef UNALIGNED_USER_EXCEPTION
+	addi	a4, a4, 8
+#endif
+
+	__ssa8r a8
+	__src_b	a8, a5, a6		# lo-mask  F..F0..0 (BE) 0..0F..F (LE)
+	__src_b	a6, a6, a5		# hi-mask  0..0F..F (BE) F..F0..0 (LE)
+#ifdef UNALIGNED_USER_EXCEPTION
+	l32e	a5, a4, -8
+#else
+	l32i	a5, a4, 0		# load lower address word
+#endif
+	and	a5, a5, a8		# mask
+	__sh	a8, a3 			# shift value
+	or	a5, a5, a8		# or with original value
+#ifdef UNALIGNED_USER_EXCEPTION
+	s32e	a5, a4, -8
+	l32e	a8, a4, -4
+#else
+	s32i	a5, a4, 0		# store
+	l32i	a8, a4, 4		# same for upper address word
+#endif
+	__sl	a5, a3
+	and	a6, a8, a6
+	or	a6, a6, a5
+#ifdef UNALIGNED_USER_EXCEPTION
+	s32e	a6, a4, -4
+#else
+	s32i	a6, a4, 4
+#endif
+#endif
+
+.Lexit:
+#if XCHAL_HAVE_LOOPS
+	rsr	a4, lend		# check if we reached LEND
+	bne	a7, a4, 1f
+	rsr	a4, lcount		# and LCOUNT != 0
+	beqz	a4, 1f
+	addi	a4, a4, -1		# decrement LCOUNT and set
+	rsr	a7, lbeg		# set PC to LBEGIN
+	wsr	a4, lcount
+#endif
+
+1:	wsr	a7, epc1		# skip emulated instruction
+
+	/* Update icount if we're single-stepping in userspace. */
+	rsr	a4, icountlevel
+	beqz	a4, 1f
+	bgeui	a4, LOCKLEVEL + 1, 1f
+	rsr	a4, icount
+	addi	a4, a4, 1
+	wsr	a4, icount
+1:
+	movi	a4, 0
+	rsr	a3, excsave1
+	s32i	a4, a3, EXC_TABLE_FIXUP
+
+	/* Restore working register */
+
+	l32i	a0, a2, PT_SAR
+	l32i	a8, a2, PT_AREG8
+	l32i	a7, a2, PT_AREG7
+	l32i	a6, a2, PT_AREG6
+	l32i	a5, a2, PT_AREG5
+	l32i	a4, a2, PT_AREG4
+	l32i	a3, a2, PT_AREG3
+
+	/* restore SAR and return */
+
+	wsr	a0, sar
+	l32i	a0, a2, PT_AREG0
+	l32i	a2, a2, PT_AREG2
+	rfe
+
+	.align	4
+.Lsave_and_load_instruction:
+
+	/* Save some working register */
+
+	s32i	a3, a2, PT_AREG3
+	s32i	a4, a2, PT_AREG4
+	s32i	a5, a2, PT_AREG5
+	s32i	a6, a2, PT_AREG6
+	s32i	a7, a2, PT_AREG7
+	s32i	a8, a2, PT_AREG8
+
+	rsr	a4, depc
+	s32i	a4, a2, PT_AREG2
+
+	rsr	a5, sar
+	s32i	a5, a2, PT_SAR
+
+	rsr	a3, excsave1
+	movi	a4, fast_unaligned_fixup
+	s32i	a4, a3, EXC_TABLE_FIXUP
+
+	rsr	a8, excvaddr		# load unaligned memory address
+
+	/* Now, identify one of the following load/store instructions.
+	 *
+	 * The only possible danger of a double exception on the
+	 * following l32i instructions is kernel code in vmalloc
+	 * memory. The processor was just executing at the EPC_1
+	 * address, and indeed, already fetched the instruction.  That
+	 * guarantees a TLB mapping, which hasn't been replaced by
+	 * this unaligned exception handler that uses only static TLB
+	 * mappings. However, high-level interrupt handlers might
+	 * modify TLB entries, so for the generic case, we register a
+	 * TABLE_FIXUP handler here, too.
+	 */
+
+	/* a3...a6 saved on stack, a2 = SP */
+
+	/* Extract the instruction that caused the unaligned access. */
+
+	rsr	a7, epc1	# load exception address
+	movi	a3, ~3
+	and	a3, a3, a7	# mask lower bits
+
+	l32i	a4, a3, 0	# load 2 words
+	l32i	a5, a3, 4
+
+	__ssa8	a7
+	__src_b	a4, a4, a5	# a4 has the instruction
+
+	ret
+
+ENDPROC(fast_unaligned)
+
+ENTRY(fast_unaligned_fixup)
+
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	wsr	a3, excsave1
+
+	l32i	a8, a2, PT_AREG8
+	l32i	a7, a2, PT_AREG7
+	l32i	a6, a2, PT_AREG6
+	l32i	a5, a2, PT_AREG5
+	l32i	a4, a2, PT_SAR
+	l32i	a0, a2, PT_AREG2
+	wsr	a4, sar
+	wsr	a0, depc			# restore depc and a0
+	l32i	a4, a2, PT_AREG4
+
+	rsr	a0, exccause
+	s32i	a0, a2, PT_DEPC			# mark as a regular exception
+
+	rsr	a0, ps
+	bbsi.l  a0, PS_UM_BIT, 1f		# jump if user mode
+
+	rsr	a0, exccause
+	addx4	a0, a0, a3              	# find entry in table
+	l32i	a0, a0, EXC_TABLE_FAST_KERNEL   # load handler
+	l32i	a3, a2, PT_AREG3
+	jx	a0
+1:
+	rsr	a0, exccause
+	addx4	a0, a0, a3              	# find entry in table
+	l32i	a0, a0, EXC_TABLE_FAST_USER     # load handler
+	l32i	a3, a2, PT_AREG3
+	jx	a0
+
+ENDPROC(fast_unaligned_fixup)
+#endif
diff --git a/arch/xtensa/kernel/asm-offsets.c b/arch/xtensa/kernel/asm-offsets.c
new file mode 100644
index 000000000..cfbced95e
--- /dev/null
+++ b/arch/xtensa/kernel/asm-offsets.c
@@ -0,0 +1,165 @@
+/*
+ * arch/xtensa/kernel/asm-offsets.c
+ *
+ * Generates definitions from c-type structures used by assembly sources.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2005 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ */
+#define COMPILE_OFFSETS
+
+#include <asm/processor.h>
+#include <asm/coprocessor.h>
+
+#include <linux/types.h>
+#include <linux/stddef.h>
+#include <linux/thread_info.h>
+#include <linux/ptrace.h>
+#include <linux/mm.h>
+#include <linux/kbuild.h>
+#include <linux/suspend.h>
+
+#include <asm/ptrace.h>
+#include <asm/traps.h>
+#include <linux/uaccess.h>
+
+int main(void)
+{
+	/* struct pt_regs */
+	DEFINE(PT_PC, offsetof (struct pt_regs, pc));
+	DEFINE(PT_PS, offsetof (struct pt_regs, ps));
+	DEFINE(PT_DEPC, offsetof (struct pt_regs, depc));
+	DEFINE(PT_EXCCAUSE, offsetof (struct pt_regs, exccause));
+	DEFINE(PT_EXCVADDR, offsetof (struct pt_regs, excvaddr));
+	DEFINE(PT_DEBUGCAUSE, offsetof (struct pt_regs, debugcause));
+	DEFINE(PT_WMASK, offsetof (struct pt_regs, wmask));
+	DEFINE(PT_LBEG, offsetof (struct pt_regs, lbeg));
+	DEFINE(PT_LEND, offsetof (struct pt_regs, lend));
+	DEFINE(PT_LCOUNT, offsetof (struct pt_regs, lcount));
+	DEFINE(PT_SAR, offsetof (struct pt_regs, sar));
+	DEFINE(PT_ICOUNTLEVEL, offsetof (struct pt_regs, icountlevel));
+	DEFINE(PT_SYSCALL, offsetof (struct pt_regs, syscall));
+	DEFINE(PT_SCOMPARE1, offsetof(struct pt_regs, scompare1));
+	DEFINE(PT_THREADPTR, offsetof(struct pt_regs, threadptr));
+	DEFINE(PT_AREG, offsetof (struct pt_regs, areg[0]));
+	DEFINE(PT_AREG0, offsetof (struct pt_regs, areg[0]));
+	DEFINE(PT_AREG1, offsetof (struct pt_regs, areg[1]));
+	DEFINE(PT_AREG2, offsetof (struct pt_regs, areg[2]));
+	DEFINE(PT_AREG3, offsetof (struct pt_regs, areg[3]));
+	DEFINE(PT_AREG4, offsetof (struct pt_regs, areg[4]));
+	DEFINE(PT_AREG5, offsetof (struct pt_regs, areg[5]));
+	DEFINE(PT_AREG6, offsetof (struct pt_regs, areg[6]));
+	DEFINE(PT_AREG7, offsetof (struct pt_regs, areg[7]));
+	DEFINE(PT_AREG8, offsetof (struct pt_regs, areg[8]));
+	DEFINE(PT_AREG9, offsetof (struct pt_regs, areg[9]));
+	DEFINE(PT_AREG10, offsetof (struct pt_regs, areg[10]));
+	DEFINE(PT_AREG11, offsetof (struct pt_regs, areg[11]));
+	DEFINE(PT_AREG12, offsetof (struct pt_regs, areg[12]));
+	DEFINE(PT_AREG13, offsetof (struct pt_regs, areg[13]));
+	DEFINE(PT_AREG14, offsetof (struct pt_regs, areg[14]));
+	DEFINE(PT_AREG15, offsetof (struct pt_regs, areg[15]));
+	DEFINE(PT_WINDOWBASE, offsetof (struct pt_regs, windowbase));
+	DEFINE(PT_WINDOWSTART, offsetof(struct pt_regs, windowstart));
+	DEFINE(PT_KERNEL_SIZE, offsetof(struct pt_regs, areg[16]));
+	DEFINE(PT_AREG_END, offsetof (struct pt_regs, areg[XCHAL_NUM_AREGS]));
+	DEFINE(PT_USER_SIZE, offsetof(struct pt_regs, areg[XCHAL_NUM_AREGS]));
+	DEFINE(PT_XTREGS_OPT, offsetof(struct pt_regs, xtregs_opt));
+	DEFINE(XTREGS_OPT_SIZE, sizeof(xtregs_opt_t));
+
+	/* struct task_struct */
+	DEFINE(TASK_PTRACE, offsetof (struct task_struct, ptrace));
+	DEFINE(TASK_MM, offsetof (struct task_struct, mm));
+	DEFINE(TASK_ACTIVE_MM, offsetof (struct task_struct, active_mm));
+	DEFINE(TASK_PID, offsetof (struct task_struct, pid));
+	DEFINE(TASK_THREAD, offsetof (struct task_struct, thread));
+	DEFINE(TASK_THREAD_INFO, offsetof (struct task_struct, stack));
+#ifdef CONFIG_STACKPROTECTOR
+	DEFINE(TASK_STACK_CANARY, offsetof(struct task_struct, stack_canary));
+#endif
+	DEFINE(TASK_STRUCT_SIZE, sizeof (struct task_struct));
+
+	/* offsets in thread_info struct */
+	OFFSET(TI_TASK, thread_info, task);
+	OFFSET(TI_FLAGS, thread_info, flags);
+	OFFSET(TI_STSTUS, thread_info, status);
+	OFFSET(TI_CPU, thread_info, cpu);
+	OFFSET(TI_PRE_COUNT, thread_info, preempt_count);
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+	OFFSET(TI_PS_WOE_FIX_ADDR, thread_info, ps_woe_fix_addr);
+#endif
+
+	/* struct thread_info (offset from start_struct) */
+	DEFINE(THREAD_RA, offsetof (struct task_struct, thread.ra));
+	DEFINE(THREAD_SP, offsetof (struct task_struct, thread.sp));
+#if XCHAL_HAVE_EXCLUSIVE
+	DEFINE(THREAD_ATOMCTL8, offsetof (struct thread_info, atomctl8));
+#endif
+	DEFINE(THREAD_CPENABLE, offsetof(struct thread_info, cpenable));
+	DEFINE(THREAD_CPU, offsetof(struct thread_info, cpu));
+	DEFINE(THREAD_CP_OWNER_CPU, offsetof(struct thread_info, cp_owner_cpu));
+#if XTENSA_HAVE_COPROCESSORS
+	DEFINE(THREAD_XTREGS_CP0, offsetof(struct thread_info, xtregs_cp.cp0));
+	DEFINE(THREAD_XTREGS_CP1, offsetof(struct thread_info, xtregs_cp.cp1));
+	DEFINE(THREAD_XTREGS_CP2, offsetof(struct thread_info, xtregs_cp.cp2));
+	DEFINE(THREAD_XTREGS_CP3, offsetof(struct thread_info, xtregs_cp.cp3));
+	DEFINE(THREAD_XTREGS_CP4, offsetof(struct thread_info, xtregs_cp.cp4));
+	DEFINE(THREAD_XTREGS_CP5, offsetof(struct thread_info, xtregs_cp.cp5));
+	DEFINE(THREAD_XTREGS_CP6, offsetof(struct thread_info, xtregs_cp.cp6));
+	DEFINE(THREAD_XTREGS_CP7, offsetof(struct thread_info, xtregs_cp.cp7));
+#endif
+	DEFINE(THREAD_XTREGS_USER, offsetof (struct thread_info, xtregs_user));
+	DEFINE(XTREGS_USER_SIZE, sizeof(xtregs_user_t));
+
+	/* struct mm_struct */
+	DEFINE(MM_USERS, offsetof(struct mm_struct, mm_users));
+	DEFINE(MM_PGD, offsetof (struct mm_struct, pgd));
+	DEFINE(MM_CONTEXT, offsetof (struct mm_struct, context));
+
+	/* struct page */
+	DEFINE(PAGE_FLAGS, offsetof(struct page, flags));
+
+	/* constants */
+	DEFINE(_CLONE_VM, CLONE_VM);
+	DEFINE(_CLONE_UNTRACED, CLONE_UNTRACED);
+	DEFINE(PG_ARCH_1, PG_arch_1);
+
+	/* struct debug_table */
+	DEFINE(DT_DEBUG_EXCEPTION,
+	       offsetof(struct debug_table, debug_exception));
+	DEFINE(DT_DEBUG_SAVE, offsetof(struct debug_table, debug_save));
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	DEFINE(DT_DBREAKC_SAVE, offsetof(struct debug_table, dbreakc_save));
+	DEFINE(DT_ICOUNT_SAVE, offsetof(struct debug_table, icount_save));
+	DEFINE(DT_ICOUNT_LEVEL_SAVE,
+	       offsetof(struct debug_table, icount_level_save));
+#endif
+
+	/* struct exc_table */
+	DEFINE(EXC_TABLE_KSTK, offsetof(struct exc_table, kstk));
+	DEFINE(EXC_TABLE_DOUBLE_SAVE, offsetof(struct exc_table, double_save));
+	DEFINE(EXC_TABLE_FIXUP, offsetof(struct exc_table, fixup));
+	DEFINE(EXC_TABLE_PARAM, offsetof(struct exc_table, fixup_param));
+#if XTENSA_HAVE_COPROCESSORS
+	DEFINE(EXC_TABLE_COPROCESSOR_OWNER,
+	       offsetof(struct exc_table, coprocessor_owner));
+#endif
+	DEFINE(EXC_TABLE_FAST_USER,
+	       offsetof(struct exc_table, fast_user_handler));
+	DEFINE(EXC_TABLE_FAST_KERNEL,
+	       offsetof(struct exc_table, fast_kernel_handler));
+	DEFINE(EXC_TABLE_DEFAULT, offsetof(struct exc_table, default_handler));
+
+#ifdef CONFIG_HIBERNATION
+	DEFINE(PBE_ADDRESS, offsetof(struct pbe, address));
+	DEFINE(PBE_ORIG_ADDRESS, offsetof(struct pbe, orig_address));
+	DEFINE(PBE_NEXT, offsetof(struct pbe, next));
+	DEFINE(PBE_SIZE, sizeof(struct pbe));
+#endif
+
+	return 0;
+}
diff --git a/arch/xtensa/kernel/coprocessor.S b/arch/xtensa/kernel/coprocessor.S
new file mode 100644
index 000000000..ef33e76e0
--- /dev/null
+++ b/arch/xtensa/kernel/coprocessor.S
@@ -0,0 +1,302 @@
+/*
+ * arch/xtensa/kernel/coprocessor.S
+ *
+ * Xtensa processor configuration-specific table of coprocessor and
+ * other custom register layout information.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2003 - 2007 Tensilica Inc.
+ */
+
+
+#include <linux/linkage.h>
+#include <asm/asm-offsets.h>
+#include <asm/asmmacro.h>
+#include <asm/coprocessor.h>
+#include <asm/current.h>
+#include <asm/regs.h>
+
+/*
+ * Rules for coprocessor state manipulation on SMP:
+ *
+ * - a task may have live coprocessors only on one CPU.
+ *
+ * - whether coprocessor context of task T is live on some CPU is
+ *   denoted by T's thread_info->cpenable.
+ *
+ * - non-zero thread_info->cpenable means that thread_info->cp_owner_cpu
+ *   is valid in the T's thread_info. Zero thread_info->cpenable means that
+ *   coprocessor context is valid in the T's thread_info.
+ *
+ * - if a coprocessor context of task T is live on CPU X, only CPU X changes
+ *   T's thread_info->cpenable, cp_owner_cpu and coprocessor save area.
+ *   This is done by making sure that for the task T with live coprocessor
+ *   on CPU X cpenable SR is 0 when T runs on any other CPU Y.
+ *   When fast_coprocessor exception is taken on CPU Y it goes to the
+ *   C-level do_coprocessor that uses IPI to make CPU X flush T's coprocessors.
+ */
+
+#if XTENSA_HAVE_COPROCESSORS
+
+/*
+ * Macros for lazy context switch. 
+ */
+
+#define SAVE_CP_REGS(x)							\
+	.if XTENSA_HAVE_COPROCESSOR(x);					\
+		.align 4;						\
+	.Lsave_cp_regs_cp##x:						\
+		xchal_cp##x##_store a2 a3 a4 a5 a6;			\
+		ret;							\
+	.endif
+
+#define LOAD_CP_REGS(x)							\
+	.if XTENSA_HAVE_COPROCESSOR(x);					\
+		.align 4;						\
+	.Lload_cp_regs_cp##x:						\
+		xchal_cp##x##_load a2 a3 a4 a5 a6;			\
+		ret;							\
+	.endif
+
+#define CP_REGS_TAB(x)							\
+	.if XTENSA_HAVE_COPROCESSOR(x);					\
+		.long .Lsave_cp_regs_cp##x;				\
+		.long .Lload_cp_regs_cp##x;				\
+	.else;								\
+		.long 0, 0;						\
+	.endif;								\
+	.long THREAD_XTREGS_CP##x
+
+#define CP_REGS_TAB_SAVE 0
+#define CP_REGS_TAB_LOAD 4
+#define CP_REGS_TAB_OFFSET 8
+
+	__XTENSA_HANDLER
+
+	SAVE_CP_REGS(0)
+	SAVE_CP_REGS(1)
+	SAVE_CP_REGS(2)
+	SAVE_CP_REGS(3)
+	SAVE_CP_REGS(4)
+	SAVE_CP_REGS(5)
+	SAVE_CP_REGS(6)
+	SAVE_CP_REGS(7)
+
+	LOAD_CP_REGS(0)
+	LOAD_CP_REGS(1)
+	LOAD_CP_REGS(2)
+	LOAD_CP_REGS(3)
+	LOAD_CP_REGS(4)
+	LOAD_CP_REGS(5)
+	LOAD_CP_REGS(6)
+	LOAD_CP_REGS(7)
+
+	.align 4
+.Lcp_regs_jump_table:
+	CP_REGS_TAB(0)
+	CP_REGS_TAB(1)
+	CP_REGS_TAB(2)
+	CP_REGS_TAB(3)
+	CP_REGS_TAB(4)
+	CP_REGS_TAB(5)
+	CP_REGS_TAB(6)
+	CP_REGS_TAB(7)
+
+/*
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+ENTRY(fast_coprocessor)
+
+	s32i	a3, a2, PT_AREG3
+
+#ifdef CONFIG_SMP
+	/*
+	 * Check if any coprocessor context is live on another CPU
+	 * and if so go through the C-level coprocessor exception handler
+	 * to flush it to memory.
+	 */
+	GET_THREAD_INFO (a0, a2)
+	l32i	a3, a0, THREAD_CPENABLE
+	beqz	a3, .Lload_local
+
+	/*
+	 * Pairs with smp_wmb in local_coprocessor_release_all
+	 * and with both memws below.
+	 */
+	memw
+	l32i	a3, a0, THREAD_CPU
+	l32i	a0, a0, THREAD_CP_OWNER_CPU
+	beq	a0, a3, .Lload_local
+
+	rsr	a0, ps
+	l32i	a3, a2, PT_AREG3
+	bbci.l	a0, PS_UM_BIT, 1f
+	call0	user_exception
+1:	call0	kernel_exception
+#endif
+
+	/* Save remaining registers a1-a3 and SAR */
+
+.Lload_local:
+	rsr	a3, sar
+	s32i	a1, a2, PT_AREG1
+	s32i	a3, a2, PT_SAR
+	mov	a1, a2
+	rsr	a2, depc
+	s32i	a2, a1, PT_AREG2
+
+	/* The hal macros require up to 4 temporary registers. We use a3..a6. */
+
+	s32i	a4, a1, PT_AREG4
+	s32i	a5, a1, PT_AREG5
+	s32i	a6, a1, PT_AREG6
+	s32i	a7, a1, PT_AREG7
+	s32i	a8, a1, PT_AREG8
+	s32i	a9, a1, PT_AREG9
+	s32i	a10, a1, PT_AREG10
+
+	/* Find coprocessor number. Subtract first CP EXCCAUSE from EXCCAUSE */
+
+	rsr	a3, exccause
+	addi	a3, a3, -EXCCAUSE_COPROCESSOR0_DISABLED
+
+	/* Set corresponding CPENABLE bit -> (sar:cp-index, a3: 1<<cp-index)*/
+
+	ssl	a3			# SAR: 32 - coprocessor_number
+	movi	a2, 1
+	rsr	a0, cpenable
+	sll	a2, a2
+	or	a0, a0, a2
+	wsr	a0, cpenable
+	rsync
+
+	/* Get coprocessor save/load table entry (a7). */
+
+	movi	a7, .Lcp_regs_jump_table
+	addx8	a7, a3, a7
+	addx4	a7, a3, a7
+
+	/* Retrieve previous owner (a8). */
+
+	rsr	a0, excsave1		# exc_table
+	addx4	a0, a3, a0		# entry for CP
+	l32i	a8, a0, EXC_TABLE_COPROCESSOR_OWNER
+
+	/* Set new owner (a9). */
+
+	GET_THREAD_INFO (a9, a1)
+	l32i	a4, a9, THREAD_CPU
+	s32i	a9, a0, EXC_TABLE_COPROCESSOR_OWNER
+	s32i	a4, a9, THREAD_CP_OWNER_CPU
+
+	/*
+	 * Enable coprocessor for the new owner. (a2 = 1 << CP number)
+	 * This can be done before loading context into the coprocessor.
+	 */
+	l32i	a4, a9, THREAD_CPENABLE
+	or	a4, a4, a2
+
+	/*
+	 * Make sure THREAD_CP_OWNER_CPU is in memory before updating
+	 * THREAD_CPENABLE
+	 */
+	memw				# (2)
+	s32i	a4, a9, THREAD_CPENABLE
+
+	beqz	a8, 1f			# skip 'save' if no previous owner
+
+	/* Disable coprocessor for previous owner. (a2 = 1 << CP number) */
+
+	l32i	a10, a8, THREAD_CPENABLE
+	xor	a10, a10, a2
+
+	/* Get context save area and call save routine. */
+
+	l32i	a2, a7, CP_REGS_TAB_OFFSET
+	l32i	a3, a7, CP_REGS_TAB_SAVE
+	add	a2, a2, a8
+	callx0	a3
+
+	/*
+	 * Make sure coprocessor context and THREAD_CP_OWNER_CPU are in memory
+	 * before updating THREAD_CPENABLE
+	 */
+	memw				# (3)
+	s32i	a10, a8, THREAD_CPENABLE
+1:
+	/* Get context save area and call load routine. */
+
+	l32i	a2, a7, CP_REGS_TAB_OFFSET
+	l32i	a3, a7, CP_REGS_TAB_LOAD
+	add	a2, a2, a9
+	callx0	a3
+
+	/* Restore all registers and return from exception handler. */
+
+	l32i	a10, a1, PT_AREG10
+	l32i	a9, a1, PT_AREG9
+	l32i	a8, a1, PT_AREG8
+	l32i	a7, a1, PT_AREG7
+	l32i	a6, a1, PT_AREG6
+	l32i	a5, a1, PT_AREG5
+	l32i	a4, a1, PT_AREG4
+
+	l32i	a0, a1, PT_SAR
+	l32i	a3, a1, PT_AREG3
+	l32i	a2, a1, PT_AREG2
+	wsr	a0, sar
+	l32i	a0, a1, PT_AREG0
+	l32i	a1, a1, PT_AREG1
+
+	rfe
+
+ENDPROC(fast_coprocessor)
+
+	.text
+
+/*
+ * coprocessor_flush(struct thread_info*, index)
+ *                             a2        a3
+ *
+ * Save coprocessor registers for coprocessor 'index'.
+ * The register values are saved to or loaded from the coprocessor area
+ * inside the task_info structure.
+ *
+ * Note that this function doesn't update the coprocessor_owner information!
+ *
+ */
+
+ENTRY(coprocessor_flush)
+
+	abi_entry_default
+
+	movi	a4, .Lcp_regs_jump_table
+	addx8	a4, a3, a4
+	addx4	a3, a3, a4
+	l32i	a4, a3, CP_REGS_TAB_SAVE
+	beqz	a4, 1f
+	l32i	a3, a3, CP_REGS_TAB_OFFSET
+	add	a2, a2, a3
+	mov	a7, a0
+	callx0	a4
+	mov	a0, a7
+1:
+	abi_ret_default
+
+ENDPROC(coprocessor_flush)
+
+#endif /* XTENSA_HAVE_COPROCESSORS */
diff --git a/arch/xtensa/kernel/entry.S b/arch/xtensa/kernel/entry.S
new file mode 100644
index 000000000..272fff587
--- /dev/null
+++ b/arch/xtensa/kernel/entry.S
@@ -0,0 +1,2264 @@
+/*
+ * Low-level exception handling
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2004 - 2008 by Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ *
+ */
+
+#include <linux/linkage.h>
+#include <linux/pgtable.h>
+#include <asm/asm-offsets.h>
+#include <asm/asmmacro.h>
+#include <asm/processor.h>
+#include <asm/coprocessor.h>
+#include <asm/thread_info.h>
+#include <asm/asm-uaccess.h>
+#include <asm/unistd.h>
+#include <asm/ptrace.h>
+#include <asm/current.h>
+#include <asm/page.h>
+#include <asm/signal.h>
+#include <asm/tlbflush.h>
+#include <variant/tie-asm.h>
+
+/*
+ * Macro to find first bit set in WINDOWBASE from the left + 1
+ *
+ * 100....0 -> 1
+ * 010....0 -> 2
+ * 000....1 -> WSBITS
+ */
+
+	.macro ffs_ws bit mask
+
+#if XCHAL_HAVE_NSA
+	nsau    \bit, \mask			# 32-WSBITS ... 31 (32 iff 0)
+	addi    \bit, \bit, WSBITS - 32 + 1   	# uppest bit set -> return 1
+#else
+	movi    \bit, WSBITS
+#if WSBITS > 16
+	_bltui  \mask, 0x10000, 99f
+	addi    \bit, \bit, -16
+	extui   \mask, \mask, 16, 16
+#endif
+#if WSBITS > 8
+99:	_bltui  \mask, 0x100, 99f
+	addi    \bit, \bit, -8
+	srli    \mask, \mask, 8
+#endif
+99:	_bltui  \mask, 0x10, 99f
+	addi    \bit, \bit, -4
+	srli    \mask, \mask, 4
+99:	_bltui  \mask, 0x4, 99f
+	addi    \bit, \bit, -2
+	srli    \mask, \mask, 2
+99:	_bltui  \mask, 0x2, 99f
+	addi    \bit, \bit, -1
+99:
+
+#endif
+	.endm
+
+
+	.macro	irq_save flags tmp
+#if XTENSA_FAKE_NMI
+#if defined(CONFIG_DEBUG_KERNEL) && (LOCKLEVEL | TOPLEVEL) >= XCHAL_DEBUGLEVEL
+	rsr	\flags, ps
+	extui	\tmp, \flags, PS_INTLEVEL_SHIFT, PS_INTLEVEL_WIDTH
+	bgei	\tmp, LOCKLEVEL, 99f
+	rsil	\tmp, LOCKLEVEL
+99:
+#else
+	movi	\tmp, LOCKLEVEL
+	rsr	\flags, ps
+	or	\flags, \flags, \tmp
+	xsr	\flags, ps
+	rsync
+#endif
+#else
+	rsil	\flags, LOCKLEVEL
+#endif
+	.endm
+
+/* ----------------- DEFAULT FIRST LEVEL EXCEPTION HANDLERS ----------------- */
+
+/*
+ * First-level exception handler for user exceptions.
+ * Save some special registers, extra states and all registers in the AR
+ * register file that were in use in the user task, and jump to the common
+ * exception code.
+ * We save SAR (used to calculate WMASK), and WB and WS (we don't have to
+ * save them for kernel exceptions).
+ *
+ * Entry condition for user_exception:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original value in depc
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ *
+ * Entry condition for _user_exception:
+ *
+ *   a0-a3 and depc have been saved to PT_AREG0...PT_AREG3 and PT_DEPC
+ *   excsave has been restored, and
+ *   stack pointer (a1) has been set.
+ *
+ * Note: _user_exception might be at an odd address. Don't use call0..call12
+ */
+	.literal_position
+
+ENTRY(user_exception)
+
+	/* Save a1, a2, a3, and set SP. */
+
+	rsr	a0, depc
+	s32i	a1, a2, PT_AREG1
+	s32i	a0, a2, PT_AREG2
+	s32i	a3, a2, PT_AREG3
+	mov	a1, a2
+
+	.globl _user_exception
+_user_exception:
+
+	/* Save SAR and turn off single stepping */
+
+	movi	a2, 0
+	wsr	a2, depc		# terminate user stack trace with 0
+	rsr	a3, sar
+	xsr	a2, icountlevel
+	s32i	a3, a1, PT_SAR
+	s32i	a2, a1, PT_ICOUNTLEVEL
+
+#if XCHAL_HAVE_THREADPTR
+	rur	a2, threadptr
+	s32i	a2, a1, PT_THREADPTR
+#endif
+
+	/* Rotate ws so that the current windowbase is at bit0. */
+	/* Assume ws = xxwww1yyyy. Rotate ws right, so that a2 = yyyyxxwww1 */
+
+#if defined(USER_SUPPORT_WINDOWED)
+	rsr	a2, windowbase
+	rsr	a3, windowstart
+	ssr	a2
+	s32i	a2, a1, PT_WINDOWBASE
+	s32i	a3, a1, PT_WINDOWSTART
+	slli	a2, a3, 32-WSBITS
+	src	a2, a3, a2
+	srli	a2, a2, 32-WSBITS
+	s32i	a2, a1, PT_WMASK	# needed for restoring registers
+#else
+	movi	a2, 0
+	movi	a3, 1
+	s32i	a2, a1, PT_WINDOWBASE
+	s32i	a3, a1, PT_WINDOWSTART
+	s32i	a3, a1, PT_WMASK
+#endif
+
+	/* Save only live registers. */
+
+UABI_W	_bbsi.l	a2, 1, .Lsave_window_registers
+	s32i	a4, a1, PT_AREG4
+	s32i	a5, a1, PT_AREG5
+	s32i	a6, a1, PT_AREG6
+	s32i	a7, a1, PT_AREG7
+UABI_W	_bbsi.l	a2, 2, .Lsave_window_registers
+	s32i	a8, a1, PT_AREG8
+	s32i	a9, a1, PT_AREG9
+	s32i	a10, a1, PT_AREG10
+	s32i	a11, a1, PT_AREG11
+UABI_W	_bbsi.l	a2, 3, .Lsave_window_registers
+	s32i	a12, a1, PT_AREG12
+	s32i	a13, a1, PT_AREG13
+	s32i	a14, a1, PT_AREG14
+	s32i	a15, a1, PT_AREG15
+
+#if defined(USER_SUPPORT_WINDOWED)
+	/* If only one valid frame skip saving regs. */
+
+	beqi	a2, 1, common_exception
+
+	/* Save the remaining registers.
+	 * We have to save all registers up to the first '1' from
+	 * the right, except the current frame (bit 0).
+	 * Assume a2 is:  001001000110001
+	 * All register frames starting from the top field to the marked '1'
+	 * must be saved.
+	 */
+.Lsave_window_registers:
+	addi	a3, a2, -1		# eliminate '1' in bit 0: yyyyxxww0
+	neg	a3, a3			# yyyyxxww0 -> YYYYXXWW1+1
+	and	a3, a3, a2		# max. only one bit is set
+
+	/* Find number of frames to save */
+
+	ffs_ws	a0, a3			# number of frames to the '1' from left
+
+	/* Store information into WMASK:
+	 * bits 0..3: xxx1 masked lower 4 bits of the rotated windowstart,
+	 * bits 4...: number of valid 4-register frames
+	 */
+
+	slli	a3, a0, 4		# number of frames to save in bits 8..4
+	extui	a2, a2, 0, 4		# mask for the first 16 registers
+	or	a2, a3, a2
+	s32i	a2, a1, PT_WMASK	# needed when we restore the reg-file
+
+	/* Save 4 registers at a time */
+
+1:	rotw	-1
+	s32i	a0, a5, PT_AREG_END - 16
+	s32i	a1, a5, PT_AREG_END - 12
+	s32i	a2, a5, PT_AREG_END - 8
+	s32i	a3, a5, PT_AREG_END - 4
+	addi	a0, a4, -1
+	addi	a1, a5, -16
+	_bnez	a0, 1b
+
+	/* WINDOWBASE still in SAR! */
+
+	rsr	a2, sar			# original WINDOWBASE
+	movi	a3, 1
+	ssl	a2
+	sll	a3, a3
+	wsr	a3, windowstart		# set corresponding WINDOWSTART bit
+	wsr	a2, windowbase		# and WINDOWSTART
+	rsync
+
+	/* We are back to the original stack pointer (a1) */
+#endif
+	/* Now, jump to the common exception handler. */
+
+	j	common_exception
+
+ENDPROC(user_exception)
+
+/*
+ * First-level exit handler for kernel exceptions
+ * Save special registers and the live window frame.
+ * Note: Even though we changes the stack pointer, we don't have to do a
+ *	 MOVSP here, as we do that when we return from the exception.
+ *	 (See comment in the kernel exception exit code)
+ *
+ * Entry condition for kernel_exception:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ *
+ * Entry condition for _kernel_exception:
+ *
+ *   a0-a3 and depc have been saved to PT_AREG0...PT_AREG3 and PT_DEPC
+ *   excsave has been restored, and
+ *   stack pointer (a1) has been set.
+ *
+ * Note: _kernel_exception might be at an odd address. Don't use call0..call12
+ */
+
+ENTRY(kernel_exception)
+
+	/* Save a1, a2, a3, and set SP. */
+
+	rsr	a0, depc		# get a2
+	s32i	a1, a2, PT_AREG1
+	s32i	a0, a2, PT_AREG2
+	s32i	a3, a2, PT_AREG3
+	mov	a1, a2
+
+	.globl _kernel_exception
+_kernel_exception:
+
+	/* Save SAR and turn off single stepping */
+
+	movi	a2, 0
+	rsr	a3, sar
+	xsr	a2, icountlevel
+	s32i	a3, a1, PT_SAR
+	s32i	a2, a1, PT_ICOUNTLEVEL
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	/* Rotate ws so that the current windowbase is at bit0. */
+	/* Assume ws = xxwww1yyyy. Rotate ws right, so that a2 = yyyyxxwww1 */
+
+	rsr	a2, windowbase		# don't need to save these, we only
+	rsr	a3, windowstart		# need shifted windowstart: windowmask
+	ssr	a2
+	slli	a2, a3, 32-WSBITS
+	src	a2, a3, a2
+	srli	a2, a2, 32-WSBITS
+	s32i	a2, a1, PT_WMASK	# needed for kernel_exception_exit
+#endif
+
+	/* Save only the live window-frame */
+
+KABI_W	_bbsi.l	a2, 1, 1f
+	s32i	a4, a1, PT_AREG4
+	s32i	a5, a1, PT_AREG5
+	s32i	a6, a1, PT_AREG6
+	s32i	a7, a1, PT_AREG7
+KABI_W	_bbsi.l	a2, 2, 1f
+	s32i	a8, a1, PT_AREG8
+	s32i	a9, a1, PT_AREG9
+	s32i	a10, a1, PT_AREG10
+	s32i	a11, a1, PT_AREG11
+KABI_W	_bbsi.l	a2, 3, 1f
+	s32i	a12, a1, PT_AREG12
+	s32i	a13, a1, PT_AREG13
+	s32i	a14, a1, PT_AREG14
+	s32i	a15, a1, PT_AREG15
+
+#ifdef __XTENSA_WINDOWED_ABI__
+	_bnei	a2, 1, 1f
+	/* Copy spill slots of a0 and a1 to imitate movsp
+	 * in order to keep exception stack continuous
+	 */
+	l32i	a3, a1, PT_KERNEL_SIZE
+	l32i	a0, a1, PT_KERNEL_SIZE + 4
+	s32e	a3, a1, -16
+	s32e	a0, a1, -12
+#endif
+1:
+	l32i	a0, a1, PT_AREG0	# restore saved a0
+	wsr	a0, depc
+
+/*
+ * This is the common exception handler.
+ * We get here from the user exception handler or simply by falling through
+ * from the kernel exception handler.
+ * Save the remaining special registers, switch to kernel mode, and jump
+ * to the second-level exception handler.
+ *
+ */
+
+common_exception:
+
+	/* Save some registers, disable loops and clear the syscall flag. */
+
+	rsr	a2, debugcause
+	rsr	a3, epc1
+	s32i	a2, a1, PT_DEBUGCAUSE
+	s32i	a3, a1, PT_PC
+
+	movi	a2, NO_SYSCALL
+	rsr	a3, excvaddr
+	s32i	a2, a1, PT_SYSCALL
+	movi	a2, 0
+	s32i	a3, a1, PT_EXCVADDR
+#if XCHAL_HAVE_LOOPS
+	xsr	a2, lcount
+	s32i	a2, a1, PT_LCOUNT
+#endif
+
+#if XCHAL_HAVE_EXCLUSIVE
+	/* Clear exclusive access monitor set by interrupted code */
+	clrex
+#endif
+
+	/* It is now save to restore the EXC_TABLE_FIXUP variable. */
+
+	rsr	a2, exccause
+	movi	a3, 0
+	rsr	a0, excsave1
+	s32i	a2, a1, PT_EXCCAUSE
+	s32i	a3, a0, EXC_TABLE_FIXUP
+
+	/* All unrecoverable states are saved on stack, now, and a1 is valid.
+	 * Now we can allow exceptions again. In case we've got an interrupt
+	 * PS.INTLEVEL is set to LOCKLEVEL disabling furhter interrupts,
+	 * otherwise it's left unchanged.
+	 *
+	 * Set PS(EXCM = 0, UM = 0, RING = 0, OWB = 0, WOE = 1, INTLEVEL = X)
+	 */
+
+	rsr	a3, ps
+	s32i	a3, a1, PT_PS		# save ps
+
+#if XTENSA_FAKE_NMI
+	/* Correct PS needs to be saved in the PT_PS:
+	 * - in case of exception or level-1 interrupt it's in the PS,
+	 *   and is already saved.
+	 * - in case of medium level interrupt it's in the excsave2.
+	 */
+	movi	a0, EXCCAUSE_MAPPED_NMI
+	extui	a3, a3, PS_INTLEVEL_SHIFT, PS_INTLEVEL_WIDTH
+	beq	a2, a0, .Lmedium_level_irq
+	bnei	a2, EXCCAUSE_LEVEL1_INTERRUPT, .Lexception
+	beqz	a3, .Llevel1_irq	# level-1 IRQ sets ps.intlevel to 0
+
+.Lmedium_level_irq:
+	rsr	a0, excsave2
+	s32i	a0, a1, PT_PS		# save medium-level interrupt ps
+	bgei	a3, LOCKLEVEL, .Lexception
+
+.Llevel1_irq:
+	movi	a3, LOCKLEVEL
+
+.Lexception:
+KABI_W	movi	a0, PS_WOE_MASK
+KABI_W	or	a3, a3, a0
+#else
+	addi	a2, a2, -EXCCAUSE_LEVEL1_INTERRUPT
+	movi	a0, LOCKLEVEL
+	extui	a3, a3, PS_INTLEVEL_SHIFT, PS_INTLEVEL_WIDTH
+					# a3 = PS.INTLEVEL
+	moveqz	a3, a0, a2		# a3 = LOCKLEVEL iff interrupt
+KABI_W	movi	a2, PS_WOE_MASK
+KABI_W	or	a3, a3, a2
+#endif
+
+	/* restore return address (or 0 if return to userspace) */
+	rsr	a0, depc
+	wsr	a3, ps
+	rsync				# PS.WOE => rsync => overflow
+
+	/* Save lbeg, lend */
+#if XCHAL_HAVE_LOOPS
+	rsr	a4, lbeg
+	rsr	a3, lend
+	s32i	a4, a1, PT_LBEG
+	s32i	a3, a1, PT_LEND
+#endif
+
+	/* Save SCOMPARE1 */
+
+#if XCHAL_HAVE_S32C1I
+	rsr     a3, scompare1
+	s32i    a3, a1, PT_SCOMPARE1
+#endif
+
+	/* Save optional registers. */
+
+	save_xtregs_opt a1 a3 a4 a5 a6 a7 PT_XTREGS_OPT
+	
+#ifdef CONFIG_TRACE_IRQFLAGS
+	rsr		abi_tmp0, ps
+	extui		abi_tmp0, abi_tmp0, PS_INTLEVEL_SHIFT, PS_INTLEVEL_WIDTH
+	beqz		abi_tmp0, 1f
+	abi_call	trace_hardirqs_off
+1:
+#endif
+#ifdef CONFIG_CONTEXT_TRACKING_USER
+	l32i		abi_tmp0, a1, PT_PS
+	bbci.l		abi_tmp0, PS_UM_BIT, 1f
+	abi_call	user_exit_callable
+1:
+#endif
+
+	/* Go to second-level dispatcher. Set up parameters to pass to the
+	 * exception handler and call the exception handler.
+	 */
+
+	l32i		abi_arg1, a1, PT_EXCCAUSE	# pass EXCCAUSE
+	rsr		abi_tmp0, excsave1
+	addx4		abi_tmp0, abi_arg1, abi_tmp0
+	l32i		abi_tmp0, abi_tmp0, EXC_TABLE_DEFAULT	# load handler
+	mov		abi_arg0, a1			# pass stack frame
+
+	/* Call the second-level handler */
+
+	abi_callx	abi_tmp0
+
+	/* Jump here for exception exit */
+	.global common_exception_return
+common_exception_return:
+
+#if XTENSA_FAKE_NMI
+	l32i		abi_tmp0, a1, PT_EXCCAUSE
+	movi		abi_tmp1, EXCCAUSE_MAPPED_NMI
+	l32i		abi_saved1, a1, PT_PS
+	beq		abi_tmp0, abi_tmp1, .Lrestore_state
+#endif
+.Ltif_loop:
+	irq_save	abi_tmp0, abi_tmp1
+#ifdef CONFIG_TRACE_IRQFLAGS
+	abi_call	trace_hardirqs_off
+#endif
+
+	/* Jump if we are returning from kernel exceptions. */
+
+	l32i		abi_saved1, a1, PT_PS
+	GET_THREAD_INFO(abi_tmp0, a1)
+	l32i		abi_saved0, abi_tmp0, TI_FLAGS
+	_bbci.l		abi_saved1, PS_UM_BIT, .Lexit_tif_loop_kernel
+
+	/* Specific to a user exception exit:
+	 * We need to check some flags for signal handling and rescheduling,
+	 * and have to restore WB and WS, extra states, and all registers
+	 * in the register file that were in use in the user task.
+	 * Note that we don't disable interrupts here. 
+	 */
+
+	_bbsi.l		abi_saved0, TIF_NEED_RESCHED, .Lresched
+	movi		abi_tmp0, _TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NOTIFY_SIGNAL
+	bnone		abi_saved0, abi_tmp0, .Lexit_tif_loop_user
+
+	l32i		abi_tmp0, a1, PT_DEPC
+	bgeui		abi_tmp0, VALID_DOUBLE_EXCEPTION_ADDRESS, .Lrestore_state
+
+	/* Call do_signal() */
+
+#ifdef CONFIG_TRACE_IRQFLAGS
+	abi_call	trace_hardirqs_on
+#endif
+	rsil		abi_tmp0, 0
+	mov		abi_arg0, a1
+	abi_call	do_notify_resume	# int do_notify_resume(struct pt_regs*)
+	j		.Ltif_loop
+
+.Lresched:
+#ifdef CONFIG_TRACE_IRQFLAGS
+	abi_call	trace_hardirqs_on
+#endif
+	rsil		abi_tmp0, 0
+	abi_call	schedule	# void schedule (void)
+	j		.Ltif_loop
+
+.Lexit_tif_loop_kernel:
+#ifdef CONFIG_PREEMPTION
+	_bbci.l		abi_saved0, TIF_NEED_RESCHED, .Lrestore_state
+
+	/* Check current_thread_info->preempt_count */
+
+	l32i		abi_tmp1, abi_tmp0, TI_PRE_COUNT
+	bnez		abi_tmp1, .Lrestore_state
+	abi_call	preempt_schedule_irq
+#endif
+	j		.Lrestore_state
+
+.Lexit_tif_loop_user:
+#ifdef CONFIG_CONTEXT_TRACKING_USER
+	abi_call	user_enter_callable
+#endif
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	_bbci.l		abi_saved0, TIF_DB_DISABLED, 1f
+	abi_call	restore_dbreak
+1:
+#endif
+#ifdef CONFIG_DEBUG_TLB_SANITY
+	l32i		abi_tmp0, a1, PT_DEPC
+	bgeui		abi_tmp0, VALID_DOUBLE_EXCEPTION_ADDRESS, .Lrestore_state
+	abi_call	check_tlb_sanity
+#endif
+
+.Lrestore_state:
+#ifdef CONFIG_TRACE_IRQFLAGS
+	extui		abi_tmp0, abi_saved1, PS_INTLEVEL_SHIFT, PS_INTLEVEL_WIDTH
+	bgei		abi_tmp0, LOCKLEVEL, 1f
+	abi_call	trace_hardirqs_on
+1:
+#endif
+	/*
+	 * Restore optional registers.
+	 * abi_arg* are used as temporary registers here.
+	 */
+
+	load_xtregs_opt a1 abi_tmp0 abi_arg0 abi_arg1 abi_arg2 abi_arg3 PT_XTREGS_OPT
+
+	/* Restore SCOMPARE1 */
+
+#if XCHAL_HAVE_S32C1I
+	l32i		abi_tmp0, a1, PT_SCOMPARE1
+	wsr		abi_tmp0, scompare1
+#endif
+	wsr		abi_saved1, ps		/* disable interrupts */
+	_bbci.l		abi_saved1, PS_UM_BIT, kernel_exception_exit
+
+user_exception_exit:
+
+	/* Restore the state of the task and return from the exception. */
+
+#if defined(USER_SUPPORT_WINDOWED)
+	/* Switch to the user thread WINDOWBASE. Save SP temporarily in DEPC */
+
+	l32i	a2, a1, PT_WINDOWBASE
+	l32i	a3, a1, PT_WINDOWSTART
+	wsr	a1, depc		# use DEPC as temp storage
+	wsr	a3, windowstart		# restore WINDOWSTART
+	ssr	a2			# preserve user's WB in the SAR
+	wsr	a2, windowbase		# switch to user's saved WB
+	rsync
+	rsr	a1, depc		# restore stack pointer
+	l32i	a2, a1, PT_WMASK	# register frames saved (in bits 4...9)
+	rotw	-1			# we restore a4..a7
+	_bltui	a6, 16, .Lclear_regs	# only have to restore current window?
+
+	/* The working registers are a0 and a3.  We are restoring to
+	 * a4..a7.  Be careful not to destroy what we have just restored.
+	 * Note: wmask has the format YYYYM:
+	 *       Y: number of registers saved in groups of 4
+	 *       M: 4 bit mask of first 16 registers
+	 */
+
+	mov	a2, a6
+	mov	a3, a5
+
+1:	rotw	-1			# a0..a3 become a4..a7
+	addi	a3, a7, -4*4		# next iteration
+	addi	a2, a6, -16		# decrementing Y in WMASK
+	l32i	a4, a3, PT_AREG_END + 0
+	l32i	a5, a3, PT_AREG_END + 4
+	l32i	a6, a3, PT_AREG_END + 8
+	l32i	a7, a3, PT_AREG_END + 12
+	_bgeui	a2, 16, 1b
+
+	/* Clear unrestored registers (don't leak anything to user-land */
+
+.Lclear_regs:
+	rsr	a0, windowbase
+	rsr	a3, sar
+	sub	a3, a0, a3
+	beqz	a3, 2f
+	extui	a3, a3, 0, WBBITS
+
+1:	rotw	-1
+	addi	a3, a7, -1
+	movi	a4, 0
+	movi	a5, 0
+	movi	a6, 0
+	movi	a7, 0
+	bgei	a3, 1, 1b
+
+	/* We are back were we were when we started.
+	 * Note: a2 still contains WMASK (if we've returned to the original
+	 *	 frame where we had loaded a2), or at least the lower 4 bits
+	 *	 (if we have restored WSBITS-1 frames).
+	 */
+2:
+#else
+	movi	a2, 1
+#endif
+#if XCHAL_HAVE_THREADPTR
+	l32i	a3, a1, PT_THREADPTR
+	wur	a3, threadptr
+#endif
+
+	j	common_exception_exit
+
+	/* This is the kernel exception exit.
+	 * We avoided to do a MOVSP when we entered the exception, but we
+	 * have to do it here.
+	 */
+
+kernel_exception_exit:
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	/* Check if we have to do a movsp.
+	 *
+	 * We only have to do a movsp if the previous window-frame has
+	 * been spilled to the *temporary* exception stack instead of the
+	 * task's stack. This is the case if the corresponding bit in
+	 * WINDOWSTART for the previous window-frame was set before
+	 * (not spilled) but is zero now (spilled).
+	 * If this bit is zero, all other bits except the one for the
+	 * current window frame are also zero. So, we can use a simple test:
+	 * 'and' WINDOWSTART and WINDOWSTART-1:
+	 *
+	 *  (XXXXXX1[0]* - 1) AND XXXXXX1[0]* = XXXXXX0[0]*
+	 *
+	 * The result is zero only if one bit was set.
+	 *
+	 * (Note: We might have gone through several task switches before
+	 *        we come back to the current task, so WINDOWBASE might be
+	 *        different from the time the exception occurred.)
+	 */
+
+	/* Test WINDOWSTART before and after the exception.
+	 * We actually have WMASK, so we only have to test if it is 1 or not.
+	 */
+
+	l32i	a2, a1, PT_WMASK
+	_beqi	a2, 1, common_exception_exit	# Spilled before exception,jump
+
+	/* Test WINDOWSTART now. If spilled, do the movsp */
+
+	rsr     a3, windowstart
+	addi	a0, a3, -1
+	and     a3, a3, a0
+	_bnez	a3, common_exception_exit
+
+	/* Do a movsp (we returned from a call4, so we have at least a0..a7) */
+
+	addi    a0, a1, -16
+	l32i    a3, a0, 0
+	l32i    a4, a0, 4
+	s32i    a3, a1, PT_KERNEL_SIZE + 0
+	s32i    a4, a1, PT_KERNEL_SIZE + 4
+	l32i    a3, a0, 8
+	l32i    a4, a0, 12
+	s32i    a3, a1, PT_KERNEL_SIZE + 8
+	s32i    a4, a1, PT_KERNEL_SIZE + 12
+
+	/* Common exception exit.
+	 * We restore the special register and the current window frame, and
+	 * return from the exception.
+	 *
+	 * Note: We expect a2 to hold PT_WMASK
+	 */
+#else
+	movi	a2, 1
+#endif
+
+common_exception_exit:
+
+	/* Restore address registers. */
+
+	_bbsi.l	a2, 1, 1f
+	l32i	a4,  a1, PT_AREG4
+	l32i	a5,  a1, PT_AREG5
+	l32i	a6,  a1, PT_AREG6
+	l32i	a7,  a1, PT_AREG7
+	_bbsi.l	a2, 2, 1f
+	l32i	a8,  a1, PT_AREG8
+	l32i	a9,  a1, PT_AREG9
+	l32i	a10, a1, PT_AREG10
+	l32i	a11, a1, PT_AREG11
+	_bbsi.l	a2, 3, 1f
+	l32i	a12, a1, PT_AREG12
+	l32i	a13, a1, PT_AREG13
+	l32i	a14, a1, PT_AREG14
+	l32i	a15, a1, PT_AREG15
+
+	/* Restore PC, SAR */
+
+1:	l32i	a2, a1, PT_PC
+	l32i	a3, a1, PT_SAR
+	wsr	a2, epc1
+	wsr	a3, sar
+
+	/* Restore LBEG, LEND, LCOUNT */
+#if XCHAL_HAVE_LOOPS
+	l32i	a2, a1, PT_LBEG
+	l32i	a3, a1, PT_LEND
+	wsr	a2, lbeg
+	l32i	a2, a1, PT_LCOUNT
+	wsr	a3, lend
+	wsr	a2, lcount
+#endif
+
+	/* We control single stepping through the ICOUNTLEVEL register. */
+
+	l32i	a2, a1, PT_ICOUNTLEVEL
+	movi	a3, -2
+	wsr	a2, icountlevel
+	wsr	a3, icount
+
+	/* Check if it was double exception. */
+
+	l32i	a0, a1, PT_DEPC
+	l32i	a3, a1, PT_AREG3
+	l32i	a2, a1, PT_AREG2
+	_bgeui	a0, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f
+
+	/* Restore a0...a3 and return */
+
+	l32i	a0, a1, PT_AREG0
+	l32i	a1, a1, PT_AREG1
+	rfe
+
+1: 	wsr	a0, depc
+	l32i	a0, a1, PT_AREG0
+	l32i	a1, a1, PT_AREG1
+	rfde
+
+ENDPROC(kernel_exception)
+
+/*
+ * Debug exception handler.
+ *
+ * Currently, we don't support KGDB, so only user application can be debugged.
+ *
+ * When we get here,  a0 is trashed and saved to excsave[debuglevel]
+ */
+
+	.literal_position
+
+ENTRY(debug_exception)
+
+	rsr	a0, SREG_EPS + XCHAL_DEBUGLEVEL
+	bbsi.l	a0, PS_EXCM_BIT, .Ldebug_exception_in_exception	# exception mode
+
+	/* Set EPC1 and EXCCAUSE */
+
+	wsr	a2, depc		# save a2 temporarily
+	rsr	a2, SREG_EPC + XCHAL_DEBUGLEVEL
+	wsr	a2, epc1
+
+	movi	a2, EXCCAUSE_MAPPED_DEBUG
+	wsr	a2, exccause
+
+	/* Restore PS to the value before the debug exc but with PS.EXCM set.*/
+
+	movi	a2, 1 << PS_EXCM_BIT
+	or	a2, a0, a2
+	wsr	a2, ps
+
+	/* Switch to kernel/user stack, restore jump vector, and save a0 */
+
+	bbsi.l	a2, PS_UM_BIT, .Ldebug_exception_user	# jump if user mode
+	addi	a2, a1, -16 - PT_KERNEL_SIZE	# assume kernel stack
+
+.Ldebug_exception_continue:
+	l32i	a0, a3, DT_DEBUG_SAVE
+	s32i	a1, a2, PT_AREG1
+	s32i	a0, a2, PT_AREG0
+	movi	a0, 0
+	s32i	a0, a2, PT_DEPC		# mark it as a regular exception
+	xsr	a3, SREG_EXCSAVE + XCHAL_DEBUGLEVEL
+	xsr	a0, depc
+	s32i	a3, a2, PT_AREG3
+	s32i	a0, a2, PT_AREG2
+	mov	a1, a2
+
+	/* Debug exception is handled as an exception, so interrupts will
+	 * likely be enabled in the common exception handler. Disable
+	 * preemption if we have HW breakpoints to preserve DEBUGCAUSE.DBNUM
+	 * meaning.
+	 */
+#if defined(CONFIG_PREEMPT_COUNT) && defined(CONFIG_HAVE_HW_BREAKPOINT)
+	GET_THREAD_INFO(a2, a1)
+	l32i	a3, a2, TI_PRE_COUNT
+	addi	a3, a3, 1
+	s32i	a3, a2, TI_PRE_COUNT
+#endif
+
+	rsr	a2, ps
+	bbsi.l	a2, PS_UM_BIT, _user_exception
+	j	_kernel_exception
+
+.Ldebug_exception_user:
+	rsr	a2, excsave1
+	l32i	a2, a2, EXC_TABLE_KSTK	# load kernel stack pointer
+	j	.Ldebug_exception_continue
+
+.Ldebug_exception_in_exception:
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	/* Debug exception while in exception mode. This may happen when
+	 * window overflow/underflow handler or fast exception handler hits
+	 * data breakpoint, in which case save and disable all data
+	 * breakpoints, single-step faulting instruction and restore data
+	 * breakpoints.
+	 */
+
+	bbci.l	a0, PS_UM_BIT, .Ldebug_exception_in_exception	# jump if kernel mode
+
+	rsr	a0, debugcause
+	bbsi.l	a0, DEBUGCAUSE_DBREAK_BIT, .Ldebug_save_dbreak
+
+	.set	_index, 0
+	.rept	XCHAL_NUM_DBREAK
+	l32i	a0, a3, DT_DBREAKC_SAVE + _index * 4
+	wsr	a0, SREG_DBREAKC + _index
+	.set	_index, _index + 1
+	.endr
+
+	l32i	a0, a3, DT_ICOUNT_LEVEL_SAVE
+	wsr	a0, icountlevel
+
+	l32i	a0, a3, DT_ICOUNT_SAVE
+	xsr	a0, icount
+
+	l32i	a0, a3, DT_DEBUG_SAVE
+	xsr	a3, SREG_EXCSAVE + XCHAL_DEBUGLEVEL
+	rfi	XCHAL_DEBUGLEVEL
+
+.Ldebug_save_dbreak:
+	.set	_index, 0
+	.rept	XCHAL_NUM_DBREAK
+	movi	a0, 0
+	xsr	a0, SREG_DBREAKC + _index
+	s32i	a0, a3, DT_DBREAKC_SAVE + _index * 4
+	.set	_index, _index + 1
+	.endr
+
+	movi	a0, XCHAL_EXCM_LEVEL + 1
+	xsr	a0, icountlevel
+	s32i	a0, a3, DT_ICOUNT_LEVEL_SAVE
+
+	movi	a0, 0xfffffffe
+	xsr	a0, icount
+	s32i	a0, a3, DT_ICOUNT_SAVE
+
+	l32i	a0, a3, DT_DEBUG_SAVE
+	xsr	a3, SREG_EXCSAVE + XCHAL_DEBUGLEVEL
+	rfi	XCHAL_DEBUGLEVEL
+#else
+	/* Debug exception while in exception mode. Should not happen. */
+	j	.Ldebug_exception_in_exception	// FIXME!!
+#endif
+
+ENDPROC(debug_exception)
+
+/*
+ * We get here in case of an unrecoverable exception.
+ * The only thing we can do is to be nice and print a panic message.
+ * We only produce a single stack frame for panic, so ???
+ *
+ *
+ * Entry conditions:
+ *
+ *   - a0 contains the caller address; original value saved in excsave1.
+ *   - the original a0 contains a valid return address (backtrace) or 0.
+ *   - a2 contains a valid stackpointer
+ *
+ * Notes:
+ *
+ *   - If the stack pointer could be invalid, the caller has to setup a
+ *     dummy stack pointer (e.g. the stack of the init_task)
+ *
+ *   - If the return address could be invalid, the caller has to set it
+ *     to 0, so the backtrace would stop.
+ *
+ */
+	.align 4
+unrecoverable_text:
+	.ascii "Unrecoverable error in exception handler\0"
+
+	.literal_position
+
+ENTRY(unrecoverable_exception)
+
+#if XCHAL_HAVE_WINDOWED
+	movi	a0, 1
+	movi	a1, 0
+
+	wsr	a0, windowstart
+	wsr	a1, windowbase
+	rsync
+#endif
+
+	movi	a1, KERNEL_PS_WOE_MASK | LOCKLEVEL
+	wsr	a1, ps
+	rsync
+
+	movi	a1, init_task
+	movi	a0, 0
+	addi	a1, a1, PT_REGS_OFFSET
+
+	movi	abi_arg0, unrecoverable_text
+	abi_call	panic
+
+1:	j	1b
+
+ENDPROC(unrecoverable_exception)
+
+/* -------------------------- FAST EXCEPTION HANDLERS ----------------------- */
+
+	__XTENSA_HANDLER
+	.literal_position
+
+#ifdef SUPPORT_WINDOWED
+/*
+ * Fast-handler for alloca exceptions
+ *
+ *  The ALLOCA handler is entered when user code executes the MOVSP
+ *  instruction and the caller's frame is not in the register file.
+ *
+ * This algorithm was taken from the Ross Morley's RTOS Porting Layer:
+ *
+ *    /home/ross/rtos/porting/XtensaRTOS-PortingLayer-20090507/xtensa_vectors.S
+ *
+ * It leverages the existing window spill/fill routines and their support for
+ * double exceptions. The 'movsp' instruction will only cause an exception if
+ * the next window needs to be loaded. In fact this ALLOCA exception may be
+ * replaced at some point by changing the hardware to do a underflow exception
+ * of the proper size instead.
+ *
+ * This algorithm simply backs out the register changes started by the user
+ * exception handler, makes it appear that we have started a window underflow
+ * by rotating the window back and then setting the old window base (OWB) in
+ * the 'ps' register with the rolled back window base. The 'movsp' instruction
+ * will be re-executed and this time since the next window frames is in the
+ * active AR registers it won't cause an exception.
+ *
+ * If the WindowUnderflow code gets a TLB miss the page will get mapped
+ * the partial WindowUnderflow will be handled in the double exception
+ * handler.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+ENTRY(fast_alloca)
+	rsr	a0, windowbase
+	rotw	-1
+	rsr	a2, ps
+	extui	a3, a2, PS_OWB_SHIFT, PS_OWB_WIDTH
+	xor	a3, a3, a4
+	l32i	a4, a6, PT_AREG0
+	l32i	a1, a6, PT_DEPC
+	rsr	a6, depc
+	wsr	a1, depc
+	slli	a3, a3, PS_OWB_SHIFT
+	xor	a2, a2, a3
+	wsr	a2, ps
+	rsync
+
+	_bbci.l	a4, 31, 4f
+	rotw	-1
+	_bbci.l	a8, 30, 8f
+	rotw	-1
+	j	_WindowUnderflow12
+8:	j	_WindowUnderflow8
+4:	j	_WindowUnderflow4
+ENDPROC(fast_alloca)
+#endif
+
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+/*
+ * fast illegal instruction handler.
+ *
+ * This is used to fix up user PS.WOE on the exception caused
+ * by the first opcode related to register window. If PS.WOE is
+ * already set it goes directly to the common user exception handler.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ */
+
+ENTRY(fast_illegal_instruction_user)
+
+	rsr	a0, ps
+	bbsi.l	a0, PS_WOE_BIT, 1f
+	s32i	a3, a2, PT_AREG3
+	movi	a3, PS_WOE_MASK
+	or	a0, a0, a3
+	wsr	a0, ps
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+	GET_THREAD_INFO(a3, a2)
+	rsr	a0, epc1
+	s32i	a0, a3, TI_PS_WOE_FIX_ADDR
+#endif
+	l32i	a3, a2, PT_AREG3
+	l32i	a0, a2, PT_AREG0
+	rsr	a2, depc
+	rfe
+1:
+	call0	user_exception
+
+ENDPROC(fast_illegal_instruction_user)
+#endif
+
+	/*
+ * fast system calls.
+ *
+ * WARNING:  The kernel doesn't save the entire user context before
+ * handling a fast system call.  These functions are small and short,
+ * usually offering some functionality not available to user tasks.
+ *
+ * BE CAREFUL TO PRESERVE THE USER'S CONTEXT.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ */
+
+ENTRY(fast_syscall_user)
+
+	/* Skip syscall. */
+
+	rsr	a0, epc1
+	addi	a0, a0, 3
+	wsr	a0, epc1
+
+	l32i	a0, a2, PT_DEPC
+	bgeui	a0, VALID_DOUBLE_EXCEPTION_ADDRESS, fast_syscall_unrecoverable
+
+	rsr	a0, depc			# get syscall-nr
+	_beqz	a0, fast_syscall_spill_registers
+	_beqi	a0, __NR_xtensa, fast_syscall_xtensa
+
+	call0	user_exception
+
+ENDPROC(fast_syscall_user)
+
+ENTRY(fast_syscall_unrecoverable)
+
+	/* Restore all states. */
+
+	l32i    a0, a2, PT_AREG0        # restore a0
+	xsr     a2, depc                # restore a2, depc
+
+	wsr     a0, excsave1
+	call0	unrecoverable_exception
+
+ENDPROC(fast_syscall_unrecoverable)
+
+/*
+ * sysxtensa syscall handler
+ *
+ * int sysxtensa (SYS_XTENSA_ATOMIC_SET,     ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_ADD,     ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_EXG_ADD, ptr, val,    unused);
+ * int sysxtensa (SYS_XTENSA_ATOMIC_CMP_SWP, ptr, oldval, newval);
+ *        a2            a6                   a3    a4      a5
+ *
+ * Entry condition:
+ *
+ *   a0:	a2 (syscall-nr), original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in a0 and DEPC
+ *   a3:	a3
+ *   a4..a15:	unchanged
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ *
+ * Note: we don't have to save a2; a2 holds the return value
+ */
+
+	.literal_position
+
+#ifdef CONFIG_FAST_SYSCALL_XTENSA
+
+ENTRY(fast_syscall_xtensa)
+
+	s32i	a7, a2, PT_AREG7	# we need an additional register
+	movi	a7, 4			# sizeof(unsigned int)
+	access_ok a3, a7, a0, a2, .Leac	# a0: scratch reg, a2: sp
+
+	_bgeui	a6, SYS_XTENSA_COUNT, .Lill
+	_bnei	a6, SYS_XTENSA_ATOMIC_CMP_SWP, .Lnswp
+
+	/* Fall through for ATOMIC_CMP_SWP. */
+
+.Lswp:	/* Atomic compare and swap */
+
+EX(.Leac) l32i	a0, a3, 0		# read old value
+	bne	a0, a4, 1f		# same as old value? jump
+EX(.Leac) s32i	a5, a3, 0		# different, modify value
+	l32i	a7, a2, PT_AREG7	# restore a7
+	l32i	a0, a2, PT_AREG0	# restore a0
+	movi	a2, 1			# and return 1
+	rfe
+
+1:	l32i	a7, a2, PT_AREG7	# restore a7
+	l32i	a0, a2, PT_AREG0	# restore a0
+	movi	a2, 0			# return 0 (note that we cannot set
+	rfe
+
+.Lnswp:	/* Atomic set, add, and exg_add. */
+
+EX(.Leac) l32i	a7, a3, 0		# orig
+	addi	a6, a6, -SYS_XTENSA_ATOMIC_SET
+	add	a0, a4, a7		# + arg
+	moveqz	a0, a4, a6		# set
+	addi	a6, a6, SYS_XTENSA_ATOMIC_SET
+EX(.Leac) s32i	a0, a3, 0		# write new value
+
+	mov	a0, a2
+	mov	a2, a7
+	l32i	a7, a0, PT_AREG7	# restore a7
+	l32i	a0, a0, PT_AREG0	# restore a0
+	rfe
+
+.Leac:	l32i	a7, a2, PT_AREG7	# restore a7
+	l32i	a0, a2, PT_AREG0	# restore a0
+	movi	a2, -EFAULT
+	rfe
+
+.Lill:	l32i	a7, a2, PT_AREG7	# restore a7
+	l32i	a0, a2, PT_AREG0	# restore a0
+	movi	a2, -EINVAL
+	rfe
+
+ENDPROC(fast_syscall_xtensa)
+
+#else /* CONFIG_FAST_SYSCALL_XTENSA */
+
+ENTRY(fast_syscall_xtensa)
+
+	l32i    a0, a2, PT_AREG0        # restore a0
+	movi	a2, -ENOSYS
+	rfe
+
+ENDPROC(fast_syscall_xtensa)
+
+#endif /* CONFIG_FAST_SYSCALL_XTENSA */
+
+
+/* fast_syscall_spill_registers.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ * Note: We assume the stack pointer is EXC_TABLE_KSTK in the fixup handler.
+ */
+
+#if defined(CONFIG_FAST_SYSCALL_SPILL_REGISTERS) && \
+		defined(USER_SUPPORT_WINDOWED)
+
+ENTRY(fast_syscall_spill_registers)
+
+	/* Register a FIXUP handler (pass current wb as a parameter) */
+
+	xsr	a3, excsave1
+	movi	a0, fast_syscall_spill_registers_fixup
+	s32i	a0, a3, EXC_TABLE_FIXUP
+	rsr	a0, windowbase
+	s32i	a0, a3, EXC_TABLE_PARAM
+	xsr	a3, excsave1		# restore a3 and excsave_1
+
+	/* Save a3, a4 and SAR on stack. */
+
+	rsr	a0, sar
+	s32i	a3, a2, PT_AREG3
+	s32i	a0, a2, PT_SAR
+
+	/* The spill routine might clobber a4, a7, a8, a11, a12, and a15. */
+
+	s32i	a4, a2, PT_AREG4
+	s32i	a7, a2, PT_AREG7
+	s32i	a8, a2, PT_AREG8
+	s32i	a11, a2, PT_AREG11
+	s32i	a12, a2, PT_AREG12
+	s32i	a15, a2, PT_AREG15
+
+	/*
+	 * Rotate ws so that the current windowbase is at bit 0.
+	 * Assume ws = xxxwww1yy (www1 current window frame).
+	 * Rotate ws right so that a4 = yyxxxwww1.
+	 */
+
+	rsr	a0, windowbase
+	rsr	a3, windowstart		# a3 = xxxwww1yy
+	ssr	a0			# holds WB
+	slli	a0, a3, WSBITS
+	or	a3, a3, a0		# a3 = xxxwww1yyxxxwww1yy
+	srl	a3, a3			# a3 = 00xxxwww1yyxxxwww1
+
+	/* We are done if there are no more than the current register frame. */
+
+	extui	a3, a3, 1, WSBITS-1	# a3 = 0yyxxxwww
+	movi	a0, (1 << (WSBITS-1))
+	_beqz	a3, .Lnospill		# only one active frame? jump
+
+	/* We want 1 at the top, so that we return to the current windowbase */
+
+	or	a3, a3, a0		# 1yyxxxwww
+
+	/* Skip empty frames - get 'oldest' WINDOWSTART-bit. */
+
+	wsr	a3, windowstart		# save shifted windowstart
+	neg	a0, a3
+	and	a3, a0, a3		# first bit set from right: 000010000
+
+	ffs_ws	a0, a3			# a0: shifts to skip empty frames
+	movi	a3, WSBITS
+	sub	a0, a3, a0		# WSBITS-a0:number of 0-bits from right
+	ssr	a0			# save in SAR for later.
+
+	rsr	a3, windowbase
+	add	a3, a3, a0
+	wsr	a3, windowbase
+	rsync
+
+	rsr	a3, windowstart
+	srl	a3, a3			# shift windowstart
+
+	/* WB is now just one frame below the oldest frame in the register
+	   window. WS is shifted so the oldest frame is in bit 0, thus, WB
+	   and WS differ by one 4-register frame. */
+
+	/* Save frames. Depending what call was used (call4, call8, call12),
+	 * we have to save 4,8. or 12 registers.
+	 */
+
+
+.Lloop: _bbsi.l	a3, 1, .Lc4
+	_bbci.l	a3, 2, .Lc12
+
+.Lc8:	s32e	a4, a13, -16
+	l32e	a4, a5, -12
+	s32e	a8, a4, -32
+	s32e	a5, a13, -12
+	s32e	a6, a13, -8
+	s32e	a7, a13, -4
+	s32e	a9, a4, -28
+	s32e	a10, a4, -24
+	s32e	a11, a4, -20
+	srli	a11, a3, 2		# shift windowbase by 2
+	rotw	2
+	_bnei	a3, 1, .Lloop
+	j	.Lexit
+
+.Lc4:	s32e	a4, a9, -16
+	s32e	a5, a9, -12
+	s32e	a6, a9, -8
+	s32e	a7, a9, -4
+
+	srli	a7, a3, 1
+	rotw	1
+	_bnei	a3, 1, .Lloop
+	j	.Lexit
+
+.Lc12:	_bbci.l	a3, 3, .Linvalid_mask	# bit 2 shouldn't be zero!
+
+	/* 12-register frame (call12) */
+
+	l32e	a0, a5, -12
+	s32e	a8, a0, -48
+	mov	a8, a0
+
+	s32e	a9, a8, -44
+	s32e	a10, a8, -40
+	s32e	a11, a8, -36
+	s32e	a12, a8, -32
+	s32e	a13, a8, -28
+	s32e	a14, a8, -24
+	s32e	a15, a8, -20
+	srli	a15, a3, 3
+
+	/* The stack pointer for a4..a7 is out of reach, so we rotate the
+	 * window, grab the stackpointer, and rotate back.
+	 * Alternatively, we could also use the following approach, but that
+	 * makes the fixup routine much more complicated:
+	 * rotw	1
+	 * s32e	a0, a13, -16
+	 * ...
+	 * rotw 2
+	 */
+
+	rotw	1
+	mov	a4, a13
+	rotw	-1
+
+	s32e	a4, a8, -16
+	s32e	a5, a8, -12
+	s32e	a6, a8, -8
+	s32e	a7, a8, -4
+
+	rotw	3
+
+	_beqi	a3, 1, .Lexit
+	j	.Lloop
+
+.Lexit:
+
+	/* Done. Do the final rotation and set WS */
+
+	rotw	1
+	rsr	a3, windowbase
+	ssl	a3
+	movi	a3, 1
+	sll	a3, a3
+	wsr	a3, windowstart
+.Lnospill:
+
+	/* Advance PC, restore registers and SAR, and return from exception. */
+
+	l32i	a3, a2, PT_SAR
+	l32i	a0, a2, PT_AREG0
+	wsr	a3, sar
+	l32i	a3, a2, PT_AREG3
+
+	/* Restore clobbered registers. */
+
+	l32i	a4, a2, PT_AREG4
+	l32i	a7, a2, PT_AREG7
+	l32i	a8, a2, PT_AREG8
+	l32i	a11, a2, PT_AREG11
+	l32i	a12, a2, PT_AREG12
+	l32i	a15, a2, PT_AREG15
+
+	movi	a2, 0
+	rfe
+
+.Linvalid_mask:
+
+	/* We get here because of an unrecoverable error in the window
+	 * registers, so set up a dummy frame and kill the user application.
+	 * Note: We assume EXC_TABLE_KSTK contains a valid stack pointer.
+	 */
+
+	movi	a0, 1
+	movi	a1, 0
+
+	wsr	a0, windowstart
+	wsr	a1, windowbase
+	rsync
+
+	movi	a0, 0
+
+	rsr	a3, excsave1
+	l32i	a1, a3, EXC_TABLE_KSTK
+
+	movi	a4, KERNEL_PS_WOE_MASK | LOCKLEVEL
+	wsr	a4, ps
+	rsync
+
+	movi	abi_arg0, SIGSEGV
+	abi_call	make_task_dead
+
+	/* shouldn't return, so panic */
+
+	wsr	a0, excsave1
+	call0	unrecoverable_exception		# should not return
+1:	j	1b
+
+
+ENDPROC(fast_syscall_spill_registers)
+
+/* Fixup handler.
+ *
+ * We get here if the spill routine causes an exception, e.g. tlb miss.
+ * We basically restore WINDOWBASE and WINDOWSTART to the condition when
+ * we entered the spill routine and jump to the user exception handler.
+ *
+ * Note that we only need to restore the bits in windowstart that have not
+ * been spilled yet by the _spill_register routine. Luckily, a3 contains a
+ * rotated windowstart with only those bits set for frames that haven't been
+ * spilled yet. Because a3 is rotated such that bit 0 represents the register
+ * frame for the current windowbase - 1, we need to rotate a3 left by the
+ * value of the current windowbase + 1 and move it to windowstart.
+ *
+ * a0: value of depc, original value in depc
+ * a2: trashed, original value in EXC_TABLE_DOUBLE_SAVE
+ * a3: exctable, original value in excsave1
+ */
+
+ENTRY(fast_syscall_spill_registers_fixup)
+
+	rsr	a2, windowbase	# get current windowbase (a2 is saved)
+	xsr	a0, depc	# restore depc and a0
+	ssl	a2		# set shift (32 - WB)
+
+	/* We need to make sure the current registers (a0-a3) are preserved.
+	 * To do this, we simply set the bit for the current window frame
+	 * in WS, so that the exception handlers save them to the task stack.
+	 *
+	 * Note: we use a3 to set the windowbase, so we take a special care
+	 * of it, saving it in the original _spill_registers frame across
+	 * the exception handler call.
+	 */
+
+	xsr	a3, excsave1	# get spill-mask
+	slli	a3, a3, 1	# shift left by one
+	addi	a3, a3, 1	# set the bit for the current window frame
+
+	slli	a2, a3, 32-WSBITS
+	src	a2, a3, a2	# a2 = xxwww1yyxxxwww1yy......
+	wsr	a2, windowstart	# set corrected windowstart
+
+	srli	a3, a3, 1
+	rsr	a2, excsave1
+	l32i	a2, a2, EXC_TABLE_DOUBLE_SAVE	# restore a2
+	xsr	a2, excsave1
+	s32i	a3, a2, EXC_TABLE_DOUBLE_SAVE	# save a3
+	l32i	a3, a2, EXC_TABLE_PARAM	# original WB (in user task)
+	xsr	a2, excsave1
+
+	/* Return to the original (user task) WINDOWBASE.
+	 * We leave the following frame behind:
+	 * a0, a1, a2	same
+	 * a3:		trashed (saved in EXC_TABLE_DOUBLE_SAVE)
+	 * depc:	depc (we have to return to that address)
+	 * excsave_1:	exctable
+	 */
+
+	wsr	a3, windowbase
+	rsync
+
+	/* We are now in the original frame when we entered _spill_registers:
+	 *  a0: return address
+	 *  a1: used, stack pointer
+	 *  a2: kernel stack pointer
+	 *  a3: available
+	 *  depc: exception address
+	 *  excsave: exctable
+	 * Note: This frame might be the same as above.
+	 */
+
+	/* Setup stack pointer. */
+
+	addi	a2, a2, -PT_USER_SIZE
+	s32i	a0, a2, PT_AREG0
+
+	/* Make sure we return to this fixup handler. */
+
+	movi	a3, fast_syscall_spill_registers_fixup_return
+	s32i	a3, a2, PT_DEPC		# setup depc
+
+	/* Jump to the exception handler. */
+
+	rsr	a3, excsave1
+	rsr	a0, exccause
+	addx4	a0, a0, a3              	# find entry in table
+	l32i	a0, a0, EXC_TABLE_FAST_USER     # load handler
+	l32i	a3, a3, EXC_TABLE_DOUBLE_SAVE
+	jx	a0
+
+ENDPROC(fast_syscall_spill_registers_fixup)
+
+ENTRY(fast_syscall_spill_registers_fixup_return)
+
+	/* When we return here, all registers have been restored (a2: DEPC) */
+
+	wsr	a2, depc		# exception address
+
+	/* Restore fixup handler. */
+
+	rsr	a2, excsave1
+	s32i	a3, a2, EXC_TABLE_DOUBLE_SAVE
+	movi	a3, fast_syscall_spill_registers_fixup
+	s32i	a3, a2, EXC_TABLE_FIXUP
+	rsr	a3, windowbase
+	s32i	a3, a2, EXC_TABLE_PARAM
+	l32i	a2, a2, EXC_TABLE_KSTK
+
+	/* Load WB at the time the exception occurred. */
+
+	rsr	a3, sar			# WB is still in SAR
+	neg	a3, a3
+	wsr	a3, windowbase
+	rsync
+
+	rsr	a3, excsave1
+	l32i	a3, a3, EXC_TABLE_DOUBLE_SAVE
+
+	rfde
+
+ENDPROC(fast_syscall_spill_registers_fixup_return)
+
+#else /* CONFIG_FAST_SYSCALL_SPILL_REGISTERS */
+
+ENTRY(fast_syscall_spill_registers)
+
+	l32i    a0, a2, PT_AREG0        # restore a0
+	movi	a2, -ENOSYS
+	rfe
+
+ENDPROC(fast_syscall_spill_registers)
+
+#endif /* CONFIG_FAST_SYSCALL_SPILL_REGISTERS */
+
+#ifdef CONFIG_MMU
+/*
+ * We should never get here. Bail out!
+ */
+
+ENTRY(fast_second_level_miss_double_kernel)
+
+1:
+	call0	unrecoverable_exception		# should not return
+1:	j	1b
+
+ENDPROC(fast_second_level_miss_double_kernel)
+
+/* First-level entry handler for user, kernel, and double 2nd-level
+ * TLB miss exceptions.  Note that for now, user and kernel miss
+ * exceptions share the same entry point and are handled identically.
+ *
+ * An old, less-efficient C version of this function used to exist.
+ * We include it below, interleaved as comments, for reference.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+ENTRY(fast_second_level_miss)
+
+	/* Save a1 and a3. Note: we don't expect a double exception. */
+
+	s32i	a1, a2, PT_AREG1
+	s32i	a3, a2, PT_AREG3
+
+	/* We need to map the page of PTEs for the user task.  Find
+	 * the pointer to that page.  Also, it's possible for tsk->mm
+	 * to be NULL while tsk->active_mm is nonzero if we faulted on
+	 * a vmalloc address.  In that rare case, we must use
+	 * active_mm instead to avoid a fault in this handler.  See
+	 *
+	 * http://mail.nl.linux.org/linux-mm/2002-08/msg00258.html
+	 *   (or search Internet on "mm vs. active_mm")
+	 *
+	 *	if (!mm)
+	 *		mm = tsk->active_mm;
+	 *	pgd = pgd_offset (mm, regs->excvaddr);
+	 *	pmd = pmd_offset (pgd, regs->excvaddr);
+	 *	pmdval = *pmd;
+	 */
+
+	GET_CURRENT(a1,a2)
+	l32i	a0, a1, TASK_MM		# tsk->mm
+	beqz	a0, .Lfast_second_level_miss_no_mm
+
+.Lfast_second_level_miss_continue:
+	rsr	a3, excvaddr		# fault address
+	_PGD_OFFSET(a0, a3, a1)
+	l32i	a0, a0, 0		# read pmdval
+	beqz	a0, .Lfast_second_level_miss_no_pmd
+
+	/* Read ptevaddr and convert to top of page-table page.
+	 *
+	 * 	vpnval = read_ptevaddr_register() & PAGE_MASK;
+	 * 	vpnval += DTLB_WAY_PGTABLE;
+	 *	pteval = mk_pte (virt_to_page(pmd_val(pmdval)), PAGE_KERNEL);
+	 *	write_dtlb_entry (pteval, vpnval);
+	 *
+	 * The messy computation for 'pteval' above really simplifies
+	 * into the following:
+	 *
+	 * pteval = ((pmdval - PAGE_OFFSET + PHYS_OFFSET) & PAGE_MASK)
+	 *                 | PAGE_DIRECTORY
+	 */
+
+	movi	a1, (PHYS_OFFSET - PAGE_OFFSET) & 0xffffffff
+	add	a0, a0, a1		# pmdval - PAGE_OFFSET
+	extui	a1, a0, 0, PAGE_SHIFT	# ... & PAGE_MASK
+	xor	a0, a0, a1
+
+	movi	a1, _PAGE_DIRECTORY
+	or	a0, a0, a1		# ... | PAGE_DIRECTORY
+
+	/*
+	 * We utilize all three wired-ways (7-9) to hold pmd translations.
+	 * Memory regions are mapped to the DTLBs according to bits 28 and 29.
+	 * This allows to map the three most common regions to three different
+	 * DTLBs:
+	 *  0,1 -> way 7	program (0040.0000) and virtual (c000.0000)
+	 *  2   -> way 8	shared libaries (2000.0000)
+	 *  3   -> way 0	stack (3000.0000)
+	 */
+
+	extui	a3, a3, 28, 2		# addr. bit 28 and 29	0,1,2,3
+	rsr	a1, ptevaddr
+	addx2	a3, a3, a3		# ->			0,3,6,9
+	srli	a1, a1, PAGE_SHIFT
+	extui	a3, a3, 2, 2		# ->			0,0,1,2
+	slli	a1, a1, PAGE_SHIFT	# ptevaddr & PAGE_MASK
+	addi	a3, a3, DTLB_WAY_PGD
+	add	a1, a1, a3		# ... + way_number
+
+.Lfast_second_level_miss_wdtlb:
+	wdtlb	a0, a1
+	dsync
+
+	/* Exit critical section. */
+.Lfast_second_level_miss_skip_wdtlb:
+	rsr	a3, excsave1
+	movi	a0, 0
+	s32i	a0, a3, EXC_TABLE_FIXUP
+
+	/* Restore the working registers, and return. */
+
+	l32i	a0, a2, PT_AREG0
+	l32i	a1, a2, PT_AREG1
+	l32i	a3, a2, PT_AREG3
+	l32i	a2, a2, PT_DEPC
+
+	bgeui	a2, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f
+
+	/* Restore excsave1 and return. */
+
+	rsr	a2, depc
+	rfe
+
+	/* Return from double exception. */
+
+1:	xsr	a2, depc
+	esync
+	rfde
+
+.Lfast_second_level_miss_no_mm:
+	l32i	a0, a1, TASK_ACTIVE_MM	# unlikely case mm == 0
+	bnez	a0, .Lfast_second_level_miss_continue
+
+	/* Even more unlikely case active_mm == 0.
+	 * We can get here with NMI in the middle of context_switch that
+	 * touches vmalloc area.
+	 */
+	movi	a0, init_mm
+	j	.Lfast_second_level_miss_continue
+
+.Lfast_second_level_miss_no_pmd:
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+	/* Special case for cache aliasing.
+	 * We (should) only get here if a clear_user_page, copy_user_page
+	 * or the aliased cache flush functions got preemptively interrupted 
+	 * by another task. Re-establish temporary mapping to the 
+	 * TLBTEMP_BASE areas.
+	 */
+
+	/* We shouldn't be in a double exception */
+
+	l32i	a0, a2, PT_DEPC
+	bgeui	a0, VALID_DOUBLE_EXCEPTION_ADDRESS, .Lfast_second_level_miss_slow
+
+	/* Make sure the exception originated in the special functions */
+
+	movi	a0, __tlbtemp_mapping_start
+	rsr	a3, epc1
+	bltu	a3, a0, .Lfast_second_level_miss_slow
+	movi	a0, __tlbtemp_mapping_end
+	bgeu	a3, a0, .Lfast_second_level_miss_slow
+
+	/* Check if excvaddr was in one of the TLBTEMP_BASE areas. */
+
+	movi	a3, TLBTEMP_BASE_1
+	rsr	a0, excvaddr
+	bltu	a0, a3, .Lfast_second_level_miss_slow
+
+	addi	a1, a0, -TLBTEMP_SIZE
+	bgeu	a1, a3, .Lfast_second_level_miss_slow
+
+	/* Check if we have to restore an ITLB mapping. */
+
+	movi	a1, __tlbtemp_mapping_itlb
+	rsr	a3, epc1
+	sub	a3, a3, a1
+
+	/* Calculate VPN */
+
+	movi	a1, PAGE_MASK
+	and	a1, a1, a0
+
+	/* Jump for ITLB entry */
+
+	bgez	a3, 1f
+
+	/* We can use up to two TLBTEMP areas, one for src and one for dst. */
+
+	extui	a3, a0, PAGE_SHIFT + DCACHE_ALIAS_ORDER, 1
+	add	a1, a3, a1
+
+	/* PPN is in a6 for the first TLBTEMP area and in a7 for the second. */
+
+	mov	a0, a6
+	movnez	a0, a7, a3
+	j	.Lfast_second_level_miss_wdtlb
+
+	/* ITLB entry. We only use dst in a6. */
+
+1:	witlb	a6, a1
+	isync
+	j	.Lfast_second_level_miss_skip_wdtlb
+
+
+#endif	// DCACHE_WAY_SIZE > PAGE_SIZE
+
+	/* Invalid PGD, default exception handling */
+.Lfast_second_level_miss_slow:
+
+	rsr	a1, depc
+	s32i	a1, a2, PT_AREG2
+	mov	a1, a2
+
+	rsr	a2, ps
+	bbsi.l	a2, PS_UM_BIT, 1f
+	call0	_kernel_exception
+1:	call0	_user_exception
+
+ENDPROC(fast_second_level_miss)
+
+/*
+ * StoreProhibitedException
+ *
+ * Update the pte and invalidate the itlb mapping for this pte.
+ *
+ * Entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original in DEPC
+ *   a3:	a3
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	dispatch table
+ *
+ *   PT_DEPC >= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception, DEPC
+ *	     <  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ */
+
+ENTRY(fast_store_prohibited)
+
+	/* Save a1 and a3. */
+
+	s32i	a1, a2, PT_AREG1
+	s32i	a3, a2, PT_AREG3
+
+	GET_CURRENT(a1,a2)
+	l32i	a0, a1, TASK_MM		# tsk->mm
+	beqz	a0, .Lfast_store_no_mm
+
+.Lfast_store_continue:
+	rsr	a1, excvaddr		# fault address
+	_PGD_OFFSET(a0, a1, a3)
+	l32i	a0, a0, 0
+	beqz	a0, .Lfast_store_slow
+
+	/*
+	 * Note that we test _PAGE_WRITABLE_BIT only if PTE is present
+	 * and is not PAGE_NONE. See pgtable.h for possible PTE layouts.
+	 */
+
+	_PTE_OFFSET(a0, a1, a3)
+	l32i	a3, a0, 0		# read pteval
+	movi	a1, _PAGE_CA_INVALID
+	ball	a3, a1, .Lfast_store_slow
+	bbci.l	a3, _PAGE_WRITABLE_BIT, .Lfast_store_slow
+
+	movi	a1, _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_HW_WRITE
+	or	a3, a3, a1
+	rsr	a1, excvaddr
+	s32i	a3, a0, 0
+
+	/* We need to flush the cache if we have page coloring. */
+#if (DCACHE_WAY_SIZE > PAGE_SIZE) && XCHAL_DCACHE_IS_WRITEBACK
+	dhwb	a0, 0
+#endif
+	pdtlb	a0, a1
+	wdtlb	a3, a0
+
+	/* Exit critical section. */
+
+	movi	a0, 0
+	rsr	a3, excsave1
+	s32i	a0, a3, EXC_TABLE_FIXUP
+
+	/* Restore the working registers, and return. */
+
+	l32i	a3, a2, PT_AREG3
+	l32i	a1, a2, PT_AREG1
+	l32i	a0, a2, PT_AREG0
+	l32i	a2, a2, PT_DEPC
+
+	bgeui	a2, VALID_DOUBLE_EXCEPTION_ADDRESS, 1f
+	rsr	a2, depc
+	rfe
+
+	/* Double exception. Restore FIXUP handler and return. */
+
+1:	xsr	a2, depc
+	esync
+	rfde
+
+.Lfast_store_no_mm:
+	l32i	a0, a1, TASK_ACTIVE_MM	# unlikely case mm == 0
+	j	.Lfast_store_continue
+
+	/* If there was a problem, handle fault in C */
+.Lfast_store_slow:
+	rsr	a1, excvaddr
+	pdtlb	a0, a1
+	bbci.l	a0, DTLB_HIT_BIT, 1f
+	idtlb	a0
+1:
+	rsr	a3, depc	# still holds a2
+	s32i	a3, a2, PT_AREG2
+	mov	a1, a2
+
+	rsr	a2, ps
+	bbsi.l	a2, PS_UM_BIT, 1f
+	call0	_kernel_exception
+1:	call0	_user_exception
+
+ENDPROC(fast_store_prohibited)
+
+#endif /* CONFIG_MMU */
+
+	.text
+/*
+ * System Calls.
+ *
+ * void system_call (struct pt_regs* regs, int exccause)
+ *                            a2                 a3
+ */
+	.literal_position
+
+ENTRY(system_call)
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	abi_entry_default
+#elif defined(__XTENSA_CALL0_ABI__)
+	abi_entry(12)
+
+	s32i	a0, sp, 0
+	s32i	abi_saved0, sp, 4
+	s32i	abi_saved1, sp, 8
+	mov	abi_saved0, a2
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+	/* regs->syscall = regs->areg[2] */
+
+	l32i	a7, abi_saved0, PT_AREG2
+	s32i	a7, abi_saved0, PT_SYSCALL
+
+	GET_THREAD_INFO(a4, a1)
+	l32i	abi_saved1, a4, TI_FLAGS
+	movi	a4, _TIF_WORK_MASK
+	and	abi_saved1, abi_saved1, a4
+	beqz	abi_saved1, 1f
+
+	mov	abi_arg0, abi_saved0
+	abi_call	do_syscall_trace_enter
+	beqz	abi_rv, .Lsyscall_exit
+	l32i	a7, abi_saved0, PT_SYSCALL
+
+1:
+	/* syscall = sys_call_table[syscall_nr] */
+
+	movi	a4, sys_call_table
+	movi	a5, __NR_syscalls
+	movi	abi_rv, -ENOSYS
+	bgeu	a7, a5, 1f
+
+	addx4	a4, a7, a4
+	l32i	abi_tmp0, a4, 0
+
+	/* Load args: arg0 - arg5 are passed via regs. */
+
+	l32i	abi_arg0, abi_saved0, PT_AREG6
+	l32i	abi_arg1, abi_saved0, PT_AREG3
+	l32i	abi_arg2, abi_saved0, PT_AREG4
+	l32i	abi_arg3, abi_saved0, PT_AREG5
+	l32i	abi_arg4, abi_saved0, PT_AREG8
+	l32i	abi_arg5, abi_saved0, PT_AREG9
+
+	abi_callx	abi_tmp0
+
+1:	/* regs->areg[2] = return_value */
+
+	s32i	abi_rv, abi_saved0, PT_AREG2
+	bnez	abi_saved1, 1f
+.Lsyscall_exit:
+#if defined(__XTENSA_WINDOWED_ABI__)
+	abi_ret_default
+#elif defined(__XTENSA_CALL0_ABI__)
+	l32i	a0, sp, 0
+	l32i	abi_saved0, sp, 4
+	l32i	abi_saved1, sp, 8
+	abi_ret(12)
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+1:
+	mov	abi_arg0, abi_saved0
+	abi_call	do_syscall_trace_leave
+	j	.Lsyscall_exit
+
+ENDPROC(system_call)
+
+/*
+ * Spill live registers on the kernel stack macro.
