diff options
Diffstat (limited to 'arch/x86/um')
70 files changed, 4539 insertions, 0 deletions
diff --git a/arch/x86/um/Kconfig b/arch/x86/um/Kconfig new file mode 100644 index 000000000..44b12e45f --- /dev/null +++ b/arch/x86/um/Kconfig @@ -0,0 +1,39 @@ +# SPDX-License-Identifier: GPL-2.0 + +menu "Host processor type and features" + +source "arch/x86/Kconfig.cpu" + +endmenu + +config UML_X86 + def_bool y + select ARCH_USE_QUEUED_RWLOCKS + select ARCH_SUPPORTS_ATOMIC_RMW + select ARCH_USE_QUEUED_SPINLOCKS + select DCACHE_WORD_ACCESS + select HAVE_EFFICIENT_UNALIGNED_ACCESS + select UML_SUBARCH_SUPPORTS_SMP if X86_CX8 + +config 64BIT + bool "64-bit kernel" if "$(SUBARCH)" = "x86" + default "$(SUBARCH)" != "i386" + +config X86_32 + def_bool !64BIT + select ARCH_32BIT_OFF_T + select ARCH_WANT_IPC_PARSE_VERSION + select MODULES_USE_ELF_REL + select CLONE_BACKWARDS + select OLD_SIGSUSPEND3 + select OLD_SIGACTION + +config X86_64 + def_bool 64BIT + select MODULES_USE_ELF_RELA + +config ARCH_HAS_SC_SIGNALS + def_bool !64BIT + +config GENERIC_HWEIGHT + def_bool y diff --git a/arch/x86/um/Makefile b/arch/x86/um/Makefile new file mode 100644 index 000000000..f9ea75bf4 --- /dev/null +++ b/arch/x86/um/Makefile @@ -0,0 +1,50 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) +# + +ifeq ($(CONFIG_X86_32),y) + BITS := 32 +else + BITS := 64 +endif + +obj-y = bugs_$(BITS).o delay.o fault.o \ + ptrace.o ptrace_$(BITS).o ptrace_user.o setjmp_$(BITS).o signal.o \ + stub_segv.o \ + sys_call_table_$(BITS).o sysrq_$(BITS).o tls_$(BITS).o \ + subarch.o os-Linux/ + +ifeq ($(CONFIG_X86_32),y) + +obj-y += syscalls_32.o + +subarch-y = ../lib/string_32.o ../lib/atomic64_32.o ../lib/atomic64_cx8_32.o +subarch-y += ../lib/cmpxchg8b_emu.o ../lib/atomic64_386_32.o +subarch-y += ../lib/checksum_32.o +subarch-y += ../kernel/sys_ia32.o + +else + +obj-y += mem_64.o syscalls_64.o vdso/ + +subarch-y = ../lib/csum-partial_64.o ../lib/memcpy_64.o \ + ../lib/memmove_64.o ../lib/memset_64.o + +endif + +subarch-$(CONFIG_MODULES) += ../kernel/module.o + +USER_OBJS := bugs_$(BITS).o ptrace_user.o fault.o + +$(obj)/user-offsets.s: c_flags = -Wp,-MD,$(depfile) $(USER_CFLAGS) \ + -Iarch/x86/include/generated +targets += user-offsets.s + +include/generated/user_constants.h: $(obj)/user-offsets.s FORCE + $(call filechk,offsets,__USER_CONSTANT_H__) + +UNPROFILE_OBJS := stub_segv.o +CFLAGS_stub_segv.o := $(CFLAGS_NO_HARDENING) + +include $(srctree)/arch/um/scripts/Makefile.rules diff --git a/arch/x86/um/asm/apic.h b/arch/x86/um/asm/apic.h new file mode 100644 index 000000000..876dee84a --- /dev/null +++ b/arch/x86/um/asm/apic.h @@ -0,0 +1,4 @@ +#ifndef __UM_APIC_H +#define __UM_APIC_H + +#endif diff --git a/arch/x86/um/asm/arch_hweight.h b/arch/x86/um/asm/arch_hweight.h new file mode 100644 index 000000000..0d2d5fbb3 --- /dev/null +++ b/arch/x86/um/asm/arch_hweight.h @@ -0,0 +1,7 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_UM_HWEIGHT_H +#define _ASM_UM_HWEIGHT_H + +#include <asm-generic/bitops/arch_hweight.h> + +#endif diff --git a/arch/x86/um/asm/barrier.h b/arch/x86/um/asm/barrier.h new file mode 100644 index 000000000..b51aefd6e --- /dev/null +++ b/arch/x86/um/asm/barrier.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_UM_BARRIER_H_ +#define _ASM_UM_BARRIER_H_ + +#include <asm/cpufeatures.h> +#include <asm/alternative.h> + +/* + * Force strict CPU ordering. + * And yes, this is required on UP too when we're talking + * to devices. + */ +#ifdef CONFIG_X86_32 + +#define mb() alternative("lock addl $0,0(%%esp)", "mfence", X86_FEATURE_XMM2) +#define rmb() alternative("lock addl $0,0(%%esp)", "lfence", X86_FEATURE_XMM2) +#define wmb() alternative("lock addl $0,0(%%esp)", "sfence", X86_FEATURE_XMM) + +#else /* CONFIG_X86_32 */ + +#define mb() asm volatile("mfence" : : : "memory") +#define rmb() asm volatile("lfence" : : : "memory") +#define wmb() asm volatile("sfence" : : : "memory") + +#endif /* CONFIG_X86_32 */ + +#include <asm-generic/barrier.h> + +#endif diff --git a/arch/x86/um/asm/checksum.h b/arch/x86/um/asm/checksum.h new file mode 100644 index 000000000..ddc144657 --- /dev/null +++ b/arch/x86/um/asm/checksum.h @@ -0,0 +1,122 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_CHECKSUM_H +#define __UM_CHECKSUM_H + +#include <linux/string.h> +#include <linux/in6.h> +#include <linux/uaccess.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 + */ +extern __wsum csum_partial(const void *buff, int len, __wsum sum); + +/* Do not call this directly. Declared for export type visibility. */ +extern __visible __wsum csum_partial_copy_generic(const void *src, void *dst, int len); + +/** + * csum_fold - Fold and invert a 32bit checksum. + * sum: 32bit unfolded sum + * + * Fold a 32bit running checksum to 16bit and invert it. This is usually + * the last step before putting a checksum into a packet. + * Make sure not to mix with 64bit checksums. + */ +static inline __sum16 csum_fold(__wsum sum) +{ + __asm__( + " addl %1,%0\n" + " adcl $0xffff,%0" + : "=r" (sum) + : "r" ((__force u32)sum << 16), + "0" ((__force u32)sum & 0xffff0000) + ); + return (__force __sum16)(~(__force u32)sum >> 16); +} + +/** + * csum_tcpup_nofold - Compute an IPv4 pseudo header checksum. + * @saddr: source address + * @daddr: destination address + * @len: length of packet + * @proto: ip protocol of packet + * @sum: initial sum to be added in (32bit unfolded) + * + * Returns the pseudo header checksum the input data. Result is + * 32bit unfolded. + */ +static inline __wsum +csum_tcpudp_nofold(__be32 saddr, __be32 daddr, __u32 len, + __u8 proto, __wsum sum) +{ + asm(" addl %1, %0\n" + " adcl %2, %0\n" + " adcl %3, %0\n" + " adcl $0, %0\n" + : "=r" (sum) + : "g" (daddr), "g" (saddr), "g" ((len + proto) << 8), "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)); +} + +/** + * ip_fast_csum - Compute the IPv4 header checksum efficiently. + * iph: ipv4 header + * ihl: length of header / 4 + */ +static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl) +{ + unsigned int sum; + + asm( " movl (%1), %0\n" + " subl $4, %2\n" + " jbe 2f\n" + " addl 4(%1), %0\n" + " adcl 8(%1), %0\n" + " adcl 12(%1), %0\n" + "1: adcl 16(%1), %0\n" + " lea 4(%1), %1\n" + " decl %2\n" + " jne 1b\n" + " adcl $0, %0\n" + " movl %0, %2\n" + " shrl $16, %0\n" + " addw %w2, %w0\n" + " adcl $0, %0\n" + " notl %0\n" + "2:" + /* Since the input registers which are loaded with iph and ipl + are modified, we must also specify them as outputs, or gcc + will assume they contain their original values. */ + : "=r" (sum), "=r" (iph), "=r" (ihl) + : "1" (iph), "2" (ihl) + : "memory"); + return (__force __sum16)sum; +} + +#ifdef CONFIG_X86_32 +# include "checksum_32.h" +#else +# include "checksum_64.h" +#endif + +#endif diff --git a/arch/x86/um/asm/checksum_32.h b/arch/x86/um/asm/checksum_32.h new file mode 100644 index 000000000..0b13c2947 --- /dev/null +++ b/arch/x86/um/asm/checksum_32.h @@ -0,0 +1,38 @@ +/* + * Licensed under the GPL + */ + +#ifndef __UM_SYSDEP_CHECKSUM_H +#define __UM_SYSDEP_CHECKSUM_H + +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) +{ + __asm__( + "addl 0(%1), %0 ;\n" + "adcl 4(%1), %0 ;\n" + "adcl 8(%1), %0 ;\n" + "adcl 12(%1), %0 ;\n" + "adcl 0(%2), %0 ;\n" + "adcl 4(%2), %0 ;\n" + "adcl 8(%2), %0 ;\n" + "adcl 12(%2), %0 ;\n" + "adcl %3, %0 ;\n" + "adcl %4, %0 ;\n" + "adcl $0, %0 ;\n" + : "=&r" (sum) + : "r" (saddr), "r" (daddr), + "r"(htonl(len)), "r"(htonl(proto)), "0"(sum)); + + return csum_fold(sum); +} + +#endif diff --git a/arch/x86/um/asm/checksum_64.h b/arch/x86/um/asm/checksum_64.h new file mode 100644 index 000000000..7b6cd1921 --- /dev/null +++ b/arch/x86/um/asm/checksum_64.h @@ -0,0 +1,19 @@ +/* + * Licensed under the GPL + */ + +#ifndef __UM_SYSDEP_CHECKSUM_H +#define __UM_SYSDEP_CHECKSUM_H + +static inline unsigned add32_with_carry(unsigned a, unsigned b) +{ + asm("addl %2,%0\n\t" + "adcl $0,%0" + : "=r" (a) + : "0" (a), "r" (b)); + return a; +} + +extern __sum16 ip_compute_csum(const void *buff, int len); + +#endif diff --git a/arch/x86/um/asm/desc.h b/arch/x86/um/asm/desc.h new file mode 100644 index 000000000..34de4e93d --- /dev/null +++ b/arch/x86/um/asm/desc.h @@ -0,0 +1,17 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_DESC_H +#define __UM_DESC_H + +/* Taken from asm-i386/desc.h, it's the only thing we need. The rest wouldn't + * compile, and has never been used. */ +#define LDT_empty(info) (\ + (info)->base_addr == 0 && \ + (info)->limit == 0 && \ + (info)->contents == 0 && \ + (info)->read_exec_only == 1 && \ + (info)->seg_32bit == 0 && \ + (info)->limit_in_pages == 0 && \ + (info)->seg_not_present == 1 && \ + (info)->useable == 0 ) + +#endif diff --git a/arch/x86/um/asm/elf.h b/arch/x86/um/asm/elf.h new file mode 100644 index 000000000..22d0111b5 --- /dev/null +++ b/arch/x86/um/asm/elf.h @@ -0,0 +1,186 @@ +/* + * Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ +#ifndef __UM_ELF_X86_H +#define __UM_ELF_X86_H + +#include <asm/user.h> +#include <skas.h> + +#define CORE_DUMP_USE_REGSET + +#ifdef CONFIG_X86_32 + +#define R_386_NONE 0 +#define R_386_32 1 +#define R_386_PC32 2 +#define R_386_GOT32 3 +#define R_386_PLT32 4 +#define R_386_COPY 5 +#define R_386_GLOB_DAT 6 +#define R_386_JMP_SLOT 7 +#define R_386_RELATIVE 8 +#define R_386_GOTOFF 9 +#define R_386_GOTPC 10 +#define R_386_NUM 11 + +/* + * This is used to ensure we don't load something for the wrong architecture. + */ +#define elf_check_arch(x) \ + (((x)->e_machine == EM_386) || ((x)->e_machine == EM_486)) + +#define ELF_CLASS ELFCLASS32 +#define ELF_DATA ELFDATA2LSB +#define ELF_ARCH EM_386 + +#define ELF_PLAT_INIT(regs, load_addr) do { \ + PT_REGS_BX(regs) = 0; \ + PT_REGS_CX(regs) = 0; \ + PT_REGS_DX(regs) = 0; \ + PT_REGS_SI(regs) = 0; \ + PT_REGS_DI(regs) = 0; \ + PT_REGS_BP(regs) = 0; \ + PT_REGS_AX(regs) = 0; \ +} while (0) + +/* Shamelessly stolen from include/asm-i386/elf.h */ + +#define ELF_CORE_COPY_REGS(pr_reg, regs) do { \ + pr_reg[0] = PT_REGS_BX(regs); \ + pr_reg[1] = PT_REGS_CX(regs); \ + pr_reg[2] = PT_REGS_DX(regs); \ + pr_reg[3] = PT_REGS_SI(regs); \ + pr_reg[4] = PT_REGS_DI(regs); \ + pr_reg[5] = PT_REGS_BP(regs); \ + pr_reg[6] = PT_REGS_AX(regs); \ + pr_reg[7] = PT_REGS_DS(regs); \ + pr_reg[8] = PT_REGS_ES(regs); \ + /* fake once used fs and gs selectors? */ \ + pr_reg[9] = PT_REGS_DS(regs); \ + pr_reg[10] = PT_REGS_DS(regs); \ + pr_reg[11] = PT_REGS_SYSCALL_NR(regs); \ + pr_reg[12] = PT_REGS_IP(regs); \ + pr_reg[13] = PT_REGS_CS(regs); \ + pr_reg[14] = PT_REGS_EFLAGS(regs); \ + pr_reg[15] = PT_REGS_SP(regs); \ + pr_reg[16] = PT_REGS_SS(regs); \ +} while (0); + +#define ELF_PLATFORM_FALLBACK "i586" + +#else + +/* x86-64 relocation types, taken from asm-x86_64/elf.h */ +#define R_X86_64_NONE 0 /* No reloc */ +#define R_X86_64_64 1 /* Direct 64 bit */ +#define R_X86_64_PC32 2 /* PC relative 32 bit signed */ +#define R_X86_64_GOT32 3 /* 32 bit GOT entry */ +#define R_X86_64_PLT32 4 /* 32 bit PLT address */ +#define R_X86_64_COPY 5 /* Copy symbol at runtime */ +#define R_X86_64_GLOB_DAT 6 /* Create GOT entry */ +#define R_X86_64_JUMP_SLOT 7 /* Create PLT entry */ +#define R_X86_64_RELATIVE 8 /* Adjust by program base */ +#define R_X86_64_GOTPCREL 9 /* 32 bit signed pc relative + offset to GOT */ +#define R_X86_64_32 10 /* Direct 32 bit zero extended */ +#define R_X86_64_32S 11 /* Direct 32 bit sign extended */ +#define R_X86_64_16 12 /* Direct 16 bit zero extended */ +#define R_X86_64_PC16 13 /* 16 bit sign extended pc relative */ +#define R_X86_64_8 14 /* Direct 8 bit sign extended */ +#define