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

1 files changed, 91 insertions, 0 deletions

diff --git a/drivers/rapidio/Kconfig b/drivers/rapidio/Kconfig
new file mode 100644
index 000000000..f1d742ac8
--- /dev/null
+++ b/drivers/rapidio/Kconfig
@@ -0,0 +1,91 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# RapidIO configuration
+#
+
+config HAVE_RAPIDIO
+	bool
+
+menuconfig RAPIDIO
+	tristate "RapidIO support"
+	depends on HAVE_RAPIDIO || PCI
+	help
+	  If you say Y here, the kernel will include drivers and
+	  infrastructure code to support RapidIO interconnect devices.
+
+source "drivers/rapidio/devices/Kconfig"
+
+config RAPIDIO_DISC_TIMEOUT
+	int "Discovery timeout duration (seconds)"
+	depends on RAPIDIO
+	default "30"
+	help
+	  Amount of time a discovery node waits for a host to complete
+	  enumeration before giving up.
+
+config RAPIDIO_ENABLE_RX_TX_PORTS
+	bool "Enable RapidIO Input/Output Ports"
+	depends on RAPIDIO
+	help
+	  The RapidIO specification describes a Output port transmit
+	  enable and a Input port receive enable. The recommended state
+	  for Input ports and Output ports should be disabled. When
+	  this switch is set the RapidIO subsystem will enable all
+	  ports for Input/Output direction to allow other traffic
+	  than Maintenance transfers.
+
+config RAPIDIO_DMA_ENGINE
+	bool "DMA Engine support for RapidIO"
+	depends on RAPIDIO
+	depends on DMADEVICES
+	select DMA_ENGINE
+	help
+	  Say Y here if you want to use DMA Engine frameork for RapidIO data
+	  transfers to/from target RIO devices. RapidIO uses NREAD and
+	  NWRITE (NWRITE_R, SWRITE) requests to transfer data between local
+	  memory and memory on remote target device. You need a DMA controller
+	  capable to perform data transfers to/from RapidIO.
+
+	  If you are unsure about this, say Y here.
+
+config RAPIDIO_DEBUG
+	bool "RapidIO subsystem debug messages"
+	depends on RAPIDIO
+	help
+	  Say Y here if you want the RapidIO subsystem to produce a bunch of
+	  debug messages to the system log. Select this if you are having a
+	  problem with the RapidIO subsystem and want to see more of what is
+	  going on.
+
+	  If you are unsure about this, say N here.
+
+config RAPIDIO_ENUM_BASIC
+	tristate "Basic Enumeration method"
+	depends on RAPIDIO
+	help
+	  This option includes basic RapidIO fabric enumeration and discovery
+	  mechanism similar to one described in RapidIO specification Annex 1.
+
+config RAPIDIO_CHMAN
+	tristate "RapidIO Channelized Messaging driver"
+	depends on RAPIDIO
+	help
+	  This option includes RapidIO channelized messaging driver which
+	  provides socket-like interface to allow sharing of single RapidIO
+	  messaging mailbox between multiple user-space applications.
+	  See "Documentation/driver-api/rapidio/rio_cm.rst" for driver description.
+
+config RAPIDIO_MPORT_CDEV
+	tristate "RapidIO /dev mport device driver"
+	depends on RAPIDIO
+	help
+	  This option includes generic RapidIO mport device driver which
+	  allows to user space applications to perform RapidIO-specific
+	  operations through selected RapidIO mport.
+
+menu "RapidIO Switch drivers"
+	depends on RAPIDIO
+
+source "drivers/rapidio/switches/Kconfig"
+
+endmenu