+ *
+ * Entry condition: ps.woe is set, ps.excm is cleared
+ * Exit condition: windowstart has single bit set
+ * May clobber: a12, a13
+ */
+	.macro	spill_registers_kernel
+
+#if XCHAL_NUM_AREGS > 16
+	call12	1f
+	_j	2f
+	retw
+	.align	4
+1:
+	_entry	a1, 48
+	addi	a12, a0, 3
+#if XCHAL_NUM_AREGS > 32
+	.rept	(XCHAL_NUM_AREGS - 32) / 12
+	_entry	a1, 48
+	mov	a12, a0
+	.endr
+#endif
+	_entry	a1, 16
+#if XCHAL_NUM_AREGS % 12 == 0
+	mov	a8, a8
+#elif XCHAL_NUM_AREGS % 12 == 4
+	mov	a12, a12
+#elif XCHAL_NUM_AREGS % 12 == 8
+	mov	a4, a4
+#endif
+	retw
+2:
+#else
+	mov	a12, a12
+#endif
+	.endm
+
+/*
+ * Task switch.
+ *
+ * struct task*  _switch_to (struct task* prev, struct task* next)
+ *         a2                              a2                 a3
+ */
+
+ENTRY(_switch_to)
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	abi_entry(XTENSA_SPILL_STACK_RESERVE)
+#elif defined(__XTENSA_CALL0_ABI__)
+	abi_entry(16)
+
+	s32i	a12, sp, 0
+	s32i	a13, sp, 4
+	s32i	a14, sp, 8
+	s32i	a15, sp, 12
+#else
+#error Unsupported Xtensa ABI
+#endif
+	mov	a11, a3			# and 'next' (a3)
+
+	l32i	a4, a2, TASK_THREAD_INFO
+	l32i	a5, a3, TASK_THREAD_INFO
+
+	save_xtregs_user a4 a6 a8 a9 a12 a13 THREAD_XTREGS_USER
+
+#if THREAD_RA > 1020 || THREAD_SP > 1020
+	addi	a10, a2, TASK_THREAD
+	s32i	a0, a10, THREAD_RA - TASK_THREAD	# save return address
+	s32i	a1, a10, THREAD_SP - TASK_THREAD	# save stack pointer
+#else
+	s32i	a0, a2, THREAD_RA	# save return address
+	s32i	a1, a2, THREAD_SP	# save stack pointer
+#endif
+
+#if defined(CONFIG_STACKPROTECTOR) && !defined(CONFIG_SMP)
+	movi	a6, __stack_chk_guard
+	l32i	a8, a3, TASK_STACK_CANARY
+	s32i	a8, a6, 0
+#endif
+
+	/* Disable ints while we manipulate the stack pointer. */
+
+	irq_save a14, a3
+	rsync
+
+	/* Switch CPENABLE */
+
+#if (XTENSA_HAVE_COPROCESSORS || XTENSA_HAVE_IO_PORTS)
+	l32i	a3, a5, THREAD_CPENABLE
+#ifdef CONFIG_SMP
+	beqz	a3, 1f
+	memw			# pairs with memw (2) in fast_coprocessor
+	l32i	a6, a5, THREAD_CP_OWNER_CPU
+	l32i	a7, a5, THREAD_CPU
+	beq	a6, a7, 1f	# load 0 into CPENABLE if current CPU is not the owner
+	movi	a3, 0
+1:
+#endif
+	wsr	a3, cpenable
+#endif
+
+#if XCHAL_HAVE_EXCLUSIVE
+	l32i	a3, a5, THREAD_ATOMCTL8
+	getex	a3
+	s32i	a3, a4, THREAD_ATOMCTL8
+#endif
+
+	/* Flush register file. */
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	spill_registers_kernel
+#endif
+
+	/* Set kernel stack (and leave critical section)
+	 * Note: It's save to set it here. The stack will not be overwritten
+	 *       because the kernel stack will only be loaded again after
+	 *       we return from kernel space.
+	 */
+
+	rsr	a3, excsave1		# exc_table
+	addi	a7, a5, PT_REGS_OFFSET
+	s32i	a7, a3, EXC_TABLE_KSTK
+
+	/* restore context of the task 'next' */
+
+	l32i	a0, a11, THREAD_RA	# restore return address
+	l32i	a1, a11, THREAD_SP	# restore stack pointer
+
+	load_xtregs_user a5 a6 a8 a9 a12 a13 THREAD_XTREGS_USER
+
+	wsr	a14, ps
+	rsync
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	abi_ret(XTENSA_SPILL_STACK_RESERVE)
+#elif defined(__XTENSA_CALL0_ABI__)
+	l32i	a12, sp, 0
+	l32i	a13, sp, 4
+	l32i	a14, sp, 8
+	l32i	a15, sp, 12
+	abi_ret(16)
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+ENDPROC(_switch_to)
+
+ENTRY(ret_from_fork)
+
+	/* void schedule_tail (struct task_struct *prev)
+	 * Note: prev is still in abi_arg0 (return value from fake call frame)
+	 */
+	abi_call	schedule_tail
+
+	mov		abi_arg0, a1
+	abi_call	do_syscall_trace_leave
+	j		common_exception_return
+
+ENDPROC(ret_from_fork)
+
+/*
+ * Kernel thread creation helper
+ * On entry, set up by copy_thread: abi_saved0 = thread_fn,
+ * abi_saved1 = thread_fn arg. Left from _switch_to: abi_arg0 = prev
+ */
+ENTRY(ret_from_kernel_thread)
+
+	abi_call	schedule_tail
+	mov		abi_arg0, abi_saved1
+	abi_callx	abi_saved0
+	j		common_exception_return
+
+ENDPROC(ret_from_kernel_thread)
+
+#ifdef CONFIG_HIBERNATION
+
+	.section	.bss, "aw"
+	.align	4
+.Lsaved_regs:
+#if defined(__XTENSA_WINDOWED_ABI__)
+	.fill	2, 4
+#elif defined(__XTENSA_CALL0_ABI__)
+	.fill	6, 4
+#else
+#error Unsupported Xtensa ABI
+#endif
+	.align	XCHAL_NCP_SA_ALIGN
+.Lsaved_user_regs:
+	.fill	XTREGS_USER_SIZE, 1
+
+	.previous
+
+ENTRY(swsusp_arch_suspend)
+
+	abi_entry_default
+
+	movi		a2, .Lsaved_regs
+	movi		a3, .Lsaved_user_regs
+	s32i		a0, a2, 0
+	s32i		a1, a2, 4
+	save_xtregs_user a3 a4 a5 a6 a7 a8 0
+#if defined(__XTENSA_WINDOWED_ABI__)
+	spill_registers_kernel
+#elif defined(__XTENSA_CALL0_ABI__)
+	s32i		a12, a2, 8
+	s32i		a13, a2, 12
+	s32i		a14, a2, 16
+	s32i		a15, a2, 20
+#else
+#error Unsupported Xtensa ABI
+#endif
+	abi_call	swsusp_save
+	mov		a2, abi_rv
+	abi_ret_default
+
+ENDPROC(swsusp_arch_suspend)
+
+ENTRY(swsusp_arch_resume)
+
+	abi_entry_default
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	spill_registers_kernel
+#endif
+
+	movi		a2, restore_pblist
+	l32i		a2, a2, 0
+
+.Lcopy_pbe:
+	l32i		a3, a2, PBE_ADDRESS
+	l32i		a4, a2, PBE_ORIG_ADDRESS
+
+	__loopi		a3, a9, PAGE_SIZE, 16
+	l32i		a5, a3, 0
+	l32i		a6, a3, 4
+	l32i		a7, a3, 8
+	l32i		a8, a3, 12
+	addi		a3, a3, 16
+	s32i		a5, a4, 0
+	s32i		a6, a4, 4
+	s32i		a7, a4, 8
+	s32i		a8, a4, 12
+	addi		a4, a4, 16
+	__endl		a3, a9
+
+	l32i		a2, a2, PBE_NEXT
+	bnez		a2, .Lcopy_pbe
+
+	movi		a2, .Lsaved_regs
+	movi		a3, .Lsaved_user_regs
+	l32i		a0, a2, 0
+	l32i		a1, a2, 4
+	load_xtregs_user a3 a4 a5 a6 a7 a8 0
+#if defined(__XTENSA_CALL0_ABI__)
+	l32i		a12, a2, 8
+	l32i		a13, a2, 12
+	l32i		a14, a2, 16
+	l32i		a15, a2, 20
+#endif
+	movi		a2, 0
+	abi_ret_default
+
+ENDPROC(swsusp_arch_resume)
+
+#endif
diff --git a/arch/xtensa/kernel/head.S b/arch/xtensa/kernel/head.S
new file mode 100644
index 000000000..4b0c5c5e6
--- /dev/null
+++ b/arch/xtensa/kernel/head.S
@@ -0,0 +1,383 @@
+/*
+ * arch/xtensa/kernel/head.S
+ *
+ * Xtensa Processor startup code.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2008 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ * Kevin Chea
+ */
+
+#include <asm/asmmacro.h>
+#include <asm/processor.h>
+#include <asm/page.h>
+#include <asm/cacheasm.h>
+#include <asm/initialize_mmu.h>
+#include <asm/mxregs.h>
+
+#include <linux/init.h>
+#include <linux/linkage.h>
+
+/*
+ * This module contains the entry code for kernel images. It performs the
+ * minimal setup needed to call the generic C routines.
+ *
+ * Prerequisites:
+ *
+ * - The kernel image has been loaded to the actual address where it was
+ *   compiled to.
+ * - a2 contains either 0 or a pointer to a list of boot parameters.
+ *   (see setup.c for more details)
+ *
+ */
+
+/*
+ *  _start
+ *
+ *  The bootloader passes a pointer to a list of boot parameters in a2.
+ */
+
+	/* The first bytes of the kernel image must be an instruction, so we
+	 * manually allocate and define the literal constant we need for a jx
+	 * instruction.
+	 */
+
+	__HEAD
+	.begin	no-absolute-literals
+
+ENTRY(_start)
+
+	/* Preserve the pointer to the boot parameter list in EXCSAVE_1 */
+	wsr     a2, excsave1
+	_j	_SetupOCD
+
+	.align	4
+	.literal_position
+_SetupOCD:
+	/*
+	 * Initialize WB, WS, and clear PS.EXCM (to allow loop instructions).
+	 * Set Interrupt Level just below XCHAL_DEBUGLEVEL to allow
+	 * xt-gdb to single step via DEBUG exceptions received directly
+	 * by ocd.
+	 */
+#if XCHAL_HAVE_WINDOWED
+	movi	a1, 1
+	movi	a0, 0
+	wsr	a1, windowstart
+	wsr	a0, windowbase
+	rsync
+#endif
+
+	movi	a1, LOCKLEVEL
+	wsr	a1, ps
+	rsync
+
+	.global _SetupMMU
+_SetupMMU:
+	Offset = _SetupMMU - _start
+
+#ifdef CONFIG_INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	initialize_mmu
+#if defined(CONFIG_MMU) && XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY
+	rsr	a2, excsave1
+	movi	a3, XCHAL_KSEG_PADDR
+	bltu	a2, a3, 1f
+	sub	a2, a2, a3
+	movi	a3, XCHAL_KSEG_SIZE
+	bgeu	a2, a3, 1f
+	movi	a3, XCHAL_KSEG_CACHED_VADDR
+	add	a2, a2, a3
+	wsr	a2, excsave1
+1:
+#endif
+#endif
+
+	movi	a0, _startup
+	jx	a0
+
+ENDPROC(_start)
+	.end	no-absolute-literals
+
+	__REF
+	.literal_position
+
+ENTRY(_startup)
+
+	/* Set a0 to 0 for the remaining initialization. */
+
+	movi	a0, 0
+
+#if XCHAL_HAVE_VECBASE
+	movi    a2, VECBASE_VADDR
+	wsr	a2, vecbase
+#endif
+
+	/* Clear debugging registers. */
+
+#if XCHAL_HAVE_DEBUG
+#if XCHAL_NUM_IBREAK > 0
+	wsr	a0, ibreakenable
+#endif
+	wsr	a0, icount
+	movi	a1, 15
+	wsr	a0, icountlevel
+
+	.set	_index, 0
+	.rept	XCHAL_NUM_DBREAK
+	wsr	a0, SREG_DBREAKC + _index
+	.set	_index, _index + 1
+	.endr
+#endif
+
+	/* Clear CCOUNT (not really necessary, but nice) */
+
+	wsr	a0, ccount	# not really necessary, but nice
+
+	/* Disable zero-loops. */
+
+#if XCHAL_HAVE_LOOPS
+	wsr	a0, lcount
+#endif
+
+	/* Disable all timers. */
+
+	.set	_index, 0
+	.rept	XCHAL_NUM_TIMERS
+	wsr	a0, SREG_CCOMPARE + _index
+	.set	_index, _index + 1
+	.endr
+
+	/* Interrupt initialization. */
+
+	movi	a2, XCHAL_INTTYPE_MASK_SOFTWARE | XCHAL_INTTYPE_MASK_EXTERN_EDGE
+	wsr	a0, intenable
+	wsr	a2, intclear
+
+	/* Disable coprocessors. */
+
+#if XCHAL_HAVE_CP
+	wsr	a0, cpenable
+#endif
+
+	/*  Initialize the caches.
+	 *  a2, a3 are just working registers (clobbered).
+	 */
+
+#if XCHAL_DCACHE_LINE_LOCKABLE
+	___unlock_dcache_all a2 a3
+#endif
+
+#if XCHAL_ICACHE_LINE_LOCKABLE
+	___unlock_icache_all a2 a3
+#endif
+
+	___invalidate_dcache_all a2 a3
+	___invalidate_icache_all a2 a3
+
+	isync
+
+	initialize_cacheattr
+
+#ifdef CONFIG_HAVE_SMP
+	movi	a2, CCON	# MX External Register to Configure Cache
+	movi	a3, 1
+	wer	a3, a2
+#endif
+
+	/* Setup stack and enable window exceptions (keep irqs disabled) */
+
+	movi	a1, start_info
+	l32i	a1, a1, 0
+
+	/* Disable interrupts. */
+	/* Enable window exceptions if kernel is built with windowed ABI. */
+	movi	a2, KERNEL_PS_WOE_MASK | LOCKLEVEL
+	wsr	a2, ps
+	rsync
+
+#ifdef CONFIG_SMP
+	/*
+	 * Notice that we assume with SMP that cores have PRID
+	 * supported by the cores.
+	 */
+	rsr	a2, prid
+	bnez	a2, .Lboot_secondary
+
+#endif  /* CONFIG_SMP */
+
+	/* Unpack data sections
+	 *
+	 * The linker script used to build the Linux kernel image
+	 * creates a table located at __boot_reloc_table_start
+	 * that contains the information what data needs to be unpacked.
+	 *
+	 * Uses a2-a7.
+	 */
+
+	movi	a2, __boot_reloc_table_start
+	movi	a3, __boot_reloc_table_end
+
+1:	beq	a2, a3, 3f	# no more entries?
+	l32i	a4, a2, 0	# start destination (in RAM)
+	l32i	a5, a2, 4	# end destination (in RAM)
+	l32i	a6, a2, 8	# start source (in ROM)
+	addi	a2, a2, 12	# next entry
+	beq	a4, a5, 1b	# skip, empty entry
+	beq	a4, a6, 1b	# skip, source and dest. are the same
+
+2:	l32i	a7, a6, 0	# load word
+	addi	a6, a6, 4
+	s32i	a7, a4, 0	# store word
+	addi	a4, a4, 4
+	bltu	a4, a5, 2b
+	j	1b
+
+3:
+	/* All code and initialized data segments have been copied.
+	 * Now clear the BSS segment.
+	 */
+
+	movi	a2, __bss_start	# start of BSS
+	movi	a3, __bss_stop	# end of BSS
+
+	__loopt	a2, a3, a4, 2
+	s32i	a0, a2, 0
+	__endla	a2, a3, 4
+
+#if XCHAL_DCACHE_IS_WRITEBACK
+
+	/* After unpacking, flush the writeback cache to memory so the
+	 * instructions/data are available.
+	 */
+
+	___flush_dcache_all a2 a3
+#endif
+	memw
+	isync
+	___invalidate_icache_all a2 a3
+	isync
+
+#ifdef CONFIG_XIP_KERNEL
+	/* Setup bootstrap CPU stack in XIP kernel */
+
+	movi	a1, start_info
+	l32i	a1, a1, 0
+#endif
+
+	movi	abi_arg0, 0
+	xsr	abi_arg0, excsave1
+
+	/* init_arch kick-starts the linux kernel */
+
+	abi_call	init_arch
+	abi_call	start_kernel
+
+should_never_return:
+	j	should_never_return
+
+#ifdef CONFIG_SMP
+.Lboot_secondary:
+
+	movi	a2, cpu_start_ccount
+1:
+	memw
+	l32i	a3, a2, 0
+	beqi	a3, 0, 1b
+	movi	a3, 0
+	s32i	a3, a2, 0
+1:
+	memw
+	l32i	a3, a2, 0
+	beqi	a3, 0, 1b
+	wsr	a3, ccount
+	movi	a3, 0
+	s32i	a3, a2, 0
+	memw
+
+	movi	abi_arg0, 0
+	wsr	abi_arg0, excsave1
+
+	abi_call	secondary_start_kernel
+	j	should_never_return
+
+#endif  /* CONFIG_SMP */
+
+ENDPROC(_startup)
+
+#ifdef CONFIG_HOTPLUG_CPU
+
+ENTRY(cpu_restart)
+
+#if XCHAL_DCACHE_IS_WRITEBACK
+	___flush_invalidate_dcache_all a2 a3
+#else
+	___invalidate_dcache_all a2 a3
+#endif
+	memw
+	movi	a2, CCON	# MX External Register to Configure Cache
+	movi	a3, 0
+	wer	a3, a2
+	extw
+
+	rsr	a0, prid
+	neg	a2, a0
+	movi	a3, cpu_start_id
+	memw
+	s32i	a2, a3, 0
+#if XCHAL_DCACHE_IS_WRITEBACK
+	dhwbi	a3, 0
+#endif
+1:
+	memw
+	l32i	a2, a3, 0
+	dhi	a3, 0
+	bne	a2, a0, 1b
+
+	/*
+	 * Initialize WB, WS, and clear PS.EXCM (to allow loop instructions).
+	 * Set Interrupt Level just below XCHAL_DEBUGLEVEL to allow
+	 * xt-gdb to single step via DEBUG exceptions received directly
+	 * by ocd.
+	 */
+	movi	a1, 1
+	movi	a0, 0
+	wsr	a1, windowstart
+	wsr	a0, windowbase
+	rsync
+
+	movi	a1, LOCKLEVEL
+	wsr	a1, ps
+	rsync
+
+	j	_startup
+
+ENDPROC(cpu_restart)
+
+#endif  /* CONFIG_HOTPLUG_CPU */
+
+/*
+ * DATA section
+ */
+
+	__REFDATA
+	.align  4
+ENTRY(start_info)
+	.long	init_thread_union + KERNEL_STACK_SIZE
+
+/*
+ * BSS section
+ */
+	
+__PAGE_ALIGNED_BSS
+#ifdef CONFIG_MMU
+ENTRY(swapper_pg_dir)
+	.fill	PAGE_SIZE, 1, 0
+END(swapper_pg_dir)
+#endif
diff --git a/arch/xtensa/kernel/hibernate.c b/arch/xtensa/kernel/hibernate.c
new file mode 100644
index 000000000..06984327d
--- /dev/null
+++ b/arch/xtensa/kernel/hibernate.c
@@ -0,0 +1,25 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include <linux/mm.h>
+#include <linux/suspend.h>
+#include <asm/coprocessor.h>
+
+int pfn_is_nosave(unsigned long pfn)
+{
+	unsigned long nosave_begin_pfn = PFN_DOWN(__pa(&__nosave_begin));
+	unsigned long nosave_end_pfn = PFN_UP(__pa(&__nosave_end));
+
+	return	(pfn >= nosave_begin_pfn) && (pfn < nosave_end_pfn);
+}
+
+void notrace save_processor_state(void)
+{
+	WARN_ON(num_online_cpus() != 1);
+#if XTENSA_HAVE_COPROCESSORS
+	local_coprocessors_flush_release_all();
+#endif
+}
+
+void notrace restore_processor_state(void)
+{
+}
diff --git a/arch/xtensa/kernel/hw_breakpoint.c b/arch/xtensa/kernel/hw_breakpoint.c
new file mode 100644
index 000000000..1eeecd58e
--- /dev/null
+++ b/arch/xtensa/kernel/hw_breakpoint.c
@@ -0,0 +1,309 @@
+/*
+ * Xtensa hardware breakpoints/watchpoints handling functions
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2016 Cadence Design Systems Inc.
+ */
+
+#include <linux/hw_breakpoint.h>
+#include <linux/log2.h>
+#include <linux/percpu.h>
+#include <linux/perf_event.h>
+#include <asm/core.h>
+#include <asm/hw_breakpoint.h>
+
+/* Breakpoint currently in use for each IBREAKA. */
+static DEFINE_PER_CPU(struct perf_event *, bp_on_reg[XCHAL_NUM_IBREAK]);
+
+/* Watchpoint currently in use for each DBREAKA. */
+static DEFINE_PER_CPU(struct perf_event *, wp_on_reg[XCHAL_NUM_DBREAK]);
+
+int hw_breakpoint_slots(int type)
+{
+	switch (type) {
+	case TYPE_INST:
+		return XCHAL_NUM_IBREAK;
+	case TYPE_DATA:
+		return XCHAL_NUM_DBREAK;
+	default:
+		pr_warn("unknown slot type: %d\n", type);
+		return 0;
+	}
+}
+
+int arch_check_bp_in_kernelspace(struct arch_hw_breakpoint *hw)
+{
+	unsigned int len;
+	unsigned long va;
+
+	va = hw->address;
+	len = hw->len;
+
+	return (va >= TASK_SIZE) && ((va + len - 1) >= TASK_SIZE);
+}
+
+/*
+ * Construct an arch_hw_breakpoint from a perf_event.
+ */
+int hw_breakpoint_arch_parse(struct perf_event *bp,
+			     const struct perf_event_attr *attr,
+			     struct arch_hw_breakpoint *hw)
+{
+	/* Type */
+	switch (attr->bp_type) {
+	case HW_BREAKPOINT_X:
+		hw->type = XTENSA_BREAKPOINT_EXECUTE;
+		break;
+	case HW_BREAKPOINT_R:
+		hw->type = XTENSA_BREAKPOINT_LOAD;
+		break;
+	case HW_BREAKPOINT_W:
+		hw->type = XTENSA_BREAKPOINT_STORE;
+		break;
+	case HW_BREAKPOINT_RW:
+		hw->type = XTENSA_BREAKPOINT_LOAD | XTENSA_BREAKPOINT_STORE;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	/* Len */
+	hw->len = attr->bp_len;
+	if (hw->len < 1 || hw->len > 64 || !is_power_of_2(hw->len))
+		return -EINVAL;
+
+	/* Address */
+	hw->address = attr->bp_addr;
+	if (hw->address & (hw->len - 1))
+		return -EINVAL;
+
+	return 0;
+}
+
+int hw_breakpoint_exceptions_notify(struct notifier_block *unused,
+				    unsigned long val, void *data)
+{
+	return NOTIFY_DONE;
+}
+
+static void xtensa_wsr(unsigned long v, u8 sr)
+{
+	/* We don't have indexed wsr and creating instruction dynamically
+	 * doesn't seem worth it given how small XCHAL_NUM_IBREAK and
+	 * XCHAL_NUM_DBREAK are. Thus the switch. In case build breaks here
+	 * the switch below needs to be extended.
+	 */
+	BUILD_BUG_ON(XCHAL_NUM_IBREAK > 2);
+	BUILD_BUG_ON(XCHAL_NUM_DBREAK > 2);
+
+	switch (sr) {
+#if XCHAL_NUM_IBREAK > 0
+	case SREG_IBREAKA + 0:
+		xtensa_set_sr(v, SREG_IBREAKA + 0);
+		break;
+#endif
+#if XCHAL_NUM_IBREAK > 1
+	case SREG_IBREAKA + 1:
+		xtensa_set_sr(v, SREG_IBREAKA + 1);
+		break;
+#endif
+
+#if XCHAL_NUM_DBREAK > 0
+	case SREG_DBREAKA + 0:
+		xtensa_set_sr(v, SREG_DBREAKA + 0);
+		break;
+	case SREG_DBREAKC + 0:
+		xtensa_set_sr(v, SREG_DBREAKC + 0);
+		break;
+#endif
+#if XCHAL_NUM_DBREAK > 1
+	case SREG_DBREAKA + 1:
+		xtensa_set_sr(v, SREG_DBREAKA + 1);
+		break;
+
+	case SREG_DBREAKC + 1:
+		xtensa_set_sr(v, SREG_DBREAKC + 1);
+		break;
+#endif
+	}
+}
+
+static int alloc_slot(struct perf_event **slot, size_t n,
+		      struct perf_event *bp)
+{
+	size_t i;
+
+	for (i = 0; i < n; ++i) {
+		if (!slot[i]) {
+			slot[i] = bp;
+			return i;
+		}
+	}
+	return -EBUSY;
+}
+
+static void set_ibreak_regs(int reg, struct perf_event *bp)
+{
+	struct arch_hw_breakpoint *info = counter_arch_bp(bp);
+	unsigned long ibreakenable;
+
+	xtensa_wsr(info->address, SREG_IBREAKA + reg);
+	ibreakenable = xtensa_get_sr(SREG_IBREAKENABLE);
+	xtensa_set_sr(ibreakenable | (1 << reg), SREG_IBREAKENABLE);
+}
+
+static void set_dbreak_regs(int reg, struct perf_event *bp)
+{
+	struct arch_hw_breakpoint *info = counter_arch_bp(bp);
+	unsigned long dbreakc = DBREAKC_MASK_MASK & -info->len;
+
+	if (info->type & XTENSA_BREAKPOINT_LOAD)
+		dbreakc |= DBREAKC_LOAD_MASK;
+	if (info->type & XTENSA_BREAKPOINT_STORE)
+		dbreakc |= DBREAKC_STOR_MASK;
+
+	xtensa_wsr(info->address, SREG_DBREAKA + reg);
+	xtensa_wsr(dbreakc, SREG_DBREAKC + reg);
+}
+
+int arch_install_hw_breakpoint(struct perf_event *bp)
+{
+	int i;
+
+	if (counter_arch_bp(bp)->type == XTENSA_BREAKPOINT_EXECUTE) {
+		/* Breakpoint */
+		i = alloc_slot(this_cpu_ptr(bp_on_reg), XCHAL_NUM_IBREAK, bp);
+		if (i < 0)
+			return i;
+		set_ibreak_regs(i, bp);
+
+	} else {
+		/* Watchpoint */
+		i = alloc_slot(this_cpu_ptr(wp_on_reg), XCHAL_NUM_DBREAK, bp);
+		if (i < 0)
+			return i;
+		set_dbreak_regs(i, bp);
+	}
+	return 0;
+}
+
+static int free_slot(struct perf_event **slot, size_t n,
+		     struct perf_event *bp)
+{
+	size_t i;
+
+	for (i = 0; i < n; ++i) {
+		if (slot[i] == bp) {
+			slot[i] = NULL;
+			return i;
+		}
+	}
+	return -EBUSY;
+}
+
+void arch_uninstall_hw_breakpoint(struct perf_event *bp)
+{
+	struct arch_hw_breakpoint *info = counter_arch_bp(bp);
+	int i;
+
+	if (info->type == XTENSA_BREAKPOINT_EXECUTE) {
+		unsigned long ibreakenable;
+
+		/* Breakpoint */
+		i = free_slot(this_cpu_ptr(bp_on_reg), XCHAL_NUM_IBREAK, bp);
+		if (i >= 0) {
+			ibreakenable = xtensa_get_sr(SREG_IBREAKENABLE);
+			xtensa_set_sr(ibreakenable & ~(1 << i),
+				      SREG_IBREAKENABLE);
+		}
+	} else {
+		/* Watchpoint */
+		i = free_slot(this_cpu_ptr(wp_on_reg), XCHAL_NUM_DBREAK, bp);
+		if (i >= 0)
+			xtensa_wsr(0, SREG_DBREAKC + i);
+	}
+}
+
+void hw_breakpoint_pmu_read(struct perf_event *bp)
+{
+}
+
+void flush_ptrace_hw_breakpoint(struct task_struct *tsk)
+{
+	int i;
+	struct thread_struct *t = &tsk->thread;
+
+	for (i = 0; i < XCHAL_NUM_IBREAK; ++i) {
+		if (t->ptrace_bp[i]) {
+			unregister_hw_breakpoint(t->ptrace_bp[i]);
+			t->ptrace_bp[i] = NULL;
+		}
+	}
+	for (i = 0; i < XCHAL_NUM_DBREAK; ++i) {
+		if (t->ptrace_wp[i]) {
+			unregister_hw_breakpoint(t->ptrace_wp[i]);
+			t->ptrace_wp[i] = NULL;
+		}
+	}
+}
+
+/*
+ * Set ptrace breakpoint pointers to zero for this task.
+ * This is required in order to prevent child processes from unregistering
+ * breakpoints held by their parent.
+ */
+void clear_ptrace_hw_breakpoint(struct task_struct *tsk)
+{
+	memset(tsk->thread.ptrace_bp, 0, sizeof(tsk->thread.ptrace_bp));
+	memset(tsk->thread.ptrace_wp, 0, sizeof(tsk->thread.ptrace_wp));
+}
+
+void restore_dbreak(void)
+{
+	int i;
+
+	for (i = 0; i < XCHAL_NUM_DBREAK; ++i) {
+		struct perf_event *bp = this_cpu_ptr(wp_on_reg)[i];
+
+		if (bp)
+			set_dbreak_regs(i, bp);
+	}
+	clear_thread_flag(TIF_DB_DISABLED);
+}
+
+int check_hw_breakpoint(struct pt_regs *regs)
+{
+	if (regs->debugcause & BIT(DEBUGCAUSE_IBREAK_BIT)) {
+		int i;
+		struct perf_event **bp = this_cpu_ptr(bp_on_reg);
+
+		for (i = 0; i < XCHAL_NUM_IBREAK; ++i) {
+			if (bp[i] && !bp[i]->attr.disabled &&
+			    regs->pc == bp[i]->attr.bp_addr)
+				perf_bp_event(bp[i], regs);
+		}
+		return 0;
+	} else if (regs->debugcause & BIT(DEBUGCAUSE_DBREAK_BIT)) {
+		struct perf_event **bp = this_cpu_ptr(wp_on_reg);
+		int dbnum = (regs->debugcause & DEBUGCAUSE_DBNUM_MASK) >>
+			DEBUGCAUSE_DBNUM_SHIFT;
+
+		if (dbnum < XCHAL_NUM_DBREAK && bp[dbnum]) {
+			if (user_mode(regs)) {
+				perf_bp_event(bp[dbnum], regs);
+			} else {
+				set_thread_flag(TIF_DB_DISABLED);
+				xtensa_wsr(0, SREG_DBREAKC + dbnum);
+			}
+		} else {
+			WARN_ONCE(1,
+				  "Wrong/unconfigured DBNUM reported in DEBUGCAUSE: %d\n",
+				  dbnum);
+		}
+		return 0;
+	}
+	return -ENOENT;
+}
diff --git a/arch/xtensa/kernel/irq.c b/arch/xtensa/kernel/irq.c
new file mode 100644
index 000000000..6f01f5308
--- /dev/null
+++ b/arch/xtensa/kernel/irq.c
@@ -0,0 +1,195 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * linux/arch/xtensa/kernel/irq.c
+ *
+ * Xtensa built-in interrupt controller and some generic functions copied
+ * from i386.
+ *
+ * Copyright (C) 2002 - 2013 Tensilica, Inc.
+ * Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
+ *
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Kevin Chea
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/seq_file.h>
+#include <linux/interrupt.h>
+#include <linux/irq.h>
+#include <linux/kernel_stat.h>
+#include <linux/irqchip.h>
+#include <linux/irqchip/xtensa-mx.h>
+#include <linux/irqchip/xtensa-pic.h>
+#include <linux/irqdomain.h>
+#include <linux/of.h>
+
+#include <asm/mxregs.h>
+#include <linux/uaccess.h>
+#include <asm/platform.h>
+#include <asm/traps.h>
+
+DECLARE_PER_CPU(unsigned long, nmi_count);
+
+asmlinkage void do_IRQ(int hwirq, struct pt_regs *regs)
+{
+#ifdef CONFIG_DEBUG_STACKOVERFLOW
+	/* Debugging check for stack overflow: is there less than 1KB free? */
+	{
+		unsigned long sp = current_stack_pointer;
+
+		sp &= THREAD_SIZE - 1;
+
+		if (unlikely(sp < (sizeof(thread_info) + 1024)))
+			printk("Stack overflow in do_IRQ: %ld\n",
+			       sp - sizeof(struct thread_info));
+	}
+#endif
+	generic_handle_domain_irq(NULL, hwirq);
+}
+
+int arch_show_interrupts(struct seq_file *p, int prec)
+{
+	unsigned cpu __maybe_unused;
+#ifdef CONFIG_SMP
+	show_ipi_list(p, prec);
+#endif
+#if XTENSA_FAKE_NMI
+	seq_printf(p, "%*s:", prec, "NMI");
+	for_each_online_cpu(cpu)
+		seq_printf(p, " %10lu", per_cpu(nmi_count, cpu));
+	seq_puts(p, " Non-maskable interrupts\n");
+#endif
+	return 0;
+}
+
+int xtensa_irq_domain_xlate(const u32 *intspec, unsigned int intsize,
+		unsigned long int_irq, unsigned long ext_irq,
+		unsigned long *out_hwirq, unsigned int *out_type)
+{
+	if (WARN_ON(intsize < 1 || intsize > 2))
+		return -EINVAL;
+	if (intsize == 2 && intspec[1] == 1) {
+		int_irq = xtensa_map_ext_irq(ext_irq);
+		if (int_irq < XCHAL_NUM_INTERRUPTS)
+			*out_hwirq = int_irq;
+		else
+			return -EINVAL;
+	} else {
+		*out_hwirq = int_irq;
+	}
+	*out_type = IRQ_TYPE_NONE;
+	return 0;
+}
+
+int xtensa_irq_map(struct irq_domain *d, unsigned int irq,
+		irq_hw_number_t hw)
+{
+	struct irq_chip *irq_chip = d->host_data;
+	u32 mask = 1 << hw;
+
+	if (mask & XCHAL_INTTYPE_MASK_SOFTWARE) {
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_simple_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_EXTERN_EDGE) {
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_edge_irq, "edge");
+		irq_clear_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_EXTERN_LEVEL) {
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_level_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	} else if (mask & XCHAL_INTTYPE_MASK_TIMER) {
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_percpu_irq, "timer");
+		irq_clear_status_flags(irq, IRQ_LEVEL);
+#ifdef XCHAL_INTTYPE_MASK_PROFILING
+	} else if (mask & XCHAL_INTTYPE_MASK_PROFILING) {
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_percpu_irq, "profiling");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+#endif
+	} else {/* XCHAL_INTTYPE_MASK_WRITE_ERROR */
+		/* XCHAL_INTTYPE_MASK_NMI */
+		irq_set_chip_and_handler_name(irq, irq_chip,
+				handle_level_irq, "level");
+		irq_set_status_flags(irq, IRQ_LEVEL);
+	}
+	return 0;
+}
+
+unsigned xtensa_map_ext_irq(unsigned ext_irq)
+{
+	unsigned mask = XCHAL_INTTYPE_MASK_EXTERN_EDGE |
+		XCHAL_INTTYPE_MASK_EXTERN_LEVEL;
+	unsigned i;
+
+	for (i = 0; mask; ++i, mask >>= 1) {
+		if ((mask & 1) && ext_irq-- == 0)
+			return i;
+	}
+	return XCHAL_NUM_INTERRUPTS;
+}
+
+unsigned xtensa_get_ext_irq_no(unsigned irq)
+{
+	unsigned mask = (XCHAL_INTTYPE_MASK_EXTERN_EDGE |
+		XCHAL_INTTYPE_MASK_EXTERN_LEVEL) &
+		((1u << irq) - 1);
+	return hweight32(mask);
+}
+
+void __init init_IRQ(void)
+{
+#ifdef CONFIG_USE_OF
+	irqchip_init();
+#else
+#ifdef CONFIG_HAVE_SMP
+	xtensa_mx_init_legacy(NULL);
+#else
+	xtensa_pic_init_legacy(NULL);
+#endif
+#endif
+
+#ifdef CONFIG_SMP
+	ipi_init();
+#endif
+}
+
+#ifdef CONFIG_HOTPLUG_CPU
+/*
+ * The CPU has been marked offline.  Migrate IRQs off this CPU.  If
+ * the affinity settings do not allow other CPUs, force them onto any
+ * available CPU.
+ */
+void migrate_irqs(void)
+{
+	unsigned int i, cpu = smp_processor_id();
+
+	for_each_active_irq(i) {
+		struct irq_data *data = irq_get_irq_data(i);
+		const struct cpumask *mask;
+		unsigned int newcpu;
+
+		if (irqd_is_per_cpu(data))
+			continue;
+
+		mask = irq_data_get_affinity_mask(data);
+		if (!cpumask_test_cpu(cpu, mask))
+			continue;
+
+		newcpu = cpumask_any_and(mask, cpu_online_mask);
+
+		if (newcpu >= nr_cpu_ids) {
+			pr_info_ratelimited("IRQ%u no longer affine to CPU%u\n",
+					    i, cpu);
+
+			irq_set_affinity(i, cpu_all_mask);
+		} else {
+			irq_set_affinity(i, mask);
+		}
+	}
+}
+#endif /* CONFIG_HOTPLUG_CPU */
diff --git a/arch/xtensa/kernel/jump_label.c b/arch/xtensa/kernel/jump_label.c
new file mode 100644
index 000000000..ad1841cec
--- /dev/null
+++ b/arch/xtensa/kernel/jump_label.c
@@ -0,0 +1,95 @@
+// SPDX-License-Identifier: GPL-2.0
+// Copyright (C) 2018 Cadence Design Systems Inc.
+
+#include <linux/cpu.h>
+#include <linux/jump_label.h>
+#include <linux/kernel.h>
+#include <linux/memory.h>
+#include <linux/stop_machine.h>
+#include <linux/types.h>
+
+#include <asm/cacheflush.h>
+
+#define J_OFFSET_MASK 0x0003ffff
+#define J_SIGN_MASK (~(J_OFFSET_MASK >> 1))
+
+#if defined(__XTENSA_EL__)
+#define J_INSN 0x6
+#define NOP_INSN 0x0020f0
+#elif defined(__XTENSA_EB__)
+#define J_INSN 0x60000000
+#define NOP_INSN 0x0f020000
+#else
+#error Unsupported endianness.
+#endif
+
+struct patch {
+	atomic_t cpu_count;
+	unsigned long addr;
+	size_t sz;
+	const void *data;
+};
+
+static void local_patch_text(unsigned long addr, const void *data, size_t sz)
+{
+	memcpy((void *)addr, data, sz);
+	local_flush_icache_range(addr, addr + sz);
+}
+
+static int patch_text_stop_machine(void *data)
+{
+	struct patch *patch = data;
+
+	if (atomic_inc_return(&patch->cpu_count) == num_online_cpus()) {
+		local_patch_text(patch->addr, patch->data, patch->sz);
+		atomic_inc(&patch->cpu_count);
+	} else {
+		while (atomic_read(&patch->cpu_count) <= num_online_cpus())
+			cpu_relax();
+		__invalidate_icache_range(patch->addr, patch->sz);
+	}
+	return 0;
+}
+
+static void patch_text(unsigned long addr, const void *data, size_t sz)
+{
+	if (IS_ENABLED(CONFIG_SMP)) {
+		struct patch patch = {
+			.cpu_count = ATOMIC_INIT(0),
+			.addr = addr,
+			.sz = sz,
+			.data = data,
+		};
+		stop_machine_cpuslocked(patch_text_stop_machine,
+					&patch, cpu_online_mask);
+	} else {
+		unsigned long flags;
+
+		local_irq_save(flags);
+		local_patch_text(addr, data, sz);
+		local_irq_restore(flags);
+	}
+}
+
+void arch_jump_label_transform(struct jump_entry *e,
+			       enum jump_label_type type)
+{
+	u32 d = (jump_entry_target(e) - (jump_entry_code(e) + 4));
+	u32 insn;
+
+	/* Jump only works within 128K of the J instruction. */
+	BUG_ON(!((d & J_SIGN_MASK) == 0 ||
+		 (d & J_SIGN_MASK) == J_SIGN_MASK));
+
+	if (type == JUMP_LABEL_JMP) {
+#if defined(__XTENSA_EL__)
+		insn = ((d & J_OFFSET_MASK) << 6) | J_INSN;
+#elif defined(__XTENSA_EB__)
+		insn = ((d & J_OFFSET_MASK) << 8) | J_INSN;
+#endif
+	} else {
+		insn = NOP_INSN;
+	}
+
+	patch_text(jump_entry_code(e), &insn, JUMP_LABEL_NOP_SIZE);
+}
diff --git a/arch/xtensa/kernel/mcount.S b/arch/xtensa/kernel/mcount.S
new file mode 100644
index 000000000..309b32982
--- /dev/null
+++ b/arch/xtensa/kernel/mcount.S
@@ -0,0 +1,86 @@
+/*
+ * arch/xtensa/kernel/mcount.S
+ *
+ * Xtensa specific mcount support
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2013 Tensilica Inc.
+ */
+
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/ftrace.h>
+
+/*
+ * Entry condition:
+ *
+ *   a2:	a0 of the caller in windowed ABI
+ *   a10:	a0 of the caller in call0 ABI
+ *
+ * In call0 ABI the function _mcount is called with the special ABI:
+ * its argument is in a10 and all the usual argument registers (a2 - a7)
+ * must be preserved in addition to callee-saved a12 - a15.
+ */
+
+ENTRY(_mcount)
+#if defined(__XTENSA_WINDOWED_ABI__)
+	abi_entry_default
+
+	movi	a4, ftrace_trace_function
+	l32i	a4, a4, 0
+	movi	a3, ftrace_stub
+	bne	a3, a4, 1f
+	abi_ret_default
+
+1: 	xor	a7, a2, a1
+	movi	a3, 0x3fffffff
+	and	a7, a7, a3
+	xor	a7, a7, a1
+
+	xor	a6, a0, a1
+	and	a6, a6, a3
+	xor	a6, a6, a1
+	addi	a6, a6, -MCOUNT_INSN_SIZE
+	callx4	a4
+
+	abi_ret_default
+#elif defined(__XTENSA_CALL0_ABI__)
+	abi_entry_default
+
+	movi	a9, ftrace_trace_function
+	l32i	a9, a9, 0
+	movi	a11, ftrace_stub
+	bne	a9, a11, 1f
+	abi_ret_default
+
+1:	abi_entry(28)
+	s32i	a0, sp, 0
+	s32i	a2, sp, 4
+	s32i	a3, sp, 8
+	s32i	a4, sp, 12
+	s32i	a5, sp, 16
+	s32i	a6, sp, 20
+	s32i	a7, sp, 24
+	addi	a2, a10, -MCOUNT_INSN_SIZE
+	callx0	a9
+	l32i	a0, sp, 0
+	l32i	a2, sp, 4
+	l32i	a3, sp, 8
+	l32i	a4, sp, 12
+	l32i	a5, sp, 16
+	l32i	a6, sp, 20
+	l32i	a7, sp, 24
+	abi_ret(28)
+#else
+#error Unsupported Xtensa ABI
+#endif
+ENDPROC(_mcount)
+EXPORT_SYMBOL(_mcount)
+
+ENTRY(ftrace_stub)
+	abi_entry_default
+	abi_ret_default
+ENDPROC(ftrace_stub)
diff --git a/arch/xtensa/kernel/module.c b/arch/xtensa/kernel/module.c
new file mode 100644
index 000000000..902845dda
--- /dev/null
+++ b/arch/xtensa/kernel/module.c
@@ -0,0 +1,189 @@
+/*
+ * arch/xtensa/kernel/module.c
+ *
+ * Module support.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2006 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ *
+ */
+
+#include <linux/module.h>
+#include <linux/moduleloader.h>
+#include <linux/elf.h>
+#include <linux/vmalloc.h>
+#include <linux/fs.h>
+#include <linux/string.h>
+#include <linux/kernel.h>
+#include <linux/cache.h>
+
+static int
+decode_calln_opcode (unsigned char *location)
+{
+#ifdef __XTENSA_EB__
+	return (location[0] & 0xf0) == 0x50;
+#endif
+#ifdef __XTENSA_EL__
+	return (location[0] & 0xf) == 0x5;
+#endif
+}
+
+static int
+decode_l32r_opcode (unsigned char *location)
+{
+#ifdef __XTENSA_EB__
+	return (location[0] & 0xf0) == 0x10;
+#endif
+#ifdef __XTENSA_EL__
+	return (location[0] & 0xf) == 0x1;
+#endif
+}
+
+int apply_relocate_add(Elf32_Shdr *sechdrs,
+		       const char *strtab,
+		       unsigned int symindex,
+		       unsigned int relsec,
+		       struct module *mod)
+{
+	unsigned int i;
+	Elf32_Rela *rela = (void *)sechdrs[relsec].sh_addr;
+	Elf32_Sym *sym;
+	unsigned char *location;
+	uint32_t value;
+
+	pr_debug("Applying relocate section %u to %u\n", relsec,
+		 sechdrs[relsec].sh_info);
+
+	for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rela); i++) {
+		location = (char *)sechdrs[sechdrs[relsec].sh_info].sh_addr
+			+ rela[i].r_offset;
+		sym = (Elf32_Sym *)sechdrs[symindex].sh_addr
+			+ ELF32_R_SYM(rela[i].r_info);
+		value = sym->st_value + rela[i].r_addend;
+
+		switch (ELF32_R_TYPE(rela[i].r_info)) {
+		case R_XTENSA_NONE:
+		case R_XTENSA_DIFF8:
+		case R_XTENSA_DIFF16:
+		case R_XTENSA_DIFF32:
+		case R_XTENSA_ASM_EXPAND:
+			break;
+
+		case R_XTENSA_32:
+		case R_XTENSA_PLT:
+			*(uint32_t *)location += value;
+			break;
+
+		case R_XTENSA_SLOT0_OP:
+			if (decode_calln_opcode(location)) {
+				value -= ((unsigned long)location & -4) + 4;
+				if ((value & 3) != 0 ||
+				    ((value + (1 << 19)) >> 20) != 0) {
+					pr_err("%s: relocation out of range, "
+					       "section %d reloc %d "
+					       "sym '%s'\n",
+					       mod->name, relsec, i,
+					       strtab + sym->st_name);
+					return -ENOEXEC;
+				}
+				value = (signed int)value >> 2;
+#ifdef __XTENSA_EB__
+				location[0] = ((location[0] & ~0x3) |
+					    ((value >> 16) & 0x3));
+				location[1] = (value >> 8) & 0xff;
+				location[2] = value & 0xff;
+#endif
+#ifdef __XTENSA_EL__
+				location[0] = ((location[0] & ~0xc0) |
+					    ((value << 6) & 0xc0));
+				location[1] = (value >> 2) & 0xff;
+				location[2] = (value >> 10) & 0xff;
+#endif
+			} else if (decode_l32r_opcode(location)) {
+				value -= (((unsigned long)location + 3) & -4);
+				if ((value & 3) != 0 ||
+				    (signed int)value >> 18 != -1) {
+					pr_err("%s: relocation out of range, "
+					       "section %d reloc %d "
+					       "sym '%s'\n",
+					       mod->name, relsec, i,
+					       strtab + sym->st_name);
+					return -ENOEXEC;
+				}
+				value = (signed int)value >> 2;
+
+#ifdef __XTENSA_EB__
+				location[1] = (value >> 8) & 0xff;
+				location[2] = value & 0xff;
+#endif
+#ifdef __XTENSA_EL__
+				location[1] = value & 0xff;
+				location[2] = (value >> 8) & 0xff;
+#endif
+			}
+			/* FIXME: Ignore any other opcodes.  The Xtensa
+			   assembler currently assumes that the linker will
+			   always do relaxation and so all PC-relative
+			   operands need relocations.  (The assembler also
+			   writes out the tentative PC-relative values,
+			   assuming no link-time relaxation, so it is usually
+			   safe to ignore the relocations.)  If the
+			   assembler's "--no-link-relax" flag can be made to
+			   work, and if all kernel modules can be assembled
+			   with that flag, then unexpected relocations could
+			   be detected here.  */
+			break;
+
+		case R_XTENSA_SLOT1_OP:
+		case R_XTENSA_SLOT2_OP:
+		case R_XTENSA_SLOT3_OP:
+		case R_XTENSA_SLOT4_OP:
+		case R_XTENSA_SLOT5_OP:
+		case R_XTENSA_SLOT6_OP:
+		case R_XTENSA_SLOT7_OP:
+		case R_XTENSA_SLOT8_OP:
+		case R_XTENSA_SLOT9_OP:
+		case R_XTENSA_SLOT10_OP:
+		case R_XTENSA_SLOT11_OP:
+		case R_XTENSA_SLOT12_OP:
+		case R_XTENSA_SLOT13_OP:
+		case R_XTENSA_SLOT14_OP:
+			pr_err("%s: unexpected FLIX relocation: %u\n",
+			       mod->name,
+			       ELF32_R_TYPE(rela[i].r_info));
+			return -ENOEXEC;
+
+		case R_XTENSA_SLOT0_ALT:
+		case R_XTENSA_SLOT1_ALT:
+		case R_XTENSA_SLOT2_ALT:
+		case R_XTENSA_SLOT3_ALT:
+		case R_XTENSA_SLOT4_ALT:
+		case R_XTENSA_SLOT5_ALT:
+		case R_XTENSA_SLOT6_ALT:
+		case R_XTENSA_SLOT7_ALT:
+		case R_XTENSA_SLOT8_ALT:
+		case R_XTENSA_SLOT9_ALT:
+		case R_XTENSA_SLOT10_ALT:
+		case R_XTENSA_SLOT11_ALT:
+		case R_XTENSA_SLOT12_ALT:
+		case R_XTENSA_SLOT13_ALT:
+		case R_XTENSA_SLOT14_ALT:
+			pr_err("%s: unexpected ALT relocation: %u\n",
+			       mod->name,
+			       ELF32_R_TYPE(rela[i].r_info));
+			return -ENOEXEC;
+
+		default:
+			pr_err("%s: unexpected relocation: %u\n",
+			       mod->name,
+			       ELF32_R_TYPE(rela[i].r_info));
+			return -ENOEXEC;
+		}
+	}
+	return 0;
+}
diff --git a/arch/xtensa/kernel/mxhead.S b/arch/xtensa/kernel/mxhead.S
new file mode 100644
index 000000000..b702c0908
--- /dev/null
+++ b/arch/xtensa/kernel/mxhead.S
@@ -0,0 +1,64 @@
+/*
+ * Xtensa Secondary Processors startup code.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ *
+ * Joe Taylor <joe@tensilica.com>
+ * Chris Zankel <chris@zankel.net>
+ * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Pete Delaney <piet@tensilica.com>
+ */
+
+#include <linux/linkage.h>
+
+#include <asm/cacheasm.h>
+#include <asm/initialize_mmu.h>
+#include <asm/mxregs.h>
+#include <asm/regs.h>
+
+
+	.section .SecondaryResetVector.text, "ax"
+
+
+ENTRY(_SecondaryResetVector)
+	_j _SetupOCD
+
+	.begin  no-absolute-literals
+	.literal_position
+
+_SetupOCD:
+	/*
+	 * Initialize WB, WS, and clear PS.EXCM (to allow loop instructions).
+	 * Set Interrupt Level just below XCHAL_DEBUGLEVEL to allow
+	 * xt-gdb to single step via DEBUG exceptions received directly
+	 * by ocd.
+	 */
+#if XCHAL_HAVE_WINDOWED
+	movi	a1, 1
+	movi	a0, 0
+	wsr	a1, windowstart
+	wsr	a0, windowbase
+	rsync
+#endif
+
+	movi	a1, LOCKLEVEL
+	wsr	a1, ps
+	rsync
+
+_SetupMMU:
+#ifdef CONFIG_INITIALIZE_XTENSA_MMU_INSIDE_VMLINUX
+	initialize_mmu
+#endif
+
+	/*
+	 * Start Secondary Processors with NULL pointer to boot params.
+	 */
+	movi	a2, 0				#  a2 == NULL
+	movi	a3, _startup
+	jx	a3
+
+	.end    no-absolute-literals
diff --git a/arch/xtensa/kernel/pci-dma.c b/arch/xtensa/kernel/pci-dma.c
new file mode 100644
index 000000000..94955caa4
--- /dev/null
+++ b/arch/xtensa/kernel/pci-dma.c
@@ -0,0 +1,98 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * DMA coherent memory allocation.
+ *
+ * Copyright (C) 2002 - 2005 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ *
+ * Based on version for i386.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ */
+
+#include <linux/dma-map-ops.h>
+#include <linux/dma-direct.h>
+#include <linux/gfp.h>
+#include <linux/highmem.h>
+#include <linux/mm.h>
+#include <linux/types.h>
+#include <asm/cacheflush.h>
+#include <asm/io.h>
+#include <asm/platform.h>
+
+static void do_cache_op(phys_addr_t paddr, size_t size,
+			void (*fn)(unsigned long, unsigned long))
+{
+	unsigned long off = paddr & (PAGE_SIZE - 1);
+	unsigned long pfn = PFN_DOWN(paddr);
+	struct page *page = pfn_to_page(pfn);
+
+	if (!PageHighMem(page))
+		fn((unsigned long)phys_to_virt(paddr), size);
+	else
+		while (size > 0) {
+			size_t sz = min_t(size_t, size, PAGE_SIZE - off);
+			void *vaddr = kmap_atomic(page);
+
+			fn((unsigned long)vaddr + off, sz);
+			kunmap_atomic(vaddr);
+			off = 0;
+			++page;
+			size -= sz;
+		}
+}
+
+void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size,
+		enum dma_data_direction dir)
+{
+	switch (dir) {
+	case DMA_BIDIRECTIONAL:
+	case DMA_FROM_DEVICE:
+		do_cache_op(paddr, size, __invalidate_dcache_range);
+		break;
+
+	case DMA_NONE:
+		BUG();
+		break;
+
+	default:
+		break;
+	}
+}
+
+void arch_sync_dma_for_device(phys_addr_t paddr, size_t size,
+		enum dma_data_direction dir)
+{
+	switch (dir) {
+	case DMA_BIDIRECTIONAL:
+	case DMA_TO_DEVICE:
+		if (XCHAL_DCACHE_IS_WRITEBACK)
+			do_cache_op(paddr, size, __flush_dcache_range);
+		break;
+
+	case DMA_NONE:
+		BUG();
+		break;
+
+	default:
+		break;
+	}
+}
+
+void arch_dma_prep_coherent(struct page *page, size_t size)
+{
+	__invalidate_dcache_range((unsigned long)page_address(page), size);
+}
+
+/*
+ * Memory caching is platform-dependent in noMMU xtensa configurations.
+ * This function should be implemented in platform code in order to enable
+ * coherent DMA memory operations when CONFIG_MMU is not enabled.
+ */
+#ifdef CONFIG_MMU
+void *arch_dma_set_uncached(void *p, size_t size)
+{
+	return p + XCHAL_KSEG_BYPASS_VADDR - XCHAL_KSEG_CACHED_VADDR;
+}
+#endif /* CONFIG_MMU */
diff --git a/arch/xtensa/kernel/pci.c b/arch/xtensa/kernel/pci.c
new file mode 100644
index 000000000..305031551
--- /dev/null
+++ b/arch/xtensa/kernel/pci.c
@@ -0,0 +1,90 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * arch/xtensa/kernel/pci.c
+ *
+ * PCI bios-type initialisation for PCI machines
+ *
+ * Copyright (C) 2001-2005 Tensilica Inc.
+ *
+ * Based largely on work from Cort (ppc/kernel/pci.c)
+ * IO functions copied from sparc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ */
+
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <linux/string.h>
+#include <linux/init.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/memblock.h>
+
+#include <asm/pci-bridge.h>
+#include <asm/platform.h>
+
+/*
+ * We need to avoid collisions with `mirrored' VGA ports
+ * and other strange ISA hardware, so we always want the
+ * addresses to be allocated in the 0x000-0x0ff region
+ * modulo 0x400.
+ *
+ * Why? Because some silly external IO cards only decode
+ * the low 10 bits of the IO address. The 0x00-0xff region
+ * is reserved for motherboard devices that decode all 16
+ * bits, so it's ok to allocate at, say, 0x2800-0x28ff,
+ * but we want to try to avoid allocating at 0x2900-0x2bff
+ * which might have be mirrored at 0x0100-0x03ff..
+ */
+resource_size_t
+pcibios_align_resource(void *data, const struct resource *res,
+		       const struct resource *empty_res,
+		       resource_size_t size, resource_size_t align)
+{
+	struct pci_dev *dev = data;
+	resource_size_t start = res->start;
+
+	if (res->flags & IORESOURCE_IO) {
+		if (size > 0x100) {
+			pr_err("PCI: I/O Region %s/%d too large (%u bytes)\n",
+					pci_name(dev), dev->resource - res,
+					size);
+		}
+
+		if (start & 0x300)
+			start = (start + 0x3ff) & ~0x3ff;
+	} else if (res->flags & IORESOURCE_MEM) {
+		start = pci_align_resource(dev, res, empty_res, size, align);
+	}
+
+	return start;
+}
+
+void pcibios_fixup_bus(struct pci_bus *bus)
+{
+	if (bus->parent) {
+		/* This is a subordinate bridge */
+		pci_read_bridge_bases(bus);
+	}
+}
+
+/*
+ * Platform support for /proc/bus/pci/X/Y mmap()s.
+ *  -- paulus.
+ */
+
+int pci_iobar_pfn(struct pci_dev *pdev, int bar, struct vm_area_struct *vma)
+{
+	struct pci_controller *pci_ctrl = (struct pci_controller*) pdev->sysdata;
+	resource_size_t ioaddr = pci_resource_start(pdev, bar);
+
+	if (!pci_ctrl)
+		return -EINVAL;		/* should never happen */
+
+	/* Convert to an offset within this PCI controller */
+	ioaddr -= (unsigned long)pci_ctrl->io_space.base;
+
+	vma->vm_pgoff += (ioaddr + pci_ctrl->io_space.start) >> PAGE_SHIFT;
+	return 0;
+}
diff --git a/arch/xtensa/kernel/perf_event.c b/arch/xtensa/kernel/perf_event.c
new file mode 100644
index 000000000..223f1d452
--- /dev/null
+++ b/arch/xtensa/kernel/perf_event.c
@@ -0,0 +1,449 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Xtensa Performance Monitor Module driver
+ * See Tensilica Debug User's Guide for PMU registers documentation.
+ *
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+
+#include <linux/interrupt.h>
+#include <linux/irqdomain.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/perf_event.h>
+#include <linux/platform_device.h>
+
+#include <asm/core.h>
+#include <asm/processor.h>
+#include <asm/stacktrace.h>
+
+#define XTENSA_HWVERSION_RG_2015_0	260000
+
+#if XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RG_2015_0
+#define XTENSA_PMU_ERI_BASE		0x00101000
+#else
+#define XTENSA_PMU_ERI_BASE		0x00001000
+#endif
+
+/* Global control/status for all perf counters */
+#define XTENSA_PMU_PMG			XTENSA_PMU_ERI_BASE
+/* Perf counter values */
+#define XTENSA_PMU_PM(i)		(XTENSA_PMU_ERI_BASE + 0x80 + (i) * 4)
+/* Perf counter control registers */
+#define XTENSA_PMU_PMCTRL(i)		(XTENSA_PMU_ERI_BASE + 0x100 + (i) * 4)
+/* Perf counter status registers */
+#define XTENSA_PMU_PMSTAT(i)		(XTENSA_PMU_ERI_BASE + 0x180 + (i) * 4)
+
+#define XTENSA_PMU_PMG_PMEN		0x1
+
+#define XTENSA_PMU_COUNTER_MASK		0xffffffffULL
+#define XTENSA_PMU_COUNTER_MAX		0x7fffffff
+
+#define XTENSA_PMU_PMCTRL_INTEN		0x00000001
+#define XTENSA_PMU_PMCTRL_KRNLCNT	0x00000008
+#define XTENSA_PMU_PMCTRL_TRACELEVEL	0x000000f0
+#define XTENSA_PMU_PMCTRL_SELECT_SHIFT	8
+#define XTENSA_PMU_PMCTRL_SELECT	0x00001f00
+#define XTENSA_PMU_PMCTRL_MASK_SHIFT	16
+#define XTENSA_PMU_PMCTRL_MASK		0xffff0000
+
+#define XTENSA_PMU_MASK(select, mask) \
+	(((select) << XTENSA_PMU_PMCTRL_SELECT_SHIFT) | \
+	 ((mask) << XTENSA_PMU_PMCTRL_MASK_SHIFT) | \
+	 XTENSA_PMU_PMCTRL_TRACELEVEL | \
+	 XTENSA_PMU_PMCTRL_INTEN)
+
+#define XTENSA_PMU_PMSTAT_OVFL		0x00000001
+#define XTENSA_PMU_PMSTAT_INTASRT	0x00000010
+
+struct xtensa_pmu_events {
+	/* Array of events currently on this core */
+	struct perf_event *event[XCHAL_NUM_PERF_COUNTERS];
+	/* Bitmap of used hardware counters */
+	unsigned long used_mask[BITS_TO_LONGS(XCHAL_NUM_PERF_COUNTERS)];
+};
+static DEFINE_PER_CPU(struct xtensa_pmu_events, xtensa_pmu_events);
+
+static const u32 xtensa_hw_ctl[] = {
+	[PERF_COUNT_HW_CPU_CYCLES]		= XTENSA_PMU_MASK(0, 0x1),
+	[PERF_COUNT_HW_INSTRUCTIONS]		= XTENSA_PMU_MASK(2, 0xffff),
+	[PERF_COUNT_HW_CACHE_REFERENCES]	= XTENSA_PMU_MASK(10, 0x1),
+	[PERF_COUNT_HW_CACHE_MISSES]		= XTENSA_PMU_MASK(12, 0x1),
+	/* Taken and non-taken branches + taken loop ends */
+	[PERF_COUNT_HW_BRANCH_INSTRUCTIONS]	= XTENSA_PMU_MASK(2, 0x490),
+	/* Instruction-related + other global stall cycles */
+	[PERF_COUNT_HW_STALLED_CYCLES_FRONTEND]	= XTENSA_PMU_MASK(4, 0x1ff),
+	/* Data-related global stall cycles */
+	[PERF_COUNT_HW_STALLED_CYCLES_BACKEND]	= XTENSA_PMU_MASK(3, 0x1ff),
+};
+
+#define C(_x) PERF_COUNT_HW_CACHE_##_x
+
+static const u32 xtensa_cache_ctl[][C(OP_MAX)][C(RESULT_MAX)] = {
+	[C(L1D)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= XTENSA_PMU_MASK(10, 0x1),
+			[C(RESULT_MISS)]	= XTENSA_PMU_MASK(10, 0x2),
+		},
+		[C(OP_WRITE)] = {
+			[C(RESULT_ACCESS)]	= XTENSA_PMU_MASK(11, 0x1),
+			[C(RESULT_MISS)]	= XTENSA_PMU_MASK(11, 0x2),
+		},
+	},
+	[C(L1I)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= XTENSA_PMU_MASK(8, 0x1),
+			[C(RESULT_MISS)]	= XTENSA_PMU_MASK(8, 0x2),
+		},
+	},
+	[C(DTLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= XTENSA_PMU_MASK(9, 0x1),
+			[C(RESULT_MISS)]	= XTENSA_PMU_MASK(9, 0x8),
+		},
+	},
+	[C(ITLB)] = {
+		[C(OP_READ)] = {
+			[C(RESULT_ACCESS)]	= XTENSA_PMU_MASK(7, 0x1),
+			[C(RESULT_MISS)]	= XTENSA_PMU_MASK(7, 0x8),
+		},
+	},
+};
+
+static int xtensa_pmu_cache_event(u64 config)
+{
+	unsigned int cache_type, cache_op, cache_result;
+	int ret;
+
+	cache_type = (config >>  0) & 0xff;
+	cache_op = (config >>  8) & 0xff;
+	cache_result = (config >> 16) & 0xff;
+
+	if (cache_type >= ARRAY_SIZE(xtensa_cache_ctl) ||
+	    cache_op >= C(OP_MAX) ||
+	    cache_result >= C(RESULT_MAX))
+		return -EINVAL;
+
+	ret = xtensa_cache_ctl[cache_type][cache_op][cache_result];
+
+	if (ret == 0)
+		return -EINVAL;
+
+	return ret;
+}
+
+static inline uint32_t xtensa_pmu_read_counter(int idx)
+{
+	return get_er(XTENSA_PMU_PM(idx));
+}
+
+static inline void xtensa_pmu_write_counter(int idx, uint32_t v)
+{
+	set_er(v, XTENSA_PMU_PM(idx));
+}
+
+static void xtensa_perf_event_update(struct perf_event *event,
+				     struct hw_perf_event *hwc, int idx)
+{
+	uint64_t prev_raw_count, new_raw_count;
+	int64_t delta;
+
+	do {
+		prev_raw_count = local64_read(&hwc->prev_count);
+		new_raw_count = xtensa_pmu_read_counter(event->hw.idx);
+	} while (local64_cmpxchg(&hwc->prev_count, prev_raw_count,
+				 new_raw_count) != prev_raw_count);
+
+	delta = (new_raw_count - prev_raw_count) & XTENSA_PMU_COUNTER_MASK;
+
+	local64_add(delta, &event->count);
+	local64_sub(delta, &hwc->period_left);
+}
+
+static bool xtensa_perf_event_set_period(struct perf_event *event,
+					 struct hw_perf_event *hwc, int idx)
+{
+	bool rc = false;
+	s64 left;
+
+	if (!is_sampling_event(event)) {
+		left = XTENSA_PMU_COUNTER_MAX;
+	} else {
+		s64 period = hwc->sample_period;
+
+		left = local64_read(&hwc->period_left);
+		if (left <= -period) {
+			left = period;
+			local64_set(&hwc->period_left, left);
+			hwc->last_period = period;
+			rc = true;
+		} else if (left <= 0) {
+			left += period;
+			local64_set(&hwc->period_left, left);
+			hwc->last_period = period;
+			rc = true;
+		}
+		if (left > XTENSA_PMU_COUNTER_MAX)
+			left = XTENSA_PMU_COUNTER_MAX;
+	}
+
+	local64_set(&hwc->prev_count, -left);
+	xtensa_pmu_write_counter(idx, -left);
+	perf_event_update_userpage(event);
+
+	return rc;
+}
+
+static void xtensa_pmu_enable(struct pmu *pmu)
+{
+	set_er(get_er(XTENSA_PMU_PMG) | XTENSA_PMU_PMG_PMEN, XTENSA_PMU_PMG);
+}
+
+static void xtensa_pmu_disable(struct pmu *pmu)
+{
+	set_er(get_er(XTENSA_PMU_PMG) & ~XTENSA_PMU_PMG_PMEN, XTENSA_PMU_PMG);
+}
+
+static int xtensa_pmu_event_init(struct perf_event *event)
+{
+	int ret;
+
+	switch (event->attr.type) {
+	case PERF_TYPE_HARDWARE:
+		if (event->attr.config >= ARRAY_SIZE(xtensa_hw_ctl) ||
+		    xtensa_hw_ctl[event->attr.config] == 0)
+			return -EINVAL;
+		event->hw.config = xtensa_hw_ctl[event->attr.config];
+		return 0;
+
+	case PERF_TYPE_HW_CACHE:
+		ret = xtensa_pmu_cache_event(event->attr.config);
+		if (ret < 0)
+			return ret;
+		event->hw.config = ret;
+		return 0;
+
+	case PERF_TYPE_RAW:
+		/* Not 'previous counter' select */
+		if ((event->attr.config & XTENSA_PMU_PMCTRL_SELECT) ==
+		    (1 << XTENSA_PMU_PMCTRL_SELECT_SHIFT))
+			return -EINVAL;
+		event->hw.config = (event->attr.config &
+				    (XTENSA_PMU_PMCTRL_KRNLCNT |
+				     XTENSA_PMU_PMCTRL_TRACELEVEL |
+				     XTENSA_PMU_PMCTRL_SELECT |
+				     XTENSA_PMU_PMCTRL_MASK)) |
+			XTENSA_PMU_PMCTRL_INTEN;
+		return 0;
+
+	default:
+		return -ENOENT;
+	}
+}
+
+/*
+ * Starts/Stops a counter present on the PMU. The PMI handler
+ * should stop the counter when perf_event_overflow() returns
+ * !0. ->start() will be used to continue.
+ */
+static void xtensa_pmu_start(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (WARN_ON_ONCE(idx == -1))
+		return;
+
+	if (flags & PERF_EF_RELOAD) {
+		WARN_ON_ONCE(!(event->hw.state & PERF_HES_UPTODATE));
+		xtensa_perf_event_set_period(event, hwc, idx);
+	}
+
+	hwc->state = 0;
+
+	set_er(hwc->config, XTENSA_PMU_PMCTRL(idx));
+}
+
+static void xtensa_pmu_stop(struct perf_event *event, int flags)
+{
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (!(hwc->state & PERF_HES_STOPPED)) {
+		set_er(0, XTENSA_PMU_PMCTRL(idx));
+		set_er(get_er(XTENSA_PMU_PMSTAT(idx)),
+		       XTENSA_PMU_PMSTAT(idx));
+		hwc->state |= PERF_HES_STOPPED;
+	}
+
+	if ((flags & PERF_EF_UPDATE) &&
+	    !(event->hw.state & PERF_HES_UPTODATE)) {
+		xtensa_perf_event_update(event, &event->hw, idx);
+		event->hw.state |= PERF_HES_UPTODATE;
+	}
+}
+
+/*
+ * Adds/Removes a counter to/from the PMU, can be done inside
+ * a transaction, see the ->*_txn() methods.
+ */
+static int xtensa_pmu_add(struct perf_event *event, int flags)
+{
+	struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
+	struct hw_perf_event *hwc = &event->hw;
+	int idx = hwc->idx;
+
+	if (__test_and_set_bit(idx, ev->used_mask)) {
+		idx = find_first_zero_bit(ev->used_mask,
+					  XCHAL_NUM_PERF_COUNTERS);
+		if (idx == XCHAL_NUM_PERF_COUNTERS)
+			return -EAGAIN;
+
+		__set_bit(idx, ev->used_mask);
+		hwc->idx = idx;
+	}
+	ev->event[idx] = event;
+
+	hwc->state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
+
+	if (flags & PERF_EF_START)
+		xtensa_pmu_start(event, PERF_EF_RELOAD);
+
+	perf_event_update_userpage(event);
+	return 0;
+}
+
+static void xtensa_pmu_del(struct perf_event *event, int flags)
+{
+	struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
+
+	xtensa_pmu_stop(event, PERF_EF_UPDATE);
+	__clear_bit(event->hw.idx, ev->used_mask);
+	perf_event_update_userpage(event);
+}
+
+static void xtensa_pmu_read(struct perf_event *event)
+{
+	xtensa_perf_event_update(event, &event->hw, event->hw.idx);
+}
+
+static int callchain_trace(struct stackframe *frame, void *data)
+{
+	struct perf_callchain_entry_ctx *entry = data;
+
+	perf_callchain_store(entry, frame->pc);
+	return 0;
+}
+
+void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry,
+			   struct pt_regs *regs)
+{
+	xtensa_backtrace_kernel(regs, entry->max_stack,
+				callchain_trace, NULL, entry);
+}
+
+void perf_callchain_user(struct perf_callchain_entry_ctx *entry,
+			 struct pt_regs *regs)
+{
+	xtensa_backtrace_user(regs, entry->max_stack,
+			      callchain_trace, entry);
+}
+
+void perf_event_print_debug(void)
+{
+	unsigned long flags;
+	unsigned i;
+
+	local_irq_save(flags);
+	pr_info("CPU#%d: PMG: 0x%08lx\n", smp_processor_id(),
+		get_er(XTENSA_PMU_PMG));
+	for (i = 0; i < XCHAL_NUM_PERF_COUNTERS; ++i)
+		pr_info("PM%d: 0x%08lx, PMCTRL%d: 0x%08lx, PMSTAT%d: 0x%08lx\n",
+			i, get_er(XTENSA_PMU_PM(i)),
+			i, get_er(XTENSA_PMU_PMCTRL(i)),
+			i, get_er(XTENSA_PMU_PMSTAT(i)));
+	local_irq_restore(flags);
+}
+
+irqreturn_t xtensa_pmu_irq_handler(int irq, void *dev_id)
+{
+	irqreturn_t rc = IRQ_NONE;
+	struct xtensa_pmu_events *ev = this_cpu_ptr(&xtensa_pmu_events);
+	unsigned i;
+
+	for_each_set_bit(i, ev->used_mask, XCHAL_NUM_PERF_COUNTERS) {
+		uint32_t v = get_er(XTENSA_PMU_PMSTAT(i));
+		struct perf_event *event = ev->event[i];
+		struct hw_perf_event *hwc = &event->hw;
+		u64 last_period;
+
+		if (!(v & XTENSA_PMU_PMSTAT_OVFL))
+			continue;
+
+		set_er(v, XTENSA_PMU_PMSTAT(i));
+		xtensa_perf_event_update(event, hwc, i);
+		last_period = hwc->last_period;
+		if (xtensa_perf_event_set_period(event, hwc, i)) {
+			struct perf_sample_data data;
+			struct pt_regs *regs = get_irq_regs();
+
+			perf_sample_data_init(&data, 0, last_period);
+			perf_event_overflow(event, &data, regs);
+		}
+
+		rc = IRQ_HANDLED;
+	}
+	return rc;
+}
+
+static struct pmu xtensa_pmu = {
+	.pmu_enable = xtensa_pmu_enable,
+	.pmu_disable = xtensa_pmu_disable,
+	.event_init = xtensa_pmu_event_init,
+	.add = xtensa_pmu_add,
+	.del = xtensa_pmu_del,
+	.start = xtensa_pmu_start,
+	.stop = xtensa_pmu_stop,
+	.read = xtensa_pmu_read,
+};
+
+static int xtensa_pmu_setup(unsigned int cpu)
+{
+	unsigned i;
+
+	set_er(0, XTENSA_PMU_PMG);
+	for (i = 0; i < XCHAL_NUM_PERF_COUNTERS; ++i) {
+		set_er(0, XTENSA_PMU_PMCTRL(i));
+		set_er(get_er(XTENSA_PMU_PMSTAT(i)), XTENSA_PMU_PMSTAT(i));
+	}
+	return 0;
+}
+
+static int __init xtensa_pmu_init(void)
+{
+	int ret;
+	int irq = irq_create_mapping(NULL, XCHAL_PROFILING_INTERRUPT);
+
+	ret = cpuhp_setup_state(CPUHP_AP_PERF_XTENSA_STARTING,
+				"perf/xtensa:starting", xtensa_pmu_setup,
+				NULL);
+	if (ret) {
+		pr_err("xtensa_pmu: failed to register CPU-hotplug.\n");
+		return ret;
+	}
+#if XTENSA_FAKE_NMI
+	enable_irq(irq);
+#else
+	ret = request_irq(irq, xtensa_pmu_irq_handler, IRQF_PERCPU,
+			  "pmu", NULL);
+	if (ret < 0)
+		return ret;
+#endif
+
+	ret = perf_pmu_register(&xtensa_pmu, "cpu", PERF_TYPE_RAW);
+	if (ret)
+		free_irq(irq, NULL);
+
+	return ret;
+}
+early_initcall(xtensa_pmu_init);
diff --git a/arch/xtensa/kernel/platform.c b/arch/xtensa/kernel/platform.c
new file mode 100644
index 000000000..f14713060
--- /dev/null
+++ b/arch/xtensa/kernel/platform.c
@@ -0,0 +1,45 @@
+/*
+ * arch/xtensa/kernel/platform.c
+ *
+ * Default platform functions.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2005 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ */
+
+#include <linux/printk.h>
+#include <linux/types.h>
+#include <linux/units.h>
+#include <asm/platform.h>
+#include <asm/timex.h>
+
+/*
+ * Default functions that are used if no platform specific function is defined.
+ * (Please, refer to arch/xtensa/include/asm/platform.h for more information)
+ */
+
+void __weak __init platform_init(bp_tag_t *first)
+{
+}
+
+void __weak __init platform_setup(char **cmd)
+{
+}
+
+void __weak platform_idle(void)
+{
+	__asm__ __volatile__ ("waiti 0" ::: "memory");
+}
+
+#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
+void __weak platform_calibrate_ccount(void)
+{
+	pr_err("ERROR: Cannot calibrate cpu frequency! Assuming 10 MHz.\n");
+	ccount_freq = 10 * HZ_PER_MHZ;
+}
+#endif
diff --git a/arch/xtensa/kernel/process.c b/arch/xtensa/kernel/process.c
new file mode 100644
index 000000000..94d43f44b
--- /dev/null
+++ b/arch/xtensa/kernel/process.c
@@ -0,0 +1,400 @@
+/*
+ * arch/xtensa/kernel/process.c
+ *
+ * Xtensa Processor version.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ *
+ * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ * Chris Zankel <chris@zankel.net>
+ * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Kevin Chea
+ */
+
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/sched/debug.h>
+#include <linux/sched/task.h>
+#include <linux/sched/task_stack.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/smp.h>
+#include <linux/stddef.h>
+#include <linux/unistd.h>
+#include <linux/ptrace.h>
+#include <linux/elf.h>
+#include <linux/hw_breakpoint.h>
+#include <linux/init.h>
+#include <linux/prctl.h>
+#include <linux/init_task.h>
+#include <linux/module.h>
+#include <linux/mqueue.h>
+#include <linux/fs.h>
+#include <linux/slab.h>
+#include <linux/rcupdate.h>
+
+#include <linux/uaccess.h>
+#include <asm/io.h>
+#include <asm/processor.h>
+#include <asm/platform.h>
+#include <asm/mmu.h>
+#include <asm/irq.h>
+#include <linux/atomic.h>
+#include <asm/asm-offsets.h>
+#include <asm/regs.h>
+#include <asm/hw_breakpoint.h>
+#include <asm/sections.h>
+#include <asm/traps.h>
+
+extern void ret_from_fork(void);
+extern void ret_from_kernel_thread(void);
+
+void (*pm_power_off)(void) = NULL;
+EXPORT_SYMBOL(pm_power_off);
+
+
+#ifdef CONFIG_STACKPROTECTOR
+#include <linux/stackprotector.h>
+unsigned long __stack_chk_guard __read_mostly;
+EXPORT_SYMBOL(__stack_chk_guard);
+#endif
+
+#if XTENSA_HAVE_COPROCESSORS
+
+void local_coprocessors_flush_release_all(void)
+{
+	struct thread_info **coprocessor_owner;
+	struct thread_info *unique_owner[XCHAL_CP_MAX];
+	int n = 0;
+	int i, j;
+
+	coprocessor_owner = this_cpu_ptr(&exc_table)->coprocessor_owner;
+	xtensa_set_sr(XCHAL_CP_MASK, cpenable);
+
+	for (i = 0; i < XCHAL_CP_MAX; i++) {
+		struct thread_info *ti = coprocessor_owner[i];
+
+		if (ti) {
+			coprocessor_flush(ti, i);
+
+			for (j = 0; j < n; j++)
+				if (unique_owner[j] == ti)
+					break;
+			if (j == n)
+				unique_owner[n++] = ti;
+
+			coprocessor_owner[i] = NULL;
+		}
+	}
+	for (i = 0; i < n; i++) {
+		/* pairs with memw (1) in fast_coprocessor and memw in switch_to */
+		smp_wmb();
+		unique_owner[i]->cpenable = 0;
+	}
+	xtensa_set_sr(0, cpenable);
+}
+
+static void local_coprocessor_release_all(void *info)
+{
+	struct thread_info *ti = info;
+	struct thread_info **coprocessor_owner;
+	int i;
+
+	coprocessor_owner = this_cpu_ptr(&exc_table)->coprocessor_owner;
+
+	/* Walk through all cp owners and release it for the requested one. */
+
+	for (i = 0; i < XCHAL_CP_MAX; i++) {
+		if (coprocessor_owner[i] == ti)
+			coprocessor_owner[i] = NULL;
+	}
+	/* pairs with memw (1) in fast_coprocessor and memw in switch_to */
+	smp_wmb();
+	ti->cpenable = 0;
+	if (ti == current_thread_info())
+		xtensa_set_sr(0, cpenable);
+}
+
+void coprocessor_release_all(struct thread_info *ti)
+{
+	if (ti->cpenable) {
+		/* pairs with memw (2) in fast_coprocessor */
+		smp_rmb();
+		smp_call_function_single(ti->cp_owner_cpu,
+					 local_coprocessor_release_all,
+					 ti, true);
+	}
+}
+
+static void local_coprocessor_flush_all(void *info)
+{
+	struct thread_info *ti = info;
+	struct thread_info **coprocessor_owner;
+	unsigned long old_cpenable;
+	int i;
+
+	coprocessor_owner = this_cpu_ptr(&exc_table)->coprocessor_owner;
+	old_cpenable = xtensa_xsr(ti->cpenable, cpenable);
+
+	for (i = 0; i < XCHAL_CP_MAX; i++) {
+		if (coprocessor_owner[i] == ti)
+			coprocessor_flush(ti, i);
+	}
+	xtensa_set_sr(old_cpenable, cpenable);
+}
+
+void coprocessor_flush_all(struct thread_info *ti)
+{
+	if (ti->cpenable) {
+		/* pairs with memw (2) in fast_coprocessor */
+		smp_rmb();
+		smp_call_function_single(ti->cp_owner_cpu,
+					 local_coprocessor_flush_all,
+					 ti, true);
+	}
+}
+
+static void local_coprocessor_flush_release_all(void *info)
+{
+	local_coprocessor_flush_all(info);
+	local_coprocessor_release_all(info);
+}
+
+void coprocessor_flush_release_all(struct thread_info *ti)
+{
+	if (ti->cpenable) {
+		/* pairs with memw (2) in fast_coprocessor */
+		smp_rmb();
+		smp_call_function_single(ti->cp_owner_cpu,
+					 local_coprocessor_flush_release_all,
+					 ti, true);
+	}
+}
+
+#endif
+
+
+/*
+ * Powermanagement idle function, if any is provided by the platform.
+ */
+void arch_cpu_idle(void)
+{
+	platform_idle();
+	raw_local_irq_disable();
+}
+
+/*
+ * This is called when the thread calls exit().
+ */
+void exit_thread(struct task_struct *tsk)
+{
+#if XTENSA_HAVE_COPROCESSORS
+	coprocessor_release_all(task_thread_info(tsk));
+#endif
+}
+
+/*
+ * Flush thread state. This is called when a thread does an execve()
+ * Note that we flush coprocessor registers for the case execve fails.
+ */
+void flush_thread(void)
+{
+#if XTENSA_HAVE_COPROCESSORS
+	struct thread_info *ti = current_thread_info();
+	coprocessor_flush_release_all(ti);
+#endif
+	flush_ptrace_hw_breakpoint(current);
+}
+
+/*
+ * this gets called so that we can store coprocessor state into memory and
+ * copy the current task into the new thread.
+ */
+int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
+{
+#if XTENSA_HAVE_COPROCESSORS
+	coprocessor_flush_all(task_thread_info(src));
+#endif
+	*dst = *src;
+	return 0;
+}
+
+/*
+ * Copy thread.
+ *
+ * There are two modes in which this function is called:
+ * 1) Userspace thread creation,
+ *    regs != NULL, usp_thread_fn is userspace stack pointer.
+ *    It is expected to copy parent regs (in case CLONE_VM is not set
+ *    in the clone_flags) and set up passed usp in the childregs.
+ * 2) Kernel thread creation,
+ *    regs == NULL, usp_thread_fn is the function to run in the new thread
+ *    and thread_fn_arg is its parameter.
+ *    childregs are not used for the kernel threads.
+ *
+ * The stack layout for the new thread looks like this:
+ *
+ *	+------------------------+
+ *	|       childregs        |
+ *	+------------------------+ <- thread.sp = sp in dummy-frame
+ *	|      dummy-frame       |    (saved in dummy-frame spill-area)
+ *	+------------------------+
+ *
+ * We create a dummy frame to return to either ret_from_fork or
+ *   ret_from_kernel_thread:
+ *   a0 points to ret_from_fork/ret_from_kernel_thread (simulating a call4)
+ *   sp points to itself (thread.sp)
+ *   a2, a3 are unused for userspace threads,
+ *   a2 points to thread_fn, a3 holds thread_fn arg for kernel threads.
+ *
+ * Note: This is a pristine frame, so we don't need any spill region on top of
+ *       childregs.
+ *
+ * The fun part:  if we're keeping the same VM (i.e. cloning a thread,
+ * not an entire process), we're normally given a new usp, and we CANNOT share
+ * any live address register windows.  If we just copy those live frames over,
+ * the two threads (parent and child) will overflow the same frames onto the
+ * parent stack at different times, likely corrupting the parent stack (esp.
+ * if the parent returns from functions that called clone() and calls new
+ * ones, before the child overflows its now old copies of its parent windows).
+ * One solution is to spill windows to the parent stack, but that's fairly
+ * involved.  Much simpler to just not copy those live frames across.
+ */
+
+int copy_thread(struct task_struct *p, const struct kernel_clone_args *args)
+{
+	u64 clone_flags = args->flags;
+	unsigned long usp_thread_fn = args->stack;
+	unsigned long tls = args->tls;
+	struct pt_regs *childregs = task_pt_regs(p);
+
+#if (XTENSA_HAVE_COPROCESSORS || XTENSA_HAVE_IO_PORTS)
+	struct thread_info *ti;
+#endif
+
+#if defined(__XTENSA_WINDOWED_ABI__)
+	/* Create a call4 dummy-frame: a0 = 0, a1 = childregs. */
+	SPILL_SLOT(childregs, 1) = (unsigned long)childregs;
+	SPILL_SLOT(childregs, 0) = 0;
+
+	p->thread.sp = (unsigned long)childregs;
+#elif defined(__XTENSA_CALL0_ABI__)
+	/* Reserve 16 bytes for the _switch_to stack frame. */
+	p->thread.sp = (unsigned long)childregs - 16;
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+	if (!args->fn) {
+		struct pt_regs *regs = current_pt_regs();
+		unsigned long usp = usp_thread_fn ?
+			usp_thread_fn : regs->areg[1];
+
+		p->thread.ra = MAKE_RA_FOR_CALL(
+				(unsigned long)ret_from_fork, 0x1);
+
+		*childregs = *regs;
+		childregs->areg[1] = usp;
+		childregs->areg[2] = 0;
+
+		/* When sharing memory with the parent thread, the child
+		   usually starts on a pristine stack, so we have to reset
+		   windowbase, windowstart and wmask.
+		   (Note that such a new thread is required to always create
+		   an initial call4 frame)
+		   The exception is vfork, where the new thread continues to
+		   run on the parent's stack until it calls execve. This could
+		   be a call8 or call12, which requires a legal stack frame
+		   of the previous caller for the overflow handlers to work.
+		   (Note that it's always legal to overflow live registers).
+		   In this case, ensure to spill at least the stack pointer
+		   of that frame. */
+
+		if (clone_flags & CLONE_VM) {
+			/* check that caller window is live and same stack */
+			int len = childregs->wmask & ~0xf;
+			if (regs->areg[1] == usp && len != 0) {
+				int callinc = (regs->areg[0] >> 30) & 3;
+				int caller_ars = XCHAL_NUM_AREGS - callinc * 4;
+				put_user(regs->areg[caller_ars+1],
+					 (unsigned __user*)(usp - 12));
+			}
+			childregs->wmask = 1;
+			childregs->windowstart = 1;
+			childregs->windowbase = 0;
+		}
+
+		if (clone_flags & CLONE_SETTLS)
+			childregs->threadptr = tls;
+	} else {
+		p->thread.ra = MAKE_RA_FOR_CALL(
+				(unsigned long)ret_from_kernel_thread, 1);
+
+		/* pass parameters to ret_from_kernel_thread: */
+#if defined(__XTENSA_WINDOWED_ABI__)
+		/*
+		 * a2 = thread_fn, a3 = thread_fn arg.
+		 * Window underflow will load registers from the
+		 * spill slots on the stack on return from _switch_to.
+		 */
+		SPILL_SLOT(childregs, 2) = (unsigned long)args->fn;
+		SPILL_SLOT(childregs, 3) = (unsigned long)args->fn_arg;
+#elif defined(__XTENSA_CALL0_ABI__)
+		/*
+		 * a12 = thread_fn, a13 = thread_fn arg.
+		 * _switch_to epilogue will load registers from the stack.
+		 */
+		((unsigned long *)p->thread.sp)[0] = (unsigned long)args->fn;
+		((unsigned long *)p->thread.sp)[1] = (unsigned long)args->fn_arg;
+#else
+#error Unsupported Xtensa ABI
+#endif
+
+		/* Childregs are only used when we're going to userspace
+		 * in which case start_thread will set them up.
+		 */
+	}
+
+#if (XTENSA_HAVE_COPROCESSORS || XTENSA_HAVE_IO_PORTS)
+	ti = task_thread_info(p);
+	ti->cpenable = 0;
+#endif
+
+	clear_ptrace_hw_breakpoint(p);
+
+	return 0;
+}
+
+
+/*
+ * These bracket the sleeping functions..
+ */
+
+unsigned long __get_wchan(struct task_struct *p)
+{
+	unsigned long sp, pc;
+	unsigned long stack_page = (unsigned long) task_stack_page(p);
+	int count = 0;
+
+	sp = p->thread.sp;
+	pc = MAKE_PC_FROM_RA(p->thread.ra, _text);
+
+	do {
+		if (sp < stack_page + sizeof(struct task_struct) ||
+		    sp >= (stack_page + THREAD_SIZE) ||
+		    pc == 0)
+			return 0;
+		if (!in_sched_functions(pc))
+			return pc;
+
+		/* Stack layout: sp-4: ra, sp-3: sp' */
+
+		pc = MAKE_PC_FROM_RA(SPILL_SLOT(sp, 0), _text);
+		sp = SPILL_SLOT(sp, 1);
+	} while (count++ < 16);
+	return 0;
+}
diff --git a/arch/xtensa/kernel/ptrace.c b/arch/xtensa/kernel/ptrace.c
new file mode 100644
index 000000000..364e4fdab
--- /dev/null
+++ b/arch/xtensa/kernel/ptrace.c
@@ -0,0 +1,583 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007  Tensilica Inc.