R_X86_64_PC8 15 /* 8 bit sign extended pc relative */ +#define R_X86_64_PC64 24 /* Place relative 64-bit signed */ + +/* + * This is used to ensure we don't load something for the wrong architecture. + */ +#define elf_check_arch(x) \ + ((x)->e_machine == EM_X86_64) + +#define ELF_CLASS ELFCLASS64 +#define ELF_DATA ELFDATA2LSB +#define ELF_ARCH EM_X86_64 + +#define ELF_PLAT_INIT(regs, load_addr) do { \ + PT_REGS_BX(regs) = 0; \ + PT_REGS_CX(regs) = 0; \ + PT_REGS_DX(regs) = 0; \ + PT_REGS_SI(regs) = 0; \ + PT_REGS_DI(regs) = 0; \ + PT_REGS_BP(regs) = 0; \ + PT_REGS_AX(regs) = 0; \ + PT_REGS_R8(regs) = 0; \ + PT_REGS_R9(regs) = 0; \ + PT_REGS_R10(regs) = 0; \ + PT_REGS_R11(regs) = 0; \ + PT_REGS_R12(regs) = 0; \ + PT_REGS_R13(regs) = 0; \ + PT_REGS_R14(regs) = 0; \ + PT_REGS_R15(regs) = 0; \ +} while (0) + +#define ELF_CORE_COPY_REGS(pr_reg, _regs) \ + (pr_reg)[0] = (_regs)->regs.gp[0]; \ + (pr_reg)[1] = (_regs)->regs.gp[1]; \ + (pr_reg)[2] = (_regs)->regs.gp[2]; \ + (pr_reg)[3] = (_regs)->regs.gp[3]; \ + (pr_reg)[4] = (_regs)->regs.gp[4]; \ + (pr_reg)[5] = (_regs)->regs.gp[5]; \ + (pr_reg)[6] = (_regs)->regs.gp[6]; \ + (pr_reg)[7] = (_regs)->regs.gp[7]; \ + (pr_reg)[8] = (_regs)->regs.gp[8]; \ + (pr_reg)[9] = (_regs)->regs.gp[9]; \ + (pr_reg)[10] = (_regs)->regs.gp[10]; \ + (pr_reg)[11] = (_regs)->regs.gp[11]; \ + (pr_reg)[12] = (_regs)->regs.gp[12]; \ + (pr_reg)[13] = (_regs)->regs.gp[13]; \ + (pr_reg)[14] = (_regs)->regs.gp[14]; \ + (pr_reg)[15] = (_regs)->regs.gp[15]; \ + (pr_reg)[16] = (_regs)->regs.gp[16]; \ + (pr_reg)[17] = (_regs)->regs.gp[17]; \ + (pr_reg)[18] = (_regs)->regs.gp[18]; \ + (pr_reg)[19] = (_regs)->regs.gp[19]; \ + (pr_reg)[20] = (_regs)->regs.gp[20]; \ + (pr_reg)[21] = (_regs)->regs.gp[21]; \ + (pr_reg)[22] = (_regs)->regs.gp[22]; \ + (pr_reg)[23] = 0; \ + (pr_reg)[24] = 0; \ + (pr_reg)[25] = 0; \ + (pr_reg)[26] = 0; + +#define ELF_PLATFORM_FALLBACK "x86_64" + +#define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1 +struct linux_binprm; +extern int arch_setup_additional_pages(struct linux_binprm *bprm, + int uses_interp); + +extern unsigned long um_vdso_addr; +#define AT_SYSINFO_EHDR 33 +#define ARCH_DLINFO NEW_AUX_ENT(AT_SYSINFO_EHDR, um_vdso_addr) + +#endif + +typedef unsigned long elf_greg_t; + +#define ELF_NGREG (sizeof(struct user_regs_struct) / sizeof(elf_greg_t)) +typedef elf_greg_t elf_gregset_t[ELF_NGREG]; + +typedef struct user_i387_struct elf_fpregset_t; + +struct task_struct; + +#define ELF_EXEC_PAGESIZE 4096 + +#define ELF_ET_DYN_BASE (TASK_SIZE / 3 * 2) + +extern long elf_aux_hwcap; +#define ELF_HWCAP (elf_aux_hwcap) + +extern char *elf_aux_platform; +#define ELF_PLATFORM (elf_aux_platform ?: ELF_PLATFORM_FALLBACK) + +#define SET_PERSONALITY(ex) do {} while(0) + +#endif diff --git a/arch/x86/um/asm/irq_vectors.h b/arch/x86/um/asm/irq_vectors.h new file mode 100644 index 000000000..272a81e0c --- /dev/null +++ b/arch/x86/um/asm/irq_vectors.h @@ -0,0 +1,10 @@ +/* + * Copyright (C) 2002 Jeff Dike (jdike@karaya.com) + * Licensed under the GPL + */ + +#ifndef __UM_IRQ_VECTORS_H +#define __UM_IRQ_VECTORS_H + +#endif + diff --git a/arch/x86/um/asm/processor.h b/arch/x86/um/asm/processor.h new file mode 100644 index 000000000..e222d2ae2 --- /dev/null +++ b/arch/x86/um/asm/processor.h @@ -0,0 +1,43 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_PROCESSOR_H +#define __UM_PROCESSOR_H +#include <linux/time-internal.h> + +/* include faultinfo structure */ +#include <sysdep/faultinfo.h> + +#ifdef CONFIG_X86_32 +# include "processor_32.h" +#else +# include "processor_64.h" +#endif + +#define KSTK_EIP(tsk) KSTK_REG(tsk, HOST_IP) +#define KSTK_ESP(tsk) KSTK_REG(tsk, HOST_SP) +#define KSTK_EBP(tsk) KSTK_REG(tsk, HOST_BP) + +#define ARCH_IS_STACKGROW(address) \ + (address + 65536 + 32 * sizeof(unsigned long) >= UPT_SP(¤t->thread.regs.regs)) + +#include <asm/user.h> + +/* PAUSE is a good thing to insert into busy-wait loops. */ +static __always_inline void native_pause(void) +{ + __asm__ __volatile__("pause": : :"memory"); +} + +static __always_inline void cpu_relax(void) +{ + if (time_travel_mode == TT_MODE_INFCPU || + time_travel_mode == TT_MODE_EXTERNAL) + time_travel_ndelay(1); + else + native_pause(); +} + +#define task_pt_regs(t) (&(t)->thread.regs) + +#include <asm/processor-generic.h> + +#endif diff --git a/arch/x86/um/asm/processor_32.h b/arch/x86/um/asm/processor_32.h new file mode 100644 index 000000000..5fb1b8449 --- /dev/null +++ b/arch/x86/um/asm/processor_32.h @@ -0,0 +1,53 @@ +/* + * Copyright (C) 2002 Jeff Dike (jdike@karaya.com) + * Licensed under the GPL + */ + +#ifndef __UM_PROCESSOR_I386_H +#define __UM_PROCESSOR_I386_H + +#include <linux/string.h> +#include <asm/segment.h> +#include <asm/ldt.h> + +extern int host_has_cmov; + +struct uml_tls_struct { + struct user_desc tls; + unsigned flushed:1; + unsigned present:1; +}; + +struct arch_thread { + struct uml_tls_struct tls_array[GDT_ENTRY_TLS_ENTRIES]; + unsigned long debugregs[8]; + int debugregs_seq; + struct faultinfo faultinfo; +}; + +#define INIT_ARCH_THREAD { \ + .tls_array = { [ 0 ... GDT_ENTRY_TLS_ENTRIES - 1 ] = \ + { .present = 0, .flushed = 0 } }, \ + .debugregs = { [ 0 ... 7 ] = 0 }, \ + .debugregs_seq = 0, \ + .faultinfo = { 0, 0, 0 } \ +} + +#define STACKSLOTS_PER_LINE 8 + +static inline void arch_flush_thread(struct arch_thread *thread) +{ + /* Clear any TLS still hanging */ + memset(&thread->tls_array, 0, sizeof(thread->tls_array)); +} + +static inline void arch_copy_thread(struct arch_thread *from, + struct arch_thread *to) +{ + memcpy(&to->tls_array, &from->tls_array, sizeof(from->tls_array)); +} + +#define current_sp() ({ void *sp; __asm__("movl %%esp, %0" : "=r" (sp) : ); sp; }) +#define current_bp() ({ unsigned long bp; __asm__("movl %%ebp, %0" : "=r" (bp) : ); bp; }) + +#endif diff --git a/arch/x86/um/asm/processor_64.h b/arch/x86/um/asm/processor_64.h new file mode 100644 index 000000000..f90159508 --- /dev/null +++ b/arch/x86/um/asm/processor_64.h @@ -0,0 +1,34 @@ +/* + * Copyright 2003 PathScale, Inc. + * + * Licensed under the GPL + */ + +#ifndef __UM_PROCESSOR_X86_64_H +#define __UM_PROCESSOR_X86_64_H + +struct arch_thread { + unsigned long debugregs[8]; + int debugregs_seq; + struct faultinfo faultinfo; +}; + +#define INIT_ARCH_THREAD { .debugregs = { [ 0 ... 7 ] = 0 }, \ + .debugregs_seq = 0, \ + .faultinfo = { 0, 0, 0 } } + +#define STACKSLOTS_PER_LINE 4 + +static inline void arch_flush_thread(struct arch_thread *thread) +{ +} + +static inline void arch_copy_thread(struct arch_thread *from, + struct arch_thread *to) +{ +} + +#define current_sp() ({ void *sp; __asm__("movq %%rsp, %0" : "=r" (sp) : ); sp; }) +#define current_bp() ({ unsigned long bp; __asm__("movq %%rbp, %0" : "=r" (bp) : ); bp; }) + +#endif diff --git a/arch/x86/um/asm/ptrace.h b/arch/x86/um/asm/ptrace.h new file mode 100644 index 000000000..2641d28d1 --- /dev/null +++ b/arch/x86/um/asm/ptrace.h @@ -0,0 +1,103 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_X86_PTRACE_H +#define __UM_X86_PTRACE_H + +/* This is here because signal.c needs the REGSET_FP_LEGACY definition */ +enum { + REGSET_GENERAL, +#ifdef CONFIG_X86_32 + REGSET_FP_LEGACY, +#endif + REGSET_FP, + REGSET_XSTATE, +}; + +#include <linux/compiler.h> +#ifndef CONFIG_X86_32 +#define __FRAME_OFFSETS /* Needed to get the R* macros */ +#endif +#include <asm/ptrace-generic.h> + +#define user_mode(r) UPT_IS_USER(&(r)->regs) + +#define PT_REGS_AX(r) UPT_AX(&(r)->regs) +#define PT_REGS_BX(r) UPT_BX(&(r)->regs) +#define PT_REGS_CX(r) UPT_CX(&(r)->regs) +#define PT_REGS_DX(r) UPT_DX(&(r)->regs) + +#define PT_REGS_SI(r) UPT_SI(&(r)->regs) +#define PT_REGS_DI(r) UPT_DI(&(r)->regs) +#define PT_REGS_BP(r) UPT_BP(&(r)->regs) +#define PT_REGS_EFLAGS(r) UPT_EFLAGS(&(r)->regs) + +#define PT_REGS_CS(r) UPT_CS(&(r)->regs) +#define PT_REGS_SS(r) UPT_SS(&(r)->regs) +#define PT_REGS_DS(r) UPT_DS(&(r)->regs) +#define PT_REGS_ES(r) UPT_ES(&(r)->regs) + +#define PT_REGS_ORIG_SYSCALL(r) PT_REGS_AX(r) +#define PT_REGS_SYSCALL_RET(r) PT_REGS_AX(r) + +#define PT_FIX_EXEC_STACK(sp) do ; while(0) + +#define profile_pc(regs) PT_REGS_IP(regs) + +#define UPT_RESTART_SYSCALL(r) (UPT_IP(r) -= 2) +#define PT_REGS_SET_SYSCALL_RETURN(r, res) (PT_REGS_AX(r) = (res)) + +static inline long regs_return_value(struct pt_regs *regs) +{ + return PT_REGS_AX(regs); +} + +/* + * Forward declaration to avoid including sysdep/tls.h, which causes a + * circular include, and compilation failures. + */ +struct user_desc; + +#ifdef CONFIG_X86_32 + +extern int ptrace_get_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc); + +extern int ptrace_set_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc); + +extern int arch_switch_tls(struct task_struct *to); + +#else + +#define PT_REGS_R8(r) UPT_R8(&(r)->regs) +#define PT_REGS_R9(r) UPT_R9(&(r)->regs) +#define PT_REGS_R10(r) UPT_R10(&(r)->regs) +#define PT_REGS_R11(r) UPT_R11(&(r)->regs) +#define PT_REGS_R12(r) UPT_R12(&(r)->regs) +#define PT_REGS_R13(r) UPT_R13(&(r)->regs) +#define PT_REGS_R14(r) UPT_R14(&(r)->regs) +#define PT_REGS_R15(r) UPT_R15(&(r)->regs) + +#include <asm/errno.h> + +static inline int ptrace_get_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc) +{ + return -ENOSYS; +} + +static inline int ptrace_set_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc) +{ + return -ENOSYS; +} + +extern long arch_prctl(struct task_struct *task, int option, + unsigned long __user *addr); + +#endif + +#define user_stack_pointer(regs) PT_REGS_SP(regs) + +extern void arch_switch_to(struct task_struct *to); + +#endif /* __UM_X86_PTRACE_H */ diff --git a/arch/x86/um/asm/required-features.h b/arch/x86/um/asm/required-features.h new file mode 100644 index 000000000..dfb967b2d --- /dev/null +++ b/arch/x86/um/asm/required-features.h @@ -0,0 +1,9 @@ +#ifndef __UM_REQUIRED_FEATURES_H +#define __UM_REQUIRED_FEATURES_H + +/* + * Nothing to see, just need something for the i386 and x86_64 asm + * headers to include. + */ + +#endif diff --git a/arch/x86/um/asm/segment.h b/arch/x86/um/asm/segment.h new file mode 100644 index 000000000..2ef507bc6 --- /dev/null +++ b/arch/x86/um/asm/segment.h @@ -0,0 +1,11 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_SEGMENT_H +#define __UM_SEGMENT_H + +extern int host_gdt_entry_tls_min; + +#define GDT_ENTRY_TLS_ENTRIES 3 +#define GDT_ENTRY_TLS_MIN host_gdt_entry_tls_min +#define GDT_ENTRY_TLS_MAX (GDT_ENTRY_TLS_MIN + GDT_ENTRY_TLS_ENTRIES - 1) + +#endif diff --git a/arch/x86/um/asm/spinlock.h b/arch/x86/um/asm/spinlock.h new file mode 100644 index 000000000..20fc77514 --- /dev/null +++ b/arch/x86/um/asm/spinlock.h @@ -0,0 +1,8 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __ASM_UM_SPINLOCK_H +#define __ASM_UM_SPINLOCK_H + +#include <asm/qspinlock.h> +#include <asm/qrwlock.h> + +#endif /* __ASM_UM_SPINLOCK_H */ diff --git a/arch/x86/um/asm/syscall.h b/arch/x86/um/asm/syscall.h new file mode 100644 index 000000000..d6208d0fa --- /dev/null +++ b/arch/x86/um/asm/syscall.