+ *
+ * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
+ * Chris Zankel <chris@zankel.net>
+ * Scott Foehner<sfoehner@yahoo.com>,
+ * Kevin Chea
+ * Marc Gauthier<marc@tensilica.com> <marc@alumni.uwaterloo.ca>
+ */
+
+#include <linux/audit.h>
+#include <linux/errno.h>
+#include <linux/hw_breakpoint.h>
+#include <linux/kernel.h>
+#include <linux/mm.h>
+#include <linux/perf_event.h>
+#include <linux/ptrace.h>
+#include <linux/regset.h>
+#include <linux/sched.h>
+#include <linux/sched/task_stack.h>
+#include <linux/seccomp.h>
+#include <linux/security.h>
+#include <linux/signal.h>
+#include <linux/smp.h>
+#include <linux/uaccess.h>
+
+#define CREATE_TRACE_POINTS
+#include <trace/events/syscalls.h>
+
+#include <asm/coprocessor.h>
+#include <asm/elf.h>
+#include <asm/page.h>
+#include <asm/ptrace.h>
+
+static int gpr_get(struct task_struct *target,
+		   const struct user_regset *regset,
+		   struct membuf to)
+{
+	struct pt_regs *regs = task_pt_regs(target);
+	struct user_pt_regs newregs = {
+		.pc = regs->pc,
+		.ps = regs->ps & ~(1 << PS_EXCM_BIT),
+		.lbeg = regs->lbeg,
+		.lend = regs->lend,
+		.lcount = regs->lcount,
+		.sar = regs->sar,
+		.threadptr = regs->threadptr,
+		.windowbase = regs->windowbase,
+		.windowstart = regs->windowstart,
+		.syscall = regs->syscall,
+	};
+
+	memcpy(newregs.a,
+	       regs->areg + XCHAL_NUM_AREGS - regs->windowbase * 4,
+	       regs->windowbase * 16);
+	memcpy(newregs.a + regs->windowbase * 4,
+	       regs->areg,
+	       (WSBITS - regs->windowbase) * 16);
+
+	return membuf_write(&to, &newregs, sizeof(newregs));
+}
+
+static int gpr_set(struct task_struct *target,
+		   const struct user_regset *regset,
+		   unsigned int pos, unsigned int count,
+		   const void *kbuf, const void __user *ubuf)
+{
+	int ret;
+	struct user_pt_regs newregs = {0};
+	struct pt_regs *regs;
+	const u32 ps_mask = PS_CALLINC_MASK | PS_OWB_MASK;
+
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf, &newregs, 0, -1);
+	if (ret)
+		return ret;
+
+	if (newregs.windowbase >= XCHAL_NUM_AREGS / 4)
+		return -EINVAL;
+
+	regs = task_pt_regs(target);
+	regs->pc = newregs.pc;
+	regs->ps = (regs->ps & ~ps_mask) | (newregs.ps & ps_mask);
+	regs->lbeg = newregs.lbeg;
+	regs->lend = newregs.lend;
+	regs->lcount = newregs.lcount;
+	regs->sar = newregs.sar;
+	regs->threadptr = newregs.threadptr;
+
+	if (newregs.syscall)
+		regs->syscall = newregs.syscall;
+
+	if (newregs.windowbase != regs->windowbase ||
+	    newregs.windowstart != regs->windowstart) {
+		u32 rotws, wmask;
+
+		rotws = (((newregs.windowstart |
+			   (newregs.windowstart << WSBITS)) >>
+			  newregs.windowbase) &
+			 ((1 << WSBITS) - 1)) & ~1;
+		wmask = ((rotws ? WSBITS + 1 - ffs(rotws) : 0) << 4) |
+			(rotws & 0xF) | 1;
+		regs->windowbase = newregs.windowbase;
+		regs->windowstart = newregs.windowstart;
+		regs->wmask = wmask;
+	}
+
+	memcpy(regs->areg + XCHAL_NUM_AREGS - newregs.windowbase * 4,
+	       newregs.a, newregs.windowbase * 16);
+	memcpy(regs->areg, newregs.a + newregs.windowbase * 4,
+	       (WSBITS - newregs.windowbase) * 16);
+
+	return 0;
+}
+
+static int tie_get(struct task_struct *target,
+		   const struct user_regset *regset,
+		   struct membuf to)
+{
+	int ret;
+	struct pt_regs *regs = task_pt_regs(target);
+	struct thread_info *ti = task_thread_info(target);
+	elf_xtregs_t *newregs = kzalloc_obj(elf_xtregs_t);
+
+	if (!newregs)
+		return -ENOMEM;
+
+	newregs->opt = regs->xtregs_opt;
+	newregs->user = ti->xtregs_user;
+
+#if XTENSA_HAVE_COPROCESSORS
+	/* Flush all coprocessor registers to memory. */
+	coprocessor_flush_all(ti);
+	newregs->cp0 = ti->xtregs_cp.cp0;
+	newregs->cp1 = ti->xtregs_cp.cp1;
+	newregs->cp2 = ti->xtregs_cp.cp2;
+	newregs->cp3 = ti->xtregs_cp.cp3;
+	newregs->cp4 = ti->xtregs_cp.cp4;
+	newregs->cp5 = ti->xtregs_cp.cp5;
+	newregs->cp6 = ti->xtregs_cp.cp6;
+	newregs->cp7 = ti->xtregs_cp.cp7;
+#endif
+	ret = membuf_write(&to, newregs, sizeof(*newregs));
+	kfree(newregs);
+	return ret;
+}
+
+static int tie_set(struct task_struct *target,
+		   const struct user_regset *regset,
+		   unsigned int pos, unsigned int count,
+		   const void *kbuf, const void __user *ubuf)
+{
+	int ret;
+	struct pt_regs *regs = task_pt_regs(target);
+	struct thread_info *ti = task_thread_info(target);
+	elf_xtregs_t *newregs = kzalloc_obj(elf_xtregs_t);
+
+	if (!newregs)
+		return -ENOMEM;
+
+	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
+				 newregs, 0, -1);
+
+	if (ret)
+		goto exit;
+	regs->xtregs_opt = newregs->opt;
+	ti->xtregs_user = newregs->user;
+
+#if XTENSA_HAVE_COPROCESSORS
+	/* Flush all coprocessors before we overwrite them. */
+	coprocessor_flush_release_all(ti);
+	ti->xtregs_cp.cp0 = newregs->cp0;
+	ti->xtregs_cp.cp1 = newregs->cp1;
+	ti->xtregs_cp.cp2 = newregs->cp2;
+	ti->xtregs_cp.cp3 = newregs->cp3;
+	ti->xtregs_cp.cp4 = newregs->cp4;
+	ti->xtregs_cp.cp5 = newregs->cp5;
+	ti->xtregs_cp.cp6 = newregs->cp6;
+	ti->xtregs_cp.cp7 = newregs->cp7;
+#endif
+exit:
+	kfree(newregs);
+	return ret;
+}
+
+enum xtensa_regset {
+	REGSET_GPR,
+	REGSET_TIE,
+};
+
+static const struct user_regset xtensa_regsets[] = {
+	[REGSET_GPR] = {
+		USER_REGSET_NOTE_TYPE(PRSTATUS),
+		.n = sizeof(struct user_pt_regs) / sizeof(u32),
+		.size = sizeof(u32),
+		.align = sizeof(u32),
+		.regset_get = gpr_get,
+		.set = gpr_set,
+	},
+	[REGSET_TIE] = {
+		USER_REGSET_NOTE_TYPE(PRFPREG),
+		.n = sizeof(elf_xtregs_t) / sizeof(u32),
+		.size = sizeof(u32),
+		.align = sizeof(u32),
+		.regset_get = tie_get,
+		.set = tie_set,
+	},
+};
+
+static const struct user_regset_view user_xtensa_view = {
+	.name = "xtensa",
+	.e_machine = EM_XTENSA,
+	.regsets = xtensa_regsets,
+	.n = ARRAY_SIZE(xtensa_regsets)
+};
+
+const struct user_regset_view *task_user_regset_view(struct task_struct *task)
+{
+	return &user_xtensa_view;
+}
+
+void user_enable_single_step(struct task_struct *child)
+{
+	set_tsk_thread_flag(child, TIF_SINGLESTEP);
+}
+
+void user_disable_single_step(struct task_struct *child)
+{
+	clear_tsk_thread_flag(child, TIF_SINGLESTEP);
+}
+
+/*
+ * Called by kernel/ptrace.c when detaching to disable single stepping.
+ */
+
+void ptrace_disable(struct task_struct *child)
+{
+	/* Nothing to do.. */
+}
+
+static int ptrace_getregs(struct task_struct *child, void __user *uregs)
+{
+	return copy_regset_to_user(child, &user_xtensa_view, REGSET_GPR,
+				   0, sizeof(xtensa_gregset_t), uregs);
+}
+
+static int ptrace_setregs(struct task_struct *child, void __user *uregs)
+{
+	return copy_regset_from_user(child, &user_xtensa_view, REGSET_GPR,
+				     0, sizeof(xtensa_gregset_t), uregs);
+}
+
+static int ptrace_getxregs(struct task_struct *child, void __user *uregs)
+{
+	return copy_regset_to_user(child, &user_xtensa_view, REGSET_TIE,
+				   0, sizeof(elf_xtregs_t), uregs);
+}
+
+static int ptrace_setxregs(struct task_struct *child, void __user *uregs)
+{
+	return copy_regset_from_user(child, &user_xtensa_view, REGSET_TIE,
+				     0, sizeof(elf_xtregs_t), uregs);
+}
+
+static int ptrace_peekusr(struct task_struct *child, long regno,
+			  long __user *ret)
+{
+	struct pt_regs *regs;
+	unsigned long tmp;
+
+	regs = task_pt_regs(child);
+	tmp = 0;  /* Default return value. */
+
+	switch(regno) {
+	case REG_AR_BASE ... REG_AR_BASE + XCHAL_NUM_AREGS - 1:
+		tmp = regs->areg[regno - REG_AR_BASE];
+		break;
+
+	case REG_A_BASE ... REG_A_BASE + 15:
+		tmp = regs->areg[regno - REG_A_BASE];
+		break;
+
+	case REG_PC:
+		tmp = regs->pc;
+		break;
+
+	case REG_PS:
+		/* Note: PS.EXCM is not set while user task is running;
+		 * its being set in regs is for exception handling
+		 * convenience.
+		 */
+		tmp = (regs->ps & ~(1 << PS_EXCM_BIT));
+		break;
+
+	case REG_WB:
+		break;		/* tmp = 0 */
+
+	case REG_WS:
+		{
+			unsigned long wb = regs->windowbase;
+			unsigned long ws = regs->windowstart;
+			tmp = ((ws >> wb) | (ws << (WSBITS - wb))) &
+				((1 << WSBITS) - 1);
+			break;
+		}
+	case REG_LBEG:
+		tmp = regs->lbeg;
+		break;
+
+	case REG_LEND:
+		tmp = regs->lend;
+		break;
+
+	case REG_LCOUNT:
+		tmp = regs->lcount;
+		break;
+
+	case REG_SAR:
+		tmp = regs->sar;
+		break;
+
+	case SYSCALL_NR:
+		tmp = regs->syscall;
+		break;
+
+	default:
+		return -EIO;
+	}
+	return put_user(tmp, ret);
+}
+
+static int ptrace_pokeusr(struct task_struct *child, long regno, long val)
+{
+	struct pt_regs *regs;
+	regs = task_pt_regs(child);
+
+	switch (regno) {
+	case REG_AR_BASE ... REG_AR_BASE + XCHAL_NUM_AREGS - 1:
+		regs->areg[regno - REG_AR_BASE] = val;
+		break;
+
+	case REG_A_BASE ... REG_A_BASE + 15:
+		regs->areg[regno - REG_A_BASE] = val;
+		break;
+
+	case REG_PC:
+		regs->pc = val;
+		break;
+
+	case SYSCALL_NR:
+		regs->syscall = val;
+		break;
+
+	default:
+		return -EIO;
+	}
+	return 0;
+}
+
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+static void ptrace_hbptriggered(struct perf_event *bp,
+				struct perf_sample_data *data,
+				struct pt_regs *regs)
+{
+	int i;
+	struct arch_hw_breakpoint *bkpt = counter_arch_bp(bp);
+
+	if (bp->attr.bp_type & HW_BREAKPOINT_X) {
+		for (i = 0; i < XCHAL_NUM_IBREAK; ++i)
+			if (current->thread.ptrace_bp[i] == bp)
+				break;
+		i <<= 1;
+	} else {
+		for (i = 0; i < XCHAL_NUM_DBREAK; ++i)
+			if (current->thread.ptrace_wp[i] == bp)
+				break;
+		i = (i << 1) | 1;
+	}
+
+	force_sig_ptrace_errno_trap(i, (void __user *)bkpt->address);
+}
+
+static struct perf_event *ptrace_hbp_create(struct task_struct *tsk, int type)
+{
+	struct perf_event_attr attr;
+
+	ptrace_breakpoint_init(&attr);
+
+	/* Initialise fields to sane defaults. */
+	attr.bp_addr	= 0;
+	attr.bp_len	= 1;
+	attr.bp_type	= type;
+	attr.disabled	= 1;
+
+	return register_user_hw_breakpoint(&attr, ptrace_hbptriggered, NULL,
+					   tsk);
+}
+
+/*
+ * Address bit 0 choose instruction (0) or data (1) break register, bits
+ * 31..1 are the register number.
+ * Both PTRACE_GETHBPREGS and PTRACE_SETHBPREGS transfer two 32-bit words:
+ * address (0) and control (1).
+ * Instruction breakpoint contorl word is 0 to clear breakpoint, 1 to set.
+ * Data breakpoint control word bit 31 is 'trigger on store', bit 30 is
+ * 'trigger on load, bits 29..0 are length. Length 0 is used to clear a
+ * breakpoint. To set a breakpoint length must be a power of 2 in the range
+ * 1..64 and the address must be length-aligned.
+ */
+
+static long ptrace_gethbpregs(struct task_struct *child, long addr,
+			      long __user *datap)
+{
+	struct perf_event *bp;
+	u32 user_data[2] = {0};
+	bool dbreak = addr & 1;
+	unsigned idx = addr >> 1;
+
+	if ((!dbreak && idx >= XCHAL_NUM_IBREAK) ||
+	    (dbreak && idx >= XCHAL_NUM_DBREAK))
+		return -EINVAL;
+
+	if (dbreak)
+		bp = child->thread.ptrace_wp[idx];
+	else
+		bp = child->thread.ptrace_bp[idx];
+
+	if (bp) {
+		user_data[0] = bp->attr.bp_addr;
+		user_data[1] = bp->attr.disabled ? 0 : bp->attr.bp_len;
+		if (dbreak) {
+			if (bp->attr.bp_type & HW_BREAKPOINT_R)
+				user_data[1] |= DBREAKC_LOAD_MASK;
+			if (bp->attr.bp_type & HW_BREAKPOINT_W)
+				user_data[1] |= DBREAKC_STOR_MASK;
+		}
+	}
+
+	if (copy_to_user(datap, user_data, sizeof(user_data)))
+		return -EFAULT;
+
+	return 0;
+}
+
+static long ptrace_sethbpregs(struct task_struct *child, long addr,
+			      long __user *datap)
+{
+	struct perf_event *bp;
+	struct perf_event_attr attr;
+	u32 user_data[2];
+	bool dbreak = addr & 1;
+	unsigned idx = addr >> 1;
+	int bp_type = 0;
+
+	if ((!dbreak && idx >= XCHAL_NUM_IBREAK) ||
+	    (dbreak && idx >= XCHAL_NUM_DBREAK))
+		return -EINVAL;
+
+	if (copy_from_user(user_data, datap, sizeof(user_data)))
+		return -EFAULT;
+
+	if (dbreak) {
+		bp = child->thread.ptrace_wp[idx];
+		if (user_data[1] & DBREAKC_LOAD_MASK)
+			bp_type |= HW_BREAKPOINT_R;
+		if (user_data[1] & DBREAKC_STOR_MASK)
+			bp_type |= HW_BREAKPOINT_W;
+	} else {
+		bp = child->thread.ptrace_bp[idx];
+		bp_type = HW_BREAKPOINT_X;
+	}
+
+	if (!bp) {
+		bp = ptrace_hbp_create(child,
+				       bp_type ? bp_type : HW_BREAKPOINT_RW);
+		if (IS_ERR(bp))
+			return PTR_ERR(bp);
+		if (dbreak)
+			child->thread.ptrace_wp[idx] = bp;
+		else
+			child->thread.ptrace_bp[idx] = bp;
+	}
+
+	attr = bp->attr;
+	attr.bp_addr = user_data[0];
+	attr.bp_len = user_data[1] & ~(DBREAKC_LOAD_MASK | DBREAKC_STOR_MASK);
+	attr.bp_type = bp_type;
+	attr.disabled = !attr.bp_len;
+
+	return modify_user_hw_breakpoint(bp, &attr);
+}
+#endif
+
+long arch_ptrace(struct task_struct *child, long request,
+		 unsigned long addr, unsigned long data)
+{
+	int ret = -EPERM;
+	void __user *datap = (void __user *) data;
+
+	switch (request) {
+	case PTRACE_PEEKUSR:	/* read register specified by addr. */
+		ret = ptrace_peekusr(child, addr, datap);
+		break;
+
+	case PTRACE_POKEUSR:	/* write register specified by addr. */
+		ret = ptrace_pokeusr(child, addr, data);
+		break;
+
+	case PTRACE_GETREGS:
+		ret = ptrace_getregs(child, datap);
+		break;
+
+	case PTRACE_SETREGS:
+		ret = ptrace_setregs(child, datap);
+		break;
+
+	case PTRACE_GETXTREGS:
+		ret = ptrace_getxregs(child, datap);
+		break;
+
+	case PTRACE_SETXTREGS:
+		ret = ptrace_setxregs(child, datap);
+		break;
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	case PTRACE_GETHBPREGS:
+		ret = ptrace_gethbpregs(child, addr, datap);
+		break;
+
+	case PTRACE_SETHBPREGS:
+		ret = ptrace_sethbpregs(child, addr, datap);
+		break;
+#endif
+	default:
+		ret = ptrace_request(child, request, addr, data);
+		break;
+	}
+
+	return ret;
+}
+
+int do_syscall_trace_enter(struct pt_regs *regs)
+{
+	if (regs->syscall == NO_SYSCALL)
+		regs->areg[2] = -ENOSYS;
+
+	if (test_thread_flag(TIF_SYSCALL_TRACE) &&
+	    !ptrace_report_syscall_permit_entry(regs)) {
+		regs->areg[2] = -ENOSYS;
+		regs->syscall = NO_SYSCALL;
+		return 0;
+	}
+
+	if (regs->syscall == NO_SYSCALL || !seccomp_permit_syscall()) {
+		do_syscall_trace_leave(regs);
+		return 0;
+	}
+
+	if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
+		trace_sys_enter(regs, syscall_get_nr(current, regs));
+
+	audit_syscall_entry(regs->syscall, regs->areg[6],
+			    regs->areg[3], regs->areg[4],
+			    regs->areg[5]);
+	return 1;
+}
+
+void do_syscall_trace_leave(struct pt_regs *regs)
+{
+	int step;
+
+	audit_syscall_exit(regs);
+
+	if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
+		trace_sys_exit(regs, regs_return_value(regs));
+
+	step = test_thread_flag(TIF_SINGLESTEP);
+
+	if (step || test_thread_flag(TIF_SYSCALL_TRACE))
+		ptrace_report_syscall_exit(regs, step);
+}
diff --git a/arch/xtensa/kernel/s32c1i_selftest.c b/arch/xtensa/kernel/s32c1i_selftest.c
new file mode 100644
index 000000000..8362388c8
--- /dev/null
+++ b/arch/xtensa/kernel/s32c1i_selftest.c
@@ -0,0 +1,127 @@
+/*
+ * S32C1I selftest.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2016 Cadence Design Systems Inc.
+ */
+
+#include <linux/init.h>
+#include <linux/kernel.h>
+
+#include <asm/traps.h>
+
+#if XCHAL_HAVE_S32C1I
+
+static int __initdata rcw_word, rcw_probe_pc, rcw_exc;
+
+/*
+ * Basic atomic compare-and-swap, that records PC of S32C1I for probing.
+ *
+ * If *v == cmp, set *v = set.  Return previous *v.
+ */
+static inline int probed_compare_swap(int *v, int cmp, int set)
+{
+	int tmp;
+
+	__asm__ __volatile__(
+			"	movi	%1, 1f\n"
+			"	s32i	%1, %4, 0\n"
+			"	wsr	%2, scompare1\n"
+			"1:	s32c1i	%0, %3, 0\n"
+			: "=a" (set), "=&a" (tmp)
+			: "a" (cmp), "a" (v), "a" (&rcw_probe_pc), "0" (set)
+			: "memory"
+			);
+	return set;
+}
+
+/* Handle probed exception */
+
+static void __init do_probed_exception(struct pt_regs *regs)
+{
+	if (regs->pc == rcw_probe_pc) {	/* exception on s32c1i ? */
+		regs->pc += 3;		/* skip the s32c1i instruction */
+		rcw_exc = regs->exccause;
+	} else {
+		do_unhandled(regs);
+	}
+}
+
+/* Simple test of S32C1I (soc bringup assist) */
+
+static int __init check_s32c1i(void)
+{
+	int n, cause1, cause2;
+	void *handbus, *handdata, *handaddr; /* temporarily saved handlers */
+
+	rcw_probe_pc = 0;
+	handbus  = trap_set_handler(EXCCAUSE_LOAD_STORE_ERROR,
+			do_probed_exception);
+	handdata = trap_set_handler(EXCCAUSE_LOAD_STORE_DATA_ERROR,
+			do_probed_exception);
+	handaddr = trap_set_handler(EXCCAUSE_LOAD_STORE_ADDR_ERROR,
+			do_probed_exception);
+
+	/* First try an S32C1I that does not store: */
+	rcw_exc = 0;
+	rcw_word = 1;
+	n = probed_compare_swap(&rcw_word, 0, 2);
+	cause1 = rcw_exc;
+
+	/* took exception? */
+	if (cause1 != 0) {
+		/* unclean exception? */
+		if (n != 2 || rcw_word != 1)
+			panic("S32C1I exception error");
+	} else if (rcw_word != 1 || n != 1) {
+		panic("S32C1I compare error");
+	}
+
+	/* Then an S32C1I that stores: */
+	rcw_exc = 0;
+	rcw_word = 0x1234567;
+	n = probed_compare_swap(&rcw_word, 0x1234567, 0xabcde);
+	cause2 = rcw_exc;
+
+	if (cause2 != 0) {
+		/* unclean exception? */
+		if (n != 0xabcde || rcw_word != 0x1234567)
+			panic("S32C1I exception error (b)");
+	} else if (rcw_word != 0xabcde || n != 0x1234567) {
+		panic("S32C1I store error");
+	}
+
+	/* Verify consistency of exceptions: */
+	if (cause1 || cause2) {
+		pr_warn("S32C1I took exception %d, %d\n", cause1, cause2);
+		/* If emulation of S32C1I upon bus error gets implemented,
+		 * we can get rid of this panic for single core (not SMP)
+		 */
+		panic("S32C1I exceptions not currently supported");
+	}
+	if (cause1 != cause2)
+		panic("inconsistent S32C1I exceptions");
+
+	trap_set_handler(EXCCAUSE_LOAD_STORE_ERROR, handbus);
+	trap_set_handler(EXCCAUSE_LOAD_STORE_DATA_ERROR, handdata);
+	trap_set_handler(EXCCAUSE_LOAD_STORE_ADDR_ERROR, handaddr);
+	return 0;
+}
+
+#else /* XCHAL_HAVE_S32C1I */
+
+/* This condition should not occur with a commercially deployed processor.
+ * Display reminder for early engr test or demo chips / FPGA bitstreams
+ */
+static int __init check_s32c1i(void)
+{
+	pr_warn("Processor configuration lacks atomic compare-and-swap support!\n");
+	return 0;
+}
+
+#endif /* XCHAL_HAVE_S32C1I */
+
+early_initcall(check_s32c1i);
diff --git a/arch/xtensa/kernel/setup.c b/arch/xtensa/kernel/setup.c
new file mode 100644
index 000000000..f72e28036
--- /dev/null
+++ b/arch/xtensa/kernel/setup.c
@@ -0,0 +1,722 @@
+/*
+ * arch/xtensa/kernel/setup.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1995  Linus Torvalds
+ * Copyright (C) 2001 - 2005  Tensilica Inc.
+ * Copyright (C) 2014 - 2016  Cadence Design Systems Inc.
+ *
+ * Chris Zankel	<chris@zankel.net>
+ * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
+ * Kevin Chea
+ * Marc Gauthier<marc@tensilica.com> <marc@alumni.uwaterloo.ca>
+ */
+
+#include <linux/errno.h>
+#include <linux/init.h>
+#include <linux/mm.h>
+#include <linux/proc_fs.h>
+#include <linux/kernel.h>
+#include <linux/percpu.h>
+#include <linux/reboot.h>
+#include <linux/cpu.h>
+#include <linux/of.h>
+#include <linux/of_fdt.h>
+
+#if defined(CONFIG_VGA_CONSOLE) || defined(CONFIG_DUMMY_CONSOLE)
+# include <linux/console.h>
+#endif
+
+#ifdef CONFIG_PROC_FS
+# include <linux/seq_file.h>
+#endif
+
+#include <asm/bootparam.h>
+#include <asm/kasan.h>
+#include <asm/mmu_context.h>
+#include <asm/page.h>
+#include <asm/param.h>
+#include <asm/platform.h>
+#include <asm/processor.h>
+#include <asm/sections.h>
+#include <asm/setup.h>
+#include <asm/smp.h>
+#include <asm/sysmem.h>
+#include <asm/timex.h>
+#include <asm/traps.h>
+
+#ifdef CONFIG_BLK_DEV_INITRD
+extern unsigned long initrd_start;
+extern unsigned long initrd_end;
+extern int initrd_below_start_ok;
+#endif
+
+#ifdef CONFIG_USE_OF
+static void *dtb_start __initdata = __dtb_start;
+#endif
+
+extern unsigned long loops_per_jiffy;
+
+/* Command line specified as configuration option. */
+
+static char __initdata command_line[COMMAND_LINE_SIZE];
+
+#ifdef CONFIG_CMDLINE_BOOL
+static char default_command_line[COMMAND_LINE_SIZE] __initdata = CONFIG_CMDLINE;
+#endif
+
+#ifdef CONFIG_PARSE_BOOTPARAM
+/*
+ * Boot parameter parsing.
+ *
+ * The Xtensa port uses a list of variable-sized tags to pass data to
+ * the kernel. The first tag must be a BP_TAG_FIRST tag for the list
+ * to be recognised. The list is terminated with a zero-sized
+ * BP_TAG_LAST tag.
+ */
+
+typedef struct tagtable {
+	u32 tag;
+	int (*parse)(const bp_tag_t*);
+} tagtable_t;
+
+#define __tagtable(tag, fn) static tagtable_t __tagtable_##fn 		\
+	__section(".taglist") __attribute__((used)) = { tag, fn }
+
+/* parse current tag */
+
+static int __init parse_tag_mem(const bp_tag_t *tag)
+{
+	struct bp_meminfo *mi = (struct bp_meminfo *)(tag->data);
+
+	if (mi->type != MEMORY_TYPE_CONVENTIONAL)
+		return -1;
+
+	return memblock_add(mi->start, mi->end - mi->start);
+}
+
+__tagtable(BP_TAG_MEMORY, parse_tag_mem);
+
+#ifdef CONFIG_BLK_DEV_INITRD
+
+static int __init parse_tag_initrd(const bp_tag_t* tag)
+{
+	struct bp_meminfo *mi = (struct bp_meminfo *)(tag->data);
+
+	initrd_start = (unsigned long)__va(mi->start);
+	initrd_end = (unsigned long)__va(mi->end);
+
+	return 0;
+}
+
+__tagtable(BP_TAG_INITRD, parse_tag_initrd);
+
+#endif /* CONFIG_BLK_DEV_INITRD */
+
+#ifdef CONFIG_USE_OF
+
+static int __init parse_tag_fdt(const bp_tag_t *tag)
+{
+	dtb_start = __va(tag->data[0]);
+	return 0;
+}
+
+__tagtable(BP_TAG_FDT, parse_tag_fdt);
+
+#endif /* CONFIG_USE_OF */
+
+static int __init parse_tag_cmdline(const bp_tag_t* tag)
+{
+	strscpy(command_line, (char *)(tag->data), COMMAND_LINE_SIZE);
+	return 0;
+}
+
+__tagtable(BP_TAG_COMMAND_LINE, parse_tag_cmdline);
+
+static int __init parse_bootparam(const bp_tag_t* tag)
+{
+	extern tagtable_t __tagtable_begin, __tagtable_end;
+	tagtable_t *t;
+
+	/* Boot parameters must start with a BP_TAG_FIRST tag. */
+
+	if (tag->id != BP_TAG_FIRST) {
+		pr_warn("Invalid boot parameters!\n");
+		return 0;
+	}
+
+	tag = (bp_tag_t*)((unsigned long)tag + sizeof(bp_tag_t) + tag->size);
+
+	/* Parse all tags. */
+
+	while (tag != NULL && tag->id != BP_TAG_LAST) {
+		for (t = &__tagtable_begin; t < &__tagtable_end; t++) {
+			if (tag->id == t->tag) {
+				t->parse(tag);
+				break;
+			}
+		}
+		if (t == &__tagtable_end)
+			pr_warn("Ignoring tag 0x%08x\n", tag->id);
+		tag = (bp_tag_t*)((unsigned long)(tag + 1) + tag->size);
+	}
+
+	return 0;
+}
+#else
+static int __init parse_bootparam(const bp_tag_t *tag)
+{
+	pr_info("Ignoring boot parameters at %p\n", tag);
+	return 0;
+}
+#endif
+
+#ifdef CONFIG_USE_OF
+
+#if !XCHAL_HAVE_PTP_MMU || XCHAL_HAVE_SPANNING_WAY
+unsigned long xtensa_kio_paddr = XCHAL_KIO_DEFAULT_PADDR;
+EXPORT_SYMBOL(xtensa_kio_paddr);
+
+static int __init xtensa_dt_io_area(unsigned long node, const char *uname,
+		int depth, void *data)
+{
+	const __be32 *ranges;
+	int len;
+
+	if (depth > 1)
+		return 0;
+
+	if (!of_flat_dt_is_compatible(node, "simple-bus"))
+		return 0;
+
+	ranges = of_get_flat_dt_prop(node, "ranges", &len);
+	if (!ranges)
+		return 1;
+	if (len == 0)
+		return 1;
+
+	xtensa_kio_paddr = of_read_ulong(ranges+1, 1);
+	/* round down to nearest 256MB boundary */
+	xtensa_kio_paddr &= 0xf0000000;
+
+	init_kio();
+
+	return 1;
+}
+#else
+static int __init xtensa_dt_io_area(unsigned long node, const char *uname,
+		int depth, void *data)
+{
+	return 1;
+}
+#endif
+
+void __init early_init_devtree(void *params)
+{
+	early_init_dt_scan(params, __pa(params));
+	of_scan_flat_dt(xtensa_dt_io_area, NULL);
+
+	if (!command_line[0])
+		strscpy(command_line, boot_command_line, COMMAND_LINE_SIZE);
+}
+
+#endif /* CONFIG_USE_OF */
+
+/*
+ * Initialize architecture. (Early stage)
+ */
+
+void __init init_arch(bp_tag_t *bp_start)
+{
+	/* Initialize basic exception handling if configuration may need it */
+
+	if (IS_ENABLED(CONFIG_KASAN) ||
+	    IS_ENABLED(CONFIG_XTENSA_LOAD_STORE))
+		early_trap_init();
+
+	/* Initialize MMU. */
+
+	init_mmu();
+
+	/* Initialize initial KASAN shadow map */
+
+	kasan_early_init();
+
+	/* Parse boot parameters */
+
+	if (bp_start)
+		parse_bootparam(bp_start);
+
+#ifdef CONFIG_USE_OF
+	early_init_devtree(dtb_start);
+#endif
+
+#ifdef CONFIG_CMDLINE_BOOL
+	if (!command_line[0])
+		strscpy(command_line, default_command_line, COMMAND_LINE_SIZE);
+#endif
+
+	/* Early hook for platforms */
+
+	platform_init(bp_start);
+}
+
+/*
+ * Initialize system. Setup memory and reserve regions.
+ */
+
+static inline int __init_memblock mem_reserve(unsigned long start,
+					      unsigned long end)
+{
+	return memblock_reserve(start, end - start);
+}
+
+void __init setup_arch(char **cmdline_p)
+{
+	pr_info("config ID: %08x:%08x\n",
+		xtensa_get_sr(SREG_EPC), xtensa_get_sr(SREG_EXCSAVE));
+	if (xtensa_get_sr(SREG_EPC) != XCHAL_HW_CONFIGID0 ||
+	    xtensa_get_sr(SREG_EXCSAVE) != XCHAL_HW_CONFIGID1)
+		pr_info("built for config ID: %08x:%08x\n",
+			XCHAL_HW_CONFIGID0, XCHAL_HW_CONFIGID1);
+
+	*cmdline_p = command_line;
+	platform_setup(cmdline_p);
+	strscpy(boot_command_line, *cmdline_p, COMMAND_LINE_SIZE);
+
+	/* Reserve some memory regions */
+
+#ifdef CONFIG_BLK_DEV_INITRD
+	if (initrd_start < initrd_end &&
+	    !mem_reserve(__pa(initrd_start), __pa(initrd_end)))
+		initrd_below_start_ok = 1;
+	else
+		initrd_start = 0;
+#endif
+
+	mem_reserve(__pa(_stext), __pa(_end));
+#ifdef CONFIG_XIP_KERNEL
+#ifdef CONFIG_VECTORS_ADDR
+	mem_reserve(__pa(_xip_text_start), __pa(_xip_text_end));
+#endif
+	mem_reserve(__pa(_xip_start), __pa(_xip_end));
+#endif
+
+#ifdef CONFIG_VECTORS_ADDR
+#ifdef SUPPORT_WINDOWED
+	mem_reserve(__pa(_WindowVectors_text_start),
+		    __pa(_WindowVectors_text_end));
+#endif
+
+	mem_reserve(__pa(_DebugInterruptVector_text_start),
+		    __pa(_DebugInterruptVector_text_end));
+
+	mem_reserve(__pa(_KernelExceptionVector_text_start),
+		    __pa(_KernelExceptionVector_text_end));
+
+	mem_reserve(__pa(_UserExceptionVector_text_start),
+		    __pa(_UserExceptionVector_text_end));
+
+	mem_reserve(__pa(_DoubleExceptionVector_text_start),
+		    __pa(_DoubleExceptionVector_text_end));
+
+	mem_reserve(__pa(_exception_text_start),
+		    __pa(_exception_text_end));
+#if XCHAL_EXCM_LEVEL >= 2
+	mem_reserve(__pa(_Level2InterruptVector_text_start),
+		    __pa(_Level2InterruptVector_text_end));
+#endif
+#if XCHAL_EXCM_LEVEL >= 3
+	mem_reserve(__pa(_Level3InterruptVector_text_start),
+		    __pa(_Level3InterruptVector_text_end));
+#endif
+#if XCHAL_EXCM_LEVEL >= 4
+	mem_reserve(__pa(_Level4InterruptVector_text_start),
+		    __pa(_Level4InterruptVector_text_end));
+#endif
+#if XCHAL_EXCM_LEVEL >= 5
+	mem_reserve(__pa(_Level5InterruptVector_text_start),
+		    __pa(_Level5InterruptVector_text_end));
+#endif
+#if XCHAL_EXCM_LEVEL >= 6
+	mem_reserve(__pa(_Level6InterruptVector_text_start),
+		    __pa(_Level6InterruptVector_text_end));
+#endif
+
+#endif /* CONFIG_VECTORS_ADDR */
+
+#ifdef CONFIG_SECONDARY_RESET_VECTOR
+	mem_reserve(__pa(_SecondaryResetVector_text_start),
+		    __pa(_SecondaryResetVector_text_end));
+#endif
+	parse_early_param();
+	bootmem_init();
+	kasan_init();
+	unflatten_and_copy_device_tree();
+
+#ifdef CONFIG_SMP
+	smp_init_cpus();
+#endif
+
+	paging_init();
+	zones_init();
+
+#ifdef CONFIG_VT
+# if defined(CONFIG_VGA_CONSOLE)
+	conswitchp = &vga_con;
+# endif
+#endif
+}
+
+static DEFINE_PER_CPU(struct cpu, cpu_data);
+
+static int __init topology_init(void)
+{
+	int i;
+
+	for_each_possible_cpu(i) {
+		struct cpu *cpu = &per_cpu(cpu_data, i);
+		cpu->hotpluggable = !!i;
+		register_cpu(cpu, i);
+	}
+
+	return 0;
+}
+subsys_initcall(topology_init);
+
+void cpu_reset(void)
+{
+#if XCHAL_HAVE_PTP_MMU && IS_ENABLED(CONFIG_MMU)
+	local_irq_disable();
+	/*
+	 * We have full MMU: all autoload ways, ways 7, 8 and 9 of DTLB must
+	 * be flushed.
+	 * Way 4 is not currently used by linux.
+	 * Ways 5 and 6 shall not be touched on MMUv2 as they are hardwired.
+	 * Way 5 shall be flushed and way 6 shall be set to identity mapping
+	 * on MMUv3.
+	 */
+	local_flush_tlb_all();
+	invalidate_page_directory();
+#if XCHAL_HAVE_SPANNING_WAY
+	/* MMU v3 */
+	{
+		unsigned long vaddr = (unsigned long)cpu_reset;
+		unsigned long paddr = __pa(vaddr);
+		unsigned long tmpaddr = vaddr + SZ_512M;
+		unsigned long tmp0, tmp1, tmp2, tmp3;
+
+		/*
+		 * Find a place for the temporary mapping. It must not be
+		 * in the same 512MB region with vaddr or paddr, otherwise
+		 * there may be multihit exception either on entry to the
+		 * temporary mapping, or on entry to the identity mapping.
+		 * (512MB is the biggest page size supported by TLB.)
+		 */
+		while (((tmpaddr ^ paddr) & -SZ_512M) == 0)
+			tmpaddr += SZ_512M;
+
+		/* Invalidate mapping in the selected temporary area */
+		if (itlb_probe(tmpaddr) & BIT(ITLB_HIT_BIT))
+			invalidate_itlb_entry(itlb_probe(tmpaddr));
+		if (itlb_probe(tmpaddr + PAGE_SIZE) & BIT(ITLB_HIT_BIT))
+			invalidate_itlb_entry(itlb_probe(tmpaddr + PAGE_SIZE));
+
+		/*
+		 * Map two consecutive pages starting at the physical address
+		 * of this function to the temporary mapping area.
+		 */
+		write_itlb_entry(__pte((paddr & PAGE_MASK) |
+				       _PAGE_HW_VALID |
+				       _PAGE_HW_EXEC |
+				       _PAGE_CA_BYPASS),
+				 tmpaddr & PAGE_MASK);
+		write_itlb_entry(__pte(((paddr & PAGE_MASK) + PAGE_SIZE) |
+				       _PAGE_HW_VALID |
+				       _PAGE_HW_EXEC |
+				       _PAGE_CA_BYPASS),
+				 (tmpaddr & PAGE_MASK) + PAGE_SIZE);
+
+		/* Reinitialize TLB */
+		__asm__ __volatile__ ("movi	%0, 1f\n\t"
+				      "movi	%3, 2f\n\t"
+				      "add	%0, %0, %4\n\t"
+				      "add	%3, %3, %5\n\t"
+				      "jx	%0\n"
+				      /*
+				       * No literal, data or stack access
+				       * below this point
+				       */
+				      "1:\n\t"
+				      /* Initialize *tlbcfg */
+				      "movi	%0, 0\n\t"
+				      "wsr	%0, itlbcfg\n\t"
+				      "wsr	%0, dtlbcfg\n\t"
+				      /* Invalidate TLB way 5 */
+				      "movi	%0, 4\n\t"
+				      "movi	%1, 5\n"
+				      "1:\n\t"
+				      "iitlb	%1\n\t"
+				      "idtlb	%1\n\t"
+				      "add	%1, %1, %6\n\t"
+				      "addi	%0, %0, -1\n\t"
+				      "bnez	%0, 1b\n\t"
+				      /* Initialize TLB way 6 */
+				      "movi	%0, 7\n\t"
+				      "addi	%1, %9, 3\n\t"
+				      "addi	%2, %9, 6\n"
+				      "1:\n\t"
+				      "witlb	%1, %2\n\t"
+				      "wdtlb	%1, %2\n\t"
+				      "add	%1, %1, %7\n\t"
+				      "add	%2, %2, %7\n\t"
+				      "addi	%0, %0, -1\n\t"
+				      "bnez	%0, 1b\n\t"
+				      "isync\n\t"
+				      /* Jump to identity mapping */
+				      "jx	%3\n"
+				      "2:\n\t"
+				      /* Complete way 6 initialization */
+				      "witlb	%1, %2\n\t"
+				      "wdtlb	%1, %2\n\t"
+				      /* Invalidate temporary mapping */
+				      "sub	%0, %9, %7\n\t"
+				      "iitlb	%0\n\t"
+				      "add	%0, %0, %8\n\t"
+				      "iitlb	%0"
+				      : "=&a"(tmp0), "=&a"(tmp1), "=&a"(tmp2),
+					"=&a"(tmp3)
+				      : "a"(tmpaddr - vaddr),
+					"a"(paddr - vaddr),
+					"a"(SZ_128M), "a"(SZ_512M),
+					"a"(PAGE_SIZE),
+					"a"((tmpaddr + SZ_512M) & PAGE_MASK)
+				      : "memory");
+	}
+#endif
+#endif
+	__asm__ __volatile__ ("movi	a2, 0\n\t"
+			      "wsr	a2, icountlevel\n\t"
+			      "movi	a2, 0\n\t"
+			      "wsr	a2, icount\n\t"
+#if XCHAL_NUM_IBREAK > 0
+			      "wsr	a2, ibreakenable\n\t"
+#endif
+#if XCHAL_HAVE_LOOPS
+			      "wsr	a2, lcount\n\t"
+#endif
+			      "movi	a2, 0x1f\n\t"
+			      "wsr	a2, ps\n\t"
+			      "isync\n\t"
+			      "jx	%0\n\t"
+			      :
+			      : "a" (XCHAL_RESET_VECTOR_VADDR)
+			      : "a2");
+	for (;;)
+		;
+}
+
+void machine_restart(char * cmd)
+{
+	local_irq_disable();
+	smp_send_stop();
+	do_kernel_restart(cmd);
+	pr_err("Reboot failed -- System halted\n");
+	while (1)
+		cpu_relax();
+}
+
+void machine_halt(void)
+{
+	local_irq_disable();
+	smp_send_stop();
+	do_kernel_power_off();
+	while (1)
+		cpu_relax();
+}
+
+void machine_power_off(void)
+{
+	local_irq_disable();
+	smp_send_stop();
+	do_kernel_power_off();
+	while (1)
+		cpu_relax();
+}
+#ifdef CONFIG_PROC_FS
+
+/*
+ * Display some core information through /proc/cpuinfo.
+ */
+
+static int
+c_show(struct seq_file *f, void *slot)
+{
+	/* high-level stuff */
+	seq_printf(f, "CPU count\t: %u\n"
+		      "CPU list\t: %*pbl\n"
+		      "vendor_id\t: Tensilica\n"
+		      "model\t\t: Xtensa " XCHAL_HW_VERSION_NAME "\n"
+		      "core ID\t\t: " XCHAL_CORE_ID "\n"
+		      "build ID\t: 0x%x\n"
+		      "config ID\t: %08x:%08x\n"
+		      "byte order\t: %s\n"
+		      "cpu MHz\t\t: %lu.%02lu\n"
+		      "bogomips\t: %lu.%02lu\n",
+		      num_online_cpus(),
+		      cpumask_pr_args(cpu_online_mask),
+		      XCHAL_BUILD_UNIQUE_ID,
+		      xtensa_get_sr(SREG_EPC), xtensa_get_sr(SREG_EXCSAVE),
+		      XCHAL_HAVE_BE ?  "big" : "little",
+		      ccount_freq/1000000,
+		      (ccount_freq/10000) % 100,
+		      loops_per_jiffy/(500000/HZ),
+		      (loops_per_jiffy/(5000/HZ)) % 100);
+	seq_puts(f, "flags\t\t: "
+#if XCHAL_HAVE_NMI
+		     "nmi "
+#endif
+#if XCHAL_HAVE_DEBUG
+		     "debug "
+# if XCHAL_HAVE_OCD
+		     "ocd "
+# endif
+#if XCHAL_HAVE_TRAX
+		     "trax "
+#endif
+#if XCHAL_NUM_PERF_COUNTERS
+		     "perf "
+#endif
+#endif
+#if XCHAL_HAVE_DENSITY
+	    	     "density "
+#endif
+#if XCHAL_HAVE_BOOLEANS
+		     "boolean "
+#endif
+#if XCHAL_HAVE_LOOPS
+		     "loop "
+#endif
+#if XCHAL_HAVE_NSA
+		     "nsa "
+#endif
+#if XCHAL_HAVE_MINMAX
+		     "minmax "
+#endif
+#if XCHAL_HAVE_SEXT
+		     "sext "
+#endif
+#if XCHAL_HAVE_CLAMPS
+		     "clamps "
+#endif
+#if XCHAL_HAVE_MAC16
+		     "mac16 "
+#endif
+#if XCHAL_HAVE_MUL16
+		     "mul16 "
+#endif
+#if XCHAL_HAVE_MUL32
+		     "mul32 "
+#endif
+#if XCHAL_HAVE_MUL32_HIGH
+		     "mul32h "
+#endif
+#if XCHAL_HAVE_FP
+		     "fpu "
+#endif
+#if XCHAL_HAVE_S32C1I
+		     "s32c1i "
+#endif
+#if XCHAL_HAVE_EXCLUSIVE
+		     "exclusive "
+#endif
+		     "\n");
+
+	/* Registers. */
+	seq_printf(f,"physical aregs\t: %d\n"
+		     "misc regs\t: %d\n"
+		     "ibreak\t\t: %d\n"
+		     "dbreak\t\t: %d\n"
+		     "perf counters\t: %d\n",
+		     XCHAL_NUM_AREGS,
+		     XCHAL_NUM_MISC_REGS,
+		     XCHAL_NUM_IBREAK,
+		     XCHAL_NUM_DBREAK,
+		     XCHAL_NUM_PERF_COUNTERS);
+
+
+	/* Interrupt. */
+	seq_printf(f,"num ints\t: %d\n"
+		     "ext ints\t: %d\n"
+		     "int levels\t: %d\n"
+		     "timers\t\t: %d\n"
+		     "debug level\t: %d\n",
+		     XCHAL_NUM_INTERRUPTS,
+		     XCHAL_NUM_EXTINTERRUPTS,
+		     XCHAL_NUM_INTLEVELS,
+		     XCHAL_NUM_TIMERS,
+		     XCHAL_DEBUGLEVEL);
+
+	/* Cache */
+	seq_printf(f,"icache line size: %d\n"
+		     "icache ways\t: %d\n"
+		     "icache size\t: %d\n"
+		     "icache flags\t: "
+#if XCHAL_ICACHE_LINE_LOCKABLE
+		     "lock "
+#endif
+		     "\n"
+		     "dcache line size: %d\n"
+		     "dcache ways\t: %d\n"
+		     "dcache size\t: %d\n"
+		     "dcache flags\t: "
+#if XCHAL_DCACHE_IS_WRITEBACK
+		     "writeback "
+#endif
+#if XCHAL_DCACHE_LINE_LOCKABLE
+		     "lock "
+#endif
+		     "\n",
+		     XCHAL_ICACHE_LINESIZE,
+		     XCHAL_ICACHE_WAYS,
+		     XCHAL_ICACHE_SIZE,
+		     XCHAL_DCACHE_LINESIZE,
+		     XCHAL_DCACHE_WAYS,
+		     XCHAL_DCACHE_SIZE);
+
+	return 0;
+}
+
+/*
+ * We show only CPU #0 info.
+ */
+static void *
+c_start(struct seq_file *f, loff_t *pos)
+{
+	return (*pos == 0) ? (void *)1 : NULL;
+}
+
+static void *
+c_next(struct seq_file *f, void *v, loff_t *pos)
+{
+	++*pos;
+	return c_start(f, pos);
+}
+
+static void
+c_stop(struct seq_file *f, void *v)
+{
+}
+
+const struct seq_operations cpuinfo_op =
+{
+	.start	= c_start,
+	.next	= c_next,
+	.stop	= c_stop,
+	.show	= c_show,
+};
+
+#endif /* CONFIG_PROC_FS */
diff --git a/arch/xtensa/kernel/signal.c b/arch/xtensa/kernel/signal.c
new file mode 100644
index 000000000..81f0b106c
--- /dev/null
+++ b/arch/xtensa/kernel/signal.c
@@ -0,0 +1,535 @@
+/*
+ * arch/xtensa/kernel/signal.c
+ *
+ * Default platform functions.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2005, 2006 Tensilica Inc.
+ * Copyright (C) 1991, 1992  Linus Torvalds
+ * 1997-11-28  Modified for POSIX.1b signals by Richard Henderson
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Joe Taylor <joe@tensilica.com>
+ */
+
+#include <linux/signal.h>
+#include <linux/errno.h>
+#include <linux/ptrace.h>
+#include <linux/personality.h>
+#include <linux/resume_user_mode.h>
+#include <linux/sched/task_stack.h>
+
+#include <asm/ucontext.h>
+#include <linux/uaccess.h>
+#include <asm/cacheflush.h>
+#include <asm/coprocessor.h>
+#include <asm/processor.h>
+#include <asm/syscall.h>
+#include <asm/unistd.h>
+
+extern struct task_struct *coproc_owners[];
+
+struct rt_sigframe
+{
+	struct siginfo info;
+	struct ucontext uc;
+	struct {
+		xtregs_opt_t opt;
+		xtregs_user_t user;
+#if XTENSA_HAVE_COPROCESSORS
+		xtregs_coprocessor_t cp;
+#endif
+	} xtregs;
+	unsigned char retcode[6];
+	unsigned int window[4];
+};
+
+#if defined(USER_SUPPORT_WINDOWED)
+/*
+ * Flush register windows stored in pt_regs to stack.
+ * Returns 1 for errors.
+ */
+
+static int
+flush_window_regs_user(struct pt_regs *regs)
+{
+	const unsigned long ws = regs->windowstart;
+	const unsigned long wb = regs->windowbase;
+	unsigned long sp = 0;
+	unsigned long wm;
+	int err = 1;
+	int base;
+
+	/* Return if no other frames. */
+
+	if (regs->wmask == 1)
+		return 0;
+
+	/* Rotate windowmask and skip empty frames. */
+
+	wm = (ws >> wb) | (ws << (XCHAL_NUM_AREGS / 4 - wb));
+	base = (XCHAL_NUM_AREGS / 4) - (regs->wmask >> 4);
+		
+	/* For call8 or call12 frames, we need the previous stack pointer. */
+
+	if ((regs->wmask & 2) == 0)
+		if (__get_user(sp, (int*)(regs->areg[base * 4 + 1] - 12)))
+			goto errout;
+
+	/* Spill frames to stack. */
+
+	while (base < XCHAL_NUM_AREGS / 4) {
+
+		int m = (wm >> base);
+		int inc = 0;
+
+		/* Save registers a4..a7 (call8) or a4...a11 (call12) */
+
+		if (m & 2) {			/* call4 */
+			inc = 1;
+
+		} else if (m & 4) {		/* call8 */
+			if (copy_to_user(&SPILL_SLOT_CALL8(sp, 4),
+					 &regs->areg[(base + 1) * 4], 16))
+				goto errout;
+			inc = 2;
+
+		} else if (m & 8) {	/* call12 */
+			if (copy_to_user(&SPILL_SLOT_CALL12(sp, 4),
+					 &regs->areg[(base + 1) * 4], 32))
+				goto errout;
+			inc = 3;
+		}
+
+		/* Save current frame a0..a3 under next SP */
+
+		sp = regs->areg[((base + inc) * 4 + 1) % XCHAL_NUM_AREGS];
+		if (copy_to_user(&SPILL_SLOT(sp, 0), &regs->areg[base * 4], 16))
+			goto errout;
+
+		/* Get current stack pointer for next loop iteration. */
+
+		sp = regs->areg[base * 4 + 1];
+		base += inc;
+	}
+
+	regs->wmask = 1;
+	regs->windowstart = 1 << wb;
+
+	return 0;
+
+errout:
+	return err;
+}
+#else
+static int
+flush_window_regs_user(struct pt_regs *regs)
+{
+	return 0;
+}
+#endif
+
+/*
+ * Note: We don't copy double exception 'regs', we have to finish double exc. 
+ * first before we return to signal handler! This dbl.exc.handler might cause 
+ * another double exception, but I think we are fine as the situation is the 
+ * same as if we had returned to the signal handerl and got an interrupt 
+ * immediately...
+ */
+
+static int
+setup_sigcontext(struct rt_sigframe __user *frame, struct pt_regs *regs)
+{
+	struct sigcontext __user *sc = &frame->uc.uc_mcontext;
+	struct thread_info *ti = current_thread_info();
+	int err = 0;
+
+#define COPY(x)	err |= __put_user(regs->x, &sc->sc_##x)
+	COPY(pc);
+	COPY(ps);
+	COPY(lbeg);
+	COPY(lend);
+	COPY(lcount);
+	COPY(sar);
+#undef COPY
+
+	err |= flush_window_regs_user(regs);
+	err |= __copy_to_user (sc->sc_a, regs->areg, 16 * 4);
+	err |= __put_user(0, &sc->sc_xtregs);
+
+	if (err)
+		return err;
+
+#if XTENSA_HAVE_COPROCESSORS
+	coprocessor_flush_release_all(ti);
+	err |= __copy_to_user(&frame->xtregs.cp, &ti->xtregs_cp,
+			      sizeof (frame->xtregs.cp));
+#endif
+	err |= __copy_to_user(&frame->xtregs.opt, &regs->xtregs_opt,
+			      sizeof (xtregs_opt_t));
+	err |= __copy_to_user(&frame->xtregs.user, &ti->xtregs_user,
+			      sizeof (xtregs_user_t));
+
+	err |= __put_user(err ? NULL : &frame->xtregs, &sc->sc_xtregs);
+
+	return err;
+}
+
+static int
+restore_sigcontext(struct pt_regs *regs, struct rt_sigframe __user *frame)
+{
+	struct sigcontext __user *sc = &frame->uc.uc_mcontext;
+	struct thread_info *ti = current_thread_info();
+	unsigned int err = 0;
+	unsigned long ps;
+
+#define COPY(x)	err |= __get_user(regs->x, &sc->sc_##x)
+	COPY(pc);
+	COPY(lbeg);
+	COPY(lend);
+	COPY(lcount);
+	COPY(sar);
+#undef COPY
+
+	/* All registers were flushed to stack. Start with a pristine frame. */
+
+	regs->wmask = 1;
+	regs->windowbase = 0;
+	regs->windowstart = 1;
+
+	regs->syscall = NO_SYSCALL;	/* disable syscall checks */
+
+	/* For PS, restore only PS.CALLINC.
+	 * Assume that all other bits are either the same as for the signal
+	 * handler, or the user mode value doesn't matter (e.g. PS.OWB).
+	 */
+	err |= __get_user(ps, &sc->sc_ps);
+	regs->ps = (regs->ps & ~PS_CALLINC_MASK) | (ps & PS_CALLINC_MASK);
+
+	/* Additional corruption checks */
+
+	if ((regs->lcount > 0)
+	    && ((regs->lbeg > TASK_SIZE) || (regs->lend > TASK_SIZE)) )
+		err = 1;
+
+	err |= __copy_from_user(regs->areg, sc->sc_a, 16 * 4);
+
+	if (err)
+		return err;
+
+	/* The signal handler may have used coprocessors in which
+	 * case they are still enabled.  We disable them to force a
+	 * reloading of the original task's CP state by the lazy
+	 * context-switching mechanisms of CP exception handling.
+	 * Also, we essentially discard any coprocessor state that the
+	 * signal handler created. */
+
+#if XTENSA_HAVE_COPROCESSORS
+	coprocessor_release_all(ti);
+	err |= __copy_from_user(&ti->xtregs_cp, &frame->xtregs.cp,
+				sizeof (frame->xtregs.cp));
+#endif
+	err |= __copy_from_user(&ti->xtregs_user, &frame->xtregs.user,
+				sizeof (xtregs_user_t));
+	err |= __copy_from_user(&regs->xtregs_opt, &frame->xtregs.opt,
+				sizeof (xtregs_opt_t));
+
+	return err;
+}
+
+
+/*
+ * Do a signal return; undo the signal stack.
+ */
+
+asmlinkage long xtensa_rt_sigreturn(void)
+{
+	struct pt_regs *regs = current_pt_regs();
+	struct rt_sigframe __user *frame;
+	sigset_t set;
+	int ret;
+
+	/* Always make any pending restarted system calls return -EINTR */
+	current->restart_block.fn = do_no_restart_syscall;
+
+	if (regs->depc > 64)
+		panic("rt_sigreturn in double exception!\n");
+
+	frame = (struct rt_sigframe __user *) regs->areg[1];
+
+	if (!access_ok(frame, sizeof(*frame)))
+		goto badframe;
+
+	if (__copy_from_user(&set, &frame->uc.uc_sigmask, sizeof(set)))
+		goto badframe;
+
+	set_current_blocked(&set);
+
+	if (restore_sigcontext(regs, frame))
+		goto badframe;
+
+	ret = regs->areg[2];
+
+	if (restore_altstack(&frame->uc.uc_stack))
+		goto badframe;
+
+	return ret;
+
+badframe:
+	force_sig(SIGSEGV);
+	return 0;
+}
+
+
+
+/*
+ * Set up a signal frame.
+ */
+
+static int
+gen_return_code(unsigned char *codemem)
+{
+	int err = 0;
+
+	/*
+	 * The 12-bit immediate is really split up within the 24-bit MOVI
+	 * instruction.  As long as the above system call numbers fit within
+	 * 8-bits, the following code works fine. See the Xtensa ISA for
+	 * details.
+	 */
+
+#if __NR_rt_sigreturn > 255
+# error Generating the MOVI instruction below breaks!
+#endif
+
+#ifdef __XTENSA_EB__   /* Big Endian version */
+	/* Generate instruction:  MOVI a2, __NR_rt_sigreturn */
+	err |= __put_user(0x22, &codemem[0]);
+	err |= __put_user(0x0a, &codemem[1]);
+	err |= __put_user(__NR_rt_sigreturn, &codemem[2]);
+	/* Generate instruction:  SYSCALL */
+	err |= __put_user(0x00, &codemem[3]);
+	err |= __put_user(0x05, &codemem[4]);
+	err |= __put_user(0x00, &codemem[5]);
+
+#elif defined __XTENSA_EL__   /* Little Endian version */
+	/* Generate instruction:  MOVI a2, __NR_rt_sigreturn */
+	err |= __put_user(0x22, &codemem[0]);
+	err |= __put_user(0xa0, &codemem[1]);
+	err |= __put_user(__NR_rt_sigreturn, &codemem[2]);
+	/* Generate instruction:  SYSCALL */
+	err |= __put_user(0x00, &codemem[3]);
+	err |= __put_user(0x50, &codemem[4]);
+	err |= __put_user(0x00, &codemem[5]);
+#else
+# error Must use compiler for Xtensa processors.
+#endif
+
+	/* Flush generated code out of the data cache */
+
+	if (err == 0) {
+		__invalidate_icache_range((unsigned long)codemem, 6UL);
+		__flush_invalidate_dcache_range((unsigned long)codemem, 6UL);
+	}
+
+	return err;
+}
+
+
+static int setup_frame(struct ksignal *ksig, sigset_t *set,
+		       struct pt_regs *regs)
+{
+	struct rt_sigframe *frame;
+	int err = 0, sig = ksig->sig;
+	unsigned long sp, ra, tp, ps;
+	unsigned long handler = (unsigned long)ksig->ka.sa.sa_handler;
+	unsigned long handler_fdpic_GOT = 0;
+	unsigned int base;
+	bool fdpic = IS_ENABLED(CONFIG_BINFMT_ELF_FDPIC) &&
+		(current->personality & FDPIC_FUNCPTRS);
+
+	if (fdpic) {
+		unsigned long __user *fdpic_func_desc =
+			(unsigned long __user *)handler;
+		if (__get_user(handler, &fdpic_func_desc[0]) ||
+		    __get_user(handler_fdpic_GOT, &fdpic_func_desc[1]))
+			return -EFAULT;
+	}
+
+	sp = regs->areg[1];
+
+	if ((ksig->ka.sa.sa_flags & SA_ONSTACK) != 0 && sas_ss_flags(sp) == 0) {
+		sp = current->sas_ss_sp + current->sas_ss_size;
+	}
+
+	frame = (void *)((sp - sizeof(*frame)) & -16ul);
+
+	if (regs->depc > 64)
+		panic ("Double exception sys_sigreturn\n");
+
+	if (!access_ok(frame, sizeof(*frame))) {
+		return -EFAULT;
+	}
+
+	if (ksig->ka.sa.sa_flags & SA_SIGINFO) {
+		err |= copy_siginfo_to_user(&frame->info, &ksig->info);
+	}
+
+	/* Create the user context.  */
+
+	err |= __put_user(0, &frame->uc.uc_flags);
+	err |= __put_user(0, &frame->uc.uc_link);
+	err |= __save_altstack(&frame->uc.uc_stack, regs->areg[1]);
+	err |= setup_sigcontext(frame, regs);
+	err |= __copy_to_user(&frame->uc.uc_sigmask, set, sizeof(*set));
+
+	if (ksig->ka.sa.sa_flags & SA_RESTORER) {
+		if (fdpic) {
+			unsigned long __user *fdpic_func_desc =
+				(unsigned long __user *)ksig->ka.sa.sa_restorer;
+
+			err |= __get_user(ra, fdpic_func_desc);
+		} else {
+			ra = (unsigned long)ksig->ka.sa.sa_restorer;
+		}
+	} else {
+
+		/* Create sys_rt_sigreturn syscall in stack frame */
+
+		err |= gen_return_code(frame->retcode);
+		ra = (unsigned long) frame->retcode;
+	}
+
+	if (err)
+		return -EFAULT;
+
+	/*
+	 * Create signal handler execution context.
+	 * Return context not modified until this point.
+	 */
+
+	/* Set up registers for signal handler; preserve the threadptr */
+	tp = regs->threadptr;
+	ps = regs->ps;
+	start_thread(regs, handler, (unsigned long)frame);
+
+	/* Set up a stack frame for a call4 if userspace uses windowed ABI */
+	if (ps & PS_WOE_MASK) {
+		base = 4;
+		regs->areg[base] =
+			(((unsigned long) ra) & 0x3fffffff) | 0x40000000;
+		ps = (ps & ~(PS_CALLINC_MASK | PS_OWB_MASK)) |
+			(1 << PS_CALLINC_SHIFT);
+	} else {
+		base = 0;
+		regs->areg[base] = (unsigned long) ra;
+	}
+	regs->areg[base + 2] = (unsigned long) sig;
+	regs->areg[base + 3] = (unsigned long) &frame->info;
+	regs->areg[base + 4] = (unsigned long) &frame->uc;
+	regs->threadptr = tp;
+	regs->ps = ps;
+	if (fdpic)
+		regs->areg[base + 11] = handler_fdpic_GOT;
+
+	pr_debug("SIG rt deliver (%s:%d): signal=%d sp=%p pc=%08lx\n",
+		 current->comm, current->pid, sig, frame, regs->pc);
+
+	return 0;
+}
+
+/*
+ * Note that 'init' is a special process: it doesn't get signals it doesn't
+ * want to handle. Thus you cannot kill init even with a SIGKILL even by
+ * mistake.
+ *
+ * Note that we go through the signals twice: once to check the signals that
+ * the kernel can handle, and then we build all the user-level signal handling
+ * stack-frames in one go after that.
+ */
+static void do_signal(struct pt_regs *regs)
+{
+	struct ksignal ksig;
+
+	task_pt_regs(current)->icountlevel = 0;
+
+	if (get_signal(&ksig)) {
+		int ret;
+
+		/* Are we from a system call? */
+
+		if (regs->syscall != NO_SYSCALL) {
+
+			/* If so, check system call restarting.. */
+
+			switch (regs->areg[2]) {
+				case -ERESTARTNOHAND:
+				case -ERESTART_RESTARTBLOCK:
+					regs->areg[2] = -EINTR;
+					break;
+
+				case -ERESTARTSYS:
+					if (!(ksig.ka.sa.sa_flags & SA_RESTART)) {
+						regs->areg[2] = -EINTR;
+						break;
+					}
+					fallthrough;
+				case -ERESTARTNOINTR:
+					regs->areg[2] = regs->syscall;
+					regs->pc -= 3;
+					break;
+
+				default:
+					/* nothing to do */
+					if (regs->areg[2] != 0)
+					break;
+			}
+		}
+
+		/* Whee!  Actually deliver the signal.  */
+		/* Set up the stack frame */
+		ret = setup_frame(&ksig, sigmask_to_save(), regs);
+		signal_setup_done(ret, &ksig, 0);
+		if (test_thread_flag(TIF_SINGLESTEP))
+			task_pt_regs(current)->icountlevel = 1;
+
+		return;
+	}
+
+	/* Did we come from a system call? */
+	if (regs->syscall != NO_SYSCALL) {
+		/* Restart the system call - no handlers present */
+		switch (regs->areg[2]) {
+		case -ERESTARTNOHAND:
+		case -ERESTARTSYS:
+		case -ERESTARTNOINTR:
+			regs->areg[2] = regs->syscall;
+			regs->pc -= 3;
+			break;
+		case -ERESTART_RESTARTBLOCK:
+			regs->areg[2] = __NR_restart_syscall;
+			regs->pc -= 3;
+			break;
+		}
+	}
+
+	/* If there's no signal to deliver, we just restore the saved mask.  */
+	restore_saved_sigmask();
+
+	if (test_thread_flag(TIF_SINGLESTEP))
+		task_pt_regs(current)->icountlevel = 1;
+	return;
+}
+
+void do_notify_resume(struct pt_regs *regs)
+{
+	if (test_thread_flag(TIF_SIGPENDING) ||
+	    test_thread_flag(TIF_NOTIFY_SIGNAL))
+		do_signal(regs);
+
+	if (test_thread_flag(TIF_NOTIFY_RESUME))
+		resume_user_mode_work(regs);
+}
diff --git a/arch/xtensa/kernel/smp.c b/arch/xtensa/kernel/smp.c
new file mode 100644
index 000000000..e4c25d0c2
--- /dev/null
+++ b/arch/xtensa/kernel/smp.c
@@ -0,0 +1,630 @@
+/*
+ * Xtensa SMP support functions.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2008 - 2013 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Joe Taylor <joe@tensilica.com>
+ * Pete Delaney <piet@tensilica.com
+ */
+
+#include <linux/cpu.h>
+#include <linux/cpumask.h>
+#include <linux/delay.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/irqdomain.h>
+#include <linux/irq.h>
+#include <linux/kdebug.h>
+#include <linux/module.h>
+#include <linux/sched/mm.h>
+#include <linux/sched/hotplug.h>
+#include <linux/sched/task_stack.h>
+#include <linux/reboot.h>
+#include <linux/seq_file.h>
+#include <linux/smp.h>
+#include <linux/thread_info.h>
+
+#include <asm/cacheflush.h>
+#include <asm/coprocessor.h>
+#include <asm/kdebug.h>
+#include <asm/mmu_context.h>
+#include <asm/mxregs.h>
+#include <asm/platform.h>
+#include <asm/tlbflush.h>
+#include <asm/traps.h>
+
+#ifdef CONFIG_SMP
+# if XCHAL_HAVE_S32C1I == 0
+#  error "The S32C1I option is required for SMP."
+# endif
+#endif
+
+static void system_invalidate_dcache_range(unsigned long start,
+		unsigned long size);
+static void system_flush_invalidate_dcache_range(unsigned long start,
+		unsigned long size);
+
+/* IPI (Inter Process Interrupt) */
+
+#define IPI_IRQ	0
+
+static irqreturn_t ipi_interrupt(int irq, void *dev_id);
+
+void ipi_init(void)
+{
+	unsigned irq = irq_create_mapping(NULL, IPI_IRQ);
+	if (request_irq(irq, ipi_interrupt, IRQF_PERCPU, "ipi", NULL))
+		pr_err("Failed to request irq %u (ipi)\n", irq);
+}
+
+static inline unsigned int get_core_count(void)
+{
+	/* Bits 18..21 of SYSCFGID contain the core count minus 1. */
+	unsigned int syscfgid = get_er(SYSCFGID);
+	return ((syscfgid >> 18) & 0xf) + 1;
+}
+
+static inline int get_core_id(void)
+{
+	/* Bits 0...18 of SYSCFGID contain the core id  */
+	unsigned int core_id = get_er(SYSCFGID);
+	return core_id & 0x3fff;
+}
+
+void __init smp_prepare_cpus(unsigned int max_cpus)
+{
+	unsigned i;
+
+	for_each_possible_cpu(i)
+		set_cpu_present(i, true);
+}
+
+void __init smp_init_cpus(void)
+{
+	unsigned i;
+	unsigned int ncpus = get_core_count();
+	unsigned int core_id = get_core_id();
+
+	pr_info("%s: Core Count = %d\n", __func__, ncpus);
+	pr_info("%s: Core Id = %d\n", __func__, core_id);
+
+	if (ncpus > NR_CPUS) {
+		ncpus = NR_CPUS;
+		pr_info("%s: limiting core count by %d\n", __func__, ncpus);
+	}
+
+	for (i = 0; i < ncpus; ++i)
+		set_cpu_possible(i, true);
+}
+
+void __init smp_prepare_boot_cpu(void)
+{
+	unsigned int cpu = smp_processor_id();
+	BUG_ON(cpu != 0);
+	cpu_asid_cache(cpu) = ASID_USER_FIRST;
+}
+
+void __init smp_cpus_done(unsigned int max_cpus)
+{
+}
+
+static int boot_secondary_processors = 1; /* Set with xt-gdb via .xt-gdb */
+static DECLARE_COMPLETION(cpu_running);
+
+void secondary_start_kernel(void)
+{
+	struct mm_struct *mm = &init_mm;
+	unsigned int cpu = smp_processor_id();
+
+	init_mmu();
+
+#ifdef CONFIG_DEBUG_MISC
+	if (boot_secondary_processors == 0) {
+		pr_debug("%s: boot_secondary_processors:%d; Hanging cpu:%d\n",
+			__func__, boot_secondary_processors, cpu);
+		for (;;)
+			__asm__ __volatile__ ("waiti " __stringify(LOCKLEVEL));
+	}
+
+	pr_debug("%s: boot_secondary_processors:%d; Booting cpu:%d\n",
+		__func__, boot_secondary_processors, cpu);
+#endif
+	/* Init EXCSAVE1 */
+
+	secondary_trap_init();
+
+	/* All kernel threads share the same mm context. */
+
+	mmget(mm);
+	mmgrab(mm);
+	current->active_mm = mm;
+	cpumask_set_cpu(cpu, mm_cpumask(mm));
+	enter_lazy_tlb(mm, current);
+
+	trace_hardirqs_off();
+
+	calibrate_delay();
+
+	notify_cpu_starting(cpu);
+
+	secondary_init_irq();
+	local_timer_setup(cpu);
+
+	set_cpu_online(cpu, true);
+
+	local_irq_enable();
+
+	complete(&cpu_running);
+
+	cpu_startup_entry(CPUHP_AP_ONLINE_IDLE);
+}
+
+static void mx_cpu_start(void *p)
+{
+	unsigned cpu = (unsigned)p;
+	unsigned long run_stall_mask = get_er(MPSCORE);
+
+	set_er(run_stall_mask & ~(1u << cpu), MPSCORE);
+	pr_debug("%s: cpu: %d, run_stall_mask: %lx ---> %lx\n",
+			__func__, cpu, run_stall_mask, get_er(MPSCORE));
+}
+
+static void mx_cpu_stop(void *p)
+{
+	unsigned cpu = (unsigned)p;
+	unsigned long run_stall_mask = get_er(MPSCORE);
+
+	set_er(run_stall_mask | (1u << cpu), MPSCORE);
+	pr_debug("%s: cpu: %d, run_stall_mask: %lx ---> %lx\n",
+			__func__, cpu, run_stall_mask, get_er(MPSCORE));
+}
+
+#ifdef CONFIG_HOTPLUG_CPU
+unsigned long cpu_start_id __cacheline_aligned;
+#endif
+unsigned long cpu_start_ccount;
+
+static int boot_secondary(unsigned int cpu, struct task_struct *ts)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(1000);
+	unsigned long ccount;
+	int i;
+
+#ifdef CONFIG_HOTPLUG_CPU
+	WRITE_ONCE(cpu_start_id, cpu);
+	/* Pairs with the third memw in the cpu_restart */
+	mb();
+	system_flush_invalidate_dcache_range((unsigned long)&cpu_start_id,
+					     sizeof(cpu_start_id));
+#endif
+	smp_call_function_single(0, mx_cpu_start, (void *)cpu, 1);
+
+	for (i = 0; i < 2; ++i) {
+		do
+			ccount = get_ccount();
+		while (!ccount);
+
+		WRITE_ONCE(cpu_start_ccount, ccount);
+
+		do {
+			/*
+			 * Pairs with the first two memws in the
+			 * .Lboot_secondary.
+			 */
+			mb();
+			ccount = READ_ONCE(cpu_start_ccount);
+		} while (ccount && time_before(jiffies, timeout));
+
+		if (ccount) {
+			smp_call_function_single(0, mx_cpu_stop,
+						 (void *)cpu, 1);
+			WRITE_ONCE(cpu_start_ccount, 0);
+			return -EIO;
+		}
+	}
+	return 0;
+}
+
+int __cpu_up(unsigned int cpu, struct task_struct *idle)
+{
+	int ret = 0;
+
+	if (cpu_asid_cache(cpu) == 0)
+		cpu_asid_cache(cpu) = ASID_USER_FIRST;
+
+	start_info.stack = (unsigned long)task_pt_regs(idle);
+	wmb();
+
+	pr_debug("%s: Calling wakeup_secondary(cpu:%d, idle:%p, sp: %08lx)\n",
+			__func__, cpu, idle, start_info.stack);
+
+	init_completion(&cpu_running);
+	ret = boot_secondary(cpu, idle);
+	if (ret == 0) {
+		wait_for_completion_timeout(&cpu_running,
+				msecs_to_jiffies(1000));
+		if (!cpu_online(cpu))
+			ret = -EIO;
+	}
+
+	if (ret)
+		pr_err("CPU %u failed to boot\n", cpu);
+
+	return ret;
+}
+
+#ifdef CONFIG_HOTPLUG_CPU
+
+/*
+ * __cpu_disable runs on the processor to be shutdown.
+ */
+int __cpu_disable(void)
+{
+	unsigned int cpu = smp_processor_id();
+
+	/*
+	 * Take this CPU offline.  Once we clear this, we can't return,
+	 * and we must not schedule until we're ready to give up the cpu.
+	 */
+	set_cpu_online(cpu, false);
+
+#if XTENSA_HAVE_COPROCESSORS
+	/*
+	 * Flush coprocessor contexts that are active on the current CPU.
+	 */
+	local_coprocessors_flush_release_all();
+#endif
+	/*
+	 * OK - migrate IRQs away from this CPU
+	 */
+	migrate_irqs();
+
+	/*
+	 * Flush user cache and TLB mappings, and then remove this CPU
+	 * from the vm mask set of all processes.
+	 */
+	local_flush_cache_all();
+	local_flush_tlb_all();
+	invalidate_page_directory();
+
+	clear_tasks_mm_cpumask(cpu);
+
+	return 0;
+}
+
+static void platform_cpu_kill(unsigned int cpu)
+{
+	smp_call_function_single(0, mx_cpu_stop, (void *)cpu, true);
+}
+
+/*
+ * called on the thread which is asking for a CPU to be shutdown -
+ * waits until shutdown has completed, or it is timed out.
+ */
+void __cpu_die(unsigned int cpu)
+{
+	unsigned long timeout = jiffies + msecs_to_jiffies(1000);
+	while (time_before(jiffies, timeout)) {
+		system_invalidate_dcache_range((unsigned long)&cpu_start_id,
+					       sizeof(cpu_start_id));
+		/* Pairs with the second memw in the cpu_restart */
+		mb();
+		if (READ_ONCE(cpu_start_id) == -cpu) {
+			platform_cpu_kill(cpu);
+			return;
+		}
+	}
+	pr_err("CPU%u: unable to kill\n", cpu);
+}
+
+void __noreturn arch_cpu_idle_dead(void)
+{
+	cpu_die();
+}
+/*
+ * Called from the idle thread for the CPU which has been shutdown.
+ *
+ * Note that we disable IRQs here, but do not re-enable them
+ * before returning to the caller. This is also the behaviour
+ * of the other hotplug-cpu capable cores, so presumably coming
+ * out of idle fixes this.
+ */
+void __ref cpu_die(void)
+{
+	idle_task_exit();
+	local_irq_disable();
+	__asm__ __volatile__(
+			"	movi	a2, cpu_restart\n"
+			"	jx	a2\n");
+
+	BUG();
+}
+
+#endif /* CONFIG_HOTPLUG_CPU */
+
+enum ipi_msg_type {
+	IPI_RESCHEDULE = 0,
+	IPI_CALL_FUNC,
+	IPI_CPU_STOP,
+	IPI_MAX
+};
+
+static const struct {
+	const char *short_text;
+	const char *long_text;
+} ipi_text[] = {
+	{ .short_text = "RES", .long_text = "Rescheduling interrupts" },
+	{ .short_text = "CAL", .long_text = "Function call interrupts" },
+	{ .short_text = "DIE", .long_text = "CPU shutdown interrupts" },
+};
+
+struct ipi_data {
+	unsigned long ipi_count[IPI_MAX];
+};
+
+static DEFINE_PER_CPU(struct ipi_data, ipi_data);
+
+static void send_ipi_message(const struct cpumask *callmask,
+		enum ipi_msg_type msg_id)
+{
+	int index;
+	unsigned long mask = 0;
+
+	for_each_cpu(index, callmask)
+		mask |= 1 << index;
+
+	set_er(mask, MIPISET(msg_id));
+}
+
+void arch_send_call_function_ipi_mask(const struct cpumask *mask)
+{
+	send_ipi_message(mask, IPI_CALL_FUNC);
+}
+
+void arch_send_call_function_single_ipi(int cpu)
+{
+	send_ipi_message(cpumask_of(cpu), IPI_CALL_FUNC);
+}
+
+void arch_smp_send_reschedule(int cpu)
+{
+	send_ipi_message(cpumask_of(cpu), IPI_RESCHEDULE);
+}
+
+void smp_send_stop(void)
+{
+	struct cpumask targets;
+
+	cpumask_copy(&targets, cpu_online_mask);
+	cpumask_clear_cpu(smp_processor_id(), &targets);
+	send_ipi_message(&targets, IPI_CPU_STOP);
+}
+
+static void ipi_cpu_stop(unsigned int cpu)
+{
+	set_cpu_online(cpu, false);
+	machine_halt();
+}
+
+irqreturn_t ipi_interrupt(int irq, void *dev_id)
+{
+	unsigned int cpu = smp_processor_id();
+	struct ipi_data *ipi = &per_cpu(ipi_data, cpu);
+
+	for (;;) {
+		unsigned int msg;
+
+		msg = get_er(MIPICAUSE(cpu));
+		set_er(msg, MIPICAUSE(cpu));
+
+		if (!msg)
+			break;
+
+		if (msg & (1 << IPI_CALL_FUNC)) {
+			++ipi->ipi_count[IPI_CALL_FUNC];
+			generic_smp_call_function_interrupt();
+		}
+
+		if (msg & (1 << IPI_RESCHEDULE)) {
+			++ipi->ipi_count[IPI_RESCHEDULE];
+			scheduler_ipi();
+		}
+
+		if (msg & (1 << IPI_CPU_STOP)) {
+			++ipi->ipi_count[IPI_CPU_STOP];
+			ipi_cpu_stop(cpu);
+		}
+	}
+
+	return IRQ_HANDLED;
+}
+
+void show_ipi_list(struct seq_file *p, int prec)
+{
+	unsigned int cpu;
+	unsigned i;
+
+	for (i = 0; i < IPI_MAX; ++i) {
+		seq_printf(p, "%*s:", prec, ipi_text[i].short_text);
+		for_each_online_cpu(cpu)
+			seq_printf(p, " %10lu",
+					per_cpu(ipi_data, cpu).ipi_count[i]);
+		seq_printf(p, "   %s\n", ipi_text[i].long_text);
+	}
+}
+
+/* TLB flush functions */
+
+struct flush_data {
+	struct vm_area_struct *vma;
+	unsigned long addr1;
+	unsigned long addr2;
+};
+
+static void ipi_flush_tlb_all(void *arg)
+{
+	local_flush_tlb_all();
+}
+
+void flush_tlb_all(void)
+{
+	on_each_cpu(ipi_flush_tlb_all, NULL, 1);
+}
+
+static void ipi_flush_tlb_mm(void *arg)
+{
+	local_flush_tlb_mm(arg);
+}
+
+void flush_tlb_mm(struct mm_struct *mm)
+{
+	on_each_cpu(ipi_flush_tlb_mm, mm, 1);
+}
+
+static void ipi_flush_tlb_page(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_tlb_page(fd->vma, fd->addr1);
+}
+
+void flush_tlb_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	struct flush_data fd = {
+		.vma = vma,
+		.addr1 = addr,
+	};
+	on_each_cpu(ipi_flush_tlb_page, &fd, 1);
+}
+
+static void ipi_flush_tlb_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_tlb_range(fd->vma, fd->addr1, fd->addr2);
+}
+
+void flush_tlb_range(struct vm_area_struct *vma,
+		     unsigned long start, unsigned long end)
+{
+	struct flush_data fd = {
+		.vma = vma,
+		.addr1 = start,
+		.addr2 = end,
+	};
+	on_each_cpu(ipi_flush_tlb_range, &fd, 1);
+}
+
+static void ipi_flush_tlb_kernel_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_tlb_kernel_range(fd->addr1, fd->addr2);
+}
+
+void flush_tlb_kernel_range(unsigned long start, unsigned long end)
+{
+	struct flush_data fd = {
+		.addr1 = start,
+		.addr2 = end,
+	};
+	on_each_cpu(ipi_flush_tlb_kernel_range, &fd, 1);
+}
+
+/* Cache flush functions */
+
+static void ipi_flush_cache_all(void *arg)
+{
+	local_flush_cache_all();
+}
+
+void flush_cache_all(void)
+{
+	on_each_cpu(ipi_flush_cache_all, NULL, 1);
+}
+
+static void ipi_flush_cache_page(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_cache_page(fd->vma, fd->addr1, fd->addr2);
+}
+
+void flush_cache_page(struct vm_area_struct *vma,
+		     unsigned long address, unsigned long pfn)
+{
+	struct flush_data fd = {
+		.vma = vma,
+		.addr1 = address,
+		.addr2 = pfn,
+	};
+	on_each_cpu(ipi_flush_cache_page, &fd, 1);
+}
+
+static void ipi_flush_cache_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_cache_range(fd->vma, fd->addr1, fd->addr2);
+}
+
+void flush_cache_range(struct vm_area_struct *vma,
+		     unsigned long start, unsigned long end)
+{
+	struct flush_data fd = {
+		.vma = vma,
+		.addr1 = start,
+		.addr2 = end,
+	};
+	on_each_cpu(ipi_flush_cache_range, &fd, 1);
+}
+
+static void ipi_flush_icache_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	local_flush_icache_range(fd->addr1, fd->addr2);
+}
+
+void flush_icache_range(unsigned long start, unsigned long end)
+{
+	struct flush_data fd = {
+		.addr1 = start,
+		.addr2 = end,
+	};
+	on_each_cpu(ipi_flush_icache_range, &fd, 1);
+}
+EXPORT_SYMBOL(flush_icache_range);
+
+/* ------------------------------------------------------------------------- */
+
+static void ipi_invalidate_dcache_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	__invalidate_dcache_range(fd->addr1, fd->addr2);
+}
+
+static void system_invalidate_dcache_range(unsigned long start,
+		unsigned long size)
+{
+	struct flush_data fd = {
+		.addr1 = start,
+		.addr2 = size,
+	};
+	on_each_cpu(ipi_invalidate_dcache_range, &fd, 1);
+}
+
+static void ipi_flush_invalidate_dcache_range(void *arg)
+{
+	struct flush_data *fd = arg;
+	__flush_invalidate_dcache_range(fd->addr1, fd->addr2);
+}
+
+static void system_flush_invalidate_dcache_range(unsigned long start,
+		unsigned long size)
+{
+	struct flush_data fd = {
+		.addr1 = start,
+		.addr2 = size,
+	};
+	on_each_cpu(ipi_flush_invalidate_dcache_range, &fd, 1);
+}
diff --git a/arch/xtensa/kernel/stacktrace.c b/arch/xtensa/kernel/stacktrace.c
new file mode 100644
index 000000000..ed324fdf2
--- /dev/null
+++ b/arch/xtensa/kernel/stacktrace.c
@@ -0,0 +1,273 @@
+/*
+ * Kernel and userspace stack tracing.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+#include <linux/export.h>
+#include <linux/sched.h>
+#include <linux/stacktrace.h>
+
+#include <asm/ftrace.h>
+#include <asm/sections.h>
+#include <asm/stacktrace.h>
+#include <asm/traps.h>
+#include <linux/uaccess.h>
+
+#if IS_ENABLED(CONFIG_PERF_EVENTS)
+
+/* Address of common_exception_return, used to check the
+ * transition from kernel to user space.
+ */
+extern int common_exception_return;
+
+void xtensa_backtrace_user(struct pt_regs *regs, unsigned int depth,
+			   int (*ufn)(struct stackframe *frame, void *data),
+			   void *data)
+{
+	unsigned long windowstart = regs->windowstart;
+	unsigned long windowbase = regs->windowbase;
+	unsigned long a0 = regs->areg[0];
+	unsigned long a1 = regs->areg[1];
+	unsigned long pc = regs->pc;
+	struct stackframe frame;
+	int index;
+
+	if (!depth--)
+		return;
+
+	frame.pc = pc;
+	frame.sp = a1;
+
+	if (pc == 0 || pc >= TASK_SIZE || ufn(&frame, data))
+		return;
+
+	if (IS_ENABLED(CONFIG_USER_ABI_CALL0_ONLY) ||
+	    (IS_ENABLED(CONFIG_USER_ABI_CALL0_PROBE) &&
+	     !(regs->ps & PS_WOE_MASK)))
+		return;
+
+	/* Two steps:
+	 *
+	 * 1. Look through the register window for the
+	 * previous PCs in the call trace.
+	 *
+	 * 2. Look on the stack.
+	 */
+
+	/* Step 1.  */
+	/* Rotate WINDOWSTART to move the bit corresponding to
+	 * the current window to the bit #0.
+	 */
+	windowstart = (windowstart << WSBITS | windowstart) >> windowbase;
+
+	/* Look for bits that are set, they correspond to
+	 * valid windows.
+	 */
+	for (index = WSBITS - 1; (index > 0) && depth; depth--, index--)
+		if (windowstart & (1 << index)) {
+			/* Get the PC from a0 and a1. */
+			pc = MAKE_PC_FROM_RA(a0, pc);
+			/* Read a0 and a1 from the
+			 * corresponding position in AREGs.
+			 */
+			a0 = regs->areg[index * 4];
+			a1 = regs->areg[index * 4 + 1];
+
+			frame.pc = pc;
+			frame.sp = a1;
+
+			if (pc == 0 || pc >= TASK_SIZE || ufn(&frame, data))
+				return;
+		}
+
+	/* Step 2. */
+	/* We are done with the register window, we need to
+	 * look through the stack.
+	 */
+	if (!depth)
+		return;
+
+	/* Start from the a1 register. */
+	/* a1 = regs->areg[1]; */
+	while (a0 != 0 && depth--) {
+		pc = MAKE_PC_FROM_RA(a0, pc);
+
+		/* Check if the region is OK to access. */
+		if (!access_ok(&SPILL_SLOT(a1, 0), 8))
+			return;
+		/* Copy a1, a0 from user space stack frame. */
+		if (__get_user(a0, &SPILL_SLOT(a1, 0)) ||
+		    __get_user(a1, &SPILL_SLOT(a1, 1)))
+			return;
+
+		frame.pc = pc;
+		frame.sp = a1;
+
+		if (pc == 0 || pc >= TASK_SIZE || ufn(&frame, data))
+			return;
+	}
+}
+EXPORT_SYMBOL(xtensa_backtrace_user);
+
+void xtensa_backtrace_kernel(struct pt_regs *regs, unsigned int depth,
+			     int (*kfn)(struct stackframe *frame, void *data),
+			     int (*ufn)(struct stackframe *frame, void *data),
+			     void *data)
+{
+	unsigned long pc = regs->depc > VALID_DOUBLE_EXCEPTION_ADDRESS ?
+		regs->depc : regs->pc;
+	unsigned long sp_start, sp_end;
+	unsigned long a0 = regs->areg[0];
+	unsigned long a1 = regs->areg[1];
+
+	sp_start = a1 & ~(THREAD_SIZE - 1);
+	sp_end = sp_start + THREAD_SIZE;
+
+	/* Spill the register window to the stack first. */
+	spill_registers();
+
+	/* Read the stack frames one by one and create the PC
+	 * from the a0 and a1 registers saved there.
+	 */
+	while (a1 > sp_start && a1 < sp_end && depth--) {
+		struct stackframe frame;
+
+		frame.pc = pc;
+		frame.sp = a1;
+
+		if (kernel_text_address(pc) && kfn(&frame, data))
+			return;
+
+		if (pc == (unsigned long)&common_exception_return) {
+			regs = (struct pt_regs *)a1;
+			if (user_mode(regs)) {
+				if (ufn == NULL)
+					return;
+				xtensa_backtrace_user(regs, depth, ufn, data);
+				return;
+			}
+			a0 = regs->areg[0];
+			a1 = regs->areg[1];
+			continue;
+		}
+
+		sp_start = a1;
+
+		pc = MAKE_PC_FROM_RA(a0, pc);
+		a0 = SPILL_SLOT(a1, 0);
+		a1 = SPILL_SLOT(a1, 1);
+	}
+}
+EXPORT_SYMBOL(xtensa_backtrace_kernel);
+
+#endif
+
+void walk_stackframe(unsigned long *sp,
+		int (*fn)(struct stackframe *frame, void *data),
+		void *data)
+{
+	unsigned long a0, a1;
+	unsigned long sp_end;
+
+	a1 = (unsigned long)sp;
+	sp_end = ALIGN(a1, THREAD_SIZE);
+
+	spill_registers();
+
+	while (a1 < sp_end) {
+		struct stackframe frame;
+
+		sp = (unsigned long *)a1;
+
+		a0 = SPILL_SLOT(a1, 0);
+		a1 = SPILL_SLOT(a1, 1);
+
+		if (a1 <= (unsigned long)sp)
+			break;
+
+		frame.pc = MAKE_PC_FROM_RA(a0, _text);
+		frame.sp = a1;
+
+		if (fn(&frame, data))
+			return;
+	}
+}
+
+#ifdef CONFIG_STACKTRACE
+
+struct stack_trace_data {
+	struct stack_trace *trace;
+	unsigned skip;
+};
+
+static int stack_trace_cb(struct stackframe *frame, void *data)
+{
+	struct stack_trace_data *trace_data = data;
+	struct stack_trace *trace = trace_data->trace;
+
+	if (trace_data->skip) {
+		--trace_data->skip;
+		return 0;
+	}
+	if (!kernel_text_address(frame->pc))
+		return 0;
+
+	trace->entries[trace->nr_entries++] = frame->pc;
+	return trace->nr_entries >= trace->max_entries;
+}
+
+void save_stack_trace_tsk(struct task_struct *task, struct stack_trace *trace)
+{
+	struct stack_trace_data trace_data = {
+		.trace = trace,
+		.skip = trace->skip,
+	};
+	walk_stackframe(stack_pointer(task), stack_trace_cb, &trace_data);
+}
+EXPORT_SYMBOL_GPL(save_stack_trace_tsk);
+
+void save_stack_trace(struct stack_trace *trace)
+{
+	save_stack_trace_tsk(current, trace);
+}
+EXPORT_SYMBOL_GPL(save_stack_trace);
+
+#endif
+
+struct return_addr_data {
+	unsigned long addr;
+	unsigned skip;
+};
+
+static int return_address_cb(struct stackframe *frame, void *data)
+{
+	struct return_addr_data *r = data;
+
+	if (r->skip) {
+		--r->skip;
+		return 0;
+	}
+	if (!kernel_text_address(frame->pc))
+		return 0;
+	r->addr = frame->pc;
+	return 1;
+}
+
+/*
+ * level == 0 is for the return address from the caller of this function,
+ * not from this function itself.
+ */
+unsigned long return_address(unsigned level)
+{
+	struct return_addr_data r = {
+		.skip = level,
+	};
+	walk_stackframe(stack_pointer(NULL), return_address_cb, &r);
+	return r.addr;
+}
+EXPORT_SYMBOL(return_address);
diff --git a/arch/xtensa/kernel/syscall.c b/arch/xtensa/kernel/syscall.c
new file mode 100644
index 000000000..dc54f854c
--- /dev/null
+++ b/arch/xtensa/kernel/syscall.c
@@ -0,0 +1,100 @@
+/*
+ * arch/xtensa/kernel/syscall.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ * Copyright (C) 2000 Silicon Graphics, Inc.
+ * Copyright (C) 1995 - 2000 by Ralf Baechle
+ *
+ * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Chris Zankel <chris@zankel.net>
+ * Kevin Chea
+ *
+ */
+#include <linux/uaccess.h>
+#include <asm/syscall.h>
+#include <linux/linkage.h>
+#include <linux/stringify.h>
+#include <linux/errno.h>
+#include <linux/syscalls.h>
+#include <linux/file.h>
+#include <linux/fs.h>
+#include <linux/mman.h>
+#include <linux/sched/mm.h>
+#include <linux/shm.h>
+
+syscall_t sys_call_table[] /* FIXME __cacheline_aligned */= {
+#define __SYSCALL(nr, entry)	(syscall_t)entry,
+#include <asm/syscall_table.h>
+};
+
+#define COLOUR_ALIGN(addr, pgoff) \
+	((((addr) + SHMLBA - 1) & ~(SHMLBA - 1)) + \
+	 (((pgoff) << PAGE_SHIFT) & (SHMLBA - 1)))
+
+asmlinkage long xtensa_shmat(int shmid, char __user *shmaddr, int shmflg)
+{
+	unsigned long ret;
+	long err;
+
+	err = do_shmat(shmid, shmaddr, shmflg, &ret, SHMLBA);
+	if (err)
+		return err;
+	return (long)ret;
+}
+
+asmlinkage long xtensa_fadvise64_64(int fd, int advice,
+		unsigned long long offset, unsigned long long len)
+{
+	return ksys_fadvise64_64(fd, offset, len, advice);
+}
+
+#ifdef CONFIG_MMU
+unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr,
+		unsigned long len, unsigned long pgoff, unsigned long flags,
+		vm_flags_t vm_flags)
+{
+	struct vm_area_struct *vmm;
+	struct vma_iterator vmi;
+
+	if (flags & MAP_FIXED) {
+		/* We do not accept a shared mapping if it would violate
+		 * cache aliasing constraints.