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __UM_ASM_SYSCALL_H +#define __UM_ASM_SYSCALL_H + +#include <asm/syscall-generic.h> +#include <uapi/linux/audit.h> + +typedef asmlinkage long (*sys_call_ptr_t)(unsigned long, unsigned long, + unsigned long, unsigned long, + unsigned long, unsigned long); + +extern const sys_call_ptr_t sys_call_table[]; + +static inline int syscall_get_arch(struct task_struct *task) +{ +#ifdef CONFIG_X86_32 + return AUDIT_ARCH_I386; +#else + return AUDIT_ARCH_X86_64; +#endif +} + +#endif /* __UM_ASM_SYSCALL_H */ diff --git a/arch/x86/um/asm/vm-flags.h b/arch/x86/um/asm/vm-flags.h new file mode 100644 index 000000000..622d36d6d --- /dev/null +++ b/arch/x86/um/asm/vm-flags.h @@ -0,0 +1,19 @@ +/* + * Copyright (C) 2004 Jeff Dike (jdike@addtoit.com) + * Copyright 2003 PathScale, Inc. + * Licensed under the GPL + */ + +#ifndef __VM_FLAGS_X86_H +#define __VM_FLAGS_X86_H + +#ifdef CONFIG_X86_32 + +#define VMA_DATA_DEFAULT_FLAGS VMA_DATA_FLAGS_TSK_EXEC + +#else + +#define VMA_STACK_DEFAULT_FLAGS append_vma_flags(VMA_DATA_FLAGS_EXEC, VMA_GROWSDOWN_BIT) + +#endif +#endif diff --git a/arch/x86/um/bugs_32.c b/arch/x86/um/bugs_32.c new file mode 100644 index 000000000..d29929efc --- /dev/null +++ b/arch/x86/um/bugs_32.c @@ -0,0 +1,75 @@ +/* + * Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <arch.h> +#include <signal.h> +#include <kern_util.h> +#include <longjmp.h> +#include <sysdep/ptrace.h> +#include <generated/asm-offsets.h> + +/* Set during early boot */ +static int host_has_cmov = 1; +static jmp_buf cmov_test_return; + +static void cmov_sigill_test_handler(int sig) +{ + host_has_cmov = 0; + longjmp(cmov_test_return, 1); +} + +void arch_check_bugs(void) +{ + struct sigaction old, new; + + printk(UM_KERN_INFO "Checking for host processor cmov support..."); + new.sa_handler = cmov_sigill_test_handler; + + /* Make sure that SIGILL is enabled after the handler longjmps back */ + new.sa_flags = SA_NODEFER; + sigemptyset(&new.sa_mask); + sigaction(SIGILL, &new, &old); + + if (setjmp(cmov_test_return) == 0) { + unsigned long foo = 0; + __asm__ __volatile__("cmovz %0, %1" : "=r" (foo) : "0" (foo)); + printk(UM_KERN_CONT "Yes\n"); + } else + printk(UM_KERN_CONT "No\n"); + + sigaction(SIGILL, &old, &new); +} + +void arch_examine_signal(int sig, struct uml_pt_regs *regs) +{ + unsigned char tmp[2]; + + /* + * This is testing for a cmov (0x0f 0x4x) instruction causing a + * SIGILL in init. + */ + if ((sig != SIGILL) || (get_current_pid() != 1)) + return; + + if (copy_from_user_proc(tmp, (void *) UPT_IP(regs), 2)) { + printk(UM_KERN_ERR "SIGILL in init, could not read " + "instructions!\n"); + return; + } + + if ((tmp[0] != 0x0f) || ((tmp[1] & 0xf0) != 0x40)) + return; + + if (host_has_cmov == 0) + printk(UM_KERN_ERR "SIGILL caused by cmov, which this " + "processor doesn't implement. Boot a filesystem " + "compiled for older processors"); + else if (host_has_cmov == 1) + printk(UM_KERN_ERR "SIGILL caused by cmov, which this " + "processor claims to implement"); + else + printk(UM_KERN_ERR "Bad value for host_has_cmov (%d)", + host_has_cmov); +} diff --git a/arch/x86/um/bugs_64.c b/arch/x86/um/bugs_64.c new file mode 100644 index 000000000..b01295e8a --- /dev/null +++ b/arch/x86/um/bugs_64.c @@ -0,0 +1,16 @@ +/* + * Copyright 2003 PathScale, Inc. + * + * Licensed under the GPL + */ + +#include <arch.h> +#include <sysdep/ptrace.h> + +void arch_check_bugs(void) +{ +} + +void arch_examine_signal(int sig, struct uml_pt_regs *regs) +{ +} diff --git a/arch/x86/um/delay.c b/arch/x86/um/delay.c new file mode 100644 index 000000000..8d510ceb4 --- /dev/null +++ b/arch/x86/um/delay.c @@ -0,0 +1,57 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2011 Richard Weinberger <richrd@nod.at> + * Mostly copied from arch/x86/lib/delay.c + */ + +#include <linux/export.h> +#include <linux/kernel.h> +#include <linux/delay.h> +#include <asm/param.h> + +void __delay(unsigned long loops) +{ + asm volatile( + "test %0,%0\n" + "jz 3f\n" + "jmp 1f\n" + + ".align 16\n" + "1: jmp 2f\n" + + ".align 16\n" + "2: dec %0\n" + " jnz 2b\n" + "3: dec %0\n" + + : /* we don't need output */ + : "a" (loops) + ); +} +EXPORT_SYMBOL(__delay); + +inline void __const_udelay(unsigned long xloops) +{ + int d0; + + xloops *= 4; + asm("mull %%edx" + : "=d" (xloops), "=&a" (d0) + : "1" (xloops), "0" + (loops_per_jiffy * (HZ/4))); + + __delay(++xloops); +} +EXPORT_SYMBOL(__const_udelay); + +void __udelay(unsigned long usecs) +{ + __const_udelay(usecs * 0x000010c7); /* 2**32 / 1000000 (rounded up) */ +} +EXPORT_SYMBOL(__udelay); + +void __ndelay(unsigned long nsecs) +{ + __const_udelay(nsecs * 0x00005); /* 2**32 / 1000000000 (rounded up) */ +} +EXPORT_SYMBOL(__ndelay); diff --git a/arch/x86/um/fault.c b/arch/x86/um/fault.c new file mode 100644 index 000000000..0dde4d613 --- /dev/null +++ b/arch/x86/um/fault.c @@ -0,0 +1,29 @@ +/* + * Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <arch.h> +#include <sysdep/ptrace.h> + +/* These two are from asm-um/uaccess.h and linux/module.h, check them. */ +struct exception_table_entry +{ + unsigned long insn; + unsigned long fixup; +}; + +const struct exception_table_entry *search_exception_tables(unsigned long add); + +/* Compare this to arch/i386/mm/extable.c:fixup_exception() */ +int arch_fixup(unsigned long address, struct uml_pt_regs *regs) +{ + const struct exception_table_entry *fixup; + + fixup = search_exception_tables(address); + if (fixup) { + UPT_IP(regs) = fixup->fixup; + return 1; + } + return 0; +} diff --git a/arch/x86/um/mem_64.c b/arch/x86/um/mem_64.c new file mode 100644 index 000000000..c027e93d1 --- /dev/null +++ b/arch/x86/um/mem_64.c @@ -0,0 +1,11 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/mm.h> +#include <asm/elf.h> + +const char *arch_vma_name(struct vm_area_struct *vma) +{ + if (vma->vm_mm && vma->vm_start == um_vdso_addr) + return "[vdso]"; + + return NULL; +} diff --git a/arch/x86/um/os-Linux/Makefile b/arch/x86/um/os-Linux/Makefile new file mode 100644 index 000000000..77a308aaa --- /dev/null +++ b/arch/x86/um/os-Linux/Makefile @@ -0,0 +1,12 @@ +# +# Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) +# Licensed under the GPL +# + +obj-y = registers.o mcontext.o + +obj-$(CONFIG_X86_32) += tls.o + +USER_OBJS := $(obj-y) + +include $(srctree)/arch/um/scripts/Makefile.rules diff --git a/arch/x86/um/os-Linux/mcontext.c b/arch/x86/um/os-Linux/mcontext.c new file mode 100644 index 000000000..b1580df80 --- /dev/null +++ b/arch/x86/um/os-Linux/mcontext.c @@ -0,0 +1,266 @@ +// SPDX-License-Identifier: GPL-2.0 +#define __FRAME_OFFSETS +#include <linux/errno.h> +#include <linux/string.h> +#include <sys/ucontext.h> +#include <asm/ptrace.h> +/* + * musl defines struct sigcontext in <bits/signal.h>. Rename the kernel's + * copy to avoid redefinition while keeping the FP-state types available. + */ +#define sigcontext __kernel_sigcontext +#include <asm/sigcontext.h> +#undef sigcontext +#include <sysdep/ptrace.h> +#include <sysdep/mcontext.h> +#include <arch.h> + +void get_regs_from_mc(struct uml_pt_regs *regs, mcontext_t *mc) +{ +#ifdef __i386__ +#define COPY2(X,Y) regs->gp[X] = mc->gregs[REG_##Y] +#define COPY(X) regs->gp[X] = mc->gregs[REG_##X] +#define COPY_SEG(X) regs->gp[X] = mc->gregs[REG_##X] & 0xffff; +#define COPY_SEG_CPL3(X) regs->gp[X] = (mc->gregs[REG_##X] & 0xffff) | 3; + COPY_SEG(GS); COPY_SEG(FS); COPY_SEG(ES); COPY_SEG(DS); + COPY(EDI); COPY(ESI); COPY(EBP); + COPY2(UESP, ESP); /* sic */ + COPY(EBX); COPY(EDX); COPY(ECX); COPY(EAX); + COPY(EIP); COPY_SEG_CPL3(CS); COPY(EFL); COPY_SEG_CPL3(SS); +#undef COPY2 +#undef COPY +#undef COPY_SEG +#undef COPY_SEG_CPL3 +#else +#define COPY2(X,Y) regs->gp[X/sizeof(unsigned long)] = mc->gregs[REG_##Y] +#define COPY(X) regs->gp[X/sizeof(unsigned long)] = mc->gregs[REG_##X] + COPY(R8); COPY(R9); COPY(R10); COPY(R11); + COPY(R12); COPY(R13); COPY(R14); COPY(R15); + COPY(RDI); COPY(RSI); COPY(RBP); COPY(RBX); + COPY(RDX); COPY(RAX); COPY(RCX); COPY(RSP); + COPY(RIP); + COPY2(EFLAGS, EFL); + COPY2(CS, CSGSFS); + regs->gp[SS / sizeof(unsigned long)] = mc->gregs[REG_CSGSFS] >> 48; +#undef COPY2 +#undef COPY +#endif +} + +void mc_set_rip(void *_mc, void *target) +{ + mcontext_t *mc = _mc; + +#ifdef __i386__ + mc->gregs[REG_EIP] = (unsigned long)target; +#else + mc->gregs[REG_RIP] = (unsigned long)target; +#endif +} + +/* Same thing, but the copy macros are turned around. */ +void get_mc_from_regs(struct uml_pt_regs *regs, mcontext_t *mc, int single_stepping) +{ +#ifdef __i386__ +#define COPY2(X,Y) mc->gregs[REG_##Y] = regs->gp[X] +#define COPY(X) mc->gregs[REG_##X] = regs->gp[X] +#define COPY_SEG(X) mc->gregs[REG_##X] = regs->gp[X] & 0xffff; +#define COPY_SEG_CPL3(X) mc->gregs[REG_##X] = (regs->gp[X] & 0xffff) | 3; + COPY_SEG(GS); COPY_SEG(FS); COPY_SEG(ES); COPY_SEG(DS); + COPY(EDI); COPY(ESI); COPY(EBP); + COPY2(UESP, ESP); /* sic */ + COPY(EBX); COPY(EDX); COPY(ECX); COPY(EAX); + COPY(EIP); COPY_SEG_CPL3(CS); COPY(EFL); COPY_SEG_CPL3(SS); +#else +#define COPY2(X,Y) mc->gregs[REG_##Y] = regs->gp[X/sizeof(unsigned long)] +#define COPY(X) mc->gregs[REG_##X] = regs->gp[X/sizeof(unsigned long)] + COPY(R8); COPY(R9); COPY(R10); COPY(R11); + COPY(R12); COPY(R13); COPY(R14); COPY(R15); + COPY(RDI); COPY(RSI); COPY(RBP); COPY(RBX); + COPY(RDX); COPY(RAX); COPY(RCX); COPY(RSP); + COPY(RIP); + COPY2(EFLAGS, EFL); + mc->gregs[REG_CSGSFS] = mc->gregs[REG_CSGSFS] & 0xffffffffffffl; + mc->gregs[REG_CSGSFS] |= (regs->gp[SS / sizeof(unsigned long)] & 0xffff) << 48; +#endif + + if (single_stepping) + mc->gregs[REG_EFL] |= X86_EFLAGS_TF; + else + mc->gregs[REG_EFL] &= ~X86_EFLAGS_TF; +} + +#ifdef CONFIG_X86_32 +struct _xstate_64 { + struct _fpstate_64 fpstate; + struct _header xstate_hdr; + struct _ymmh_state ymmh; + /* New processor state extensions go here: */ +}; + +/* Not quite the right structures as these contain more information */ +int um_i387_from_fxsr(struct _fpstate_32 *i387, + const struct _fpstate_64 *fxsave); +int um_fxsr_from_i387(struct _fpstate_64 *fxsave, + const struct _fpstate_32 *from); +#else +#define _xstate_64 _xstate +#endif + +static struct _fpstate *get_fpstate(struct stub_data *data, + mcontext_t *mcontext, + int *fp_size) +{ + struct _fpstate *res; + + /* Assume floating point registers are on the same page */ + res = (void *)(((unsigned long)mcontext->fpregs & + (UM_KERN_PAGE_SIZE - 1)) + + (unsigned long)&data->sigstack[0]); + + if ((void *)res + sizeof(struct _fpstate) > + (void *)data->sigstack + sizeof(data->sigstack)) + return NULL; + + if (res->sw_reserved.magic1 != FP_XSTATE_MAGIC1) { + *fp_size = sizeof(struct _fpstate); + } else { + char *magic2_addr; + + magic2_addr = (void *)res; + magic2_addr += res->sw_reserved.extended_size; + magic2_addr -= FP_XSTATE_MAGIC2_SIZE; + + /* We still need to be within our stack */ + if ((void *)magic2_addr > + (void *)data->sigstack + sizeof(data->sigstack)) + return NULL; + + /* If we do not read MAGIC2, then we did something wrong */ + if (*(__u32 *)magic2_addr != FP_XSTATE_MAGIC2) + return NULL; + + /* Remove MAGIC2 from the size, we do not save/restore it */ + *fp_size = res->sw_reserved.extended_size - + FP_XSTATE_MAGIC2_SIZE; + } + + return res; +} + +int get_stub_state(struct uml_pt_regs *regs, struct stub_data *data, + unsigned long *fp_size_out) +{ + mcontext_t *mcontext; + struct _fpstate *fpstate_stub; + struct _xstate_64 *xstate_stub; + int fp_size, xstate_size; + + /* mctx_offset is verified by wait_stub_done_seccomp */ + mcontext = (void *)&data->sigstack[data->mctx_offset]; + + get_regs_from_mc(regs, mcontext); + + fpstate_stub = get_fpstate(data, mcontext, &fp_size); + if (!fpstate_stub) + return -EINVAL; + +#ifdef CONFIG_X86_32 + xstate_stub = (void *)&fpstate_stub->_fxsr_env; + xstate_size = fp_size - offsetof(struct _fpstate_32, _fxsr_env); +#else + xstate_stub = (void *)fpstate_stub; + xstate_size = fp_size; +#endif + + if (fp_size_out) + *fp_size_out = xstate_size; + + if (xstate_size > host_fp_size) + return -ENOSPC; + + memcpy(®s->fp, xstate_stub, xstate_size); + + /* We do not need to read the x86_64 FS_BASE/GS_BASE registers as + * we do not permit userspace to set them directly. + */ + +#ifdef CONFIG_X86_32 + /* Read the i387 legacy FP registers */ + if (um_fxsr_from_i387((void *)®s->fp, fpstate_stub)) + return -EINVAL; +#endif + + return 0; +} + +/* Copied because we cannot include regset.h here. */ +struct task_struct; +struct user_regset; +struct membuf { + void *p; + size_t left; +}; + +int fpregs_legacy_get(struct task_struct *target, + const struct user_regset *regset, + struct membuf to); + +int set_stub_state(struct uml_pt_regs *regs, struct stub_data *data, + int single_stepping) +{ + mcontext_t *mcontext; + struct _fpstate *fpstate_stub; + struct _xstate_64 *xstate_stub; + int fp_size, xstate_size; + + /* mctx_offset is verified by wait_stub_done_seccomp */ + mcontext = (void *)&data->sigstack[data->mctx_offset]; + + if ((unsigned long)mcontext < (unsigned long)data->sigstack || + (unsigned long)mcontext > + (unsigned long) data->sigstack + + sizeof(data->sigstack) - sizeof(*mcontext)) + return -EINVAL; + + get_mc_from_regs(regs, mcontext, single_stepping); + + fpstate_stub = get_fpstate(data, mcontext, &fp_size); + if (!fpstate_stub) + return -EINVAL; + +#ifdef CONFIG_X86_32 + xstate_stub = (void *)&fpstate_stub->_fxsr_env; + xstate_size = fp_size - offsetof(struct _fpstate_32, _fxsr_env); +#else + xstate_stub = (void *)fpstate_stub; + xstate_size = fp_size; +#endif + + memcpy(xstate_stub, ®s->fp, xstate_size); + +#ifdef __i386__ + /* + * On x86, the GDT entries are updated by arch_set_tls. + */ + + /* Store the i387 legacy FP registers which the host will use */ + if (um_i387_from_fxsr(fpstate_stub, (void *)®s->fp)) + return -EINVAL; +#else + /* + * On x86_64, we need to sync the FS_BASE/GS_BASE registers using the + * arch specific data. + */ + if (data->arch_data.fs_base != regs->gp[FS_BASE / sizeof(unsigned long)]) { + data->arch_data.fs_base = regs->gp[FS_BASE / sizeof(unsigned long)]; + data->arch_data.sync |= STUB_SYNC_FS_BASE; + } + if (data->arch_data.gs_base != regs->gp[GS_BASE / sizeof(unsigned long)]) { + data->arch_data.gs_base = regs->gp[GS_BASE / sizeof(unsigned long)]; + data->arch_data.sync |= STUB_SYNC_GS_BASE; + } +#endif + + return 0; +} diff --git a/arch/x86/um/os-Linux/registers.c b/arch/x86/um/os-Linux/registers.c new file mode 100644 index 000000000..eb1cdadc8 --- /dev/null +++ b/arch/x86/um/os-Linux/registers.c @@ -0,0 +1,109 @@ +/* + * Copyright (C) 2004 PathScale, Inc + * Copyright (C) 2004 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <errno.h> +#include <stdlib.h> +#include <sys/ptrace.h> +#ifdef __i386__ +#include <sys/user.h> +#endif +#include <longjmp.h> +#include <sysdep/ptrace_user.h> +#include <sys/uio.h> +#include <asm/sigcontext.h> +#include <linux/elf.h> +#include <registers.h> +#include <sys/mman.h> + +static unsigned long ptrace_regset; +unsigned long host_fp_size; + +int get_fp_registers(int pid, unsigned long *regs) +{ + struct iovec iov = { + .iov_base = regs, + .iov_len = host_fp_size, + }; + + if (ptrace(PTRACE_GETREGSET, pid, ptrace_regset, &iov) < 0) + return -errno; + return 0; +} + +int put_fp_registers(int pid, unsigned long *regs) +{ + struct iovec iov = { + .iov_base = regs, + .iov_len = host_fp_size, + }; + + if (ptrace(PTRACE_SETREGSET, pid, ptrace_regset, &iov) < 0) + return -errno; + return 0; +} + +int arch_init_registers(int pid) +{ + struct iovec iov = { + /* Just use plenty of space, it does not cost us anything */ + .iov_len = 2 * 1024 * 1024, + }; + int ret; + + iov.iov_base = mmap(NULL, iov.iov_len, PROT_WRITE | PROT_READ, + MAP_ANONYMOUS | MAP_PRIVATE, -1, 0); + if (iov.iov_base == MAP_FAILED) + return -ENOMEM; + + /* GDB has x86_xsave_length, which uses x86_cpuid_count */ + ptrace_regset = NT_X86_XSTATE; + ret = ptrace(PTRACE_GETREGSET, pid, ptrace_regset, &iov); + if (ret) + ret = -errno; + + if (ret == -ENODEV) { +#ifdef CONFIG_X86_32 + ptrace_regset = NT_PRXFPREG; +#else + ptrace_regset = NT_PRFPREG; +#endif + iov.iov_len = 2 * 1024 * 1024; + ret = ptrace(PTRACE_GETREGSET, pid, ptrace_regset, &iov); + if (ret) + ret = -errno; + } + + munmap(iov.iov_base, 2 * 1024 * 1024); + + host_fp_size = iov.iov_len; + + return ret; +} + +unsigned long get_thread_reg(int reg, jmp_buf *buf) +{ + switch (reg) { +#ifdef __i386__ + case HOST_IP: + return buf[0]->__eip; + case HOST_SP: + return buf[0]->__esp; + case HOST_BP: + return buf[0]->__ebp; +#else + case HOST_IP: + return buf[0]->__rip; + case HOST_SP: + return buf[0]->__rsp; + case HOST_BP: + return buf[0]->__rbp; +#endif + default: + printk(UM_KERN_ERR "get_thread_regs - unknown register %d\n", + reg); + return 0; + } +} diff --git a/arch/x86/um/os-Linux/tls.c b/arch/x86/um/os-Linux/tls.c new file mode 100644 index 000000000..eed9efe29 --- /dev/null +++ b/arch/x86/um/os-Linux/tls.c @@ -0,0 +1,69 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <errno.h> +#include <linux/unistd.h> + +#include <sys/ptrace.h> +#include <sys/syscall.h> +#include <unistd.h> + +#include <os.h> +#include <sysdep/tls.h> + +#ifndef PTRACE_GET_THREAD_AREA +#define PTRACE_GET_THREAD_AREA 25 +#endif + +#ifndef PTRACE_SET_THREAD_AREA +#define PTRACE_SET_THREAD_AREA 26 +#endif + +/* Checks whether host supports TLS, and sets *tls_min according to the value + * valid on the host. + * i386 host have it == 6; x86_64 host have it == 12, for i386 emulation. */ +void check_host_supports_tls(int *supports_tls, int *tls_min) +{ + /* Values for x86 and x86_64.*/ + int val[] = {GDT_ENTRY_TLS_MIN_I386, GDT_ENTRY_TLS_MIN_X86_64}; + int i; + + for (i = 0; i < ARRAY_SIZE(val); i++) { + user_desc_t info; + info.entry_number = val[i]; + + if (syscall(__NR_get_thread_area, &info) == 0) { + *tls_min = val[i]; + *supports_tls = 1; + return; + } else { + if (errno == EINVAL) + continue; + else if (errno == ENOSYS) + *supports_tls = 0; + return; + } + } + + *supports_tls = 0; +} + +int os_set_thread_area(user_desc_t *info, int pid) +{ + int ret; + + ret = ptrace(PTRACE_SET_THREAD_AREA, pid, info->entry_number, + (unsigned long) info); + if (ret < 0) + ret = -errno; + return ret; +} + +int os_get_thread_area(user_desc_t *info, int pid) +{ + int ret; + + ret = ptrace(PTRACE_GET_THREAD_AREA, pid, info->entry_number, + (unsigned long) info); + if (ret < 0) + ret = -errno; + return ret; +} diff --git a/arch/x86/um/ptrace.c b/arch/x86/um/ptrace.c new file mode 100644 index 000000000..2635ca259 --- /dev/null +++ b/arch/x86/um/ptrace.c @@ -0,0 +1,305 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include <linux/sched.h> +#include <linux/elf.h> +#include <linux/regset.h> +#include <asm/user32.h> +#include <asm/sigcontext.h> + +#ifdef CONFIG_X86_32 +/* + * FPU tag word conversions. + */ + +static inline unsigned short twd_i387_to_fxsr(unsigned short twd) +{ + unsigned int tmp; /* to avoid 16 bit prefixes in the code */ + + /* Transform each pair of bits into 01 (valid) or 00 (empty) */ + tmp = ~twd; + tmp = (tmp | (tmp>>1)) & 0x5555; /* 0V0V0V0V0V0V0V0V */ + /* and move the valid bits to the lower byte. */ + tmp = (tmp | (tmp >> 1)) & 0x3333; /* 00VV00VV00VV00VV */ + tmp = (tmp | (tmp >> 2)) & 0x0f0f; /* 0000VVVV0000VVVV */ + tmp = (tmp | (tmp >> 4)) & 0x00ff; /* 00000000VVVVVVVV */ + return tmp; +} + +static inline unsigned long +twd_fxsr_to_i387(const struct user_fxsr_struct *fxsave) +{ + struct _fpxreg *st = NULL; + unsigned long twd = (unsigned long) fxsave->twd; + unsigned long tag; + unsigned long ret = 0xffff0000; + int i; + +#define FPREG_ADDR(f, n) ((char *)&(f)->st_space + (n) * 16) + + for (i = 0; i < 8; i++) { + if (twd & 0x1) { + st = (struct _fpxreg *) FPREG_ADDR(fxsave, i); + + switch (st->exponent & 0x7fff) { + case 0x7fff: + tag = 2; /* Special */ + break; + case 0x0000: + if (!st->significand[0] && + !st->significand[1] && + !st->significand[2] && + !st->significand[3]) { + tag = 1; /* Zero */ + } else { + tag = 2; /* Special */ + } + break; + default: + if (st->significand[3] & 0x8000) + tag = 0; /* Valid */ + else + tag = 2; /* Special */ + break; + } + } else { + tag = 3; /* Empty */ + } + ret |= (tag << (2 * i)); + twd = twd >> 1; + } + return ret; +} + +/* + * Get/set the old 32bit i387 registers (pre-FPX) + * + * We provide simple wrappers for mcontext.c, they are only defined locally + * because mcontext.c is userspace facing and needs to a different definition + * of the structures. + */ +static int _um_i387_from_fxsr(struct membuf to, + const struct user_fxsr_struct *fxsave) +{ + int i; + + membuf_store(&to, (unsigned long)fxsave->cwd | 0xffff0000ul); + membuf_store(&to, (unsigned long)fxsave->swd | 0xffff0000ul); + membuf_store(&to, twd_fxsr_to_i387(fxsave)); + membuf_store(&to, fxsave->fip); + membuf_store(&to, fxsave->fcs | ((unsigned long)fxsave->fop << 16)); + membuf_store(&to, fxsave->foo); + membuf_store(&to, fxsave->fos); + + for (i = 0; i < 8; i++) + membuf_write(&to, (void *)fxsave->st_space + i * 16, 10); + + return 0; +} + +int um_i387_from_fxsr(struct user_i387_struct *i387, + const struct user_fxsr_struct *fxsave); + +int um_i387_from_fxsr(struct user_i387_struct *i387, + const struct user_fxsr_struct *fxsave) +{ + struct membuf to = { + .p = i387, + .left = sizeof(*i387), + }; + + return _um_i387_from_fxsr(to, fxsave); +} + +static int fpregs_legacy_get(struct task_struct *target, + const struct user_regset *regset, + struct membuf to) +{ + struct user_fxsr_struct *fxsave = (void *)target->thread.regs.regs.fp; + + return _um_i387_from_fxsr(to, fxsave); +} + +int um_fxsr_from_i387(struct user_fxsr_struct *fxsave, + const struct user_i387_struct *from); + +int um_fxsr_from_i387(struct user_fxsr_struct *fxsave, + const struct user_i387_struct *from) +{ + int i; + + fxsave->cwd = (unsigned short)(from->cwd & 0xffff); + fxsave->swd = (unsigned short)(from->swd & 0xffff); + fxsave->twd = twd_i387_to_fxsr((unsigned short)(from->twd & 0xffff)); + fxsave->fip = from->fip; + fxsave->fop = (unsigned short)((from->fcs & 0xffff0000ul) >> 16); + fxsave->fcs = (from->fcs & 0xffff); + fxsave->foo = from->foo; + fxsave->fos = from->fos; + + for (i = 0; i < 8; i++) { + memcpy((void *)fxsave->st_space + i * 16, + (void *)from->st_space + i * 10, 10); + } + + return 0; +} + +static int fpregs_legacy_set(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) +{ + struct user_fxsr_struct *fxsave = (void *)target->thread.regs.regs.fp; + const struct user_i387_struct *from; + struct user_i387_struct buf; + + if (ubuf) { + if (copy_from_user(&buf, ubuf, sizeof(buf))) + return -EFAULT; + from = &buf; + } else { + from = kbuf; + } + + return um_fxsr_from_i387(fxsave, from); +} +#endif + +static int genregs_get(struct task_struct *target, + const struct user_regset *regset, + struct membuf to) +{ + int reg; + + for (reg = 0; to.left; reg++) + membuf_store(&to, getreg(target, reg * sizeof(unsigned long))); + return 0; +} + +static int genregs_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 = 0; + + if (kbuf) { + const unsigned long *k = kbuf; + + while (count >= sizeof(*k) && !ret) { + ret = putreg(target, pos, *k++); + count -= sizeof(*k); + pos += sizeof(*k); + } + } else { + const unsigned long __user *u = ubuf; + + while (count >= sizeof(*u) && !ret) { + unsigned long word; + + ret = __get_user(word, u++); + if (ret) + break; + ret = putreg(target, pos, word); + count -= sizeof(*u); + pos += sizeof(*u); + } + } + return ret; +} + +static int generic_fpregs_active(struct task_struct *target, const struct user_regset *regset) +{ + return regset->n; +} + +static int generic_fpregs_get(struct task_struct *target, + const struct user_regset *regset, + struct membuf to) +{ + void *fpregs = task_pt_regs(target)->regs.fp; + + membuf_write(&to, fpregs, regset->size * regset->n); + return 0; +} + +static int generic_fpregs_set(struct task_struct *target, + const struct user_regset *regset, + unsigned int pos, unsigned int count, + const void *kbuf, const void __user *ubuf) +{ + void *fpregs = task_pt_regs(target)->regs.fp; + + return user_regset_copyin(&pos, &count, &kbuf, &ubuf, + fpregs, 0, regset->size * regset->n); +} + +static struct user_regset uml_regsets[] __ro_after_init = { + [REGSET_GENERAL] = { + USER_REGSET_NOTE_TYPE(PRSTATUS), + .n = sizeof(struct user_regs_struct) / sizeof(long), + .size = sizeof(long), + .align = sizeof(long), + .regset_get = genregs_get, + .set = genregs_set + }, +#ifdef CONFIG_X86_32 + /* Old FP registers, they are needed in signal frames */ + [REGSET_FP_LEGACY] = { + USER_REGSET_NOTE_TYPE(PRFPREG), + .n = sizeof(struct user_i387_ia32_struct) / sizeof(long), + .size = sizeof(long), + .align = sizeof(long), + .active = generic_fpregs_active, + .regset_get = fpregs_legacy_get, + .set = fpregs_legacy_set, + }, +#endif + [REGSET_FP] = { +#ifdef CONFIG_X86_32 + USER_REGSET_NOTE_TYPE(PRXFPREG), + .n = sizeof(struct user32_fxsr_struct) / sizeof(long), +#else + USER_REGSET_NOTE_TYPE(PRFPREG), + .n = sizeof(struct user_i387_struct) / sizeof(long), +#endif + .size = sizeof(long), + .align = sizeof(long), + .active = generic_fpregs_active, + .regset_get = generic_fpregs_get, + .set = generic_fpregs_set, + }, + [REGSET_XSTATE] = { + USER_REGSET_NOTE_TYPE(X86_XSTATE), + .size = sizeof(long), + .align = sizeof(long), + .active = generic_fpregs_active, + .regset_get = generic_fpregs_get, + .set = generic_fpregs_set, + }, + /* TODO: Add TLS regset for 32bit */ +}; + +static const struct user_regset_view user_uml_view = { +#ifdef CONFIG_X86_32 + .name = "i386", .e_machine = EM_386, +#else + .name = "x86_64", .e_machine = EM_X86_64, +#endif + .regsets = uml_regsets, .n = ARRAY_SIZE(uml_regsets) +}; + +const struct user_regset_view * +task_user_regset_view(struct task_struct *tsk) +{ + return &user_uml_view; +} + +static int __init init_regset_xstate_info(void) +{ + uml_regsets[REGSET_XSTATE].n = + host_fp_size / uml_regsets[REGSET_XSTATE].size; + + return 0; +} +arch_initcall(init_regset_xstate_info); diff --git a/arch/x86/um/ptrace_32.c b/arch/x86/um/ptrace_32.c new file mode 100644 index 000000000..3af3cb821 --- /dev/null +++ b/arch/x86/um/ptrace_32.c @@ -0,0 +1,202 @@ +/* + * Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <linux/mm.h> +#include <linux/sched.h> +#include <linux/uaccess.h> +#include <linux/regset.h> +#include <asm/ptrace-abi.h> +#include <registers.h> +#include <skas.h> + +void arch_switch_to(struct task_struct *to) +{ + int err = arch_switch_tls(to); + if (!err) + return; + + if (err != -EINVAL) + printk(KERN_WARNING "arch_switch_tls failed, errno %d, " + "not EINVAL\n", -err); + else + printk(KERN_WARNING "arch_switch_tls failed, errno = EINVAL\n"); +} + +/* determines which flags the user has access to. */ +/* 1 = access 0 = no access */ +#define FLAG_MASK 0x00044dd5 + +static const int reg_offsets[] = { + [EBX] = HOST_BX, + [ECX] = HOST_CX, + [EDX] = HOST_DX, + [ESI] = HOST_SI, + [EDI] = HOST_DI, + [EBP] = HOST_BP, + [EAX] = HOST_AX, + [DS] = HOST_DS, + [ES] = HOST_ES, + [FS] = HOST_FS, + [GS] = HOST_GS, + [EIP] = HOST_IP, + [CS] = HOST_CS, + [EFL] = HOST_EFLAGS, + [UESP] = HOST_SP, + [SS] = HOST_SS, + [ORIG_EAX] = HOST_ORIG_AX, +}; + +int putreg(struct task_struct *child, int regno, unsigned long value) +{ + regno >>= 2; + switch (regno) { + case EBX: + case ECX: + case EDX: + case ESI: + case EDI: + case EBP: + case EAX: + case EIP: + case UESP: + break; + case ORIG_EAX: + /* Update the syscall number. */ + UPT_SYSCALL_NR(&child->thread.regs.regs) = value; + break; + case FS: + if (value && (value & 3) != 3) + return -EIO; + break; + case GS: + if (value && (value & 3) != 3) + return -EIO; + break; + case DS: + case ES: + if (value && (value & 3) != 3) + return -EIO; + value &= 0xffff; + break; + case SS: + case CS: + if ((value & 3) != 3) + return -EIO; + value &= 0xffff; + break; + case EFL: + value &= FLAG_MASK; + child->thread.regs.regs.gp[HOST_EFLAGS] |= value; + return 0; + default : + panic("Bad register in putreg() : %d\n", regno); + } + child->thread.regs.regs.gp[reg_offsets[regno]] = value; + return 0; +} + +int poke_user(struct task_struct *child, long addr, long data) +{ + if ((addr & 3) || addr < 0) + return -EIO; + + if (addr < MAX_REG_OFFSET) + return putreg(child, addr, data); + else if ((addr >= offsetof(struct user, u_debugreg[0])) && + (addr <= offsetof(struct user, u_debugreg[7]))) { + addr -= offsetof(struct user, u_debugreg[0]); + addr = addr >> 2; + if ((addr == 4) || (addr == 5)) + return -EIO; + child->thread.arch.debugregs[addr] = data; + return 0; + } + return -EIO; +} + +unsigned long getreg(struct task_struct *child, int regno) +{ + unsigned long mask = ~0UL; + + regno >>= 2; + switch (regno) { + case FS: + case GS: + case DS: + case ES: + case SS: + case CS: + mask = 0xffff; + break; + case EIP: + case UESP: + case EAX: + case EBX: + case ECX: + case EDX: + case ESI: + case EDI: + case EBP: + case EFL: + case ORIG_EAX: + break; + default: + panic("Bad register in getreg() : %d\n", regno); + } + return mask & child->thread.regs.regs.gp[reg_offsets[regno]]; +} + +/* read the word at location addr in the USER area. */ +int peek_user(struct task_struct *child, long addr, long data) +{ + unsigned long tmp; + + if ((addr & 3) || addr < 0) + return -EIO; + + tmp = 0; /* Default return condition */ + if (addr < MAX_REG_OFFSET) { + tmp = getreg(child, addr); + } + else if ((addr >= offsetof(struct user, u_debugreg[0])) && + (addr <= offsetof(struct user, u_debugreg[7]))) { + addr -= offsetof(struct user, u_debugreg[0]); + addr = addr >> 2; + tmp = child->thread.arch.debugregs[addr]; + } + return put_user(tmp, (unsigned long __user *) data); +} + +long subarch_ptrace(struct task_struct *child, long request, + unsigned long addr, unsigned long data) +{ + int ret = -EIO; + void __user *datap = (void __user *) data; + switch (request) { + case PTRACE_GETFPREGS: /* Get the child FPU state. */ + return copy_regset_to_user(child, task_user_regset_view(child), + REGSET_FP_LEGACY, + 0, sizeof(struct user_i387_struct), + datap); + case PTRACE_SETFPREGS: /* Set the child FPU state. */ + return copy_regset_from_user(child, task_user_regset_view(child), + REGSET_FP_LEGACY, + 0, sizeof(struct user_i387_struct), + datap); + case PTRACE_GETFPXREGS: /* Get the child FPU state. */ + return copy_regset_to_user(child, task_user_regset_view(child), + REGSET_FP, + 0, sizeof(struct user_fxsr_struct), + datap); + case PTRACE_SETFPXREGS: /* Set the child FPU state. */ + return copy_regset_from_user(child, task_user_regset_view(child), + REGSET_FP, + 0, sizeof(struct user_fxsr_struct), + datap); + default: + ret = -EIO; + } + return ret; +} diff --git a/arch/x86/um/ptrace_64.c b/arch/x86/um/ptrace_64.c new file mode 100644 index 000000000..e0d4120a4 --- /dev/null +++ b/arch/x86/um/ptrace_64.c @@ -0,0 +1,216 @@ +/* + * Copyright 2003 PathScale, Inc. + * Copyright (C) 2003 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * + * Licensed under the GPL + */ + +#include <linux/mm.h> +#include <linux/sched.h> +#include <linux/errno.h> +#include <linux/regset.h> +#define __FRAME_OFFSETS +#include <asm/ptrace.h> +#include <linux/uaccess.h> +#include <registers.h> +#include <asm/ptrace-abi.h> + +/* + * determines which flags the user has access to. + * 1 = access 0 = no access + */ +#define FLAG_MASK 0x44dd5UL + +static const int reg_offsets[] = +{ + [R8 >> 3] = HOST_R8, + [R9 >> 3] = HOST_R9, + [R10 >> 3] = HOST_R10, + [R11 >> 3] = HOST_R11, + [R12 >> 3] = HOST_R12, + [R13 >> 3] = HOST_R13, + [R14 >> 3] = HOST_R14, + [R15 >> 3] = HOST_R15, + [RIP >> 3] = HOST_IP, + [RSP >> 3] = HOST_SP, + [RAX >> 3] = HOST_AX, + [RBX >> 3] = HOST_BX, + [RCX >> 3] = HOST_CX, + [RDX >> 3] = HOST_DX, + [RSI >> 3] = HOST_SI, + [RDI >> 3] = HOST_DI, + [RBP >> 3] = HOST_BP, + [CS >> 3] = HOST_CS, + [SS >> 3] = HOST_SS, + [FS_BASE >> 3] = HOST_FS_BASE, + [GS_BASE >> 3] = HOST_GS_BASE, + [DS >> 3] = HOST_DS, + [ES >> 3] = HOST_ES, + [FS >> 3] = HOST_FS, + [GS >> 3] = HOST_GS, + [EFLAGS >> 3] = HOST_EFLAGS, + [ORIG_RAX >> 3] = HOST_ORIG_AX, +}; + +int putreg(struct task_struct *child, int regno, unsigned long value) +{ + switch (regno) { + case R8: + case R9: + case R10: + case R11: + case R12: + case R13: + case R14: + case R15: + case RIP: + case RSP: + case RAX: + case RBX: + case RCX: + case RDX: + case RSI: + case RDI: + case RBP: + break; + + case ORIG_RAX: + /* Update the syscall number. */ + UPT_SYSCALL_NR(&child->thread.regs.regs) = value; + break; + + case FS: + case GS: + case DS: + case ES: + case SS: + case CS: + if (value && (value & 3) != 3) + return -EIO; + value &= 0xffff; + break; + + case FS_BASE: + case GS_BASE: + if (!((value >> 48) == 0 || (value >> 48) == 0xffff)) + return -EIO; + break; + + case EFLAGS: + value &= FLAG_MASK; + child->thread.regs.regs.gp[HOST_EFLAGS] |= value; + return 0; + + default: + panic("Bad register in putreg(): %d\n", regno); + } + + child->thread.regs.regs.gp[reg_offsets[regno >> 3]] = value; + return 0; +} + +int poke_user(struct task_struct *child, long addr, long data) +{ + if ((addr & 3) || addr < 0) + return -EIO; + + if (addr < MAX_REG_OFFSET) + return putreg(child, addr, data); + else if ((addr >= offsetof(struct user, u_debugreg[0])) && + (addr <= offsetof(struct user, u_debugreg[7]))) { + addr -= offsetof(struct user, u_debugreg[0]); + addr = addr >> 3; + if ((addr == 4) || (addr == 5)) + return -EIO; + child->thread.arch.debugregs[addr] = data; + return 0; + } + return -EIO; +} + +unsigned long getreg(struct task_struct *child, int regno) +{ + unsigned long mask = ~0UL; + + switch (regno) { + case R8: + case R9: + case R10: + case R11: + case R12: + case R13: + case R14: + case R15: + case RIP: + case RSP: + case RAX: + case RBX: + case RCX: + case RDX: + case RSI: + case RDI: + case RBP: + case ORIG_RAX: + case EFLAGS: + case FS_BASE: + case GS_BASE: + break; + case FS: + case GS: + case DS: + case ES: + case SS: + case CS: + mask = 0xffff; + break; + default: + panic("Bad register in getreg: %d\n", regno); + } + return mask & child->thread.regs.regs.gp[reg_offsets[regno >> 3]]; +} + +int peek_user(struct task_struct *child, long addr, long data) +{ + /* read the word at location addr in the USER area. */ + unsigned long tmp; + + if ((addr & 3) || addr < 0) + return -EIO; + + tmp = 0; /* Default return condition */ + if (addr < MAX_REG_OFFSET) + tmp = getreg(child, addr); + else if ((addr >= offsetof(struct user, u_debugreg[0])) && + (addr <= offsetof(struct user, u_debugreg[7]))) { + addr -= offsetof(struct user, u_debugreg[0]); + addr = addr >> 2; + tmp = child->thread.arch.debugregs[addr]; + } + return put_user(tmp, (unsigned long *) data); +} + +long subarch_ptrace(struct task_struct *child, long request, + unsigned long addr, unsigned long data) +{ + int ret = -EIO; + void __user *datap = (void __user *) data; + + switch (request) { + case PTRACE_GETFPREGS: /* Get the child FPU state. */ + return copy_regset_to_user(child, task_user_regset_view(child), + REGSET_FP, + 0, sizeof(struct user_i387_struct), + datap); + case PTRACE_SETFPREGS: /* Set the child FPU state. */ + return copy_regset_from_user(child, task_user_regset_view(child), + REGSET_FP, + 0, sizeof(struct user_i387_struct), + datap); + case PTRACE_ARCH_PRCTL: + /* XXX Calls ptrace on the host - needs some SMP thinking */ + ret = arch_prctl(child, data, (void __user *) addr); + break; + } + + return ret; +} diff --git a/arch/x86/um/ptrace_user.c b/arch/x86/um/ptrace_user.c new file mode 100644 index 000000000..617885b18 --- /dev/null +++ b/arch/x86/um/ptrace_user.c @@ -0,0 +1,21 @@ +/* + * Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <errno.h> +#include <ptrace_user.h> + +int ptrace_getregs(long pid, unsigned long *regs_out) +{ + if (ptrace(PTRACE_GETREGS, pid, 0, regs_out) < 0) + return -errno; + return 0; +} + +int ptrace_setregs(long pid, unsigned long *regs) +{ + if (ptrace(PTRACE_SETREGS, pid, 0, regs) < 0) + return -errno; + return 0; +} diff --git a/arch/x86/um/setjmp_32.S b/arch/x86/um/setjmp_32.S new file mode 100644 index 000000000..2d991ddbc --- /dev/null +++ b/arch/x86/um/setjmp_32.S @@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +# +# arch/i386/setjmp.S +# +# setjmp/longjmp for the i386 architecture +# + +# +# The jmp_buf is assumed to contain the following, in order: +# %ebx +# %esp +# %ebp +# %esi +# %edi +# <return address> +# + + .text + .align 4 + .globl kernel_setjmp + .type kernel_setjmp, @function +kernel_setjmp: +#ifdef _REGPARM + movl %eax,%edx +#else + movl 4(%esp),%edx +#endif + popl %ecx # Return address, and adjust the stack + xorl %eax,%eax # Return value + movl %ebx,(%edx) + movl %esp,4(%edx) # Post-return %esp! + pushl %ecx # Make the call/return stack happy + movl %ebp,8(%edx) + movl %esi,12(%edx) + movl %edi,16(%edx) + movl %ecx,20(%edx) # Return address + RET + + .size kernel_setjmp,.