+		 */
+		if ((flags & MAP_SHARED) &&
+				((addr - (pgoff << PAGE_SHIFT)) & (SHMLBA - 1)))
+			return -EINVAL;
+		return addr;
+	}
+
+	if (len > TASK_SIZE)
+		return -ENOMEM;
+	if (!addr)
+		addr = TASK_UNMAPPED_BASE;
+
+	if (flags & MAP_SHARED)
+		addr = COLOUR_ALIGN(addr, pgoff);
+	else
+		addr = PAGE_ALIGN(addr);
+
+	vma_iter_init(&vmi, current->mm, addr);
+	for_each_vma(vmi, vmm) {
+		/* At this point:  (addr < vmm->vm_end). */
+		if (addr + len <= vm_start_gap(vmm))
+			break;
+
+		addr = vmm->vm_end;
+		if (flags & MAP_SHARED)
+			addr = COLOUR_ALIGN(addr, pgoff);
+	}
+
+	if (TASK_SIZE - len < addr)
+		return -ENOMEM;
+
+	return addr;
+}
+#endif
diff --git a/arch/xtensa/kernel/syscalls/Makefile b/arch/xtensa/kernel/syscalls/Makefile
new file mode 100644
index 000000000..b265e4bc1
--- /dev/null
+++ b/arch/xtensa/kernel/syscalls/Makefile
@@ -0,0 +1,32 @@
+# SPDX-License-Identifier: GPL-2.0
+kapi := arch/$(SRCARCH)/include/generated/asm
+uapi := arch/$(SRCARCH)/include/generated/uapi/asm
+
+$(shell mkdir -p $(uapi) $(kapi))
+
+syscall := $(src)/syscall.tbl
+syshdr := $(srctree)/scripts/syscallhdr.sh
+systbl := $(srctree)/scripts/syscalltbl.sh
+
+quiet_cmd_syshdr = SYSHDR  $@
+      cmd_syshdr = $(CONFIG_SHELL) $(syshdr) --emit-nr $< $@
+
+quiet_cmd_systbl = SYSTBL  $@
+      cmd_systbl = $(CONFIG_SHELL) $(systbl) $< $@
+
+$(uapi)/unistd_32.h: $(syscall) $(syshdr) FORCE
+	$(call if_changed,syshdr)
+
+$(kapi)/syscall_table.h: $(syscall) $(systbl) FORCE
+	$(call if_changed,systbl)
+
+uapisyshdr-y		+= unistd_32.h
+kapisyshdr-y		+= syscall_table.h
+
+uapisyshdr-y	:= $(addprefix $(uapi)/, $(uapisyshdr-y))
+kapisyshdr-y	:= $(addprefix $(kapi)/, $(kapisyshdr-y))
+targets		+= $(addprefix ../../../../, $(uapisyshdr-y) $(kapisyshdr-y))
+
+PHONY += all
+all: $(uapisyshdr-y) $(kapisyshdr-y)
+	@:
diff --git a/arch/xtensa/kernel/syscalls/syscall.tbl b/arch/xtensa/kernel/syscalls/syscall.tbl
new file mode 100644
index 000000000..d354bb231
--- /dev/null
+++ b/arch/xtensa/kernel/syscalls/syscall.tbl
@@ -0,0 +1,445 @@
+# SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note
+#
+# system call numbers and entry vectors for xtensa
+#
+# The format is:
+# <number> <abi> <name> <entry point>
+#
+# The <abi> is always "common" for this file
+#
+0	common	spill				sys_ni_syscall
+1	common	xtensa				sys_ni_syscall
+2	common	available4			sys_ni_syscall
+3	common	available5			sys_ni_syscall
+4	common	available6			sys_ni_syscall
+5	common	available7			sys_ni_syscall
+6	common	available8			sys_ni_syscall
+7	common	available9			sys_ni_syscall
+# File Operations
+8	common	open				sys_open
+9	common	close				sys_close
+10	common	dup				sys_dup
+11	common	dup2				sys_dup2
+12	common	read				sys_read
+13	common	write				sys_write
+14	common	select				sys_select
+15	common	lseek				sys_lseek
+16	common	poll				sys_poll
+17	common	_llseek				sys_llseek
+18	common	epoll_wait			sys_epoll_wait
+19	common	epoll_ctl			sys_epoll_ctl
+20	common	epoll_create			sys_epoll_create
+21	common	creat				sys_creat
+22	common	truncate			sys_truncate
+23	common	ftruncate			sys_ftruncate
+24	common	readv				sys_readv
+25	common	writev				sys_writev
+26	common	fsync				sys_fsync
+27	common	fdatasync			sys_fdatasync
+28	common	truncate64			sys_truncate64
+29	common	ftruncate64			sys_ftruncate64
+30	common	pread64				sys_pread64
+31	common	pwrite64			sys_pwrite64
+32	common	link				sys_link
+33	common	rename				sys_rename
+34	common	symlink				sys_symlink
+35	common	readlink			sys_readlink
+36	common	mknod				sys_mknod
+37	common	pipe				sys_pipe
+38	common	unlink				sys_unlink
+39	common	rmdir				sys_rmdir
+40	common	mkdir				sys_mkdir
+41	common	chdir				sys_chdir
+42	common	fchdir				sys_fchdir
+43	common	getcwd				sys_getcwd
+44	common	chmod				sys_chmod
+45	common	chown				sys_chown
+46	common	stat				sys_newstat
+47	common	stat64				sys_stat64
+48	common	lchown				sys_lchown
+49	common	lstat				sys_newlstat
+50	common	lstat64				sys_lstat64
+51	common	available51			sys_ni_syscall
+52	common	fchmod				sys_fchmod
+53	common	fchown				sys_fchown
+54	common	fstat				sys_newfstat
+55	common	fstat64				sys_fstat64
+56	common	flock				sys_flock
+57	common	access				sys_access
+58	common	umask				sys_umask
+59	common	getdents			sys_getdents
+60	common	getdents64			sys_getdents64
+61	common	fcntl64				sys_fcntl64
+62	common	fallocate			sys_fallocate
+63	common	fadvise64_64			xtensa_fadvise64_64
+64	common	utime				sys_utime32
+65	common	utimes				sys_utimes_time32
+66	common	ioctl				sys_ioctl
+67	common	fcntl				sys_fcntl
+68	common	setxattr			sys_setxattr
+69	common	getxattr			sys_getxattr
+70	common	listxattr			sys_listxattr
+71	common	removexattr			sys_removexattr
+72	common	lsetxattr			sys_lsetxattr
+73	common	lgetxattr			sys_lgetxattr
+74	common	llistxattr			sys_llistxattr
+75	common	lremovexattr			sys_lremovexattr
+76	common	fsetxattr			sys_fsetxattr
+77	common	fgetxattr			sys_fgetxattr
+78	common	flistxattr			sys_flistxattr
+79	common	fremovexattr			sys_fremovexattr
+# File Map / Shared Memory Operations
+80	common	mmap2				sys_mmap_pgoff
+81	common	munmap				sys_munmap
+82	common	mprotect			sys_mprotect
+83	common	brk				sys_brk
+84	common	mlock				sys_mlock
+85	common	munlock				sys_munlock
+86	common	mlockall			sys_mlockall
+87	common	munlockall			sys_munlockall
+88	common	mremap				sys_mremap
+89	common	msync				sys_msync
+90	common	mincore				sys_mincore
+91	common	madvise				sys_madvise
+92	common	shmget				sys_shmget
+93	common	shmat				xtensa_shmat
+94	common	shmctl				sys_old_shmctl
+95	common	shmdt				sys_shmdt
+# Socket Operations
+96	common	socket				sys_socket
+97	common	setsockopt			sys_setsockopt
+98	common	getsockopt			sys_getsockopt
+99	common	shutdown			sys_shutdown
+100	common	bind				sys_bind
+101	common	connect				sys_connect
+102	common	listen				sys_listen
+103	common	accept				sys_accept
+104	common	getsockname			sys_getsockname
+105	common	getpeername			sys_getpeername
+106	common	sendmsg				sys_sendmsg
+107	common	recvmsg				sys_recvmsg
+108	common	send				sys_send
+109	common	recv				sys_recv
+110	common	sendto				sys_sendto
+111	common	recvfrom			sys_recvfrom
+112	common	socketpair			sys_socketpair
+113	common	sendfile			sys_sendfile
+114	common	sendfile64			sys_sendfile64
+115	common	sendmmsg			sys_sendmmsg
+# Process Operations
+116	common	clone				sys_clone
+117	common	execve				sys_execve
+118	common	exit				sys_exit
+119	common	exit_group			sys_exit_group
+120	common	getpid				sys_getpid
+121	common	wait4				sys_wait4
+122	common	waitid				sys_waitid
+123	common	kill				sys_kill
+124	common	tkill				sys_tkill
+125	common	tgkill				sys_tgkill
+126	common	set_tid_address			sys_set_tid_address
+127	common	gettid				sys_gettid
+128	common	setsid				sys_setsid
+129	common	getsid				sys_getsid
+130	common	prctl				sys_prctl
+131	common	personality			sys_personality
+132	common	getpriority			sys_getpriority
+133	common	setpriority			sys_setpriority
+134	common	setitimer			sys_setitimer
+135	common	getitimer			sys_getitimer
+136	common	setuid				sys_setuid
+137	common	getuid				sys_getuid
+138	common	setgid				sys_setgid
+139	common	getgid				sys_getgid
+140	common	geteuid				sys_geteuid
+141	common	getegid				sys_getegid
+142	common	setreuid			sys_setreuid
+143	common	setregid			sys_setregid
+144	common	setresuid			sys_setresuid
+145	common	getresuid			sys_getresuid
+146	common	setresgid			sys_setresgid
+147	common	getresgid			sys_getresgid
+148	common	setpgid				sys_setpgid
+149	common	getpgid				sys_getpgid
+150	common	getppid				sys_getppid
+151	common	getpgrp				sys_getpgrp
+# 152 was set_thread_area
+152	common	reserved152			sys_ni_syscall
+# 153 was get_thread_area
+153	common	reserved153			sys_ni_syscall
+154	common	times				sys_times
+155	common	acct				sys_acct
+156	common	sched_setaffinity		sys_sched_setaffinity
+157	common	sched_getaffinity		sys_sched_getaffinity
+158	common	capget				sys_capget
+159	common	capset				sys_capset
+160	common	ptrace				sys_ptrace
+161	common	semtimedop			sys_semtimedop_time32
+162	common	semget				sys_semget
+163	common	semop				sys_semop
+164	common	semctl				sys_old_semctl
+165	common	available165			sys_ni_syscall
+166	common	msgget				sys_msgget
+167	common	msgsnd				sys_msgsnd
+168	common	msgrcv				sys_msgrcv
+169	common	msgctl				sys_old_msgctl
+170	common	available170			sys_ni_syscall
+# File System
+171	common	umount2				sys_umount
+172	common	mount				sys_mount
+173	common	swapon				sys_swapon
+174	common	chroot				sys_chroot
+175	common	pivot_root			sys_pivot_root
+176	common	umount				sys_oldumount
+177	common	swapoff				sys_swapoff
+178	common	sync				sys_sync
+179	common	syncfs				sys_syncfs
+180	common	setfsuid			sys_setfsuid
+181	common	setfsgid			sys_setfsgid
+182	common	sysfs				sys_sysfs
+183	common	ustat				sys_ustat
+184	common	statfs				sys_statfs
+185	common	fstatfs				sys_fstatfs
+186	common	statfs64			sys_statfs64
+187	common	fstatfs64			sys_fstatfs64
+# System
+188	common	setrlimit			sys_setrlimit
+189	common	getrlimit			sys_getrlimit
+190	common	getrusage			sys_getrusage
+191	common	futex				sys_futex_time32
+192	common	gettimeofday			sys_gettimeofday
+193	common	settimeofday			sys_settimeofday
+194	common	adjtimex			sys_adjtimex_time32
+195	common	nanosleep			sys_nanosleep_time32
+196	common	getgroups			sys_getgroups
+197	common	setgroups			sys_setgroups
+198	common	sethostname			sys_sethostname
+199	common	setdomainname			sys_setdomainname
+200	common	syslog				sys_syslog
+201	common	vhangup				sys_vhangup
+202	common	uselib				sys_uselib
+203	common	reboot				sys_reboot
+204	common	quotactl			sys_quotactl
+# 205 was old nfsservctl
+205	common	nfsservctl			sys_ni_syscall
+206	common	_sysctl				sys_ni_syscall
+207	common	bdflush				sys_ni_syscall
+208	common	uname				sys_newuname
+209	common	sysinfo				sys_sysinfo
+210	common	init_module			sys_init_module
+211	common	delete_module			sys_delete_module
+212	common	sched_setparam			sys_sched_setparam
+213	common	sched_getparam			sys_sched_getparam
+214	common	sched_setscheduler		sys_sched_setscheduler
+215	common	sched_getscheduler		sys_sched_getscheduler
+216	common	sched_get_priority_max		sys_sched_get_priority_max
+217	common	sched_get_priority_min		sys_sched_get_priority_min
+218	common	sched_rr_get_interval		sys_sched_rr_get_interval_time32
+219	common	sched_yield			sys_sched_yield
+222	common	available222			sys_ni_syscall
+# Signal Handling
+223	common	restart_syscall			sys_restart_syscall
+224	common	sigaltstack			sys_sigaltstack
+225	common	rt_sigreturn			xtensa_rt_sigreturn
+226	common	rt_sigaction			sys_rt_sigaction
+227	common	rt_sigprocmask			sys_rt_sigprocmask
+228	common	rt_sigpending			sys_rt_sigpending
+229	common	rt_sigtimedwait			sys_rt_sigtimedwait_time32
+230	common	rt_sigqueueinfo			sys_rt_sigqueueinfo
+231	common	rt_sigsuspend			sys_rt_sigsuspend
+# Message
+232	common	mq_open				sys_mq_open
+233	common	mq_unlink			sys_mq_unlink
+234	common	mq_timedsend			sys_mq_timedsend_time32
+235	common	mq_timedreceive			sys_mq_timedreceive_time32
+236	common	mq_notify			sys_mq_notify
+237	common	mq_getsetattr			sys_mq_getsetattr
+238	common	available238			sys_ni_syscall
+239	common	io_setup			sys_io_setup
+# IO
+240	common	io_destroy			sys_io_destroy
+241	common	io_submit			sys_io_submit
+242	common	io_getevents			sys_io_getevents_time32
+243	common	io_cancel			sys_io_cancel
+244	common	clock_settime			sys_clock_settime32
+245	common	clock_gettime			sys_clock_gettime32
+246	common	clock_getres			sys_clock_getres_time32
+247	common	clock_nanosleep			sys_clock_nanosleep_time32
+# Timer
+248	common	timer_create			sys_timer_create
+249	common	timer_delete			sys_timer_delete
+250	common	timer_settime			sys_timer_settime32
+251	common	timer_gettime			sys_timer_gettime32
+252	common	timer_getoverrun		sys_timer_getoverrun
+# System
+253	common	reserved253			sys_ni_syscall
+254	common	lookup_dcookie			sys_ni_syscall
+255	common	available255			sys_ni_syscall
+256	common	add_key				sys_add_key
+257	common	request_key			sys_request_key
+258	common	keyctl				sys_keyctl
+259	common	available259			sys_ni_syscall
+260	common	readahead			sys_readahead
+261	common	remap_file_pages		sys_remap_file_pages
+262	common	migrate_pages			sys_migrate_pages
+263	common	mbind				sys_mbind
+264	common	get_mempolicy			sys_get_mempolicy
+265	common	set_mempolicy			sys_set_mempolicy
+266	common	unshare				sys_unshare
+267	common	move_pages			sys_move_pages
+268	common	splice				sys_splice
+269	common	tee				sys_tee
+270	common	vmsplice			sys_vmsplice
+271	common	available271			sys_ni_syscall
+272	common	pselect6			sys_pselect6_time32
+273	common	ppoll				sys_ppoll_time32
+274	common	epoll_pwait			sys_epoll_pwait
+275	common	epoll_create1			sys_epoll_create1
+276	common	inotify_init			sys_inotify_init
+277	common	inotify_add_watch		sys_inotify_add_watch
+278	common	inotify_rm_watch		sys_inotify_rm_watch
+279	common	inotify_init1			sys_inotify_init1
+280	common	getcpu				sys_getcpu
+281	common	kexec_load			sys_ni_syscall
+282	common	ioprio_set			sys_ioprio_set
+283	common	ioprio_get			sys_ioprio_get
+284	common	set_robust_list			sys_set_robust_list
+285	common	get_robust_list			sys_get_robust_list
+286	common	available286			sys_ni_syscall
+287	common	available287			sys_ni_syscall
+# Relative File Operations
+288	common	openat				sys_openat
+289	common	mkdirat				sys_mkdirat
+290	common	mknodat				sys_mknodat
+291	common	unlinkat			sys_unlinkat
+292	common	renameat			sys_renameat
+293	common	linkat				sys_linkat
+294	common	symlinkat			sys_symlinkat
+295	common	readlinkat			sys_readlinkat
+296	common	utimensat			sys_utimensat_time32
+297	common	fchownat			sys_fchownat
+298	common	futimesat			sys_futimesat_time32
+299	common	fstatat64			sys_fstatat64
+300	common	fchmodat			sys_fchmodat
+301	common	faccessat			sys_faccessat
+302	common	available302			sys_ni_syscall
+303	common	available303			sys_ni_syscall
+304	common	signalfd			sys_signalfd
+# 305 was timerfd
+306	common	eventfd				sys_eventfd
+307	common	recvmmsg			sys_recvmmsg_time32
+308	common	setns				sys_setns
+309	common	signalfd4			sys_signalfd4
+310	common	dup3				sys_dup3
+311	common	pipe2				sys_pipe2
+312	common	timerfd_create			sys_timerfd_create
+313	common	timerfd_settime			sys_timerfd_settime32
+314	common	timerfd_gettime			sys_timerfd_gettime32
+315	common	available315			sys_ni_syscall
+316	common	eventfd2			sys_eventfd2
+317	common	preadv				sys_preadv
+318	common	pwritev				sys_pwritev
+319	common	available319			sys_ni_syscall
+320	common	fanotify_init			sys_fanotify_init
+321	common	fanotify_mark			sys_fanotify_mark
+322	common	process_vm_readv		sys_process_vm_readv
+323	common	process_vm_writev		sys_process_vm_writev
+324	common	name_to_handle_at		sys_name_to_handle_at
+325	common	open_by_handle_at		sys_open_by_handle_at
+326	common	sync_file_range2		sys_sync_file_range2
+327	common	perf_event_open			sys_perf_event_open
+328	common	rt_tgsigqueueinfo		sys_rt_tgsigqueueinfo
+329	common	clock_adjtime			sys_clock_adjtime32
+330	common	prlimit64			sys_prlimit64
+331	common	kcmp				sys_kcmp
+332	common	finit_module			sys_finit_module
+333	common	accept4				sys_accept4
+334	common	sched_setattr			sys_sched_setattr
+335	common	sched_getattr			sys_sched_getattr
+336	common	renameat2			sys_renameat2
+337	common	seccomp				sys_seccomp
+338	common	getrandom			sys_getrandom
+339	common	memfd_create			sys_memfd_create
+340	common	bpf				sys_bpf
+341	common	execveat			sys_execveat
+342	common	userfaultfd			sys_userfaultfd
+343	common	membarrier			sys_membarrier
+344	common	mlock2				sys_mlock2
+345	common	copy_file_range			sys_copy_file_range
+346	common	preadv2				sys_preadv2
+347	common	pwritev2			sys_pwritev2
+348	common	pkey_mprotect			sys_pkey_mprotect
+349	common	pkey_alloc			sys_pkey_alloc
+350	common	pkey_free			sys_pkey_free
+351	common	statx				sys_statx
+352	common	rseq				sys_rseq
+# 353 through 402 are unassigned to sync up with generic numbers
+403	common	clock_gettime64			sys_clock_gettime
+404	common	clock_settime64			sys_clock_settime
+405	common	clock_adjtime64			sys_clock_adjtime
+406	common	clock_getres_time64		sys_clock_getres
+407	common	clock_nanosleep_time64		sys_clock_nanosleep
+408	common	timer_gettime64			sys_timer_gettime
+409	common	timer_settime64			sys_timer_settime
+410	common	timerfd_gettime64		sys_timerfd_gettime
+411	common	timerfd_settime64		sys_timerfd_settime
+412	common	utimensat_time64		sys_utimensat
+413	common	pselect6_time64			sys_pselect6
+414	common	ppoll_time64			sys_ppoll
+416	common	io_pgetevents_time64		sys_io_pgetevents
+417	common	recvmmsg_time64			sys_recvmmsg
+418	common	mq_timedsend_time64		sys_mq_timedsend
+419	common	mq_timedreceive_time64		sys_mq_timedreceive
+420	common	semtimedop_time64		sys_semtimedop
+421	common	rt_sigtimedwait_time64		sys_rt_sigtimedwait
+422	common	futex_time64			sys_futex
+423	common	sched_rr_get_interval_time64	sys_sched_rr_get_interval
+424	common	pidfd_send_signal		sys_pidfd_send_signal
+425	common	io_uring_setup			sys_io_uring_setup
+426	common	io_uring_enter			sys_io_uring_enter
+427	common	io_uring_register		sys_io_uring_register
+428	common	open_tree			sys_open_tree
+429	common	move_mount			sys_move_mount
+430	common	fsopen				sys_fsopen
+431	common	fsconfig			sys_fsconfig
+432	common	fsmount				sys_fsmount
+433	common	fspick				sys_fspick
+434	common	pidfd_open			sys_pidfd_open
+435	common	clone3				sys_clone3
+436	common	close_range			sys_close_range
+437	common	openat2				sys_openat2
+438	common	pidfd_getfd			sys_pidfd_getfd
+439	common	faccessat2			sys_faccessat2
+440	common	process_madvise			sys_process_madvise
+441	common	epoll_pwait2			sys_epoll_pwait2
+442	common	mount_setattr			sys_mount_setattr
+443	common	quotactl_fd			sys_quotactl_fd
+444	common	landlock_create_ruleset		sys_landlock_create_ruleset
+445	common	landlock_add_rule		sys_landlock_add_rule
+446	common	landlock_restrict_self		sys_landlock_restrict_self
+# 447 reserved for memfd_secret
+448	common	process_mrelease		sys_process_mrelease
+449	common  futex_waitv                     sys_futex_waitv
+450	common	set_mempolicy_home_node		sys_set_mempolicy_home_node
+451	common	cachestat			sys_cachestat
+452	common	fchmodat2			sys_fchmodat2
+453	common	map_shadow_stack		sys_map_shadow_stack
+454	common	futex_wake			sys_futex_wake
+455	common	futex_wait			sys_futex_wait
+456	common	futex_requeue			sys_futex_requeue
+457	common	statmount			sys_statmount
+458	common	listmount			sys_listmount
+459	common	lsm_get_self_attr		sys_lsm_get_self_attr
+460	common	lsm_set_self_attr		sys_lsm_set_self_attr
+461	common	lsm_list_modules		sys_lsm_list_modules
+462	common	mseal 				sys_mseal
+463	common	setxattrat			sys_setxattrat
+464	common	getxattrat			sys_getxattrat
+465	common	listxattrat			sys_listxattrat
+466	common	removexattrat			sys_removexattrat
+467	common	open_tree_attr			sys_open_tree_attr
+468	common	file_getattr			sys_file_getattr
+469	common	file_setattr			sys_file_setattr
+470	common	listns				sys_listns
+471	common	rseq_slice_yield		sys_rseq_slice_yield
+472	common	fchroot				sys_fchroot
diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c
new file mode 100644
index 000000000..09fda6a3f
--- /dev/null
+++ b/arch/xtensa/kernel/time.c
@@ -0,0 +1,209 @@
+/*
+ * arch/xtensa/kernel/time.c
+ *
+ * Timer and clock support.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2005 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ */
+
+#include <linux/clk.h>
+#include <linux/of_clk.h>
+#include <linux/errno.h>
+#include <linux/sched.h>
+#include <linux/time.h>
+#include <linux/clocksource.h>
+#include <linux/clockchips.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/irq.h>
+#include <linux/delay.h>
+#include <linux/irqdomain.h>
+#include <linux/sched_clock.h>
+
+#include <asm/timex.h>
+#include <asm/platform.h>
+
+unsigned long ccount_freq;		/* ccount Hz */
+EXPORT_SYMBOL(ccount_freq);
+
+static u64 ccount_read(struct clocksource *cs)
+{
+	return (u64)get_ccount();
+}
+
+static u64 notrace ccount_sched_clock_read(void)
+{
+	return get_ccount();
+}
+
+static struct clocksource ccount_clocksource = {
+	.name = "ccount",
+	.rating = 200,
+	.read = ccount_read,
+	.mask = CLOCKSOURCE_MASK(32),
+	.flags = CLOCK_SOURCE_IS_CONTINUOUS,
+};
+
+struct ccount_timer {
+	struct clock_event_device evt;
+	int irq_enabled;
+	char name[24];
+};
+
+static int ccount_timer_set_next_event(unsigned long delta,
+		struct clock_event_device *dev)
+{
+	unsigned long flags, next;
+	int ret = 0;
+
+	local_irq_save(flags);
+	next = get_ccount() + delta;
+	set_linux_timer(next);
+	if (next - get_ccount() > delta)
+		ret = -ETIME;
+	local_irq_restore(flags);
+
+	return ret;
+}
+
+/*
+ * There is no way to disable the timer interrupt at the device level,
+ * only at the intenable register itself. Since enable_irq/disable_irq
+ * calls are nested, we need to make sure that these calls are
+ * balanced.
+ */
+static int ccount_timer_shutdown(struct clock_event_device *evt)
+{
+	struct ccount_timer *timer =
+		container_of(evt, struct ccount_timer, evt);
+
+	if (timer->irq_enabled) {
+		disable_irq_nosync(evt->irq);
+		timer->irq_enabled = 0;
+	}
+	return 0;
+}
+
+static int ccount_timer_set_oneshot(struct clock_event_device *evt)
+{
+	struct ccount_timer *timer =
+		container_of(evt, struct ccount_timer, evt);
+
+	if (!timer->irq_enabled) {
+		enable_irq(evt->irq);
+		timer->irq_enabled = 1;
+	}
+	return 0;
+}
+
+static DEFINE_PER_CPU(struct ccount_timer, ccount_timer) = {
+	.evt = {
+		.features = CLOCK_EVT_FEAT_ONESHOT,
+		.rating = 300,
+		.set_next_event = ccount_timer_set_next_event,
+		.set_state_shutdown = ccount_timer_shutdown,
+		.set_state_oneshot = ccount_timer_set_oneshot,
+		.tick_resume = ccount_timer_set_oneshot,
+	},
+};
+
+static irqreturn_t timer_interrupt(int irq, void *dev_id)
+{
+	struct clock_event_device *evt = &this_cpu_ptr(&ccount_timer)->evt;
+
+	set_linux_timer(get_linux_timer());
+	evt->event_handler(evt);
+	return IRQ_HANDLED;
+}
+
+void local_timer_setup(unsigned cpu)
+{
+	struct ccount_timer *timer = &per_cpu(ccount_timer, cpu);
+	struct clock_event_device *clockevent = &timer->evt;
+
+	timer->irq_enabled = 1;
+	snprintf(timer->name, sizeof(timer->name), "ccount_clockevent_%u", cpu);
+	clockevent->name = timer->name;
+	clockevent->cpumask = cpumask_of(cpu);
+	clockevent->irq = irq_create_mapping(NULL, LINUX_TIMER_INT);
+	if (WARN(!clockevent->irq, "error: can't map timer irq"))
+		return;
+	clockevents_config_and_register(clockevent, ccount_freq,
+					0xf, 0xffffffff);
+}
+
+#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
+#ifdef CONFIG_OF
+static void __init calibrate_ccount(void)
+{
+	struct device_node *cpu;
+	struct clk *clk;
+
+	cpu = of_find_compatible_node(NULL, NULL, "cdns,xtensa-cpu");
+	if (cpu) {
+		clk = of_clk_get(cpu, 0);
+		of_node_put(cpu);
+		if (!IS_ERR(clk)) {
+			ccount_freq = clk_get_rate(clk);
+			return;
+		} else {
+			pr_warn("%s: CPU input clock not found\n",
+				__func__);
+		}
+	} else {
+		pr_warn("%s: CPU node not found in the device tree\n",
+			__func__);
+	}
+
+	platform_calibrate_ccount();
+}
+#else
+static inline void calibrate_ccount(void)
+{
+	platform_calibrate_ccount();
+}
+#endif
+#endif
+
+void __init time_init(void)
+{
+	int irq;
+
+	of_clk_init(NULL);
+#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
+	pr_info("Calibrating CPU frequency ");
+	calibrate_ccount();
+	pr_cont("%d.%02d MHz\n",
+		(int)ccount_freq / 1000000,
+		(int)(ccount_freq / 10000) % 100);
+#else
+	ccount_freq = CONFIG_XTENSA_CPU_CLOCK*1000000UL;
+#endif
+	WARN(!ccount_freq,
+	     "%s: CPU clock frequency is not set up correctly\n",
+	     __func__);
+	clocksource_register_hz(&ccount_clocksource, ccount_freq);
+	local_timer_setup(0);
+	irq = this_cpu_ptr(&ccount_timer)->evt.irq;
+	if (request_irq(irq, timer_interrupt, IRQF_TIMER, "timer", NULL))
+		pr_err("Failed to request irq %d (timer)\n", irq);
+	sched_clock_register(ccount_sched_clock_read, 32, ccount_freq);
+	timer_probe();
+}
+
+#ifndef CONFIG_GENERIC_CALIBRATE_DELAY
+void calibrate_delay(void)
+{
+	loops_per_jiffy = ccount_freq / HZ;
+	pr_info("Calibrating delay loop (skipped)... %lu.%02lu BogoMIPS preset\n",
+		loops_per_jiffy / (1000000 / HZ),
+		(loops_per_jiffy / (10000 / HZ)) % 100);
+}
+#endif
diff --git a/arch/xtensa/kernel/traps.c b/arch/xtensa/kernel/traps.c
new file mode 100644
index 000000000..44c07c4e0
--- /dev/null
+++ b/arch/xtensa/kernel/traps.c
@@ -0,0 +1,651 @@
+/*
+ * arch/xtensa/kernel/traps.c
+ *
+ * Exception handling.
+ *
+ * Derived from code with the following copyrights:
+ * Copyright (C) 1994 - 1999 by Ralf Baechle
+ * Modified for R3000 by Paul M. Antoine, 1995, 1996
+ * Complete output from die() by Ulf Carlsson, 1998
+ * Copyright (C) 1999 Silicon Graphics, Inc.
+ *
+ * Essentially rewritten for the Xtensa architecture port.
+ *
+ * Copyright (C) 2001 - 2013 Tensilica Inc.
+ *
+ * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
+ * Chris Zankel	<chris@zankel.net>
+ * Marc Gauthier<marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Kevin Chea
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/cpu.h>
+#include <linux/kernel.h>
+#include <linux/sched/signal.h>
+#include <linux/sched/debug.h>
+#include <linux/sched/task_stack.h>
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/stringify.h>
+#include <linux/kallsyms.h>
+#include <linux/delay.h>
+#include <linux/hardirq.h>
+#include <linux/ratelimit.h>
+#include <linux/pgtable.h>
+
+#include <asm/stacktrace.h>
+#include <asm/ptrace.h>
+#include <asm/timex.h>
+#include <linux/uaccess.h>
+#include <asm/processor.h>
+#include <asm/traps.h>
+#include <asm/hw_breakpoint.h>
+
+/*
+ * Machine specific interrupt handlers
+ */
+
+static void do_illegal_instruction(struct pt_regs *regs);
+static void do_div0(struct pt_regs *regs);
+static void do_interrupt(struct pt_regs *regs);
+#if XTENSA_FAKE_NMI
+static void do_nmi(struct pt_regs *regs);
+#endif
+#ifdef CONFIG_XTENSA_LOAD_STORE
+static void do_load_store(struct pt_regs *regs);
+#endif
+static void do_unaligned_user(struct pt_regs *regs);
+static void do_multihit(struct pt_regs *regs);
+#if XTENSA_HAVE_COPROCESSORS
+static void do_coprocessor(struct pt_regs *regs);
+#endif
+static void do_debug(struct pt_regs *regs);
+
+/*
+ * The vector table must be preceded by a save area (which
+ * implies it must be in RAM, unless one places RAM immediately
+ * before a ROM and puts the vector at the start of the ROM (!))
+ */
+
+#define KRNL		0x01
+#define USER		0x02
+
+#define COPROCESSOR(x)							\
+{ EXCCAUSE_COPROCESSOR ## x ## _DISABLED, USER|KRNL, fast_coprocessor },\
+{ EXCCAUSE_COPROCESSOR ## x ## _DISABLED, 0, do_coprocessor }
+
+typedef struct {
+	int cause;
+	int fast;
+	void* handler;
+} dispatch_init_table_t;
+
+static dispatch_init_table_t __initdata dispatch_init_table[] = {
+
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+{ EXCCAUSE_ILLEGAL_INSTRUCTION,	USER,	   fast_illegal_instruction_user },
+#endif
+{ EXCCAUSE_ILLEGAL_INSTRUCTION,	0,	   do_illegal_instruction},
+{ EXCCAUSE_SYSTEM_CALL,		USER,	   fast_syscall_user },
+{ EXCCAUSE_SYSTEM_CALL,		0,	   system_call },
+/* EXCCAUSE_INSTRUCTION_FETCH unhandled */
+#ifdef CONFIG_XTENSA_LOAD_STORE
+{ EXCCAUSE_LOAD_STORE_ERROR,	USER|KRNL, fast_load_store },
+{ EXCCAUSE_LOAD_STORE_ERROR,	0,	   do_load_store },
+#endif
+{ EXCCAUSE_LEVEL1_INTERRUPT,	0,	   do_interrupt },
+#ifdef SUPPORT_WINDOWED
+{ EXCCAUSE_ALLOCA,		USER|KRNL, fast_alloca },
+#endif
+{ EXCCAUSE_INTEGER_DIVIDE_BY_ZERO, 0,	   do_div0 },
+/* EXCCAUSE_PRIVILEGED unhandled */
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || XCHAL_UNALIGNED_STORE_EXCEPTION || \
+		IS_ENABLED(CONFIG_XTENSA_LOAD_STORE)
+#ifdef CONFIG_XTENSA_UNALIGNED_USER
+{ EXCCAUSE_UNALIGNED,		USER,	   fast_unaligned },
+#endif
+{ EXCCAUSE_UNALIGNED,		KRNL,	   fast_unaligned },
+#endif
+{ EXCCAUSE_UNALIGNED,		0,	   do_unaligned_user },
+#ifdef CONFIG_MMU
+{ EXCCAUSE_ITLB_MISS,			0,	   do_page_fault },
+{ EXCCAUSE_ITLB_MISS,			USER|KRNL, fast_second_level_miss},
+{ EXCCAUSE_DTLB_MISS,			USER|KRNL, fast_second_level_miss},
+{ EXCCAUSE_DTLB_MISS,			0,	   do_page_fault },
+{ EXCCAUSE_STORE_CACHE_ATTRIBUTE,	USER|KRNL, fast_store_prohibited },
+#endif /* CONFIG_MMU */
+#ifdef CONFIG_PFAULT
+{ EXCCAUSE_ITLB_MULTIHIT,		0,	   do_multihit },
+{ EXCCAUSE_ITLB_PRIVILEGE,		0,	   do_page_fault },
+{ EXCCAUSE_FETCH_CACHE_ATTRIBUTE,	0,	   do_page_fault },
+{ EXCCAUSE_DTLB_MULTIHIT,		0,	   do_multihit },
+{ EXCCAUSE_DTLB_PRIVILEGE,		0,	   do_page_fault },
+{ EXCCAUSE_STORE_CACHE_ATTRIBUTE,	0,	   do_page_fault },
+{ EXCCAUSE_LOAD_CACHE_ATTRIBUTE,	0,	   do_page_fault },
+#endif
+/* XCCHAL_EXCCAUSE_FLOATING_POINT unhandled */
+#if XTENSA_HAVE_COPROCESSOR(0)
+COPROCESSOR(0),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(1)
+COPROCESSOR(1),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(2)
+COPROCESSOR(2),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(3)
+COPROCESSOR(3),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(4)
+COPROCESSOR(4),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(5)
+COPROCESSOR(5),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(6)
+COPROCESSOR(6),
+#endif
+#if XTENSA_HAVE_COPROCESSOR(7)
+COPROCESSOR(7),
+#endif
+#if XTENSA_FAKE_NMI
+{ EXCCAUSE_MAPPED_NMI,			0,		do_nmi },
+#endif
+{ EXCCAUSE_MAPPED_DEBUG,		0,		do_debug },
+{ -1, -1, 0 }
+
+};
+
+/* The exception table <exc_table> serves two functions:
+ * 1. it contains three dispatch tables (fast_user, fast_kernel, default-c)
+ * 2. it is a temporary memory buffer for the exception handlers.
+ */
+
+DEFINE_PER_CPU(struct exc_table, exc_table);
+DEFINE_PER_CPU(struct debug_table, debug_table);
+
+void die(const char*, struct pt_regs*, long);
+
+static inline void
+__die_if_kernel(const char *str, struct pt_regs *regs, long err)
+{
+	if (!user_mode(regs))
+		die(str, regs, err);
+}
+
+#ifdef CONFIG_PRINT_USER_CODE_ON_UNHANDLED_EXCEPTION
+static inline void dump_user_code(struct pt_regs *regs)
+{
+	char buf[32];
+
+	if (copy_from_user(buf, (void __user *)(regs->pc & -16), sizeof(buf)) == 0) {
+		print_hex_dump(KERN_INFO, " ", DUMP_PREFIX_NONE,
+			       32, 1, buf, sizeof(buf), false);
+
+	}
+}
+#else
+static inline void dump_user_code(struct pt_regs *regs)
+{
+}
+#endif
+
+/*
+ * Unhandled Exceptions. Kill user task or panic if in kernel space.
+ */
+
+void do_unhandled(struct pt_regs *regs)
+{
+	__die_if_kernel("Caught unhandled exception - should not happen",
+			regs, SIGKILL);
+
+	/* If in user mode, send SIGILL signal to current process */
+	pr_info_ratelimited("Caught unhandled exception in '%s' "
+			    "(pid = %d, pc = %#010lx) - should not happen\n"
+			    "\tEXCCAUSE is %ld\n",
+			    current->comm, task_pid_nr(current), regs->pc,
+			    regs->exccause);
+	dump_user_code(regs);
+	force_sig(SIGILL);
+}
+
+/*
+ * Multi-hit exception. This if fatal!
+ */
+
+static void do_multihit(struct pt_regs *regs)
+{
+	die("Caught multihit exception", regs, SIGKILL);
+}
+
+/*
+ * IRQ handler.
+ */
+
+#if XTENSA_FAKE_NMI
+
+#define IS_POW2(v) (((v) & ((v) - 1)) == 0)
+
+#if !(PROFILING_INTLEVEL == XCHAL_EXCM_LEVEL && \
+      IS_POW2(XTENSA_INTLEVEL_MASK(PROFILING_INTLEVEL)))
+#warning "Fake NMI is requested for PMM, but there are other IRQs at or above its level."
+#warning "Fake NMI will be used, but there will be a bugcheck if one of those IRQs fire."
+
+static inline void check_valid_nmi(void)
+{
+	unsigned intread = xtensa_get_sr(interrupt);
+	unsigned intenable = xtensa_get_sr(intenable);
+
+	BUG_ON(intread & intenable &
+	       ~(XTENSA_INTLEVEL_ANDBELOW_MASK(PROFILING_INTLEVEL) ^
+		 XTENSA_INTLEVEL_MASK(PROFILING_INTLEVEL) ^
+		 BIT(XCHAL_PROFILING_INTERRUPT)));
+}
+
+#else
+
+static inline void check_valid_nmi(void)
+{
+}
+
+#endif
+
+irqreturn_t xtensa_pmu_irq_handler(int irq, void *dev_id);
+
+DEFINE_PER_CPU(unsigned long, nmi_count);
+
+static void do_nmi(struct pt_regs *regs)
+{
+	struct pt_regs *old_regs = set_irq_regs(regs);
+
+	nmi_enter();
+	++*this_cpu_ptr(&nmi_count);
+	check_valid_nmi();
+	xtensa_pmu_irq_handler(0, NULL);
+	nmi_exit();
+	set_irq_regs(old_regs);
+}
+#endif
+
+static void do_interrupt(struct pt_regs *regs)
+{
+	static const unsigned int_level_mask[] = {
+		0,
+		XCHAL_INTLEVEL1_MASK,
+		XCHAL_INTLEVEL2_MASK,
+		XCHAL_INTLEVEL3_MASK,
+		XCHAL_INTLEVEL4_MASK,
+		XCHAL_INTLEVEL5_MASK,
+		XCHAL_INTLEVEL6_MASK,
+		XCHAL_INTLEVEL7_MASK,
+	};
+	struct pt_regs *old_regs = set_irq_regs(regs);
+	unsigned unhandled = ~0u;
+
+	irq_enter();
+
+	for (;;) {
+		unsigned intread = xtensa_get_sr(interrupt);
+		unsigned intenable = xtensa_get_sr(intenable);
+		unsigned int_at_level = intread & intenable;
+		unsigned level;
+
+		for (level = LOCKLEVEL; level > 0; --level) {
+			if (int_at_level & int_level_mask[level]) {
+				int_at_level &= int_level_mask[level];
+				if (int_at_level & unhandled)
+					int_at_level &= unhandled;
+				else
+					unhandled |= int_level_mask[level];
+				break;
+			}
+		}
+
+		if (level == 0)
+			break;
+
+		/* clear lowest pending irq in the unhandled mask */
+		unhandled ^= (int_at_level & -int_at_level);
+		do_IRQ(__ffs(int_at_level), regs);
+	}
+
+	irq_exit();
+	set_irq_regs(old_regs);
+}
+
+static bool check_div0(struct pt_regs *regs)
+{
+	static const u8 pattern[] = {'D', 'I', 'V', '0'};
+	const u8 *p;
+	u8 buf[5];
+
+	if (user_mode(regs)) {
+		if (copy_from_user(buf, (void __user *)regs->pc + 2, 5))
+			return false;
+		p = buf;
+	} else {
+		p = (const u8 *)regs->pc + 2;
+	}
+
+	return memcmp(p, pattern, sizeof(pattern)) == 0 ||
+		memcmp(p + 1, pattern, sizeof(pattern)) == 0;
+}
+
+/*
+ * Illegal instruction. Fatal if in kernel space.
+ */
+
+static void do_illegal_instruction(struct pt_regs *regs)
+{
+#ifdef CONFIG_USER_ABI_CALL0_PROBE
+	/*
+	 * When call0 application encounters an illegal instruction fast
+	 * exception handler will attempt to set PS.WOE and retry failing
+	 * instruction.
+	 * If we get here we know that that instruction is also illegal
+	 * with PS.WOE set, so it's not related to the windowed option
+	 * hence PS.WOE may be cleared.
+	 */
+	if (regs->pc == current_thread_info()->ps_woe_fix_addr)
+		regs->ps &= ~PS_WOE_MASK;
+#endif
+	if (check_div0(regs)) {
+		do_div0(regs);
+		return;
+	}
+
+	__die_if_kernel("Illegal instruction in kernel", regs, SIGKILL);
+
+	/* If in user mode, send SIGILL signal to current process. */
+
+	pr_info_ratelimited("Illegal Instruction in '%s' (pid = %d, pc = %#010lx)\n",
+			    current->comm, task_pid_nr(current), regs->pc);
+	force_sig(SIGILL);
+}
+
+static void do_div0(struct pt_regs *regs)
+{
+	__die_if_kernel("Unhandled division by 0 in kernel", regs, SIGKILL);
+	force_sig_fault(SIGFPE, FPE_INTDIV, (void __user *)regs->pc);
+}
+
+#ifdef CONFIG_XTENSA_LOAD_STORE
+static void do_load_store(struct pt_regs *regs)
+{
+	__die_if_kernel("Unhandled load/store exception in kernel",
+			regs, SIGKILL);
+
+	pr_info_ratelimited("Load/store error to %08lx in '%s' (pid = %d, pc = %#010lx)\n",
+			    regs->excvaddr, current->comm,
+			    task_pid_nr(current), regs->pc);
+	force_sig_fault(SIGBUS, BUS_ADRERR, (void *)regs->excvaddr);
+}
+#endif
+
+/*
+ * Handle unaligned memory accesses from user space. Kill task.
+ *
+ * If CONFIG_UNALIGNED_USER is not set, we don't allow unaligned memory
+ * accesses causes from user space.
+ */
+
+static void do_unaligned_user(struct pt_regs *regs)
+{
+	__die_if_kernel("Unhandled unaligned exception in kernel",
+			regs, SIGKILL);
+
+	pr_info_ratelimited("Unaligned memory access to %08lx in '%s' "
+			    "(pid = %d, pc = %#010lx)\n",
+			    regs->excvaddr, current->comm,
+			    task_pid_nr(current), regs->pc);
+	force_sig_fault(SIGBUS, BUS_ADRALN, (void *) regs->excvaddr);
+}
+
+#if XTENSA_HAVE_COPROCESSORS
+static void do_coprocessor(struct pt_regs *regs)
+{
+	coprocessor_flush_release_all(current_thread_info());
+}
+#endif
+
+/* Handle debug events.
+ * When CONFIG_HAVE_HW_BREAKPOINT is on this handler is called with
+ * preemption disabled to avoid rescheduling and keep mapping of hardware
+ * breakpoint structures to debug registers intact, so that
+ * DEBUGCAUSE.DBNUM could be used in case of data breakpoint hit.
+ */
+static void do_debug(struct pt_regs *regs)
+{
+#ifdef CONFIG_HAVE_HW_BREAKPOINT
+	int ret = check_hw_breakpoint(regs);
+
+	preempt_enable();
+	if (ret == 0)
+		return;
+#endif
+	__die_if_kernel("Breakpoint in kernel", regs, SIGKILL);
+
+	/* If in user mode, send SIGTRAP signal to current process */
+
+	force_sig(SIGTRAP);
+}
+
+
+#define set_handler(type, cause, handler)				\
+	do {								\
+		unsigned int cpu;					\
+									\
+		for_each_possible_cpu(cpu)				\
+			per_cpu(exc_table, cpu).type[cause] = (handler);\
+	} while (0)
+
+/* Set exception C handler - for temporary use when probing exceptions */
+
+xtensa_exception_handler *
+__init trap_set_handler(int cause, xtensa_exception_handler *handler)
+{
+	void *previous = per_cpu(exc_table, 0).default_handler[cause];
+
+	set_handler(default_handler, cause, handler);
+	return previous;
+}
+
+
+static void trap_init_excsave(void)
+{
+	xtensa_set_sr(this_cpu_ptr(&exc_table), excsave1);
+}
+
+static void trap_init_debug(void)
+{
+	unsigned long debugsave = (unsigned long)this_cpu_ptr(&debug_table);
+
+	this_cpu_ptr(&debug_table)->debug_exception = debug_exception;
+	__asm__ __volatile__("wsr %0, excsave" __stringify(XCHAL_DEBUGLEVEL)
+			     :: "a"(debugsave));
+}
+
+/*
+ * Initialize dispatch tables.
+ *
+ * The exception vectors are stored compressed the __init section in the
+ * dispatch_init_table. This function initializes the following three tables
+ * from that compressed table:
+ * - fast user		first dispatch table for user exceptions
+ * - fast kernel	first dispatch table for kernel exceptions
+ * - default C-handler	C-handler called by the default fast handler.
+ *
+ * See vectors.S for more details.
+ */
+
+void __init trap_init(void)
+{
+	int i;
+
+	/* Setup default vectors. */
+
+	for (i = 0; i < EXCCAUSE_N; i++) {
+		set_handler(fast_user_handler, i, user_exception);
+		set_handler(fast_kernel_handler, i, kernel_exception);
+		set_handler(default_handler, i, do_unhandled);
+	}
+
+	/* Setup specific handlers. */
+
+	for(i = 0; dispatch_init_table[i].cause >= 0; i++) {
+		int fast = dispatch_init_table[i].fast;
+		int cause = dispatch_init_table[i].cause;
+		void *handler = dispatch_init_table[i].handler;
+
+		if (fast == 0)
+			set_handler(default_handler, cause, handler);
+		if ((fast & USER) != 0)
+			set_handler(fast_user_handler, cause, handler);
+		if ((fast & KRNL) != 0)
+			set_handler(fast_kernel_handler, cause, handler);
+	}
+
+	/* Initialize EXCSAVE_1 to hold the address of the exception table. */
+	trap_init_excsave();
+	trap_init_debug();
+}
+
+#ifdef CONFIG_SMP
+void secondary_trap_init(void)
+{
+	trap_init_excsave();
+	trap_init_debug();
+}
+#endif
+
+/*
+ * This function dumps the current valid window frame and other base registers.
+ */
+
+void show_regs(struct pt_regs * regs)
+{
+	int i;
+
+	show_regs_print_info(KERN_DEFAULT);
+
+	for (i = 0; i < 16; i++) {
+		if ((i % 8) == 0)
+			pr_info("a%02d:", i);
+		pr_cont(" %08lx", regs->areg[i]);
+	}
+	pr_cont("\n");
+	pr_info("pc: %08lx, ps: %08lx, depc: %08lx, excvaddr: %08lx\n",
+		regs->pc, regs->ps, regs->depc, regs->excvaddr);
+	pr_info("lbeg: %08lx, lend: %08lx lcount: %08lx, sar: %08lx\n",
+		regs->lbeg, regs->lend, regs->lcount, regs->sar);
+	if (user_mode(regs))
+		pr_cont("wb: %08lx, ws: %08lx, wmask: %08lx, syscall: %ld\n",
+			regs->windowbase, regs->windowstart, regs->wmask,
+			regs->syscall);
+}
+
+static int show_trace_cb(struct stackframe *frame, void *data)
+{
+	const char *loglvl = data;
+
+	if (kernel_text_address(frame->pc))
+		printk("%s [<%08lx>] %pB\n",
+			loglvl, frame->pc, (void *)frame->pc);
+	return 0;
+}
+
+static void show_trace(struct task_struct *task, unsigned long *sp,
+		       const char *loglvl)
+{
+	if (!sp)
+		sp = stack_pointer(task);
+
+	printk("%sCall Trace:\n", loglvl);
+	walk_stackframe(sp, show_trace_cb, (void *)loglvl);
+}
+
+#define STACK_DUMP_ENTRY_SIZE 4
+#define STACK_DUMP_LINE_SIZE 16
+static size_t kstack_depth_to_print = CONFIG_PRINT_STACK_DEPTH;
+
+struct stack_fragment
+{
+	size_t len;
+	size_t off;
+	u8 *sp;
+	const char *loglvl;
+};
+
+static int show_stack_fragment_cb(struct stackframe *frame, void *data)
+{
+	struct stack_fragment *sf = data;
+
+	while (sf->off < sf->len) {
+		u8 line[STACK_DUMP_LINE_SIZE];
+		size_t line_len = sf->len - sf->off > STACK_DUMP_LINE_SIZE ?
+			STACK_DUMP_LINE_SIZE : sf->len - sf->off;
+		bool arrow = sf->off == 0;
+
+		if (frame && frame->sp == (unsigned long)(sf->sp + sf->off))
+			arrow = true;
+
+		__memcpy(line, sf->sp + sf->off, line_len);
+		print_hex_dump(sf->loglvl, arrow ? "> " : "  ", DUMP_PREFIX_NONE,
+			       STACK_DUMP_LINE_SIZE, STACK_DUMP_ENTRY_SIZE,
+			       line, line_len, false);
+		sf->off += STACK_DUMP_LINE_SIZE;
+		if (arrow)
+			return 0;
+	}
+	return 1;
+}
+
+void show_stack(struct task_struct *task, unsigned long *sp, const char *loglvl)
+{
+	struct stack_fragment sf;
+
+	if (!sp)
+		sp = stack_pointer(task);
+
+	sf.len = min((-(size_t)sp) & (THREAD_SIZE - STACK_DUMP_ENTRY_SIZE),
+		     kstack_depth_to_print * STACK_DUMP_ENTRY_SIZE);
+	sf.off = 0;
+	sf.sp = (u8 *)sp;
+	sf.loglvl = loglvl;
+
+	printk("%sStack:\n", loglvl);
+	walk_stackframe(sp, show_stack_fragment_cb, &sf);
+	while (sf.off < sf.len)
+		show_stack_fragment_cb(NULL, &sf);
+	show_trace(task, sp, loglvl);
+}
+
+DEFINE_SPINLOCK(die_lock);
+
+void __noreturn die(const char * str, struct pt_regs * regs, long err)
+{
+	static int die_counter;
+
+	console_verbose();
+	spin_lock_irq(&die_lock);
+
+	pr_info("%s: sig: %ld [#%d]\n", str, err, ++die_counter);
+	show_regs(regs);
+	if (!user_mode(regs))
+		show_stack(NULL, (unsigned long *)regs->areg[1], KERN_INFO);
+
+	add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
+	spin_unlock_irq(&die_lock);
+
+	if (in_interrupt())
+		panic("Fatal exception in interrupt");
+
+	if (panic_on_oops)
+		panic("Fatal exception");
+
+	make_task_dead(err);
+}
diff --git a/arch/xtensa/kernel/vectors.S b/arch/xtensa/kernel/vectors.S
new file mode 100644
index 000000000..1073fe4a5
--- /dev/null
+++ b/arch/xtensa/kernel/vectors.S
@@ -0,0 +1,805 @@
+/*
+ * arch/xtensa/kernel/vectors.S
+ *
+ * This file contains all exception vectors (user, kernel, and double),
+ * as well as the window vectors (overflow and underflow), and the debug
+ * vector. These are the primary vectors executed by the processor if an
+ * exception occurs.
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2005 - 2008 Tensilica, Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ *
+ */
+
+/*
+ * We use a two-level table approach. The user and kernel exception vectors
+ * use a first-level dispatch table to dispatch the exception to a registered
+ * fast handler or the default handler, if no fast handler was registered.
+ * The default handler sets up a C-stack and dispatches the exception to a
+ * registerd C handler in the second-level dispatch table.
+ *
+ * Fast handler entry condition:
+ *
+ *   a0:	trashed, original value saved on stack (PT_AREG0)
+ *   a1:	a1
+ *   a2:	new stack pointer, original value in depc
+ *   a3:	dispatch table
+ *   depc:	a2, original value saved on stack (PT_DEPC)
+ *   excsave_1:	a3
+ *
+ * The value for PT_DEPC saved to stack also functions as a boolean to
+ * indicate that the exception is either a double or a regular exception:
+ *
+ *   PT_DEPC	>= VALID_DOUBLE_EXCEPTION_ADDRESS: double exception
+ *		<  VALID_DOUBLE_EXCEPTION_ADDRESS: regular exception
+ *
+ * Note:  Neither the kernel nor the user exception handler generate literals.
+ *
+ */
+
+#include <linux/linkage.h>
+#include <linux/pgtable.h>
+#include <asm/asmmacro.h>
+#include <asm/ptrace.h>
+#include <asm/current.h>
+#include <asm/asm-offsets.h>
+#include <asm/processor.h>
+#include <asm/page.h>
+#include <asm/thread_info.h>
+#include <asm/vectors.h>
+
+#define WINDOW_VECTORS_SIZE   0x180
+
+
+/*
+ * User exception vector. (Exceptions with PS.UM == 1, PS.EXCM == 0)
+ *
+ * We get here when an exception occurred while we were in userland.
+ * We switch to the kernel stack and jump to the first level handler
+ * associated to the exception cause.
+ *
+ * Note: the saved kernel stack pointer (EXC_TABLE_KSTK) is already
+ *       decremented by PT_USER_SIZE.
+ */
+
+	.section .UserExceptionVector.text, "ax"
+
+ENTRY(_UserExceptionVector)
+
+	xsr	a3, excsave1		# save a3 and get dispatch table
+	wsr	a2, depc		# save a2
+	l32i	a2, a3, EXC_TABLE_KSTK	# load kernel stack to a2
+	s32i	a0, a2, PT_AREG0	# save a0 to ESF
+	rsr	a0, exccause		# retrieve exception cause
+	s32i	a0, a2, PT_DEPC		# mark it as a regular exception
+	addx4	a0, a0, a3		# find entry in table
+	l32i	a0, a0, EXC_TABLE_FAST_USER	# load handler
+	xsr	a3, excsave1		# restore a3 and dispatch table
+	jx	a0
+
+ENDPROC(_UserExceptionVector)
+
+/*
+ * Kernel exception vector. (Exceptions with PS.UM == 0, PS.EXCM == 0)
+ *
+ * We get this exception when we were already in kernel space.
+ * We decrement the current stack pointer (kernel) by PT_KERNEL_SIZE and
+ * jump to the first-level handler associated with the exception cause.
+ *
+ * Note: we need to preserve space for the spill region.
+ */
+
+	.section .KernelExceptionVector.text, "ax"
+
+ENTRY(_KernelExceptionVector)
+
+	xsr	a3, excsave1		# save a3, and get dispatch table
+	wsr	a2, depc		# save a2
+	addi	a2, a1, -16 - PT_KERNEL_SIZE	# adjust stack pointer
+	s32i	a0, a2, PT_AREG0	# save a0 to ESF
+	rsr	a0, exccause		# retrieve exception cause
+	s32i	a0, a2, PT_DEPC		# mark it as a regular exception
+	addx4	a0, a0, a3		# find entry in table
+	l32i	a0, a0, EXC_TABLE_FAST_KERNEL	# load handler address
+	xsr	a3, excsave1		# restore a3 and dispatch table
+	jx	a0
+
+ENDPROC(_KernelExceptionVector)
+
+/*
+ * Double exception vector (Exceptions with PS.EXCM == 1)
+ * We get this exception when another exception occurs while were are
+ * already in an exception, such as window overflow/underflow exception,
+ * or 'expected' exceptions, for example memory exception when we were trying
+ * to read data from an invalid address in user space.
+ *
+ * Note that this vector is never invoked for level-1 interrupts, because such
+ * interrupts are disabled (masked) when PS.EXCM is set.
+ *
+ * We decode the exception and take the appropriate action.  However, the
+ * double exception vector is much more careful, because a lot more error
+ * cases go through the double exception vector than through the user and
+ * kernel exception vectors.
+ *
+ * Occasionally, the kernel expects a double exception to occur.  This usually
+ * happens when accessing user-space memory with the user's permissions
+ * (l32e/s32e instructions).  The kernel state, though, is not always suitable
+ * for immediate transfer of control to handle_double, where "normal" exception
+ * processing occurs. Also in kernel mode, TLB misses can occur if accessing
+ * vmalloc memory, possibly requiring repair in a double exception handler.
+ *
+ * The variable at TABLE_FIXUP offset from the pointer in EXCSAVE_1 doubles as
+ * a boolean variable and a pointer to a fixup routine. If the variable
+ * EXC_TABLE_FIXUP is non-zero, this handler jumps to that address. A value of
+ * zero indicates to use the default kernel/user exception handler.
+ * There is only one exception, when the value is identical to the exc_table
+ * label, the kernel is in trouble. This mechanism is used to protect critical
+ * sections, mainly when the handler writes to the stack to assert the stack
+ * pointer is valid. Once the fixup/default handler leaves that area, the
+ * EXC_TABLE_FIXUP variable is reset to the fixup handler or zero.
+ *
+ * Procedures wishing to use this mechanism should set EXC_TABLE_FIXUP to the
+ * nonzero address of a fixup routine before it could cause a double exception
+ * and reset it before it returns.
+ *
+ * Some other things to take care of when a fast exception handler doesn't
+ * specify a particular fixup handler but wants to use the default handlers:
+ *
+ *  - The original stack pointer (in a1) must not be modified. The fast
+ *    exception handler should only use a2 as the stack pointer.
+ *
+ *  - If the fast handler manipulates the stack pointer (in a2), it has to
+ *    register a valid fixup handler and cannot use the default handlers.
+ *
+ *  - The handler can use any other generic register from a3 to a15, but it
+ *    must save the content of these registers to stack (PT_AREG3...PT_AREGx)
+ *
+ *  - These registers must be saved before a double exception can occur.
+ *
+ *  - If we ever implement handling signals while in double exceptions, the
+ *    number of registers a fast handler has saved (excluding a0 and a1) must
+ *    be written to  PT_AREG1. (1 if only a3 is used, 2 for a3 and a4, etc. )
+ *
+ * The fixup handlers are special handlers:
+ *
+ *  - Fixup entry conditions differ from regular exceptions:
+ *
+ *	a0:	   DEPC
+ *	a1: 	   a1
+ *	a2:	   trashed, original value in EXC_TABLE_DOUBLE_SAVE
+ *	a3:	   exctable
+ *	depc:	   a0
+ *	excsave_1: a3
+ *
+ *  - When the kernel enters the fixup handler, it still assumes it is in a
+ *    critical section, so EXC_TABLE_FIXUP variable is set to exc_table.
+ *    The fixup handler, therefore, has to re-register itself as the fixup
+ *    handler before it returns from the double exception.
+ *
+ *  - Fixup handler can share the same exception frame with the fast handler.
+ *    The kernel stack pointer is not changed when entering the fixup handler.
+ *
+ *  - Fixup handlers can jump to the default kernel and user exception
+ *    handlers. Before it jumps, though, it has to setup a exception frame
+ *    on stack. Because the default handler resets the register fixup handler
+ *    the fixup handler must make sure that the default handler returns to
+ *    it instead of the exception address, so it can re-register itself as
+ *    the fixup handler.
+ *
+ * In case of a critical condition where the kernel cannot recover, we jump
+ * to unrecoverable_exception with the following entry conditions.
+ * All registers a0...a15 are unchanged from the last exception, except:
+ *
+ *	a0:	   last address before we jumped to the unrecoverable_exception.
+ *	excsave_1: a0
+ *
+ *
+ * See the handle_alloca_user and spill_registers routines for example clients.
+ *
+ * FIXME: Note: we currently don't allow signal handling coming from a double
+ *        exception, so the item markt with (*) is not required.
+ */
+
+	.section .DoubleExceptionVector.text, "ax"
+
+ENTRY(_DoubleExceptionVector)
+
+	xsr	a3, excsave1
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+
+	/* Check for kernel double exception (usually fatal). */
+
+	rsr	a2, ps
+	_bbsi.l	a2, PS_UM_BIT, 1f
+	j	.Lksp
+
+	.align	4
+	.literal_position
+1:
+	/* Check if we are currently handling a window exception. */
+	/* Note: We don't need to indicate that we enter a critical section. */
+
+	xsr	a0, depc		# get DEPC, save a0
+
+#ifdef SUPPORT_WINDOWED
+	movi	a2, WINDOW_VECTORS_VADDR
+	_bltu	a0, a2, .Lfixup
+	addi	a2, a2, WINDOW_VECTORS_SIZE
+	_bgeu	a0, a2, .Lfixup
+
+	/* Window overflow/underflow exception. Get stack pointer. */
+
+	l32i	a2, a3, EXC_TABLE_KSTK
+
+	/* Check for overflow/underflow exception, jump if overflow. */
+
+	bbci.l	a0, 6, _DoubleExceptionVector_WindowOverflow
+
+	/*
+	 * Restart window underflow exception.
+	 * Currently:
+	 *	depc = orig a0,
+	 *	a0 = orig DEPC,
+	 *	a2 = new sp based on KSTK from exc_table
+	 *	a3 = excsave_1
+	 *	excsave_1 = orig a3
+	 *
+	 * We return to the instruction in user space that caused the window
+	 * underflow exception. Therefore, we change window base to the value
+	 * before we entered the window underflow exception and prepare the
+	 * registers to return as if we were coming from a regular exception
+	 * by changing depc (in a0).
+	 * Note: We can trash the current window frame (a0...a3) and depc!
+	 */
+_DoubleExceptionVector_WindowUnderflow:
+	xsr	a3, excsave1
+	wsr	a2, depc		# save stack pointer temporarily
+	rsr	a0, ps
+	extui	a0, a0, PS_OWB_SHIFT, PS_OWB_WIDTH
+	wsr	a0, windowbase
+	rsync
+
+	/* We are now in the previous window frame. Save registers again. */
+
+	xsr	a2, depc		# save a2 and get stack pointer
+	s32i	a0, a2, PT_AREG0
+	xsr	a3, excsave1
+	rsr	a0, exccause
+	s32i	a0, a2, PT_DEPC		# mark it as a regular exception
+	addx4	a0, a0, a3
+	xsr	a3, excsave1
+	l32i	a0, a0, EXC_TABLE_FAST_USER
+	jx	a0
+
+#else
+	j	.Lfixup
+#endif
+
+	/*
+	 * We only allow the ITLB miss exception if we are in kernel space.
+	 * All other exceptions are unexpected and thus unrecoverable!
+	 */
+
+#ifdef CONFIG_MMU
+	.extern fast_second_level_miss_double_kernel
+
+.Lksp:	/* a0: a0, a1: a1, a2: a2, a3: trashed, depc: depc, excsave: a3 */
+
+	rsr	a3, exccause
+	beqi	a3, EXCCAUSE_ITLB_MISS, 1f
+	addi	a3, a3, -EXCCAUSE_DTLB_MISS
+	bnez	a3, .Lunrecoverable
+1:	movi	a3, fast_second_level_miss_double_kernel
+	jx	a3
+#else
+.equ	.Lksp,	.Lunrecoverable
+#endif
+
+	/* Critical! We can't handle this situation. PANIC! */
+
+	.extern unrecoverable_exception
+
+.Lunrecoverable_fixup:
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a0, depc
+
+.Lunrecoverable:
+	rsr	a3, excsave1
+	wsr	a0, excsave1
+	call0	unrecoverable_exception
+
+.Lfixup:/* Check for a fixup handler or if we were in a critical section. */
+
+	/* a0: depc, a1: a1, a2: trash, a3: exctable, depc: a0, excsave1: a3 */
+
+	/* Enter critical section. */
+
+	l32i	a2, a3, EXC_TABLE_FIXUP
+	s32i	a3, a3, EXC_TABLE_FIXUP
+	beq	a2, a3, .Lunrecoverable_fixup	# critical section
+	beqz	a2, .Ldflt			# no handler was registered
+
+	/* a0: depc, a1: a1, a2: trash, a3: exctable, depc: a0, excsave: a3 */
+
+	jx	a2
+
+.Ldflt:	/* Get stack pointer. */
+
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	addi	a2, a2, -PT_USER_SIZE
+
+	/* a0: depc, a1: a1, a2: kstk, a3: exctable, depc: a0, excsave: a3 */
+
+	s32i	a0, a2, PT_DEPC
+	l32i	a0, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a0, depc
+	s32i	a0, a2, PT_AREG0
+
+	/* a0: avail, a1: a1, a2: kstk, a3: exctable, depc: a2, excsave: a3 */
+
+	rsr	a0, exccause
+	addx4	a0, a0, a3
+	xsr	a3, excsave1
+	l32i	a0, a0, EXC_TABLE_FAST_USER
+	jx	a0
+
+#ifdef SUPPORT_WINDOWED
+	/*
+	 * Restart window OVERFLOW exception.
+	 * Currently:
+	 *	depc = orig a0,
+	 *	a0 = orig DEPC,
+	 *	a2 = new sp based on KSTK from exc_table
+	 *	a3 = EXCSAVE_1
+	 *	excsave_1 = orig a3
+	 *
+	 * We return to the instruction in user space that caused the window
+	 * overflow exception. Therefore, we change window base to the value
+	 * before we entered the window overflow exception and prepare the
+	 * registers to return as if we were coming from a regular exception
+	 * by changing DEPC (in a0).
+	 *
+	 * NOTE: We CANNOT trash the current window frame (a0...a3), but we
+	 * can clobber depc.
+	 *
+	 * The tricky part here is that overflow8 and overflow12 handlers
+	 * save a0, then clobber a0.  To restart the handler, we have to restore
+	 * a0 if the double exception was past the point where a0 was clobbered.
+	 *
+	 * To keep things simple, we take advantage of the fact all overflow
+	 * handlers save a0 in their very first instruction.  If DEPC was past
+	 * that instruction, we can safely restore a0 from where it was saved
+	 * on the stack.
+	 *
+	 * a0: depc, a1: a1, a2: kstk, a3: exc_table, depc: a0, excsave1: a3
+	 */
+_DoubleExceptionVector_WindowOverflow:
+	extui	a2, a0, 0, 6	# get offset into 64-byte vector handler
+	beqz	a2, 1f		# if at start of vector, don't restore
+
+	addi	a0, a0, -128
+	bbsi.l	a0, 8, 1f	# don't restore except for overflow 8 and 12
+
+	/*
+	 * This fixup handler is for the extremely unlikely case where the
+	 * overflow handler's reference thru a0 gets a hardware TLB refill
+	 * that bumps out the (distinct, aliasing) TLB entry that mapped its
+	 * prior references thru a9/a13, and where our reference now thru
+	 * a9/a13 gets a 2nd-level miss exception (not hardware TLB refill).
+	 */
+	movi	a2, window_overflow_restore_a0_fixup
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+
+	bbsi.l	a0, 7, 2f
+
+	/*
+	 * Restore a0 as saved by _WindowOverflow8().
+	 */
+
+	l32e	a0, a9, -16
+	wsr	a0, depc	# replace the saved a0
+	j	3f
+
+2:
+	/*
+	 * Restore a0 as saved by _WindowOverflow12().
+	 */
+
+	l32e	a0, a13, -16
+	wsr	a0, depc	# replace the saved a0
+3:
+	xsr	a3, excsave1
+	movi	a0, 0
+	s32i	a0, a3, EXC_TABLE_FIXUP
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+1:
+	/*
+	 * Restore WindowBase while leaving all address registers restored.
+	 * We have to use ROTW for this, because WSR.WINDOWBASE requires
+	 * an address register (which would prevent restore).
+	 *
+	 * Window Base goes from 0 ... 7 (Module 8)
+	 * Window Start is 8 bits; Ex: (0b1010 1010):0x55 from series of call4s
+	 */
+
+	rsr	a0, ps
+	extui	a0, a0, PS_OWB_SHIFT, PS_OWB_WIDTH
+	rsr	a2, windowbase
+	sub	a0, a2, a0
+	extui	a0, a0, 0, 3
+
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+	beqi	a0, 1, .L1pane
+	beqi	a0, 3, .L3pane
+
+	rsr	a0, depc
+	rotw	-2
+
+	/*
+	 * We are now in the user code's original window frame.
+	 * Process the exception as a user exception as if it was
+	 * taken by the user code.
+	 *
+	 * This is similar to the user exception vector,
+	 * except that PT_DEPC isn't set to EXCCAUSE.
+	 */
+1:
+	xsr	a3, excsave1
+	wsr	a2, depc
+	l32i	a2, a3, EXC_TABLE_KSTK
+	s32i	a0, a2, PT_AREG0
+	rsr	a0, exccause
+
+	s32i	a0, a2, PT_DEPC
+
+_DoubleExceptionVector_handle_exception:
+	addi	a0, a0, -EXCCAUSE_UNALIGNED
+	beqz	a0, 2f
+	addx4	a0, a0, a3
+	l32i	a0, a0, EXC_TABLE_FAST_USER + 4 * EXCCAUSE_UNALIGNED
+	xsr	a3, excsave1
+	jx	a0
+2:
+	movi	a0, user_exception
+	xsr	a3, excsave1
+	jx	a0
+
+.L1pane:
+	rsr	a0, depc
+	rotw	-1
+	j	1b
+
+.L3pane:
+	rsr	a0, depc
+	rotw	-3
+	j	1b
+#endif
+
+ENDPROC(_DoubleExceptionVector)
+
+#ifdef SUPPORT_WINDOWED
+
+/*
+ * Fixup handler for TLB miss in double exception handler for window owerflow.
+ * We get here with windowbase set to the window that was being spilled and
+ * a0 trashed. a0 bit 7 determines if this is a call8 (bit clear) or call12
+ * (bit set) window.
+ *
+ * We do the following here:
+ * - go to the original window retaining a0 value;
+ * - set up exception stack to return back to appropriate a0 restore code
+ *   (we'll need to rotate window back and there's no place to save this
+ *    information, use different return address for that);
+ * - handle the exception;
+ * - go to the window that was being spilled;
+ * - set up window_overflow_restore_a0_fixup as a fixup routine;
+ * - reload a0;
+ * - restore the original window;
+ * - reset the default fixup routine;
+ * - return to user. By the time we get to this fixup handler all information
+ *   about the conditions of the original double exception that happened in
+ *   the window overflow handler is lost, so we just return to userspace to
+ *   retry overflow from start.
+ *
+ * a0: value of depc, original value in depc
+ * a2: trashed, original value in EXC_TABLE_DOUBLE_SAVE
+ * a3: exctable, original value in excsave1
+ */
+
+	__XTENSA_HANDLER
+	.literal_position
+
+ENTRY(window_overflow_restore_a0_fixup)
+
+	rsr	a0, ps
+	extui	a0, a0, PS_OWB_SHIFT, PS_OWB_WIDTH
+	rsr	a2, windowbase
+	sub	a0, a2, a0
+	extui	a0, a0, 0, 3
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+
+	_beqi	a0, 1, .Lhandle_1
+	_beqi	a0, 3, .Lhandle_3
+
+	.macro	overflow_fixup_handle_exception_pane n
+
+	rsr	a0, depc
+	rotw	-\n
+
+	xsr	a3, excsave1
+	wsr	a2, depc
+	l32i	a2, a3, EXC_TABLE_KSTK
+	s32i	a0, a2, PT_AREG0
+
+	movi	a0, .Lrestore_\n
+	s32i	a0, a2, PT_DEPC
+	rsr	a0, exccause
+	j	_DoubleExceptionVector_handle_exception
+
+	.endm
+
+	overflow_fixup_handle_exception_pane 2
+.Lhandle_1:
+	overflow_fixup_handle_exception_pane 1
+.Lhandle_3:
+	overflow_fixup_handle_exception_pane 3
+
+	.macro	overflow_fixup_restore_a0_pane n
+
+	rotw	\n
+	/* Need to preserve a0 value here to be able to handle exception
+	 * that may occur on a0 reload from stack. It may occur because
+	 * TLB miss handler may not be atomic and pointer to page table
+	 * may be lost before we get here. There are no free registers,
+	 * so we need to use EXC_TABLE_DOUBLE_SAVE area.
+	 */
+	xsr	a3, excsave1
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	movi	a2, window_overflow_restore_a0_fixup
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+	bbsi.l	a0, 7, 1f
+	l32e	a0, a9, -16
+	j	2f
+1:
+	l32e	a0, a13, -16
+2:
+	rotw	-\n
+
+	.endm
+
+.Lrestore_2:
+	overflow_fixup_restore_a0_pane 2
+
+.Lset_default_fixup:
+	xsr	a3, excsave1
+	s32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	movi	a2, 0
+	s32i	a2, a3, EXC_TABLE_FIXUP
+	l32i	a2, a3, EXC_TABLE_DOUBLE_SAVE
+	xsr	a3, excsave1
+	rfe
+
+.Lrestore_1:
+	overflow_fixup_restore_a0_pane 1
+	j	.Lset_default_fixup
+.Lrestore_3:
+	overflow_fixup_restore_a0_pane 3
+	j	.Lset_default_fixup
+
+ENDPROC(window_overflow_restore_a0_fixup)
+
+#endif
+
+/*
+ * Debug interrupt vector
+ *
+ * There is not much space here, so simply jump to another handler.
+ * EXCSAVE[DEBUGLEVEL] has been set to that handler.
+ */
+
+	.section .DebugInterruptVector.text, "ax"
+
+ENTRY(_DebugInterruptVector)
+
+	xsr	a3, SREG_EXCSAVE + XCHAL_DEBUGLEVEL
+	s32i	a0, a3, DT_DEBUG_SAVE
+	l32i	a0, a3, DT_DEBUG_EXCEPTION
+	jx	a0
+
+ENDPROC(_DebugInterruptVector)
+
+
+
+/*
+ * Medium priority level interrupt vectors
+ *
+ * Each takes less than 16 (0x10) bytes, no literals, by placing
+ * the extra 8 bytes that would otherwise be required in the window
+ * vectors area where there is space.  With relocatable vectors,
+ * all vectors are within ~ 4 kB range of each other, so we can
+ * simply jump (J) to another vector without having to use JX.
+ *
+ * common_exception code gets current IRQ level in PS.INTLEVEL
+ * and preserves it for the IRQ handling time.
+ */
+
+	.macro	irq_entry_level level
+
+	.if	XCHAL_EXCM_LEVEL >= \level
+	.section .Level\level\()InterruptVector.text, "ax"
+ENTRY(_Level\level\()InterruptVector)
+	wsr	a0, excsave2
+	rsr	a0, epc\level
+	wsr	a0, epc1
+	.if	\level <= LOCKLEVEL
+	movi	a0, EXCCAUSE_LEVEL1_INTERRUPT
+	.else
+	movi	a0, EXCCAUSE_MAPPED_NMI
+	.endif
+	wsr	a0, exccause
+	rsr	a0, eps\level
+					# branch to user or kernel vector
+	j	_SimulateUserKernelVectorException
+	.endif
+
+	.endm
+
+	irq_entry_level 2
+	irq_entry_level 3
+	irq_entry_level 4
+	irq_entry_level 5
+	irq_entry_level 6
+
+#if XCHAL_EXCM_LEVEL >= 2
+	/*
+	 *  Continuation of medium priority interrupt dispatch code.
+	 *  On entry here, a0 contains PS, and EPC2 contains saved a0:
+	 */
+	__XTENSA_HANDLER
+	.align 4
+_SimulateUserKernelVectorException:
+	addi	a0, a0, (1 << PS_EXCM_BIT)
+#if !XTENSA_FAKE_NMI
+	wsr	a0, ps
+#endif
+	bbsi.l	a0, PS_UM_BIT, 1f	# branch if user mode
+	xsr	a0, excsave2		# restore a0
+	j	_KernelExceptionVector	# simulate kernel vector exception
+1:	xsr	a0, excsave2		# restore a0
+	j	_UserExceptionVector	# simulate user vector exception
+#endif
+
+
+/* Window overflow and underflow handlers.
+ * The handlers must be 64 bytes apart, first starting with the underflow
+ * handlers underflow-4 to underflow-12, then the overflow handlers
+ * overflow-4 to overflow-12.
+ *
+ * Note: We rerun the underflow handlers if we hit an exception, so
+ *	 we try to access any page that would cause a page fault early.
+ */
+
+#define ENTRY_ALIGN64(name)	\
+	.globl name;		\
+	.align 64;		\
+	name:
+
+	.section		.WindowVectors.text, "ax"
+
+
+#ifdef SUPPORT_WINDOWED
+
+/* 4-Register Window Overflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowOverflow4)
+
+	s32e	a0, a5, -16
+	s32e	a1, a5, -12
+	s32e	a2, a5,  -8
+	s32e	a3, a5,  -4
+	rfwo
+
+ENDPROC(_WindowOverflow4)
+
+/* 4-Register Window Underflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowUnderflow4)
+
+	l32e	a0, a5, -16
+	l32e	a1, a5, -12
+	l32e	a2, a5,  -8
+	l32e	a3, a5,  -4
+	rfwu
+
+ENDPROC(_WindowUnderflow4)
+
+/* 8-Register Window Overflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowOverflow8)
+
+	s32e	a0, a9, -16
+	l32e	a0, a1, -12
+	s32e	a2, a9,  -8
+	s32e	a1, a9, -12
+	s32e	a3, a9,  -4
+	s32e	a4, a0, -32
+	s32e	a5, a0, -28
+	s32e	a6, a0, -24
+	s32e	a7, a0, -20
+	rfwo
+
+ENDPROC(_WindowOverflow8)
+
+/* 8-Register Window Underflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowUnderflow8)
+
+	l32e	a1, a9, -12
+	l32e	a0, a9, -16
+	l32e	a7, a1, -12
+	l32e	a2, a9,  -8
+	l32e	a4, a7, -32
+	l32e	a3, a9,  -4
+	l32e	a5, a7, -28
+	l32e	a6, a7, -24
+	l32e	a7, a7, -20
+	rfwu
+
+ENDPROC(_WindowUnderflow8)
+
+/* 12-Register Window Overflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowOverflow12)
+
+	s32e	a0,  a13, -16
+	l32e	a0,  a1,  -12
+	s32e	a1,  a13, -12
+	s32e	a2,  a13,  -8
+	s32e	a3,  a13,  -4
+	s32e	a4,  a0,  -48
+	s32e	a5,  a0,  -44
+	s32e	a6,  a0,  -40
+	s32e	a7,  a0,  -36
+	s32e	a8,  a0,  -32
+	s32e	a9,  a0,  -28
+	s32e	a10, a0,  -24
+	s32e	a11, a0,  -20
+	rfwo
+
+ENDPROC(_WindowOverflow12)
+
+/* 12-Register Window Underflow Vector (Handler) */
+
+ENTRY_ALIGN64(_WindowUnderflow12)
+
+	l32e	a1,  a13, -12
+	l32e	a0,  a13, -16
+	l32e	a11, a1,  -12
+	l32e	a2,  a13,  -8
+	l32e	a4,  a11, -48
+	l32e	a8,  a11, -32
+	l32e	a3,  a13,  -4
+	l32e	a5,  a11, -44
+	l32e	a6,  a11, -40
+	l32e	a7,  a11, -36
+	l32e	a9,  a11, -28
+	l32e	a10, a11, -24
+	l32e	a11, a11, -20
+	rfwu
+
+ENDPROC(_WindowUnderflow12)
+
+#endif
+
+	.text
diff --git a/arch/xtensa/kernel/vmlinux.lds.S b/arch/xtensa/kernel/vmlinux.lds.S
new file mode 100644
index 000000000..f47e9bbbd
--- /dev/null
+++ b/arch/xtensa/kernel/vmlinux.lds.S
@@ -0,0 +1,388 @@
+/*
+ * arch/xtensa/kernel/vmlinux.lds.S
+ *
+ * Xtensa linker script
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2008 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Marc Gauthier <marc@tensilica.com, marc@alumni.uwaterloo.ca>
+ * Joe Taylor <joe@tensilica.com, joetylr@yahoo.com>
+ */
+
+#define RO_EXCEPTION_TABLE_ALIGN	16
+
+#include <asm-generic/vmlinux.lds.h>
+#include <asm/page.h>
+#include <asm/thread_info.h>
+
+#include <asm/core.h>
+#include <asm/vectors.h>
+
+OUTPUT_ARCH(xtensa)
+ENTRY(_start)
+
+#ifdef __XTENSA_EB__
+jiffies = jiffies_64 + 4;
+#else
+jiffies = jiffies_64;
+#endif
+
+/* Note: In the following macros, it would be nice to specify only the
+   vector name and section kind and construct "sym" and "section" using
+   CPP concatenation, but that does not work reliably.  Concatenating a
+   string with "." produces an invalid token.  CPP will not print a
+   warning because it thinks this is an assembly file, but it leaves
+   them as multiple tokens and there may or may not be whitespace
+   between them.  */
+
+/* Macro for a relocation entry */
+
+#define RELOCATE_ENTRY(sym, section)		\
+	LONG(sym ## _start);			\
+	LONG(sym ## _end);			\
+	LONG(LOADADDR(section))
+
+#if !defined(CONFIG_VECTORS_ADDR) && XCHAL_HAVE_VECBASE
+#define MERGED_VECTORS 1
+#else
+#define MERGED_VECTORS 0
+#endif
+
+/*
+ * Macro to define a section for a vector. When MERGED_VECTORS is 0
+ * code for every vector is located with other init data. At startup
+ * time head.S copies code for every vector to its final position according
+ * to description recorded in the corresponding RELOCATE_ENTRY.
+ */
+
+#define SECTION_VECTOR4(sym, section, addr, prevsec)			    \
+  section addr : AT(((LOADADDR(prevsec) + SIZEOF(prevsec)) + 3) & ~ 3)      \
+  {									    \
+    . = ALIGN(4);							    \
+    sym ## _start = ABSOLUTE(.);		 			    \
+    *(section)								    \
+    sym ## _end = ABSOLUTE(.);						    \
+  }
+
+#define SECTION_VECTOR2(section, addr)					    \
+  . = addr;								    \
+  *(section)
+
+/*
+ *  Mapping of input sections to output sections when linking.
+ */
+
+SECTIONS
+{
+  . = KERNELOFFSET;
+  /* .text section */
+
+  _text = .;
+  _stext = .;
+
+  .text :
+  {
+    /* The HEAD_TEXT section must be the first section! */
+    HEAD_TEXT
+
+#if MERGED_VECTORS
+    . = ALIGN(PAGE_SIZE);
+    _vecbase = .;
+
+#ifdef SUPPORT_WINDOWED
+    SECTION_VECTOR2 (.WindowVectors.text, WINDOW_VECTORS_VADDR)
+#endif
+#if XCHAL_EXCM_LEVEL >= 2
+    SECTION_VECTOR2 (.Level2InterruptVector.text, INTLEVEL2_VECTOR_VADDR)
+#endif
+#if XCHAL_EXCM_LEVEL >= 3
+    SECTION_VECTOR2 (.Level3InterruptVector.text, INTLEVEL3_VECTOR_VADDR)
+#endif
+#if XCHAL_EXCM_LEVEL >= 4
+    SECTION_VECTOR2 (.Level4InterruptVector.text, INTLEVEL4_VECTOR_VADDR)
+#endif
+#if XCHAL_EXCM_LEVEL >= 5
+    SECTION_VECTOR2 (.Level5InterruptVector.text, INTLEVEL5_VECTOR_VADDR)
+#endif
+#if XCHAL_EXCM_LEVEL >= 6
+    SECTION_VECTOR2 (.Level6InterruptVector.text, INTLEVEL6_VECTOR_VADDR)
+#endif
+    SECTION_VECTOR2 (.DebugInterruptVector.text, DEBUG_VECTOR_VADDR)
+    SECTION_VECTOR2 (.KernelExceptionVector.text, KERNEL_VECTOR_VADDR)
+    SECTION_VECTOR2 (.UserExceptionVector.text, USER_VECTOR_VADDR)
+    SECTION_VECTOR2 (.DoubleExceptionVector.text, DOUBLEEXC_VECTOR_VADDR)
+
+    *(.exception.text)
+    *(.xiptext)
+#endif
+
+    IRQENTRY_TEXT
+    SOFTIRQENTRY_TEXT
+    ENTRY_TEXT
+    TEXT_TEXT
+    SCHED_TEXT
+    LOCK_TEXT
+    *(.fixup)
+  }
+  _etext = .;
+  PROVIDE (etext = .);
+
+  . = ALIGN(16);
+
+  RO_DATA(4096)
+
+  /* Data section */
+
+#ifdef CONFIG_XIP_KERNEL
+  INIT_TEXT_SECTION(PAGE_SIZE)
+#else
+  _sdata = .;
+  RW_DATA(XCHAL_ICACHE_LINESIZE, PAGE_SIZE, THREAD_SIZE)
+  _edata = .;
+
+  /* Initialization code and data: */
+
+  . = ALIGN(PAGE_SIZE);
+  __init_begin = .;
+  INIT_TEXT_SECTION(PAGE_SIZE)
+
+  .init.data :
+  {
+    INIT_DATA
+  }
+#endif
+
+  .init.rodata :
+  {
+    . = ALIGN(0x4);
+    __tagtable_begin = .;
+    *(.taglist)
+    __tagtable_end = .;
+
+    . = ALIGN(16);
+    __boot_reloc_table_start = ABSOLUTE(.);
+
+#if !MERGED_VECTORS
+#ifdef SUPPORT_WINDOWED
+    RELOCATE_ENTRY(_WindowVectors_text,
+		   .WindowVectors.text);
+#endif
+#if XCHAL_EXCM_LEVEL >= 2
+    RELOCATE_ENTRY(_Level2InterruptVector_text,
+		   .Level2InterruptVector.text);
+#endif
+#if XCHAL_EXCM_LEVEL >= 3
+    RELOCATE_ENTRY(_Level3InterruptVector_text,
+		   .Level3InterruptVector.text);
+#endif
+#if XCHAL_EXCM_LEVEL >= 4
+    RELOCATE_ENTRY(_Level4InterruptVector_text,
+		   .Level4InterruptVector.text);
+#endif
+#if XCHAL_EXCM_LEVEL >= 5
+    RELOCATE_ENTRY(_Level5InterruptVector_text,
+		   .Level5InterruptVector.text);
+#endif
+#if XCHAL_EXCM_LEVEL >= 6
+    RELOCATE_ENTRY(_Level6InterruptVector_text,
+		   .Level6InterruptVector.text);
+#endif
+    RELOCATE_ENTRY(_KernelExceptionVector_text,
+		   .KernelExceptionVector.text);
+    RELOCATE_ENTRY(_UserExceptionVector_text,
+		   .UserExceptionVector.text);
+    RELOCATE_ENTRY(_DoubleExceptionVector_text,
+		   .DoubleExceptionVector.text);
+    RELOCATE_ENTRY(_DebugInterruptVector_text,
+		   .DebugInterruptVector.text);
+    RELOCATE_ENTRY(_exception_text,
+		   .exception.text);
+#ifdef CONFIG_XIP_KERNEL
+    RELOCATE_ENTRY(_xip_text, .xiptext);
+#endif
+#endif
+#ifdef CONFIG_XIP_KERNEL
+    RELOCATE_ENTRY(_xip_data, .data);
+    RELOCATE_ENTRY(_xip_init_data, .init.data);
+#endif
+#if defined(CONFIG_SECONDARY_RESET_VECTOR)
+    RELOCATE_ENTRY(_SecondaryResetVector_text,
+		   .SecondaryResetVector.text);
+#endif
+
+    __boot_reloc_table_end = ABSOLUTE(.) ;
+
+    INIT_SETUP(XCHAL_ICACHE_LINESIZE)
+    INIT_CALLS
+    CON_INITCALL
+    INIT_RAM_FS
+  }
+
+  PERCPU_SECTION(XCHAL_ICACHE_LINESIZE)
+
+  /* We need this dummy segment here */
+
+  . = ALIGN(4);
+  .dummy : { LONG(0) }
+
+#undef LAST
+#define LAST	.dummy
+
+#if !MERGED_VECTORS
+  /* The vectors are relocated to the real position at startup time */
+
+#ifdef SUPPORT_WINDOWED
+  SECTION_VECTOR4 (_WindowVectors_text,
+		  .WindowVectors.text,
+		  WINDOW_VECTORS_VADDR,
+		  LAST)
+#undef LAST
+#define LAST	.WindowVectors.text
+#endif
+  SECTION_VECTOR4 (_DebugInterruptVector_text,
+		  .DebugInterruptVector.text,
+		  DEBUG_VECTOR_VADDR,
+		  LAST)
+#undef LAST
+#define LAST	.DebugInterruptVector.text
+#if XCHAL_EXCM_LEVEL >= 2
+  SECTION_VECTOR4 (_Level2InterruptVector_text,
+		  .Level2InterruptVector.text,
+		  INTLEVEL2_VECTOR_VADDR,
+		  LAST)
+# undef LAST
+# define LAST	.Level2InterruptVector.text
+#endif
+#if XCHAL_EXCM_LEVEL >= 3
+  SECTION_VECTOR4 (_Level3InterruptVector_text,
+		  .Level3InterruptVector.text,
+		  INTLEVEL3_VECTOR_VADDR,
+		  LAST)
+# undef LAST
+# define LAST	.Level3InterruptVector.text
+#endif
+#if XCHAL_EXCM_LEVEL >= 4
+  SECTION_VECTOR4 (_Level4InterruptVector_text,
+		  .Level4InterruptVector.text,
+		  INTLEVEL4_VECTOR_VADDR,
+		  LAST)
+# undef LAST
+# define LAST	.Level4InterruptVector.text
+#endif
+#if XCHAL_EXCM_LEVEL >= 5
+  SECTION_VECTOR4 (_Level5InterruptVector_text,
+		  .Level5InterruptVector.text,
+		  INTLEVEL5_VECTOR_VADDR,
+		  LAST)
+# undef LAST
+# define LAST	.Level5InterruptVector.text
+#endif
+#if XCHAL_EXCM_LEVEL >= 6
+  SECTION_VECTOR4 (_Level6InterruptVector_text,
+		  .Level6InterruptVector.text,
+		  INTLEVEL6_VECTOR_VADDR,
+		  LAST)
+# undef LAST
+# define LAST	.Level6InterruptVector.text
+#endif
+  SECTION_VECTOR4 (_KernelExceptionVector_text,
+		  .KernelExceptionVector.text,
+		  KERNEL_VECTOR_VADDR,
+		  LAST)
+#undef LAST
+  SECTION_VECTOR4 (_UserExceptionVector_text,
+		  .UserExceptionVector.text,
+		  USER_VECTOR_VADDR,
+		  .KernelExceptionVector.text)
+  SECTION_VECTOR4 (_DoubleExceptionVector_text,
+		  .DoubleExceptionVector.text,
+		  DOUBLEEXC_VECTOR_VADDR,
+		  .UserExceptionVector.text)
+#define LAST .DoubleExceptionVector.text
+
+#endif
+#if defined(CONFIG_SECONDARY_RESET_VECTOR)
+
+  SECTION_VECTOR4 (_SecondaryResetVector_text,
+		  .SecondaryResetVector.text,
+		  RESET_VECTOR1_VADDR,
+		  LAST)
+#undef LAST
+#define LAST .SecondaryResetVector.text
+
+#endif
+#if !MERGED_VECTORS
+  SECTION_VECTOR4 (_exception_text,
+		  .exception.text,
+		  ,
+		  LAST)
+#undef LAST
+#define LAST .exception.text
+  SECTION_VECTOR4 (_xip_text,
+		  .xiptext,
+		  ,
+		  LAST)
+#undef LAST
+#define LAST .xiptext
+#endif
+  . = (LOADADDR(LAST) + SIZEOF(LAST) + 3) & ~ 3;
+
+  .dummy1 : AT(ADDR(.dummy1)) { LONG(0) }
+  . = ALIGN(PAGE_SIZE);
+
+#ifndef CONFIG_XIP_KERNEL
+  __init_end = .;
+
+  BSS_SECTION(0, 8192, 0)
+#endif
+
+  _end = .;
+
+#ifdef CONFIG_XIP_KERNEL
+  . = CONFIG_XIP_DATA_ADDR;
+
+  _xip_start = .;
+
+#undef LOAD_OFFSET
+#define LOAD_OFFSET \
+  (CONFIG_XIP_DATA_ADDR - (LOADADDR(.dummy1) + SIZEOF(.dummy1) + 3) & ~ 3)
+
+  _xip_data_start = .;
+  _sdata = .;
+  RW_DATA(XCHAL_ICACHE_LINESIZE, PAGE_SIZE, THREAD_SIZE)
+  _edata = .;
+  _xip_data_end = .;
+
+  /* Initialization data: */
+
+  STRUCT_ALIGN();
+
+  _xip_init_data_start = .;
+  __init_begin = .;
+  .init.data :
+  {
+    INIT_DATA
+  }
+  _xip_init_data_end = .;
+  __init_end = .;
+  BSS_SECTION(0, 8192, 0)
+
+  _xip_end = .;
+
+#undef LOAD_OFFSET
+#endif
+
+  DWARF_DEBUG
+
+  .xt.prop 0 : { KEEP(*(.xt.prop .xt.prop.* .gnu.linkonce.prop.*)) }
+  .xt.insn 0 : { KEEP(*(.xt.insn .xt.insn.* .gnu.linkonce.x*)) }
+  .xt.lit  0 : { KEEP(*(.xt.lit  .xt.lit.*  .gnu.linkonce.p*)) }
+
+  /* Sections to be discarded */
+  DISCARDS
+}
diff --git a/arch/xtensa/kernel/xtensa_ksyms.c b/arch/xtensa/kernel/xtensa_ksyms.c
new file mode 100644
index 000000000..ced335b4d
--- /dev/null
+++ b/arch/xtensa/kernel/xtensa_ksyms.c
@@ -0,0 +1,28 @@
+/*
+ * arch/xtensa/kernel/xtensa_ksyms.c
+ *
+ * Export Xtensa-specific functions for loadable modules.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005  Tensilica Inc.
+ *
+ * Joe Taylor <joe@tensilica.com>
+ */
+
+#include <linux/module.h>
+#include <asm/pgtable.h>
+
+unsigned int __sync_fetch_and_and_4(volatile void *p, unsigned int v)
+{
+	BUG();
+}
+EXPORT_SYMBOL(__sync_fetch_and_and_4);
+
+unsigned int __sync_fetch_and_or_4(volatile void *p, unsigned int v)
+{
+	BUG();
+}
+EXPORT_SYMBOL(__sync_fetch_and_or_4);
diff --git a/arch/xtensa/lib/Makefile b/arch/xtensa/lib/Makefile
new file mode 100644
index 000000000..ec0d8d233
--- /dev/null
+++ b/arch/xtensa/lib/Makefile
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for Xtensa-specific library files.