-kernel_setjmp + + .text + .align 4 + .globl kernel_longjmp + .type kernel_longjmp, @function +kernel_longjmp: +#ifdef _REGPARM + xchgl %eax,%edx +#else + movl 4(%esp),%edx # jmp_ptr address + movl 8(%esp),%eax # Return value +#endif + movl (%edx),%ebx + movl 4(%edx),%esp + movl 8(%edx),%ebp + movl 12(%edx),%esi + movl 16(%edx),%edi + jmp *20(%edx) + + .size kernel_longjmp,.-kernel_longjmp diff --git a/arch/x86/um/setjmp_64.S b/arch/x86/um/setjmp_64.S new file mode 100644 index 000000000..b46acb6a8 --- /dev/null +++ b/arch/x86/um/setjmp_64.S @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +# +# arch/x86_64/setjmp.S +# +# setjmp/longjmp for the x86-64 architecture +# + +# +# The jmp_buf is assumed to contain the following, in order: +# %rbx +# %rsp (post-return) +# %rbp +# %r12 +# %r13 +# %r14 +# %r15 +# <return address> +# + + .text + .align 4 + .globl kernel_setjmp + .type kernel_setjmp, @function +kernel_setjmp: + pop %rsi # Return address, and adjust the stack + xorl %eax,%eax # Return value + movq %rbx,(%rdi) + movq %rsp,8(%rdi) # Post-return %rsp! + push %rsi # Make the call/return stack happy + movq %rbp,16(%rdi) + movq %r12,24(%rdi) + movq %r13,32(%rdi) + movq %r14,40(%rdi) + movq %r15,48(%rdi) + movq %rsi,56(%rdi) # Return address + RET + + .size kernel_setjmp,.-kernel_setjmp + + .text + .align 4 + .globl kernel_longjmp + .type kernel_longjmp, @function +kernel_longjmp: + movl %esi,%eax # Return value (int) + movq (%rdi),%rbx + movq 8(%rdi),%rsp + movq 16(%rdi),%rbp + movq 24(%rdi),%r12 + movq 32(%rdi),%r13 + movq 40(%rdi),%r14 + movq 48(%rdi),%r15 + jmp *56(%rdi) + + .size kernel_longjmp,.-kernel_longjmp diff --git a/arch/x86/um/shared/sysdep/archsetjmp.h b/arch/x86/um/shared/sysdep/archsetjmp.h new file mode 100644 index 000000000..8c81d1a60 --- /dev/null +++ b/arch/x86/um/shared/sysdep/archsetjmp.h @@ -0,0 +1,13 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __X86_UM_SYSDEP_ARCHSETJMP_H +#define __X86_UM_SYSDEP_ARCHSETJMP_H + +#ifdef __i386__ +#include "archsetjmp_32.h" +#else +#include "archsetjmp_64.h" +#endif + +unsigned long get_thread_reg(int reg, jmp_buf *buf); + +#endif /* __X86_UM_SYSDEP_ARCHSETJMP_H */ diff --git a/arch/x86/um/shared/sysdep/archsetjmp_32.h b/arch/x86/um/shared/sysdep/archsetjmp_32.h new file mode 100644 index 000000000..fb08f2576 --- /dev/null +++ b/arch/x86/um/shared/sysdep/archsetjmp_32.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * arch/um/include/sysdep-i386/archsetjmp.h + */ + +#ifndef _KLIBC_ARCHSETJMP_H +#define _KLIBC_ARCHSETJMP_H + +struct __jmp_buf { + unsigned int __ebx; + unsigned int __esp; + unsigned int __ebp; + unsigned int __esi; + unsigned int __edi; + unsigned int __eip; +}; + +typedef struct __jmp_buf jmp_buf[1]; + +#define JB_IP __eip +#define JB_SP __esp + +#endif /* _SETJMP_H */ diff --git a/arch/x86/um/shared/sysdep/archsetjmp_64.h b/arch/x86/um/shared/sysdep/archsetjmp_64.h new file mode 100644 index 000000000..9b499e457 --- /dev/null +++ b/arch/x86/um/shared/sysdep/archsetjmp_64.h @@ -0,0 +1,25 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * arch/um/include/sysdep-x86_64/archsetjmp.h + */ + +#ifndef _KLIBC_ARCHSETJMP_H +#define _KLIBC_ARCHSETJMP_H + +struct __jmp_buf { + unsigned long __rbx; + unsigned long __rsp; + unsigned long __rbp; + unsigned long __r12; + unsigned long __r13; + unsigned long __r14; + unsigned long __r15; + unsigned long __rip; +}; + +typedef struct __jmp_buf jmp_buf[1]; + +#define JB_IP __rip +#define JB_SP __rsp + +#endif /* _SETJMP_H */ diff --git a/arch/x86/um/shared/sysdep/faultinfo.h b/arch/x86/um/shared/sysdep/faultinfo.h new file mode 100644 index 000000000..4390803e0 --- /dev/null +++ b/arch/x86/um/shared/sysdep/faultinfo.h @@ -0,0 +1,6 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifdef __i386__ +#include "faultinfo_32.h" +#else +#include "faultinfo_64.h" +#endif diff --git a/arch/x86/um/shared/sysdep/faultinfo_32.h b/arch/x86/um/shared/sysdep/faultinfo_32.h new file mode 100644 index 000000000..9193a7790 --- /dev/null +++ b/arch/x86/um/shared/sysdep/faultinfo_32.h @@ -0,0 +1,44 @@ +/* + * Copyright (C) 2004 Fujitsu Siemens Computers GmbH + * Author: Bodo Stroesser <bstroesser@fujitsu-siemens.com> + * Licensed under the GPL + */ + +#ifndef __FAULTINFO_I386_H +#define __FAULTINFO_I386_H + +/* this structure contains the full arch-specific faultinfo + * from the traps. + * On i386, ptrace_faultinfo unfortunately doesn't provide + * all the info, since trap_no is missing. + * All common elements are defined at the same position in + * both structures, thus making it easy to copy the + * contents without knowledge about the structure elements. + */ +struct faultinfo { + int error_code; /* in ptrace_faultinfo misleadingly called is_write */ + unsigned long cr2; /* in ptrace_faultinfo called addr */ + int trap_no; /* missing in ptrace_faultinfo */ +}; + +#define FAULT_WRITE(fi) ((fi).error_code & 2) +#define FAULT_ADDRESS(fi) ((fi).cr2) + +/* This is Page Fault */ +#define SEGV_IS_FIXABLE(fi) ((fi)->trap_no == 14) + +#define PTRACE_FULL_FAULTINFO 0 + +#define ___backtrack_faulted(_faulted) \ + asm volatile ( \ + "movl $__get_kernel_nofault_faulted_%=,%1\n" \ + "mov $0, %0\n" \ + "jmp _end_%=\n" \ + "__get_kernel_nofault_faulted_%=:\n" \ + "mov $1, %0;" \ + "_end_%=:" \ + : "=r" (_faulted), \ + "=m" (current->thread.segv_continue) :: \ + ) + +#endif diff --git a/arch/x86/um/shared/sysdep/faultinfo_64.h b/arch/x86/um/shared/sysdep/faultinfo_64.h new file mode 100644 index 000000000..61e4ca1e0 --- /dev/null +++ b/arch/x86/um/shared/sysdep/faultinfo_64.h @@ -0,0 +1,44 @@ +/* + * Copyright (C) 2004 Fujitsu Siemens Computers GmbH + * Author: Bodo Stroesser <bstroesser@fujitsu-siemens.com> + * Licensed under the GPL + */ + +#ifndef __FAULTINFO_X86_64_H +#define __FAULTINFO_X86_64_H + +/* this structure contains the full arch-specific faultinfo + * from the traps. + * On i386, ptrace_faultinfo unfortunately doesn't provide + * all the info, since trap_no is missing. + * All common elements are defined at the same position in + * both structures, thus making it easy to copy the + * contents without knowledge about the structure elements. + */ +struct faultinfo { + int error_code; /* in ptrace_faultinfo misleadingly called is_write */ + unsigned long cr2; /* in ptrace_faultinfo called addr */ + int trap_no; /* missing in ptrace_faultinfo */ +}; + +#define FAULT_WRITE(fi) ((fi).error_code & 2) +#define FAULT_ADDRESS(fi) ((fi).cr2) + +/* This is Page Fault */ +#define SEGV_IS_FIXABLE(fi) ((fi)->trap_no == 14) + +#define PTRACE_FULL_FAULTINFO 1 + +#define ___backtrack_faulted(_faulted) \ + asm volatile ( \ + "movq $__get_kernel_nofault_faulted_%=,%1\n" \ + "mov $0, %0\n" \ + "jmp _end_%=\n" \ + "__get_kernel_nofault_faulted_%=:\n" \ + "mov $1, %0;" \ + "_end_%=:" \ + : "=r" (_faulted), \ + "=m" (current->thread.segv_continue) :: \ + ) + +#endif diff --git a/arch/x86/um/shared/sysdep/mcontext.h b/arch/x86/um/shared/sysdep/mcontext.h new file mode 100644 index 000000000..6fe490cc5 --- /dev/null +++ b/arch/x86/um/shared/sysdep/mcontext.h @@ -0,0 +1,40 @@ +/* + * Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#ifndef __SYS_SIGCONTEXT_X86_H +#define __SYS_SIGCONTEXT_X86_H + +#include <stub-data.h> + +extern void get_regs_from_mc(struct uml_pt_regs *, mcontext_t *); +extern void get_mc_from_regs(struct uml_pt_regs *regs, mcontext_t *mc, + int single_stepping); + +extern int get_stub_state(struct uml_pt_regs *regs, struct stub_data *data, + unsigned long *fp_size_out); +extern int set_stub_state(struct uml_pt_regs *regs, struct stub_data *data, + int single_stepping); + +#ifdef __i386__ + +#define GET_FAULTINFO_FROM_MC(fi, mc) \ + { \ + (fi).cr2 = (mc)->cr2; \ + (fi).error_code = (mc)->gregs[REG_ERR]; \ + (fi).trap_no = (mc)->gregs[REG_TRAPNO]; \ + } + +#else + +#define GET_FAULTINFO_FROM_MC(fi, mc) \ + { \ + (fi).cr2 = (mc)->gregs[REG_CR2]; \ + (fi).error_code = (mc)->gregs[REG_ERR]; \ + (fi).trap_no = (mc)->gregs[REG_TRAPNO]; \ + } + +#endif + +#endif diff --git a/arch/x86/um/shared/sysdep/ptrace.h b/arch/x86/um/shared/sysdep/ptrace.h new file mode 100644 index 000000000..572ea2d79 --- /dev/null +++ b/arch/x86/um/shared/sysdep/ptrace.h @@ -0,0 +1,67 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __SYSDEP_X86_PTRACE_H +#define __SYSDEP_X86_PTRACE_H + +#include <generated/user_constants.h> +#include <sysdep/faultinfo.h> + +#define MAX_REG_OFFSET (UM_FRAME_SIZE) +#define MAX_REG_NR ((MAX_REG_OFFSET) / sizeof(unsigned long)) + +#define REGS_IP(r) ((r)[HOST_IP]) +#define REGS_SP(r) ((r)[HOST_SP]) +#define REGS_EFLAGS(r) ((r)[HOST_EFLAGS]) +#define REGS_AX(r) ((r)[HOST_AX]) +#define REGS_BX(r) ((r)[HOST_BX]) +#define REGS_CX(r) ((r)[HOST_CX]) +#define REGS_DX(r) ((r)[HOST_DX]) +#define REGS_SI(r) ((r)[HOST_SI]) +#define REGS_DI(r) ((r)[HOST_DI]) +#define REGS_BP(r) ((r)[HOST_BP]) +#define REGS_CS(r) ((r)[HOST_CS]) +#define REGS_SS(r) ((r)[HOST_SS]) +#define REGS_DS(r) ((r)[HOST_DS]) +#define REGS_ES(r) ((r)[HOST_ES]) + +#define UPT_IP(r) REGS_IP((r)->gp) +#define UPT_SP(r) REGS_SP((r)->gp) +#define UPT_EFLAGS(r) REGS_EFLAGS((r)->gp) +#define UPT_AX(r) REGS_AX((r)->gp) +#define UPT_BX(r) REGS_BX((r)->gp) +#define UPT_CX(r) REGS_CX((r)->gp) +#define UPT_DX(r) REGS_DX((r)->gp) +#define UPT_SI(r) REGS_SI((r)->gp) +#define UPT_DI(r) REGS_DI((r)->gp) +#define UPT_BP(r) REGS_BP((r)->gp) +#define UPT_CS(r) REGS_CS((r)->gp) +#define UPT_SS(r) REGS_SS((r)->gp) +#define UPT_DS(r) REGS_DS((r)->gp) +#define UPT_ES(r) REGS_ES((r)->gp) + +#ifdef __i386__ +#include "ptrace_32.h" +#else +#include "ptrace_64.h" +#endif + +extern unsigned long host_fp_size; + +struct uml_pt_regs { + unsigned long gp[MAX_REG_NR]; + struct faultinfo faultinfo; + long syscall; + int is_user; + + /* Dynamically sized FP registers (holds an XSTATE) */ + unsigned long fp[]; +}; + +#define EMPTY_UML_PT_REGS { } + +#define UPT_SYSCALL_NR(r) ((r)->syscall) +#define UPT_FAULTINFO(r) (&(r)->faultinfo) +#define