+#
+
+lib-y	+= memcopy.o memset.o checksum.o \
+	   ashldi3.o ashrdi3.o bswapdi2.o bswapsi2.o lshrdi3.o \
+	   divsi3.o udivsi3.o modsi3.o umodsi3.o mulsi3.o umulsidi3.o \
+	   usercopy.o strnlen_user.o
+lib-$(CONFIG_ARCH_HAS_STRNCPY_FROM_USER) += strncpy_user.o
+lib-$(CONFIG_PCI) += pci-auto.o
diff --git a/arch/xtensa/lib/ashldi3.S b/arch/xtensa/lib/ashldi3.S
new file mode 100644
index 000000000..cd6b73121
--- /dev/null
+++ b/arch/xtensa/lib/ashldi3.S
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+#ifdef __XTENSA_EB__
+#define uh a2
+#define ul a3
+#else
+#define uh a3
+#define ul a2
+#endif /* __XTENSA_EB__ */
+
+ENTRY(__ashldi3)
+
+	abi_entry_default
+	ssl	a4
+	bgei	a4, 32, .Llow_only
+	src	uh, uh, ul
+	sll	ul, ul
+	abi_ret_default
+
+.Llow_only:
+	sll	uh, ul
+	movi	ul, 0
+	abi_ret_default
+
+ENDPROC(__ashldi3)
+EXPORT_SYMBOL(__ashldi3)
diff --git a/arch/xtensa/lib/ashrdi3.S b/arch/xtensa/lib/ashrdi3.S
new file mode 100644
index 000000000..07bc6e758
--- /dev/null
+++ b/arch/xtensa/lib/ashrdi3.S
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+#ifdef __XTENSA_EB__
+#define uh a2
+#define ul a3
+#else
+#define uh a3
+#define ul a2
+#endif /* __XTENSA_EB__ */
+
+ENTRY(__ashrdi3)
+
+	abi_entry_default
+	ssr	a4
+	bgei	a4, 32, .Lhigh_only
+	src	ul, uh, ul
+	sra	uh, uh
+	abi_ret_default
+
+.Lhigh_only:
+	sra	ul, uh
+	srai	uh, uh, 31
+	abi_ret_default
+
+ENDPROC(__ashrdi3)
+EXPORT_SYMBOL(__ashrdi3)
diff --git a/arch/xtensa/lib/bswapdi2.S b/arch/xtensa/lib/bswapdi2.S
new file mode 100644
index 000000000..5d94a9352
--- /dev/null
+++ b/arch/xtensa/lib/bswapdi2.S
@@ -0,0 +1,22 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__bswapdi2)
+
+	abi_entry_default
+	ssai	8
+	srli	a4, a2, 16
+	src	a4, a4, a2
+	src	a4, a4, a4
+	src	a4, a2, a4
+	srli	a2, a3, 16
+	src	a2, a2, a3
+	src	a2, a2, a2
+	src	a2, a3, a2
+	mov	a3, a4
+	abi_ret_default
+
+ENDPROC(__bswapdi2)
+EXPORT_SYMBOL(__bswapdi2)
diff --git a/arch/xtensa/lib/bswapsi2.S b/arch/xtensa/lib/bswapsi2.S
new file mode 100644
index 000000000..fbfb8613d
--- /dev/null
+++ b/arch/xtensa/lib/bswapsi2.S
@@ -0,0 +1,17 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__bswapsi2)
+
+	abi_entry_default
+	ssai	8
+	srli	a3, a2, 16
+	src	a3, a3, a2
+	src	a3, a3, a3
+	src	a2, a2, a3
+	abi_ret_default
+
+ENDPROC(__bswapsi2)
+EXPORT_SYMBOL(__bswapsi2)
diff --git a/arch/xtensa/lib/checksum.S b/arch/xtensa/lib/checksum.S
new file mode 100644
index 000000000..ffee6f94c
--- /dev/null
+++ b/arch/xtensa/lib/checksum.S
@@ -0,0 +1,359 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * INET		An implementation of the TCP/IP protocol suite for the LINUX
+ *		operating system.  INET is implemented using the  BSD Socket
+ *		interface as the means of communication with the user level.
+ *
+ *		IP/TCP/UDP checksumming routines
+ *
+ * Xtensa version:  Copyright (C) 2001 Tensilica, Inc. by Kevin Chea
+ *                  Optimized by Joe Taylor
+ */
+
+#include <linux/errno.h>
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+/*
+ * computes a partial checksum, e.g. for TCP/UDP fragments
+ */
+
+/*
+ * unsigned int csum_partial(const unsigned char *buf, int len,
+ *                           unsigned int sum);
+ *    a2 = buf
+ *    a3 = len
+ *    a4 = sum
+ *
+ * This function assumes 2- or 4-byte alignment.  Other alignments will fail!
+ */
+
+/* ONES_ADD converts twos-complement math to ones-complement. */
+#define ONES_ADD(sum, val)	  \
+	add	sum, sum, val	; \
+	bgeu	sum, val, 99f	; \
+	addi	sum, sum, 1	; \
+99:				;
+
+.text
+ENTRY(csum_partial)
+
+	/*
+	 * Experiments with Ethernet and SLIP connections show that buf
+	 * is aligned on either a 2-byte or 4-byte boundary.
+	 */
+	abi_entry_default
+	extui	a5, a2, 0, 2
+	bnez	a5, 8f		/* branch if 2-byte aligned */
+	/* Fall-through on common case, 4-byte alignment */
+1:
+	srli	a5, a3, 5	/* 32-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a5, 2f
+#else
+	beqz	a5, 2f
+	slli	a5, a5, 5
+	add	a5, a5, a2	/* a5 = end of last 32-byte chunk */
+.Loop1:
+#endif
+	l32i	a6, a2, 0
+	l32i	a7, a2, 4
+	ONES_ADD(a4, a6)
+	ONES_ADD(a4, a7)
+	l32i	a6, a2, 8
+	l32i	a7, a2, 12
+	ONES_ADD(a4, a6)
+	ONES_ADD(a4, a7)
+	l32i	a6, a2, 16
+	l32i	a7, a2, 20
+	ONES_ADD(a4, a6)
+	ONES_ADD(a4, a7)
+	l32i	a6, a2, 24
+	l32i	a7, a2, 28
+	ONES_ADD(a4, a6)
+	ONES_ADD(a4, a7)
+	addi	a2, a2, 4*8
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a5, .Loop1
+#endif
+2:
+	extui	a5, a3, 2, 3	/* remaining 4-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a5, 3f
+#else
+	beqz	a5, 3f
+	slli	a5, a5, 2
+	add	a5, a5, a2	/* a5 = end of last 4-byte chunk */
+.Loop2:
+#endif
+	l32i	a6, a2, 0
+	ONES_ADD(a4, a6)
+	addi	a2, a2, 4
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a5, .Loop2
+#endif
+3:
+	_bbci.l	a3, 1, 5f	/* remaining 2-byte chunk */
+	l16ui	a6, a2, 0
+	ONES_ADD(a4, a6)
+	addi	a2, a2, 2
+5:
+	_bbci.l	a3, 0, 7f	/* remaining 1-byte chunk */
+6:	l8ui	a6, a2, 0
+#ifdef __XTENSA_EB__
+	slli	a6, a6, 8	/* load byte into bits 8..15 */
+#endif
+	ONES_ADD(a4, a6)
+7:
+	mov	a2, a4
+	abi_ret_default
+
+	/* uncommon case, buf is 2-byte aligned */
+8:
+	beqz	a3, 7b		/* branch if len == 0 */
+	beqi	a3, 1, 6b	/* branch if len == 1 */
+
+	extui	a5, a2, 0, 1
+	bnez	a5, 8f		/* branch if 1-byte aligned */
+
+	l16ui	a6, a2, 0	/* common case, len >= 2 */
+	ONES_ADD(a4, a6)
+	addi	a2, a2, 2	/* adjust buf */
+	addi	a3, a3, -2	/* adjust len */
+	j	1b		/* now buf is 4-byte aligned */
+
+	/* case: odd-byte aligned, len > 1
+	 * This case is dog slow, so don't give us an odd address.
+	 * (I don't think this ever happens, but just in case.)
+	 */
+8:
+	srli	a5, a3, 2	/* 4-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a5, 2f
+#else
+	beqz	a5, 2f
+	slli	a5, a5, 2
+	add	a5, a5, a2	/* a5 = end of last 4-byte chunk */
+.Loop3:
+#endif
+	l8ui	a6, a2, 0	/* bits 24..31 */
+	l16ui	a7, a2, 1	/* bits  8..23 */
+	l8ui	a8, a2, 3	/* bits  0.. 8 */
+#ifdef	__XTENSA_EB__
+	slli	a6, a6, 24
+#else
+	slli	a8, a8, 24
+#endif
+	slli	a7, a7, 8
+	or	a7, a7, a6
+	or	a7, a7, a8
+	ONES_ADD(a4, a7)
+	addi	a2, a2, 4
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a5, .Loop3
+#endif
+2:
+	_bbci.l	a3, 1, 3f	/* remaining 2-byte chunk, still odd addr */
+	l8ui	a6, a2, 0
+	l8ui	a7, a2, 1
+#ifdef	__XTENSA_EB__
+	slli	a6, a6, 8
+#else
+	slli	a7, a7, 8
+#endif
+	or	a7, a7, a6
+	ONES_ADD(a4, a7)
+	addi	a2, a2, 2
+3:
+	j	5b		/* branch to handle the remaining byte */
+
+ENDPROC(csum_partial)
+EXPORT_SYMBOL(csum_partial)
+
+/*
+ * Copy from ds while checksumming, otherwise like csum_partial
+ */
+
+/*
+unsigned int csum_partial_copy_generic (const char *src, char *dst, int len)
+	a2  = src
+	a3  = dst
+	a4  = len
+	a5  = sum
+	a8  = temp
+	a9  = temp
+	a10 = temp
+
+    This function is optimized for 4-byte aligned addresses.  Other
+    alignments work, but not nearly as efficiently.
+ */
+
+ENTRY(csum_partial_copy_generic)
+
+	abi_entry_default
+	movi	a5, -1
+	or	a10, a2, a3
+
+	/* We optimize the following alignment tests for the 4-byte
+	aligned case.  Two bbsi.l instructions might seem more optimal
+	(commented out below).  However, both labels 5: and 3: are out
+	of the imm8 range, so the assembler relaxes them into
+	equivalent bbci.l, j combinations, which is actually
+	slower. */
+
+	extui	a9, a10, 0, 2
+	beqz	a9, 1f		/* branch if both are 4-byte aligned */
+	bbsi.l	a10, 0, 5f	/* branch if one address is odd */
+	j	3f		/* one address is 2-byte aligned */
+
+/*	_bbsi.l	a10, 0, 5f */	/* branch if odd address */
+/*	_bbsi.l	a10, 1, 3f */	/* branch if 2-byte-aligned address */
+
+1:
+	/* src and dst are both 4-byte aligned */
+	srli	a10, a4, 5	/* 32-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a10, 2f
+#else
+	beqz	a10, 2f
+	slli	a10, a10, 5
+	add	a10, a10, a2	/* a10 = end of last 32-byte src chunk */
+.Loop5:
+#endif
+EX(10f)	l32i	a9, a2, 0
+EX(10f)	l32i	a8, a2, 4
+EX(10f)	s32i	a9, a3, 0
+EX(10f)	s32i	a8, a3, 4
+	ONES_ADD(a5, a9)
+	ONES_ADD(a5, a8)
+EX(10f)	l32i	a9, a2, 8
+EX(10f)	l32i	a8, a2, 12
+EX(10f)	s32i	a9, a3, 8
+EX(10f)	s32i	a8, a3, 12
+	ONES_ADD(a5, a9)
+	ONES_ADD(a5, a8)
+EX(10f)	l32i	a9, a2, 16
+EX(10f)	l32i	a8, a2, 20
+EX(10f)	s32i	a9, a3, 16
+EX(10f)	s32i	a8, a3, 20
+	ONES_ADD(a5, a9)
+	ONES_ADD(a5, a8)
+EX(10f)	l32i	a9, a2, 24
+EX(10f)	l32i	a8, a2, 28
+EX(10f)	s32i	a9, a3, 24
+EX(10f)	s32i	a8, a3, 28
+	ONES_ADD(a5, a9)
+	ONES_ADD(a5, a8)
+	addi	a2, a2, 32
+	addi	a3, a3, 32
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a10, .Loop5
+#endif
+2:
+	extui	a10, a4, 2, 3	/* remaining 4-byte chunks */
+	extui	a4, a4, 0, 2	/* reset len for general-case, 2-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a10, 3f
+#else
+	beqz	a10, 3f
+	slli	a10, a10, 2
+	add	a10, a10, a2	/* a10 = end of last 4-byte src chunk */
+.Loop6:
+#endif
+EX(10f)	l32i	a9, a2, 0
+EX(10f)	s32i	a9, a3, 0
+	ONES_ADD(a5, a9)
+	addi	a2, a2, 4
+	addi	a3, a3, 4
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a10, .Loop6
+#endif
+3:
+	/*
+	Control comes to here in two cases: (1) It may fall through
+	to here from the 4-byte alignment case to process, at most,
+	one 2-byte chunk.  (2) It branches to here from above if
+	either src or dst is 2-byte aligned, and we process all bytes
+	here, except for perhaps a trailing odd byte.  It's
+	inefficient, so align your addresses to 4-byte boundaries.
+
+	a2 = src
+	a3 = dst
+	a4 = len
+	a5 = sum
+	*/
+	srli	a10, a4, 1	/* 2-byte chunks */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a10, 4f
+#else
+	beqz	a10, 4f
+	slli	a10, a10, 1
+	add	a10, a10, a2	/* a10 = end of last 2-byte src chunk */
+.Loop7:
+#endif
+EX(10f)	l16ui	a9, a2, 0
+EX(10f)	s16i	a9, a3, 0
+	ONES_ADD(a5, a9)
+	addi	a2, a2, 2
+	addi	a3, a3, 2
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a10, .Loop7
+#endif
+4:
+	/* This section processes a possible trailing odd byte. */
+	_bbci.l	a4, 0, 8f	/* 1-byte chunk */
+EX(10f)	l8ui	a9, a2, 0
+EX(10f)	s8i	a9, a3, 0
+#ifdef __XTENSA_EB__
+	slli	a9, a9, 8	/* shift byte to bits 8..15 */
+#endif
+	ONES_ADD(a5, a9)
+8:
+	mov	a2, a5
+	abi_ret_default
+
+5:
+	/* Control branch to here when either src or dst is odd.  We
+	process all bytes using 8-bit accesses.  Grossly inefficient,
+	so don't feed us an odd address. */
+
+	srli	a10, a4, 1	/* handle in pairs for 16-bit csum */
+#if XCHAL_HAVE_LOOPS
+	loopgtz	a10, 6f
+#else
+	beqz	a10, 6f
+	slli	a10, a10, 1
+	add	a10, a10, a2	/* a10 = end of last odd-aligned, 2-byte src chunk */
+.Loop8:
+#endif
+EX(10f)	l8ui	a9, a2, 0
+EX(10f)	l8ui	a8, a2, 1
+EX(10f)	s8i	a9, a3, 0
+EX(10f)	s8i	a8, a3, 1
+#ifdef __XTENSA_EB__
+	slli	a9, a9, 8	/* combine into a single 16-bit value */
+#else				/* for checksum computation */
+	slli	a8, a8, 8
+#endif
+	or	a9, a9, a8
+	ONES_ADD(a5, a9)
+	addi	a2, a2, 2
+	addi	a3, a3, 2
+#if !XCHAL_HAVE_LOOPS
+	blt	a2, a10, .Loop8
+#endif
+6:
+	j	4b		/* process the possible trailing odd byte */
+
+ENDPROC(csum_partial_copy_generic)
+EXPORT_SYMBOL(csum_partial_copy_generic)
+
+
+# Exception handler:
+.section .fixup, "ax"
+10:
+	movi	a2, 0
+	abi_ret_default
+
+.previous
diff --git a/arch/xtensa/lib/divsi3.S b/arch/xtensa/lib/divsi3.S
new file mode 100644
index 000000000..edb3c4ad9
--- /dev/null
+++ b/arch/xtensa/lib/divsi3.S
@@ -0,0 +1,75 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__divsi3)
+
+	abi_entry_default
+#if XCHAL_HAVE_DIV32
+	quos	a2, a2, a3
+#else
+	xor	a7, a2, a3	/* sign = dividend ^ divisor */
+	do_abs	a6, a2, a4	/* udividend = abs (dividend) */
+	do_abs	a3, a3, a4	/* udivisor = abs (divisor) */
+	bltui	a3, 2, .Lle_one	/* check if udivisor <= 1 */
+	do_nsau	a5, a6, a2, a8	/* udividend_shift = nsau (udividend) */
+	do_nsau	a4, a3, a2, a8	/* udivisor_shift = nsau (udivisor) */
+	bgeu	a5, a4, .Lspecial
+
+	sub	a4, a4, a5	/* count = udivisor_shift - udividend_shift */
+	ssl	a4
+	sll	a3, a3		/* udivisor <<= count */
+	movi	a2, 0		/* quotient = 0 */
+
+	/* test-subtract-and-shift loop; one quotient bit on each iteration */
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lloopend
+#endif /* XCHAL_HAVE_LOOPS */
+.Lloop:
+	bltu	a6, a3, .Lzerobit
+	sub	a6, a6, a3
+	addi	a2, a2, 1
+.Lzerobit:
+	slli	a2, a2, 1
+	srli	a3, a3, 1
+#if !XCHAL_HAVE_LOOPS
+	addi	a4, a4, -1
+	bnez	a4, .Lloop
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lloopend:
+
+	bltu	a6, a3, .Lreturn
+	addi	a2, a2, 1	/* increment if udividend >= udivisor */
+.Lreturn:
+	neg	a5, a2
+	movltz	a2, a5, a7	/* return (sign < 0) ? -quotient : quotient */
+	abi_ret_default
+
+.Lle_one:
+	beqz	a3, .Lerror
+	neg	a2, a6		/* if udivisor == 1, then return... */
+	movgez	a2, a6, a7	/* (sign < 0) ? -udividend : udividend */
+	abi_ret_default
+
+.Lspecial:
+	bltu	a6, a3, .Lreturn0 /* if dividend < divisor, return 0 */
+	movi	a2, 1
+	movi	a4, -1
+	movltz	a2, a4, a7	/* else return (sign < 0) ? -1 : 1 */
+	abi_ret_default
+
+.Lerror:
+	/* Divide by zero: Use an illegal instruction to force an exception.
+	   The subsequent "DIV0" string can be recognized by the exception
+	   handler to identify the real cause of the exception.  */
+	ill
+	.ascii	"DIV0"
+
+.Lreturn0:
+	movi	a2, 0
+#endif /* XCHAL_HAVE_DIV32 */
+	abi_ret_default
+
+ENDPROC(__divsi3)
+EXPORT_SYMBOL(__divsi3)
diff --git a/arch/xtensa/lib/lshrdi3.S b/arch/xtensa/lib/lshrdi3.S
new file mode 100644
index 000000000..e432e1a40
--- /dev/null
+++ b/arch/xtensa/lib/lshrdi3.S
@@ -0,0 +1,29 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+#ifdef __XTENSA_EB__
+#define uh a2
+#define ul a3
+#else
+#define uh a3
+#define ul a2
+#endif /* __XTENSA_EB__ */
+
+ENTRY(__lshrdi3)
+
+	abi_entry_default
+	ssr	a4
+	bgei	a4, 32, .Lhigh_only
+	src	ul, uh, ul
+	srl	uh, uh
+	abi_ret_default
+
+.Lhigh_only:
+	srl	ul, uh
+	movi	uh, 0
+	abi_ret_default
+
+ENDPROC(__lshrdi3)
+EXPORT_SYMBOL(__lshrdi3)
diff --git a/arch/xtensa/lib/memcopy.S b/arch/xtensa/lib/memcopy.S
new file mode 100644
index 000000000..f60760396
--- /dev/null
+++ b/arch/xtensa/lib/memcopy.S
@@ -0,0 +1,542 @@
+/*
+ * arch/xtensa/lib/hal/memcopy.S -- Core HAL library functions
+ * xthal_memcpy and xthal_bcopy
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2002 - 2012 Tensilica Inc.
+ */
+
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+/*
+ * void *memcpy(void *dst, const void *src, size_t len);
+ *
+ * This function is intended to do the same thing as the standard
+ * library function memcpy() for most cases.
+ * However, where the source and/or destination references
+ * an instruction RAM or ROM or a data RAM or ROM, that
+ * source and/or destination will always be accessed with
+ * 32-bit load and store instructions (as required for these
+ * types of devices).
+ *
+ * !!!!!!!  XTFIXME:
+ * !!!!!!!  Handling of IRAM/IROM has not yet
+ * !!!!!!!  been implemented.
+ *
+ * The (general case) algorithm is as follows:
+ *   If destination is unaligned, align it by conditionally
+ *     copying 1 and 2 bytes.
+ *   If source is aligned,
+ *     do 16 bytes with a loop, and then finish up with
+ *     8, 4, 2, and 1 byte copies conditional on the length;
+ *   else (if source is unaligned),
+ *     do the same, but use SRC to align the source data.
+ *   This code tries to use fall-through branches for the common
+ *     case of aligned source and destination and multiple
+ *     of 4 (or 8) length.
+ *
+ * Register use:
+ *	a0/ return address
+ *	a1/ stack pointer
+ *	a2/ return value
+ *	a3/ src
+ *	a4/ length
+ *	a5/ dst
+ *	a6/ tmp
+ *	a7/ tmp
+ *	a8/ tmp
+ *	a9/ tmp
+ *	a10/ tmp
+ *	a11/ tmp
+ */
+
+	.text
+
+/*
+ * Byte by byte copy
+ */
+	.align	4
+	.byte	0		# 1 mod 4 alignment for LOOPNEZ
+				# (0 mod 4 alignment for LBEG)
+.Lbytecopy:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lbytecopydone
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a4, .Lbytecopydone
+	add	a7, a3, a4	# a7 = end address for source
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lnextbyte:
+	l8ui	a6, a3, 0
+	addi	a3, a3, 1
+	s8i	a6, a5, 0
+	addi	a5, a5, 1
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a7, .Lnextbyte # continue loop if $a3:src != $a7:src_end
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbytecopydone:
+	abi_ret_default
+
+/*
+ * Destination is unaligned
+ */
+
+	.align	4
+.Ldst1mod2:	# dst is only byte aligned
+	_bltui	a4, 7, .Lbytecopy	# do short copies byte by byte
+
+	# copy 1 byte
+	l8ui	a6, a3,  0
+	addi	a3, a3,  1
+	addi	a4, a4, -1
+	s8i	a6, a5,  0
+	addi	a5, a5,  1
+	_bbci.l	a5, 1, .Ldstaligned	# if dst is now aligned, then
+					# return to main algorithm
+.Ldst2mod4:	# dst 16-bit aligned
+	# copy 2 bytes
+	_bltui	a4, 6, .Lbytecopy	# do short copies byte by byte
+	l8ui	a6, a3,  0
+	l8ui	a7, a3,  1
+	addi	a3, a3,  2
+	addi	a4, a4, -2
+	s8i	a6, a5,  0
+	s8i	a7, a5,  1
+	addi	a5, a5,  2
+	j	.Ldstaligned	# dst is now aligned, return to main algorithm
+
+ENTRY(__memcpy)
+WEAK(memcpy)
+
+	abi_entry_default
+	# a2/ dst, a3/ src, a4/ len
+	mov	a5, a2		# copy dst so that a2 is return value
+.Lcommon:
+	_bbsi.l	a2, 0, .Ldst1mod2	# if dst is 1 mod 2
+	_bbsi.l	a2, 1, .Ldst2mod4	# if dst is 2 mod 4
+.Ldstaligned:	# return here from .Ldst?mod? once dst is aligned
+	srli	a7, a4, 4	# number of loop iterations with 16B
+				# per iteration
+	movi	a8, 3		# if source is not aligned,
+	_bany	a3, a8, .Lsrcunaligned	# then use shifting copy
+	/*
+	 * Destination and source are word-aligned, use word copy.
+	 */
+	# copy 16 bytes per iteration for word-aligned dst and word-aligned src
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .Loop1done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .Loop1done
+	slli	a8, a7, 4
+	add	a8, a8, a3	# a8 = end of last 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1:
+	l32i	a6, a3,  0
+	l32i	a7, a3,  4
+	s32i	a6, a5,  0
+	l32i	a6, a3,  8
+	s32i	a7, a5,  4
+	l32i	a7, a3, 12
+	s32i	a6, a5,  8
+	addi	a3, a3, 16
+	s32i	a7, a5, 12
+	addi	a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a8, .Loop1  # continue loop if a3:src != a8:src_end
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1done:
+	bbci.l	a4, 3, .L2
+	# copy 8 bytes
+	l32i	a6, a3,  0
+	l32i	a7, a3,  4
+	addi	a3, a3,  8
+	s32i	a6, a5,  0
+	s32i	a7, a5,  4
+	addi	a5, a5,  8
+.L2:
+	bbsi.l	a4, 2, .L3
+	bbsi.l	a4, 1, .L4
+	bbsi.l	a4, 0, .L5
+	abi_ret_default
+.L3:
+	# copy 4 bytes
+	l32i	a6, a3,  0
+	addi	a3, a3,  4
+	s32i	a6, a5,  0
+	addi	a5, a5,  4
+	bbsi.l	a4, 1, .L4
+	bbsi.l	a4, 0, .L5
+	abi_ret_default
+.L4:
+	# copy 2 bytes
+	l16ui	a6, a3,  0
+	addi	a3, a3,  2
+	s16i	a6, a5,  0
+	addi	a5, a5,  2
+	bbsi.l	a4, 0, .L5
+	abi_ret_default
+.L5:
+	# copy 1 byte
+	l8ui	a6, a3,  0
+	s8i	a6, a5,  0
+	abi_ret_default
+
+/*
+ * Destination is aligned, Source is unaligned
+ */
+
+	.align	4
+.Lsrcunaligned:
+	_beqz	a4, .Ldone	# avoid loading anything for zero-length copies
+	# copy 16 bytes per iteration for word-aligned dst and unaligned src
+	__ssa8	a3		# set shift amount from byte offset
+
+/* set to 1 when running on ISS (simulator) with the
+   lint or ferret client, or 0 to save a few cycles */
+#define SIM_CHECKS_ALIGNMENT	1
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT
+	and	a11, a3, a8	# save unalignment offset for below
+	sub	a3, a3, a11	# align a3
+#endif
+	l32i	a6, a3, 0	# load first word
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .Loop2done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .Loop2done
+	slli	a10, a7, 4
+	add	a10, a10, a3	# a10 = end of last 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2:
+	l32i	a7, a3,  4
+	l32i	a8, a3,  8
+	__src_b	a6, a6, a7
+	s32i	a6, a5,  0
+	l32i	a9, a3, 12
+	__src_b	a7, a7, a8
+	s32i	a7, a5,  4
+	l32i	a6, a3, 16
+	__src_b	a8, a8, a9
+	s32i	a8, a5,  8
+	addi	a3, a3, 16
+	__src_b	a9, a9, a6
+	s32i	a9, a5, 12
+	addi	a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a10, .Loop2 # continue loop if a3:src != a10:src_end
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2done:
+	bbci.l	a4, 3, .L12
+	# copy 8 bytes
+	l32i	a7, a3,  4
+	l32i	a8, a3,  8
+	__src_b	a6, a6, a7
+	s32i	a6, a5,  0
+	addi	a3, a3,  8
+	__src_b	a7, a7, a8
+	s32i	a7, a5,  4
+	addi	a5, a5,  8
+	mov	a6, a8
+.L12:
+	bbci.l	a4, 2, .L13
+	# copy 4 bytes
+	l32i	a7, a3,  4
+	addi	a3, a3,  4
+	__src_b	a6, a6, a7
+	s32i	a6, a5,  0
+	addi	a5, a5,  4
+	mov	a6, a7
+.L13:
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT
+	add	a3, a3, a11	# readjust a3 with correct misalignment
+#endif
+	bbsi.l	a4, 1, .L14
+	bbsi.l	a4, 0, .L15
+.Ldone:	abi_ret_default
+.L14:
+	# copy 2 bytes
+	l8ui	a6, a3,  0
+	l8ui	a7, a3,  1
+	addi	a3, a3,  2
+	s8i	a6, a5,  0
+	s8i	a7, a5,  1
+	addi	a5, a5,  2
+	bbsi.l	a4, 0, .L15
+	abi_ret_default
+.L15:
+	# copy 1 byte
+	l8ui	a6, a3,  0
+	s8i	a6, a5,  0
+	abi_ret_default
+
+ENDPROC(__memcpy)
+EXPORT_SYMBOL(__memcpy)
+EXPORT_SYMBOL(memcpy)
+
+/*
+ * void *memmove(void *dst, const void *src, size_t len);
+ *
+ * This function is intended to do the same thing as the standard
+ * library function memmove() for most cases.
+ * However, where the source and/or destination references
+ * an instruction RAM or ROM or a data RAM or ROM, that
+ * source and/or destination will always be accessed with
+ * 32-bit load and store instructions (as required for these
+ * types of devices).
+ *
+ * !!!!!!!  XTFIXME:
+ * !!!!!!!  Handling of IRAM/IROM has not yet
+ * !!!!!!!  been implemented.
+ *
+ * The (general case) algorithm is as follows:
+ *   If end of source doesn't overlap destination then use memcpy.
+ *   Otherwise do memcpy backwards.
+ *
+ * Register use:
+ *	a0/ return address
+ *	a1/ stack pointer
+ *	a2/ return value
+ *	a3/ src
+ *	a4/ length
+ *	a5/ dst
+ *	a6/ tmp
+ *	a7/ tmp
+ *	a8/ tmp
+ *	a9/ tmp
+ *	a10/ tmp
+ *	a11/ tmp
+ */
+
+/*
+ * Byte by byte copy
+ */
+	.align	4
+	.byte	0		# 1 mod 4 alignment for LOOPNEZ
+				# (0 mod 4 alignment for LBEG)
+.Lbackbytecopy:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lbackbytecopydone
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a4, .Lbackbytecopydone
+	sub	a7, a3, a4	# a7 = start address for source
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbacknextbyte:
+	addi	a3, a3, -1
+	l8ui	a6, a3, 0
+	addi	a5, a5, -1
+	s8i	a6, a5, 0
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a7, .Lbacknextbyte # continue loop if
+				       # $a3:src != $a7:src_start
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbackbytecopydone:
+	abi_ret_default
+
+/*
+ * Destination is unaligned
+ */
+
+	.align	4
+.Lbackdst1mod2:	# dst is only byte aligned
+	_bltui	a4, 7, .Lbackbytecopy	# do short copies byte by byte
+
+	# copy 1 byte
+	addi	a3, a3, -1
+	l8ui	a6, a3,  0
+	addi	a5, a5, -1
+	s8i	a6, a5,  0
+	addi	a4, a4, -1
+	_bbci.l	a5, 1, .Lbackdstaligned	# if dst is now aligned, then
+					# return to main algorithm
+.Lbackdst2mod4:	# dst 16-bit aligned
+	# copy 2 bytes
+	_bltui	a4, 6, .Lbackbytecopy	# do short copies byte by byte
+	addi	a3, a3, -2
+	l8ui	a6, a3,  0
+	l8ui	a7, a3,  1
+	addi	a5, a5, -2
+	s8i	a6, a5,  0
+	s8i	a7, a5,  1
+	addi	a4, a4, -2
+	j	.Lbackdstaligned	# dst is now aligned,
+					# return to main algorithm
+
+ENTRY(__memmove)
+WEAK(memmove)
+
+	abi_entry_default
+	# a2/ dst, a3/ src, a4/ len
+	mov	a5, a2		# copy dst so that a2 is return value
+.Lmovecommon:
+	sub	a6, a5, a3
+	bgeu	a6, a4, .Lcommon
+
+	add	a5, a5, a4
+	add	a3, a3, a4
+
+	_bbsi.l	a5, 0, .Lbackdst1mod2	# if dst is 1 mod 2
+	_bbsi.l	a5, 1, .Lbackdst2mod4	# if dst is 2 mod 4
+.Lbackdstaligned:	# return here from .Lbackdst?mod? once dst is aligned
+	srli	a7, a4, 4	# number of loop iterations with 16B
+				# per iteration
+	movi	a8, 3		# if source is not aligned,
+	_bany	a3, a8, .Lbacksrcunaligned	# then use shifting copy
+	/*
+	 * Destination and source are word-aligned, use word copy.
+	 */
+	# copy 16 bytes per iteration for word-aligned dst and word-aligned src
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .LbackLoop1done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .LbackLoop1done
+	slli	a8, a7, 4
+	sub	a8, a3, a8	# a8 = start of first 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.LbackLoop1:
+	addi	a3, a3, -16
+	l32i	a7, a3, 12
+	l32i	a6, a3,  8
+	addi	a5, a5, -16
+	s32i	a7, a5, 12
+	l32i	a7, a3,  4
+	s32i	a6, a5,  8
+	l32i	a6, a3,  0
+	s32i	a7, a5,  4
+	s32i	a6, a5,  0
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a8, .LbackLoop1  # continue loop if a3:src != a8:src_start
+#endif /* !XCHAL_HAVE_LOOPS */
+.LbackLoop1done:
+	bbci.l	a4, 3, .Lback2
+	# copy 8 bytes
+	addi	a3, a3, -8
+	l32i	a6, a3,  0
+	l32i	a7, a3,  4
+	addi	a5, a5, -8
+	s32i	a6, a5,  0
+	s32i	a7, a5,  4
+.Lback2:
+	bbsi.l	a4, 2, .Lback3
+	bbsi.l	a4, 1, .Lback4
+	bbsi.l	a4, 0, .Lback5
+	abi_ret_default
+.Lback3:
+	# copy 4 bytes
+	addi	a3, a3, -4
+	l32i	a6, a3,  0
+	addi	a5, a5, -4
+	s32i	a6, a5,  0
+	bbsi.l	a4, 1, .Lback4
+	bbsi.l	a4, 0, .Lback5
+	abi_ret_default
+.Lback4:
+	# copy 2 bytes
+	addi	a3, a3, -2
+	l16ui	a6, a3,  0
+	addi	a5, a5, -2
+	s16i	a6, a5,  0
+	bbsi.l	a4, 0, .Lback5
+	abi_ret_default
+.Lback5:
+	# copy 1 byte
+	addi	a3, a3, -1
+	l8ui	a6, a3,  0
+	addi	a5, a5, -1
+	s8i	a6, a5,  0
+	abi_ret_default
+
+/*
+ * Destination is aligned, Source is unaligned
+ */
+
+	.align	4
+.Lbacksrcunaligned:
+	_beqz	a4, .Lbackdone	# avoid loading anything for zero-length copies
+	# copy 16 bytes per iteration for word-aligned dst and unaligned src
+	__ssa8	a3		# set shift amount from byte offset
+#define SIM_CHECKS_ALIGNMENT	1	/* set to 1 when running on ISS with
+					 * the lint or ferret client, or 0
+					 * to save a few cycles */
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT
+	and	a11, a3, a8	# save unalignment offset for below
+	sub	a3, a3, a11	# align a3
+#endif
+	l32i	a6, a3, 0	# load first word
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .LbackLoop2done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .LbackLoop2done
+	slli	a10, a7, 4
+	sub	a10, a3, a10	# a10 = start of first 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.LbackLoop2:
+	addi	a3, a3, -16
+	l32i	a7, a3, 12
+	l32i	a8, a3,  8
+	addi	a5, a5, -16
+	__src_b	a6, a7, a6
+	s32i	a6, a5, 12
+	l32i	a9, a3,  4
+	__src_b	a7, a8, a7
+	s32i	a7, a5,  8
+	l32i	a6, a3,  0
+	__src_b	a8, a9, a8
+	s32i	a8, a5,  4
+	__src_b	a9, a6, a9
+	s32i	a9, a5,  0
+#if !XCHAL_HAVE_LOOPS
+	bne	a3, a10, .LbackLoop2 # continue loop if a3:src != a10:src_start
+#endif /* !XCHAL_HAVE_LOOPS */
+.LbackLoop2done:
+	bbci.l	a4, 3, .Lback12
+	# copy 8 bytes
+	addi	a3, a3, -8
+	l32i	a7, a3,  4
+	l32i	a8, a3,  0
+	addi	a5, a5, -8
+	__src_b	a6, a7, a6
+	s32i	a6, a5,  4
+	__src_b	a7, a8, a7
+	s32i	a7, a5,  0
+	mov	a6, a8
+.Lback12:
+	bbci.l	a4, 2, .Lback13
+	# copy 4 bytes
+	addi	a3, a3, -4
+	l32i	a7, a3,  0
+	addi	a5, a5, -4
+	__src_b	a6, a7, a6
+	s32i	a6, a5,  0
+	mov	a6, a7
+.Lback13:
+#if XCHAL_UNALIGNED_LOAD_EXCEPTION || SIM_CHECKS_ALIGNMENT
+	add	a3, a3, a11	# readjust a3 with correct misalignment
+#endif
+	bbsi.l	a4, 1, .Lback14
+	bbsi.l	a4, 0, .Lback15
+.Lbackdone:
+	abi_ret_default
+.Lback14:
+	# copy 2 bytes
+	addi	a3, a3, -2
+	l8ui	a6, a3,  0
+	l8ui	a7, a3,  1
+	addi	a5, a5, -2
+	s8i	a6, a5,  0
+	s8i	a7, a5,  1
+	bbsi.l	a4, 0, .Lback15
+	abi_ret_default
+.Lback15:
+	# copy 1 byte
+	addi	a3, a3, -1
+	addi	a5, a5, -1
+	l8ui	a6, a3,  0
+	s8i	a6, a5,  0
+	abi_ret_default
+
+ENDPROC(__memmove)
+EXPORT_SYMBOL(__memmove)
+EXPORT_SYMBOL(memmove)
diff --git a/arch/xtensa/lib/memset.S b/arch/xtensa/lib/memset.S
new file mode 100644
index 000000000..262c3f39f
--- /dev/null
+++ b/arch/xtensa/lib/memset.S
@@ -0,0 +1,155 @@
+/*
+ *  arch/xtensa/lib/memset.S
+ *
+ *  ANSI C standard library function memset
+ *  (Well, almost.  .fixup code might return zero.)
+ *
+ *  This file is subject to the terms and conditions of the GNU General
+ *  Public License.  See the file "COPYING" in the main directory of
+ *  this archive for more details.
+ *
+ *  Copyright (C) 2002 Tensilica Inc.
+ */
+
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+/*
+ * void *memset(void *dst, int c, size_t length)
+ *
+ * The algorithm is as follows:
+ *   Create a word with c in all byte positions
+ *   If the destination is aligned,
+ *     do 16B chucks with a loop, and then finish up with
+ *     8B, 4B, 2B, and 1B stores conditional on the length.
+ *   If destination is unaligned, align it by conditionally
+ *     setting 1B and 2B and then go to aligned case.
+ *   This code tries to use fall-through branches for the common
+ *     case of an aligned destination (except for the branches to
+ *     the alignment labels).
+ */
+
+.text
+ENTRY(__memset)
+WEAK(memset)
+
+	abi_entry_default
+	# a2/ dst, a3/ c, a4/ length
+	extui	a3, a3, 0, 8	# mask to just 8 bits
+	slli	a7, a3, 8	# duplicate character in all bytes of word
+	or	a3, a3, a7	# ...
+	slli	a7, a3, 16	# ...
+	or	a3, a3, a7	# ...
+	mov	a5, a2		# copy dst so that a2 is return value
+	movi	a6, 3		# for alignment tests
+	bany	a2, a6, .Ldstunaligned # if dst is unaligned
+.L0:	# return here from .Ldstunaligned when dst is aligned
+	srli	a7, a4, 4	# number of loop iterations with 16B
+				# per iteration
+	bnez	a4, .Laligned
+	abi_ret_default
+
+/*
+ * Destination is word-aligned.
+ */
+	# set 16 bytes per iteration for word-aligned dst
+	.align	4		# 1 mod 4 alignment for LOOPNEZ
+	.byte	0		# (0 mod 4 alignment for LBEG)
+.Laligned:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .Loop1done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .Loop1done
+	slli	a6, a7, 4
+	add	a6, a6, a5	# a6 = end of last 16B chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1:
+EX(10f) s32i	a3, a5,  0
+EX(10f) s32i	a3, a5,  4
+EX(10f) s32i	a3, a5,  8
+EX(10f) s32i	a3, a5, 12
+	addi	a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+	blt	a5, a6, .Loop1
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1done:
+	bbci.l	a4, 3, .L2
+	# set 8 bytes
+EX(10f) s32i	a3, a5,  0
+EX(10f) s32i	a3, a5,  4
+	addi	a5, a5,  8
+.L2:
+	bbci.l	a4, 2, .L3
+	# set 4 bytes
+EX(10f) s32i	a3, a5,  0
+	addi	a5, a5,  4
+.L3:
+	bbci.l	a4, 1, .L4
+	# set 2 bytes
+EX(10f) s16i	a3, a5,  0
+	addi	a5, a5,  2
+.L4:
+	bbci.l	a4, 0, .L5
+	# set 1 byte
+EX(10f) s8i	a3, a5,  0
+.L5:
+.Lret1:
+	abi_ret_default
+
+/*
+ * Destination is unaligned
+ */
+
+.Ldstunaligned:
+	bltui	a4, 8, .Lbyteset	# do short copies byte by byte
+	bbci.l	a5, 0, .L20		# branch if dst alignment half-aligned
+	# dst is only byte aligned
+	# set 1 byte
+EX(10f) s8i	a3, a5,  0
+	addi	a5, a5,  1
+	addi	a4, a4, -1
+	# now retest if dst aligned
+	bbci.l	a5, 1, .L0	# if now aligned, return to main algorithm
+.L20:
+	# dst half-aligned
+	# set 2 bytes
+EX(10f) s16i	a3, a5,  0
+	addi	a5, a5,  2
+	addi	a4, a4, -2
+	j	.L0		# dst is now aligned, return to main algorithm
+
+/*
+ * Byte by byte set
+ */
+	.align	4
+	.byte	0		# 1 mod 4 alignment for LOOPNEZ
+				# (0 mod 4 alignment for LBEG)
+.Lbyteset:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lbytesetdone
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a4, .Lbytesetdone
+	add	a6, a5, a4	# a6 = ending address
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbyteloop:
+EX(10f) s8i	a3, a5, 0
+	addi	a5, a5, 1
+#if !XCHAL_HAVE_LOOPS
+	blt	a5, a6, .Lbyteloop
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbytesetdone:
+	abi_ret_default
+
+ENDPROC(__memset)
+EXPORT_SYMBOL(__memset)
+EXPORT_SYMBOL(memset)
+
+	.section .fixup, "ax"
+	.align	4
+
+/* We return zero if a failure occurred. */
+
+10:
+	movi	a2, 0
+	abi_ret_default
diff --git a/arch/xtensa/lib/modsi3.S b/arch/xtensa/lib/modsi3.S
new file mode 100644
index 000000000..c5f4295c6
--- /dev/null
+++ b/arch/xtensa/lib/modsi3.S
@@ -0,0 +1,88 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__modsi3)
+
+	abi_entry_default
+#if XCHAL_HAVE_DIV32
+	rems	a2, a2, a3
+#else
+	mov	a7, a2		/* save original (signed) dividend */
+	do_abs	a2, a2, a4	/* udividend = abs (dividend) */
+	do_abs	a3, a3, a4	/* udivisor = abs (divisor) */
+	bltui	a3, 2, .Lle_one	/* check if udivisor <= 1 */
+	do_nsau	a5, a2, a6, a8	/* udividend_shift = nsau (udividend) */
+	do_nsau	a4, a3, a6, a8	/* udivisor_shift = nsau (udivisor) */
+	bgeu	a5, a4, .Lspecial
+
+	sub	a4, a4, a5	/* count = udivisor_shift - udividend_shift */
+	ssl	a4
+	sll	a3, a3		/* udivisor <<= count */
+
+	/* test-subtract-and-shift loop */
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lloopend
+#endif /* XCHAL_HAVE_LOOPS */
+.Lloop:
+	bltu	a2, a3, .Lzerobit
+	sub	a2, a2, a3
+.Lzerobit:
+	srli	a3, a3, 1
+#if !XCHAL_HAVE_LOOPS
+	addi	a4, a4, -1
+	bnez	a4, .Lloop
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lloopend:
+
+.Lspecial:
+	bltu	a2, a3, .Lreturn
+	sub	a2, a2, a3	/* subtract again if udividend >= udivisor */
+.Lreturn:
+	bgez	a7, .Lpositive
+	neg	a2, a2		/* if (dividend < 0), return -udividend */
+.Lpositive:
+	abi_ret_default
+
+.Lle_one:
+	bnez	a3, .Lreturn0
+
+	/* Divide by zero: Use an illegal instruction to force an exception.
+	   The subsequent "DIV0" string can be recognized by the exception
+	   handler to identify the real cause of the exception.  */
+	ill
+	.ascii	"DIV0"
+
+.Lreturn0:
+	movi	a2, 0
+#endif /* XCHAL_HAVE_DIV32 */
+	abi_ret_default
+
+ENDPROC(__modsi3)
+EXPORT_SYMBOL(__modsi3)
+
+#if !XCHAL_HAVE_NSA
+	.section .rodata
+	.align	4
+	.global	__nsau_data
+	.type	__nsau_data, @object
+__nsau_data:
+	.byte	8, 7, 6, 6, 5, 5, 5, 5, 4, 4, 4, 4, 4, 4, 4, 4
+	.byte	3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3
+	.byte	2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2
+	.byte	2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2
+	.byte	1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
+	.byte	1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
+	.byte	1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
+	.byte	1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.byte	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
+	.size	__nsau_data, . - __nsau_data
+#endif /* !XCHAL_HAVE_NSA */
diff --git a/arch/xtensa/lib/mulsi3.S b/arch/xtensa/lib/mulsi3.S
new file mode 100644
index 000000000..c6b4fd46b
--- /dev/null
+++ b/arch/xtensa/lib/mulsi3.S
@@ -0,0 +1,134 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+	.macro	do_addx2 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx2	\dst, \as, \at
+#else
+	slli	\tmp, \as, 1
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+	.macro	do_addx4 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx4	\dst, \as, \at
+#else
+	slli	\tmp, \as, 2
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+	.macro	do_addx8 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx8	\dst, \as, \at
+#else
+	slli	\tmp, \as, 3
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+ENTRY(__mulsi3)
+
+	abi_entry_default
+
+#if XCHAL_HAVE_MUL32
+	mull	a2, a2, a3
+
+#elif XCHAL_HAVE_MUL16
+	or	a4, a2, a3
+	srai	a4, a4, 16
+	bnez	a4, .LMUL16
+	mul16u	a2, a2, a3
+	abi_ret_default
+.LMUL16:
+	srai	a4, a2, 16
+	srai	a5, a3, 16
+	mul16u	a7, a4, a3
+	mul16u	a6, a5, a2
+	mul16u	a4, a2, a3
+	add	a7, a7, a6
+	slli	a7, a7, 16
+	add	a2, a7, a4
+
+#elif XCHAL_HAVE_MAC16
+	mul.aa.hl a2, a3
+	mula.aa.lh a2, a3
+	rsr	a5, ACCLO
+	umul.aa.ll a2, a3
+	rsr	a4, ACCLO
+	slli	a5, a5, 16
+	add	a2, a4, a5
+
+#else /* !MUL32 && !MUL16 && !MAC16 */
+
+	/* Multiply one bit at a time, but unroll the loop 4x to better
+	   exploit the addx instructions and avoid overhead.
+	   Peel the first iteration to save a cycle on init.  */
+
+	/* Avoid negative numbers.  */
+	xor	a5, a2, a3	/* Top bit is 1 if one input is negative.  */
+	do_abs	a3, a3, a6
+	do_abs	a2, a2, a6
+
+	/* Swap so the second argument is smaller.  */
+	sub	a7, a2, a3
+	mov	a4, a3
+	movgez	a4, a2, a7	/* a4 = max (a2, a3) */
+	movltz	a3, a2, a7	/* a3 = min (a2, a3) */
+
+	movi	a2, 0
+	extui	a6, a3, 0, 1
+	movnez	a2, a4, a6
+
+	do_addx2 a7, a4, a2, a7
+	extui	a6, a3, 1, 1
+	movnez	a2, a7, a6
+
+	do_addx4 a7, a4, a2, a7
+	extui	a6, a3, 2, 1
+	movnez	a2, a7, a6
+
+	do_addx8 a7, a4, a2, a7
+	extui	a6, a3, 3, 1
+	movnez	a2, a7, a6
+
+	bgeui	a3, 16, .Lmult_main_loop
+	neg	a3, a2
+	movltz	a2, a3, a5
+	abi_ret_default
+
+	.align	4
+.Lmult_main_loop:
+	srli	a3, a3, 4
+	slli	a4, a4, 4
+
+	add	a7, a4, a2
+	extui	a6, a3, 0, 1
+	movnez	a2, a7, a6
+
+	do_addx2 a7, a4, a2, a7
+	extui	a6, a3, 1, 1
+	movnez	a2, a7, a6
+
+	do_addx4 a7, a4, a2, a7
+	extui	a6, a3, 2, 1
+	movnez	a2, a7, a6
+
+	do_addx8 a7, a4, a2, a7
+	extui	a6, a3, 3, 1
+	movnez	a2, a7, a6
+
+	bgeui	a3, 16, .Lmult_main_loop
+
+	neg	a3, a2
+	movltz	a2, a3, a5
+
+#endif /* !MUL32 && !MUL16 && !MAC16 */
+
+	abi_ret_default
+
+ENDPROC(__mulsi3)
+EXPORT_SYMBOL(__mulsi3)
diff --git a/arch/xtensa/lib/pci-auto.c b/arch/xtensa/lib/pci-auto.c
new file mode 100644
index 000000000..05fc02f9e
--- /dev/null
+++ b/arch/xtensa/lib/pci-auto.c
@@ -0,0 +1,318 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * arch/xtensa/lib/pci-auto.c
+ *
+ * PCI autoconfiguration library
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ *
+ * Chris Zankel <zankel@tensilica.com, cez@zankel.net>
+ *
+ * Based on work from Matt Porter <mporter@mvista.com>
+ */
+
+#include <linux/bitfield.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/pci.h>
+
+#include <asm/pci-bridge.h>
+
+
+/*
+ *
+ * Setting up a PCI
+ *
+ * pci_ctrl->first_busno = <first bus number (0)>
+ * pci_ctrl->last_busno = <last bus number (0xff)>
+ * pci_ctrl->ops = <PCI config operations>
+ * pci_ctrl->map_irq = <function to return the interrupt number for a device>
+ *
+ * pci_ctrl->io_space.start = <IO space start address (PCI view)>
+ * pci_ctrl->io_space.end = <IO space end address (PCI view)>
+ * pci_ctrl->io_space.base = <IO space offset: address 0 from CPU space>
+ * pci_ctrl->mem_space.start = <MEM space start address (PCI view)>
+ * pci_ctrl->mem_space.end = <MEM space end address (PCI view)>
+ * pci_ctrl->mem_space.base = <MEM space offset: address 0 from CPU space>
+ *
+ * pcibios_init_resource(&pci_ctrl->io_resource, <IO space start>,
+ * 			 <IO space end>, IORESOURCE_IO, "PCI host bridge");
+ * pcibios_init_resource(&pci_ctrl->mem_resources[0], <MEM space start>,
+ * 			 <MEM space end>, IORESOURCE_MEM, "PCI host bridge");
+ *
+ * pci_ctrl->last_busno = pciauto_bus_scan(pci_ctrl,pci_ctrl->first_busno);
+ *
+ * int __init pciauto_bus_scan(struct pci_controller *pci_ctrl, int current_bus)
+ *
+ */
+
+static int pciauto_upper_iospc;
+static int pciauto_upper_memspc;
+
+static struct pci_dev pciauto_dev;
+static struct pci_bus pciauto_bus;
+
+/*
+ * Helper functions
+ */
+
+/* Initialize the bars of a PCI device.  */
+
+static void __init
+pciauto_setup_bars(struct pci_dev *dev, int bar_limit)
+{
+	int bar_size;
+	int bar, bar_nr;
+	int *upper_limit;
+	int found_mem64 = 0;
+
+	for (bar = PCI_BASE_ADDRESS_0, bar_nr = 0;
+	     bar <= bar_limit;
+	     bar+=4, bar_nr++)
+	{
+		/* Tickle the BAR and get the size */
+		pci_write_config_dword(dev, bar, 0xffffffff);
+		pci_read_config_dword(dev, bar, &bar_size);
+
+		/* If BAR is not implemented go to the next BAR */
+		if (!bar_size)
+			continue;
+
+		/* Check the BAR type and set our address mask */
+		if (bar_size & PCI_BASE_ADDRESS_SPACE_IO)
+		{
+			bar_size &= PCI_BASE_ADDRESS_IO_MASK;
+			upper_limit = &pciauto_upper_iospc;
+			pr_debug("PCI Autoconfig: BAR %d, I/O, ", bar_nr);
+		}
+		else
+		{
+			if ((bar_size & PCI_BASE_ADDRESS_MEM_TYPE_MASK) ==
+			    PCI_BASE_ADDRESS_MEM_TYPE_64)
+				found_mem64 = 1;
+
+			bar_size &= PCI_BASE_ADDRESS_MEM_MASK;
+			upper_limit = &pciauto_upper_memspc;
+			pr_debug("PCI Autoconfig: BAR %d, Mem, ", bar_nr);
+		}
+
+		/* Allocate a base address (bar_size is negative!) */
+		*upper_limit = (*upper_limit + bar_size) & bar_size;
+
+		/* Write it out and update our limit */
+		pci_write_config_dword(dev, bar, *upper_limit);
+
+		/*
+		 * If we are a 64-bit decoder then increment to the
+		 * upper 32 bits of the bar and force it to locate
+		 * in the lower 4GB of memory.
+		 */
+
+		if (found_mem64)
+			pci_write_config_dword(dev, (bar+=4), 0x00000000);
+
+		pr_debug("size=0x%x, address=0x%x\n",
+			 ~bar_size + 1, *upper_limit);
+	}
+}
+
+/* Initialize the interrupt number. */
+
+static void __init
+pciauto_setup_irq(struct pci_controller* pci_ctrl,struct pci_dev *dev,int devfn)
+{
+	u8 pin;
+	int irq = 0;
+
+	pci_read_config_byte(dev, PCI_INTERRUPT_PIN, &pin);
+
+	/* Fix illegal pin numbers. */
+
+	if (pin == 0 || pin > 4)
+		pin = 1;
+
+	if (pci_ctrl->map_irq)
+		irq = pci_ctrl->map_irq(dev, PCI_SLOT(devfn), pin);
+
+	if (irq == -1)
+		irq = 0;
+
+	pr_debug("PCI Autoconfig: Interrupt %d, pin %d\n", irq, pin);
+
+	pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq);
+}
+
+
+static void __init
+pciauto_prescan_setup_bridge(struct pci_dev *dev, int current_bus,
+			     int sub_bus, int *iosave, int *memsave)
+{
+	/* Configure bus number registers */
+	pci_write_config_byte(dev, PCI_PRIMARY_BUS, current_bus);
+	pci_write_config_byte(dev, PCI_SECONDARY_BUS, sub_bus + 1);
+	pci_write_config_byte(dev, PCI_SUBORDINATE_BUS,	0xff);
+
+	/* Round memory allocator to 1MB boundary */
+	pciauto_upper_memspc &= ~(0x100000 - 1);
+	*memsave = pciauto_upper_memspc;
+
+	/* Round I/O allocator to 4KB boundary */
+	pciauto_upper_iospc &= ~(0x1000 - 1);
+	*iosave = pciauto_upper_iospc;
+
+	/* Set up memory and I/O filter limits, assume 32-bit I/O space */
+	pci_write_config_word(dev, PCI_MEMORY_LIMIT,
+			      ((pciauto_upper_memspc - 1) & 0xfff00000) >> 16);
+	pci_write_config_byte(dev, PCI_IO_LIMIT,
+			      ((pciauto_upper_iospc - 1) & 0x0000f000) >> 8);
+	pci_write_config_word(dev, PCI_IO_LIMIT_UPPER16,
+			      ((pciauto_upper_iospc - 1) & 0xffff0000) >> 16);
+}
+
+static void __init
+pciauto_postscan_setup_bridge(struct pci_dev *dev, int current_bus, int sub_bus,
+			      int *iosave, int *memsave)
+{
+	int cmdstat;
+
+	/* Configure bus number registers */
+	pci_write_config_byte(dev, PCI_SUBORDINATE_BUS,	sub_bus);
+
+	/*
+	 * Round memory allocator to 1MB boundary.
+	 * If no space used, allocate minimum.
+	 */
+	pciauto_upper_memspc &= ~(0x100000 - 1);
+	if (*memsave == pciauto_upper_memspc)
+		pciauto_upper_memspc -= 0x00100000;
+
+	pci_write_config_word(dev, PCI_MEMORY_BASE, pciauto_upper_memspc >> 16);
+
+	/* Allocate 1MB for pre-fretch */
+	pci_write_config_word(dev, PCI_PREF_MEMORY_LIMIT,
+			      ((pciauto_upper_memspc - 1) & 0xfff00000) >> 16);
+
+	pciauto_upper_memspc -= 0x100000;
+
+	pci_write_config_word(dev, PCI_PREF_MEMORY_BASE,
+			      pciauto_upper_memspc >> 16);
+
+	/* Round I/O allocator to 4KB boundary */
+	pciauto_upper_iospc &= ~(0x1000 - 1);
+	if (*iosave == pciauto_upper_iospc)
+		pciauto_upper_iospc -= 0x1000;
+
+	pci_write_config_byte(dev, PCI_IO_BASE,
+			      (pciauto_upper_iospc & 0x0000f000) >> 8);
+	pci_write_config_word(dev, PCI_IO_BASE_UPPER16,
+			      pciauto_upper_iospc >> 16);
+
+	/* Enable memory and I/O accesses, enable bus master */
+	pci_read_config_dword(dev, PCI_COMMAND, &cmdstat);
+	pci_write_config_dword(dev, PCI_COMMAND,
+			       cmdstat |
+			       PCI_COMMAND_IO |
+			       PCI_COMMAND_MEMORY |
+			       PCI_COMMAND_MASTER);
+}
+
+/*
+ * Scan the current PCI bus.
+ */
+
+
+int __init pciauto_bus_scan(struct pci_controller *pci_ctrl, int current_bus)
+{
+	int sub_bus, pci_devfn, pci_class, cmdstat;
+	unsigned short vid;
+	unsigned char header_type;
+	struct pci_dev *dev = &pciauto_dev;
+	bool found_multi = false;
+
+	pciauto_dev.bus = &pciauto_bus;
+	pciauto_dev.sysdata = pci_ctrl;
+	pciauto_bus.ops = pci_ctrl->ops;
+
+	/*
+	 * Fetch our I/O and memory space upper boundaries used
+	 * to allocated base addresses on this pci_controller.
+	 */
+
+	if (current_bus == pci_ctrl->first_busno)
+	{
+		pciauto_upper_iospc = pci_ctrl->io_resource.end + 1;
+		pciauto_upper_memspc = pci_ctrl->mem_resources[0].end + 1;
+	}
+
+	sub_bus = current_bus;
+
+	for (pci_devfn = 0; pci_devfn < 0xff; pci_devfn++)
+	{
+		/* Skip our host bridge */
+		if ((current_bus == pci_ctrl->first_busno) && (pci_devfn == 0))
+			continue;
+
+		if (PCI_FUNC(pci_devfn) && !found_multi)
+			continue;
+
+		pciauto_bus.number = current_bus;
+		pciauto_dev.devfn = pci_devfn;
+
+		/* If config space read fails from this device, move on */
+		if (pci_read_config_byte(dev, PCI_HEADER_TYPE, &header_type))
+			continue;
+
+		if (!PCI_FUNC(pci_devfn))
+			found_multi = FIELD_GET(PCI_HEADER_TYPE_MFD, header_type);
+		pci_read_config_word(dev, PCI_VENDOR_ID, &vid);
+
+		if (vid == 0xffff || vid == 0x0000) {
+			found_multi = false;
+			continue;
+		}
+
+		pci_read_config_dword(dev, PCI_CLASS_REVISION, &pci_class);
+
+		if ((pci_class >> 16) == PCI_CLASS_BRIDGE_PCI) {
+
+			int iosave, memsave;
+
+			pr_debug("PCI Autoconfig: Found P2P bridge, device %d\n",
+				 PCI_SLOT(pci_devfn));
+
+			/* Allocate PCI I/O and/or memory space */
+			pciauto_setup_bars(dev, PCI_BASE_ADDRESS_1);
+
+			pciauto_prescan_setup_bridge(dev, current_bus, sub_bus,
+					&iosave, &memsave);
+			sub_bus = pciauto_bus_scan(pci_ctrl, sub_bus+1);
+			pciauto_postscan_setup_bridge(dev, current_bus, sub_bus,
+					&iosave, &memsave);
+			pciauto_bus.number = current_bus;
+
+			continue;
+
+		}
+
+		/*
+		 * Found a peripheral, enable some standard
+		 * settings
+		 */
+
+		pci_read_config_dword(dev, PCI_COMMAND,	&cmdstat);
+		pci_write_config_dword(dev, PCI_COMMAND,
+				cmdstat |
+					PCI_COMMAND_IO |
+					PCI_COMMAND_MEMORY |
+					PCI_COMMAND_MASTER);
+		pci_write_config_byte(dev, PCI_LATENCY_TIMER, 0x80);
+
+		/* Allocate PCI I/O and/or memory space */
+		pr_debug("PCI Autoconfig: Found Bus %d, Device %d, Function %d\n",
+			 current_bus, PCI_SLOT(pci_devfn), PCI_FUNC(pci_devfn));
+
+		pciauto_setup_bars(dev, PCI_BASE_ADDRESS_5);
+		pciauto_setup_irq(pci_ctrl, dev, pci_devfn);
+	}
+	return sub_bus;
+}
diff --git a/arch/xtensa/lib/strncpy_user.S b/arch/xtensa/lib/strncpy_user.S
new file mode 100644
index 000000000..9841d1694
--- /dev/null
+++ b/arch/xtensa/lib/strncpy_user.S
@@ -0,0 +1,217 @@
+/*
+ *  arch/xtensa/lib/strncpy_user.S
+ *
+ *  This file is subject to the terms and conditions of the GNU General
+ *  Public License.  See the file "COPYING" in the main directory of
+ *  this archive for more details.
+ *
+ *  Returns: -EFAULT if exception before terminator, N if the entire
+ *  buffer filled, else strlen.
+ *
+ *  Copyright (C) 2002 Tensilica Inc.
+ */
+
+#include <linux/errno.h>
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+/*
+ * char *__strncpy_user(char *dst, const char *src, size_t len)
+ */
+
+#ifdef __XTENSA_EB__
+# define MASK0 0xff000000
+# define MASK1 0x00ff0000
+# define MASK2 0x0000ff00
+# define MASK3 0x000000ff
+#else
+# define MASK0 0x000000ff
+# define MASK1 0x0000ff00
+# define MASK2 0x00ff0000
+# define MASK3 0xff000000
+#endif
+
+# Register use
+#   a0/ return address
+#   a1/ stack pointer
+#   a2/ return value
+#   a3/ src
+#   a4/ len
+#   a5/ mask0
+#   a6/ mask1
+#   a7/ mask2
+#   a8/ mask3
+#   a9/ tmp
+#   a10/ tmp
+#   a11/ dst
+
+.text
+ENTRY(__strncpy_user)
+
+	abi_entry_default
+	# a2/ dst, a3/ src, a4/ len
+	mov	a11, a2		# leave dst in return value register
+	beqz	a4, .Lret	# if len is zero
+	movi	a5, MASK0	# mask for byte 0
+	movi	a6, MASK1	# mask for byte 1
+	movi	a7, MASK2	# mask for byte 2
+	movi	a8, MASK3	# mask for byte 3
+	bbsi.l	a3, 0, .Lsrc1mod2 # if only  8-bit aligned
+	bbsi.l	a3, 1, .Lsrc2mod4 # if only 16-bit aligned
+.Lsrcaligned:	# return here when src is word-aligned
+	srli	a10, a4, 2	# number of loop iterations with 4B per loop
+	movi	a9, 3
+	bnone	a11, a9, .Laligned
+	j	.Ldstunaligned
+
+.Lsrc1mod2:	# src address is odd
+EX(11f)	l8ui	a9, a3, 0		# get byte 0
+	addi	a3, a3, 1		# advance src pointer
+EX(10f)	s8i	a9, a11, 0		# store byte 0
+	beqz	a9, .Lret		# if byte 0 is zero
+	addi	a11, a11, 1		# advance dst pointer
+	addi	a4, a4, -1		# decrement len
+	beqz	a4, .Lret		# if len is zero
+	bbci.l	a3, 1, .Lsrcaligned	# if src is now word-aligned
+
+.Lsrc2mod4:	# src address is 2 mod 4
+EX(11f)	l8ui	a9, a3, 0		# get byte 0
+	/* 1-cycle interlock */
+EX(10f)	s8i	a9, a11, 0		# store byte 0
+	beqz	a9, .Lret		# if byte 0 is zero
+	addi	a11, a11, 1		# advance dst pointer
+	addi	a4, a4, -1		# decrement len
+	beqz	a4, .Lret		# if len is zero
+EX(11f)	l8ui	a9, a3, 1		# get byte 0
+	addi	a3, a3, 2		# advance src pointer
+EX(10f)	s8i	a9, a11, 0		# store byte 0
+	beqz	a9, .Lret		# if byte 0 is zero
+	addi	a11, a11, 1		# advance dst pointer
+	addi	a4, a4, -1		# decrement len
+	bnez	a4, .Lsrcaligned	# if len is nonzero
+.Lret:
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+
+/*
+ * dst is word-aligned, src is word-aligned
+ */
+	.align	4		# 1 mod 4 alignment for LOOPNEZ
+	.byte	0		# (0 mod 4 alignment for LBEG)
+.Laligned:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a10, .Loop1done
+#else
+	beqz	a10, .Loop1done
+	slli	a10, a10, 2
+	add	a10, a10, a11	# a10 = end of last 4B chunck
+#endif
+.Loop1:
+EX(11f)	l32i	a9, a3, 0		# get word from src
+	addi	a3, a3, 4		# advance src pointer
+	bnone	a9, a5, .Lz0		# if byte 0 is zero
+	bnone	a9, a6, .Lz1		# if byte 1 is zero
+	bnone	a9, a7, .Lz2		# if byte 2 is zero
+EX(10f)	s32i	a9, a11, 0		# store word to dst
+	bnone	a9, a8, .Lz3		# if byte 3 is zero
+	addi	a11, a11, 4		# advance dst pointer
+#if !XCHAL_HAVE_LOOPS
+	blt	a11, a10, .Loop1
+#endif
+
+.Loop1done:
+	bbci.l	a4, 1, .L100
+	# copy 2 bytes
+EX(11f)	l16ui	a9, a3, 0
+	addi	a3, a3, 2		# advance src pointer
+#ifdef __XTENSA_EB__
+	bnone	a9, a7, .Lz0		# if byte 2 is zero
+	bnone	a9, a8, .Lz1		# if byte 3 is zero
+#else
+	bnone	a9, a5, .Lz0		# if byte 0 is zero
+	bnone	a9, a6, .Lz1		# if byte 1 is zero
+#endif
+EX(10f)	s16i	a9, a11, 0
+	addi	a11, a11, 2		# advance dst pointer
+.L100:
+	bbci.l	a4, 0, .Lret
+EX(11f)	l8ui	a9, a3, 0
+	/* slot */
+EX(10f)	s8i	a9, a11, 0
+	beqz	a9, .Lret		# if byte is zero
+	addi	a11, a11, 1-3		# advance dst ptr 1, but also cancel
+					# the effect of adding 3 in .Lz3 code
+	/* fall thru to .Lz3 and "retw" */
+
+.Lz3:	# byte 3 is zero
+	addi	a11, a11, 3		# advance dst pointer
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+.Lz0:	# byte 0 is zero
+#ifdef __XTENSA_EB__
+	movi	a9, 0
+#endif /* __XTENSA_EB__ */
+EX(10f)	s8i	a9, a11, 0
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+.Lz1:	# byte 1 is zero
+#ifdef __XTENSA_EB__
+	extui   a9, a9, 16, 16
+#endif /* __XTENSA_EB__ */
+EX(10f)	s16i	a9, a11, 0
+	addi	a11, a11, 1		# advance dst pointer
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+.Lz2:	# byte 2 is zero
+#ifdef __XTENSA_EB__
+	extui   a9, a9, 16, 16
+#endif /* __XTENSA_EB__ */
+EX(10f)	s16i	a9, a11, 0
+	movi	a9, 0
+EX(10f)	s8i	a9, a11, 2
+	addi	a11, a11, 2		# advance dst pointer
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+
+	.align	4		# 1 mod 4 alignment for LOOPNEZ
+	.byte	0		# (0 mod 4 alignment for LBEG)
+.Ldstunaligned:
+/*
+ * for now just use byte copy loop
+ */
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lunalignedend
+#else
+	beqz	a4, .Lunalignedend
+	add	a10, a11, a4		# a10 = ending address
+#endif /* XCHAL_HAVE_LOOPS */
+.Lnextbyte:
+EX(11f)	l8ui	a9, a3, 0
+	addi	a3, a3, 1
+EX(10f)	s8i	a9, a11, 0
+	beqz	a9, .Lunalignedend
+	addi	a11, a11, 1
+#if !XCHAL_HAVE_LOOPS
+	blt	a11, a10, .Lnextbyte
+#endif
+
+.Lunalignedend:
+	sub	a2, a11, a2		# compute strlen
+	abi_ret_default
+
+ENDPROC(__strncpy_user)
+EXPORT_SYMBOL(__strncpy_user)
+
+	.section .fixup, "ax"
+	.align	4
+
+	/* For now, just return -EFAULT.  Future implementations might
+	 * like to clear remaining kernel space, like the fixup
+	 * implementation in memset().  Thus, we differentiate between
+	 * load/store fixups. */
+
+10:
+11:
+	movi	a2, -EFAULT
+	abi_ret_default
diff --git a/arch/xtensa/lib/strnlen_user.S b/arch/xtensa/lib/strnlen_user.S
new file mode 100644
index 000000000..cdcf57474
--- /dev/null
+++ b/arch/xtensa/lib/strnlen_user.S
@@ -0,0 +1,142 @@
+/*
+ *  arch/xtensa/lib/strnlen_user.S
+ *
+ *  This file is subject to the terms and conditions of the GNU General
+ *  Public License.  See the file "COPYING" in the main directory of
+ *  this archive for more details.
+ *
+ *  Returns strnlen, including trailing zero terminator.
+ *  Zero indicates error.
+ *
+ *  Copyright (C) 2002 Tensilica Inc.
+ */
+
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+/*
+ * size_t __strnlen_user(const char *s, size_t len)
+ */
+
+#ifdef __XTENSA_EB__
+# define MASK0 0xff000000
+# define MASK1 0x00ff0000
+# define MASK2 0x0000ff00
+# define MASK3 0x000000ff
+#else
+# define MASK0 0x000000ff
+# define MASK1 0x0000ff00
+# define MASK2 0x00ff0000
+# define MASK3 0xff000000
+#endif
+
+# Register use:
+#   a2/ src
+#   a3/ len
+#   a4/ tmp
+#   a5/ mask0
+#   a6/ mask1
+#   a7/ mask2
+#   a8/ mask3
+#   a9/ tmp
+#   a10/ tmp
+
+.text
+ENTRY(__strnlen_user)
+
+	abi_entry_default
+	# a2/ s, a3/ len
+	addi	a4, a2, -4	# because we overincrement at the end;
+				# we compensate with load offsets of 4
+	movi	a5, MASK0	# mask for byte 0
+	movi	a6, MASK1	# mask for byte 1
+	movi	a7, MASK2	# mask for byte 2
+	movi	a8, MASK3	# mask for byte 3
+	bbsi.l	a2, 0, .L1mod2	# if only  8-bit aligned
+	bbsi.l	a2, 1, .L2mod4	# if only 16-bit aligned
+
+/*
+ * String is word-aligned.
+ */
+.Laligned:
+	srli	a10, a3, 2	# number of loop iterations with 4B per loop
+#if XCHAL_HAVE_LOOPS
+	loopnez	a10, .Ldone
+#else
+	beqz	a10, .Ldone
+	slli	a10, a10, 2
+	add	a10, a10, a4	# a10 = end of last 4B chunk
+#endif /* XCHAL_HAVE_LOOPS */
+.Loop:
+EX(10f)	l32i	a9, a4, 4		# get next word of string
+	addi	a4, a4, 4		# advance string pointer
+	bnone	a9, a5, .Lz0		# if byte 0 is zero
+	bnone	a9, a6, .Lz1		# if byte 1 is zero
+	bnone	a9, a7, .Lz2		# if byte 2 is zero
+	bnone	a9, a8, .Lz3		# if byte 3 is zero
+#if !XCHAL_HAVE_LOOPS
+	blt	a4, a10, .Loop
+#endif
+
+.Ldone:
+EX(10f)	l32i	a9, a4, 4	# load 4 bytes for remaining checks
+
+	bbci.l	a3, 1, .L100
+	# check two more bytes (bytes 0, 1 of word)
+	addi	a4, a4, 2	# advance string pointer
+	bnone	a9, a5, .Lz0	# if byte 0 is zero
+	bnone	a9, a6, .Lz1	# if byte 1 is zero
+.L100:
+	bbci.l	a3, 0, .L101
+	# check one more byte (byte 2 of word)
+	# Actually, we don't need to check.  Zero or nonzero, we'll add one.
+	# Do not add an extra one for the NULL terminator since we have
+	#  exhausted the original len parameter.
+	addi	a4, a4, 1	# advance string pointer
+.L101:
+	sub	a2, a4, a2	# compute length
+	abi_ret_default
+
+# NOTE that in several places below, we point to the byte just after
+# the zero byte in order to include the NULL terminator in the count.
+
+.Lz3:	# byte 3 is zero
+	addi	a4, a4, 3	# point to zero byte
+.Lz0:	# byte 0 is zero
+	addi	a4, a4, 1	# point just beyond zero byte
+	sub	a2, a4, a2	# subtract to get length
+	abi_ret_default
+.Lz1:	# byte 1 is zero
+	addi	a4, a4, 1+1	# point just beyond zero byte
+	sub	a2, a4, a2	# subtract to get length
+	abi_ret_default
+.Lz2:	# byte 2 is zero
+	addi	a4, a4, 2+1	# point just beyond zero byte
+	sub	a2, a4, a2	# subtract to get length
+	abi_ret_default
+
+.L1mod2:	# address is odd
+EX(10f)	l8ui	a9, a4, 4		# get byte 0
+	addi	a4, a4, 1		# advance string pointer
+	beqz	a9, .Lz3		# if byte 0 is zero
+	bbci.l	a4, 1, .Laligned	# if string pointer is now word-aligned
+
+.L2mod4:	# address is 2 mod 4
+	addi	a4, a4, 2	# advance ptr for aligned access
+EX(10f)	l32i	a9, a4, 0	# get word with first two bytes of string
+	bnone	a9, a7, .Lz2	# if byte 2 (of word, not string) is zero
+	bany	a9, a8, .Laligned # if byte 3 (of word, not string) is nonzero
+	# byte 3 is zero
+	addi	a4, a4, 3+1	# point just beyond zero byte
+	sub	a2, a4, a2	# subtract to get length
+	abi_ret_default
+
+ENDPROC(__strnlen_user)
+EXPORT_SYMBOL(__strnlen_user)
+
+	.section .fixup, "ax"
+	.align	4
+10:
+	movi	a2, 0
+	abi_ret_default
diff --git a/arch/xtensa/lib/udivsi3.S b/arch/xtensa/lib/udivsi3.S
new file mode 100644
index 000000000..59ea2dfc3
--- /dev/null
+++ b/arch/xtensa/lib/udivsi3.S
@@ -0,0 +1,69 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__udivsi3)
+
+	abi_entry_default
+#if XCHAL_HAVE_DIV32
+	quou	a2, a2, a3
+#else
+	bltui	a3, 2, .Lle_one	/* check if the divisor <= 1 */
+
+	mov	a6, a2		/* keep dividend in a6 */
+	do_nsau	a5, a6, a2, a7	/* dividend_shift = nsau (dividend) */
+	do_nsau	a4, a3, a2, a7	/* divisor_shift = nsau (divisor) */
+	bgeu	a5, a4, .Lspecial
+
+	sub	a4, a4, a5	/* count = divisor_shift - dividend_shift */
+	ssl	a4
+	sll	a3, a3		/* divisor <<= count */
+	movi	a2, 0		/* quotient = 0 */
+
+	/* test-subtract-and-shift loop; one quotient bit on each iteration */
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lloopend
+#endif /* XCHAL_HAVE_LOOPS */
+.Lloop:
+	bltu	a6, a3, .Lzerobit
+	sub	a6, a6, a3
+	addi	a2, a2, 1
+.Lzerobit:
+	slli	a2, a2, 1
+	srli	a3, a3, 1
+#if !XCHAL_HAVE_LOOPS
+	addi	a4, a4, -1
+	bnez	a4, .Lloop
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lloopend:
+
+	bltu	a6, a3, .Lreturn
+	addi	a2, a2, 1	/* increment quotient if dividend >= divisor */
+.Lreturn:
+	abi_ret_default
+
+.Lle_one:
+	beqz	a3, .Lerror	/* if divisor == 1, return the dividend */
+	abi_ret_default
+
+.Lspecial:
+	/* return dividend >= divisor */
+	bltu	a6, a3, .Lreturn0
+	movi	a2, 1
+	abi_ret_default
+
+.Lerror:
+	/* Divide by zero: Use an illegal instruction to force an exception.
+	   The subsequent "DIV0" string can be recognized by the exception
+	   handler to identify the real cause of the exception.  */
+	ill
+	.ascii	"DIV0"
+
+.Lreturn0:
+	movi	a2, 0
+#endif /* XCHAL_HAVE_DIV32 */
+	abi_ret_default
+
+ENDPROC(__udivsi3)
+EXPORT_SYMBOL(__udivsi3)
diff --git a/arch/xtensa/lib/umodsi3.S b/arch/xtensa/lib/umodsi3.S
new file mode 100644
index 000000000..d39a7e56a
--- /dev/null
+++ b/arch/xtensa/lib/umodsi3.S
@@ -0,0 +1,58 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+ENTRY(__umodsi3)
+
+	abi_entry_default
+#if XCHAL_HAVE_DIV32
+	remu	a2, a2, a3
+#else
+	bltui	a3, 2, .Lle_one	/* check if the divisor is <= 1 */
+
+	do_nsau	a5, a2, a6, a7	/* dividend_shift = nsau (dividend) */
+	do_nsau	a4, a3, a6, a7	/* divisor_shift = nsau (divisor) */
+	bgeu	a5, a4, .Lspecial
+
+	sub	a4, a4, a5	/* count = divisor_shift - dividend_shift */
+	ssl	a4
+	sll	a3, a3		/* divisor <<= count */
+
+	/* test-subtract-and-shift loop */
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lloopend
+#endif /* XCHAL_HAVE_LOOPS */
+.Lloop:
+	bltu	a2, a3, .Lzerobit
+	sub	a2, a2, a3
+.Lzerobit:
+	srli	a3, a3, 1
+#if !XCHAL_HAVE_LOOPS
+	addi	a4, a4, -1
+	bnez	a4, .Lloop
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lloopend:
+
+.Lspecial:
+	bltu	a2, a3, .Lreturn
+	sub	a2, a2, a3	/* subtract once more if dividend >= divisor */
+.Lreturn:
+	abi_ret_default
+
+.Lle_one:
+	bnez	a3, .Lreturn0
+
+	/* Divide by zero: Use an illegal instruction to force an exception.
+	   The subsequent "DIV0" string can be recognized by the exception
+	   handler to identify the real cause of the exception.  */
+	ill
+	.ascii	"DIV0"
+
+.Lreturn0:
+	movi	a2, 0
+#endif /* XCHAL_HAVE_DIV32 */
+	abi_ret_default
+
+ENDPROC(__umodsi3)
+EXPORT_SYMBOL(__umodsi3)
diff --git a/arch/xtensa/lib/umulsidi3.S b/arch/xtensa/lib/umulsidi3.S
new file mode 100644
index 000000000..5da501b57
--- /dev/null
+++ b/arch/xtensa/lib/umulsidi3.S
@@ -0,0 +1,233 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later WITH GCC-exception-2.0 */
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+#if XCHAL_HAVE_MUL16 || XCHAL_HAVE_MUL32 || XCHAL_HAVE_MAC16
+#define XCHAL_NO_MUL 0
+#else
+#define XCHAL_NO_MUL 1
+#endif
+
+ENTRY(__umulsidi3)
+
+#ifdef __XTENSA_CALL0_ABI__
+	abi_entry(32)
+	s32i	a12, sp, 16
+	s32i	a13, sp, 20
+	s32i	a14, sp, 24
+	s32i	a15, sp, 28
+#elif XCHAL_NO_MUL
+	/* This is not really a leaf function; allocate enough stack space
+	   to allow CALL12s to a helper function.  */
+	abi_entry(32)
+#else
+	abi_entry_default
+#endif
+
+#ifdef __XTENSA_EB__
+#define wh a2
+#define wl a3
+#else
+#define wh a3
+#define wl a2
+#endif /* __XTENSA_EB__ */
+
+	/* This code is taken from the mulsf3 routine in ieee754-sf.S.
+	   See more comments there.  */
+
+#if XCHAL_HAVE_MUL32_HIGH
+	mull	a6, a2, a3
+	muluh	wh, a2, a3
+	mov	wl, a6
+
+#else /* ! MUL32_HIGH */
+
+#if defined(__XTENSA_CALL0_ABI__) && XCHAL_NO_MUL
+	/* a0 and a8 will be clobbered by calling the multiply function
+	   but a8 is not used here and need not be saved.  */
+	s32i	a0, sp, 0
+#endif
+
+#if XCHAL_HAVE_MUL16 || XCHAL_HAVE_MUL32
+
+#define a2h a4
+#define a3h a5
+
+	/* Get the high halves of the inputs into registers.  */
+	srli	a2h, a2, 16
+	srli	a3h, a3, 16
+
+#define a2l a2
+#define a3l a3
+
+#if XCHAL_HAVE_MUL32 && !XCHAL_HAVE_MUL16
+	/* Clear the high halves of the inputs.  This does not matter
+	   for MUL16 because the high bits are ignored.  */
+	extui	a2, a2, 0, 16
+	extui	a3, a3, 0, 16
+#endif
+#endif /* MUL16 || MUL32 */
+
+
+#if XCHAL_HAVE_MUL16
+
+#define do_mul(dst, xreg, xhalf, yreg, yhalf) \
+	mul16u	dst, xreg ## xhalf, yreg ## yhalf
+
+#elif XCHAL_HAVE_MUL32
+
+#define do_mul(dst, xreg, xhalf, yreg, yhalf) \
+	mull	dst, xreg ## xhalf, yreg ## yhalf
+
+#elif XCHAL_HAVE_MAC16
+
+/* The preprocessor insists on inserting a space when concatenating after
+   a period in the definition of do_mul below.  These macros are a workaround
+   using underscores instead of periods when doing the concatenation.  */
+#define umul_aa_ll umul.aa.ll
+#define umul_aa_lh umul.aa.lh
+#define umul_aa_hl umul.aa.hl
+#define umul_aa_hh umul.aa.hh
+
+#define do_mul(dst, xreg, xhalf, yreg, yhalf) \
+	umul_aa_ ## xhalf ## yhalf	xreg, yreg; \
+	rsr	dst, ACCLO
+
+#else /* no multiply hardware */
+
+#define set_arg_l(dst, src) \
+	extui	dst, src, 0, 16
+#define set_arg_h(dst, src) \
+	srli	dst, src, 16
+
+#ifdef __XTENSA_CALL0_ABI__
+#define do_mul(dst, xreg, xhalf, yreg, yhalf) \
+	set_arg_ ## xhalf (a13, xreg); \
+	set_arg_ ## yhalf (a14, yreg); \
+	call0	.Lmul_mulsi3; \
+	mov	dst, a12
+#else
+#define do_mul(dst, xreg, xhalf, yreg, yhalf) \
+	set_arg_ ## xhalf (a14, xreg); \
+	set_arg_ ## yhalf (a15, yreg); \
+	call12	.Lmul_mulsi3; \
+	mov	dst, a14
+#endif /* __XTENSA_CALL0_ABI__ */
+
+#endif /* no multiply hardware */
+
+	/* Add pp1 and pp2 into a6 with carry-out in a9.  */
+	do_mul(a6, a2, l, a3, h)	/* pp 1 */
+	do_mul(a11, a2, h, a3, l)	/* pp 2 */
+	movi	a9, 0
+	add	a6, a6, a11
+	bgeu	a6, a11, 1f
+	addi	a9, a9, 1
+1:
+	/* Shift the high half of a9/a6 into position in a9.  Note that
+	   this value can be safely incremented without any carry-outs.  */
+	ssai	16
+	src	a9, a9, a6
+
+	/* Compute the low word into a6.  */
+	do_mul(a11, a2, l, a3, l)	/* pp 0 */
+	sll	a6, a6
+	add	a6, a6, a11
+	bgeu	a6, a11, 1f
+	addi	a9, a9, 1
+1:
+	/* Compute the high word into wh.  */
+	do_mul(wh, a2, h, a3, h)	/* pp 3 */
+	add	wh, wh, a9
+	mov	wl, a6
+
+#endif /* !MUL32_HIGH */
+
+#if defined(__XTENSA_CALL0_ABI__) && XCHAL_NO_MUL
+	/* Restore the original return address.  */
+	l32i	a0, sp, 0
+#endif
+#ifdef __XTENSA_CALL0_ABI__
+	l32i	a12, sp, 16
+	l32i	a13, sp, 20
+	l32i	a14, sp, 24
+	l32i	a15, sp, 28
+	abi_ret(32)
+#else
+	abi_ret_default
+#endif
+
+#if XCHAL_NO_MUL
+
+	.macro	do_addx2 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx2	\dst, \as, \at
+#else
+	slli	\tmp, \as, 1
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+	.macro	do_addx4 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx4	\dst, \as, \at
+#else
+	slli	\tmp, \as, 2
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+	.macro	do_addx8 dst, as, at, tmp
+#if XCHAL_HAVE_ADDX
+	addx8	\dst, \as, \at
+#else
+	slli	\tmp, \as, 3
+	add	\dst, \tmp, \at
+#endif
+	.endm
+
+	/* For Xtensa processors with no multiply hardware, this simplified
+	   version of _mulsi3 is used for multiplying 16-bit chunks of
+	   the floating-point mantissas.  When using CALL0, this function
+	   uses a custom ABI: the inputs are passed in a13 and a14, the
+	   result is returned in a12, and a8 and a15 are clobbered.  */
+	.align	4
+.Lmul_mulsi3:
+	abi_entry_default
+
+	.macro mul_mulsi3_body dst, src1, src2, tmp1, tmp2
+	movi	\dst, 0
+1:	add	\tmp1, \src2, \dst
+	extui	\tmp2, \src1, 0, 1
+	movnez	\dst, \tmp1, \tmp2
+
+	do_addx2 \tmp1, \src2, \dst, \tmp1
+	extui	\tmp2, \src1, 1, 1
+	movnez	\dst, \tmp1, \tmp2
+
+	do_addx4 \tmp1, \src2, \dst, \tmp1
+	extui	\tmp2, \src1, 2, 1
+	movnez	\dst, \tmp1, \tmp2
+
+	do_addx8 \tmp1, \src2, \dst, \tmp1
+	extui	\tmp2, \src1, 3, 1
+	movnez	\dst, \tmp1, \tmp2
+
+	srli	\src1, \src1, 4
+	slli	\src2, \src2, 4
+	bnez	\src1, 1b
+	.endm
+
+#ifdef __XTENSA_CALL0_ABI__
+	mul_mulsi3_body a12, a13, a14, a15, a8
+#else
+	/* The result will be written into a2, so save that argument in a4.  */
+	mov	a4, a2
+	mul_mulsi3_body a2, a4, a3, a5, a6
+#endif
+	abi_ret_default
+#endif /* XCHAL_NO_MUL */
+
+ENDPROC(__umulsidi3)
+EXPORT_SYMBOL(__umulsidi3)
diff --git a/arch/xtensa/lib/usercopy.S b/arch/xtensa/lib/usercopy.S
new file mode 100644
index 000000000..2c665c0b4
--- /dev/null
+++ b/arch/xtensa/lib/usercopy.S
@@ -0,0 +1,301 @@
+/*
+ *  arch/xtensa/lib/usercopy.S
+ *
+ *  Copy to/from user space (derived from arch/xtensa/lib/hal/memcopy.S)
+ *
+ *  DO NOT COMBINE this function with <arch/xtensa/lib/hal/memcopy.S>.
+ *  It needs to remain separate and distinct.  The hal files are part
+ *  of the Xtensa link-time HAL, and those files may differ per
+ *  processor configuration.  Patching the kernel for another
+ *  processor configuration includes replacing the hal files, and we
+ *  could lose the special functionality for accessing user-space
+ *  memory during such a patch.  We sacrifice a little code space here
+ *  in favor to simplify code maintenance.
+ *
+ *  This file is subject to the terms and conditions of the GNU General
+ *  Public License.  See the file "COPYING" in the main directory of
+ *  this archive for more details.
+ *
+ *  Copyright (C) 2002 Tensilica Inc.
+ */
+
+
+/*
+ * size_t __xtensa_copy_user (void *dst, const void *src, size_t len);
+ *
+ * The returned value is the number of bytes not copied.  Implies zero
+ * is success.
+ *
+ * The general case algorithm is as follows:
+ *   If the destination and source are both aligned,
+ *     do 16B chunks with a loop, and then finish up with
+ *     8B, 4B, 2B, and 1B copies conditional on the length.
+ *   If destination is aligned and source unaligned,
+ *     do the same, but use SRC to align the source data.
+ *   If destination is unaligned, align it by conditionally
+ *     copying 1B and 2B and then retest.
+ *   This code tries to use fall-through braches for the common
+ *     case of aligned destinations (except for the branches to
+ *     the alignment label).
+ *
+ * Register use:
+ *	a0/ return address
+ *	a1/ stack pointer
+ *	a2/ return value
+ *	a3/ src
+ *	a4/ length
+ *	a5/ dst
+ *	a6/ tmp
+ *	a7/ tmp
+ *	a8/ tmp
+ *	a9/ tmp
+ *	a10/ tmp
+ *	a11/ original length
+ */
+
+#include <linux/linkage.h>
+#include <asm/asmmacro.h>
+#include <asm/core.h>
+
+	.text
+ENTRY(__xtensa_copy_user)
+
+#if !XCHAL_HAVE_LOOPS && defined(__XTENSA_CALL0_ABI__)
+#define STACK_SIZE 4
+#else
+#define STACK_SIZE 0
+#endif
+	abi_entry(STACK_SIZE)
+	# a2/ dst, a3/ src, a4/ len
+	mov	a5, a2		# copy dst so that a2 is return value
+	mov	a11, a4		# preserve original len for error case
+.Lcommon:
+	bbsi.l	a2, 0, .Ldst1mod2 # if dst is 1 mod 2
+	bbsi.l	a2, 1, .Ldst2mod4 # if dst is 2 mod 4
+.Ldstaligned:	# return here from .Ldstunaligned when dst is aligned
+	srli	a7, a4, 4	# number of loop iterations with 16B
+				# per iteration
+	movi	a8, 3		  # if source is also aligned,
+	bnone	a3, a8, .Laligned # then use word copy
+	__ssa8	a3		# set shift amount from byte offset
+	bnez	a4, .Lsrcunaligned
+	movi	a2, 0		# return success for len==0
+	abi_ret(STACK_SIZE)
+
+/*
+ * Destination is unaligned
+ */
+
+.Ldst1mod2:	# dst is only byte aligned
+	bltui	a4, 7, .Lbytecopy	# do short copies byte by byte
+
+	# copy 1 byte
+EX(10f)	l8ui	a6, a3, 0
+	addi	a3, a3,  1
+EX(10f)	s8i	a6, a5,  0
+	addi	a5, a5,  1
+	addi	a4, a4, -1
+	bbci.l	a5, 1, .Ldstaligned	# if dst is now aligned, then
+					# return to main algorithm
+.Ldst2mod4:	# dst 16-bit aligned
+	# copy 2 bytes
+	bltui	a4, 6, .Lbytecopy	# do short copies byte by byte
+EX(10f)	l8ui	a6, a3, 0
+EX(10f)	l8ui	a7, a3, 1
+	addi	a3, a3,  2
+EX(10f)	s8i	a6, a5,  0
+EX(10f)	s8i	a7, a5,  1
+	addi	a5, a5,  2
+	addi	a4, a4, -2
+	j	.Ldstaligned	# dst is now aligned, return to main algorithm
+
+/*
+ * Byte by byte copy
+ */
+	.align	4
+	.byte	0		# 1 mod 4 alignment for LOOPNEZ
+				# (0 mod 4 alignment for LBEG)
+.Lbytecopy:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a4, .Lbytecopydone
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a4, .Lbytecopydone
+	add	a7, a3, a4	# a7 = end address for source
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lnextbyte:
+EX(10f)	l8ui	a6, a3, 0
+	addi	a3, a3, 1
+EX(10f)	s8i	a6, a5, 0
+	addi	a5, a5, 1
+#if !XCHAL_HAVE_LOOPS
+	blt	a3, a7, .Lnextbyte
+#endif /* !XCHAL_HAVE_LOOPS */
+.Lbytecopydone:
+	movi	a2, 0		# return success for len bytes copied
+	abi_ret(STACK_SIZE)
+
+/*
+ * Destination and source are word-aligned.
+ */
+	# copy 16 bytes per iteration for word-aligned dst and word-aligned src
+	.align	4		# 1 mod 4 alignment for LOOPNEZ
+	.byte	0		# (0 mod 4 alignment for LBEG)
+.Laligned:
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .Loop1done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .Loop1done
+	slli	a8, a7, 4
+	add	a8, a8, a3	# a8 = end of last 16B source chunk
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1:
+EX(10f)	l32i	a6, a3,  0
+EX(10f)	l32i	a7, a3,  4
+EX(10f)	s32i	a6, a5,  0
+EX(10f)	l32i	a6, a3,  8
+EX(10f)	s32i	a7, a5,  4
+EX(10f)	l32i	a7, a3, 12
+EX(10f)	s32i	a6, a5,  8
+	addi	a3, a3, 16
+EX(10f)	s32i	a7, a5, 12
+	addi	a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+	blt	a3, a8, .Loop1
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop1done:
+	bbci.l	a4, 3, .L2
+	# copy 8 bytes
+EX(10f)	l32i	a6, a3,  0
+EX(10f)	l32i	a7, a3,  4
+	addi	a3, a3,  8
+EX(10f)	s32i	a6, a5,  0
+EX(10f)	s32i	a7, a5,  4
+	addi	a5, a5,  8
+.L2:
+	bbci.l	a4, 2, .L3
+	# copy 4 bytes
+EX(10f)	l32i	a6, a3,  0
+	addi	a3, a3,  4
+EX(10f)	s32i	a6, a5,  0
+	addi	a5, a5,  4
+.L3:
+	bbci.l	a4, 1, .L4
+	# copy 2 bytes
+EX(10f)	l16ui	a6, a3,  0
+	addi	a3, a3,  2
+EX(10f)	s16i	a6, a5,  0
+	addi	a5, a5,  2
+.L4:
+	bbci.l	a4, 0, .L5
+	# copy 1 byte
+EX(10f)	l8ui	a6, a3,  0
+EX(10f)	s8i	a6, a5,  0
+.L5:
+	movi	a2, 0		# return success for len bytes copied
+	abi_ret(STACK_SIZE)
+
+/*
+ * Destination is aligned, Source is unaligned
+ */
+
+	.align	4
+	.byte	0		# 1 mod 4 alignement for LOOPNEZ
+				# (0 mod 4 alignment for LBEG)
+.Lsrcunaligned:
+	# copy 16 bytes per iteration for word-aligned dst and unaligned src
+	and	a10, a3, a8	# save unalignment offset for below
+	sub	a3, a3, a10	# align a3 (to avoid sim warnings only; not needed for hardware)
+EX(10f)	l32i	a6, a3, 0	# load first word
+#if XCHAL_HAVE_LOOPS
+	loopnez	a7, .Loop2done
+#else /* !XCHAL_HAVE_LOOPS */
+	beqz	a7, .Loop2done
+#if defined(__XTENSA_CALL0_ABI__)
+	s32i	a10, a1, 0
+	slli	a10, a7, 4
+	add	a10, a10, a3	# a10 = end of last 16B source chunk
+#else
+	slli	a12, a7, 4
+	add	a12, a12, a3	# a12 = end of last 16B source chunk
+#endif
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2:
+EX(10f)	l32i	a7, a3,  4
+EX(10f)	l32i	a8, a3,  8
+	__src_b	a6, a6, a7
+EX(10f)	s32i	a6, a5,  0
+EX(10f)	l32i	a9, a3, 12
+	__src_b	a7, a7, a8
+EX(10f)	s32i	a7, a5,  4
+EX(10f)	l32i	a6, a3, 16
+	__src_b	a8, a8, a9
+EX(10f)	s32i	a8, a5,  8
+	addi	a3, a3, 16
+	__src_b	a9, a9, a6
+EX(10f)	s32i	a9, a5, 12
+	addi	a5, a5, 16
+#if !XCHAL_HAVE_LOOPS
+#if defined(__XTENSA_CALL0_ABI__)
+	blt	a3, a10, .Loop2
+	l32i	a10, a1, 0
+#else
+	blt	a3, a12, .Loop2
+#endif
+#endif /* !XCHAL_HAVE_LOOPS */
+.Loop2done:
+	bbci.l	a4, 3, .L12
+	# copy 8 bytes
+EX(10f)	l32i	a7, a3,  4
+EX(10f)	l32i	a8, a3,  8
+	__src_b	a6, a6, a7
+EX(10f)	s32i	a6, a5,  0
+	addi	a3, a3,  8
+	__src_b	a7, a7, a8
+EX(10f)	s32i	a7, a5,  4
+	addi	a5, a5,  8
+	mov	a6, a8
+.L12:
+	bbci.l	a4, 2, .L13
+	# copy 4 bytes
+EX(10f)	l32i	a7, a3,  4
+	addi	a3, a3,  4
+	__src_b	a6, a6, a7
+EX(10f)	s32i	a6, a5,  0
+	addi	a5, a5,  4
+	mov	a6, a7
+.L13:
+	add	a3, a3, a10	# readjust a3 with correct misalignment
+	bbci.l	a4, 1, .L14
+	# copy 2 bytes
+EX(10f)	l8ui	a6, a3,  0
+EX(10f)	l8ui	a7, a3,  1
+	addi	a3, a3,  2
+EX(10f)	s8i	a6, a5,  0
+EX(10f)	s8i	a7, a5,  1
+	addi	a5, a5,  2
+.L14:
+	bbci.l	a4, 0, .L15
+	# copy 1 byte
+EX(10f)	l8ui	a6, a3,  0
+EX(10f)	s8i	a6, a5,  0
+.L15:
+	movi	a2, 0		# return success for len bytes copied
+	abi_ret(STACK_SIZE)
+
+ENDPROC(__xtensa_copy_user)
+EXPORT_SYMBOL(__xtensa_copy_user)
+
+	.section .fixup, "ax"
+	.align	4
+
+/* a2 = original dst; a5 = current dst; a11= original len
+ * bytes_copied = a5 - a2
+ * retval = bytes_not_copied = original len - bytes_copied
+ * retval = a11 - (a5 - a2)
+ */
+
+
+10:
+	sub	a2, a5, a2	/* a2 <-- bytes copied */
+	sub	a2, a11, a2	/* a2 <-- bytes not copied */
+	abi_ret(STACK_SIZE)
diff --git a/arch/xtensa/mm/Makefile b/arch/xtensa/mm/Makefile
new file mode 100644
index 000000000..44153a335
--- /dev/null
+++ b/arch/xtensa/mm/Makefile
@@ -0,0 +1,14 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for the Linux/Xtensa-specific parts of the memory manager.
+#
+
+obj-y			:= init.o misc.o
+obj-$(CONFIG_PFAULT)	+= fault.o
+obj-$(CONFIG_MMU)	+= cache.o ioremap.o mmu.o tlb.o
+obj-$(CONFIG_HIGHMEM)	+= highmem.o
+obj-$(CONFIG_KASAN)	+= kasan_init.o
+
+KASAN_SANITIZE_fault.o := n
+KASAN_SANITIZE_kasan_init.o := n
+KASAN_SANITIZE_mmu.o := n
diff --git a/arch/xtensa/mm/cache.c b/arch/xtensa/mm/cache.c
new file mode 100644
index 000000000..5354df52d
--- /dev/null
+++ b/arch/xtensa/mm/cache.c
@@ -0,0 +1,335 @@
+/*
+ * arch/xtensa/mm/cache.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001-2006 Tensilica Inc.
+ *
+ * Chris Zankel	<chris@zankel.net>
+ * Joe Taylor
+ * Marc Gauthier
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/memblock.h>
+#include <linux/swap.h>
+#include <linux/pagemap.h>
+#include <linux/pgtable.h>
+
+#include <asm/bootparam.h>
+#include <asm/mmu_context.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+#include <asm/page.h>
+
+/* 
+ * Note:
+ * The kernel provides one architecture bit PG_arch_1 in the page flags that 
+ * can be used for cache coherency.
+ *
+ * I$-D$ coherency.
+ *
+ * The Xtensa architecture doesn't keep the instruction cache coherent with
+ * the data cache. We use the architecture bit to indicate if the caches
+ * are coherent. The kernel clears this bit whenever a page is added to the
+ * page cache. At that time, the caches might not be in sync. We, therefore,
+ * define this flag as 'clean' if set.
+ *
+ * D-cache aliasing.
+ *
+ * With cache aliasing, we have to always flush the cache when pages are
+ * unmapped (see tlb_start_vma(). So, we use this flag to indicate a dirty
+ * page.
+ * 
+ *
+ *
+ */
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+static inline void kmap_invalidate_coherent(struct page *page,
+					    unsigned long vaddr)
+{
+	if (!DCACHE_ALIAS_EQ(page_to_phys(page), vaddr)) {
+		unsigned long kvaddr;
+
+		if (!PageHighMem(page)) {
+			kvaddr = (unsigned long)page_to_virt(page);
+
+			__invalidate_dcache_page(kvaddr);
+		} else {
+			kvaddr = TLBTEMP_BASE_1 +
+				(page_to_phys(page) & DCACHE_ALIAS_MASK);
+
+			preempt_disable();
+			__invalidate_dcache_page_alias(kvaddr,
+						       page_to_phys(page));
+			preempt_enable();
+		}
+	}
+}
+
+static inline void *coherent_kvaddr(struct page *page, unsigned long base,
+				    unsigned long vaddr, unsigned long *paddr)
+{
+	*paddr = page_to_phys(page);
+	return (void *)(base + (vaddr & DCACHE_ALIAS_MASK));
+}
+
+void clear_user_highpage(struct page *page, unsigned long vaddr)
+{
+	struct folio *folio = page_folio(page);
+	unsigned long paddr;
+	void *kvaddr = coherent_kvaddr(page, TLBTEMP_BASE_1, vaddr, &paddr);
+
+	preempt_disable();
+	kmap_invalidate_coherent(page, vaddr);
+	set_bit(PG_arch_1, folio_flags(folio, 0));
+	clear_page_alias(kvaddr, paddr);
+	preempt_enable();
+}
+EXPORT_SYMBOL(clear_user_highpage);
+
+void copy_user_highpage(struct page *dst, struct page *src,
+			unsigned long vaddr, struct vm_area_struct *vma)
+{
+	struct folio *folio = page_folio(dst);
+	unsigned long dst_paddr, src_paddr;
+	void *dst_vaddr = coherent_kvaddr(dst, TLBTEMP_BASE_1, vaddr,
+					  &dst_paddr);
+	void *src_vaddr = coherent_kvaddr(src, TLBTEMP_BASE_2, vaddr,
+					  &src_paddr);
+
+	preempt_disable();
+	kmap_invalidate_coherent(dst, vaddr);
+	set_bit(PG_arch_1, folio_flags(folio, 0));
+	copy_page_alias(dst_vaddr, src_vaddr, dst_paddr, src_paddr);
+	preempt_enable();
+}
+EXPORT_SYMBOL(copy_user_highpage);
+
+/*
+ * Any time the kernel writes to a user page cache page, or it is about to
+ * read from a page cache page this routine is called.
+ *
+ */
+
+void flush_dcache_folio(struct folio *folio)
+{
+	struct address_space *mapping = folio_flush_mapping(folio);
+
+	/*
+	 * If we have a mapping but the page is not mapped to user-space
+	 * yet, we simply mark this page dirty and defer flushing the 
+	 * caches until update_mmu().
+	 */
+
+	if (mapping && !mapping_mapped(mapping)) {
+		if (!test_bit(PG_arch_1, &folio->flags.f))
+			set_bit(PG_arch_1, &folio->flags.f);
+		return;
+
+	} else {
+		unsigned long phys = folio_pfn(folio) * PAGE_SIZE;
+		unsigned long temp = folio_pos(folio);
+		unsigned int i, nr = folio_nr_pages(folio);
+		unsigned long alias = !(DCACHE_ALIAS_EQ(temp, phys));
+		unsigned long virt;
+
+		/* 
+		 * Flush the page in kernel space and user space.
+		 * Note that we can omit that step if aliasing is not
+		 * an issue, but we do have to synchronize I$ and D$
+		 * if we have a mapping.
+		 */
+
+		if (!alias && !mapping)
+			return;
+
+		preempt_disable();
+		for (i = 0; i < nr; i++) {
+			virt = TLBTEMP_BASE_1 + (phys & DCACHE_ALIAS_MASK);
+			__flush_invalidate_dcache_page_alias(virt, phys);
+
+			virt = TLBTEMP_BASE_1 + (temp & DCACHE_ALIAS_MASK);
+
+			if (alias)
+				__flush_invalidate_dcache_page_alias(virt, phys);
+
+			if (mapping)
+				__invalidate_icache_page_alias(virt, phys);
+			phys += PAGE_SIZE;
+			temp += PAGE_SIZE;
+		}
+		preempt_enable();
+	}
+
+	/* There shouldn't be an entry in the cache for this page anymore. */
+}
+EXPORT_SYMBOL(flush_dcache_folio);
+
+/*
+ * For now, flush the whole cache. FIXME??
+ */
+
+void local_flush_cache_range(struct vm_area_struct *vma,
+		       unsigned long start, unsigned long end)
+{
+	__flush_invalidate_dcache_all();
+	__invalidate_icache_all();
+}
+EXPORT_SYMBOL(local_flush_cache_range);
+
+/* 
+ * Remove any entry in the cache for this page. 
+ *
+ * Note that this function is only called for user pages, so use the
+ * alias versions of the cache flush functions.
+ */
+
+void local_flush_cache_page(struct vm_area_struct *vma, unsigned long address,
+		      unsigned long pfn)
+{
+	/* Note that we have to use the 'alias' address to avoid multi-hit */
+
+	unsigned long phys = page_to_phys(pfn_to_page(pfn));
+	unsigned long virt = TLBTEMP_BASE_1 + (address & DCACHE_ALIAS_MASK);
+
+	preempt_disable();
+	__flush_invalidate_dcache_page_alias(virt, phys);
+	__invalidate_icache_page_alias(virt, phys);
+	preempt_enable();
+}
+EXPORT_SYMBOL(local_flush_cache_page);
+
+#endif /* DCACHE_WAY_SIZE > PAGE_SIZE */
+
+void update_mmu_cache_range(struct vm_fault *vmf, struct vm_area_struct *vma,
+		unsigned long addr, pte_t *ptep, unsigned int nr)
+{
+	unsigned long pfn = pte_pfn(*ptep);
+	struct folio *folio;
+	unsigned int i;
+
+	if (!pfn_valid(pfn))
+		return;
+
+	folio = page_folio(pfn_to_page(pfn));
+
+	/* Invalidate old entries in TLBs */
+	for (i = 0; i < nr; i++)
+		flush_tlb_page(vma, addr + i * PAGE_SIZE);
+	nr = folio_nr_pages(folio);
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+	if (!folio_test_reserved(folio) && test_bit(PG_arch_1, &folio->flags.f)) {
+		unsigned long phys = folio_pfn(folio) * PAGE_SIZE;
+		unsigned long tmp;
+
+		preempt_disable();
+		for (i = 0; i < nr; i++) {
+			tmp = TLBTEMP_BASE_1 + (phys & DCACHE_ALIAS_MASK);
+			__flush_invalidate_dcache_page_alias(tmp, phys);
+			tmp = TLBTEMP_BASE_1 + (addr & DCACHE_ALIAS_MASK);
+			__flush_invalidate_dcache_page_alias(tmp, phys);
+			__invalidate_icache_page_alias(tmp, phys);
+			phys += PAGE_SIZE;
+		}
+		preempt_enable();
+
+		clear_bit(PG_arch_1, &folio->flags.f);
+	}
+#else
+	if (!folio_test_reserved(folio) && !test_bit(PG_arch_1, &folio->flags.f)
+	    && (vma->vm_flags & VM_EXEC) != 0) {
+		for (i = 0; i < nr; i++) {
+			void *paddr = kmap_local_folio(folio, i * PAGE_SIZE);
+			__flush_dcache_page((unsigned long)paddr);
+			__invalidate_icache_page((unsigned long)paddr);
+			kunmap_local(paddr);
+		}
+		set_bit(PG_arch_1, &folio->flags.f);
+	}
+#endif
+}
+
+/*
+ * access_process_vm() has called get_user_pages(), which has done a
+ * flush_dcache_page() on the page.
+ */
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+void copy_to_user_page(struct vm_area_struct *vma, struct page *page,
+		unsigned long vaddr, void *dst, const void *src,
+		unsigned long len)
+{
+	unsigned long phys = page_to_phys(page);
+	unsigned long alias = !(DCACHE_ALIAS_EQ(vaddr, phys));
+
+	/* Flush and invalidate user page if aliased. */
+
+	if (alias) {
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+		preempt_disable();
+		__flush_invalidate_dcache_page_alias(t, phys);
+		preempt_enable();
+	}
+
+	/* Copy data */
+	
+	memcpy(dst, src, len);
+
+	/*
+	 * Flush and invalidate kernel page if aliased and synchronize 
+	 * data and instruction caches for executable pages. 
+	 */
+
+	if (alias) {
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+
+		preempt_disable();
+		__flush_invalidate_dcache_range((unsigned long) dst, len);
+		if ((vma->vm_flags & VM_EXEC) != 0)
+			__invalidate_icache_page_alias(t, phys);
+		preempt_enable();
+
+	} else if ((vma->vm_flags & VM_EXEC) != 0) {
+		__flush_dcache_range((unsigned long)dst,len);
+		__invalidate_icache_range((unsigned long) dst, len);
+	}
+}
+
+extern void copy_from_user_page(struct vm_area_struct *vma, struct page *page,
+		unsigned long vaddr, void *dst, const void *src,
+		unsigned long len)
+{
+	unsigned long phys = page_to_phys(page);
+	unsigned long alias = !(DCACHE_ALIAS_EQ(vaddr, phys));
+
+	/*
+	 * Flush user page if aliased. 
+	 * (Note: a simply flush would be sufficient) 
+	 */
+
+	if (alias) {
+		unsigned long t = TLBTEMP_BASE_1 + (vaddr & DCACHE_ALIAS_MASK);
+		preempt_disable();
+		__flush_invalidate_dcache_page_alias(t, phys);
+		preempt_enable();
+	}
+
+	memcpy(dst, src, len);
+}
+
+#endif
diff --git a/arch/xtensa/mm/fault.c b/arch/xtensa/mm/fault.c
new file mode 100644
index 000000000..16e11b6f6
--- /dev/null
+++ b/arch/xtensa/mm/fault.c
@@ -0,0 +1,256 @@
+// TODO VM_EXEC flag work-around, cache aliasing
+/*
+ * arch/xtensa/mm/fault.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2010 Tensilica Inc.
+ *
+ * Chris Zankel <chris@zankel.net>
+ * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
+ */
+
+#include <linux/mm.h>
+#include <linux/extable.h>
+#include <linux/hardirq.h>
+#include <linux/perf_event.h>
+#include <linux/uaccess.h>
+#include <asm/mmu_context.h>
+#include <asm/cacheflush.h>
+#include <asm/hardirq.h>
+#include <asm/traps.h>
+
+void bad_page_fault(struct pt_regs*, unsigned long, int);
+
+static void vmalloc_fault(struct pt_regs *regs, unsigned int address)
+{
+#ifdef CONFIG_MMU
+	/* Synchronize this task's top level page-table
+	 * with the 'reference' page table.
+	 */
+	struct mm_struct *act_mm = current->active_mm;
+	int index = pgd_index(address);
+	pgd_t *pgd, *pgd_k;
+	p4d_t *p4d, *p4d_k;
+	pud_t *pud, *pud_k;
+	pmd_t *pmd, *pmd_k;
+	pte_t *pte_k;
+
+	if (act_mm == NULL)
+		goto bad_page_fault;
+
+	pgd = act_mm->pgd + index;
+	pgd_k = init_mm.pgd + index;
+
+	if (!pgd_present(*pgd_k))
+		goto bad_page_fault;
+
+	pgd_val(*pgd) = pgd_val(*pgd_k);
+
+	p4d = p4d_offset(pgd, address);
+	p4d_k = p4d_offset(pgd_k, address);
+	if (!p4d_present(*p4d) || !p4d_present(*p4d_k))
+		goto bad_page_fault;
+
+	pud = pud_offset(p4d, address);
+	pud_k = pud_offset(p4d_k, address);
+	if (!pud_present(*pud) || !pud_present(*pud_k))
+		goto bad_page_fault;
+
+	pmd = pmd_offset(pud, address);
+	pmd_k = pmd_offset(pud_k, address);
+	if (!pmd_present(*pmd) || !pmd_present(*pmd_k))
+		goto bad_page_fault;
+
+	pmd_val(*pmd) = pmd_val(*pmd_k);
+	pte_k = pte_offset_kernel(pmd_k, address);
+
+	if (!pte_present(*pte_k))
+		goto bad_page_fault;
+	return;
+
+bad_page_fault:
+	bad_page_fault(regs, address, SIGKILL);
+#else
+	WARN_ONCE(1, "%s in noMMU configuration\n", __func__);
+#endif
+}
+/*
+ * This routine handles page faults.  It determines the address,
+ * and the problem, and then passes it off to one of the appropriate
+ * routines.
+ *
+ * Note: does not handle Miss and MultiHit.
+ */
+
+void do_page_fault(struct pt_regs *regs)
+{
+	struct vm_area_struct * vma;
+	struct mm_struct *mm = current->mm;
+	unsigned int exccause = regs->exccause;
+	unsigned int address = regs->excvaddr;
+	int code;
+
+	int is_write, is_exec;
+	vm_fault_t fault;
+	unsigned int flags = FAULT_FLAG_DEFAULT;
+
+	code = SEGV_MAPERR;
+
+	/* We fault-in kernel-space virtual memory on-demand. The
+	 * 'reference' page table is init_mm.pgd.
+	 */
+	if (address >= TASK_SIZE && !user_mode(regs)) {
+		vmalloc_fault(regs, address);
+		return;
+	}
+
+	/* If we're in an interrupt or have no user
+	 * context, we must not take the fault..
+	 */
+	if (faulthandler_disabled() || !mm) {
+		bad_page_fault(regs, address, SIGSEGV);
+		return;
+	}
+
+	is_write = (exccause == EXCCAUSE_STORE_CACHE_ATTRIBUTE) ? 1 : 0;
+	is_exec =  (exccause == EXCCAUSE_ITLB_PRIVILEGE ||
+		    exccause == EXCCAUSE_ITLB_MISS ||
+		    exccause == EXCCAUSE_FETCH_CACHE_ATTRIBUTE) ? 1 : 0;
+
+	pr_debug("[%s:%d:%08x:%d:%08lx:%s%s]\n",
+		 current->comm, current->pid,
+		 address, exccause, regs->pc,
+		 is_write ? "w" : "", is_exec ? "x" : "");
+
+	if (user_mode(regs))
+		flags |= FAULT_FLAG_USER;
+
+	perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS, 1, regs, address);
+
+retry:
+	vma = lock_mm_and_find_vma(mm, address, regs);
+	if (!vma)
+		goto bad_area_nosemaphore;
+
+	/* Ok, we have a good vm_area for this memory access, so
+	 * we can handle it..
+	 */
+
+	code = SEGV_ACCERR;
+
+	if (is_write) {
+		if (!(vma->vm_flags & VM_WRITE))
+			goto bad_area;
+		flags |= FAULT_FLAG_WRITE;
+	} else if (is_exec) {
+		if (!(vma->vm_flags & VM_EXEC))
+			goto bad_area;
+	} else	/* Allow read even from write-only pages. */
+		if (!(vma->vm_flags & (VM_READ | VM_WRITE)))
+			goto bad_area;
+
+	/* If for any reason at all we couldn't handle the fault,
+	 * make sure we exit gracefully rather than endlessly redo
+	 * the fault.
+	 */
+	fault = handle_mm_fault(vma, address, flags, regs);
+
+	if (fault_signal_pending(fault, regs)) {
+		if (!user_mode(regs))
+			bad_page_fault(regs, address, SIGKILL);
+		return;
+	}
+
+	/* The fault is fully completed (including releasing mmap lock) */
+	if (fault & VM_FAULT_COMPLETED)
+		return;
+
+	if (unlikely(fault & VM_FAULT_ERROR)) {
+		if (fault & VM_FAULT_OOM)
+			goto out_of_memory;
+		else if (fault & VM_FAULT_SIGSEGV)
+			goto bad_area;
+		else if (fault & VM_FAULT_SIGBUS)
+			goto do_sigbus;
+		BUG();
+	}
+
+	if (fault & VM_FAULT_RETRY) {
+		flags |= FAULT_FLAG_TRIED;
+
+		/* No need to mmap_read_unlock(mm) as we would
+		 * have already released it in __lock_page_or_retry
+		 * in mm/filemap.c.
+		 */
+
+		goto retry;
+	}
+
+	mmap_read_unlock(mm);
+	return;
+
+	/* Something tried to access memory that isn't in our memory map..
+	 * Fix it, but check if it's kernel or user first..
+	 */
+bad_area:
+	mmap_read_unlock(mm);
+bad_area_nosemaphore:
+	if (user_mode(regs)) {
+		force_sig_fault(SIGSEGV, code, (void *) address);
+		return;
+	}
+	bad_page_fault(regs, address, SIGSEGV);
+	return;
+
+
+	/* We ran out of memory, or some other thing happened to us that made
+	 * us unable to handle the page fault gracefully.
+	 */
+out_of_memory:
+	mmap_read_unlock(mm);
+	if (!user_mode(regs))
+		bad_page_fault(regs, address, SIGKILL);
+	else
+		pagefault_out_of_memory();
+	return;
+
+do_sigbus:
+	mmap_read_unlock(mm);
+
+	/* Send a sigbus, regardless of whether we were in kernel
+	 * or user mode.
+	 */
+	force_sig_fault(SIGBUS, BUS_ADRERR, (void *) address);
+
+	/* Kernel mode? Handle exceptions or die */
+	if (!user_mode(regs))
+		bad_page_fault(regs, address, SIGBUS);
+	return;
+}
+
+
+void
+bad_page_fault(struct pt_regs *regs, unsigned long address, int sig)
+{
+	extern void __noreturn die(const char*, struct pt_regs*, long);
+	const struct exception_table_entry *entry;
+
+	/* Are we prepared to handle this kernel fault?  */
+	if ((entry = search_exception_tables(regs->pc)) != NULL) {
+		pr_debug("%s: Exception at pc=%#010lx (%lx)\n",
+			 current->comm, regs->pc, entry->fixup);
+		regs->pc = entry->fixup;
+		return;
+	}
+
+	/* Oops. The kernel tried to access some bad page. We'll have to
+	 * terminate things with extreme prejudice.
+	 */
+	pr_alert("Unable to handle kernel paging request at virtual "
+		 "address %08lx\n pc = %08lx, ra = %08lx\n",
+		 address, regs->pc, regs->areg[0]);
+	die("Oops", regs, sig);
+}
diff --git a/arch/xtensa/mm/highmem.c b/arch/xtensa/mm/highmem.c
new file mode 100644
index 000000000..35c4f7d4a
--- /dev/null
+++ b/arch/xtensa/mm/highmem.c
@@ -0,0 +1,59 @@
+/*
+ * High memory support for Xtensa architecture
+ *
+ * This file is subject to the terms and conditions of the GNU General
+ * Public License.  See the file "COPYING" in the main directory of
+ * this archive for more details.
+ *
+ * Copyright (C) 2014 Cadence Design Systems Inc.
+ */
+
+#include <linux/export.h>
+#include <linux/highmem.h>
+#include <asm/tlbflush.h>
+
+#if DCACHE_WAY_SIZE > PAGE_SIZE
+unsigned int last_pkmap_nr_arr[DCACHE_N_COLORS];
+wait_queue_head_t pkmap_map_wait_arr[DCACHE_N_COLORS];
+
+static void __init kmap_waitqueues_init(void)
+{
+	unsigned int i;
+
+	for (i = 0; i < ARRAY_SIZE(pkmap_map_wait_arr); ++i)
+		init_waitqueue_head(pkmap_map_wait_arr + i);
+}
+
+static inline enum fixed_addresses kmap_idx(int type, unsigned long color)
+{
+	int idx = (type + KM_MAX_IDX * smp_processor_id()) * DCACHE_N_COLORS;
+
+	/*
+	 * The fixmap operates top down, so the color offset needs to be
+	 * reverse as well.
+	 */
+	return idx + DCACHE_N_COLORS - 1 - color;
+}
+
+enum fixed_addresses kmap_local_map_idx(int type, unsigned long pfn)
+{
+	return kmap_idx(type, DCACHE_ALIAS(pfn << PAGE_SHIFT));
+}
+
+enum fixed_addresses kmap_local_unmap_idx(int type, unsigned long addr)
+{
+	return kmap_idx(type, DCACHE_ALIAS(addr));
+}
+
+#else
+static inline void kmap_waitqueues_init(void) { }
+#endif
+
+void __init kmap_init(void)
+{
+	/* Check if this memory layout is broken because PKMAP overlaps
+	 * page table.
+	 */
+	BUILD_BUG_ON(PKMAP_BASE < TLBTEMP_BASE_1 + TLBTEMP_SIZE);
+	kmap_waitqueues_init();
+}
diff --git a/arch/xtensa/mm/init.c b/arch/xtensa/mm/init.c
new file mode 100644
index 000000000..fe83a6833
--- /dev/null
+++ b/arch/xtensa/mm/init.c
@@ -0,0 +1,202 @@
+/*
+ * arch/xtensa/mm/init.c
+ *
+ * Derived from MIPS, PPC.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2005 Tensilica Inc.
+ * Copyright (C) 2014 - 2016 Cadence Design Systems Inc.
+ *
+ * Chris Zankel	<chris@zankel.net>
+ * Joe Taylor	<joe@tensilica.com, joetylr@yahoo.com>
+ * Marc Gauthier
+ * Kevin Chea
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/memblock.h>
+#include <linux/gfp.h>
+#include <linux/highmem.h>
+#include <linux/swap.h>
+#include <linux/mman.h>
+#include <linux/nodemask.h>
+#include <linux/mm.h>
+#include <linux/of_fdt.h>
+#include <linux/dma-map-ops.h>
+
+#include <asm/bootparam.h>
+#include <asm/page.h>
+#include <asm/sections.h>
+#include <asm/sysmem.h>
+
+/*
+ * Initialize the bootmem system and give it all low memory we have available.
+ */
+
+void __init bootmem_init(void)
+{
+	/* Reserve all memory below PHYS_OFFSET, as memory
+	 * accounting doesn't work for pages below that address.
+	 *
+	 * If PHYS_OFFSET is zero reserve page at address 0:
+	 * successfull allocations should never return NULL.
+	 */
+	memblock_reserve(0, PHYS_OFFSET ? PHYS_OFFSET : 1);
+
+	early_init_fdt_scan_reserved_mem();
+
+	if (!memblock_phys_mem_size())
+		panic("No memory found!\n");
+
+	min_low_pfn = PFN_UP(memblock_start_of_DRAM());
+	min_low_pfn = max(min_low_pfn, PFN_UP(PHYS_OFFSET));
+	max_pfn = PFN_DOWN(memblock_end_of_DRAM());
+	max_low_pfn = min(max_pfn, MAX_LOW_PFN);
+
+	early_memtest((phys_addr_t)min_low_pfn << PAGE_SHIFT,
+		      (phys_addr_t)max_low_pfn << PAGE_SHIFT);
+
+	memblock_set_current_limit(PFN_PHYS(max_low_pfn));
+	dma_contiguous_reserve(PFN_PHYS(max_low_pfn));
+
+	memblock_dump_all();
+}
+
+static void __init print_vm_layout(void)
+{
+	pr_info("virtual kernel memory layout:\n"
+#ifdef CONFIG_KASAN
+		"    kasan   : 0x%08lx - 0x%08lx  (%5lu MB)\n"
+#endif
+#ifdef CONFIG_MMU
+		"    vmalloc : 0x%08lx - 0x%08lx  (%5lu MB)\n"
+#endif
+#ifdef CONFIG_HIGHMEM
+		"    pkmap   : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+		"    fixmap  : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+#endif
+		"    lowmem  : 0x%08lx - 0x%08lx  (%5lu MB)\n"
+		"    .text   : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+		"    .rodata : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+		"    .data   : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+		"    .init   : 0x%08lx - 0x%08lx  (%5lu kB)\n"
+		"    .bss    : 0x%08lx - 0x%08lx  (%5lu kB)\n",
+#ifdef CONFIG_KASAN
+		KASAN_SHADOW_START, KASAN_SHADOW_START + KASAN_SHADOW_SIZE,
+		KASAN_SHADOW_SIZE >> 20,
+#endif
+#ifdef CONFIG_MMU
+		VMALLOC_START, VMALLOC_END,
+		(VMALLOC_END - VMALLOC_START) >> 20,
+#ifdef CONFIG_HIGHMEM
+		PKMAP_BASE, PKMAP_BASE + LAST_PKMAP * PAGE_SIZE,
+		(LAST_PKMAP*PAGE_SIZE) >> 10,
+		FIXADDR_START, FIXADDR_END,
+		(FIXADDR_END - FIXADDR_START) >> 10,
+#endif
+		PAGE_OFFSET, PAGE_OFFSET +
+		(max_low_pfn - min_low_pfn) * PAGE_SIZE,
+#else
+		min_low_pfn * PAGE_SIZE, max_low_pfn * PAGE_SIZE,
+#endif
+		((max_low_pfn - min_low_pfn) * PAGE_SIZE) >> 20,
+		(unsigned long)_text, (unsigned long)_etext,
+		(unsigned long)(_etext - _text) >> 10,
+		(unsigned long)__start_rodata, (unsigned long)__end_rodata,
+		(unsigned long)(__end_rodata - __start_rodata) >> 10,
+		(unsigned long)_sdata, (unsigned long)_edata,
+		(unsigned long)(_edata - _sdata) >> 10,
+		(unsigned long)__init_begin, (unsigned long)__init_end,
+		(unsigned long)(__init_end - __init_begin) >> 10,
+		(unsigned long)__bss_start, (unsigned long)__bss_stop,
+		(unsigned long)(__bss_stop - __bss_start) >> 10);
+}
+
+void __init arch_zone_limits_init(unsigned long *max_zone_pfns)
+{
+	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
+#ifdef CONFIG_HIGHMEM
+	max_zone_pfns[ZONE_HIGHMEM] = max_pfn;
+#endif
+}
+
+void __init zones_init(void)
+{
+	print_vm_layout();
+}
+
+static void __init parse_memmap_one(char *p)
+{
+	char *oldp;
+	unsigned long start_at, mem_size;
+
+	if (!p)
+		return;
+
+	oldp = p;
+	mem_size = memparse(p, &p);
+	if (p == oldp)
+		return;
+
+	switch (*p) {
+	case '@':
+		start_at = memparse(p + 1, &p);
+		memblock_add(start_at, mem_size);
+		break;
+
+	case '$':
+		start_at = memparse(p + 1, &p);
+		memblock_reserve(start_at, mem_size);
+		break;
+
+	case 0:
+		memblock_reserve(mem_size, -mem_size);
+		break;
+
+	default:
+		pr_warn("Unrecognized memmap syntax: %s\n", p);
+		break;
+	}
+}
+
+static int __init parse_memmap_opt(char *str)
+{
+	while (str) {
+		char *k = strchr(str, ',');
+
+		if (k)
+			*k++ = 0;
+
+		parse_memmap_one(str);
+		str = k;
+	}
+
+	return 0;
+}
+early_param("memmap", parse_memmap_opt);
+
+#ifdef CONFIG_MMU
+static const pgprot_t protection_map[16] = {
+	[VM_NONE]					= PAGE_NONE,
+	[VM_READ]					= PAGE_READONLY,
+	[VM_WRITE]					= PAGE_COPY,
+	[VM_WRITE | VM_READ]				= PAGE_COPY,
+	[VM_EXEC]					= PAGE_READONLY_EXEC,
+	[VM_EXEC | VM_READ]				= PAGE_READONLY_EXEC,
+	[VM_EXEC | VM_WRITE]				= PAGE_COPY_EXEC,
+	[VM_EXEC | VM_WRITE | VM_READ]			= PAGE_COPY_EXEC,
+	[VM_SHARED]					= PAGE_NONE,
+	[VM_SHARED | VM_READ]				= PAGE_READONLY,
+	[VM_SHARED | VM_WRITE]				= PAGE_SHARED,
+	[VM_SHARED | VM_WRITE | VM_READ]		= PAGE_SHARED,
+	[VM_SHARED | VM_EXEC]				= PAGE_READONLY_EXEC,
+	[VM_SHARED | VM_EXEC | VM_READ]			= PAGE_READONLY_EXEC,
+	[VM_SHARED | VM_EXEC | VM_WRITE]		= PAGE_SHARED_EXEC,
+	[VM_SHARED | VM_EXEC | VM_WRITE | VM_READ]	= PAGE_SHARED_EXEC
+};
+DECLARE_VM_GET_PAGE_PROT
+#endif
diff --git a/arch/xtensa/mm/ioremap.c b/arch/xtensa/mm/ioremap.c
new file mode 100644
index 000000000..26f238fa9
--- /dev/null
+++ b/arch/xtensa/mm/ioremap.c
@@ -0,0 +1,35 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * ioremap implementation.
+ *
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+
+#include <linux/io.h>
+#include <linux/pgtable.h>
+#include <asm/cacheflush.h>
+#include <asm/io.h>
+
+void __iomem *ioremap_prot(phys_addr_t phys_addr, size_t size,
+			   pgprot_t prot)
+{
+	unsigned long pfn = __phys_to_pfn((phys_addr));
+	WARN_ON(pfn_valid(pfn));
+
+	return generic_ioremap_prot(phys_addr, size, prot);
+}
+EXPORT_SYMBOL(ioremap_prot);
+
+void iounmap(volatile void __iomem *addr)
+{
+	unsigned long va = (unsigned long) addr;
+
+	if ((va >= XCHAL_KIO_CACHED_VADDR &&
+	      va - XCHAL_KIO_CACHED_VADDR < XCHAL_KIO_SIZE) ||
+	    (va >= XCHAL_KIO_BYPASS_VADDR &&
+	      va - XCHAL_KIO_BYPASS_VADDR < XCHAL_KIO_SIZE))
+		return;
+
+	generic_iounmap(addr);
+}
+EXPORT_SYMBOL(iounmap);
diff --git a/arch/xtensa/mm/kasan_init.c b/arch/xtensa/mm/kasan_init.c
new file mode 100644
index 000000000..0524b9ed5
--- /dev/null
+++ b/arch/xtensa/mm/kasan_init.c
@@ -0,0 +1,98 @@
+/*
+ * Xtensa KASAN shadow map initialization
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2017 Cadence Design Systems Inc.
+ */
+
+#include <linux/memblock.h>
+#include <linux/init_task.h>
+#include <linux/kasan.h>
+#include <linux/kernel.h>
+#include <asm/initialize_mmu.h>
+#include <asm/tlbflush.h>
+
+void __init kasan_early_init(void)
+{
+	unsigned long vaddr = KASAN_SHADOW_START;
+	pmd_t *pmd = pmd_off_k(vaddr);
+	int i;
+
+	for (i = 0; i < PTRS_PER_PTE; ++i)
+		set_pte(kasan_early_shadow_pte + i,
+			mk_pte(virt_to_page(kasan_early_shadow_page),
+				PAGE_KERNEL));
+
+	for (vaddr = 0; vaddr < KASAN_SHADOW_SIZE; vaddr += PMD_SIZE, ++pmd) {
+		BUG_ON(!pmd_none(*pmd));
+		set_pmd(pmd, __pmd((unsigned long)kasan_early_shadow_pte));
+	}
+}
+
+static void __init populate(void *start, void *end)
+{
+	unsigned long n_pages = (end - start) / PAGE_SIZE;
+	unsigned long n_pmds = n_pages / PTRS_PER_PTE;
+	unsigned long i, j;
+	unsigned long vaddr = (unsigned long)start;
+	pmd_t *pmd = pmd_off_k(vaddr);
+	pte_t *pte = memblock_alloc_or_panic(n_pages * sizeof(pte_t), PAGE_SIZE);
+
+	pr_debug("%s: %p - %p\n", __func__, start, end);
+
+	for (i = j = 0; i < n_pmds; ++i) {
+		int k;
+
+		for (k = 0; k < PTRS_PER_PTE; ++k, ++j) {
+			phys_addr_t phys =
+				memblock_phys_alloc_range(PAGE_SIZE, PAGE_SIZE,
+							  0,
+							  MEMBLOCK_ALLOC_ANYWHERE);
+
+			if (!phys)
+				panic("Failed to allocate page table page\n");
+
+			set_pte(pte + j, pfn_pte(PHYS_PFN(phys), PAGE_KERNEL));
+		}
+	}
+
+	for (i = 0; i < n_pmds ; ++i, pte += PTRS_PER_PTE)
+		set_pmd(pmd + i, __pmd((unsigned long)pte));
+
+	local_flush_tlb_all();
+	memset(start, 0, end - start);
+}
+
+void __init kasan_init(void)
+{
+	int i;
+
+	BUILD_BUG_ON(KASAN_SHADOW_OFFSET != KASAN_SHADOW_START -
+		     (KASAN_START_VADDR >> KASAN_SHADOW_SCALE_SHIFT));
+	BUILD_BUG_ON(VMALLOC_START < KASAN_START_VADDR);
+
+	/*
+	 * Replace shadow map pages that cover addresses from VMALLOC area
+	 * start to the end of KSEG with clean writable pages.
+	 */
+	populate(kasan_mem_to_shadow((void *)VMALLOC_START),
+		 kasan_mem_to_shadow((void *)XCHAL_KSEG_BYPASS_VADDR));
+
+	/*
+	 * Write protect kasan_early_shadow_page and zero-initialize it again.
+	 */
+	for (i = 0; i < PTRS_PER_PTE; ++i)
+		set_pte(kasan_early_shadow_pte + i,
+			mk_pte(virt_to_page(kasan_early_shadow_page),
+				PAGE_KERNEL_RO));
+
+	local_flush_tlb_all();
+	memset(kasan_early_shadow_page, 0, PAGE_SIZE);
+
+	/* At this point kasan is fully initialized. Enable error messages. */
+	current->kasan_depth = 0;
+	kasan_init_generic();
+}
diff --git a/arch/xtensa/mm/misc.S b/arch/xtensa/mm/misc.S
new file mode 100644
index 000000000..ec36f73c4
--- /dev/null
+++ b/arch/xtensa/mm/misc.S
@@ -0,0 +1,463 @@
+/*
+ * arch/xtensa/mm/misc.S
+ *
+ * Miscellaneous assembly functions.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2007 Tensilica Inc.
+ *
+ * Chris Zankel	<chris@zankel.net>
+ */
+
+
+#include <linux/linkage.h>
+#include <linux/pgtable.h>
+#include <asm/page.h>
+#include <asm/asmmacro.h>
+#include <asm/cacheasm.h>
+#include <asm/tlbflush.h>
+
+
+/*
+ * clear_page and clear_user_page are the same for non-cache-aliased configs.
+ *
+ * clear_page (unsigned long page)
+ *                    a2
+ */
+
+ENTRY(clear_page)
+
+	abi_entry_default
+
+	movi	a3, 0
+	__loopi	a2, a7, PAGE_SIZE, 32
+	s32i	a3, a2, 0
+	s32i	a3, a2, 4
+	s32i	a3, a2, 8
+	s32i	a3, a2, 12
+	s32i	a3, a2, 16
+	s32i	a3, a2, 20
+	s32i	a3, a2, 24
+	s32i	a3, a2, 28
+	__endla	a2, a7, 32
+
+	abi_ret_default
+
+ENDPROC(clear_page)
+EXPORT_SYMBOL(clear_page)
+
+/*
+ * copy_page and copy_user_page are the same for non-cache-aliased configs.
+ *
+ * copy_page (void *to, void *from)
+ *               a2          a3
+ */
+
+ENTRY(copy_page)
+
+	abi_entry_default
+
+	__loopi a2, a4, PAGE_SIZE, 32
+
+	l32i    a8, a3, 0
+	l32i    a9, a3, 4
+	s32i    a8, a2, 0
+	s32i    a9, a2, 4
+
+	l32i    a8, a3, 8
+	l32i    a9, a3, 12
+	s32i    a8, a2, 8
+	s32i    a9, a2, 12
+
+	l32i    a8, a3, 16
+	l32i    a9, a3, 20
+	s32i    a8, a2, 16
+	s32i    a9, a2, 20
+
+	l32i    a8, a3, 24
+	l32i    a9, a3, 28
+	s32i    a8, a2, 24
+	s32i    a9, a2, 28
+
+	addi    a2, a2, 32
+	addi    a3, a3, 32
+
+	__endl  a2, a4
+
+	abi_ret_default
+
+ENDPROC(copy_page)
+EXPORT_SYMBOL(copy_page)
+
+#ifdef CONFIG_MMU
+/*
+ * If we have to deal with cache aliasing, we use temporary memory mappings
+ * to ensure that the source and destination pages have the same color as
+ * the virtual address. We use way 0 and 1 for temporary mappings in such cases.
+ *
+ * The temporary DTLB entries shouldn't be flushed by interrupts, but are
+ * flushed by preemptive task switches. Special code in the 
+ * fast_second_level_miss handler re-established the temporary mapping. 
+ * It requires that the PPNs for the destination and source addresses are
+ * in a6, and a7, respectively.
+ */
+
+/* TLB miss exceptions are treated special in the following region */
+
+ENTRY(__tlbtemp_mapping_start)
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+/*
+ * clear_page_alias(void *addr, unsigned long paddr)
+ *                     a2              a3
+ */
+
+ENTRY(clear_page_alias)
+
+	abi_entry_default
+
+	movi	a5, PAGE_OFFSET
+	addi	a6, a3, (PAGE_KERNEL | _PAGE_HW_WRITE)
+	mov	a4, a2
+	wdtlb	a6, a2
+	dsync
+
+	movi	a3, 0
+	__loopi	a2, a7, PAGE_SIZE, 32
+	s32i	a3, a2, 0
+	s32i	a3, a2, 4
+	s32i	a3, a2, 8
+	s32i	a3, a2, 12
+	s32i	a3, a2, 16
+	s32i	a3, a2, 20
+	s32i	a3, a2, 24
+	s32i	a3, a2, 28
+	__endla	a2, a7, 32
+
+	/* We need to invalidate the temporary dtlb entry. */
+
+	idtlb	a4
+	dsync
+
+	abi_ret_default
+
+ENDPROC(clear_page_alias)
+
+/*
+ * copy_page_alias(void *to, void *from,
+ *			a2	  a3
+ *                 unsigned long to_paddr, unsigned long from_paddr)
+ *	        		 a4			 a5
+ */
+
+ENTRY(copy_page_alias)
+
+	abi_entry_default
+
+	/* Setup a temporary DTLB for destination. */
+
+	addi	a6, a4, (PAGE_KERNEL | _PAGE_HW_WRITE)
+	wdtlb	a6, a2
+	dsync
+
+	/* Setup a temporary DTLB for source. */
+
+	addi	a7, a5, PAGE_KERNEL
+	addi	a8, a3, 1				# way1
+
+	wdtlb	a7, a8
+	dsync
+
+1:	__loopi a2, a4, PAGE_SIZE, 32
+
+	l32i    a8, a3, 0
+	l32i    a9, a3, 4
+	s32i    a8, a2, 0
+	s32i    a9, a2, 4
+
+	l32i    a8, a3, 8
+	l32i    a9, a3, 12
+	s32i    a8, a2, 8
+	s32i    a9, a2, 12
+
+	l32i    a8, a3, 16
+	l32i    a9, a3, 20
+	s32i    a8, a2, 16
+	s32i    a9, a2, 20
+
+	l32i    a8, a3, 24
+	l32i    a9, a3, 28
+	s32i    a8, a2, 24
+	s32i    a9, a2, 28
+
+	addi    a2, a2, 32
+	addi    a3, a3, 32
+
+	__endl  a2, a4
+
+	/* We need to invalidate any temporary mapping! */
+
+	addi	a2, a2, -PAGE_SIZE
+	idtlb	a2
+	dsync
+
+	addi	a3, a3, -PAGE_SIZE+1
+	idtlb	a3
+	dsync
+
+	abi_ret_default
+
+ENDPROC(copy_page_alias)
+
+#endif
+
+#if (DCACHE_WAY_SIZE > PAGE_SIZE)
+
+/*
+ * void __flush_invalidate_dcache_page_alias (addr, phys)
+ *                                             a2    a3
+ */
+
+ENTRY(__flush_invalidate_dcache_page_alias)
+
+	abi_entry_default
+
+	movi	a7, 0			# required for exception handler
+	addi	a6, a3, (PAGE_KERNEL | _PAGE_HW_WRITE)
+	mov	a4, a2
+	wdtlb	a6, a2
+	dsync
+
+	___flush_invalidate_dcache_page a2 a3
+
+	idtlb	a4
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__flush_invalidate_dcache_page_alias)
+
+/*
+ * void __invalidate_dcache_page_alias (addr, phys)
+ *                                       a2    a3
+ */
+
+ENTRY(__invalidate_dcache_page_alias)
+
+	abi_entry_default
+
+	movi	a7, 0			# required for exception handler
+	addi	a6, a3, (PAGE_KERNEL | _PAGE_HW_WRITE)
+	mov	a4, a2
+	wdtlb	a6, a2
+	dsync
+
+	___invalidate_dcache_page a2 a3
+
+	idtlb	a4
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_dcache_page_alias)
+#endif
+
+ENTRY(__tlbtemp_mapping_itlb)
+
+#if (ICACHE_WAY_SIZE > PAGE_SIZE)
+	
+ENTRY(__invalidate_icache_page_alias)
+
+	abi_entry_default
+
+	addi	a6, a3, (PAGE_KERNEL_EXEC | _PAGE_HW_WRITE)
+	mov	a4, a2
+	witlb	a6, a2
+	isync
+
+	___invalidate_icache_page a2 a3
+
+	iitlb	a4
+	isync
+	abi_ret_default
+
+ENDPROC(__invalidate_icache_page_alias)
+
+#endif
+
+/* End of special treatment in tlb miss exception */
+
+ENTRY(__tlbtemp_mapping_end)
+
+#endif /* CONFIG_MMU
+
+/*
+ * void __invalidate_icache_page(ulong start)
+ */
+
+ENTRY(__invalidate_icache_page)
+
+	abi_entry_default
+
+	___invalidate_icache_page a2 a3
+	isync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_icache_page)
+
+/*
+ * void __invalidate_dcache_page(ulong start)
+ */
+
+ENTRY(__invalidate_dcache_page)
+
+	abi_entry_default
+
+	___invalidate_dcache_page a2 a3
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_dcache_page)
+
+/*
+ * void __flush_invalidate_dcache_page(ulong start)
+ */
+
+ENTRY(__flush_invalidate_dcache_page)
+
+	abi_entry_default
+
+	___flush_invalidate_dcache_page a2 a3
+
+	dsync
+	abi_ret_default
+
+ENDPROC(__flush_invalidate_dcache_page)
+
+/*
+ * void __flush_dcache_page(ulong start)
+ */
+
+ENTRY(__flush_dcache_page)
+
+	abi_entry_default
+
+	___flush_dcache_page a2 a3
+
+	dsync
+	abi_ret_default
+
+ENDPROC(__flush_dcache_page)
+
+/*
+ * void __invalidate_icache_range(ulong start, ulong size)
+ */
+
+ENTRY(__invalidate_icache_range)
+
+	abi_entry_default
+
+	___invalidate_icache_range a2 a3 a4
+	isync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_icache_range)
+EXPORT_SYMBOL(__invalidate_icache_range)
+
+/*
+ * void __flush_invalidate_dcache_range(ulong start, ulong size)
+ */
+
+ENTRY(__flush_invalidate_dcache_range)
+
+	abi_entry_default
+
+	___flush_invalidate_dcache_range a2 a3 a4
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__flush_invalidate_dcache_range)
+
+/*
+ * void _flush_dcache_range(ulong start, ulong size)
+ */
+
+ENTRY(__flush_dcache_range)
+
+	abi_entry_default
+
+	___flush_dcache_range a2 a3 a4
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__flush_dcache_range)
+EXPORT_SYMBOL(__flush_dcache_range)
+
+/*
+ * void _invalidate_dcache_range(ulong start, ulong size)
+ */
+
+ENTRY(__invalidate_dcache_range)
+
+	abi_entry_default
+
+	___invalidate_dcache_range a2 a3 a4
+
+	abi_ret_default
+
+ENDPROC(__invalidate_dcache_range)
+EXPORT_SYMBOL(__invalidate_dcache_range)
+
+/*
+ * void _invalidate_icache_all(void)
+ */
+
+ENTRY(__invalidate_icache_all)
+
+	abi_entry_default
+
+	___invalidate_icache_all a2 a3
+	isync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_icache_all)
+
+/*
+ * void _flush_invalidate_dcache_all(void)
+ */
+
+ENTRY(__flush_invalidate_dcache_all)
+
+	abi_entry_default
+
+	___flush_invalidate_dcache_all a2 a3
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__flush_invalidate_dcache_all)
+
+/*
+ * void _invalidate_dcache_all(void)
+ */
+
+ENTRY(__invalidate_dcache_all)
+
+	abi_entry_default
+
+	___invalidate_dcache_all a2 a3
+	dsync
+
+	abi_ret_default
+
+ENDPROC(__invalidate_dcache_all)
diff --git a/arch/xtensa/mm/mmu.c b/arch/xtensa/mm/mmu.c
new file mode 100644
index 000000000..92e158c69
--- /dev/null
+++ b/arch/xtensa/mm/mmu.c
@@ -0,0 +1,119 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * xtensa mmu stuff
+ *
+ * Extracted from init.c
+ */
+#include <linux/memblock.h>
+#include <linux/percpu.h>
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/slab.h>
+#include <linux/cache.h>
+
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+#include <asm/mmu_context.h>
+#include <asm/page.h>
+#include <asm/initialize_mmu.h>
+#include <asm/io.h>
+
+DEFINE_PER_CPU(unsigned long, asid_cache) = ASID_USER_FIRST;
+
+#if defined(CONFIG_HIGHMEM)
+static void * __init init_pmd(unsigned long vaddr, unsigned long n_pages)
+{
+	pmd_t *pmd = pmd_off_k(vaddr);
+	pte_t *pte;
+	unsigned long i;
+
+	n_pages = ALIGN(n_pages, PTRS_PER_PTE);
+
+	pr_debug("%s: vaddr: 0x%08lx, n_pages: %ld\n",
+		 __func__, vaddr, n_pages);
+
+	pte = memblock_alloc_low(n_pages * sizeof(pte_t), PAGE_SIZE);
+	if (!pte)
+		panic("%s: Failed to allocate %lu bytes align=%lx\n",
+		      __func__, n_pages * sizeof(pte_t), PAGE_SIZE);
+
+	for (i = 0; i < n_pages; ++i)
+		pte_clear(NULL, 0, pte + i);
+
+	for (i = 0; i < n_pages; i += PTRS_PER_PTE, ++pmd) {
+		pte_t *cur_pte = pte + i;
+
+		BUG_ON(!pmd_none(*pmd));
+		set_pmd(pmd, __pmd(((unsigned long)cur_pte) & PAGE_MASK));
+		BUG_ON(cur_pte != pte_offset_kernel(pmd, 0));
+		pr_debug("%s: pmd: 0x%p, pte: 0x%p\n",
+			 __func__, pmd, cur_pte);
+	}
+	return pte;
+}
+
+static void __init fixedrange_init(void)
+{
+	BUILD_BUG_ON(FIXADDR_START < TLBTEMP_BASE_1 + TLBTEMP_SIZE);
+	init_pmd(FIXADDR_START, __end_of_fixed_addresses);
+}
+#endif
+
+void __init paging_init(void)
+{
+#ifdef CONFIG_HIGHMEM
+	fixedrange_init();
+	pkmap_page_table = init_pmd(PKMAP_BASE, LAST_PKMAP);
+	kmap_init();
+#endif
+}
+
+/*
+ * Flush the mmu and reset associated register to default values.
+ */
+void init_mmu(void)
+{
+#if !(XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY)
+	/*
+	 * Writing zeros to the instruction and data TLBCFG special
+	 * registers ensure that valid values exist in the register.
+	 *
+	 * For existing PGSZID<w> fields, zero selects the first element
+	 * of the page-size array.  For nonexistent PGSZID<w> fields,
+	 * zero is the best value to write.  Also, when changing PGSZID<w>
+	 * fields, the corresponding TLB must be flushed.
+	 */
+	set_itlbcfg_register(0);
+	set_dtlbcfg_register(0);
+#endif
+	init_kio();
+	local_flush_tlb_all();
+
+	/* Set rasid register to a known value. */
+
+	set_rasid_register(ASID_INSERT(ASID_USER_FIRST));
+
+	/* Set PTEVADDR special register to the start of the page
+	 * table, which is in kernel mappable space (ie. not
+	 * statically mapped).  This register's value is undefined on
+	 * reset.
+	 */
+	set_ptevaddr_register(XCHAL_PAGE_TABLE_VADDR);
+}
+
+void init_kio(void)
+{
+#if XCHAL_HAVE_PTP_MMU && XCHAL_HAVE_SPANNING_WAY && defined(CONFIG_USE_OF)
+	/*
+	 * Update the IO area mapping in case xtensa_kio_paddr has changed
+	 */
+	write_dtlb_entry(__pte(xtensa_kio_paddr + CA_WRITEBACK),
+			XCHAL_KIO_CACHED_VADDR + 6);
+	write_itlb_entry(__pte(xtensa_kio_paddr + CA_WRITEBACK),
+			XCHAL_KIO_CACHED_VADDR + 6);
+	write_dtlb_entry(__pte(xtensa_kio_paddr + CA_BYPASS),
+			XCHAL_KIO_BYPASS_VADDR + 6);
+	write_itlb_entry(__pte(xtensa_kio_paddr + CA_BYPASS),
+			XCHAL_KIO_BYPASS_VADDR + 6);
+#endif
+}
diff --git a/arch/xtensa/mm/tlb.c b/arch/xtensa/mm/tlb.c
new file mode 100644
index 000000000..0a1a815dc
--- /dev/null
+++ b/arch/xtensa/mm/tlb.c
@@ -0,0 +1,295 @@
+/*
+ * arch/xtensa/mm/tlb.c
+ *
+ * Logic that manipulates the Xtensa MMU.  Derived from MIPS.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 - 2003 Tensilica Inc.
+ *
+ * Joe Taylor
+ * Chris Zankel	<chris@zankel.net>
+ * Marc Gauthier
+ */
+
+#include <linux/mm.h>
+#include <asm/processor.h>
+#include <asm/mmu_context.h>
+#include <asm/tlb.h>
+#include <asm/tlbflush.h>
+#include <asm/cacheflush.h>
+
+
+static inline void __flush_itlb_all (void)
+{
+	int w, i;
+
+	for (w = 0; w < ITLB_ARF_WAYS; w++) {
+		for (i = 0; i < (1 << XCHAL_ITLB_ARF_ENTRIES_LOG2); i++) {
+			int e = w + (i << PAGE_SHIFT);
+			invalidate_itlb_entry_no_isync(e);
+		}
+	}
+	asm volatile ("isync\n");
+}
+
+static inline void __flush_dtlb_all (void)
+{
+	int w, i;
+
+	for (w = 0; w < DTLB_ARF_WAYS; w++) {
+		for (i = 0; i < (1 << XCHAL_DTLB_ARF_ENTRIES_LOG2); i++) {
+			int e = w + (i << PAGE_SHIFT);
+			invalidate_dtlb_entry_no_isync(e);
+		}
+	}
+	asm volatile ("isync\n");
+}
+
+
+void local_flush_tlb_all(void)
+{
+	__flush_itlb_all();
+	__flush_dtlb_all();
+}
+
+/* If mm is current, we simply assign the current task a new ASID, thus,
+ * invalidating all previous tlb entries. If mm is someone else's user mapping,
+ * wie invalidate the context, thus, when that user mapping is swapped in,
+ * a new context will be assigned to it.
+ */
+
+void local_flush_tlb_mm(struct mm_struct *mm)
+{
+	int cpu = smp_processor_id();
+
+	if (mm == current->active_mm) {
+		unsigned long flags;
+		local_irq_save(flags);
+		mm->context.asid[cpu] = NO_CONTEXT;
+		activate_context(mm, cpu);
+		local_irq_restore(flags);
+	} else {
+		mm->context.asid[cpu] = NO_CONTEXT;
+		mm->context.cpu = -1;
+	}
+}
+
+
+#define _ITLB_ENTRIES (ITLB_ARF_WAYS << XCHAL_ITLB_ARF_ENTRIES_LOG2)
+#define _DTLB_ENTRIES (DTLB_ARF_WAYS << XCHAL_DTLB_ARF_ENTRIES_LOG2)
+#if _ITLB_ENTRIES > _DTLB_ENTRIES
+# define _TLB_ENTRIES _ITLB_ENTRIES
+#else
+# define _TLB_ENTRIES _DTLB_ENTRIES
+#endif
+
+void local_flush_tlb_range(struct vm_area_struct *vma,
+		unsigned long start, unsigned long end)
+{
+	int cpu = smp_processor_id();
+	struct mm_struct *mm = vma->vm_mm;
+	unsigned long flags;
+
+	if (mm->context.asid[cpu] == NO_CONTEXT)
+		return;
+
+	pr_debug("[tlbrange<%02lx,%08lx,%08lx>]\n",
+		 (unsigned long)mm->context.asid[cpu], start, end);
+	local_irq_save(flags);
+
+	if (end-start + (PAGE_SIZE-1) <= _TLB_ENTRIES << PAGE_SHIFT) {
+		int oldpid = get_rasid_register();
+
+		set_rasid_register(ASID_INSERT(mm->context.asid[cpu]));
+		start &= PAGE_MASK;
+		if (vma->vm_flags & VM_EXEC)
+			while(start < end) {
+				invalidate_itlb_mapping(start);
+				invalidate_dtlb_mapping(start);
+				start += PAGE_SIZE;
+			}
+		else
+			while(start < end) {
+				invalidate_dtlb_mapping(start);
+				start += PAGE_SIZE;
+			}
+
+		set_rasid_register(oldpid);
+	} else {
+		local_flush_tlb_mm(mm);
+	}
+	local_irq_restore(flags);
+}
+
+void local_flush_tlb_page(struct vm_area_struct *vma, unsigned long page)
+{
+	int cpu = smp_processor_id();
+	struct mm_struct* mm = vma->vm_mm;
+	unsigned long flags;
+	int oldpid;
+
+	if (mm->context.asid[cpu] == NO_CONTEXT)
+		return;
+
+	local_irq_save(flags);
+
+	oldpid = get_rasid_register();
+	set_rasid_register(ASID_INSERT(mm->context.asid[cpu]));
+
+	if (vma->vm_flags & VM_EXEC)
+		invalidate_itlb_mapping(page);
+	invalidate_dtlb_mapping(page);
+
+	set_rasid_register(oldpid);
+
+	local_irq_restore(flags);
+}
+
+void local_flush_tlb_kernel_range(unsigned long start, unsigned long end)
+{
+	if (end > start && start >= TASK_SIZE && end <= PAGE_OFFSET &&
+	    end - start < _TLB_ENTRIES << PAGE_SHIFT) {
+		start &= PAGE_MASK;
+		while (start < end) {
+			invalidate_itlb_mapping(start);
+			invalidate_dtlb_mapping(start);
+			start += PAGE_SIZE;
+		}
+	} else {
+		local_flush_tlb_all();
+	}
+}
+
+void update_mmu_tlb_range(struct vm_area_struct *vma,
+			unsigned long address, pte_t *ptep, unsigned int nr)
+{
+	local_flush_tlb_range(vma, address, address + PAGE_SIZE * nr);
+}
+
+#ifdef CONFIG_DEBUG_TLB_SANITY
+
+static unsigned get_pte_for_vaddr(unsigned vaddr)
+{
+	struct task_struct *task = get_current();
+	struct mm_struct *mm = task->mm;
+	pgd_t *pgd;
+	p4d_t *p4d;
+	pud_t *pud;
+	pmd_t *pmd;
+	pte_t *pte;
+	unsigned int pteval;
+
+	if (!mm)
+		mm = task->active_mm;
+	pgd = pgd_offset(mm, vaddr);
+	if (pgd_none_or_clear_bad(pgd))
+		return 0;
+	p4d = p4d_offset(pgd, vaddr);
+	if (p4d_none_or_clear_bad(p4d))
+		return 0;
+	pud = pud_offset(p4d, vaddr);
+	if (pud_none_or_clear_bad(pud))
+		return 0;
+	pmd = pmd_offset(pud, vaddr);
+	if (pmd_none_or_clear_bad(pmd))
+		return 0;
+	pte = pte_offset_map(pmd, vaddr);
+	if (!pte)
+		return 0;
+	pteval = pte_val(*pte);
+	pte_unmap(pte);
+	return pteval;
+}
+
+enum {
+	TLB_SUSPICIOUS	= 1,
+	TLB_INSANE	= 2,
+};
+
+static void tlb_insane(void)
+{
+	BUG_ON(1);
+}
+
+static void tlb_suspicious(void)
+{
+	WARN_ON(1);
+}
+
+/*
+ * Check that TLB entries with kernel ASID (1) have kernel VMA (>= TASK_SIZE),
+ * and TLB entries with user ASID (>=4) have VMA < TASK_SIZE.
+ *
+ * Check that valid TLB entries either have the same PA as the PTE, or PTE is
+ * marked as non-present. Non-present PTE and the page with non-zero refcount
+ * and zero mapcount is normal for batched TLB flush operation. Zero refcount
+ * means that the page was freed prematurely. Non-zero mapcount is unusual,
+ * but does not necessary means an error, thus marked as suspicious.
+ */
+static int check_tlb_entry(unsigned w, unsigned e, bool dtlb)
+{
+	unsigned tlbidx = w | (e << PAGE_SHIFT);
+	unsigned r0 = dtlb ?
+		read_dtlb_virtual(tlbidx) : read_itlb_virtual(tlbidx);
+	unsigned r1 = dtlb ?
+		read_dtlb_translation(tlbidx) : read_itlb_translation(tlbidx);
+	unsigned vpn = (r0 & PAGE_MASK) | (e << PAGE_SHIFT);
+	unsigned pte = get_pte_for_vaddr(vpn);
+	unsigned mm_asid = (get_rasid_register() >> 8) & ASID_MASK;
+	unsigned tlb_asid = r0 & ASID_MASK;
+	bool kernel = tlb_asid == 1;
+	int rc = 0;
+
+	if (tlb_asid > 0 && ((vpn < TASK_SIZE) == kernel)) {
+		pr_err("%cTLB: way: %u, entry: %u, VPN %08x in %s PTE\n",
+				dtlb ? 'D' : 'I', w, e, vpn,
+				kernel ? "kernel" : "user");
+		rc |= TLB_INSANE;
+	}
+
+	if (tlb_asid == mm_asid) {
+		if ((pte ^ r1) & PAGE_MASK) {
+			pr_err("%cTLB: way: %u, entry: %u, mapping: %08x->%08x, PTE: %08x\n",
+					dtlb ? 'D' : 'I', w, e, r0, r1, pte);
+			if (pte == 0 || !pte_present(__pte(pte))) {
+				struct page *p = pfn_to_page(r1 >> PAGE_SHIFT);
+				struct folio *f = page_folio(p);
+
+				pr_err("folio refcount: %d, mapcount: %d\n",
+					folio_ref_count(f), folio_mapcount(f));
+				if (!folio_ref_count(f))
+					rc |= TLB_INSANE;
+				else if (folio_mapped(f))
+					rc |= TLB_SUSPICIOUS;
+			} else {
+				rc |= TLB_INSANE;
+			}
+		}
+	}
+	return rc;
+}
+
+void check_tlb_sanity(void)
+{
+	unsigned long flags;
+	unsigned w, e;
+	int bug = 0;
+
+	local_irq_save(flags);
+	for (w = 0; w < DTLB_ARF_WAYS; ++w)
+		for (e = 0; e < (1 << XCHAL_DTLB_ARF_ENTRIES_LOG2); ++e)
+			bug |= check_tlb_entry(w, e, true);
+	for (w = 0; w < ITLB_ARF_WAYS; ++w)
+		for (e = 0; e < (1 << XCHAL_ITLB_ARF_ENTRIES_LOG2); ++e)
+			bug |= check_tlb_entry(w, e, false);
+	if (bug & TLB_INSANE)
+		tlb_insane();
+	if (bug & TLB_SUSPICIOUS)
+		tlb_suspicious();
+	local_irq_restore(flags);
+}
+
+#endif /* CONFIG_DEBUG_TLB_SANITY */
diff --git a/arch/xtensa/platforms/Makefile b/arch/xtensa/platforms/Makefile
new file mode 100644
index 000000000..e2e7e0726
--- /dev/null
+++ b/arch/xtensa/platforms/Makefile
@@ -0,0 +1,4 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-$(CONFIG_XTENSA_PLATFORM_XT2000)	+= xt2000/
+obj-$(CONFIG_XTENSA_PLATFORM_ISS)	+= iss/
+obj-$(CONFIG_XTENSA_PLATFORM_XTFPGA)	+= xtfpga/
diff --git a/arch/xtensa/platforms/iss/Makefile b/arch/xtensa/platforms/iss/Makefile
new file mode 100644
index 000000000..f3dd5e72a
--- /dev/null
+++ b/arch/xtensa/platforms/iss/Makefile
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# $Id: Makefile,v 1.1.1.1 2002/08/28 16:10:14 aroll Exp $
+#
+# Makefile for the Xtensa Instruction Set Simulator (ISS)
+# "prom monitor" library routines under Linux.
+#
+
+obj-y			= setup.o
+obj-$(CONFIG_TTY)	+= console.o
+obj-$(CONFIG_NET)	+= network.o
+obj-$(CONFIG_BLK_DEV_SIMDISK) += simdisk.o
diff --git a/arch/xtensa/platforms/iss/console.c b/arch/xtensa/platforms/iss/console.c
new file mode 100644
index 000000000..8e54625cb
--- /dev/null
+++ b/arch/xtensa/platforms/iss/console.c
@@ -0,0 +1,198 @@
+/*
+ * arch/xtensa/platforms/iss/console.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001-2005 Tensilica Inc.
+ *   Authors	Christian Zankel, Joe Taylor
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/sched.h>
+#include <linux/console.h>
+#include <linux/init.h>
+#include <linux/mm.h>
+#include <linux/major.h>
+#include <linux/param.h>
+#include <linux/seq_file.h>
+#include <linux/serial.h>
+
+#include <linux/uaccess.h>
+#include <asm/irq.h>
+
+#include <platform/simcall.h>
+
+#include <linux/tty.h>
+#include <linux/tty_flip.h>
+
+#define SERIAL_MAX_NUM_LINES 1
+#define SERIAL_TIMER_VALUE (HZ / 10)
+
+static void rs_poll(struct timer_list *);
+
+static struct tty_driver *serial_driver;
+static struct tty_port serial_port;
+static DEFINE_TIMER(serial_timer, rs_poll);
+
+static int rs_open(struct tty_struct *tty, struct file * filp)
+{
+	if (tty->count == 1)
+		mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE);
+
+	return 0;
+}
+
+static void rs_close(struct tty_struct *tty, struct file * filp)
+{
+	if (tty->count == 1)
+		timer_delete_sync(&serial_timer);
+}
+
+
+static ssize_t rs_write(struct tty_struct * tty, const u8 *buf, size_t count)
+{
+	/* see drivers/char/serialX.c to reference original version */
+
+	simc_write(1, buf, count);
+	return count;
+}
+
+static void rs_poll(struct timer_list *unused)
+{
+	struct tty_port *port = &serial_port;
+	int i = 0;
+	int rd = 1;
+	u8 c;
+
+	while (simc_poll(0)) {
+		rd = simc_read(0, &c, 1);
+		if (rd <= 0)
+			break;
+		tty_insert_flip_char(port, c, TTY_NORMAL);
+		i++;
+	}
+
+	if (i)
+		tty_flip_buffer_push(port);
+	if (rd)
+		mod_timer(&serial_timer, jiffies + SERIAL_TIMER_VALUE);
+}
+
+static unsigned int rs_write_room(struct tty_struct *tty)
+{
+	/* Let's say iss can always accept 2K characters.. */
+	return 2 * 1024;
+}
+
+static int rs_proc_show(struct seq_file *m, void *v)
+{
+	seq_printf(m, "serinfo:1.0 driver:0.1\n");
+	return 0;
+}
+
+static const struct tty_operations serial_ops = {
+	.open = rs_open,
+	.close = rs_close,
+	.write = rs_write,
+	.write_room = rs_write_room,
+	.proc_show = rs_proc_show,
+};
+
+static int __init rs_init(void)
+{
+	struct tty_driver *driver;
+	int ret;
+
+	driver = tty_alloc_driver(SERIAL_MAX_NUM_LINES, TTY_DRIVER_REAL_RAW);
+	if (IS_ERR(driver))
+		return PTR_ERR(driver);
+
+	tty_port_init(&serial_port);
+
+	/* Initialize the tty_driver structure */
+
+	driver->driver_name = "iss_serial";
+	driver->name = "ttyS";
+	driver->major = TTY_MAJOR;
+	driver->minor_start = 64;
+	driver->type = TTY_DRIVER_TYPE_SERIAL;
+	driver->subtype = SERIAL_TYPE_NORMAL;
+	driver->init_termios = tty_std_termios;
+	driver->init_termios.c_cflag =
+		B9600 | CS8 | CREAD | HUPCL | CLOCAL;
+
+	tty_set_operations(driver, &serial_ops);
+	tty_port_link_device(&serial_port, driver, 0);
+
+	ret = tty_register_driver(driver);
+	if (ret) {
+		pr_err("Couldn't register serial driver\n");
+		tty_driver_kref_put(driver);
+		tty_port_destroy(&serial_port);
+
+		return ret;
+	}
+
+	serial_driver = driver;
+
+	return 0;
+}
+
+
+static __exit void rs_exit(void)
+{
+	tty_unregister_driver(serial_driver);
+	tty_driver_kref_put(serial_driver);
+	tty_port_destroy(&serial_port);
+}
+
+
+/* We use `late_initcall' instead of just `__initcall' as a workaround for
+ * the fact that (1) simcons_tty_init can't be called before tty_init,
+ * (2) tty_init is called via `module_init', (3) if statically linked,
+ * module_init == device_init, and (4) there's no ordering of init lists.
+ * We can do this easily because simcons is always statically linked, but
+ * other tty drivers that depend on tty_init and which must use
+ * `module_init' to declare their init routines are likely to be broken.
+ */
+
+late_initcall(rs_init);
+
+
+#ifdef CONFIG_SERIAL_CONSOLE
+
+static void iss_console_write(struct console *co, const char *s, unsigned count)
+{
+	count = s ? strnlen(s, count) : 0;
+	if (count)
+		simc_write(1, s, count);
+}
+
+static struct tty_driver* iss_console_device(struct console *c, int *index)
+{
+	*index = c->index;
+	return serial_driver;
+}
+
+
+static struct console sercons = {
+	.name = "ttyS",
+	.write = iss_console_write,
+	.device = iss_console_device,
+	.flags = CON_PRINTBUFFER,
+	.index = -1
+};
+
+static int __init iss_console_init(void)
+{
+	register_console(&sercons);
+	return 0;
+}
+
+console_initcall(iss_console_init);
+
+#endif /* CONFIG_SERIAL_CONSOLE */
+
diff --git a/arch/xtensa/platforms/iss/include/platform/serial.h b/arch/xtensa/platforms/iss/include/platform/serial.h
new file mode 100644
index 000000000..16aec542d
--- /dev/null
+++ b/arch/xtensa/platforms/iss/include/platform/serial.h
@@ -0,0 +1,15 @@
+/*
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2012 Tensilica Inc.
+ */
+
+#ifndef __ASM_XTENSA_ISS_SERIAL_H
+#define __ASM_XTENSA_ISS_SERIAL_H
+
+/* Have no meaning on ISS, but needed for 8250_early.c */
+#define BASE_BAUD 0
+
+#endif /* __ASM_XTENSA_ISS_SERIAL_H */
diff --git a/arch/xtensa/platforms/iss/include/platform/simcall-gdbio.h b/arch/xtensa/platforms/iss/include/platform/simcall-gdbio.h
new file mode 100644
index 000000000..e642860e2
--- /dev/null
+++ b/arch/xtensa/platforms/iss/include/platform/simcall-gdbio.h
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2021 Cadence Design Systems Inc. */
+
+#ifndef _XTENSA_PLATFORM_ISS_SIMCALL_GDBIO_H
+#define _XTENSA_PLATFORM_ISS_SIMCALL_GDBIO_H
+
+/*
+ *  System call like services offered by the GDBIO host.
+ */
+
+#define SYS_open	-2
+#define SYS_close	-3
+#define SYS_read	-4
+#define SYS_write	-5
+#define SYS_lseek	-6
+
+static int errno;
+
+static inline int __simc(int a, int b, int c, int d)
+{
+	register int a1 asm("a2") = a;
+	register int b1 asm("a6") = b;
+	register int c1 asm("a3") = c;
+	register int d1 asm("a4") = d;
+	__asm__ __volatile__ (
+			"break 1, 14\n"
+			: "+r"(a1), "+r"(c1)
+			: "r"(b1), "r"(d1)
+			: "memory");
+	errno = c1;
+	return a1;
+}
+
+#endif /* _XTENSA_PLATFORM_ISS_SIMCALL_GDBIO_H */
diff --git a/arch/xtensa/platforms/iss/include/platform/simcall-iss.h b/arch/xtensa/platforms/iss/include/platform/simcall-iss.h
new file mode 100644
index 000000000..5a1e7a1f1
--- /dev/null
+++ b/arch/xtensa/platforms/iss/include/platform/simcall-iss.h
@@ -0,0 +1,73 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2021 Cadence Design Systems Inc. */
+
+#ifndef _XTENSA_PLATFORM_ISS_SIMCALL_ISS_H
+#define _XTENSA_PLATFORM_ISS_SIMCALL_ISS_H
+
+/*
+ *  System call like services offered by the simulator host.
+ */
+
+#define SYS_nop		0	/* unused */
+#define SYS_exit	1	/*x*/
+#define SYS_fork	2
+#define SYS_read	3	/*x*/
+#define SYS_write	4	/*x*/
+#define SYS_open	5	/*x*/
+#define SYS_close	6	/*x*/
+#define SYS_rename	7	/*x 38 - waitpid */
+#define SYS_creat	8	/*x*/
+#define SYS_link	9	/*x (not implemented on WIN32) */
+#define SYS_unlink	10	/*x*/
+#define SYS_execv	11	/* n/a - execve */
+#define SYS_execve	12	/* 11 - chdir */
+#define SYS_pipe	13	/* 42 - time */
+#define SYS_stat	14	/* 106 - mknod */
+#define SYS_chmod	15
+#define SYS_chown	16	/* 202 - lchown */
+#define SYS_utime	17	/* 30 - break */
+#define SYS_wait	18	/* n/a - oldstat */
+#define SYS_lseek	19	/*x*/
+#define SYS_getpid	20
+#define SYS_isatty	21	/* n/a - mount */
+#define SYS_fstat	22	/* 108 - oldumount */
+#define SYS_time	23	/* 13 - setuid */
+#define SYS_gettimeofday 24	/*x 78 - getuid (not implemented on WIN32) */
+#define SYS_times	25	/*X 43 - stime (Xtensa-specific implementation) */
+#define SYS_socket      26
+#define SYS_sendto      27
+#define SYS_recvfrom    28
+#define SYS_select_one  29      /* not compatible select, one file descriptor at the time */
+#define SYS_bind        30
+#define SYS_ioctl	31
+
+#define SYS_iss_argc	1000	/* returns value of argc */
+#define SYS_iss_argv_size 1001	/* bytes needed for argv & arg strings */
+#define SYS_iss_set_argv 1002	/* saves argv & arg strings at given addr */
+
+/*
+ * SYS_select_one specifiers
+ */
+
+#define  XTISS_SELECT_ONE_READ    1
+#define  XTISS_SELECT_ONE_WRITE   2
+#define  XTISS_SELECT_ONE_EXCEPT  3
+
+static int errno;
+
+static inline int __simc(int a, int b, int c, int d)
+{
+	register int a1 asm("a2") = a;
+	register int b1 asm("a3") = b;
+	register int c1 asm("a4") = c;
+	register int d1 asm("a5") = d;
+	__asm__ __volatile__ (
+			"simcall\n"
+			: "+r"(a1), "+r"(b1)
+			: "r"(c1), "r"(d1)
+			: "memory");
+	errno = b1;
+	return a1;
+}
+
+#endif /* _XTENSA_PLATFORM_ISS_SIMCALL_ISS_H */
diff --git a/arch/xtensa/platforms/iss/include/platform/simcall.h b/arch/xtensa/platforms/iss/include/platform/simcall.h
new file mode 100644
index 000000000..e1ec50ce3
--- /dev/null
+++ b/arch/xtensa/platforms/iss/include/platform/simcall.h
@@ -0,0 +1,110 @@
+/*
+ * include/asm-xtensa/platform-iss/simcall.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 Tensilica Inc.
+ * Copyright (C) 2017 - 2021 Cadence Design Systems Inc.
+ */
+
+#ifndef _XTENSA_PLATFORM_ISS_SIMCALL_H
+#define _XTENSA_PLATFORM_ISS_SIMCALL_H
+
+#include <linux/bug.h>
+
+#ifdef CONFIG_XTENSA_SIMCALL_ISS
+#include <platform/simcall-iss.h>
+#endif
+#ifdef CONFIG_XTENSA_SIMCALL_GDBIO
+#include <platform/simcall-gdbio.h>
+#endif
+
+static inline int simc_exit(int exit_code)
+{
+#ifdef SYS_exit
+	return __simc(SYS_exit, exit_code, 0, 0);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+	return -1;
+#endif
+}
+
+static inline int simc_open(const char *file, int flags, int mode)
+{
+	return __simc(SYS_open, (int) file, flags, mode);
+}
+
+static inline int simc_close(int fd)
+{
+	return __simc(SYS_close, fd, 0, 0);
+}
+
+static inline int simc_ioctl(int fd, int request, void *arg)
+{
+#ifdef SYS_ioctl
+	return __simc(SYS_ioctl, fd, request, (int) arg);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+	return -1;
+#endif
+}
+
+static inline int simc_read(int fd, void *buf, size_t count)
+{
+	return __simc(SYS_read, fd, (int) buf, count);
+}
+
+static inline int simc_write(int fd, const void *buf, size_t count)
+{
+	return __simc(SYS_write, fd, (int) buf, count);
+}
+
+static inline int simc_poll(int fd)
+{
+#ifdef SYS_select_one
+	long timeval[2] = { 0, 0 };
+
+	return __simc(SYS_select_one, fd, XTISS_SELECT_ONE_READ, (int)&timeval);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+	return -1;
+#endif
+}
+
+static inline int simc_lseek(int fd, uint32_t off, int whence)
+{
+	return __simc(SYS_lseek, fd, off, whence);
+}
+
+static inline int simc_argc(void)
+{
+#ifdef SYS_iss_argc
+	return __simc(SYS_iss_argc, 0, 0, 0);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+	return 0;
+#endif
+}
+
+static inline int simc_argv_size(void)
+{
+#ifdef SYS_iss_argv_size
+	return __simc(SYS_iss_argv_size, 0, 0, 0);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+	return 0;
+#endif
+}
+
+static inline void simc_argv(void *buf)
+{
+#ifdef SYS_iss_set_argv
+	__simc(SYS_iss_set_argv, (int)buf, 0, 0);
+#else
+	WARN_ONCE(1, "%s: not implemented\n", __func__);
+#endif
+}
+
+#endif /* _XTENSA_PLATFORM_ISS_SIMCALL_H */
diff --git a/arch/xtensa/platforms/iss/network.c b/arch/xtensa/platforms/iss/network.c
new file mode 100644
index 000000000..832579143
--- /dev/null
+++ b/arch/xtensa/platforms/iss/network.c
@@ -0,0 +1,642 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *
+ * arch/xtensa/platforms/iss/network.c
+ *
+ * Platform specific initialization.
+ *
+ * Authors: Chris Zankel <chris@zankel.net>
+ * Based on work form the UML team.
+ *
+ * Copyright 2005 Tensilica Inc.
+ */
+
+#define pr_fmt(fmt) "%s: " fmt, __func__
+
+#include <linux/hex.h>
+#include <linux/list.h>
+#include <linux/irq.h>
+#include <linux/spinlock.h>
+#include <linux/slab.h>
+#include <linux/timer.h>
+#include <linux/if_ether.h>
+#include <linux/inetdevice.h>
+#include <linux/init.h>
+#include <linux/if_tun.h>
+#include <linux/etherdevice.h>
+#include <linux/interrupt.h>
+#include <linux/ioctl.h>
+#include <linux/memblock.h>
+#include <linux/ethtool.h>
+#include <linux/rtnetlink.h>
+#include <linux/platform_device.h>
+
+#include <platform/simcall.h>
+
+#define DRIVER_NAME "iss-netdev"
+#define ETH_MAX_PACKET 1500
+#define ETH_HEADER_OTHER 14
+#define ISS_NET_TIMER_VALUE (HZ / 10)
+
+/* ------------------------------------------------------------------------- */
+
+/* We currently only support the TUNTAP transport protocol. */
+
+#define TRANSPORT_TUNTAP_NAME "tuntap"
+#define TRANSPORT_TUNTAP_MTU ETH_MAX_PACKET
+
+struct tuntap_info {
+	char dev_name[IFNAMSIZ];
+	int fd;
+};
+
+/* ------------------------------------------------------------------------- */
+
+
+struct iss_net_private;
+
+struct iss_net_ops {
+	int (*open)(struct iss_net_private *lp);
+	void (*close)(struct iss_net_private *lp);
+	int (*read)(struct iss_net_private *lp, struct sk_buff **skb);
+	int (*write)(struct iss_net_private *lp, struct sk_buff **skb);
+	unsigned short (*protocol)(struct sk_buff *skb);
+	int (*poll)(struct iss_net_private *lp);
+};
+
+/* This structure contains out private information for the driver. */
+
+struct iss_net_private {
+	spinlock_t lock;
+	struct net_device *dev;
+	struct platform_device pdev;
+	struct timer_list tl;
+	struct rtnl_link_stats64 stats;
+
+	struct timer_list timer;
+	unsigned int timer_val;
+
+	int index;
+	int mtu;
+
+	struct {
+		union {
+			struct tuntap_info tuntap;
+		} info;
+
+		const struct iss_net_ops *net_ops;
+	} tp;
+
+};
+
+/* ================================ HELPERS ================================ */
+
+
+static char *split_if_spec(char *str, ...)
+{
+	char **arg, *end;
+	va_list ap;
+
+	va_start(ap, str);
+	while ((arg = va_arg(ap, char**)) != NULL) {
+		if (*str == '\0') {
+			va_end(ap);
+			return NULL;
+		}
+		end = strchr(str, ',');
+		if (end != str)
+			*arg = str;
+		if (end == NULL) {
+			va_end(ap);
+			return NULL;
+		}
+		*end++ = '\0';
+		str = end;
+	}
+	va_end(ap);
+	return str;
+}
+
+/* Set Ethernet address of the specified device. */
+
+static void setup_etheraddr(struct net_device *dev, char *str)
+{
+	u8 addr[ETH_ALEN];
+
+	if (str == NULL)
+		goto random;
+
+	if (!mac_pton(str, addr)) {
+		pr_err("%s: failed to parse '%s' as an ethernet address\n",
+		       dev->name, str);
+		goto random;
+	}
+	if (is_multicast_ether_addr(addr)) {
+		pr_err("%s: attempt to assign a multicast ethernet address\n",
+		       dev->name);
+		goto random;
+	}
+	if (!is_valid_ether_addr(addr)) {
+		pr_err("%s: attempt to assign an invalid ethernet address\n",
+		       dev->name);
+		goto random;
+	}
+	if (!is_local_ether_addr(addr))
+		pr_warn("%s: assigning a globally valid ethernet address\n",
+			dev->name);
+	eth_hw_addr_set(dev, addr);
+	return;
+
+random:
+	pr_info("%s: choosing a random ethernet address\n",
+		dev->name);
+	eth_hw_addr_random(dev);
+}
+
+/* ======================= TUNTAP TRANSPORT INTERFACE ====================== */
+
+static int tuntap_open(struct iss_net_private *lp)
+{
+	struct ifreq ifr;
+	char *dev_name = lp->tp.info.tuntap.dev_name;
+	int err = -EINVAL;
+	int fd;
+
+	fd = simc_open("/dev/net/tun", 02, 0); /* O_RDWR */
+	if (fd < 0) {
+		pr_err("%s: failed to open /dev/net/tun, returned %d (errno = %d)\n",
+		       lp->dev->name, fd, errno);
+		return fd;
+	}
+
+	memset(&ifr, 0, sizeof(ifr));
+	ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
+	strscpy(ifr.ifr_name, dev_name, sizeof(ifr.ifr_name));
+
+	err = simc_ioctl(fd, TUNSETIFF, &ifr);
+	if (err < 0) {
+		pr_err("%s: failed to set interface %s, returned %d (errno = %d)\n",
+		       lp->dev->name, dev_name, err, errno);
+		simc_close(fd);
+		return err;
+	}
+
+	lp->tp.info.tuntap.fd = fd;
+	return err;
+}
+
+static void tuntap_close(struct iss_net_private *lp)
+{
+	simc_close(lp->tp.info.tuntap.fd);
+	lp->tp.info.tuntap.fd = -1;
+}
+
+static int tuntap_read(struct iss_net_private *lp, struct sk_buff **skb)
+{
+	return simc_read(lp->tp.info.tuntap.fd,
+			(*skb)->data, (*skb)->dev->mtu + ETH_HEADER_OTHER);
+}
+
+static int tuntap_write(struct iss_net_private *lp, struct sk_buff **skb)
+{
+	return simc_write(lp->tp.info.tuntap.fd, (*skb)->data, (*skb)->len);
+}
+
+static unsigned short tuntap_protocol(struct sk_buff *skb)
+{
+	return eth_type_trans(skb, skb->dev);
+}
+
+static int tuntap_poll(struct iss_net_private *lp)
+{
+	return simc_poll(lp->tp.info.tuntap.fd);
+}
+
+static const struct iss_net_ops tuntap_ops = {
+	.open		= tuntap_open,
+	.close		= tuntap_close,
+	.read		= tuntap_read,
+	.write		= tuntap_write,
+	.protocol	= tuntap_protocol,
+	.poll		= tuntap_poll,
+};
+
+/*
+ * ethX=tuntap,[mac address],device name
+ */
+
+static int tuntap_probe(struct iss_net_private *lp, int index, char *init)
+{
+	struct net_device *dev = lp->dev;
+	char *dev_name = NULL, *mac_str = NULL, *rem = NULL;
+
+	/* Transport should be 'tuntap': ethX=tuntap,mac,dev_name */
+
+	if (strncmp(init, TRANSPORT_TUNTAP_NAME,
+		    sizeof(TRANSPORT_TUNTAP_NAME) - 1))
+		return 0;
+
+	init += sizeof(TRANSPORT_TUNTAP_NAME) - 1;
+	if (*init == ',') {
+		rem = split_if_spec(init + 1, &mac_str, &dev_name, NULL);
+		if (rem != NULL) {
+			pr_err("%s: extra garbage on specification : '%s'\n",
+			       dev->name, rem);
+			return 0;
+		}
+	} else if (*init != '\0') {
+		pr_err("%s: invalid argument: %s. Skipping device!\n",
+		       dev->name, init);
+		return 0;
+	}
+
+	if (!dev_name) {
+		pr_err("%s: missing tuntap device name\n", dev->name);
+		return 0;
+	}
+
+	strscpy(lp->tp.info.tuntap.dev_name, dev_name,
+		sizeof(lp->tp.info.tuntap.dev_name));
+
+	setup_etheraddr(dev, mac_str);
+
+	lp->mtu = TRANSPORT_TUNTAP_MTU;
+
+	lp->tp.info.tuntap.fd = -1;
+	lp->tp.net_ops = &tuntap_ops;
+
+	return 1;
+}
+
+/* ================================ ISS NET ================================ */
+
+static int iss_net_rx(struct net_device *dev)
+{
+	struct iss_net_private *lp = netdev_priv(dev);
+	int pkt_len;
+	struct sk_buff *skb;
+
+	/* Check if there is any new data. */
+
+	if (lp->tp.net_ops->poll(lp) == 0)
+		return 0;
+
+	/* Try to allocate memory, if it fails, try again next round. */
+
+	skb = dev_alloc_skb(dev->mtu + 2 + ETH_HEADER_OTHER);
+	if (skb == NULL) {
+		spin_lock_bh(&lp->lock);
+		lp->stats.rx_dropped++;
+		spin_unlock_bh(&lp->lock);
+		return 0;
+	}
+
+	skb_reserve(skb, 2);
+
+	/* Setup skb */
+
+	skb->dev = dev;
+	skb_reset_mac_header(skb);
+	pkt_len = lp->tp.net_ops->read(lp, &skb);
+	skb_put(skb, pkt_len);
+
+	if (pkt_len > 0) {
+		skb_trim(skb, pkt_len);
+		skb->protocol = lp->tp.net_ops->protocol(skb);
+
+		spin_lock_bh(&lp->lock);
+		lp->stats.rx_bytes += skb->len;
+		lp->stats.rx_packets++;
+		spin_unlock_bh(&lp->lock);
+		netif_rx(skb);
+		return pkt_len;
+	}
+	kfree_skb(skb);
+	return pkt_len;
+}
+
+static int iss_net_poll(struct iss_net_private *lp)
+{
+	int err, ret = 0;
+
+	if (!netif_running(lp->dev))
+		return 0;
+
+	while ((err = iss_net_rx(lp->dev)) > 0)
+		ret++;
+
+	if (err < 0) {
+		pr_err("Device '%s' read returned %d, shutting it down\n",
+		       lp->dev->name, err);
+		dev_close(lp->dev);
+	} else {
+		/* FIXME reactivate_fd(lp->fd, ISS_ETH_IRQ); */
+	}
+
+	return ret;
+}
+
+
+static void iss_net_timer(struct timer_list *t)
+{
+	struct iss_net_private *lp = timer_container_of(lp, t, timer);
+
+	iss_net_poll(lp);
+	mod_timer(&lp->timer, jiffies + lp->timer_val);
+}
+
+
+static int iss_net_open(struct net_device *dev)
+{
+	struct iss_net_private *lp = netdev_priv(dev);
+	int err;
+
+	err = lp->tp.net_ops->open(lp);
+	if (err < 0)
+		return err;
+
+	netif_start_queue(dev);
+
+	/* clear buffer - it can happen that the host side of the interface
+	 * is full when we get here. In this case, new data is never queued,
+	 * SIGIOs never arrive, and the net never works.
+	 */
+	while ((err = iss_net_rx(dev)) > 0)
+		;
+
+	timer_setup(&lp->timer, iss_net_timer, 0);
+	lp->timer_val = ISS_NET_TIMER_VALUE;
+	mod_timer(&lp->timer, jiffies + lp->timer_val);
+
+	return err;
+}
+
+static int iss_net_close(struct net_device *dev)
+{
+	struct iss_net_private *lp = netdev_priv(dev);
+
+	netif_stop_queue(dev);
+	timer_delete_sync(&lp->timer);
+	lp->tp.net_ops->close(lp);
+
+	return 0;
+}
+
+static int iss_net_start_xmit(struct sk_buff *skb, struct net_device *dev)
+{
+	struct iss_net_private *lp = netdev_priv(dev);
+	int len;
+
+	netif_stop_queue(dev);
+
+	len = lp->tp.net_ops->write(lp, &skb);
+
+	if (len == skb->len) {
+		spin_lock_bh(&lp->lock);
+		lp->stats.tx_packets++;
+		lp->stats.tx_bytes += skb->len;
+		spin_unlock_bh(&lp->lock);
+		netif_trans_update(dev);
+		netif_start_queue(dev);
+
+		/* this is normally done in the interrupt when tx finishes */
+		netif_wake_queue(dev);
+
+	} else if (len == 0) {
+		netif_start_queue(dev);
+		spin_lock_bh(&lp->lock);
+		lp->stats.tx_dropped++;
+		spin_unlock_bh(&lp->lock);
+
+	} else {
+		netif_start_queue(dev);
+		pr_err("%s: %s failed(%d)\n", dev->name, __func__, len);
+	}
+
+
+	dev_kfree_skb(skb);
+	return NETDEV_TX_OK;
+}
+
+
+static void iss_net_get_stats64(struct net_device *dev,
+				struct rtnl_link_stats64 *stats)
+{
+	struct iss_net_private *lp = netdev_priv(dev);
+
+	spin_lock_bh(&lp->lock);
+	*stats = lp->stats;
+	spin_unlock_bh(&lp->lock);
+}
+
+static void iss_net_set_multicast_list(struct net_device *dev)
+{
+}
+
+static void iss_net_tx_timeout(struct net_device *dev, unsigned int txqueue)
+{
+}
+
+static int iss_net_change_mtu(struct net_device *dev, int new_mtu)
+{
+	return -EINVAL;
+}
+
+static void iss_net_user_timer_expire(struct timer_list *unused)
+{
+}
+
+
+static struct platform_driver iss_net_driver = {
+	.driver = {
+		.name  = DRIVER_NAME,
+	},
+};
+
+static int driver_registered;
+
+static const struct net_device_ops iss_netdev_ops = {
+	.ndo_open		= iss_net_open,
+	.ndo_stop		= iss_net_close,
+	.ndo_get_stats64	= iss_net_get_stats64,
+	.ndo_start_xmit		= iss_net_start_xmit,
+	.ndo_validate_addr	= eth_validate_addr,
+	.ndo_change_mtu		= iss_net_change_mtu,
+	.ndo_set_mac_address	= eth_mac_addr,
+	.ndo_tx_timeout		= iss_net_tx_timeout,
+	.ndo_set_rx_mode	= iss_net_set_multicast_list,
+};
+
+static void iss_net_pdev_release(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct iss_net_private *lp =
+		container_of(pdev, struct iss_net_private, pdev);
+
+	free_netdev(lp->dev);
+}
+
+static void iss_net_configure(int index, char *init)
+{
+	struct net_device *dev;
+	struct iss_net_private *lp;
+
+	dev = alloc_etherdev(sizeof(*lp));
+	if (dev == NULL) {
+		pr_err("eth_configure: failed to allocate device\n");
+		return;
+	}
+
+	/* Initialize private element. */
+
+	lp = netdev_priv(dev);
+	*lp = (struct iss_net_private) {
+		.dev			= dev,
+		.index			= index,
+	};
+
+	spin_lock_init(&lp->lock);
+	/*
+	 * If this name ends up conflicting with an existing registered
+	 * netdevice, that is OK, register_netdev{,ice}() will notice this
+	 * and fail.
+	 */
+	snprintf(dev->name, sizeof(dev->name), "eth%d", index);
+
+	/*
+	 * Try all transport protocols.
+	 * Note: more protocols can be added by adding '&& !X_init(lp, eth)'.
+	 */
+
+	if (!tuntap_probe(lp, index, init)) {
+		pr_err("%s: invalid arguments. Skipping device!\n",
+		       dev->name);
+		goto err_free_netdev;
+	}
+
+	pr_info("Netdevice %d (%pM)\n", index, dev->dev_addr);
+
+	/* sysfs register */
+
+	if (!driver_registered) {
+		if (platform_driver_register(&iss_net_driver))
+			goto err_free_netdev;
+		driver_registered = 1;
+	}
+
+	lp->pdev.id = index;
+	lp->pdev.name = DRIVER_NAME;
+	lp->pdev.dev.release = iss_net_pdev_release;
+	if (platform_device_register(&lp->pdev))
+		goto err_free_netdev;
+	SET_NETDEV_DEV(dev, &lp->pdev.dev);
+
+	dev->netdev_ops = &iss_netdev_ops;
+	dev->mtu = lp->mtu;
+	dev->watchdog_timeo = (HZ >> 1);
+	dev->irq = -1;
+
+	rtnl_lock();
+	if (register_netdevice(dev)) {
+		rtnl_unlock();
+		pr_err("%s: error registering net device!\n", dev->name);
+		platform_device_unregister(&lp->pdev);
+		/* dev is freed by the iss_net_pdev_release callback */
+		return;
+	}
+	rtnl_unlock();
+
+	timer_setup(&lp->tl, iss_net_user_timer_expire, 0);
+
+	return;
+
+err_free_netdev:
+	free_netdev(dev);
+}
+
+/* ------------------------------------------------------------------------- */
+
+/* Filled in during early boot */
+
+struct list_head eth_cmd_line = LIST_HEAD_INIT(eth_cmd_line);
+
+struct iss_net_init {
+	struct list_head list;
+	char *init;		/* init string */
+	int index;
+};
+
+/*
+ * Parse the command line and look for 'ethX=...' fields, and register all
+ * those fields. They will be later initialized in iss_net_init.
+ */
+
+static int __init iss_net_setup(char *str)
+{
+	struct iss_net_init *device = NULL;
+	struct iss_net_init *new;
+	struct list_head *ele;
+	char *end;
+	int rc;
+	unsigned n;
+
+	end = strchr(str, '=');
+	if (!end) {
+		pr_err("Expected '=' after device number\n");
+		return 1;
+	}
+	*end = 0;
+	rc = kstrtouint(str, 0, &n);
+	*end = '=';
+	if (rc < 0) {
+		pr_err("Failed to parse '%s'\n", str);
+		return 1;
+	}
+	str = end;
+
+	list_for_each(ele, &eth_cmd_line) {
+		device = list_entry(ele, struct iss_net_init, list);
+		if (device->index == n)
+			break;
+	}
+
+	if (device && device->index == n) {
+		pr_err("Device %u already configured\n", n);
+		return 1;
+	}
+
+	new = memblock_alloc(sizeof(*new), SMP_CACHE_BYTES);
+	if (new == NULL) {
+		pr_err("Alloc_bootmem failed\n");
+		return 1;
+	}
+
+	INIT_LIST_HEAD(&new->list);
+	new->index = n;
+	new->init = str + 1;
+
+	list_add_tail(&new->list, &eth_cmd_line);
+	return 1;
+}
+
+__setup("eth", iss_net_setup);
+
+/*
+ * Initialize all ISS Ethernet devices previously registered in iss_net_setup.
+ */
+
+static int iss_net_init(void)
+{
+	struct list_head *ele, *next;
+
+	/* Walk through all Ethernet devices specified in the command line. */
+
+	list_for_each_safe(ele, next, &eth_cmd_line) {
+		struct iss_net_init *eth;
+		eth = list_entry(ele, struct iss_net_init, list);
+		iss_net_configure(eth->index, eth->init);
+	}
+
+	return 1;
+}
+device_initcall(iss_net_init);
diff --git a/arch/xtensa/platforms/iss/setup.c b/arch/xtensa/platforms/iss/setup.c
new file mode 100644
index 000000000..21283acab
--- /dev/null
+++ b/arch/xtensa/platforms/iss/setup.c
@@ -0,0 +1,88 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *
+ * arch/xtensa/platform-iss/setup.c
+ *
+ * Platform specific initialization.
+ *
+ * Authors: Chris Zankel <chris@zankel.net>
+ *          Joe Taylor <joe@tensilica.com>
+ *
+ * Copyright 2001 - 2005 Tensilica Inc.
+ * Copyright 2017 Cadence Design Systems Inc.
+ */
+#include <linux/init.h>
+#include <linux/kernel.h>
+#include <linux/notifier.h>
+#include <linux/panic_notifier.h>
+#include <linux/printk.h>
+#include <linux/reboot.h>
+#include <linux/string.h>
+
+#include <asm/platform.h>
+#include <asm/setup.h>
+
+#include <platform/simcall.h>
+
+
+static int iss_power_off(struct sys_off_data *unused)
+{
+	pr_info(" ** Called platform_power_off() **\n");
+	simc_exit(0);
+	return NOTIFY_DONE;
+}
+
+static int iss_restart(struct sys_off_data *unused)
+{
+	/* Flush and reset the mmu, simulate a processor reset, and
+	 * jump to the reset vector. */
+	cpu_reset();
+
+	return NOTIFY_DONE;
+}
+
+static int
+iss_panic_event(struct notifier_block *this, unsigned long event, void *ptr)
+{
+	simc_exit(1);
+	return NOTIFY_DONE;
+}
+
+static struct notifier_block iss_panic_block = {
+	.notifier_call = iss_panic_event,
+};
+
+void __init platform_setup(char **p_cmdline)
+{
+	static void *argv[COMMAND_LINE_SIZE / sizeof(void *)] __initdata;
+	static char cmdline[COMMAND_LINE_SIZE] __initdata;
+	int argc = simc_argc();
+	int argv_size = simc_argv_size();
+
+	if (argc > 1) {
+		if (argv_size > sizeof(argv)) {
+			pr_err("%s: command line too long: argv_size = %d\n",
+			       __func__, argv_size);
+		} else {
+			int i;
+
+			cmdline[0] = 0;
+			simc_argv((void *)argv);
+
+			for (i = 1; i < argc; ++i) {
+				if (i > 1)
+					strcat(cmdline, " ");
+				strcat(cmdline, argv[i]);
+			}
+			*p_cmdline = cmdline;
+		}
+	}
+
+	atomic_notifier_chain_register(&panic_notifier_list, &iss_panic_block);
+	register_sys_off_handler(SYS_OFF_MODE_RESTART,
+				 SYS_OFF_PRIO_PLATFORM,
+				 iss_restart, NULL);
+	register_sys_off_handler(SYS_OFF_MODE_POWER_OFF,
+				 SYS_OFF_PRIO_PLATFORM,
+				 iss_power_off, NULL);
+}
diff --git a/arch/xtensa/platforms/iss/simdisk.c b/arch/xtensa/platforms/iss/simdisk.c
new file mode 100644
index 000000000..a2ddb49c2
--- /dev/null
+++ b/arch/xtensa/platforms/iss/simdisk.c
@@ -0,0 +1,404 @@
+/*
+ * arch/xtensa/platforms/iss/simdisk.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001-2013 Tensilica Inc.
+ *   Authors	Victor Prupis
+ */
+
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/string.h>
+#include <linux/string_choices.h>
+#include <linux/blkdev.h>
+#include <linux/bio.h>
+#include <linux/proc_fs.h>
+#include <linux/uaccess.h>
+#include <platform/simcall.h>
+
+#define SIMDISK_MAJOR 240
+#define SIMDISK_MINORS 1
+#define MAX_SIMDISK_COUNT 10
+
+struct simdisk {
+	const char *filename;
+	spinlock_t lock;
+	struct gendisk *gd;
+	struct proc_dir_entry *procfile;
+	int users;
+	unsigned long size;
+	int fd;
+};
+
+
+static int simdisk_count = CONFIG_BLK_DEV_SIMDISK_COUNT;
+module_param(simdisk_count, int, S_IRUGO);
+MODULE_PARM_DESC(simdisk_count, "Number of simdisk units.");
+
+static int n_files;
+static char *filename[MAX_SIMDISK_COUNT] = {
+#ifdef CONFIG_SIMDISK0_FILENAME
+	CONFIG_SIMDISK0_FILENAME,
+#ifdef CONFIG_SIMDISK1_FILENAME
+	CONFIG_SIMDISK1_FILENAME,
+#endif
+#endif
+};
+
+/*
+ * The simdisk code can be built either into the kernel or as a loadable module.
+ * When built-in, CONFIG_SIMDISK{0,1}_FILENAME can be used to specify the
+ * initial simdisk filenames and additional filenames can be provided on the
+ * kernel command line. These arguments are parsed during early boot when slab
+ * is not yet available, but the command line itself is preserved for the
+ * lifetime of the kernel, so the incoming pointer is stored directly.
+ * When built as a loadable module, each value is copied with kstrdup() and all
+ * allocated memory is freed in simdisk_param_free_filename() when the module is
+ * unloaded.
+ */
+static int simdisk_param_set_filename(const char *val,
+		const struct kernel_param *kp)
+{
+	char *str;
+
+	if (n_files >= ARRAY_SIZE(filename))
+		return -EINVAL;
+
+#ifdef MODULE
+	str = kstrdup(val, GFP_KERNEL);
+	if (!str)
+		return -ENOMEM;
+#else
+	str = (char *)val;
+#endif
+
+	filename[n_files++] = str;
+	return 0;
+}
+
+static void simdisk_param_free_filename(void *arg)
+{
+	for (int i = 0; i < n_files; i++)
+		kfree(filename[i]);
+}
+
+static const struct kernel_param_ops simdisk_param_ops_filename = {
+	.set = simdisk_param_set_filename,
+	.free = simdisk_param_free_filename,
+};
+module_param_cb(filename, &simdisk_param_ops_filename, NULL, 0);
+MODULE_PARM_DESC(filename, "Backing storage filename.");
+
+static int simdisk_major = SIMDISK_MAJOR;
+
+static void simdisk_transfer(struct simdisk *dev, unsigned long sector,
+		unsigned long nsect, char *buffer, int write)
+{
+	unsigned long offset = sector << SECTOR_SHIFT;
+	unsigned long nbytes = nsect << SECTOR_SHIFT;
+
+	if (offset > dev->size || dev->size - offset < nbytes) {
+		pr_notice("Beyond-end %s (%ld %ld)\n",
+				str_write_read(write), offset, nbytes);
+		return;
+	}
+
+	spin_lock(&dev->lock);
+	while (nbytes > 0) {
+		unsigned long io;
+
+		simc_lseek(dev->fd, offset, SEEK_SET);
+		READ_ONCE(*buffer);
+		if (write)
+			io = simc_write(dev->fd, buffer, nbytes);
+		else
+			io = simc_read(dev->fd, buffer, nbytes);
+		if (io == -1) {
+			pr_err("SIMDISK: IO error %d\n", errno);
+			break;
+		}
+		buffer += io;
+		offset += io;
+		nbytes -= io;
+	}
+	spin_unlock(&dev->lock);
+}
+
+static void simdisk_submit_bio(struct bio *bio)
+{
+	struct simdisk *dev = bio->bi_bdev->bd_disk->private_data;
+	struct bio_vec bvec;
+	struct bvec_iter iter;
+	sector_t sector = bio->bi_iter.bi_sector;
+
+	bio_for_each_segment(bvec, bio, iter) {
+		char *buffer = bvec_kmap_local(&bvec);
+		unsigned len = bvec.bv_len >> SECTOR_SHIFT;
+
+		simdisk_transfer(dev, sector, len, buffer,
+				bio_data_dir(bio) == WRITE);
+		sector += len;
+		kunmap_local(buffer);
+	}
+
+	bio_endio(bio);
+}
+
+static int simdisk_open(struct gendisk *disk, blk_mode_t mode)
+{
+	struct simdisk *dev = disk->private_data;
+
+	spin_lock(&dev->lock);
+	++dev->users;
+	spin_unlock(&dev->lock);
+	return 0;
+}
+
+static void simdisk_release(struct gendisk *disk)
+{
+	struct simdisk *dev = disk->private_data;
+	spin_lock(&dev->lock);
+	--dev->users;
+	spin_unlock(&dev->lock);
+}
+
+static const struct block_device_operations simdisk_ops = {
+	.owner		= THIS_MODULE,
+	.submit_bio	= simdisk_submit_bio,
+	.open		= simdisk_open,
+	.release	= simdisk_release,
+};
+
+static struct simdisk *sddev;
+static struct proc_dir_entry *simdisk_procdir;
+
+static int simdisk_attach(struct simdisk *dev, const char *filename)
+{
+	int err = 0;
+
+	filename = kstrdup(filename, GFP_KERNEL);
+	if (filename == NULL)
+		return -ENOMEM;
+
+	spin_lock(&dev->lock);
+
+	if (dev->fd != -1) {
+		err = -EBUSY;
+		goto out;
+	}
+	dev->fd = simc_open(filename, O_RDWR, 0);
+	if (dev->fd == -1) {
+		pr_err("SIMDISK: Can't open %s: %d\n", filename, errno);
+		err = -ENODEV;
+		goto out;
+	}
+	dev->size = simc_lseek(dev->fd, 0, SEEK_END);
+	set_capacity(dev->gd, dev->size >> SECTOR_SHIFT);
+	dev->filename = filename;
+	pr_info("SIMDISK: %s=%s\n", dev->gd->disk_name, dev->filename);
+out:
+	if (err)
+		kfree(filename);
+	spin_unlock(&dev->lock);
+
+	return err;
+}
+
+static int simdisk_detach(struct simdisk *dev)
+{
+	int err = 0;
+
+	spin_lock(&dev->lock);
+
+	if (dev->users != 0) {
+		err = -EBUSY;
+	} else if (dev->fd != -1) {
+		if (simc_close(dev->fd)) {
+			pr_err("SIMDISK: error closing %s: %d\n",
+					dev->filename, errno);
+			err = -EIO;
+		} else {
+			pr_info("SIMDISK: %s detached from %s\n",
+					dev->gd->disk_name, dev->filename);
+			dev->fd = -1;
+			kfree(dev->filename);
+			dev->filename = NULL;
+		}
+	}
+	spin_unlock(&dev->lock);
+	return err;
+}
+
+static ssize_t proc_read_simdisk(struct file *file, char __user *buf,
+			size_t size, loff_t *ppos)
+{
+	struct simdisk *dev = pde_data(file_inode(file));
+	const char *s = dev->filename;
+	if (s) {
+		ssize_t len = strlen(s);
+		char *temp = kmalloc(len + 2, GFP_KERNEL);
+
+		if (!temp)
+			return -ENOMEM;
+
+		len = scnprintf(temp, len + 2, "%s\n", s);
+		len = simple_read_from_buffer(buf, size, ppos,
+					      temp, len);
+
+		kfree(temp);
+		return len;
+	}
+	return simple_read_from_buffer(buf, size, ppos, "\n", 1);
+}
+
+static ssize_t proc_write_simdisk(struct file *file, const char __user *buf,
+			size_t count, loff_t *ppos)
+{
+	char *tmp;
+	struct simdisk *dev = pde_data(file_inode(file));
+	int err;
+
+	if (count == 0 || count > PAGE_SIZE)
+		return -EINVAL;
+
+	tmp = memdup_user_nul(buf, count);
+	if (IS_ERR(tmp))
+		return PTR_ERR(tmp);
+
+	err = simdisk_detach(dev);
+	if (err != 0)
+		goto out_free;
+
+	if (count > 0 && tmp[count - 1] == '\n')
+		tmp[count - 1] = 0;
+
+	if (tmp[0])
+		err = simdisk_attach(dev, tmp);
+
+	if (err == 0)
+		err = count;
+out_free:
+	kfree(tmp);
+	return err;
+}
+
+static const struct proc_ops simdisk_proc_ops = {
+	.proc_read	= proc_read_simdisk,
+	.proc_write	= proc_write_simdisk,
+	.proc_lseek	= default_llseek,
+};
+
+static int __init simdisk_setup(struct simdisk *dev, int which,
+		struct proc_dir_entry *procdir)
+{
+	struct queue_limits lim = {
+		.features		= BLK_FEAT_ROTATIONAL,
+	};
+	char tmp[2] = { '0' + which, 0 };
+	int err;
+
+	dev->fd = -1;
+	dev->filename = NULL;
+	spin_lock_init(&dev->lock);
+	dev->users = 0;
+
+	dev->gd = blk_alloc_disk(&lim, NUMA_NO_NODE);
+	if (IS_ERR(dev->gd)) {
+		err = PTR_ERR(dev->gd);
+		goto out;
+	}
+	dev->gd->major = simdisk_major;
+	dev->gd->first_minor = which;
+	dev->gd->minors = SIMDISK_MINORS;
+	dev->gd->fops = &simdisk_ops;
+	dev->gd->private_data = dev;
+	snprintf(dev->gd->disk_name, 32, "simdisk%d", which);
+	set_capacity(dev->gd, 0);
+	err = add_disk(dev->gd);
+	if (err)
+		goto out_cleanup_disk;
+
+	dev->procfile = proc_create_data(tmp, 0644, procdir, &simdisk_proc_ops, dev);
+
+	return 0;
+
+out_cleanup_disk:
+	put_disk(dev->gd);
+out:
+	return err;
+}
+
+static int __init simdisk_init(void)
+{
+	int i;
+
+	if (register_blkdev(simdisk_major, "simdisk") < 0) {
+		pr_err("SIMDISK: register_blkdev: %d\n", simdisk_major);
+		return -EIO;
+	}
+	pr_info("SIMDISK: major: %d\n", simdisk_major);
+
+	if (n_files > simdisk_count)
+		simdisk_count = n_files;
+	if (simdisk_count > MAX_SIMDISK_COUNT)
+		simdisk_count = MAX_SIMDISK_COUNT;
+
+	sddev = kmalloc_objs(*sddev, simdisk_count);
+	if (sddev == NULL)
+		goto out_unregister;
+
+	simdisk_procdir = proc_mkdir("simdisk", 0);
+	if (simdisk_procdir == NULL)
+		goto out_free_unregister;
+
+	for (i = 0; i < simdisk_count; ++i) {
+		if (simdisk_setup(sddev + i, i, simdisk_procdir) == 0) {
+			if (filename[i] != NULL && filename[i][0] != 0 &&
+					(n_files == 0 || i < n_files))
+				simdisk_attach(sddev + i, filename[i]);
+		}
+	}
+
+	return 0;
+
+out_free_unregister:
+	kfree(sddev);
+out_unregister:
+	unregister_blkdev(simdisk_major, "simdisk");
+	return -ENOMEM;
+}
+module_init(simdisk_init);
+
+static void simdisk_teardown(struct simdisk *dev, int which,
+		struct proc_dir_entry *procdir)
+{
+	char tmp[2] = { '0' + which, 0 };
+
+	simdisk_detach(dev);
+	if (dev->gd) {
+		del_gendisk(dev->gd);
+		put_disk(dev->gd);
+	}
+	remove_proc_entry(tmp, procdir);
+}
+
+static void __exit simdisk_exit(void)
+{
+	int i;
+
+	for (i = 0; i < simdisk_count; ++i)
+		simdisk_teardown(sddev + i, i, simdisk_procdir);
+	remove_proc_entry("simdisk", 0);
+	kfree(sddev);
+	unregister_blkdev(simdisk_major, "simdisk");
+}
+module_exit(simdisk_exit);
+
+MODULE_ALIAS_BLOCKDEV_MAJOR(SIMDISK_MAJOR);
+
+MODULE_LICENSE("GPL");
diff --git a/arch/xtensa/platforms/xt2000/Makefile b/arch/xtensa/platforms/xt2000/Makefile
new file mode 100644
index 000000000..53eaeba5e
--- /dev/null
+++ b/arch/xtensa/platforms/xt2000/Makefile
@@ -0,0 +1,6 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# Makefile for the Tensilica XT2000 Emulation Board
+#
+
+obj-y 			= setup.o
diff --git a/arch/xtensa/platforms/xt2000/include/platform/hardware.h b/arch/xtensa/platforms/xt2000/include/platform/hardware.h
new file mode 100644
index 000000000..9f213f573
--- /dev/null
+++ b/arch/xtensa/platforms/xt2000/include/platform/hardware.h
@@ -0,0 +1,43 @@
+/*
+ * platform/hardware.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 Tensilica Inc.
+ */
+
+/*
+ * This file contains the hardware configuration of the XT2000 board.
+ */
+
+#ifndef _XTENSA_XT2000_HARDWARE_H
+#define _XTENSA_XT2000_HARDWARE_H
+
+#include <asm/core.h>
+
+/*
+ * On-board components.
+ */
+
+#define SONIC83934_INTNUM	XCHAL_EXTINT3_NUM
+#define SONIC83934_ADDR		IOADDR(0x0d030000)
+
+/*
+ * V3-PCI
+ */
+
+/* The XT2000 uses the V3 as a cascaded interrupt controller for the PCI bus */
+
+#define IRQ_PCI_A		(XCHAL_NUM_INTERRUPTS + 0)
+#define IRQ_PCI_B		(XCHAL_NUM_INTERRUPTS + 1)
+#define IRQ_PCI_C		(XCHAL_NUM_INTERRUPTS + 2)
+
+/*
+ * Various other components.
+ */
+
+#define XT2000_LED_ADDR		IOADDR(0x0d040000)
+
+#endif /* _XTENSA_XT2000_HARDWARE_H */
diff --git a/arch/xtensa/platforms/xt2000/include/platform/serial.h b/arch/xtensa/platforms/xt2000/include/platform/serial.h
new file mode 100644
index 000000000..cde804827
--- /dev/null
+++ b/arch/xtensa/platforms/xt2000/include/platform/serial.h
@@ -0,0 +1,28 @@
+/*
+ * platform/serial.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_XT2000_SERIAL_H
+#define _XTENSA_XT2000_SERIAL_H
+
+#include <asm/core.h>
+#include <asm/io.h>
+
+/*  National-Semi PC16552D DUART:  */
+
+#define DUART16552_1_INTNUM	XCHAL_EXTINT4_NUM
+#define DUART16552_2_INTNUM	XCHAL_EXTINT5_NUM
+
+#define DUART16552_1_ADDR	IOADDR(0x0d050020)	/* channel 1 */
+#define DUART16552_2_ADDR	IOADDR(0x0d050000)	/* channel 2 */
+
+#define DUART16552_XTAL_FREQ	18432000	/* crystal frequency in Hz */
+#define BASE_BAUD ( DUART16552_XTAL_FREQ / 16 )
+
+#endif /* _XTENSA_XT2000_SERIAL_H */
diff --git a/arch/xtensa/platforms/xt2000/setup.c b/arch/xtensa/platforms/xt2000/setup.c
new file mode 100644
index 000000000..eda4c15c6
--- /dev/null
+++ b/arch/xtensa/platforms/xt2000/setup.c
@@ -0,0 +1,147 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * arch/xtensa/platforms/xt2000/setup.c
+ *
+ * Platform specific functions for the XT2000 board.
+ *
+ * Authors:	Chris Zankel <chris@zankel.net>
+ *		Joe Taylor <joe@tensilica.com>
+ *
+ * Copyright 2001 - 2004 Tensilica Inc.
+ */
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/errno.h>
+#include <linux/reboot.h>
+#include <linux/kdev_t.h>
+#include <linux/types.h>
+#include <linux/major.h>
+#include <linux/console.h>
+#include <linux/delay.h>
+#include <linux/stringify.h>
+#include <linux/platform_device.h>
+#include <linux/serial.h>
+#include <linux/serial_8250.h>
+#include <linux/timer.h>
+
+#include <asm/processor.h>
+#include <asm/platform.h>
+#include <asm/bootparam.h>
+#include <platform/hardware.h>
+#include <platform/serial.h>
+
+/* Assumes s points to an 8-chr string.  No checking for NULL. */
+
+static void led_print (int f, char *s)
+{
+	unsigned long* led_addr = (unsigned long*) (XT2000_LED_ADDR + 0xE0) + f;
+	int i;
+	for (i = f; i < 8; i++)
+		if ((*led_addr++ = *s++) == 0)
+		    break;
+}
+
+static int xt2000_power_off(struct sys_off_data *unused)
+{
+	led_print (0, "POWEROFF");
+	local_irq_disable();
+	while (1);
+	return NOTIFY_DONE;
+}
+
+static int xt2000_restart(struct sys_off_data *unused)
+{
+	/* Flush and reset the mmu, simulate a processor reset, and
+	 * jump to the reset vector. */
+	cpu_reset();
+
+	return NOTIFY_DONE;
+}
+
+void __init platform_setup(char** cmdline)
+{
+	led_print (0, "LINUX   ");
+}
+
+/* Heartbeat. Let the LED blink. */
+
+static void xt2000_heartbeat(struct timer_list *unused);
+
+static DEFINE_TIMER(heartbeat_timer, xt2000_heartbeat);
+
+static void xt2000_heartbeat(struct timer_list *unused)
+{
+	static int i;
+
+	led_print(7, i ? "." : " ");
+	i ^= 1;
+	mod_timer(&heartbeat_timer, jiffies + HZ / 2);
+}
+
+//#define RS_TABLE_SIZE 2
+
+#define _SERIAL_PORT(_base,_irq)					\
+{									\
+	.mapbase	= (_base),					\
+	.membase	= (void*)(_base),				\
+	.irq		= (_irq),					\
+	.uartclk	= DUART16552_XTAL_FREQ,				\
+	.iotype		= UPIO_MEM,					\
+	.flags		= UPF_BOOT_AUTOCONF,				\
+	.regshift	= 2,						\
+}
+
+static struct plat_serial8250_port xt2000_serial_data[] = {
+#if XCHAL_HAVE_BE
+	_SERIAL_PORT(DUART16552_1_ADDR + 3, DUART16552_1_INTNUM),
+	_SERIAL_PORT(DUART16552_2_ADDR + 3, DUART16552_2_INTNUM),
+#else
+	_SERIAL_PORT(DUART16552_1_ADDR, DUART16552_1_INTNUM),
+	_SERIAL_PORT(DUART16552_2_ADDR, DUART16552_2_INTNUM),
+#endif
+	{ }
+};
+
+static struct platform_device xt2000_serial8250_device = {
+	.name		= "serial8250",
+	.id		= PLAT8250_DEV_PLATFORM,
+	.dev		= {
+	    .platform_data = xt2000_serial_data,
+	},
+};
+
+static struct resource xt2000_sonic_res[] = {
+	{
+		.start = SONIC83934_ADDR,
+		.end   = SONIC83934_ADDR + 0xff,
+		.flags = IORESOURCE_MEM,
+	},
+	{
+		.start = SONIC83934_INTNUM,
+		.end = SONIC83934_INTNUM,
+		.flags = IORESOURCE_IRQ,
+	},
+};
+
+static struct platform_device xt2000_sonic_device = {
+	.name		= "xtsonic",
+	.num_resources	= ARRAY_SIZE(xt2000_sonic_res),
+	.resource		= xt2000_sonic_res,
+};
+
+static int __init xt2000_setup_devinit(void)
+{
+	platform_device_register(&xt2000_serial8250_device);
+	platform_device_register(&xt2000_sonic_device);
+	mod_timer(&heartbeat_timer, jiffies + HZ / 2);
+	register_sys_off_handler(SYS_OFF_MODE_RESTART,
+				 SYS_OFF_PRIO_PLATFORM,
+				 xt2000_restart, NULL);
+	register_sys_off_handler(SYS_OFF_MODE_POWER_OFF,
+				 SYS_OFF_PRIO_DEFAULT,
+				 xt2000_power_off, NULL);
+	return 0;
+}
+
+device_initcall(xt2000_setup_devinit);
diff --git a/arch/xtensa/platforms/xtfpga/Makefile b/arch/xtensa/platforms/xtfpga/Makefile
new file mode 100644
index 000000000..0600371b2
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/Makefile
@@ -0,0 +1,11 @@
+# SPDX-License-Identifier: GPL-2.0-only
+# Makefile for the Tensilica xtavnet Emulation Board
+#
+# Note! Dependencies are done automagically by 'make dep', which also
+# removes any old dependencies. DON'T put your own dependencies here
+# unless it's something special (ie not a .c file).
+#
+# Note 2! The CFLAGS definitions are in the main makefile...
+
+obj-y			+= setup.o
+obj-$(CONFIG_XTFPGA_LCD) += lcd.o
diff --git a/arch/xtensa/platforms/xtfpga/include/platform/hardware.h b/arch/xtensa/platforms/xtfpga/include/platform/hardware.h
new file mode 100644
index 000000000..30d9cb6cf
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/include/platform/hardware.h
@@ -0,0 +1,60 @@
+/*
+ * arch/xtensa/platform/xtavnet/include/platform/hardware.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2006 Tensilica Inc.
+ */
+
+/*
+ * This file contains the hardware configuration of the XTAVNET boards.
+ */
+
+#include <asm/types.h>
+
+#ifndef __XTENSA_XTAVNET_HARDWARE_H
+#define __XTENSA_XTAVNET_HARDWARE_H
+
+/* Default assignment of LX60 devices to external interrupts. */
+
+#ifdef CONFIG_XTENSA_MX
+#define DUART16552_INTNUM	XCHAL_EXTINT3_NUM
+#define OETH_IRQ		XCHAL_EXTINT4_NUM
+#define C67X00_IRQ		XCHAL_EXTINT8_NUM
+#else
+#define DUART16552_INTNUM	XCHAL_EXTINT0_NUM
+#define OETH_IRQ		XCHAL_EXTINT1_NUM
+#define C67X00_IRQ		XCHAL_EXTINT5_NUM
+#endif
+
+/*
+ *  Device addresses and parameters.
+ */
+
+/* UART */
+#define DUART16552_PADDR	(XCHAL_KIO_PADDR + 0x0D050020)
+
+/* Misc. */
+#define XTFPGA_FPGAREGS_VADDR	IOADDR(0x0D020000)
+/* Clock frequency in Hz (read-only):  */
+#define XTFPGA_CLKFRQ_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x04)
+/* Setting of 8 DIP switches:  */
+#define DIP_SWITCHES_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x0C)
+/* Software reset (write 0xdead):  */
+#define XTFPGA_SWRST_VADDR	(XTFPGA_FPGAREGS_VADDR + 0x10)
+
+/*  OpenCores Ethernet controller:  */
+				/* regs + RX/TX descriptors */
+#define OETH_REGS_PADDR		(XCHAL_KIO_PADDR + 0x0D030000)
+#define OETH_REGS_SIZE		0x1000
+#define OETH_SRAMBUFF_PADDR	(XCHAL_KIO_PADDR + 0x0D800000)
+
+				/* 5*rx buffs + 5*tx buffs */
+#define OETH_SRAMBUFF_SIZE	(5 * 0x600 + 5 * 0x600)
+
+#define C67X00_PADDR		(XCHAL_KIO_PADDR + 0x0D0D0000)
+#define C67X00_SIZE		0x10
+
+#endif /* __XTENSA_XTAVNET_HARDWARE_H */
diff --git a/arch/xtensa/platforms/xtfpga/include/platform/lcd.h b/arch/xtensa/platforms/xtfpga/include/platform/lcd.h
new file mode 100644
index 000000000..4c8541ed1
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/include/platform/lcd.h
@@ -0,0 +1,35 @@
+/*
+ * arch/xtensa/platform/xtavnet/include/platform/lcd.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ */
+
+#ifndef __XTENSA_XTAVNET_LCD_H
+#define __XTENSA_XTAVNET_LCD_H
+
+#ifdef CONFIG_XTFPGA_LCD
+/* Display string STR at position POS on the LCD. */
+void lcd_disp_at_pos(char *str, unsigned char pos);
+
+/* Shift the contents of the LCD display left or right. */
+void lcd_shiftleft(void);
+void lcd_shiftright(void);
+#else
+static inline void lcd_disp_at_pos(char *str, unsigned char pos)
+{
+}
+
+static inline void lcd_shiftleft(void)
+{
+}
+
+static inline void lcd_shiftright(void)
+{
+}
+#endif
+
+#endif
diff --git a/arch/xtensa/platforms/xtfpga/include/platform/serial.h b/arch/xtensa/platforms/xtfpga/include/platform/serial.h
new file mode 100644
index 000000000..14d8f7bee
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/include/platform/serial.h
@@ -0,0 +1,18 @@
+/*
+ * arch/xtensa/platform/xtavnet/include/platform/serial.h
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ */
+
+#ifndef __ASM_XTENSA_XTAVNET_SERIAL_H
+#define __ASM_XTENSA_XTAVNET_SERIAL_H
+
+#include <platform/hardware.h>
+
+#define BASE_BAUD (*(long *)XTFPGA_CLKFRQ_VADDR / 16)
+
+#endif /* __ASM_XTENSA_XTAVNET_SERIAL_H */
diff --git a/arch/xtensa/platforms/xtfpga/lcd.c b/arch/xtensa/platforms/xtfpga/lcd.c
new file mode 100644
index 000000000..2f7eb66c2
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/lcd.c
@@ -0,0 +1,89 @@
+/*
+ * Driver for the LCD display on the Tensilica XTFPGA board family.
+ * http://www.mytechcorp.com/cfdata/productFile/File1/MOC-16216B-B-A0A04.pdf
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 2001, 2006 Tensilica Inc.
+ * Copyright (C) 2015 Cadence Design Systems Inc.
+ */
+
+#include <linux/delay.h>
+#include <linux/init.h>
+#include <linux/io.h>
+
+#include <platform/hardware.h>
+#include <platform/lcd.h>
+
+/* LCD instruction and data addresses. */
+#define LCD_INSTR_ADDR		((char *)IOADDR(CONFIG_XTFPGA_LCD_BASE_ADDR))
+#define LCD_DATA_ADDR		(LCD_INSTR_ADDR + 4)
+
+#define LCD_CLEAR		0x1
+#define LCD_DISPLAY_ON		0xc
+
+/* 8bit and 2 lines display */
+#define LCD_DISPLAY_MODE8BIT	0x38
+#define LCD_DISPLAY_MODE4BIT	0x28
+#define LCD_DISPLAY_POS		0x80
+#define LCD_SHIFT_LEFT		0x18
+#define LCD_SHIFT_RIGHT		0x1c
+
+static void lcd_put_byte(u8 *addr, u8 data)
+{
+#ifdef CONFIG_XTFPGA_LCD_8BIT_ACCESS
+	WRITE_ONCE(*addr, data);
+#else
+	WRITE_ONCE(*addr, data & 0xf0);
+	WRITE_ONCE(*addr, (data << 4) & 0xf0);
+#endif
+}
+
+static int __init lcd_init(void)
+{
+	WRITE_ONCE(*LCD_INSTR_ADDR, LCD_DISPLAY_MODE8BIT);
+	mdelay(5);
+	WRITE_ONCE(*LCD_INSTR_ADDR, LCD_DISPLAY_MODE8BIT);
+	udelay(200);
+	WRITE_ONCE(*LCD_INSTR_ADDR, LCD_DISPLAY_MODE8BIT);
+	udelay(50);
+#ifndef CONFIG_XTFPGA_LCD_8BIT_ACCESS
+	WRITE_ONCE(*LCD_INSTR_ADDR, LCD_DISPLAY_MODE4BIT);
+	udelay(50);
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_DISPLAY_MODE4BIT);
+	udelay(50);
+#endif
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_DISPLAY_ON);
+	udelay(50);
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_CLEAR);
+	mdelay(10);
+	lcd_disp_at_pos("XTENSA LINUX", 0);
+	return 0;
+}
+
+void lcd_disp_at_pos(char *str, unsigned char pos)
+{
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_DISPLAY_POS | pos);
+	udelay(100);
+	while (*str != 0) {
+		lcd_put_byte(LCD_DATA_ADDR, *str);
+		udelay(200);
+		str++;
+	}
+}
+
+void lcd_shiftleft(void)
+{
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_SHIFT_LEFT);
+	udelay(50);
+}
+
+void lcd_shiftright(void)
+{
+	lcd_put_byte(LCD_INSTR_ADDR, LCD_SHIFT_RIGHT);
+	udelay(50);
+}
+
+arch_initcall(lcd_init);
diff --git a/arch/xtensa/platforms/xtfpga/setup.c b/arch/xtensa/platforms/xtfpga/setup.c
new file mode 100644
index 000000000..67d7f37f1
--- /dev/null
+++ b/arch/xtensa/platforms/xtfpga/setup.c
@@ -0,0 +1,301 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ *
+ * arch/xtensa/platform/xtavnet/setup.c
+ *
+ * ...
+ *
+ * Authors:	Chris Zankel <chris@zankel.net>
+ *		Joe Taylor <joe@tensilica.com>
+ *
+ * Copyright 2001 - 2006 Tensilica Inc.
+ */
+#include <linux/stddef.h>
+#include <linux/kernel.h>
+#include <linux/init.h>
+#include <linux/io.h>
+#include <linux/errno.h>
+#include <linux/reboot.h>
+#include <linux/kdev_t.h>
+#include <linux/types.h>
+#include <linux/major.h>
+#include <linux/console.h>
+#include <linux/delay.h>
+#include <linux/of.h>
+#include <linux/clk-provider.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+
+#include <asm/timex.h>
+#include <asm/processor.h>
+#include <asm/platform.h>
+#include <asm/bootparam.h>
+#include <platform/lcd.h>
+#include <platform/hardware.h>
+
+static int xtfpga_power_off(struct sys_off_data *unused)
+{
+	lcd_disp_at_pos("POWEROFF", 0);
+	local_irq_disable();
+	while (1)
+		cpu_relax();
+	return NOTIFY_DONE;
+}
+
+static int xtfpga_restart(struct sys_off_data *unused)
+{
+	/* Try software reset first. */
+	WRITE_ONCE(*(u32 *)XTFPGA_SWRST_VADDR, 0xdead);
+
+	/* If software reset did not work, flush and reset the mmu,
+	 * simulate a processor reset, and jump to the reset vector.
+	 */
+	cpu_reset();
+
+	return NOTIFY_DONE;
+}
+
+#ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
+
+void __init platform_calibrate_ccount(void)
+{
+	ccount_freq = *(long *)XTFPGA_CLKFRQ_VADDR;
+}
+
+#endif
+
+static void __init xtfpga_register_handlers(void)
+{
+	register_sys_off_handler(SYS_OFF_MODE_RESTART,
+				 SYS_OFF_PRIO_PLATFORM,
+				 xtfpga_restart, NULL);
+	register_sys_off_handler(SYS_OFF_MODE_POWER_OFF,
+				 SYS_OFF_PRIO_DEFAULT,
+				 xtfpga_power_off, NULL);
+}
+
+#ifdef CONFIG_USE_OF
+
+static void __init xtfpga_clk_setup(struct device_node *np)
+{
+	void __iomem *base = of_iomap(np, 0);
+	struct clk *clk;
+	u32 freq;
+
+	if (!base) {
+		pr_err("%pOFn: invalid address\n", np);
+		return;
+	}
+
+	freq = __raw_readl(base);
+	iounmap(base);
+	clk = clk_register_fixed_rate(NULL, np->name, NULL, 0, freq);
+
+	if (IS_ERR(clk)) {
+		pr_err("%pOFn: clk registration failed\n", np);
+		return;
+	}
+
+	if (of_clk_add_provider(np, of_clk_src_simple_get, clk)) {
+		pr_err("%pOFn: clk provider registration failed\n", np);
+		return;
+	}
+}
+CLK_OF_DECLARE(xtfpga_clk, "cdns,xtfpga-clock", xtfpga_clk_setup);
+
+#define MAC_LEN 6
+static void __init update_local_mac(struct device_node *node)
+{
+	struct property *newmac;
+	const u8* macaddr;
+	int prop_len;
+
+	macaddr = of_get_property(node, "local-mac-address", &prop_len);
+	if (macaddr == NULL || prop_len != MAC_LEN)
+		return;
+
+	newmac = kzalloc(sizeof(*newmac) + MAC_LEN, GFP_KERNEL);
+	if (newmac == NULL)
+		return;
+
+	newmac->value = newmac + 1;
+	newmac->length = MAC_LEN;
+	newmac->name = kstrdup("local-mac-address", GFP_KERNEL);
+	if (newmac->name == NULL) {
+		kfree(newmac);
+		return;
+	}
+
+	memcpy(newmac->value, macaddr, MAC_LEN);
+	((u8*)newmac->value)[5] = (*(u32*)DIP_SWITCHES_VADDR) & 0x3f;
+	of_update_property(node, newmac);
+}
+
+static int __init machine_setup(void)
+{
+	struct device_node *eth = NULL;
+
+	if ((eth = of_find_compatible_node(eth, NULL, "opencores,ethoc")))
+		update_local_mac(eth);
+	of_node_put(eth);
+
+	xtfpga_register_handlers();
+
+	return 0;
+}
+arch_initcall(machine_setup);
+
+#else
+
+#include <linux/serial_8250.h>
+#include <linux/if.h>
+#include <net/ethoc.h>
+#include <linux/usb/c67x00.h>
+
+/*----------------------------------------------------------------------------
+ *  Ethernet -- OpenCores Ethernet MAC (ethoc driver)
+ */
+
+static struct resource ethoc_res[] = {
+	[0] = { /* register space */
+		.start = OETH_REGS_PADDR,
+		.end   = OETH_REGS_PADDR + OETH_REGS_SIZE - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[1] = { /* buffer space */
+		.start = OETH_SRAMBUFF_PADDR,
+		.end   = OETH_SRAMBUFF_PADDR + OETH_SRAMBUFF_SIZE - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[2] = { /* IRQ number */
+		.start = XTENSA_PIC_LINUX_IRQ(OETH_IRQ),
+		.end   = XTENSA_PIC_LINUX_IRQ(OETH_IRQ),
+		.flags = IORESOURCE_IRQ,
+	},
+};
+
+static struct ethoc_platform_data ethoc_pdata = {
+	/*
+	 * The MAC address for these boards is 00:50:c2:13:6f:xx.
+	 * The last byte (here as zero) is read from the DIP switches on the
+	 * board.
+	 */
+	.hwaddr = { 0x00, 0x50, 0xc2, 0x13, 0x6f, 0 },
+	.phy_id = -1,
+	.big_endian = XCHAL_HAVE_BE,
+};
+
+static struct platform_device ethoc_device = {
+	.name = "ethoc",
+	.id = -1,
+	.num_resources = ARRAY_SIZE(ethoc_res),
+	.resource = ethoc_res,
+	.dev = {
+		.platform_data = &ethoc_pdata,
+	},
+};
+
+/*----------------------------------------------------------------------------
+ *  USB Host/Device -- Cypress CY7C67300
+ */
+
+static struct resource c67x00_res[] = {
+	[0] = { /* register space */
+		.start = C67X00_PADDR,
+		.end   = C67X00_PADDR + C67X00_SIZE - 1,
+		.flags = IORESOURCE_MEM,
+	},
+	[1] = { /* IRQ number */
+		.start = XTENSA_PIC_LINUX_IRQ(C67X00_IRQ),
+		.end   = XTENSA_PIC_LINUX_IRQ(C67X00_IRQ),
+		.flags = IORESOURCE_IRQ,
+	},
+};
+
+static struct c67x00_platform_data c67x00_pdata = {
+	.sie_config = C67X00_SIE1_HOST | C67X00_SIE2_UNUSED,
+	.hpi_regstep = 4,
+};
+
+static struct platform_device c67x00_device = {
+	.name = "c67x00",
+	.id = -1,
+	.num_resources = ARRAY_SIZE(c67x00_res),
+	.resource = c67x00_res,
+	.dev = {
+		.platform_data = &c67x00_pdata,
+	},
+};
+
+/*----------------------------------------------------------------------------
+ *  UART
+ */
+
+static struct resource serial_resource = {
+	.start	= DUART16552_PADDR,
+	.end	= DUART16552_PADDR + 0x1f,
+	.flags	= IORESOURCE_MEM,
+};
+
+static struct plat_serial8250_port serial_platform_data[] = {
+	[0] = {
+		.mapbase	= DUART16552_PADDR,
+		.irq		= XTENSA_PIC_LINUX_IRQ(DUART16552_INTNUM),
+		.flags		= UPF_BOOT_AUTOCONF | UPF_SKIP_TEST |
+				  UPF_IOREMAP,
+		.iotype		= XCHAL_HAVE_BE ? UPIO_MEM32BE : UPIO_MEM32,
+		.regshift	= 2,
+		.uartclk	= 0,    /* set in xtavnet_init() */
+	},
+	{ },
+};
+
+static struct platform_device xtavnet_uart = {
+	.name		= "serial8250",
+	.id		= PLAT8250_DEV_PLATFORM,
+	.dev		= {
+		.platform_data	= serial_platform_data,
+	},
+	.num_resources	= 1,
+	.resource	= &serial_resource,
+};
+
+/* platform devices */
+static struct platform_device *platform_devices[] __initdata = {
+	&ethoc_device,
+	&c67x00_device,
+	&xtavnet_uart,
+};
+
+
+static int __init xtavnet_init(void)
+{
+	/* Ethernet MAC address.  */
+	ethoc_pdata.hwaddr[5] = *(u32 *)DIP_SWITCHES_VADDR;
+
+	/* Clock rate varies among FPGA bitstreams; board specific FPGA register
+	 * reports the actual clock rate.
+	 */
+	serial_platform_data[0].uartclk = *(long *)XTFPGA_CLKFRQ_VADDR;
+
+
+	/* register platform devices */
+	platform_add_devices(platform_devices, ARRAY_SIZE(platform_devices));
+
+	/* ETHOC driver is a bit quiet; at least display Ethernet MAC, so user
+	 * knows whether they set it correctly on the DIP switches.
+	 */
+	pr_info("XTFPGA: Ethernet MAC %pM\n", ethoc_pdata.hwaddr);
+	ethoc_pdata.eth_clkfreq = *(long *)XTFPGA_CLKFRQ_VADDR;
+
+	xtfpga_register_handlers();
+
+	return 0;
+}
+
+/*
+ * Register to be done during do_initcalls().
+ */
+arch_initcall(xtavnet_init);
+
+#endif /* CONFIG_USE_OF */
diff --git a/arch/xtensa/variants/csp/include/variant/core.h b/arch/xtensa/variants/csp/include/variant/core.h
new file mode 100644
index 000000000..ccd81f07e
--- /dev/null
+++ b/arch/xtensa/variants/csp/include/variant/core.h
@@ -0,0 +1,575 @@
+/* 
+ * xtensa/config/core-isa.h -- HAL definitions that are dependent on Xtensa
+ *				processor CORE configuration
+ *
+ *  See <xtensa/config/core.h>, which includes this file, for more details.
+ */
+
+/* Xtensa processor core configuration information.
+
+   Copyright (c) 1999-2015 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	8	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_LOOP_BUFFER_SIZE		0	/* zero-ov. loop instr buffer size */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_DEPBITS		0	/* DEPBITS instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		1	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	0	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		4	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MX			0	/* MX core (Tensilica internal) */
+#define XCHAL_HAVE_MP_INTERRUPTS	0	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		0	/* core RunStall control port */
+#define XCHAL_HAVE_PSO			0	/* Power Shut-Off */
+#define XCHAL_HAVE_PSO_CDM		0	/* core/debug/mem pwr domains */
+#define XCHAL_HAVE_PSO_FULL_RETENTION	0	/* all regs preserved on PSO */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		1	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			8	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+
+#define XCHAL_HAVE_FUSION		 0	/* Fusion*/
+#define XCHAL_HAVE_FUSION_FP	 0	        /* Fusion FP option */
+#define XCHAL_HAVE_FUSION_LOW_POWER 0	/* Fusion Low Power option */
+#define XCHAL_HAVE_FUSION_AES	 0	        /* Fusion BLE/Wifi AES-128 CCM option */
+#define XCHAL_HAVE_FUSION_CONVENC	 0       /* Fusion Conv Encode option */
+#define XCHAL_HAVE_FUSION_LFSR_CRC	 0	/* Fusion LFSR-CRC option */
+#define XCHAL_HAVE_FUSION_BITOPS	 0	/* Fusion Bit Operations Support option */
+#define XCHAL_HAVE_FUSION_AVS	 0	/* Fusion AVS option */
+#define XCHAL_HAVE_FUSION_16BIT_BASEBAND	 0	/* Fusion 16-bit Baseband option */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4		0	/* HiFi4 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4_VFPU		0	/* HiFi4 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI3		0	/* HiFi3 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI3_VFPU		0	/* HiFi3 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2EP		0	/* HiFi2EP */
+#define XCHAL_HAVE_HIFI_MINI		0	
+
+
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector or user floating-point pkg */
+#define XCHAL_HAVE_USER_DPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_USER_SPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_FP                 0      /* single prec floating point */
+#define XCHAL_HAVE_FP_DIV             0  /* FP with DIV instructions */
+#define XCHAL_HAVE_FP_RECIP           0        /* FP with RECIP instructions */
+#define XCHAL_HAVE_FP_SQRT            0 /* FP with SQRT instructions */
+#define XCHAL_HAVE_FP_RSQRT           0        /* FP with RSQRT instructions */
+#define XCHAL_HAVE_DFP                        0     /* double precision FP pkg */
+#define XCHAL_HAVE_DFP_DIV            0 /* DFP with DIV instructions */
+#define XCHAL_HAVE_DFP_RECIP          0       /* DFP with RECIP instructions*/
+#define XCHAL_HAVE_DFP_SQRT           0        /* DFP with SQRT instructions */
+#define XCHAL_HAVE_DFP_RSQRT          0       /* DFP with RSQRT instructions*/
+#define XCHAL_HAVE_DFP_ACCEL		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_DFP_accel		XCHAL_HAVE_DFP_ACCEL				/* for backward compatibility */
+
+#define XCHAL_HAVE_DFPU_SINGLE_ONLY    0                 	/* DFPU Coprocessor, single precision only */
+#define XCHAL_HAVE_DFPU_SINGLE_DOUBLE  0               	/* DFPU Coprocessor, single and double precision */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_PDX4		        0	/* PDX4 */
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+#define XCHAL_HAVE_CONNXD2_DUALLSFLIX   0	/* ConnX D2 & Dual LoadStore Flix */
+#define XCHAL_HAVE_BBE16		0	/* ConnX BBE16 pkg */
+#define XCHAL_HAVE_BBE16_RSQRT		0	/* BBE16 & vector recip sqrt */
+#define XCHAL_HAVE_BBE16_VECDIV		0	/* BBE16 & vector divide */
+#define XCHAL_HAVE_BBE16_DESPREAD	0	/* BBE16 & despread */
+#define XCHAL_HAVE_BBENEP		0	/* ConnX BBENEP pkgs */
+#define XCHAL_HAVE_BSP3			0	/* ConnX BSP3 pkg */
+#define XCHAL_HAVE_BSP3_TRANSPOSE	0	/* BSP3 & transpose32x32 */
+#define XCHAL_HAVE_SSP16		0	/* ConnX SSP16 pkg */
+#define XCHAL_HAVE_SSP16_VITERBI	0	/* SSP16 & viterbi */
+#define XCHAL_HAVE_TURBO16		0	/* ConnX Turbo16 pkg */
+#define XCHAL_HAVE_BBP16		0	/* ConnX BBP16 pkg */
+#define XCHAL_HAVE_FLIX3		0	/* basic 3-way FLIX option */
+#define XCHAL_HAVE_GRIVPEP              0   /*  GRIVPEP is General Release of IVPEP */
+#define XCHAL_HAVE_GRIVPEP_HISTOGRAM    0   /* Histogram option on GRIVPEP */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_LOADSTORE_UNITS	1	/* load/store units */
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		8	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		16	/* data width in bytes */
+#define XCHAL_DATA_PIPE_DELAY		1	/* d-side pipeline delay
+						   (1 = 5-stage, 2 = 7-stage) */
+#define XCHAL_CLOCK_GATING_GLOBAL	0	/* global clock gating */
+#define XCHAL_CLOCK_GATING_FUNCUNIT	0	/* funct. unit clock gating */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		1100002	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"xt_lnx"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_BUILD_UNIQUE_ID		0x00057D54	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC1B3FFFE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x1C857D54	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX6.0.2"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2600	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		2	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		260002	/* major*100+minor */
+#define XCHAL_HW_REL_LX6		1
+#define XCHAL_HW_REL_LX6_0		1
+#define XCHAL_HW_REL_LX6_0_2		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2600	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	2	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		260002	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2600	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	2	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		260002	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		64	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		64	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		6	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		6	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		65536	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	0	/* MP coherence feature */
+
+#define XCHAL_HAVE_PREFETCH		0	/* PREFCTL register */
+#define XCHAL_HAVE_PREFETCH_L1		0	/* prefetch to L1 dcache */
+#define XCHAL_PREFETCH_CASTOUT_LINES	0	/* dcache pref. castout bufsz */
+#define XCHAL_PREFETCH_ENTRIES		0	/* cache prefetch entries */
+#define XCHAL_PREFETCH_BLOCK_ENTRIES	0	/* prefetch block streams */
+#define XCHAL_HAVE_CACHE_BLOCKOPS	0	/* block prefetch for caches */
+#define XCHAL_HAVE_ICACHE_TEST		1	/* Icache test instructions */
+#define XCHAL_HAVE_DCACHE_TEST		1	/* Dcache test instructions */
+#define XCHAL_HAVE_ICACHE_DYN_WAYS	0	/* Icache dynamic way support */
+#define XCHAL_HAVE_DCACHE_DYN_WAYS	0	/* Dcache dynamic way support */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		8
+#define XCHAL_DCACHE_SETWIDTH		6
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		4
+#define XCHAL_DCACHE_WAYS		4
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	16
+#define XCHAL_DCACHE_ACCESS_SIZE	16
+
+#define XCHAL_DCACHE_BANKS		1	/* number of banks */
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+/*  Whether MEMCTL register has anything useful  */
+#define XCHAL_USE_MEMCTL		(((XCHAL_LOOP_BUFFER_SIZE > 0)	||	\
+					   XCHAL_DCACHE_IS_COHERENT	||	\
+					   XCHAL_HAVE_ICACHE_DYN_WAYS	||	\
+					   XCHAL_HAVE_DCACHE_DYN_WAYS)	&&	\
+					   (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RE_2012_0))
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+#define XCHAL_HAVE_IMEM_LOADSTORE	1	/* can load/store to IROM/IRAM*/
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		16	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		3	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F00BF
+#define XCHAL_INTLEVEL2_MASK		0x00001140
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00008000
+#define XCHAL_INTLEVEL5_MASK		0x00002000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F00BF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F11FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F1FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F9FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		2
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		2
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		4
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_PROFILING
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F1000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000033F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+#define XCHAL_INTTYPE_MASK_PROFILING	0x00008000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+#define XCHAL_PROFILING_INTERRUPT	15	/* profiling interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL4_NUM		15
+#define XCHAL_INTLEVEL5_NUM		13
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 2, 3.)  */
+
+
+/*
+ *  External interrupt mapping.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL BInterrupt pin number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 2) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		21	/* (intlevel 3) */
+/*  EXTERNAL BInterrupt pin numbers mapped to each core interrupt number:  */
+#define XCHAL_INT0_EXTNUM		0	/* (intlevel 1) */
+#define XCHAL_INT1_EXTNUM		1	/* (intlevel 1) */
+#define XCHAL_INT2_EXTNUM		2	/* (intlevel 1) */
+#define XCHAL_INT3_EXTNUM		3	/* (intlevel 1) */
+#define XCHAL_INT4_EXTNUM		4	/* (intlevel 1) */
+#define XCHAL_INT5_EXTNUM		5	/* (intlevel 1) */
+#define XCHAL_INT8_EXTNUM		6	/* (intlevel 2) */
+#define XCHAL_INT9_EXTNUM		7	/* (intlevel 3) */
+#define XCHAL_INT12_EXTNUM		8	/* (intlevel 2) */
+#define XCHAL_INT14_EXTNUM		9	/* (intlevel 7) */
+#define XCHAL_INT16_EXTNUM		10	/* (intlevel 1) */
+#define XCHAL_INT17_EXTNUM		11	/* (intlevel 1) */
+#define XCHAL_INT18_EXTNUM		12	/* (intlevel 1) */
+#define XCHAL_INT19_EXTNUM		13	/* (intlevel 1) */
+#define XCHAL_INT20_EXTNUM		14	/* (intlevel 1) */
+#define XCHAL_INT21_EXTNUM		15	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) or TX */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_HALT			0	/* halt architecture option */
+#define XCHAL_HAVE_BOOTLOADER		0	/* boot loader (for TX) */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0x00002000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00002000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0x00001000
+#define XCHAL_RESET_VECTOR1_PADDR	0x00001000
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0x00002340
+#define XCHAL_USER_VECTOR_PADDR		0x00002340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0x00002300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00002300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0x000023C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000023C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0x00002000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00002000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0x00002180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00002180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0x000021C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x000021C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0x00002200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x00002200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0x00002240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x00002240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0x00002280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x00002280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0x000022C0
+#define XCHAL_NMI_VECTOR_PADDR		0x000022C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG MODULE
+  ----------------------------------------------------------------------*/
+
+/*  Misc  */
+#define XCHAL_HAVE_DEBUG_ERI		1	/* ERI to debug module */
+#define XCHAL_HAVE_DEBUG_APB		1	/* APB to debug module */
+#define XCHAL_HAVE_DEBUG_JTAG		1	/* JTAG to debug module */
+
+/*  On-Chip Debug (OCD)  */
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	0	/* faster OCD option (to LX4) */
+#define XCHAL_HAVE_OCD_LS32DDR		1	/* L32DDR/S32DDR (faster OCD) */
+
+/*  TRAX (in core)  */
+#define XCHAL_HAVE_TRAX			1	/* TRAX in debug module */
+#define XCHAL_TRAX_MEM_SIZE		262144	/* TRAX memory size in bytes */
+#define XCHAL_TRAX_MEM_SHAREABLE	1	/* start/end regs; ready sig. */
+#define XCHAL_TRAX_ATB_WIDTH		0	/* ATB width (bits), 0=no ATB */
+#define XCHAL_TRAX_TIME_WIDTH		0	/* timestamp bitwidth, 0=none */
+
+/*  Perf counters  */
+#define XCHAL_NUM_PERF_COUNTERS		8	/* performance counters */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		1	/* one way maps I+D 4GB vaddr */
+#define XCHAL_SPANNING_WAY		6	/* TLB spanning way number */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/csp/include/variant/tie-asm.h b/arch/xtensa/variants/csp/include/variant/tie-asm.h
new file mode 100644
index 000000000..ba773c4a2
--- /dev/null
+++ b/arch/xtensa/variants/csp/include/variant/tie-asm.h
@@ -0,0 +1,194 @@
+/* 
+ * tie-asm.h -- compile-time HAL assembler definitions dependent on CORE & TIE
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file contains assembly-language definitions (assembly
+   macros, etc.) for this specific Xtensa processor's TIE extensions
+   and options.  It is customized to this Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+#define XTHAL_SAS_ANYOT	0x0003	/* both of the above */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+#define XTHAL_SAS_ANYCC	0x000C	/* both of the above */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+#define XTHAL_SAS_ANYABI	0x0070	/* all of the above three */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+#define XTHAL_SAS3(optie,ccuse,abi)	( ((optie) & XTHAL_SAS_ANYOT)  \
+					| ((ccuse) & XTHAL_SAS_ANYCC)  \
+					| ((abi)   & XTHAL_SAS_ANYABI) )
+
+
+    /*
+      *  Macro to store all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger store sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to store.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to store, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any store.
+      */
+    .macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	rur.THREADPTR	\at1		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	rsr.ACCLO	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.ACCHI	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	rsr.BR	\at1		// boolean option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.SCOMPARE1	\at1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rsr.M0	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rsr.M1	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rsr.M2	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	rsr.M3	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+20
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_store
+
+    /*
+      *  Macro to load all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger load sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to load.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to load, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any load.
+      */
+    .macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.THREADPTR	\at1		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.ACCLO	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.ACCHI	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.BR	\at1		// boolean option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.SCOMPARE1	\at1		// conditional store option
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wsr.M0	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wsr.M1	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wsr.M2	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+20
+	wsr.M3	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_load
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+#define XCHAL_SA_NUM_ATMPS	1
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/csp/include/variant/tie.h b/arch/xtensa/variants/csp/include/variant/tie.h
new file mode 100644
index 000000000..3ce391cfe
--- /dev/null
+++ b/arch/xtensa/variants/csp/include/variant/tie.h
@@ -0,0 +1,161 @@
+/* 
+ * tie.h -- compile-time HAL definitions dependent on CORE & TIE configuration
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file describes this specific Xtensa processor's TIE extensions
+   that extend basic Xtensa core functionality.  It is customized to this
+   Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			1	/* number of coprocessors */
+#define XCHAL_CP_MAX			8	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x80	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x80	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP7_NAME			"XTIOP"
+#define XCHAL_CP7_IDENT			XTIOP
+#define XCHAL_CP7_SA_SIZE		0	/* size of state save area */
+#define XCHAL_CP7_SA_ALIGN		1	/* min alignment of save area */
+#define XCHAL_CP_ID_XTIOP           	7	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP1_SA_SIZE		0
+#define XCHAL_CP1_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		36
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		48	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		4	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	9
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             br, 4, 4, 4,0x0204,  sr,4  , 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	0
+#define XCHAL_CP1_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+/* Byte length of instruction from its first byte, per FLIX.  */
+#define XCHAL_BYTE0_FORMAT_LENGTHS	\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/dc232b/include/variant/core.h b/arch/xtensa/variants/dc232b/include/variant/core.h
new file mode 100644
index 000000000..525bd3d90
--- /dev/null
+++ b/arch/xtensa/variants/dc232b/include/variant/core.h
@@ -0,0 +1,424 @@
+/*
+ * Xtensa processor core configuration information.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (c) 1999-2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	3	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		0	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	1	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		0	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			8	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector floating-point pkg */
+#define XCHAL_HAVE_FP			0	/* floating point pkg */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		4	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		4	/* data width in bytes */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+
+#define XCHAL_SW_VERSION		701001	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"dc232b"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_CORE_DESCRIPTION		"Diamond 232L Standard Core Rev.B (LE)"
+#define XCHAL_BUILD_UNIQUE_ID		0x0000BEEF	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC56307FE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x0D40BEEF	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX2.1.1"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2210	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		1	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		221001	/* major*100+minor */
+#define XCHAL_HW_REL_LX2		1
+#define XCHAL_HW_REL_LX2_1		1
+#define XCHAL_HW_REL_LX2_1_1		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2210	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	1	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		221001	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2210	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	1	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		221001	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		16384	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		7
+#define XCHAL_DCACHE_SETWIDTH		7
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		4
+#define XCHAL_DCACHE_WAYS		4
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		17	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		3	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F80FF
+#define XCHAL_INTLEVEL2_MASK		0x00000100
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00001000
+#define XCHAL_INTLEVEL5_MASK		0x00002000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F80FF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F81FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F8FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F9FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		1
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		4
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		1
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F8000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000133F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL2_NUM		8
+#define XCHAL_INTLEVEL4_NUM		12
+#define XCHAL_INTLEVEL5_NUM		13
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 3.)  */
+
+
+/*
+ *  External interrupt vectors/levels.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL interrupt number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 4) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		15	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT16_NUM		21	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0xD0000000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00000000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0xD8000500
+#define XCHAL_RESET_VECTOR1_PADDR	0x00000500
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0xD0000340
+#define XCHAL_USER_VECTOR_PADDR		0x00000340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0xD0000300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00000300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0xD00003C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000003C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0xD0000000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00000000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0xD0000180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00000180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0xD00001C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x000001C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0xD0000200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x00000200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0xD0000240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x00000240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0xD0000280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x00000280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0xD00002C0
+#define XCHAL_NMI_VECTOR_PADDR		0x000002C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	1	/* faster OCD option */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		0	/* one way maps I+D 4GB vaddr */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/dc232b/include/variant/tie-asm.h b/arch/xtensa/variants/dc232b/include/variant/tie-asm.h
new file mode 100644
index 000000000..ed4f53f52
--- /dev/null
+++ b/arch/xtensa/variants/dc232b/include/variant/tie-asm.h
@@ -0,0 +1,122 @@
+/*
+ * This header file contains assembly-language definitions (assembly
+ * macros, etc.) for this specific Xtensa processor's TIE extensions
+ * and options.  It is customized to this Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1999-2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+
+
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (1 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-8, 4, 4
+	rsr	\at1, ACCLO		// MAC16 accumulator
+	rsr	\at2, ACCHI
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	s32i	\at2, \ptr, .Lxchal_ofs_ + 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-16, 4, 4
+	rsr	\at1, M0		// MAC16 registers
+	rsr	\at2, M1
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	s32i	\at2, \ptr, .Lxchal_ofs_ + 4
+	rsr	\at1, M2
+	rsr	\at2, M3
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 8
+	s32i	\at2, \ptr, .Lxchal_ofs_ + 12
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 16
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rsr	\at1, SCOMPARE1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rur	\at1, THREADPTR		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_store
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (1 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-8, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	l32i	\at2, \ptr, .Lxchal_ofs_ + 4
+	wsr	\at1, ACCLO		// MAC16 accumulator
+	wsr	\at2, ACCHI
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-16, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	l32i	\at2, \ptr, .Lxchal_ofs_ + 4
+	wsr	\at1, M0		// MAC16 registers
+	wsr	\at2, M1
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 8
+	l32i	\at2, \ptr, .Lxchal_ofs_ + 12
+	wsr	\at1, M2
+	wsr	\at2, M3
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 16
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wsr	\at1, SCOMPARE1		// conditional store option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wur	\at1, THREADPTR		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_load
+
+
+
+#define XCHAL_NCP_NUM_ATMPS	2
+
+
+#define XCHAL_SA_NUM_ATMPS	2
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/dc232b/include/variant/tie.h b/arch/xtensa/variants/dc232b/include/variant/tie.h
new file mode 100644
index 000000000..018e81af4
--- /dev/null
+++ b/arch/xtensa/variants/dc232b/include/variant/tie.h
@@ -0,0 +1,131 @@
+/*
+ * This header file describes this specific Xtensa processor's TIE extensions
+ * that extend basic Xtensa core functionality.  It is customized to this
+ * Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1999-2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			1	/* number of coprocessors */
+#define XCHAL_CP_MAX			8	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x80	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x80	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP7_NAME			"XTIOP"
+#define XCHAL_CP7_IDENT			XTIOP
+#define XCHAL_CP7_SA_SIZE		0	/* size of state save area */
+#define XCHAL_CP7_SA_ALIGN		1	/* min alignment of save area */
+#define XCHAL_CP_ID_XTIOP		7	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP1_SA_SIZE		0
+#define XCHAL_CP1_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		32
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		32	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		4	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	8
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	0
+#define XCHAL_CP1_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/dc233c/include/variant/core.h b/arch/xtensa/variants/dc233c/include/variant/core.h
new file mode 100644
index 000000000..3a2e53b94
--- /dev/null
+++ b/arch/xtensa/variants/dc233c/include/variant/core.h
@@ -0,0 +1,475 @@
+/* 
+ * xtensa/config/core-isa.h -- HAL definitions that are dependent on Xtensa
+ *				processor CORE configuration
+ *
+ *  See <xtensa/config/core.h>, which includes this file, for more details.
+ */
+
+/* Xtensa processor core configuration information.
+
+   Copyright (c) 1999-2010 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	3	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		0	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	1	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MP_INTERRUPTS	0	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		0	/* core RunStall control port */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		0	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			8	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector floating-point pkg */
+#define XCHAL_HAVE_FP			0	/* floating point pkg */
+#define XCHAL_HAVE_DFP			0	/* double precision FP pkg */
+#define XCHAL_HAVE_DFP_accel		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2EP	0	/* HiFi2EP */
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+#define XCHAL_HAVE_BBE16		0	/* ConnX BBE16 pkg */
+#define XCHAL_HAVE_BBE16_RSQRT		0	/* BBE16 & vector recip sqrt */
+#define XCHAL_HAVE_BBE16_VECDIV		0	/* BBE16 & vector divide */
+#define XCHAL_HAVE_BBE16_DESPREAD	0	/* BBE16 & despread */
+#define XCHAL_HAVE_BSP3			0	/* ConnX BSP3 pkg */
+#define XCHAL_HAVE_SSP16		0	/* ConnX SSP16 pkg */
+#define XCHAL_HAVE_SSP16_VITERBI	0	/* SSP16 & viterbi */
+#define XCHAL_HAVE_TURBO16		0	/* ConnX Turbo16 pkg */
+#define XCHAL_HAVE_BBP16		0	/* ConnX BBP16 pkg */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		4	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		4	/* data width in bytes */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		900001	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"dc233c"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_CORE_DESCRIPTION		"dc233c"
+#define XCHAL_BUILD_UNIQUE_ID		0x00004B21	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC56707FE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x14404B21	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX4.0.1"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2400	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		1	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		240001	/* major*100+minor */
+#define XCHAL_HW_REL_LX4		1
+#define XCHAL_HW_REL_LX4_0		1
+#define XCHAL_HW_REL_LX4_0_1		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2400	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	1	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		240001	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2400	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	1	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		240001	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		16384	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	0	/* MP coherence feature */
+
+#define XCHAL_HAVE_PREFETCH		0	/* PREFCTL register */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		7
+#define XCHAL_DCACHE_SETWIDTH		7
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		4
+#define XCHAL_DCACHE_WAYS		4
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	4
+#define XCHAL_DCACHE_ACCESS_SIZE	4
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+#define XCHAL_HAVE_IMEM_LOADSTORE	1	/* can load/store to IROM/IRAM*/
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		17	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		3	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F80FF
+#define XCHAL_INTLEVEL2_MASK		0x00000100
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00001000
+#define XCHAL_INTLEVEL5_MASK		0x00002000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F80FF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F81FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F8FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F9FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		1
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		4
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		1
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F8000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000133F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL2_NUM		8
+#define XCHAL_INTLEVEL4_NUM		12
+#define XCHAL_INTLEVEL5_NUM		13
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 3.)  */
+
+
+/*
+ *  External interrupt vectors/levels.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL interrupt number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 4) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		15	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT16_NUM		21	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) or TX */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_HALT			0	/* halt architecture option */
+#define XCHAL_HAVE_BOOTLOADER		0	/* boot loader (for TX) */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0x00002000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00002000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0x00001000
+#define XCHAL_RESET_VECTOR1_PADDR	0x00001000
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0x00002340
+#define XCHAL_USER_VECTOR_PADDR		0x00002340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0x00002300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00002300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0x000023C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000023C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0x00002000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00002000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0x00002180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00002180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0x000021C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x000021C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0x00002200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x00002200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0x00002240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x00002240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0x00002280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x00002280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0x000022C0
+#define XCHAL_NMI_VECTOR_PADDR		0x000022C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	1	/* faster OCD option */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		1	/* one way maps I+D 4GB vaddr */
+#define XCHAL_SPANNING_WAY		6	/* TLB spanning way number */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/dc233c/include/variant/tie-asm.h b/arch/xtensa/variants/dc233c/include/variant/tie-asm.h
new file mode 100644
index 000000000..5dbd981ea
--- /dev/null
+++ b/arch/xtensa/variants/dc233c/include/variant/tie-asm.h
@@ -0,0 +1,193 @@
+/* 
+ * tie-asm.h -- compile-time HAL assembler definitions dependent on CORE & TIE
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file contains assembly-language definitions (assembly
+   macros, etc.) for this specific Xtensa processor's TIE extensions
+   and options.  It is customized to this Xtensa processor configuration.
+
+   Copyright (c) 1999-2010 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+#define XTHAL_SAS_ANYOT	0x0003	/* both of the above */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+#define XTHAL_SAS_ANYCC	0x000C	/* both of the above */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+#define XTHAL_SAS_ANYABI 0x0070	/* all of the above three */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+#define XTHAL_SAS3(optie,ccuse,abi)	( ((optie) & XTHAL_SAS_ANYOT)  \
+					| ((ccuse) & XTHAL_SAS_ANYCC)  \
+					| ((abi)   & XTHAL_SAS_ANYABI) )
+
+
+
+    /*
+     *  Macro to save all non-coprocessor (extra) custom TIE and optional state
+     *  (not including zero-overhead loop registers).
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 4 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters:
+     *      continue	If macro invoked as part of a larger store sequence, set to 1
+     *			if this is not the first in the sequence.  Defaults to 0.
+     *      ofs		Offset from start of larger sequence (from value of first ptr
+     *			in sequence) at which to store.  Defaults to next available space
+     *			(or 0 if <continue> is 0).
+     *      select	Select what category(ies) of registers to store, as a bitmask
+     *			(see XTHAL_SAS_xxx constants).  Defaults to all registers.
+     *      alloc	Select what category(ies) of registers to allocate; if any
+     *			category is selected here that is not in <select>, space for
+     *			the corresponding registers is skipped without doing any store.
+     */
+    .macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global register used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	rur.THREADPTR	\at1		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	rsr	\at1, ACCLO			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr	\at1, ACCHI			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	rsr	\at1, M0			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr	\at1, M1			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rsr	\at1, M2			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rsr	\at1, M3			// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rsr	\at1, SCOMPARE1			// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.endif
+    .endm	// xchal_ncp_store
+
+    /*
+     *  Macro to restore all non-coprocessor (extra) custom TIE and optional state
+     *  (not including zero-overhead loop registers).
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 4 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters:
+     *      continue	If macro invoked as part of a larger load sequence, set to 1
+     *			if this is not the first in the sequence.  Defaults to 0.
+     *      ofs		Offset from start of larger sequence (from value of first ptr
+     *			in sequence) at which to load.  Defaults to next available space
+     *			(or 0 if <continue> is 0).
+     *      select	Select what category(ies) of registers to load, as a bitmask
+     *			(see XTHAL_SAS_xxx constants).  Defaults to all registers.
+     *      alloc	Select what category(ies) of registers to allocate; if any
+     *			category is selected here that is not in <select>, space for
+     *			the corresponding registers is skipped without doing any load.
+     */
+    .macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global register used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.THREADPTR	\at1		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr	\at1, ACCLO			// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr	\at1, ACCHI			// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr	\at1, M0			// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr	\at1, M1			// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wsr	\at1, M2			// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wsr	\at1, M3			// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wsr	\at1, SCOMPARE1			// conditional store option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.endif
+    .endm	// xchal_ncp_load
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+
+
+#define XCHAL_SA_NUM_ATMPS	1
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/dc233c/include/variant/tie.h b/arch/xtensa/variants/dc233c/include/variant/tie.h
new file mode 100644
index 000000000..815e52bc3
--- /dev/null
+++ b/arch/xtensa/variants/dc233c/include/variant/tie.h
@@ -0,0 +1,150 @@
+/* 
+ * tie.h -- compile-time HAL definitions dependent on CORE & TIE configuration
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file describes this specific Xtensa processor's TIE extensions
+   that extend basic Xtensa core functionality.  It is customized to this
+   Xtensa processor configuration.
+
+   Copyright (c) 1999-2010 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			1	/* number of coprocessors */
+#define XCHAL_CP_MAX			8	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x80	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x80	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP7_NAME			"XTIOP"
+#define XCHAL_CP7_IDENT			XTIOP
+#define XCHAL_CP7_SA_SIZE		0	/* size of state save area */
+#define XCHAL_CP7_SA_ALIGN		1	/* min alignment of save area */
+#define XCHAL_CP_ID_XTIOP		7	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP1_SA_SIZE		0
+#define XCHAL_CP1_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		32
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		32	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		4	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	8
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	0
+#define XCHAL_CP1_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/de212/include/variant/core.h b/arch/xtensa/variants/de212/include/variant/core.h
new file mode 100644
index 000000000..59e91e47e
--- /dev/null
+++ b/arch/xtensa/variants/de212/include/variant/core.h
@@ -0,0 +1,594 @@
+/* 
+ * xtensa/config/core-isa.h -- HAL definitions that are dependent on Xtensa
+ *				processor CORE configuration
+ *
+ *  See <xtensa/config/core.h>, which includes this file, for more details.
+ */
+
+/* Xtensa processor core configuration information.
+
+   Copyright (c) 1999-2015 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	3	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_LOOP_BUFFER_SIZE		0	/* zero-ov. loop instr buffer size */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_DEPBITS		0	/* DEPBITS instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		0	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	0	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MX			0	/* MX core (Tensilica internal) */
+#define XCHAL_HAVE_MP_INTERRUPTS	0	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		0	/* core RunStall control port */
+#define XCHAL_HAVE_PSO			0	/* Power Shut-Off */
+#define XCHAL_HAVE_PSO_CDM		0	/* core/debug/mem pwr domains */
+#define XCHAL_HAVE_PSO_FULL_RETENTION	0	/* all regs preserved on PSO */
+#define XCHAL_HAVE_THREADPTR		0	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		0	/* boolean registers */
+#define XCHAL_HAVE_CP			0	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			0	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+
+#define XCHAL_HAVE_FUSION		 0	/* Fusion*/
+#define XCHAL_HAVE_FUSION_FP	 0	        /* Fusion FP option */
+#define XCHAL_HAVE_FUSION_LOW_POWER 0	/* Fusion Low Power option */
+#define XCHAL_HAVE_FUSION_AES	 0	        /* Fusion BLE/Wifi AES-128 CCM option */
+#define XCHAL_HAVE_FUSION_CONVENC	 0       /* Fusion Conv Encode option */
+#define XCHAL_HAVE_FUSION_LFSR_CRC	 0	/* Fusion LFSR-CRC option */
+#define XCHAL_HAVE_FUSION_BITOPS	 0	/* Fusion Bit Operations Support option */
+#define XCHAL_HAVE_FUSION_AVS	 0	/* Fusion AVS option */
+#define XCHAL_HAVE_FUSION_16BIT_BASEBAND	 0	/* Fusion 16-bit Baseband option */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4		0	/* HiFi4 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4_VFPU		0	/* HiFi4 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI3		0	/* HiFi3 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI3_VFPU		0	/* HiFi3 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2EP		0	/* HiFi2EP */
+#define XCHAL_HAVE_HIFI_MINI		0	
+
+
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector or user floating-point pkg */
+#define XCHAL_HAVE_USER_DPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_USER_SPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_FP                 0      /* single prec floating point */
+#define XCHAL_HAVE_FP_DIV             0  /* FP with DIV instructions */
+#define XCHAL_HAVE_FP_RECIP           0        /* FP with RECIP instructions */
+#define XCHAL_HAVE_FP_SQRT            0 /* FP with SQRT instructions */
+#define XCHAL_HAVE_FP_RSQRT           0        /* FP with RSQRT instructions */
+#define XCHAL_HAVE_DFP                        0     /* double precision FP pkg */
+#define XCHAL_HAVE_DFP_DIV            0 /* DFP with DIV instructions */
+#define XCHAL_HAVE_DFP_RECIP          0       /* DFP with RECIP instructions*/
+#define XCHAL_HAVE_DFP_SQRT           0        /* DFP with SQRT instructions */
+#define XCHAL_HAVE_DFP_RSQRT          0       /* DFP with RSQRT instructions*/
+#define XCHAL_HAVE_DFP_ACCEL		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_DFP_accel		XCHAL_HAVE_DFP_ACCEL				/* for backward compatibility */
+
+#define XCHAL_HAVE_DFPU_SINGLE_ONLY    0                 	/* DFPU Coprocessor, single precision only */
+#define XCHAL_HAVE_DFPU_SINGLE_DOUBLE  0               	/* DFPU Coprocessor, single and double precision */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_PDX4		        0	/* PDX4 */
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+#define XCHAL_HAVE_CONNXD2_DUALLSFLIX   0	/* ConnX D2 & Dual LoadStore Flix */
+#define XCHAL_HAVE_BBE16		0	/* ConnX BBE16 pkg */
+#define XCHAL_HAVE_BBE16_RSQRT		0	/* BBE16 & vector recip sqrt */
+#define XCHAL_HAVE_BBE16_VECDIV		0	/* BBE16 & vector divide */
+#define XCHAL_HAVE_BBE16_DESPREAD	0	/* BBE16 & despread */
+#define XCHAL_HAVE_BBENEP		0	/* ConnX BBENEP pkgs */
+#define XCHAL_HAVE_BSP3			0	/* ConnX BSP3 pkg */
+#define XCHAL_HAVE_BSP3_TRANSPOSE	0	/* BSP3 & transpose32x32 */
+#define XCHAL_HAVE_SSP16		0	/* ConnX SSP16 pkg */
+#define XCHAL_HAVE_SSP16_VITERBI	0	/* SSP16 & viterbi */
+#define XCHAL_HAVE_TURBO16		0	/* ConnX Turbo16 pkg */
+#define XCHAL_HAVE_BBP16		0	/* ConnX BBP16 pkg */
+#define XCHAL_HAVE_FLIX3		0	/* basic 3-way FLIX option */
+#define XCHAL_HAVE_GRIVPEP              0   /*  GRIVPEP is General Release of IVPEP */
+#define XCHAL_HAVE_GRIVPEP_HISTOGRAM    0   /* Histogram option on GRIVPEP */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_LOADSTORE_UNITS	1	/* load/store units */
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		4	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		4	/* data width in bytes */
+#define XCHAL_DATA_PIPE_DELAY		1	/* d-side pipeline delay
+						   (1 = 5-stage, 2 = 7-stage) */
+#define XCHAL_CLOCK_GATING_GLOBAL	0	/* global clock gating */
+#define XCHAL_CLOCK_GATING_FUNCUNIT	0	/* funct. unit clock gating */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		1100002	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"de212"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_BUILD_UNIQUE_ID		0x0005A985	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC283DFFE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x1C85A985	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX6.0.2"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2600	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		2	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		260002	/* major*100+minor */
+#define XCHAL_HW_REL_LX6		1
+#define XCHAL_HW_REL_LX6_0		1
+#define XCHAL_HW_REL_LX6_0_2		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2600	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	2	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		260002	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2600	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	2	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		260002	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		8192	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		8192	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	0	/* MP coherence feature */
+
+#define XCHAL_HAVE_PREFETCH		0	/* PREFCTL register */
+#define XCHAL_HAVE_PREFETCH_L1		0	/* prefetch to L1 dcache */
+#define XCHAL_PREFETCH_CASTOUT_LINES	0	/* dcache pref. castout bufsz */
+#define XCHAL_PREFETCH_ENTRIES		0	/* cache prefetch entries */
+#define XCHAL_PREFETCH_BLOCK_ENTRIES	0	/* prefetch block streams */
+#define XCHAL_HAVE_CACHE_BLOCKOPS	0	/* block prefetch for caches */
+#define XCHAL_HAVE_ICACHE_TEST		1	/* Icache test instructions */
+#define XCHAL_HAVE_DCACHE_TEST		1	/* Dcache test instructions */
+#define XCHAL_HAVE_ICACHE_DYN_WAYS	0	/* Icache dynamic way support */
+#define XCHAL_HAVE_DCACHE_DYN_WAYS	0	/* Dcache dynamic way support */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		7
+#define XCHAL_DCACHE_SETWIDTH		7
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		2
+#define XCHAL_DCACHE_WAYS		2
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	4
+#define XCHAL_DCACHE_ACCESS_SIZE	4
+
+#define XCHAL_DCACHE_BANKS		1	/* number of banks */
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+/*  Whether MEMCTL register has anything useful  */
+#define XCHAL_USE_MEMCTL		(((XCHAL_LOOP_BUFFER_SIZE > 0)	||	\
+					   XCHAL_DCACHE_IS_COHERENT	||	\
+					   XCHAL_HAVE_ICACHE_DYN_WAYS	||	\
+					   XCHAL_HAVE_DCACHE_DYN_WAYS)	&&	\
+					   (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RE_2012_0))
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		1	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		1	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			1	/* number of core XLMI ports */
+
+/*  Instruction RAM 0:  */
+#define XCHAL_INSTRAM0_VADDR		0x40000000	/* virtual address */
+#define XCHAL_INSTRAM0_PADDR		0x40000000	/* physical address */
+#define XCHAL_INSTRAM0_SIZE		131072	/* size in bytes */
+#define XCHAL_INSTRAM0_ECC_PARITY	0	/* ECC/parity type, 0=none */
+
+/*  Data RAM 0:  */
+#define XCHAL_DATARAM0_VADDR		0x3FFE0000	/* virtual address */
+#define XCHAL_DATARAM0_PADDR		0x3FFE0000	/* physical address */
+#define XCHAL_DATARAM0_SIZE		131072	/* size in bytes */
+#define XCHAL_DATARAM0_ECC_PARITY	0	/* ECC/parity type, 0=none */
+#define XCHAL_DATARAM0_BANKS		1	/* number of banks */
+
+/*  XLMI Port 0:  */
+#define XCHAL_XLMI0_VADDR		0x3FFC0000	/* virtual address */
+#define XCHAL_XLMI0_PADDR		0x3FFC0000	/* physical address */
+#define XCHAL_XLMI0_SIZE		131072	/* size in bytes */
+#define XCHAL_XLMI0_ECC_PARITY	0	/* ECC/parity type, 0=none */
+
+#define XCHAL_HAVE_IMEM_LOADSTORE	1	/* can load/store to IROM/IRAM*/
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		17	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		3	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F80FF
+#define XCHAL_INTLEVEL2_MASK		0x00000100
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00001000
+#define XCHAL_INTLEVEL5_MASK		0x00002000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F80FF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F81FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F8FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F9FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		1
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		4
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		1
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F8000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000133F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+#define XCHAL_INTTYPE_MASK_PROFILING	0x00000000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL2_NUM		8
+#define XCHAL_INTLEVEL4_NUM		12
+#define XCHAL_INTLEVEL5_NUM		13
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 3.)  */
+
+
+/*
+ *  External interrupt mapping.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL BInterrupt pin number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 4) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		15	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT16_NUM		21	/* (intlevel 3) */
+/*  EXTERNAL BInterrupt pin numbers mapped to each core interrupt number:  */
+#define XCHAL_INT0_EXTNUM		0	/* (intlevel 1) */
+#define XCHAL_INT1_EXTNUM		1	/* (intlevel 1) */
+#define XCHAL_INT2_EXTNUM		2	/* (intlevel 1) */
+#define XCHAL_INT3_EXTNUM		3	/* (intlevel 1) */
+#define XCHAL_INT4_EXTNUM		4	/* (intlevel 1) */
+#define XCHAL_INT5_EXTNUM		5	/* (intlevel 1) */
+#define XCHAL_INT8_EXTNUM		6	/* (intlevel 2) */
+#define XCHAL_INT9_EXTNUM		7	/* (intlevel 3) */
+#define XCHAL_INT12_EXTNUM		8	/* (intlevel 4) */
+#define XCHAL_INT14_EXTNUM		9	/* (intlevel 7) */
+#define XCHAL_INT15_EXTNUM		10	/* (intlevel 1) */
+#define XCHAL_INT16_EXTNUM		11	/* (intlevel 1) */
+#define XCHAL_INT17_EXTNUM		12	/* (intlevel 1) */
+#define XCHAL_INT18_EXTNUM		13	/* (intlevel 1) */
+#define XCHAL_INT19_EXTNUM		14	/* (intlevel 1) */
+#define XCHAL_INT20_EXTNUM		15	/* (intlevel 1) */
+#define XCHAL_INT21_EXTNUM		16	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) or TX */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_HALT			0	/* halt architecture option */
+#define XCHAL_HAVE_BOOTLOADER		0	/* boot loader (for TX) */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0x60000000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x60000000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0x50000000
+#define XCHAL_RESET_VECTOR0_PADDR	0x50000000
+#define XCHAL_RESET_VECTOR1_VADDR	0x40000400
+#define XCHAL_RESET_VECTOR1_PADDR	0x40000400
+#define XCHAL_RESET_VECTOR_VADDR	0x50000000
+#define XCHAL_RESET_VECTOR_PADDR	0x50000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0x60000340
+#define XCHAL_USER_VECTOR_PADDR		0x60000340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0x60000300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x60000300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0x600003C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x600003C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0x60000000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x60000000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0x60000180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x60000180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0x600001C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x600001C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0x60000200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x60000200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0x60000240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x60000240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0x60000280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x60000280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0x600002C0
+#define XCHAL_NMI_VECTOR_PADDR		0x600002C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG MODULE
+  ----------------------------------------------------------------------*/
+
+/*  Misc  */
+#define XCHAL_HAVE_DEBUG_ERI		1	/* ERI to debug module */
+#define XCHAL_HAVE_DEBUG_APB		0	/* APB to debug module */
+#define XCHAL_HAVE_DEBUG_JTAG		1	/* JTAG to debug module */
+
+/*  On-Chip Debug (OCD)  */
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	0	/* faster OCD option (to LX4) */
+#define XCHAL_HAVE_OCD_LS32DDR		1	/* L32DDR/S32DDR (faster OCD) */
+
+/*  TRAX (in core)  */
+#define XCHAL_HAVE_TRAX			1	/* TRAX in debug module */
+#define XCHAL_TRAX_MEM_SIZE		262144	/* TRAX memory size in bytes */
+#define XCHAL_TRAX_MEM_SHAREABLE	0	/* start/end regs; ready sig. */
+#define XCHAL_TRAX_ATB_WIDTH		0	/* ATB width (bits), 0=no ATB */
+#define XCHAL_TRAX_TIME_WIDTH		0	/* timestamp bitwidth, 0=none */
+
+/*  Perf counters  */
+#define XCHAL_NUM_PERF_COUNTERS		0	/* performance counters */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		1	/* one way maps I+D 4GB vaddr */
+#define XCHAL_SPANNING_WAY		0	/* TLB spanning way number */
+#define XCHAL_HAVE_IDENTITY_MAP		1	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	1	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		0	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+
+#define XCHAL_MMU_ASID_BITS		0	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			1	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		0	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/de212/include/variant/tie-asm.h b/arch/xtensa/variants/de212/include/variant/tie-asm.h
new file mode 100644
index 000000000..77755354f
--- /dev/null
+++ b/arch/xtensa/variants/de212/include/variant/tie-asm.h
@@ -0,0 +1,170 @@
+/* 
+ * tie-asm.h -- compile-time HAL assembler definitions dependent on CORE & TIE
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file contains assembly-language definitions (assembly
+   macros, etc.) for this specific Xtensa processor's TIE extensions
+   and options.  It is customized to this Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+#define XTHAL_SAS_ANYOT	0x0003	/* both of the above */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+#define XTHAL_SAS_ANYCC	0x000C	/* both of the above */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+#define XTHAL_SAS_ANYABI	0x0070	/* all of the above three */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+#define XTHAL_SAS3(optie,ccuse,abi)	( ((optie) & XTHAL_SAS_ANYOT)  \
+					| ((ccuse) & XTHAL_SAS_ANYCC)  \
+					| ((abi)   & XTHAL_SAS_ANYABI) )
+
+
+    /*
+      *  Macro to store all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger store sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to store.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to store, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any store.
+      */
+    .macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	rsr.ACCLO	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.ACCHI	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	rsr.SCOMPARE1	\at1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.M0	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rsr.M1	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rsr.M2	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rsr.M3	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.endif
+    .endm	// xchal_ncp_store
+
+    /*
+      *  Macro to load all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger load sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to load.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to load, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any load.
+      */
+    .macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.ACCLO	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.ACCHI	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.SCOMPARE1	\at1		// conditional store option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.M0	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wsr.M1	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wsr.M2	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wsr.M3	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1004, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 20
+	.endif
+    .endm	// xchal_ncp_load
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+#define XCHAL_SA_NUM_ATMPS	1
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/de212/include/variant/tie.h b/arch/xtensa/variants/de212/include/variant/tie.h
new file mode 100644
index 000000000..b8a061a3f
--- /dev/null
+++ b/arch/xtensa/variants/de212/include/variant/tie.h
@@ -0,0 +1,136 @@
+/* 
+ * tie.h -- compile-time HAL definitions dependent on CORE & TIE configuration
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file describes this specific Xtensa processor's TIE extensions
+   that extend basic Xtensa core functionality.  It is customized to this
+   Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			0	/* number of coprocessors */
+#define XCHAL_CP_MAX			0	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x00	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x00	/* bitmask of only port CPs */
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		28
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		32	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		4	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	7
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	0
+#define XCHAL_CP1_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+/* Byte length of instruction from its first byte, per FLIX.  */
+#define XCHAL_BYTE0_FORMAT_LENGTHS	\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/fsf/include/variant/core.h b/arch/xtensa/variants/fsf/include/variant/core.h
new file mode 100644
index 000000000..2f337605c
--- /dev/null
+++ b/arch/xtensa/variants/fsf/include/variant/core.h
@@ -0,0 +1,359 @@
+/*
+ * Xtensa processor core configuration information.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1999-2006 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_H
+#define _XTENSA_CORE_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			1	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			64	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		6	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	3	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		0	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			0	/* SEXT instruction */
+#define XCHAL_HAVE_CLAMPS		0	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		0	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		0	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		0	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	1	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		0	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		0	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		0	/* boolean registers */
+#define XCHAL_HAVE_CP			0	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			0	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		0	/* MAC16 package */
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector floating-point pkg */
+#define XCHAL_HAVE_FP			0	/* floating point pkg */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	4	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		4	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		4	/* data width in bytes */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+
+#define XCHAL_CORE_ID			"fsf"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_BUILD_UNIQUE_ID		0x00006700	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC103C3FF	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x0C006700	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX2.0.0"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2200	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		0	/* minor ver# of targeted hw */
+#define XTHAL_HW_REL_LX2		1
+#define XTHAL_HW_REL_LX2_0		1
+#define XTHAL_HW_REL_LX2_0_0		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2200	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	0	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2200	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	0	/* minor v of latest tgt hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		16	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		16	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		4	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		4	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		8192	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		8192	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	0	/* writeback feature */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		8
+#define XCHAL_DCACHE_SETWIDTH		8
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		2
+#define XCHAL_DCACHE_WAYS		2
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	0
+#define XCHAL_DCACHE_LINE_LOCKABLE	0
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			0	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		17	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		10	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		4	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		1	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x000064F9
+#define XCHAL_INTLEVEL2_MASK		0x00008902
+#define XCHAL_INTLEVEL3_MASK		0x00011204
+#define XCHAL_INTLEVEL4_MASK		0x00000000
+#define XCHAL_INTLEVEL5_MASK		0x00000000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00000000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x000064F9
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x0000EDFB
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x0001FFFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x0001FFFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x0001FFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x0001FFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x0001FFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		2
+#define XCHAL_INT2_LEVEL		3
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		1
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		1
+#define XCHAL_INT11_LEVEL		2
+#define XCHAL_INT12_LEVEL		3
+#define XCHAL_INT13_LEVEL		1
+#define XCHAL_INT14_LEVEL		1
+#define XCHAL_INT15_LEVEL		2
+#define XCHAL_INT16_LEVEL		3
+#define XCHAL_DEBUGLEVEL		4	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	0	/* OCD external db interrupt */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_SOFTWARE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFFE0000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x0001E000
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x00000380
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000007F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00001C00
+#define XCHAL_INTTYPE_MASK_NMI		0x00000000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		10	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		11	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		12	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+/*  (There are many interrupts each at level(s) 1, 2, 3.)  */
+
+
+/*
+ *  External interrupt vectors/levels.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL interrupt number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 2) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 3) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		6	/* (intlevel 1) */
+#define XCHAL_EXTINT7_NUM		7	/* (intlevel 1) */
+#define XCHAL_EXTINT8_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT9_NUM		9	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000020
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000020
+#define XCHAL_USER_VECTOR_VADDR		0xD0000220
+#define XCHAL_USER_VECTOR_PADDR		0x00000220
+#define XCHAL_KERNEL_VECTOR_VADDR	0xD0000200
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00000200
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0xD0000290
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x00000290
+#define XCHAL_WINDOW_VECTORS_VADDR	0xD0000000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00000000
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0xD0000240
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00000240
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0xD0000250
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x00000250
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0xFE000520
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0xFE000520
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL4_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL4_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	1	/* faster OCD option */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See <xtensa/config/core-matmap.h> header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		0	/* one way maps I+D 4GB vaddr */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/fsf/include/variant/tie-asm.h b/arch/xtensa/variants/fsf/include/variant/tie-asm.h
new file mode 100644
index 000000000..68a73bf4f
--- /dev/null
+++ b/arch/xtensa/variants/fsf/include/variant/tie-asm.h
@@ -0,0 +1,70 @@
+/*
+ * This header file contains assembly-language definitions (assembly
+ * macros, etc.) for this specific Xtensa processor's TIE extensions
+ * and options.  It is customized to this Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1999-2008 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+
+
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (1 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rur	\at1, THREADPTR		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_store
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (1 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wur	\at1, THREADPTR		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_load
+
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+
+#define XCHAL_SA_NUM_ATMPS	1
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/fsf/include/variant/tie.h b/arch/xtensa/variants/fsf/include/variant/tie.h
new file mode 100644
index 000000000..244cdea4d
--- /dev/null
+++ b/arch/xtensa/variants/fsf/include/variant/tie.h
@@ -0,0 +1,72 @@
+/*
+ * This header file describes this specific Xtensa processor's TIE extensions
+ * that extend basic Xtensa core functionality.  It is customized to this
+ * Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License.  See the file "COPYING" in the main directory of this archive
+ * for more details.
+ *
+ * Copyright (C) 1999-2007 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			0	/* number of coprocessors */
+#define XCHAL_CP_MAX			0	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x00	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x00	/* bitmask of only port CPs */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_NCP_SA_SIZE		0
+#define XCHAL_NCP_SA_ALIGN		1
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP1_SA_SIZE		0
+#define XCHAL_CP1_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+#define XCHAL_CP7_SA_SIZE		0
+#define XCHAL_CP7_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		0
+#define XCHAL_NCP_SA_ALIGN		1
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		0	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		1	/* actual minimum alignment */
+
+#define XCHAL_NCP_SA_NUM	0
+#define XCHAL_NCP_SA_LIST(s)
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)
+#define XCHAL_CP1_SA_NUM	0
+#define XCHAL_CP1_SA_LIST(s)
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,3
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/test_kc705_be/include/variant/core.h b/arch/xtensa/variants/test_kc705_be/include/variant/core.h
new file mode 100644
index 000000000..a4ebdf719
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_be/include/variant/core.h
@@ -0,0 +1,575 @@
+/* 
+ * xtensa/config/core-isa.h -- HAL definitions that are dependent on Xtensa
+ *				processor CORE configuration
+ *
+ *  See <xtensa/config/core.h>, which includes this file, for more details.
+ */
+
+/* Xtensa processor core configuration information.
+
+   Copyright (c) 1999-2015 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			1	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	8	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_LOOP_BUFFER_SIZE		0	/* zero-ov. loop instr buffer size */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_DEPBITS		0	/* DEPBITS instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		1	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	0	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MX			0	/* MX core (Tensilica internal) */
+#define XCHAL_HAVE_MP_INTERRUPTS	0	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		0	/* core RunStall control port */
+#define XCHAL_HAVE_PSO			0	/* Power Shut-Off */
+#define XCHAL_HAVE_PSO_CDM		0	/* core/debug/mem pwr domains */
+#define XCHAL_HAVE_PSO_FULL_RETENTION	0	/* all regs preserved on PSO */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		1	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			8	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+
+#define XCHAL_HAVE_FUSION		 0	/* Fusion*/
+#define XCHAL_HAVE_FUSION_FP	 0	        /* Fusion FP option */
+#define XCHAL_HAVE_FUSION_LOW_POWER 0	/* Fusion Low Power option */
+#define XCHAL_HAVE_FUSION_AES	 0	        /* Fusion BLE/Wifi AES-128 CCM option */
+#define XCHAL_HAVE_FUSION_CONVENC	 0       /* Fusion Conv Encode option */
+#define XCHAL_HAVE_FUSION_LFSR_CRC	 0	/* Fusion LFSR-CRC option */
+#define XCHAL_HAVE_FUSION_BITOPS	 0	/* Fusion Bit Operations Support option */
+#define XCHAL_HAVE_FUSION_AVS	 0	/* Fusion AVS option */
+#define XCHAL_HAVE_FUSION_16BIT_BASEBAND	 0	/* Fusion 16-bit Baseband option */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4		0	/* HiFi4 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI4_VFPU		0	/* HiFi4 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI3		0	/* HiFi3 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI3_VFPU		0	/* HiFi3 Audio Engine VFPU option */
+#define XCHAL_HAVE_HIFI2		1	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2_MUL32X24	1	/* HiFi2 and 32x24 MACs */
+#define XCHAL_HAVE_HIFI2EP		1	/* HiFi2EP */
+#define XCHAL_HAVE_HIFI_MINI		0	
+
+
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector or user floating-point pkg */
+#define XCHAL_HAVE_USER_DPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_USER_SPFPU         0       /* user DP floating-point pkg */
+#define XCHAL_HAVE_FP                 0      /* single prec floating point */
+#define XCHAL_HAVE_FP_DIV             0  /* FP with DIV instructions */
+#define XCHAL_HAVE_FP_RECIP           0        /* FP with RECIP instructions */
+#define XCHAL_HAVE_FP_SQRT            0 /* FP with SQRT instructions */
+#define XCHAL_HAVE_FP_RSQRT           0        /* FP with RSQRT instructions */
+#define XCHAL_HAVE_DFP                        0     /* double precision FP pkg */
+#define XCHAL_HAVE_DFP_DIV            0 /* DFP with DIV instructions */
+#define XCHAL_HAVE_DFP_RECIP          0       /* DFP with RECIP instructions*/
+#define XCHAL_HAVE_DFP_SQRT           0        /* DFP with SQRT instructions */
+#define XCHAL_HAVE_DFP_RSQRT          0       /* DFP with RSQRT instructions*/
+#define XCHAL_HAVE_DFP_ACCEL		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_DFP_accel		XCHAL_HAVE_DFP_ACCEL				/* for backward compatibility */
+
+#define XCHAL_HAVE_DFPU_SINGLE_ONLY    0                 	/* DFPU Coprocessor, single precision only */
+#define XCHAL_HAVE_DFPU_SINGLE_DOUBLE  0               	/* DFPU Coprocessor, single and double precision */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_PDX4		        0	/* PDX4 */
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+#define XCHAL_HAVE_CONNXD2_DUALLSFLIX   0	/* ConnX D2 & Dual LoadStore Flix */
+#define XCHAL_HAVE_BBE16		0	/* ConnX BBE16 pkg */
+#define XCHAL_HAVE_BBE16_RSQRT		0	/* BBE16 & vector recip sqrt */
+#define XCHAL_HAVE_BBE16_VECDIV		0	/* BBE16 & vector divide */
+#define XCHAL_HAVE_BBE16_DESPREAD	0	/* BBE16 & despread */
+#define XCHAL_HAVE_BBENEP		0	/* ConnX BBENEP pkgs */
+#define XCHAL_HAVE_BSP3			0	/* ConnX BSP3 pkg */
+#define XCHAL_HAVE_BSP3_TRANSPOSE	0	/* BSP3 & transpose32x32 */
+#define XCHAL_HAVE_SSP16		0	/* ConnX SSP16 pkg */
+#define XCHAL_HAVE_SSP16_VITERBI	0	/* SSP16 & viterbi */
+#define XCHAL_HAVE_TURBO16		0	/* ConnX Turbo16 pkg */
+#define XCHAL_HAVE_BBP16		0	/* ConnX BBP16 pkg */
+#define XCHAL_HAVE_FLIX3		0	/* basic 3-way FLIX option */
+#define XCHAL_HAVE_GRIVPEP              0   /*  GRIVPEP is General Release of IVPEP */
+#define XCHAL_HAVE_GRIVPEP_HISTOGRAM    0   /* Histogram option on GRIVPEP */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_LOADSTORE_UNITS	1	/* load/store units */
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		8	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		8	/* data width in bytes */
+#define XCHAL_DATA_PIPE_DELAY		1	/* d-side pipeline delay
+						   (1 = 5-stage, 2 = 7-stage) */
+#define XCHAL_CLOCK_GATING_GLOBAL	0	/* global clock gating */
+#define XCHAL_CLOCK_GATING_FUNCUNIT	0	/* funct. unit clock gating */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		1100002	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"test_kc705_be"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_BUILD_UNIQUE_ID		0x00058D8C	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC1B3FFFF	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x1C858D8C	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX6.0.2"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2600	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		2	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		260002	/* major*100+minor */
+#define XCHAL_HW_REL_LX6		1
+#define XCHAL_HW_REL_LX6_0		1
+#define XCHAL_HW_REL_LX6_0_2		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2600	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	2	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		260002	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2600	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	2	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		260002	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		16384	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	0	/* MP coherence feature */
+
+#define XCHAL_HAVE_PREFETCH		1	/* PREFCTL register */
+#define XCHAL_HAVE_PREFETCH_L1		0	/* prefetch to L1 dcache */
+#define XCHAL_PREFETCH_CASTOUT_LINES	1	/* dcache pref. castout bufsz */
+#define XCHAL_PREFETCH_ENTRIES		8	/* cache prefetch entries */
+#define XCHAL_PREFETCH_BLOCK_ENTRIES	0	/* prefetch block streams */
+#define XCHAL_HAVE_CACHE_BLOCKOPS	0	/* block prefetch for caches */
+#define XCHAL_HAVE_ICACHE_TEST		1	/* Icache test instructions */
+#define XCHAL_HAVE_DCACHE_TEST		1	/* Dcache test instructions */
+#define XCHAL_HAVE_ICACHE_DYN_WAYS	0	/* Icache dynamic way support */
+#define XCHAL_HAVE_DCACHE_DYN_WAYS	0	/* Dcache dynamic way support */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		7
+#define XCHAL_DCACHE_SETWIDTH		7
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		4
+#define XCHAL_DCACHE_WAYS		4
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	8
+#define XCHAL_DCACHE_ACCESS_SIZE	8
+
+#define XCHAL_DCACHE_BANKS		1	/* number of banks */
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+/*  Whether MEMCTL register has anything useful  */
+#define XCHAL_USE_MEMCTL		(((XCHAL_LOOP_BUFFER_SIZE > 0)	||	\
+					   XCHAL_DCACHE_IS_COHERENT	||	\
+					   XCHAL_HAVE_ICACHE_DYN_WAYS	||	\
+					   XCHAL_HAVE_DCACHE_DYN_WAYS)	&&	\
+					   (XCHAL_HW_MIN_VERSION >= XTENSA_HWVERSION_RE_2012_0))
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+#define XCHAL_HAVE_IMEM_LOADSTORE	1	/* can load/store to IROM/IRAM*/
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		16	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		4	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F00BF
+#define XCHAL_INTLEVEL2_MASK		0x00000140
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00008000
+#define XCHAL_INTLEVEL5_MASK		0x00003000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F00BF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F01FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F0FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F8FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		2
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		5
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		4
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_PROFILING
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F0000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000133F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+#define XCHAL_INTTYPE_MASK_PROFILING	0x00008000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+#define XCHAL_PROFILING_INTERRUPT	15	/* profiling interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL4_NUM		15
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 2, 3, 5.)  */
+
+
+/*
+ *  External interrupt mapping.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL BInterrupt pin number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 5) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		21	/* (intlevel 3) */
+/*  EXTERNAL BInterrupt pin numbers mapped to each core interrupt number:  */
+#define XCHAL_INT0_EXTNUM		0	/* (intlevel 1) */
+#define XCHAL_INT1_EXTNUM		1	/* (intlevel 1) */
+#define XCHAL_INT2_EXTNUM		2	/* (intlevel 1) */
+#define XCHAL_INT3_EXTNUM		3	/* (intlevel 1) */
+#define XCHAL_INT4_EXTNUM		4	/* (intlevel 1) */
+#define XCHAL_INT5_EXTNUM		5	/* (intlevel 1) */
+#define XCHAL_INT8_EXTNUM		6	/* (intlevel 2) */
+#define XCHAL_INT9_EXTNUM		7	/* (intlevel 3) */
+#define XCHAL_INT12_EXTNUM		8	/* (intlevel 5) */
+#define XCHAL_INT14_EXTNUM		9	/* (intlevel 7) */
+#define XCHAL_INT16_EXTNUM		10	/* (intlevel 1) */
+#define XCHAL_INT17_EXTNUM		11	/* (intlevel 1) */
+#define XCHAL_INT18_EXTNUM		12	/* (intlevel 1) */
+#define XCHAL_INT19_EXTNUM		13	/* (intlevel 1) */
+#define XCHAL_INT20_EXTNUM		14	/* (intlevel 1) */
+#define XCHAL_INT21_EXTNUM		15	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) or TX */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_HALT			0	/* halt architecture option */
+#define XCHAL_HAVE_BOOTLOADER		0	/* boot loader (for TX) */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0x00002000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00002000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0x00001000
+#define XCHAL_RESET_VECTOR1_PADDR	0x00001000
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0x00002340
+#define XCHAL_USER_VECTOR_PADDR		0x00002340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0x00002300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00002300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0x000023C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000023C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0x00002000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00002000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0x00002180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00002180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0x000021C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x000021C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0x00002200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x00002200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0x00002240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x00002240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0x00002280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x00002280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0x000022C0
+#define XCHAL_NMI_VECTOR_PADDR		0x000022C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG MODULE
+  ----------------------------------------------------------------------*/
+
+/*  Misc  */
+#define XCHAL_HAVE_DEBUG_ERI		1	/* ERI to debug module */
+#define XCHAL_HAVE_DEBUG_APB		0	/* APB to debug module */
+#define XCHAL_HAVE_DEBUG_JTAG		1	/* JTAG to debug module */
+
+/*  On-Chip Debug (OCD)  */
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	0	/* faster OCD option (to LX4) */
+#define XCHAL_HAVE_OCD_LS32DDR		1	/* L32DDR/S32DDR (faster OCD) */
+
+/*  TRAX (in core)  */
+#define XCHAL_HAVE_TRAX			1	/* TRAX in debug module */
+#define XCHAL_TRAX_MEM_SIZE		262144	/* TRAX memory size in bytes */
+#define XCHAL_TRAX_MEM_SHAREABLE	1	/* start/end regs; ready sig. */
+#define XCHAL_TRAX_ATB_WIDTH		0	/* ATB width (bits), 0=no ATB */
+#define XCHAL_TRAX_TIME_WIDTH		0	/* timestamp bitwidth, 0=none */
+
+/*  Perf counters  */
+#define XCHAL_NUM_PERF_COUNTERS		8	/* performance counters */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		1	/* one way maps I+D 4GB vaddr */
+#define XCHAL_SPANNING_WAY		6	/* TLB spanning way number */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
+
diff --git a/arch/xtensa/variants/test_kc705_be/include/variant/tie-asm.h b/arch/xtensa/variants/test_kc705_be/include/variant/tie-asm.h
new file mode 100644
index 000000000..b87fe1a51
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_be/include/variant/tie-asm.h
@@ -0,0 +1,308 @@
+/* 
+ * tie-asm.h -- compile-time HAL assembler definitions dependent on CORE & TIE
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file contains assembly-language definitions (assembly
+   macros, etc.) for this specific Xtensa processor's TIE extensions
+   and options.  It is customized to this Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+#define XTHAL_SAS_ANYOT	0x0003	/* both of the above */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+#define XTHAL_SAS_ANYCC	0x000C	/* both of the above */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+#define XTHAL_SAS_ANYABI	0x0070	/* all of the above three */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+#define XTHAL_SAS3(optie,ccuse,abi)	( ((optie) & XTHAL_SAS_ANYOT)  \
+					| ((ccuse) & XTHAL_SAS_ANYCC)  \
+					| ((abi)   & XTHAL_SAS_ANYABI) )
+
+
+    /*
+      *  Macro to store all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger store sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to store.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to store, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any store.
+      */
+    .macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	rur.THREADPTR	\at1		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	rsr.ACCLO	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.ACCHI	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	rsr.BR	\at1		// boolean option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.SCOMPARE1	\at1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rsr.M0	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rsr.M1	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rsr.M2	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	rsr.M3	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+20
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_store
+
+    /*
+      *  Macro to load all non-coprocessor (extra) custom TIE and optional state
+      *  (not including zero-overhead loop registers).
+      *  Required parameters:
+      *      ptr         Save area pointer address register (clobbered)
+      *                  (register must contain a 4 byte aligned address).
+      *      at1..at4    Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+      *                  registers are clobbered, the remaining are unused).
+      *  Optional parameters:
+      *      continue    If macro invoked as part of a larger load sequence, set to 1
+      *                  if this is not the first in the sequence.  Defaults to 0.
+      *      ofs         Offset from start of larger sequence (from value of first ptr
+      *                  in sequence) at which to load.  Defaults to next available space
+      *                  (or 0 if <continue> is 0).
+      *      select      Select what category(ies) of registers to load, as a bitmask
+      *                  (see XTHAL_SAS_xxx constants).  Defaults to all registers.
+      *      alloc       Select what category(ies) of registers to allocate; if any
+      *                  category is selected here that is not in <select>, space for
+      *                  the corresponding registers is skipped without doing any load.
+      */
+    .macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.THREADPTR	\at1		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.ACCLO	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.ACCHI	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.BR	\at1		// boolean option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.SCOMPARE1	\at1		// conditional store option
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wsr.M0	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wsr.M1	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wsr.M2	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+20
+	wsr.M3	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_load
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+    /* 
+     *  Macro to store the state of TIE coprocessor AudioEngineLX.
+     *  Required parameters:
+     *      ptr         Save area pointer address register (clobbered)
+     *                  (register must contain a 8 byte aligned address).
+     *      at1..at4    Four temporary address registers (first XCHAL_CP1_NUM_ATMPS
+     *                  registers are clobbered, the remaining are unused).
+     *  Optional parameters are the same as for xchal_ncp_store.
+     */
+#define xchal_cp_AudioEngineLX_store	xchal_cp1_store
+    .macro	xchal_cp1_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start \continue, \ofs
+	// Custom caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	rur.AE_OVF_SAR	\at1		// ureg 240
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rur.AE_BITHEAD	\at1		// ureg 241
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rur.AE_TS_FTS_BU_BP	\at1		// ureg 242
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rur.AE_SD_NO	\at1		// ureg 243
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rur.AE_CBEGIN0	\at1		// ureg 246
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	rur.AE_CEND0	\at1		// ureg 247
+	s32i	\at1, \ptr, .Lxchal_ofs_+20
+	ae_sp24x2s.i	aep0, \ptr, .Lxchal_ofs_+24
+	ae_sp24x2s.i	aep1, \ptr, .Lxchal_ofs_+32
+	ae_sp24x2s.i	aep2, \ptr, .Lxchal_ofs_+40
+	ae_sp24x2s.i	aep3, \ptr, .Lxchal_ofs_+48
+	ae_sp24x2s.i	aep4, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	ae_sp24x2s.i	aep5, \ptr, .Lxchal_ofs_+0
+	ae_sp24x2s.i	aep6, \ptr, .Lxchal_ofs_+8
+	ae_sp24x2s.i	aep7, \ptr, .Lxchal_ofs_+16
+	ae_sq56s.i	aeq0, \ptr, .Lxchal_ofs_+24
+	ae_sq56s.i	aeq1, \ptr, .Lxchal_ofs_+32
+	ae_sq56s.i	aeq2, \ptr, .Lxchal_ofs_+40
+	ae_sq56s.i	aeq3, \ptr, .Lxchal_ofs_+48
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 64
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 56
+	.elseif ((XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 120
+	.endif
+    .endm	// xchal_cp1_store
+
+    /* 
+     *  Macro to load the state of TIE coprocessor AudioEngineLX.
+     *  Required parameters:
+     *      ptr         Save area pointer address register (clobbered)
+     *                  (register must contain a 8 byte aligned address).
+     *      at1..at4    Four temporary address registers (first XCHAL_CP1_NUM_ATMPS
+     *                  registers are clobbered, the remaining are unused).
+     *  Optional parameters are the same as for xchal_ncp_load.
+     */
+#define xchal_cp_AudioEngineLX_load	xchal_cp1_load
+    .macro	xchal_cp1_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start \continue, \ofs
+	// Custom caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.AE_OVF_SAR	\at1		// ureg 240
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wur.AE_BITHEAD	\at1		// ureg 241
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wur.AE_TS_FTS_BU_BP	\at1		// ureg 242
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wur.AE_SD_NO	\at1		// ureg 243
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wur.AE_CBEGIN0	\at1		// ureg 246
+	l32i	\at1, \ptr, .Lxchal_ofs_+20
+	wur.AE_CEND0	\at1		// ureg 247
+	ae_lp24x2.i	aep0, \ptr, .Lxchal_ofs_+24
+	ae_lp24x2.i	aep1, \ptr, .Lxchal_ofs_+32
+	ae_lp24x2.i	aep2, \ptr, .Lxchal_ofs_+40
+	ae_lp24x2.i	aep3, \ptr, .Lxchal_ofs_+48
+	ae_lp24x2.i	aep4, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	ae_lp24x2.i	aep5, \ptr, .Lxchal_ofs_+0
+	ae_lp24x2.i	aep6, \ptr, .Lxchal_ofs_+8
+	ae_lp24x2.i	aep7, \ptr, .Lxchal_ofs_+16
+	addi	\ptr, \ptr, 24
+	ae_lq56.i	aeq0, \ptr, .Lxchal_ofs_+0
+	ae_lq56.i	aeq1, \ptr, .Lxchal_ofs_+8
+	ae_lq56.i	aeq2, \ptr, .Lxchal_ofs_+16
+	ae_lq56.i	aeq3, \ptr, .Lxchal_ofs_+24
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 88
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 32
+	.elseif ((XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 120
+	.endif
+    .endm	// xchal_cp1_load
+
+#define XCHAL_CP1_NUM_ATMPS	1
+#define XCHAL_SA_NUM_ATMPS	1
+
+	/*  Empty macros for unconfigured coprocessors:  */
+	.macro xchal_cp0_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp0_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
+
diff --git a/arch/xtensa/variants/test_kc705_be/include/variant/tie.h b/arch/xtensa/variants/test_kc705_be/include/variant/tie.h
new file mode 100644
index 000000000..4e1b4baac
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_be/include/variant/tie.h
@@ -0,0 +1,182 @@
+/* 
+ * tie.h -- compile-time HAL definitions dependent on CORE & TIE configuration
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file describes this specific Xtensa processor's TIE extensions
+   that extend basic Xtensa core functionality.  It is customized to this
+   Xtensa processor configuration.
+
+   Copyright (c) 1999-2015 Cadence Design Systems Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			2	/* number of coprocessors */
+#define XCHAL_CP_MAX			8	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x82	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x80	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP1_NAME			"AudioEngineLX"
+#define XCHAL_CP1_IDENT			AudioEngineLX
+#define XCHAL_CP1_SA_SIZE		120	/* size of state save area */
+#define XCHAL_CP1_SA_ALIGN		8	/* min alignment of save area */
+#define XCHAL_CP_ID_AUDIOENGINELX   	1	/* coprocessor ID (0..7) */
+#define XCHAL_CP7_NAME			"XTIOP"
+#define XCHAL_CP7_IDENT			XTIOP
+#define XCHAL_CP7_SA_SIZE		0	/* size of state save area */
+#define XCHAL_CP7_SA_ALIGN		1	/* min alignment of save area */
+#define XCHAL_CP_ID_XTIOP           	7	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		36
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		160	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		8	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	9
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             br, 4, 4, 4,0x0204,  sr,4  , 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	18
+#define XCHAL_CP1_SA_LIST(s)	\
+ XCHAL_SA_REG(s,0,0,1,0,     ae_ovf_sar, 8, 4, 4,0x03F0,  ur,240,  7,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,     ae_bithead, 4, 4, 4,0x03F1,  ur,241, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,ae_ts_fts_bu_bp, 4, 4, 4,0x03F2,  ur,242, 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,       ae_sd_no, 4, 4, 4,0x03F3,  ur,243, 28,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,     ae_cbegin0, 4, 4, 4,0x03F6,  ur,246, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,       ae_cend0, 4, 4, 4,0x03F7,  ur,247, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep0, 8, 8, 8,0x0060, aep,0  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep1, 8, 8, 8,0x0061, aep,1  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep2, 8, 8, 8,0x0062, aep,2  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep3, 8, 8, 8,0x0063, aep,3  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep4, 8, 8, 8,0x0064, aep,4  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep5, 8, 8, 8,0x0065, aep,5  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep6, 8, 8, 8,0x0066, aep,6  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep7, 8, 8, 8,0x0067, aep,7  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq0, 8, 8, 8,0x0068, aeq,0  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq1, 8, 8, 8,0x0069, aeq,1  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq2, 8, 8, 8,0x006A, aeq,2  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq3, 8, 8, 8,0x006B, aeq,3  , 56,0,0,0)
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,8
+/* Byte length of instruction from its first byte, per FLIX.  */
+#define XCHAL_BYTE0_FORMAT_LENGTHS	\
+	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,\
+	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,\
+	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,\
+	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,\
+	2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,\
+	2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,\
+	2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,\
+	3,3,3,3,3,3,3,3,3,3,3,3,3,3,3,3, 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8
+
+#endif /*_XTENSA_CORE_TIE_H*/
+
diff --git a/arch/xtensa/variants/test_kc705_hifi/include/variant/core.h b/arch/xtensa/variants/test_kc705_hifi/include/variant/core.h
new file mode 100644
index 000000000..4a2222979
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_hifi/include/variant/core.h
@@ -0,0 +1,531 @@
+/* 
+ * xtensa/config/core-isa.h -- HAL definitions that are dependent on Xtensa
+ *				processor CORE configuration
+ *
+ *  See <xtensa/config/core.h>, which includes this file, for more details.
+ */
+
+/* Xtensa processor core configuration information.
+
+   Copyright (c) 1999-2014 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	8	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_LOOP_BUFFER_SIZE		0	/* zero-ov. loop instr buffer size */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		1	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		1	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	0	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MX			0	/* MX core (Tensilica internal) */
+#define XCHAL_HAVE_MP_INTERRUPTS	0	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		0	/* core RunStall control port */
+#define XCHAL_HAVE_PSO			0	/* Power Shut-Off */
+#define XCHAL_HAVE_PSO_CDM		0	/* core/debug/mem pwr domains */
+#define XCHAL_HAVE_PSO_FULL_RETENTION	0	/* all regs preserved on PSO */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		1	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			8	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		1	/* MAC16 package */
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector floating-point pkg */
+#define XCHAL_HAVE_FP			0	/* single prec floating point */
+#define XCHAL_HAVE_FP_DIV		0	/* FP with DIV instructions */
+#define XCHAL_HAVE_FP_RECIP		0	/* FP with RECIP instructions */
+#define XCHAL_HAVE_FP_SQRT		0	/* FP with SQRT instructions */
+#define XCHAL_HAVE_FP_RSQRT		0	/* FP with RSQRT instructions */
+#define XCHAL_HAVE_DFP			0	/* double precision FP pkg */
+#define XCHAL_HAVE_DFP_DIV		0	/* DFP with DIV instructions */
+#define XCHAL_HAVE_DFP_RECIP		0	/* DFP with RECIP instructions*/
+#define XCHAL_HAVE_DFP_SQRT		0	/* DFP with SQRT instructions */
+#define XCHAL_HAVE_DFP_RSQRT		0	/* DFP with RSQRT instructions*/
+#define XCHAL_HAVE_DFP_accel		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI3		1	/* HiFi3 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2		0	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2EP		0	/* HiFi2EP */
+#define XCHAL_HAVE_HIFI_MINI		0	
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+#define XCHAL_HAVE_BBE16		0	/* ConnX BBE16 pkg */
+#define XCHAL_HAVE_BBE16_RSQRT		0	/* BBE16 & vector recip sqrt */
+#define XCHAL_HAVE_BBE16_VECDIV		0	/* BBE16 & vector divide */
+#define XCHAL_HAVE_BBE16_DESPREAD	0	/* BBE16 & despread */
+#define XCHAL_HAVE_BBENEP		0	/* ConnX BBENEP pkgs */
+#define XCHAL_HAVE_BSP3			0	/* ConnX BSP3 pkg */
+#define XCHAL_HAVE_BSP3_TRANSPOSE	0	/* BSP3 & transpose32x32 */
+#define XCHAL_HAVE_SSP16		0	/* ConnX SSP16 pkg */
+#define XCHAL_HAVE_SSP16_VITERBI	0	/* SSP16 & viterbi */
+#define XCHAL_HAVE_TURBO16		0	/* ConnX Turbo16 pkg */
+#define XCHAL_HAVE_BBP16		0	/* ConnX BBP16 pkg */
+#define XCHAL_HAVE_FLIX3		0	/* basic 3-way FLIX option */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_LOADSTORE_UNITS	1	/* load/store units */
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		8	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		8	/* data width in bytes */
+#define XCHAL_DATA_PIPE_DELAY		1	/* d-side pipeline delay
+						   (1 = 5-stage, 2 = 7-stage) */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		1000004	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"test_kc705_hifi"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_BUILD_UNIQUE_ID		0x0004983D	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC1B3FFFE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x1904983D	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX5.0.4"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2500	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		4	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		250004	/* major*100+minor */
+#define XCHAL_HW_REL_LX5		1
+#define XCHAL_HW_REL_LX5_0		1
+#define XCHAL_HW_REL_LX5_0_4		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2500	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	4	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		250004	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2500	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	4	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		250004	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		16384	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	0	/* MP coherence feature */
+
+#define XCHAL_HAVE_PREFETCH		1	/* PREFCTL register */
+#define XCHAL_HAVE_PREFETCH_L1		0	/* prefetch to L1 dcache */
+#define XCHAL_PREFETCH_CASTOUT_LINES	1	/* dcache pref. castout bufsz */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		7
+#define XCHAL_DCACHE_SETWIDTH		7
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		4
+#define XCHAL_DCACHE_WAYS		4
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	1
+#define XCHAL_DCACHE_LINE_LOCKABLE	1
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	8
+#define XCHAL_DCACHE_ACCESS_SIZE	8
+
+#define XCHAL_DCACHE_BANKS		1	/* number of banks */
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+#define XCHAL_HAVE_IMEM_LOADSTORE	1	/* can load/store to IROM/IRAM*/
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			1	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		3	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		22	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	5	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		16	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		6	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		3	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x001F00BF
+#define XCHAL_INTLEVEL2_MASK		0x00000140
+#define XCHAL_INTLEVEL3_MASK		0x00200E00
+#define XCHAL_INTLEVEL4_MASK		0x00009000
+#define XCHAL_INTLEVEL5_MASK		0x00002000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00004000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x001F00BF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x001F01FF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x003F0FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x003F9FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x003FBFFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x003FFFFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		2
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		2
+#define XCHAL_INT9_LEVEL		3
+#define XCHAL_INT10_LEVEL		3
+#define XCHAL_INT11_LEVEL		3
+#define XCHAL_INT12_LEVEL		4
+#define XCHAL_INT13_LEVEL		5
+#define XCHAL_INT14_LEVEL		7
+#define XCHAL_INT15_LEVEL		4
+#define XCHAL_INT16_LEVEL		1
+#define XCHAL_INT17_LEVEL		1
+#define XCHAL_INT18_LEVEL		1
+#define XCHAL_INT19_LEVEL		1
+#define XCHAL_INT20_LEVEL		1
+#define XCHAL_INT21_LEVEL		3
+#define XCHAL_DEBUGLEVEL		6	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+#define XCHAL_NMILEVEL			7	/* NMI "level" (for use with
+						   EXCSAVE/EPS/EPC_n, RFI n) */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT12_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT13_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT14_TYPE 	XTHAL_INTTYPE_NMI
+#define XCHAL_INT15_TYPE 	XTHAL_INTTYPE_PROFILING
+#define XCHAL_INT16_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT17_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT18_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT19_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT20_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT21_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFC00000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000880
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x003F0000
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x0000133F
+#define XCHAL_INTTYPE_MASK_TIMER	0x00002440
+#define XCHAL_INTTYPE_MASK_NMI		0x00004000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+#define XCHAL_INTTYPE_MASK_PROFILING	0x00008000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		10	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		13	/* CCOMPARE2 */
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_NMI_INTERRUPT		14	/* non-maskable interrupt */
+#define XCHAL_PROFILING_INTERRUPT	15	/* profiling interrupt */
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+#define XCHAL_INTLEVEL5_NUM		13
+#define XCHAL_INTLEVEL7_NUM		14
+/*  (There are many interrupts each at level(s) 1, 2, 3, 4.)  */
+
+
+/*
+ *  External interrupt mapping.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL BInterrupt pin number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		8	/* (intlevel 2) */
+#define XCHAL_EXTINT7_NUM		9	/* (intlevel 3) */
+#define XCHAL_EXTINT8_NUM		12	/* (intlevel 4) */
+#define XCHAL_EXTINT9_NUM		14	/* (intlevel 7) */
+#define XCHAL_EXTINT10_NUM		16	/* (intlevel 1) */
+#define XCHAL_EXTINT11_NUM		17	/* (intlevel 1) */
+#define XCHAL_EXTINT12_NUM		18	/* (intlevel 1) */
+#define XCHAL_EXTINT13_NUM		19	/* (intlevel 1) */
+#define XCHAL_EXTINT14_NUM		20	/* (intlevel 1) */
+#define XCHAL_EXTINT15_NUM		21	/* (intlevel 3) */
+/*  EXTERNAL BInterrupt pin numbers mapped to each core interrupt number:  */
+#define XCHAL_INT0_EXTNUM		0	/* (intlevel 1) */
+#define XCHAL_INT1_EXTNUM		1	/* (intlevel 1) */
+#define XCHAL_INT2_EXTNUM		2	/* (intlevel 1) */
+#define XCHAL_INT3_EXTNUM		3	/* (intlevel 1) */
+#define XCHAL_INT4_EXTNUM		4	/* (intlevel 1) */
+#define XCHAL_INT5_EXTNUM		5	/* (intlevel 1) */
+#define XCHAL_INT8_EXTNUM		6	/* (intlevel 2) */
+#define XCHAL_INT9_EXTNUM		7	/* (intlevel 3) */
+#define XCHAL_INT12_EXTNUM		8	/* (intlevel 4) */
+#define XCHAL_INT14_EXTNUM		9	/* (intlevel 7) */
+#define XCHAL_INT16_EXTNUM		10	/* (intlevel 1) */
+#define XCHAL_INT17_EXTNUM		11	/* (intlevel 1) */
+#define XCHAL_INT18_EXTNUM		12	/* (intlevel 1) */
+#define XCHAL_INT19_EXTNUM		13	/* (intlevel 1) */
+#define XCHAL_INT20_EXTNUM		14	/* (intlevel 1) */
+#define XCHAL_INT21_EXTNUM		15	/* (intlevel 3) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) or TX */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_HALT			0	/* halt architecture option */
+#define XCHAL_HAVE_BOOTLOADER		0	/* boot loader (for TX) */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0x00002000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00002000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0x00001000
+#define XCHAL_RESET_VECTOR1_PADDR	0x00001000
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0x00002340
+#define XCHAL_USER_VECTOR_PADDR		0x00002340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0x00002300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00002300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0x000023C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000023C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0x00002000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00002000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000180
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0x00002180
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00002180
+#define XCHAL_INTLEVEL3_VECOFS		0x000001C0
+#define XCHAL_INTLEVEL3_VECTOR_VADDR	0x000021C0
+#define XCHAL_INTLEVEL3_VECTOR_PADDR	0x000021C0
+#define XCHAL_INTLEVEL4_VECOFS		0x00000200
+#define XCHAL_INTLEVEL4_VECTOR_VADDR	0x00002200
+#define XCHAL_INTLEVEL4_VECTOR_PADDR	0x00002200
+#define XCHAL_INTLEVEL5_VECOFS		0x00000240
+#define XCHAL_INTLEVEL5_VECTOR_VADDR	0x00002240
+#define XCHAL_INTLEVEL5_VECTOR_PADDR	0x00002240
+#define XCHAL_INTLEVEL6_VECOFS		0x00000280
+#define XCHAL_INTLEVEL6_VECTOR_VADDR	0x00002280
+#define XCHAL_INTLEVEL6_VECTOR_PADDR	0x00002280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL6_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL6_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL6_VECTOR_PADDR
+#define XCHAL_NMI_VECOFS		0x000002C0
+#define XCHAL_NMI_VECTOR_VADDR		0x000022C0
+#define XCHAL_NMI_VECTOR_PADDR		0x000022C0
+#define XCHAL_INTLEVEL7_VECOFS		XCHAL_NMI_VECOFS
+#define XCHAL_INTLEVEL7_VECTOR_VADDR	XCHAL_NMI_VECTOR_VADDR
+#define XCHAL_INTLEVEL7_VECTOR_PADDR	XCHAL_NMI_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG MODULE
+  ----------------------------------------------------------------------*/
+
+/*  Misc  */
+#define XCHAL_HAVE_DEBUG_ERI		1	/* ERI to debug module */
+#define XCHAL_HAVE_DEBUG_APB		0	/* APB to debug module */
+#define XCHAL_HAVE_DEBUG_JTAG		1	/* JTAG to debug module */
+
+/*  On-Chip Debug (OCD)  */
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		2	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		2	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	0	/* faster OCD option (to LX4) */
+#define XCHAL_HAVE_OCD_LS32DDR		1	/* L32DDR/S32DDR (faster OCD) */
+
+/*  TRAX (in core)  */
+#define XCHAL_HAVE_TRAX			1	/* TRAX in debug module */
+#define XCHAL_TRAX_MEM_SIZE		262144	/* TRAX memory size in bytes */
+#define XCHAL_TRAX_MEM_SHAREABLE	1	/* start/end regs; ready sig. */
+#define XCHAL_TRAX_ATB_WIDTH		0	/* ATB width (bits), 0=no ATB */
+#define XCHAL_TRAX_TIME_WIDTH		0	/* timestamp bitwidth, 0=none */
+
+/*  Perf counters  */
+#define XCHAL_NUM_PERF_COUNTERS		8	/* performance counters */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		1	/* one way maps I+D 4GB vaddr */
+#define XCHAL_SPANNING_WAY		6	/* TLB spanning way number */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
diff --git a/arch/xtensa/variants/test_kc705_hifi/include/variant/tie-asm.h b/arch/xtensa/variants/test_kc705_hifi/include/variant/tie-asm.h
new file mode 100644
index 000000000..378163c0a
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_hifi/include/variant/tie-asm.h
@@ -0,0 +1,328 @@
+/* 
+ * tie-asm.h -- compile-time HAL assembler definitions dependent on CORE & TIE
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file contains assembly-language definitions (assembly
+   macros, etc.) for this specific Xtensa processor's TIE extensions
+   and options.  It is customized to this Xtensa processor configuration.
+
+   Copyright (c) 1999-2014 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+#define XTHAL_SAS_ANYOT	0x0003	/* both of the above */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+#define XTHAL_SAS_ANYCC	0x000C	/* both of the above */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+#define XTHAL_SAS_ANYABI 0x0070	/* all of the above three */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+#define XTHAL_SAS3(optie,ccuse,abi)	( ((optie) & XTHAL_SAS_ANYOT)  \
+					| ((ccuse) & XTHAL_SAS_ANYCC)  \
+					| ((abi)   & XTHAL_SAS_ANYABI) )
+
+
+
+    /*
+     *  Macro to save all non-coprocessor (extra) custom TIE and optional state
+     *  (not including zero-overhead loop registers).
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 4 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters:
+     *      continue	If macro invoked as part of a larger store sequence, set to 1
+     *			if this is not the first in the sequence.  Defaults to 0.
+     *      ofs		Offset from start of larger sequence (from value of first ptr
+     *			in sequence) at which to store.  Defaults to next available space
+     *			(or 0 if <continue> is 0).
+     *      select	Select what category(ies) of registers to store, as a bitmask
+     *			(see XTHAL_SAS_xxx constants).  Defaults to all registers.
+     *      alloc	Select what category(ies) of registers to allocate; if any
+     *			category is selected here that is not in <select>, space for
+     *			the corresponding registers is skipped without doing any store.
+     */
+    .macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global register used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	rur.THREADPTR	\at1		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	rsr.ACCLO	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.ACCHI	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	rsr.M0	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rsr.M1	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rsr.M2	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rsr.M3	\at1		// MAC16 option
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rsr.BR	\at1		// boolean option
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	rsr.SCOMPARE1	\at1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_+20
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_store
+
+    /*
+     *  Macro to restore all non-coprocessor (extra) custom TIE and optional state
+     *  (not including zero-overhead loop registers).
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 4 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_NCP_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters:
+     *      continue	If macro invoked as part of a larger load sequence, set to 1
+     *			if this is not the first in the sequence.  Defaults to 0.
+     *      ofs		Offset from start of larger sequence (from value of first ptr
+     *			in sequence) at which to load.  Defaults to next available space
+     *			(or 0 if <continue> is 0).
+     *      select	Select what category(ies) of registers to load, as a bitmask
+     *			(see XTHAL_SAS_xxx constants).  Defaults to all registers.
+     *      alloc	Select what category(ies) of registers to allocate; if any
+     *			category is selected here that is not in <select>, space for
+     *			the corresponding registers is skipped without doing any load.
+     */
+    .macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start	\continue, \ofs
+	// Optional global register used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.THREADPTR	\at1		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1020, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	// Optional caller-saved registers used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.ACCLO	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.ACCHI	\at1		// MAC16 option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1016, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 8
+	.endif
+	// Optional caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wsr.M0	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wsr.M1	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wsr.M2	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wsr.M3	\at1		// MAC16 option
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wsr.BR	\at1		// boolean option
+	l32i	\at1, \ptr, .Lxchal_ofs_+20
+	wsr.SCOMPARE1	\at1		// conditional store option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.elseif ((XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 1000, 4, 4
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 24
+	.endif
+    .endm	// xchal_ncp_load
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+
+
+
+    /*
+     *  Macro to save the state of TIE coprocessor AudioEngineLX.
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 8 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_CP1_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters are the same as for xchal_ncp_store.
+     */
+#define xchal_cp_AudioEngineLX_store	xchal_cp1_store
+    .macro	xchal_cp1_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start \continue, \ofs
+	// Custom caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	rur.AE_OVF_SAR	\at1		// ureg 240
+	s32i	\at1, \ptr, .Lxchal_ofs_+0
+	rur.AE_BITHEAD	\at1		// ureg 241
+	s32i	\at1, \ptr, .Lxchal_ofs_+4
+	rur.AE_TS_FTS_BU_BP	\at1		// ureg 242
+	s32i	\at1, \ptr, .Lxchal_ofs_+8
+	rur.AE_CW_SD_NO	\at1		// ureg 243
+	s32i	\at1, \ptr, .Lxchal_ofs_+12
+	rur.AE_CBEGIN0	\at1		// ureg 246
+	s32i	\at1, \ptr, .Lxchal_ofs_+16
+	rur.AE_CEND0	\at1		// ureg 247
+	s32i	\at1, \ptr, .Lxchal_ofs_+20
+	AE_S64.I	aed0, \ptr, .Lxchal_ofs_+24
+	AE_S64.I	aed1, \ptr, .Lxchal_ofs_+32
+	AE_S64.I	aed2, \ptr, .Lxchal_ofs_+40
+	AE_S64.I	aed3, \ptr, .Lxchal_ofs_+48
+	AE_S64.I	aed4, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	AE_S64.I	aed5, \ptr, .Lxchal_ofs_+0
+	AE_S64.I	aed6, \ptr, .Lxchal_ofs_+8
+	AE_S64.I	aed7, \ptr, .Lxchal_ofs_+16
+	AE_S64.I	aed8, \ptr, .Lxchal_ofs_+24
+	AE_S64.I	aed9, \ptr, .Lxchal_ofs_+32
+	AE_S64.I	aed10, \ptr, .Lxchal_ofs_+40
+	AE_S64.I	aed11, \ptr, .Lxchal_ofs_+48
+	AE_S64.I	aed12, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	AE_S64.I	aed13, \ptr, .Lxchal_ofs_+0
+	AE_S64.I	aed14, \ptr, .Lxchal_ofs_+8
+	AE_S64.I	aed15, \ptr, .Lxchal_ofs_+16
+	AE_SALIGN64.I	u0, \ptr, .Lxchal_ofs_+24
+	AE_SALIGN64.I	u1, \ptr, .Lxchal_ofs_+32
+	AE_SALIGN64.I	u2, \ptr, .Lxchal_ofs_+40
+	AE_SALIGN64.I	u3, \ptr, .Lxchal_ofs_+48
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 128
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 56
+	.elseif ((XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 184
+	.endif
+    .endm	// xchal_cp1_store
+
+    /*
+     *  Macro to restore the state of TIE coprocessor AudioEngineLX.
+     *  Required parameters:
+     *      ptr		Save area pointer address register (clobbered)
+     *			(register must contain a 8 byte aligned address).
+     *      at1..at4	Four temporary address registers (first XCHAL_CP1_NUM_ATMPS
+     *			registers are clobbered, the remaining are unused).
+     *  Optional parameters are the same as for xchal_ncp_load.
+     */
+#define xchal_cp_AudioEngineLX_load	xchal_cp1_load
+    .macro	xchal_cp1_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL alloc=0
+	xchal_sa_start \continue, \ofs
+	// Custom caller-saved registers not used by default by the compiler:
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\select)
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	l32i	\at1, \ptr, .Lxchal_ofs_+0
+	wur.AE_OVF_SAR	\at1		// ureg 240
+	l32i	\at1, \ptr, .Lxchal_ofs_+4
+	wur.AE_BITHEAD	\at1		// ureg 241
+	l32i	\at1, \ptr, .Lxchal_ofs_+8
+	wur.AE_TS_FTS_BU_BP	\at1		// ureg 242
+	l32i	\at1, \ptr, .Lxchal_ofs_+12
+	wur.AE_CW_SD_NO	\at1		// ureg 243
+	l32i	\at1, \ptr, .Lxchal_ofs_+16
+	wur.AE_CBEGIN0	\at1		// ureg 246
+	l32i	\at1, \ptr, .Lxchal_ofs_+20
+	wur.AE_CEND0	\at1		// ureg 247
+	AE_L64.I	aed0, \ptr, .Lxchal_ofs_+24
+	AE_L64.I	aed1, \ptr, .Lxchal_ofs_+32
+	AE_L64.I	aed2, \ptr, .Lxchal_ofs_+40
+	AE_L64.I	aed3, \ptr, .Lxchal_ofs_+48
+	AE_L64.I	aed4, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	AE_L64.I	aed5, \ptr, .Lxchal_ofs_+0
+	AE_L64.I	aed6, \ptr, .Lxchal_ofs_+8
+	AE_L64.I	aed7, \ptr, .Lxchal_ofs_+16
+	AE_L64.I	aed8, \ptr, .Lxchal_ofs_+24
+	AE_L64.I	aed9, \ptr, .Lxchal_ofs_+32
+	AE_L64.I	aed10, \ptr, .Lxchal_ofs_+40
+	AE_L64.I	aed11, \ptr, .Lxchal_ofs_+48
+	AE_L64.I	aed12, \ptr, .Lxchal_ofs_+56
+	addi	\ptr, \ptr, 64
+	AE_L64.I	aed13, \ptr, .Lxchal_ofs_+0
+	AE_L64.I	aed14, \ptr, .Lxchal_ofs_+8
+	AE_L64.I	aed15, \ptr, .Lxchal_ofs_+16
+	AE_LALIGN64.I	u0, \ptr, .Lxchal_ofs_+24
+	AE_LALIGN64.I	u1, \ptr, .Lxchal_ofs_+32
+	AE_LALIGN64.I	u2, \ptr, .Lxchal_ofs_+40
+	AE_LALIGN64.I	u3, \ptr, .Lxchal_ofs_+48
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 128
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 56
+	.elseif ((XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~(\alloc)) == 0
+	xchal_sa_align	\ptr, 0, 0, 8, 8
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 184
+	.endif
+    .endm	// xchal_cp1_load
+
+#define XCHAL_CP1_NUM_ATMPS	1
+#define XCHAL_SA_NUM_ATMPS	1
+
+	/*  Empty macros for unconfigured coprocessors:  */
+	.macro xchal_cp0_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp0_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
diff --git a/arch/xtensa/variants/test_kc705_hifi/include/variant/tie.h b/arch/xtensa/variants/test_kc705_hifi/include/variant/tie.h
new file mode 100644
index 000000000..e39fea643
--- /dev/null
+++ b/arch/xtensa/variants/test_kc705_hifi/include/variant/tie.h
@@ -0,0 +1,189 @@
+/* 
+ * tie.h -- compile-time HAL definitions dependent on CORE & TIE configuration
+ *
+ *  NOTE:  This header file is not meant to be included directly.
+ */
+
+/* This header file describes this specific Xtensa processor's TIE extensions
+   that extend basic Xtensa core functionality.  It is customized to this
+   Xtensa processor configuration.
+
+   Copyright (c) 1999-2014 Tensilica Inc.
+
+   Permission is hereby granted, free of charge, to any person obtaining
+   a copy of this software and associated documentation files (the
+   "Software"), to deal in the Software without restriction, including
+   without limitation the rights to use, copy, modify, merge, publish,
+   distribute, sublicense, and/or sell copies of the Software, and to
+   permit persons to whom the Software is furnished to do so, subject to
+   the following conditions:
+
+   The above copyright notice and this permission notice shall be included
+   in all copies or substantial portions of the Software.
+
+   THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+   EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+   MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+   IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+   CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+   TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+   SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.  */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			2	/* number of coprocessors */
+#define XCHAL_CP_MAX			8	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x82	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x80	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP1_NAME			"AudioEngineLX"
+#define XCHAL_CP1_IDENT			AudioEngineLX
+#define XCHAL_CP1_SA_SIZE		184	/* size of state save area */
+#define XCHAL_CP1_SA_ALIGN		8	/* min alignment of save area */
+#define XCHAL_CP_ID_AUDIOENGINELX	1	/* coprocessor ID (0..7) */
+#define XCHAL_CP7_NAME			"XTIOP"
+#define XCHAL_CP7_IDENT			XTIOP
+#define XCHAL_CP7_SA_SIZE		0	/* size of state save area */
+#define XCHAL_CP7_SA_ALIGN		1	/* min alignment of save area */
+#define XCHAL_CP_ID_XTIOP		7	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		36
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		240	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		8	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	9
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acclo, 4, 4, 4,0x0210,  sr,16 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,0,0,1,          acchi, 4, 4, 4,0x0211,  sr,17 ,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m0, 4, 4, 4,0x0220,  sr,32 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m1, 4, 4, 4,0x0221,  sr,33 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m2, 4, 4, 4,0x0222,  sr,34 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             m3, 4, 4, 4,0x0223,  sr,35 , 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,             br, 4, 4, 4,0x0204,  sr,4  , 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	26
+#define XCHAL_CP1_SA_LIST(s)	\
+ XCHAL_SA_REG(s,0,0,1,0,     ae_ovf_sar, 8, 4, 4,0x03F0,  ur,240,  8,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,     ae_bithead, 4, 4, 4,0x03F1,  ur,241, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,ae_ts_fts_bu_bp, 4, 4, 4,0x03F2,  ur,242, 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,    ae_cw_sd_no, 4, 4, 4,0x03F3,  ur,243, 29,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,     ae_cbegin0, 4, 4, 4,0x03F6,  ur,246, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,       ae_cend0, 4, 4, 4,0x03F7,  ur,247, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed0, 8, 8, 8,0x1010, aed,0  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed1, 8, 8, 8,0x1011, aed,1  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed2, 8, 8, 8,0x1012, aed,2  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed3, 8, 8, 8,0x1013, aed,3  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed4, 8, 8, 8,0x1014, aed,4  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed5, 8, 8, 8,0x1015, aed,5  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed6, 8, 8, 8,0x1016, aed,6  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed7, 8, 8, 8,0x1017, aed,7  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed8, 8, 8, 8,0x1018, aed,8  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aed9, 8, 8, 8,0x1019, aed,9  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed10, 8, 8, 8,0x101A, aed,10 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed11, 8, 8, 8,0x101B, aed,11 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed12, 8, 8, 8,0x101C, aed,12 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed13, 8, 8, 8,0x101D, aed,13 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed14, 8, 8, 8,0x101E, aed,14 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,          aed15, 8, 8, 8,0x101F, aed,15 , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,             u0, 8, 8, 8,0x1020,   u,0  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,             u1, 8, 8, 8,0x1021,   u,1  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,             u2, 8, 8, 8,0x1022,   u,2  , 64,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,             u3, 8, 8, 8,0x1023,   u,3  , 64,0,0,0)
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8
+/* Byte length of instruction from its first byte, per FLIX.  */
+#define XCHAL_BYTE0_FORMAT_LENGTHS	\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8,\
+	3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8, 3,3,3,3,3,3,3,3,2,2,2,2,2,2,8,8
+
+#endif /*_XTENSA_CORE_TIE_H*/
diff --git a/arch/xtensa/variants/test_mmuhifi_c3/include/variant/core.h b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/core.h
new file mode 100644
index 000000000..7c5d3a22d
--- /dev/null
+++ b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/core.h
@@ -0,0 +1,383 @@
+/*
+ * Xtensa processor core configuration information.
+ *
+ * This file is subject to the terms and conditions of version 2.1 of the GNU
+ * Lesser General Public License as published by the Free Software Foundation.
+ *
+ * Copyright (c) 1999-2009 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_CONFIGURATION_H
+#define _XTENSA_CORE_CONFIGURATION_H
+
+
+/****************************************************************************
+	    Parameters Useful for Any Code, USER or PRIVILEGED
+ ****************************************************************************/
+
+/*
+ *  Note:  Macros of the form XCHAL_HAVE_*** have a value of 1 if the option is
+ *  configured, and a value of 0 otherwise.  These macros are always defined.
+ */
+
+
+/*----------------------------------------------------------------------
+				ISA
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_BE			0	/* big-endian byte ordering */
+#define XCHAL_HAVE_WINDOWED		1	/* windowed registers option */
+#define XCHAL_NUM_AREGS			32	/* num of physical addr regs */
+#define XCHAL_NUM_AREGS_LOG2		5	/* log2(XCHAL_NUM_AREGS) */
+#define XCHAL_MAX_INSTRUCTION_SIZE	8	/* max instr bytes (3..8) */
+#define XCHAL_HAVE_DEBUG		1	/* debug option */
+#define XCHAL_HAVE_DENSITY		1	/* 16-bit instructions */
+#define XCHAL_HAVE_LOOPS		1	/* zero-overhead loops */
+#define XCHAL_HAVE_NSA			1	/* NSA/NSAU instructions */
+#define XCHAL_HAVE_MINMAX		1	/* MIN/MAX instructions */
+#define XCHAL_HAVE_SEXT			1	/* SEXT instruction */
+#define XCHAL_HAVE_CLAMPS		1	/* CLAMPS instruction */
+#define XCHAL_HAVE_MUL16		1	/* MUL16S/MUL16U instructions */
+#define XCHAL_HAVE_MUL32		1	/* MULL instruction */
+#define XCHAL_HAVE_MUL32_HIGH		0	/* MULUH/MULSH instructions */
+#define XCHAL_HAVE_DIV32		0	/* QUOS/QUOU/REMS/REMU instructions */
+#define XCHAL_HAVE_L32R			1	/* L32R instruction */
+#define XCHAL_HAVE_ABSOLUTE_LITERALS	1	/* non-PC-rel (extended) L32R */
+#define XCHAL_HAVE_CONST16		0	/* CONST16 instruction */
+#define XCHAL_HAVE_ADDX			1	/* ADDX#/SUBX# instructions */
+#define XCHAL_HAVE_WIDE_BRANCHES	0	/* B*.W18 or B*.W15 instr's */
+#define XCHAL_HAVE_PREDICTED_BRANCHES	0	/* B[EQ/EQZ/NE/NEZ]T instr's */
+#define XCHAL_HAVE_CALL4AND12		1	/* (obsolete option) */
+#define XCHAL_HAVE_ABS			1	/* ABS instruction */
+/*#define XCHAL_HAVE_POPC		0*/	/* POPC instruction */
+/*#define XCHAL_HAVE_CRC		0*/	/* CRC instruction */
+#define XCHAL_HAVE_RELEASE_SYNC		1	/* L32AI/S32RI instructions */
+#define XCHAL_HAVE_S32C1I		1	/* S32C1I instruction */
+#define XCHAL_HAVE_SPECULATION		0	/* speculation */
+#define XCHAL_HAVE_FULL_RESET		1	/* all regs/state reset */
+#define XCHAL_NUM_CONTEXTS		1	/* */
+#define XCHAL_NUM_MISC_REGS		2	/* num of scratch regs (0..4) */
+#define XCHAL_HAVE_TAP_MASTER		0	/* JTAG TAP control instr's */
+#define XCHAL_HAVE_PRID			1	/* processor ID register */
+#define XCHAL_HAVE_EXTERN_REGS		1	/* WER/RER instructions */
+#define XCHAL_HAVE_MP_INTERRUPTS	1	/* interrupt distributor port */
+#define XCHAL_HAVE_MP_RUNSTALL		1	/* core RunStall control port */
+#define XCHAL_HAVE_THREADPTR		1	/* THREADPTR register */
+#define XCHAL_HAVE_BOOLEANS		1	/* boolean registers */
+#define XCHAL_HAVE_CP			1	/* CPENABLE reg (coprocessor) */
+#define XCHAL_CP_MAXCFG			2	/* max allowed cp id plus one */
+#define XCHAL_HAVE_MAC16		0	/* MAC16 package */
+#define XCHAL_HAVE_VECTORFPU2005	0	/* vector floating-point pkg */
+#define XCHAL_HAVE_FP			0	/* floating point pkg */
+#define XCHAL_HAVE_DFP			0	/* double precision FP pkg */
+#define XCHAL_HAVE_DFP_accel		0	/* double precision FP acceleration pkg */
+#define XCHAL_HAVE_VECTRA1		0	/* Vectra I  pkg */
+#define XCHAL_HAVE_VECTRALX		0	/* Vectra LX pkg */
+#define XCHAL_HAVE_HIFIPRO		0	/* HiFiPro Audio Engine pkg */
+#define XCHAL_HAVE_HIFI2		1	/* HiFi2 Audio Engine pkg */
+#define XCHAL_HAVE_CONNXD2		0	/* ConnX D2 pkg */
+
+
+/*----------------------------------------------------------------------
+				MISC
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_WRITEBUFFER_ENTRIES	8	/* size of write buffer */
+#define XCHAL_INST_FETCH_WIDTH		8	/* instr-fetch width in bytes */
+#define XCHAL_DATA_WIDTH		8	/* data width in bytes */
+/*  In T1050, applies to selected core load and store instructions (see ISA): */
+#define XCHAL_UNALIGNED_LOAD_EXCEPTION	1	/* unaligned loads cause exc. */
+#define XCHAL_UNALIGNED_STORE_EXCEPTION	1	/* unaligned stores cause exc.*/
+#define XCHAL_UNALIGNED_LOAD_HW		0	/* unaligned loads work in hw */
+#define XCHAL_UNALIGNED_STORE_HW	0	/* unaligned stores work in hw*/
+
+#define XCHAL_SW_VERSION		800000	/* sw version of this header */
+
+#define XCHAL_CORE_ID			"test_mmuhifi_c3"	/* alphanum core name
+						   (CoreID) set in the Xtensa
+						   Processor Generator */
+
+#define XCHAL_CORE_DESCRIPTION		"test_mmuhifi_c3"
+#define XCHAL_BUILD_UNIQUE_ID		0x00005A6A	/* 22-bit sw build ID */
+
+/*
+ *  These definitions describe the hardware targeted by this software.
+ */
+#define XCHAL_HW_CONFIGID0		0xC1B3CBFE	/* ConfigID hi 32 bits*/
+#define XCHAL_HW_CONFIGID1		0x10405A6A	/* ConfigID lo 32 bits*/
+#define XCHAL_HW_VERSION_NAME		"LX3.0.0"	/* full version name */
+#define XCHAL_HW_VERSION_MAJOR		2300	/* major ver# of targeted hw */
+#define XCHAL_HW_VERSION_MINOR		0	/* minor ver# of targeted hw */
+#define XCHAL_HW_VERSION		230000	/* major*100+minor */
+#define XCHAL_HW_REL_LX3		1
+#define XCHAL_HW_REL_LX3_0		1
+#define XCHAL_HW_REL_LX3_0_0		1
+#define XCHAL_HW_CONFIGID_RELIABLE	1
+/*  If software targets a *range* of hardware versions, these are the bounds: */
+#define XCHAL_HW_MIN_VERSION_MAJOR	2300	/* major v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION_MINOR	0	/* minor v of earliest tgt hw */
+#define XCHAL_HW_MIN_VERSION		230000	/* earliest targeted hw */
+#define XCHAL_HW_MAX_VERSION_MAJOR	2300	/* major v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION_MINOR	0	/* minor v of latest tgt hw */
+#define XCHAL_HW_MAX_VERSION		230000	/* latest targeted hw */
+
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_ICACHE_LINESIZE		32	/* I-cache line size in bytes */
+#define XCHAL_DCACHE_LINESIZE		32	/* D-cache line size in bytes */
+#define XCHAL_ICACHE_LINEWIDTH		5	/* log2(I line size in bytes) */
+#define XCHAL_DCACHE_LINEWIDTH		5	/* log2(D line size in bytes) */
+
+#define XCHAL_ICACHE_SIZE		16384	/* I-cache size in bytes or 0 */
+#define XCHAL_DCACHE_SIZE		16384	/* D-cache size in bytes or 0 */
+
+#define XCHAL_DCACHE_IS_WRITEBACK	1	/* writeback feature */
+#define XCHAL_DCACHE_IS_COHERENT	1	/* MP coherence feature */
+
+
+
+
+/****************************************************************************
+    Parameters Useful for PRIVILEGED (Supervisory or Non-Virtualized) Code
+ ****************************************************************************/
+
+
+#ifndef XTENSA_HAL_NON_PRIVILEGED_ONLY
+
+/*----------------------------------------------------------------------
+				CACHE
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_PIF			1	/* any outbound PIF present */
+
+/*  If present, cache size in bytes == (ways * 2^(linewidth + setwidth)).  */
+
+/*  Number of cache sets in log2(lines per way):  */
+#define XCHAL_ICACHE_SETWIDTH		8
+#define XCHAL_DCACHE_SETWIDTH		8
+
+/*  Cache set associativity (number of ways):  */
+#define XCHAL_ICACHE_WAYS		2
+#define XCHAL_DCACHE_WAYS		2
+
+/*  Cache features:  */
+#define XCHAL_ICACHE_LINE_LOCKABLE	0
+#define XCHAL_DCACHE_LINE_LOCKABLE	0
+#define XCHAL_ICACHE_ECC_PARITY		0
+#define XCHAL_DCACHE_ECC_PARITY		0
+
+/*  Cache access size in bytes (affects operation of SICW instruction):  */
+#define XCHAL_ICACHE_ACCESS_SIZE	8
+#define XCHAL_DCACHE_ACCESS_SIZE	8
+
+/*  Number of encoded cache attr bits (see <xtensa/hal.h> for decoded bits):  */
+#define XCHAL_CA_BITS			4
+
+
+/*----------------------------------------------------------------------
+			INTERNAL I/D RAM/ROMs and XLMI
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_NUM_INSTROM		0	/* number of core instr. ROMs */
+#define XCHAL_NUM_INSTRAM		0	/* number of core instr. RAMs */
+#define XCHAL_NUM_DATAROM		0	/* number of core data ROMs */
+#define XCHAL_NUM_DATARAM		0	/* number of core data RAMs */
+#define XCHAL_NUM_URAM			0	/* number of core unified RAMs*/
+#define XCHAL_NUM_XLMI			0	/* number of core XLMI ports */
+
+
+/*----------------------------------------------------------------------
+			INTERRUPTS and TIMERS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_INTERRUPTS		1	/* interrupt option */
+#define XCHAL_HAVE_HIGHPRI_INTERRUPTS	1	/* med/high-pri. interrupts */
+#define XCHAL_HAVE_NMI			0	/* non-maskable interrupt */
+#define XCHAL_HAVE_CCOUNT		1	/* CCOUNT reg. (timer option) */
+#define XCHAL_NUM_TIMERS		2	/* number of CCOMPAREn regs */
+#define XCHAL_NUM_INTERRUPTS		12	/* number of interrupts */
+#define XCHAL_NUM_INTERRUPTS_LOG2	4	/* ceil(log2(NUM_INTERRUPTS)) */
+#define XCHAL_NUM_EXTINTERRUPTS		9	/* num of external interrupts */
+#define XCHAL_NUM_INTLEVELS		2	/* number of interrupt levels
+						   (not including level zero) */
+#define XCHAL_EXCM_LEVEL		1	/* level masked by PS.EXCM */
+	/* (always 1 in XEA1; levels 2 .. EXCM_LEVEL are "medium priority") */
+
+/*  Masks of interrupts at each interrupt level:  */
+#define XCHAL_INTLEVEL1_MASK		0x00000FFF
+#define XCHAL_INTLEVEL2_MASK		0x00000000
+#define XCHAL_INTLEVEL3_MASK		0x00000000
+#define XCHAL_INTLEVEL4_MASK		0x00000000
+#define XCHAL_INTLEVEL5_MASK		0x00000000
+#define XCHAL_INTLEVEL6_MASK		0x00000000
+#define XCHAL_INTLEVEL7_MASK		0x00000000
+
+/*  Masks of interrupts at each range 1..n of interrupt levels:  */
+#define XCHAL_INTLEVEL1_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL2_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL3_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL4_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL5_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL6_ANDBELOW_MASK	0x00000FFF
+#define XCHAL_INTLEVEL7_ANDBELOW_MASK	0x00000FFF
+
+/*  Level of each interrupt:  */
+#define XCHAL_INT0_LEVEL		1
+#define XCHAL_INT1_LEVEL		1
+#define XCHAL_INT2_LEVEL		1
+#define XCHAL_INT3_LEVEL		1
+#define XCHAL_INT4_LEVEL		1
+#define XCHAL_INT5_LEVEL		1
+#define XCHAL_INT6_LEVEL		1
+#define XCHAL_INT7_LEVEL		1
+#define XCHAL_INT8_LEVEL		1
+#define XCHAL_INT9_LEVEL		1
+#define XCHAL_INT10_LEVEL		1
+#define XCHAL_INT11_LEVEL		1
+#define XCHAL_DEBUGLEVEL		2	/* debug interrupt level */
+#define XCHAL_HAVE_DEBUG_EXTERN_INT	1	/* OCD external db interrupt */
+
+/*  Type of each interrupt:  */
+#define XCHAL_INT0_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT1_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT2_TYPE 	XTHAL_INTTYPE_EXTERN_EDGE
+#define XCHAL_INT3_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT4_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT5_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT6_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT7_TYPE 	XTHAL_INTTYPE_SOFTWARE
+#define XCHAL_INT8_TYPE 	XTHAL_INTTYPE_TIMER
+#define XCHAL_INT9_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT10_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+#define XCHAL_INT11_TYPE 	XTHAL_INTTYPE_EXTERN_LEVEL
+
+/*  Masks of interrupts for each type of interrupt:  */
+#define XCHAL_INTTYPE_MASK_UNCONFIGURED	0xFFFFF000
+#define XCHAL_INTTYPE_MASK_SOFTWARE	0x00000080
+#define XCHAL_INTTYPE_MASK_EXTERN_EDGE	0x00000004
+#define XCHAL_INTTYPE_MASK_EXTERN_LEVEL	0x00000E3B
+#define XCHAL_INTTYPE_MASK_TIMER	0x00000140
+#define XCHAL_INTTYPE_MASK_NMI		0x00000000
+#define XCHAL_INTTYPE_MASK_WRITE_ERROR	0x00000000
+
+/*  Interrupt numbers assigned to specific interrupt sources:  */
+#define XCHAL_TIMER0_INTERRUPT		6	/* CCOMPARE0 */
+#define XCHAL_TIMER1_INTERRUPT		8	/* CCOMPARE1 */
+#define XCHAL_TIMER2_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+#define XCHAL_TIMER3_INTERRUPT		XTHAL_TIMER_UNCONFIGURED
+
+/*  Interrupt numbers for levels at which only one interrupt is configured:  */
+/*  (There are many interrupts each at level(s) 1.)  */
+
+
+/*
+ *  External interrupt vectors/levels.
+ *  These macros describe how Xtensa processor interrupt numbers
+ *  (as numbered internally, eg. in INTERRUPT and INTENABLE registers)
+ *  map to external BInterrupt<n> pins, for those interrupts
+ *  configured as external (level-triggered, edge-triggered, or NMI).
+ *  See the Xtensa processor databook for more details.
+ */
+
+/*  Core interrupt numbers mapped to each EXTERNAL interrupt number:  */
+#define XCHAL_EXTINT0_NUM		0	/* (intlevel 1) */
+#define XCHAL_EXTINT1_NUM		1	/* (intlevel 1) */
+#define XCHAL_EXTINT2_NUM		2	/* (intlevel 1) */
+#define XCHAL_EXTINT3_NUM		3	/* (intlevel 1) */
+#define XCHAL_EXTINT4_NUM		4	/* (intlevel 1) */
+#define XCHAL_EXTINT5_NUM		5	/* (intlevel 1) */
+#define XCHAL_EXTINT6_NUM		9	/* (intlevel 1) */
+#define XCHAL_EXTINT7_NUM		10	/* (intlevel 1) */
+#define XCHAL_EXTINT8_NUM		11	/* (intlevel 1) */
+
+
+/*----------------------------------------------------------------------
+			EXCEPTIONS and VECTORS
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_XEA_VERSION		2	/* Xtensa Exception Architecture
+						   number: 1 == XEA1 (old)
+							   2 == XEA2 (new)
+							   0 == XEAX (extern) */
+#define XCHAL_HAVE_XEA1			0	/* Exception Architecture 1 */
+#define XCHAL_HAVE_XEA2			1	/* Exception Architecture 2 */
+#define XCHAL_HAVE_XEAX			0	/* External Exception Arch. */
+#define XCHAL_HAVE_EXCEPTIONS		1	/* exception option */
+#define XCHAL_HAVE_MEM_ECC_PARITY	0	/* local memory ECC/parity */
+#define XCHAL_HAVE_VECTOR_SELECT	1	/* relocatable vectors */
+#define XCHAL_HAVE_VECBASE		1	/* relocatable vectors */
+#define XCHAL_VECBASE_RESET_VADDR	0xD0000000  /* VECBASE reset value */
+#define XCHAL_VECBASE_RESET_PADDR	0x00000000
+#define XCHAL_RESET_VECBASE_OVERLAP	0
+
+#define XCHAL_RESET_VECTOR0_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR0_PADDR	0xFE000000
+#define XCHAL_RESET_VECTOR1_VADDR	0xD8000500
+#define XCHAL_RESET_VECTOR1_PADDR	0x00000500
+#define XCHAL_RESET_VECTOR_VADDR	0xFE000000
+#define XCHAL_RESET_VECTOR_PADDR	0xFE000000
+#define XCHAL_USER_VECOFS		0x00000340
+#define XCHAL_USER_VECTOR_VADDR		0xD0000340
+#define XCHAL_USER_VECTOR_PADDR		0x00000340
+#define XCHAL_KERNEL_VECOFS		0x00000300
+#define XCHAL_KERNEL_VECTOR_VADDR	0xD0000300
+#define XCHAL_KERNEL_VECTOR_PADDR	0x00000300
+#define XCHAL_DOUBLEEXC_VECOFS		0x000003C0
+#define XCHAL_DOUBLEEXC_VECTOR_VADDR	0xD00003C0
+#define XCHAL_DOUBLEEXC_VECTOR_PADDR	0x000003C0
+#define XCHAL_WINDOW_OF4_VECOFS		0x00000000
+#define XCHAL_WINDOW_UF4_VECOFS		0x00000040
+#define XCHAL_WINDOW_OF8_VECOFS		0x00000080
+#define XCHAL_WINDOW_UF8_VECOFS		0x000000C0
+#define XCHAL_WINDOW_OF12_VECOFS	0x00000100
+#define XCHAL_WINDOW_UF12_VECOFS	0x00000140
+#define XCHAL_WINDOW_VECTORS_VADDR	0xD0000000
+#define XCHAL_WINDOW_VECTORS_PADDR	0x00000000
+#define XCHAL_INTLEVEL2_VECOFS		0x00000280
+#define XCHAL_INTLEVEL2_VECTOR_VADDR	0xD0000280
+#define XCHAL_INTLEVEL2_VECTOR_PADDR	0x00000280
+#define XCHAL_DEBUG_VECOFS		XCHAL_INTLEVEL2_VECOFS
+#define XCHAL_DEBUG_VECTOR_VADDR	XCHAL_INTLEVEL2_VECTOR_VADDR
+#define XCHAL_DEBUG_VECTOR_PADDR	XCHAL_INTLEVEL2_VECTOR_PADDR
+
+
+/*----------------------------------------------------------------------
+				DEBUG
+  ----------------------------------------------------------------------*/
+
+#define XCHAL_HAVE_OCD			1	/* OnChipDebug option */
+#define XCHAL_NUM_IBREAK		0	/* number of IBREAKn regs */
+#define XCHAL_NUM_DBREAK		0	/* number of DBREAKn regs */
+#define XCHAL_HAVE_OCD_DIR_ARRAY	0	/* faster OCD option */
+
+
+/*----------------------------------------------------------------------
+				MMU
+  ----------------------------------------------------------------------*/
+
+/*  See core-matmap.h header file for more details.  */
+
+#define XCHAL_HAVE_TLBS			1	/* inverse of HAVE_CACHEATTR */
+#define XCHAL_HAVE_SPANNING_WAY		0	/* one way maps I+D 4GB vaddr */
+#define XCHAL_HAVE_IDENTITY_MAP		0	/* vaddr == paddr always */
+#define XCHAL_HAVE_CACHEATTR		0	/* CACHEATTR register present */
+#define XCHAL_HAVE_MIMIC_CACHEATTR	0	/* region protection */
+#define XCHAL_HAVE_XLT_CACHEATTR	0	/* region prot. w/translation */
+#define XCHAL_HAVE_PTP_MMU		1	/* full MMU (with page table
+						   [autorefill] and protection)
+						   usable for an MMU-based OS */
+/*  If none of the above last 4 are set, it's a custom TLB configuration.  */
+#define XCHAL_ITLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+#define XCHAL_DTLB_ARF_ENTRIES_LOG2	2	/* log2(autorefill way size) */
+
+#define XCHAL_MMU_ASID_BITS		8	/* number of bits in ASIDs */
+#define XCHAL_MMU_RINGS			4	/* number of rings (1..4) */
+#define XCHAL_MMU_RING_BITS		2	/* num of bits in RING field */
+
+#endif /* !XTENSA_HAL_NON_PRIVILEGED_ONLY */
+
+
+#endif /* _XTENSA_CORE_CONFIGURATION_H */
diff --git a/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie-asm.h b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie-asm.h
new file mode 100644
index 000000000..6e9d69caf
--- /dev/null
+++ b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie-asm.h
@@ -0,0 +1,182 @@
+/*
+ * This header file contains assembly-language definitions (assembly
+ * macros, etc.) for this specific Xtensa processor's TIE extensions
+ * and options.  It is customized to this Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of version 2.1 of the GNU
+ * Lesser General Public License as published by the Free Software Foundation.
+ *
+ * Copyright (C) 1999-2009 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_ASM_H
+#define _XTENSA_CORE_TIE_ASM_H
+
+/*  Selection parameter values for save-area save/restore macros:  */
+/*  Option vs. TIE:  */
+#define XTHAL_SAS_TIE	0x0001	/* custom extension or coprocessor */
+#define XTHAL_SAS_OPT	0x0002	/* optional (and not a coprocessor) */
+/*  Whether used automatically by compiler:  */
+#define XTHAL_SAS_NOCC	0x0004	/* not used by compiler w/o special opts/code */
+#define XTHAL_SAS_CC	0x0008	/* used by compiler without special opts/code */
+/*  ABI handling across function calls:  */
+#define XTHAL_SAS_CALR	0x0010	/* caller-saved */
+#define XTHAL_SAS_CALE	0x0020	/* callee-saved */
+#define XTHAL_SAS_GLOB	0x0040	/* global across function calls (in thread) */
+/*  Misc  */
+#define XTHAL_SAS_ALL	0xFFFF	/* include all default NCP contents */
+
+
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (8 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rsr	\at1, BR		// boolean option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rsr	\at1, SCOMPARE1		// conditional store option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	rur	\at1, THREADPTR		// threadptr option
+	s32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_store
+
+/* Macro to save all non-coprocessor (extra) custom TIE and optional state
+ * (not including zero-overhead loop registers).
+ * Save area ptr (clobbered):  ptr  (8 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_NCP_NUM_ATMPS needed)
+ */
+	.macro xchal_ncp_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start	\continue, \ofs
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wsr	\at1, BR		// boolean option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wsr	\at1, SCOMPARE1		// conditional store option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.ifeq (XTHAL_SAS_OPT | XTHAL_SAS_CC | XTHAL_SAS_GLOB) & ~\select
+	xchal_sa_align	\ptr, 0, 1024-4, 4, 4
+	l32i	\at1, \ptr, .Lxchal_ofs_ + 0
+	wur	\at1, THREADPTR		// threadptr option
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 4
+	.endif
+	.endm	// xchal_ncp_load
+
+
+
+#define XCHAL_NCP_NUM_ATMPS	1
+
+
+
+/* Macro to save the state of TIE coprocessor AudioEngineLX.
+ * Save area ptr (clobbered):  ptr  (8 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_CP1_NUM_ATMPS needed)
+ */
+#define xchal_cp_AudioEngineLX_store	xchal_cp1_store
+/* #define xchal_cp_AudioEngineLX_store_a2	xchal_cp1_store a2 a3 a4 a5 a6 */
+	.macro	xchal_cp1_store  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start \continue, \ofs
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 0, 1, 8
+	rur240	\at1		// AE_OVF_SAR
+	s32i	\at1, \ptr, 0
+	rur241	\at1		// AE_BITHEAD
+	s32i	\at1, \ptr, 4
+	rur242	\at1		// AE_TS_FTS_BU_BP
+	s32i	\at1, \ptr, 8
+	rur243	\at1		// AE_SD_NO
+	s32i	\at1, \ptr, 12
+	AE_SP24X2S.I aep0, \ptr,  16
+	AE_SP24X2S.I aep1, \ptr,  24
+	AE_SP24X2S.I aep2, \ptr,  32
+	AE_SP24X2S.I aep3, \ptr,  40
+	AE_SP24X2S.I aep4, \ptr,  48
+	AE_SP24X2S.I aep5, \ptr,  56
+	addi	\ptr, \ptr, 64
+	AE_SP24X2S.I aep6, \ptr,  0
+	AE_SP24X2S.I aep7, \ptr,  8
+	AE_SQ56S.I aeq0, \ptr,  16
+	AE_SQ56S.I aeq1, \ptr,  24
+	AE_SQ56S.I aeq2, \ptr,  32
+	AE_SQ56S.I aeq3, \ptr,  40
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 64
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 112
+	.endif
+	.endm	// xchal_cp1_store
+
+/* Macro to restore the state of TIE coprocessor AudioEngineLX.
+ * Save area ptr (clobbered):  ptr  (8 byte aligned)
+ * Scratch regs  (clobbered):  at1..at4  (only first XCHAL_CP1_NUM_ATMPS needed)
+ */
+#define xchal_cp_AudioEngineLX_load	xchal_cp1_load
+/* #define xchal_cp_AudioEngineLX_load_a2	xchal_cp1_load a2 a3 a4 a5 a6 */
+	.macro	xchal_cp1_load  ptr at1 at2 at3 at4  continue=0 ofs=-1 select=XTHAL_SAS_ALL
+	xchal_sa_start \continue, \ofs
+	.ifeq (XTHAL_SAS_TIE | XTHAL_SAS_NOCC | XTHAL_SAS_CALR) & ~\select
+	xchal_sa_align	\ptr, 0, 0, 1, 8
+	l32i	\at1, \ptr, 0
+	wur240	\at1		// AE_OVF_SAR
+	l32i	\at1, \ptr, 4
+	wur241	\at1		// AE_BITHEAD
+	l32i	\at1, \ptr, 8
+	wur242	\at1		// AE_TS_FTS_BU_BP
+	l32i	\at1, \ptr, 12
+	wur243	\at1		// AE_SD_NO
+	addi	\ptr, \ptr, 80
+	AE_LQ56.I aeq0, \ptr,  0
+	AE_LQ56.I aeq1, \ptr,  8
+	AE_LQ56.I aeq2, \ptr,  16
+	AE_LQ56.I aeq3, \ptr,  24
+	AE_LP24X2.I aep0, \ptr,  -64
+	AE_LP24X2.I aep1, \ptr,  -56
+	AE_LP24X2.I aep2, \ptr,  -48
+	AE_LP24X2.I aep3, \ptr,  -40
+	AE_LP24X2.I aep4, \ptr,  -32
+	AE_LP24X2.I aep5, \ptr,  -24
+	AE_LP24X2.I aep6, \ptr,  -16
+	AE_LP24X2.I aep7, \ptr,  -8
+	.set	.Lxchal_pofs_, .Lxchal_pofs_ + 80
+	.set	.Lxchal_ofs_, .Lxchal_ofs_ + 112
+	.endif
+	.endm	// xchal_cp1_load
+
+#define XCHAL_CP1_NUM_ATMPS	1
+#define XCHAL_SA_NUM_ATMPS	1
+
+	/*  Empty macros for unconfigured coprocessors:  */
+	.macro xchal_cp0_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp0_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp2_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp3_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp4_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp5_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp6_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_store	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+	.macro xchal_cp7_load	p a b c d continue=0 ofs=-1 select=-1 ; .endm
+
+#endif /*_XTENSA_CORE_TIE_ASM_H*/
diff --git a/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie.h b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie.h
new file mode 100644
index 000000000..fb1f5f004
--- /dev/null
+++ b/arch/xtensa/variants/test_mmuhifi_c3/include/variant/tie.h
@@ -0,0 +1,140 @@
+/*
+ * This header file describes this specific Xtensa processor's TIE extensions
+ * that extend basic Xtensa core functionality.  It is customized to this
+ * Xtensa processor configuration.
+ *
+ * This file is subject to the terms and conditions of version 2.1 of the GNU
+ * Lesser General Public License as published by the Free Software Foundation.
+ *
+ * Copyright (C) 1999-2009 Tensilica Inc.
+ */
+
+#ifndef _XTENSA_CORE_TIE_H
+#define _XTENSA_CORE_TIE_H
+
+#define XCHAL_CP_NUM			1	/* number of coprocessors */
+#define XCHAL_CP_MAX			2	/* max CP ID + 1 (0 if none) */
+#define XCHAL_CP_MASK			0x02	/* bitmask of all CPs by ID */
+#define XCHAL_CP_PORT_MASK		0x00	/* bitmask of only port CPs */
+
+/*  Basic parameters of each coprocessor:  */
+#define XCHAL_CP1_NAME			"AudioEngineLX"
+#define XCHAL_CP1_IDENT			AudioEngineLX
+#define XCHAL_CP1_SA_SIZE		112	/* size of state save area */
+#define XCHAL_CP1_SA_ALIGN		8	/* min alignment of save area */
+#define XCHAL_CP_ID_AUDIOENGINELX	1	/* coprocessor ID (0..7) */
+
+/*  Filler info for unassigned coprocessors, to simplify arrays etc:  */
+#define XCHAL_CP0_SA_SIZE		0
+#define XCHAL_CP0_SA_ALIGN		1
+#define XCHAL_CP2_SA_SIZE		0
+#define XCHAL_CP2_SA_ALIGN		1
+#define XCHAL_CP3_SA_SIZE		0
+#define XCHAL_CP3_SA_ALIGN		1
+#define XCHAL_CP4_SA_SIZE		0
+#define XCHAL_CP4_SA_ALIGN		1
+#define XCHAL_CP5_SA_SIZE		0
+#define XCHAL_CP5_SA_ALIGN		1
+#define XCHAL_CP6_SA_SIZE		0
+#define XCHAL_CP6_SA_ALIGN		1
+#define XCHAL_CP7_SA_SIZE		0
+#define XCHAL_CP7_SA_ALIGN		1
+
+/*  Save area for non-coprocessor optional and custom (TIE) state:  */
+#define XCHAL_NCP_SA_SIZE		12
+#define XCHAL_NCP_SA_ALIGN		4
+
+/*  Total save area for optional and custom state (NCP + CPn):  */
+#define XCHAL_TOTAL_SA_SIZE		128	/* with 16-byte align padding */
+#define XCHAL_TOTAL_SA_ALIGN		8	/* actual minimum alignment */
+
+/*
+ * Detailed contents of save areas.
+ * NOTE:  caller must define the XCHAL_SA_REG macro (not defined here)
+ * before expanding the XCHAL_xxx_SA_LIST() macros.
+ *
+ * XCHAL_SA_REG(s,ccused,abikind,kind,opt,name,galign,align,asize,
+ *		dbnum,base,regnum,bitsz,gapsz,reset,x...)
+ *
+ *	s = passed from XCHAL_*_LIST(s), eg. to select how to expand
+ *	ccused = set if used by compiler without special options or code
+ *	abikind = 0 (caller-saved), 1 (callee-saved), or 2 (thread-global)
+ *	kind = 0 (special reg), 1 (TIE user reg), or 2 (TIE regfile reg)
+ *	opt = 0 (custom TIE extension or coprocessor), or 1 (optional reg)
+ *	name = lowercase reg name (no quotes)
+ *	galign = group byte alignment (power of 2) (galign >= align)
+ *	align = register byte alignment (power of 2)
+ *	asize = allocated size in bytes (asize*8 == bitsz + gapsz + padsz)
+ *	  (not including any pad bytes required to galign this or next reg)
+ *	dbnum = unique target number f/debug (see <xtensa-libdb-macros.h>)
+ *	base = reg shortname w/o index (or sr=special, ur=TIE user reg)
+ *	regnum = reg index in regfile, or special/TIE-user reg number
+ *	bitsz = number of significant bits (regfile width, or ur/sr mask bits)
+ *	gapsz = intervening bits, if bitsz bits not stored contiguously
+ *	(padsz = pad bits at end [TIE regfile] or at msbits [ur,sr] of asize)
+ *	reset = register reset value (or 0 if undefined at reset)
+ *	x = reserved for future use (0 until then)
+ *
+ *  To filter out certain registers, e.g. to expand only the non-global
+ *  registers used by the compiler, you can do something like this:
+ *
+ *  #define XCHAL_SA_REG(s,ccused,p...)	SELCC##ccused(p)
+ *  #define SELCC0(p...)
+ *  #define SELCC1(abikind,p...)	SELAK##abikind(p)
+ *  #define SELAK0(p...)		REG(p)
+ *  #define SELAK1(p...)		REG(p)
+ *  #define SELAK2(p...)
+ *  #define REG(kind,tie,name,galn,aln,asz,csz,dbnum,base,rnum,bsz,rst,x...) \
+ *		...what you want to expand...
+ */
+
+#define XCHAL_NCP_SA_NUM	3
+#define XCHAL_NCP_SA_LIST(s)	\
+ XCHAL_SA_REG(s,0,0,0,1,             br, 4, 4, 4,0x0204,  sr,4  , 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,0,1,      scompare1, 4, 4, 4,0x020C,  sr,12 , 32,0,0,0) \
+ XCHAL_SA_REG(s,1,2,1,1,      threadptr, 4, 4, 4,0x03E7,  ur,231, 32,0,0,0)
+
+#define XCHAL_CP0_SA_NUM	0
+#define XCHAL_CP0_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP1_SA_NUM	16
+#define XCHAL_CP1_SA_LIST(s)	\
+ XCHAL_SA_REG(s,0,0,1,0,     ae_ovf_sar, 8, 4, 4,0x03F0,  ur,240,  7,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,     ae_bithead, 4, 4, 4,0x03F1,  ur,241, 32,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,ae_ts_fts_bu_bp, 4, 4, 4,0x03F2,  ur,242, 16,0,0,0) \
+ XCHAL_SA_REG(s,0,0,1,0,       ae_sd_no, 4, 4, 4,0x03F3,  ur,243, 28,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep0, 8, 8, 8,0x0060, aep,0  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep1, 8, 8, 8,0x0061, aep,1  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep2, 8, 8, 8,0x0062, aep,2  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep3, 8, 8, 8,0x0063, aep,3  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep4, 8, 8, 8,0x0064, aep,4  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep5, 8, 8, 8,0x0065, aep,5  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep6, 8, 8, 8,0x0066, aep,6  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aep7, 8, 8, 8,0x0067, aep,7  , 48,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq0, 8, 8, 8,0x0068, aeq,0  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq1, 8, 8, 8,0x0069, aeq,1  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq2, 8, 8, 8,0x006A, aeq,2  , 56,0,0,0) \
+ XCHAL_SA_REG(s,0,0,2,0,           aeq3, 8, 8, 8,0x006B, aeq,3  , 56,0,0,0)
+
+#define XCHAL_CP2_SA_NUM	0
+#define XCHAL_CP2_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP3_SA_NUM	0
+#define XCHAL_CP3_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP4_SA_NUM	0
+#define XCHAL_CP4_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP5_SA_NUM	0
+#define XCHAL_CP5_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP6_SA_NUM	0
+#define XCHAL_CP6_SA_LIST(s)	/* empty */
+
+#define XCHAL_CP7_SA_NUM	0
+#define XCHAL_CP7_SA_LIST(s)	/* empty */
+
+/* Byte length of instruction from its first nibble (op0 field), per FLIX.  */
+#define XCHAL_OP0_FORMAT_LENGTHS	3,3,3,3,3,3,3,3,2,2,2,2,2,2,3,8
+
+#endif /*_XTENSA_CORE_TIE_H*/