UPT_IS_USER(r) ((r)->is_user) + +extern int arch_init_registers(int pid); + +#endif /* __SYSDEP_X86_PTRACE_H */ diff --git a/arch/x86/um/shared/sysdep/ptrace_32.h b/arch/x86/um/shared/sysdep/ptrace_32.h new file mode 100644 index 000000000..2392470ca --- /dev/null +++ b/arch/x86/um/shared/sysdep/ptrace_32.h @@ -0,0 +1,16 @@ +/* + * Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#ifndef __SYSDEP_I386_PTRACE_H +#define __SYSDEP_I386_PTRACE_H + +#define UPT_SYSCALL_ARG1(r) UPT_BX(r) +#define UPT_SYSCALL_ARG2(r) UPT_CX(r) +#define UPT_SYSCALL_ARG3(r) UPT_DX(r) +#define UPT_SYSCALL_ARG4(r) UPT_SI(r) +#define UPT_SYSCALL_ARG5(r) UPT_DI(r) +#define UPT_SYSCALL_ARG6(r) UPT_BP(r) + +#endif diff --git a/arch/x86/um/shared/sysdep/ptrace_64.h b/arch/x86/um/shared/sysdep/ptrace_64.h new file mode 100644 index 000000000..e73573ac8 --- /dev/null +++ b/arch/x86/um/shared/sysdep/ptrace_64.h @@ -0,0 +1,58 @@ +/* + * Copyright 2003 PathScale, Inc. + * Copyright (C) 2003 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * + * Licensed under the GPL + */ + +#ifndef __SYSDEP_X86_64_PTRACE_H +#define __SYSDEP_X86_64_PTRACE_H + +#define REGS_R8(r) ((r)[HOST_R8]) +#define REGS_R9(r) ((r)[HOST_R9]) +#define REGS_R10(r) ((r)[HOST_R10]) +#define REGS_R11(r) ((r)[HOST_R11]) +#define REGS_R12(r) ((r)[HOST_R12]) +#define REGS_R13(r) ((r)[HOST_R13]) +#define REGS_R14(r) ((r)[HOST_R14]) +#define REGS_R15(r) ((r)[HOST_R15]) + +#define HOST_FS_BASE 21 +#define HOST_GS_BASE 22 +#define HOST_DS 23 +#define HOST_ES 24 +#define HOST_FS 25 +#define HOST_GS 26 + +/* Also defined in asm/ptrace-x86_64.h, but not in libc headers. So, these + * are already defined for kernel code, but not for userspace code. + */ +#ifndef FS_BASE +/* These aren't defined in ptrace.h, but exist in struct user_regs_struct, + * which is what x86_64 ptrace actually uses. + */ +#define FS_BASE (HOST_FS_BASE * sizeof(long)) +#define GS_BASE (HOST_GS_BASE * sizeof(long)) +#define DS (HOST_DS * sizeof(long)) +#define ES (HOST_ES * sizeof(long)) +#define FS (HOST_FS * sizeof(long)) +#define GS (HOST_GS * sizeof(long)) +#endif + +#define UPT_R8(r) REGS_R8((r)->gp) +#define UPT_R9(r) REGS_R9((r)->gp) +#define UPT_R10(r) REGS_R10((r)->gp) +#define UPT_R11(r) REGS_R11((r)->gp) +#define UPT_R12(r) REGS_R12((r)->gp) +#define UPT_R13(r) REGS_R13((r)->gp) +#define UPT_R14(r) REGS_R14((r)->gp) +#define UPT_R15(r) REGS_R15((r)->gp) + +#define UPT_SYSCALL_ARG1(r) UPT_DI(r) +#define UPT_SYSCALL_ARG2(r) UPT_SI(r) +#define UPT_SYSCALL_ARG3(r) UPT_DX(r) +#define UPT_SYSCALL_ARG4(r) UPT_R10(r) +#define UPT_SYSCALL_ARG5(r) UPT_R8(r) +#define UPT_SYSCALL_ARG6(r) UPT_R9(r) + +#endif diff --git a/arch/x86/um/shared/sysdep/ptrace_user.h b/arch/x86/um/shared/sysdep/ptrace_user.h new file mode 100644 index 000000000..98da23120 --- /dev/null +++ b/arch/x86/um/shared/sysdep/ptrace_user.h @@ -0,0 +1,20 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include <generated/user_constants.h> + +#define PT_OFFSET(r) ((r) * sizeof(long)) + +#define PT_SYSCALL_NR(regs) ((regs)[HOST_ORIG_AX]) +#define PT_SYSCALL_NR_OFFSET PT_OFFSET(HOST_ORIG_AX) + +#define PT_SYSCALL_RET_OFFSET PT_OFFSET(HOST_AX) + +#define REGS_IP_INDEX HOST_IP +#define REGS_SP_INDEX HOST_SP + +/* + * glibc before 2.27 does not include PTRACE_SYSEMU_SINGLESTEP in its enum, + * ensure we have a definition by (re-)defining it here. + */ +#ifndef PTRACE_SYSEMU_SINGLESTEP +#define PTRACE_SYSEMU_SINGLESTEP 32 +#endif diff --git a/arch/x86/um/shared/sysdep/stub-data.h b/arch/x86/um/shared/sysdep/stub-data.h new file mode 100644 index 000000000..82b1b7f8a --- /dev/null +++ b/arch/x86/um/shared/sysdep/stub-data.h @@ -0,0 +1,23 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __ARCH_STUB_DATA_H +#define __ARCH_STUB_DATA_H + +#ifdef __i386__ +#include <generated/asm-offsets.h> +#include <asm/ldt.h> + +struct stub_data_arch { + int sync; + struct user_desc tls[UM_KERN_GDT_ENTRY_TLS_ENTRIES]; +}; +#else +#define STUB_SYNC_FS_BASE (1 << 0) +#define STUB_SYNC_GS_BASE (1 << 1) +struct stub_data_arch { + int sync; + unsigned long fs_base; + unsigned long gs_base; +}; +#endif + +#endif /* __ARCH_STUB_DATA_H */ diff --git a/arch/x86/um/shared/sysdep/stub.h b/arch/x86/um/shared/sysdep/stub.h new file mode 100644 index 000000000..4fa58f5b4 --- /dev/null +++ b/arch/x86/um/shared/sysdep/stub.h @@ -0,0 +1,17 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include <asm/unistd.h> +#include <sys/mman.h> +#include <signal.h> +#include <as-layout.h> +#include <stub-data.h> + +#ifdef __i386__ +#include "stub_32.h" +#else +#include "stub_64.h" +#endif + +extern void stub_segv_handler(int, siginfo_t *, void *); +extern void stub_syscall_handler(void); +extern void stub_signal_interrupt(int, siginfo_t *, void *); +extern void stub_signal_restorer(void); diff --git a/arch/x86/um/shared/sysdep/stub_32.h b/arch/x86/um/shared/sysdep/stub_32.h new file mode 100644 index 000000000..9dc2efaf5 --- /dev/null +++ b/arch/x86/um/shared/sysdep/stub_32.h @@ -0,0 +1,147 @@ +/* + * Copyright (C) 2004 Jeff Dike (jdike@addtoit.com) + * Licensed under the GPL + */ + +#ifndef __SYSDEP_STUB_H +#define __SYSDEP_STUB_H + +#include <stddef.h> +#include <asm/ptrace.h> +#include <generated/asm-offsets.h> + +#define STUB_MMAP_NR __NR_mmap2 +#define MMAP_OFFSET(o) ((o) >> UM_KERN_PAGE_SHIFT) + +static __always_inline long stub_syscall0(long syscall) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall1(long syscall, long arg1) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall), "b" (arg1) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall2(long syscall, long arg1, long arg2) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall), "b" (arg1), + "c" (arg2) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall3(long syscall, long arg1, long arg2, + long arg3) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall), "b" (arg1), + "c" (arg2), "d" (arg3) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall4(long syscall, long arg1, long arg2, + long arg3, long arg4) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall), "b" (arg1), + "c" (arg2), "d" (arg3), "S" (arg4) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall5(long syscall, long arg1, long arg2, + long arg3, long arg4, long arg5) +{ + long ret; + + __asm__ volatile ("int $0x80" : "=a" (ret) : "0" (syscall), "b" (arg1), + "c" (arg2), "d" (arg3), "S" (arg4), "D" (arg5) + : "memory"); + + return ret; +} + +static __always_inline long stub_syscall6(long syscall, long arg1, long arg2, + long arg3, long arg4, long arg5, + long arg6) +{ + struct syscall_args { + int ebx, ebp; + } args = { arg1, arg6 }; + long ret; + + __asm__ volatile ("pushl %%ebp;" + "movl 0x4(%%ebx),%%ebp;" + "movl (%%ebx),%%ebx;" + "int $0x80;" + "popl %%ebp" + : "=a" (ret) + : "0" (syscall), "b" (&args), + "c" (arg2), "d" (arg3), "S" (arg4), "D" (arg5) + : "memory"); + + return ret; +} + +static __always_inline void trap_myself(void) +{ + __asm("int3"); +} + +static __always_inline void *get_stub_data(void) +{ + unsigned long ret; + + asm volatile ( + "call _here_%=;" + "_here_%=:" + "popl %0;" + "andl %1, %0 ;" + "addl %2, %0 ;" + : "=a" (ret) + : "g" (~(UM_KERN_PAGE_SIZE - 1)), + "g" (UM_KERN_PAGE_SIZE)); + + return (void *)ret; +} + +#define stub_start(fn) \ + asm volatile ( \ + "subl %0,%%esp ;" \ + "movl %1, %%eax ; " \ + "call *%%eax ;" \ + :: "i" (STUB_SIZE), \ + "i" (&fn)) + +static __always_inline void +stub_seccomp_restore_state(struct stub_data_arch *arch) +{ + for (int i = 0; i < sizeof(arch->tls) / sizeof(arch->tls[0]); i++) { + if (arch->sync & (1 << i)) + stub_syscall1(__NR_set_thread_area, + (unsigned long) &arch->tls[i]); + } + + arch->sync = 0; +} + +#endif diff --git a/arch/x86/um/shared/sysdep/stub_64.h b/arch/x86/um/shared/sysdep/stub_64.h new file mode 100644 index 000000000..9fd56954e --- /dev/null +++ b/arch/x86/um/shared/sysdep/stub_64.h @@ -0,0 +1,154 @@ +/* + * Copyright (C) 2004 Jeff Dike (jdike@addtoit.com) + * Licensed under the GPL + */ + +#ifndef __SYSDEP_STUB_H +#define __SYSDEP_STUB_H + +#include <stddef.h> +#include <sysdep/ptrace_user.h> +#include <generated/asm-offsets.h> +#include <linux/stddef.h> +#include <asm/prctl.h> + +#define STUB_MMAP_NR __NR_mmap +#define MMAP_OFFSET(o) (o) + +#define __syscall_clobber "r11","rcx","memory" +#define __syscall "syscall" + +static __always_inline long stub_syscall0(long syscall) +{ + long ret; + + __asm__ volatile (__syscall + : "=a" (ret) + : "0" (syscall) : __syscall_clobber ); + + return ret; +} + +static __always_inline long stub_syscall1(long syscall, long arg1) +{ + long ret; + + __asm__ volatile (__syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1) : __syscall_clobber ); + + return ret; +} + +static __always_inline long stub_syscall2(long syscall, long arg1, long arg2) +{ + long ret; + + __asm__ volatile (__syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1), "S" (arg2) : __syscall_clobber ); + + return ret; +} + +static __always_inline long stub_syscall3(long syscall, long arg1, long arg2, + long arg3) +{ + long ret; + + __asm__ volatile (__syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1), "S" (arg2), "d" (arg3) + : __syscall_clobber ); + + return ret; +} + +static __always_inline long stub_syscall4(long syscall, long arg1, long arg2, long arg3, + long arg4) +{ + long ret; + + __asm__ volatile ("movq %5,%%r10 ; " __syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1), "S" (arg2), "d" (arg3), + "g" (arg4) + : __syscall_clobber, "r10" ); + + return ret; +} + +static __always_inline long stub_syscall5(long syscall, long arg1, long arg2, + long arg3, long arg4, long arg5) +{ + long ret; + + __asm__ volatile ("movq %5,%%r10 ; movq %6,%%r8 ; " __syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1), "S" (arg2), "d" (arg3), + "g" (arg4), "g" (arg5) + : __syscall_clobber, "r10", "r8" ); + + return ret; +} + +static __always_inline long stub_syscall6(long syscall, long arg1, long arg2, + long arg3, long arg4, long arg5, + long arg6) +{ + long ret; + + __asm__ volatile ("movq %5,%%r10 ; movq %6,%%r8 ; movq %7,%%r9 ; " + __syscall + : "=a" (ret) + : "0" (syscall), "D" (arg1), "S" (arg2), "d" (arg3), + "g" (arg4), "g" (arg5), "g" (arg6) + : __syscall_clobber, "r10", "r8", "r9"); + + return ret; +} + +static __always_inline void trap_myself(void) +{ + __asm("int3"); +} + +static __always_inline void *get_stub_data(void) +{ + unsigned long ret; + + asm volatile ( + "lea 0(%%rip), %0;" + "andq %1, %0 ;" + "addq %2, %0 ;" + : "=a" (ret) + : "g" (~(UM_KERN_PAGE_SIZE - 1)), + "g" (UM_KERN_PAGE_SIZE)); + + return (void *)ret; +} + +#define stub_start(fn) \ + asm volatile ( \ + "subq %0,%%rsp ;" \ + "movq %1,%%rax ;" \ + "call *%%rax ;" \ + :: "i" (STUB_SIZE), \ + "i" (&fn)) + +static __always_inline void +stub_seccomp_restore_state(struct stub_data_arch *arch) +{ + /* + * We could use _writefsbase_u64/_writegsbase_u64 if the host reports + * support in the hwcaps (HWCAP2_FSGSBASE). + */ + if (arch->sync & STUB_SYNC_FS_BASE) + stub_syscall2(__NR_arch_prctl, ARCH_SET_FS, arch->fs_base); + if (arch->sync & STUB_SYNC_GS_BASE) + stub_syscall2(__NR_arch_prctl, ARCH_SET_GS, arch->gs_base); + + arch->sync = 0; +} + +#endif diff --git a/arch/x86/um/shared/sysdep/tls.h b/arch/x86/um/shared/sysdep/tls.h new file mode 100644 index 000000000..b968016aa --- /dev/null +++ b/arch/x86/um/shared/sysdep/tls.h @@ -0,0 +1,40 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _SYSDEP_TLS_H +#define _SYSDEP_TLS_H + +#ifdef __UM_HOST__ + +/* Change name to avoid conflicts with the original one from <asm/ldt.h>, which + * may be named user_desc (but in 2.4 and in header matching its API was named + * modify_ldt_ldt_s). */ + +typedef struct um_dup_user_desc { + unsigned int entry_number; + unsigned int base_addr; + unsigned int limit; + unsigned int seg_32bit:1; + unsigned int contents:2; + unsigned int read_exec_only:1; + unsigned int limit_in_pages:1; + unsigned int seg_not_present:1; + unsigned int useable:1; +#ifdef __x86_64__ + unsigned int lm:1; +#endif +} user_desc_t; + +#else /* __UM_HOST__ */ + +typedef struct user_desc user_desc_t; + +#endif /* __UM_HOST__ */ + +extern int os_set_thread_area(user_desc_t *info, int pid); +extern int os_get_thread_area(user_desc_t *info, int pid); + +#ifdef __i386__ +#define GDT_ENTRY_TLS_MIN_I386 6 +#define GDT_ENTRY_TLS_MIN_X86_64 12 +#endif + +#endif /* _SYSDEP_TLS_H */ diff --git a/arch/x86/um/signal.c b/arch/x86/um/signal.c new file mode 100644 index 000000000..2934e170b --- /dev/null +++ b/arch/x86/um/signal.c @@ -0,0 +1,470 @@ +/* + * Copyright (C) 2003 PathScale, Inc. + * Copyright (C) 2003 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + + +#include <linux/personality.h> +#include <linux/ptrace.h> +#include <linux/kernel.h> +#include <linux/syscalls.h> +#include <asm/unistd.h> +#include <linux/uaccess.h> +#include <asm/ucontext.h> +#include <frame_kern.h> +#include <registers.h> +#include <skas.h> + +#include <linux/regset.h> +#include <asm/sigframe.h> + +#ifdef CONFIG_X86_32 +struct _xstate_64 { + struct _fpstate_64 fpstate; + struct _header xstate_hdr; + struct _ymmh_state ymmh; + /* New processor state extensions go here: */ +}; +#else +#define _xstate_64 _xstate +#endif + +static int copy_sc_from_user(struct pt_regs *regs, + struct sigcontext __user *from) +{ + struct _xstate_64 __user *from_fp64; + struct sigcontext sc; + int err; + + /* Always make any pending restarted system calls return -EINTR */ + current->restart_block.fn = do_no_restart_syscall; + + err = copy_from_user(&sc, from, sizeof(sc)); + if (err) + return err; + +#define GETREG(regno, regname) regs->regs.gp[HOST_##regno] = sc.regname + +#ifdef CONFIG_X86_32 + GETREG(GS, gs); + GETREG(FS, fs); + GETREG(ES, es); + GETREG(DS, ds); +#endif + GETREG(DI, di); + GETREG(SI, si); + GETREG(BP, bp); + GETREG(SP, sp); + GETREG(BX, bx); + GETREG(DX, dx); + GETREG(CX, cx); + GETREG(AX, ax); + GETREG(IP, ip); + +#ifdef CONFIG_X86_64 + GETREG(R8, r8); + GETREG(R9, r9); + GETREG(R10, r10); + GETREG(R11, r11); + GETREG(R12, r12); + GETREG(R13, r13); + GETREG(R14, r14); + GETREG(R15, r15); +#endif + + GETREG(CS, cs); + GETREG(EFLAGS, flags); +#ifdef CONFIG_X86_32 + GETREG(SS, ss); +#endif + +#undef GETREG + +#ifdef CONFIG_X86_32 + from_fp64 = ((void __user *)sc.fpstate) + + offsetof(struct _fpstate_32, _fxsr_env); +#else + from_fp64 = (void __user *)sc.fpstate; +#endif + + err = copy_from_user(regs->regs.fp, from_fp64, host_fp_size); + if (err) + return 1; + +#ifdef CONFIG_X86_32 + /* Data is duplicated and this copy is the important one */ + err = copy_regset_from_user(current, + task_user_regset_view(current), + REGSET_FP_LEGACY, 0, + sizeof(struct user_i387_struct), + (void __user *)sc.fpstate); + if (err < 0) + return err; +#endif + + return 0; +} + +static int copy_sc_to_user(struct sigcontext __user *to, + struct _xstate __user *to_fp, struct pt_regs *regs, + unsigned long mask) +{ + struct _xstate_64 __user *to_fp64; + struct sigcontext sc; + struct faultinfo * fi = ¤t->thread.arch.faultinfo; + int err; + memset(&sc, 0, sizeof(struct sigcontext)); + +#define PUTREG(regno, regname) sc.regname = regs->regs.gp[HOST_##regno] + +#ifdef CONFIG_X86_32 + PUTREG(GS, gs); + PUTREG(FS, fs); + PUTREG(ES, es); + PUTREG(DS, ds); +#endif + PUTREG(DI, di); + PUTREG(SI, si); + PUTREG(BP, bp); + PUTREG(SP, sp); + PUTREG(BX, bx); + PUTREG(DX, dx); + PUTREG(CX, cx); + PUTREG(AX, ax); +#ifdef CONFIG_X86_64 + PUTREG(R8, r8); + PUTREG(R9, r9); + PUTREG(R10, r10); + PUTREG(R11, r11); + PUTREG(R12, r12); + PUTREG(R13, r13); + PUTREG(R14, r14); + PUTREG(R15, r15); +#endif + + sc.cr2 = fi->cr2; + sc.err = fi->error_code; + sc.trapno = fi->trap_no; + PUTREG(IP, ip); + PUTREG(CS, cs); + PUTREG(EFLAGS, flags); +#ifdef CONFIG_X86_32 + PUTREG(SP, sp_at_signal); + PUTREG(SS, ss); +#endif +#undef PUTREG + sc.oldmask = mask; + sc.fpstate = (unsigned long)to_fp; + + err = copy_to_user(to, &sc, sizeof(struct sigcontext)); + if (err) + return 1; + +#ifdef CONFIG_X86_32 + err = copy_regset_to_user(current, + task_user_regset_view(current), + REGSET_FP_LEGACY, 0, + sizeof(struct _fpstate_32), to_fp); + if (err < 0) + return err; + + __put_user(X86_FXSR_MAGIC, &to_fp->fpstate.magic); + + BUILD_BUG_ON(offsetof(struct _xstate, xstate_hdr) != + offsetof(struct _xstate_64, xstate_hdr) + + offsetof(struct _fpstate_32, _fxsr_env)); + to_fp64 = (void __user *)to_fp + + offsetof(struct _fpstate_32, _fxsr_env); +#else + to_fp64 = to_fp; +#endif /* CONFIG_X86_32 */ + + if (copy_to_user(to_fp64, regs->regs.fp, host_fp_size)) + return 1; + + /* + * Put magic/size values for userspace. We do not bother to verify them + * later on, however, userspace needs them should it try to read the + * XSTATE data. And ptrace does not fill in these parts. + * + * Skip this if we do not have an XSTATE frame. + */ + if (host_fp_size <= sizeof(to_fp64->fpstate)) + return 0; + + BUILD_BUG_ON(sizeof(int) != FP_XSTATE_MAGIC2_SIZE); +#ifdef CONFIG_X86_32 + __put_user(offsetof(struct _fpstate_32, _fxsr_env) + + host_fp_size + FP_XSTATE_MAGIC2_SIZE, + &to_fp64->fpstate.sw_reserved.extended_size); +#else + __put_user(host_fp_size + FP_XSTATE_MAGIC2_SIZE, + &to_fp64->fpstate.sw_reserved.extended_size); +#endif + __put_user(host_fp_size, &to_fp64->fpstate.sw_reserved.xstate_size); + + __put_user(FP_XSTATE_MAGIC1, &to_fp64->fpstate.sw_reserved.magic1); + __put_user(FP_XSTATE_MAGIC2, + (int __user *)((void __user *)to_fp64 + host_fp_size)); + + return 0; +} + +#ifdef CONFIG_X86_32 +static int copy_ucontext_to_user(struct ucontext __user *uc, + struct _xstate __user *fp, sigset_t *set, + unsigned long sp) +{ + int err = 0; + + err |= __save_altstack(&uc->uc_stack, sp); + err |= copy_sc_to_user(&uc->uc_mcontext, fp, ¤t->thread.regs, 0); + err |= copy_to_user(&uc->uc_sigmask, set, sizeof(*set)); + return err; +} + +int setup_signal_stack_sc(unsigned long stack_top, struct ksignal *ksig, + struct pt_regs *regs, sigset_t *mask) +{ + size_t math_size = offsetof(struct _fpstate_32, _fxsr_env) + + host_fp_size + FP_XSTATE_MAGIC2_SIZE; + struct sigframe __user *frame; + void __user *restorer; + int err = 0, sig = ksig->sig; + unsigned long fp_to; + + /* This is the same calculation as i386 - ((sp + 4) & 15) == 0 */ + stack_top = ((stack_top + 4) & -16UL) - 4; + frame = (struct sigframe __user *) stack_top - 1; + if (!access_ok(frame, sizeof(*frame))) + return 1; + + /* Add required space for math frame */ + frame = (struct sigframe __user *)((unsigned long)frame - math_size); + + restorer = frame->retcode; + if (ksig->ka.sa.sa_flags & SA_RESTORER) + restorer = ksig->ka.sa.sa_restorer; + + err |= __put_user(restorer, (void __user * __user *)&frame->pretcode); + err |= __put_user(sig, &frame->sig); + + fp_to = (unsigned long)frame + sizeof(*frame); + + err |= copy_sc_to_user(&frame->sc, + (struct _xstate __user *)fp_to, + regs, mask->sig[0]); + if (_NSIG_WORDS > 1) + err |= __copy_to_user(&frame->extramask, &mask->sig[1], + sizeof(frame->extramask)); + + /* + * This is popl %eax ; movl $,%eax ; int $0x80 + * + * WE DO NOT USE IT ANY MORE! It's only left here for historical + * reasons and because gdb uses it as a signature to notice + * signal handler stack frames. + */ + err |= __put_user(0xb858, (short __user *)(frame->retcode+0)); + err |= __put_user(__NR_sigreturn, (int __user *)(frame->retcode+2)); + err |= __put_user(0x80cd, (short __user *)(frame->retcode+6)); + + if (err) + return err; + + PT_REGS_SP(regs) = (unsigned long) frame; + PT_REGS_IP(regs) = (unsigned long) ksig->ka.sa.sa_handler; + PT_REGS_AX(regs) = (unsigned long) sig; + PT_REGS_DX(regs) = (unsigned long) 0; + PT_REGS_CX(regs) = (unsigned long) 0; + return 0; +} + +int setup_signal_stack_si(unsigned long stack_top, struct ksignal *ksig, + struct pt_regs *regs, sigset_t *mask) +{ + size_t math_size = offsetof(struct _fpstate_32, _fxsr_env) + + host_fp_size + FP_XSTATE_MAGIC2_SIZE; + struct rt_sigframe __user *frame; + void __user *restorer; + int err = 0, sig = ksig->sig; + unsigned long fp_to; + + stack_top &= -8UL; + frame = (struct rt_sigframe __user *) stack_top - 1; + if (!access_ok(frame, sizeof(*frame))) + return 1; + + /* Add required space for math frame */ + frame = (struct rt_sigframe __user *)((unsigned long)frame - math_size); + + restorer = frame->retcode; + if (ksig->ka.sa.sa_flags & SA_RESTORER) + restorer = ksig->ka.sa.sa_restorer; + + err |= __put_user(restorer, (void __user * __user *)&frame->pretcode); + err |= __put_user(sig, &frame->sig); + err |= __put_user(&frame->info, (void __user * __user *)&frame->pinfo); + err |= __put_user(&frame->uc, (void __user * __user *)&frame->puc); + err |= copy_siginfo_to_user(&frame->info, &ksig->info); + + fp_to = (unsigned long)frame + sizeof(*frame); + + err |= copy_ucontext_to_user(&frame->uc, (struct _xstate __user *)fp_to, + mask, PT_REGS_SP(regs)); + + /* + * This is movl $,%eax ; int $0x80 + * + * WE DO NOT USE IT ANY MORE! It's only left here for historical + * reasons and because gdb uses it as a signature to notice + * signal handler stack frames. + */ + err |= __put_user(0xb8, (char __user *)(frame->retcode+0)); + err |= __put_user(__NR_rt_sigreturn, (int __user *)(frame->retcode+1)); + err |= __put_user(0x80cd, (short __user *)(frame->retcode+5)); + + if (err) + return err; + + PT_REGS_SP(regs) = (unsigned long) frame; + PT_REGS_IP(regs) = (unsigned long) ksig->ka.sa.sa_handler; + PT_REGS_AX(regs) = (unsigned long) sig; + PT_REGS_DX(regs) = (unsigned long) &frame->info; + PT_REGS_CX(regs) = (unsigned long) &frame->uc; + return 0; +} + +SYSCALL_DEFINE0(sigreturn) +{ + unsigned long sp = PT_REGS_SP(¤t->thread.regs); + struct sigframe __user *frame = (struct sigframe __user *)(sp - 8); + sigset_t set; + struct sigcontext __user *sc = &frame->sc; + int sig_size = (_NSIG_WORDS - 1) * sizeof(unsigned long); + + if (copy_from_user(&set.sig[0], &sc->oldmask, sizeof(set.sig[0])) || + copy_from_user(&set.sig[1], frame->extramask, sig_size)) + goto segfault; + + set_current_blocked(&set); + + if (copy_sc_from_user(¤t->thread.regs, sc)) + goto segfault; + + /* Avoid ERESTART handling */ + PT_REGS_SYSCALL_NR(¤t->thread.regs) = -1; + return PT_REGS_SYSCALL_RET(¤t->thread.regs); + + segfault: + force_sig(SIGSEGV); + return 0; +} + +#else + +int setup_signal_stack_si(unsigned long stack_top, struct ksignal *ksig, + struct pt_regs *regs, sigset_t *set) +{ + unsigned long math_size = host_fp_size + FP_XSTATE_MAGIC2_SIZE; + struct rt_sigframe __user *frame; + int err = 0, sig = ksig->sig; + unsigned long fp_to; + + frame = (void __user *)stack_top - sizeof(struct rt_sigframe); + + /* Add required space for math frame */ + frame = (void __user *)((unsigned long)frame - math_size); + + /* ABI requires 16 byte boundary alignment */ + frame = (void __user *)round_down((unsigned long)frame, 16); + + /* Subtract 128 for a red zone and 8 for proper alignment */ + frame = (struct rt_sigframe __user *) ((unsigned long) frame - 128 - 8); + + if (!access_ok(frame, sizeof(*frame) + math_size)) + goto out; + + if (ksig->ka.sa.sa_flags & SA_SIGINFO) { + err |= copy_siginfo_to_user(&frame->info, &ksig->info); + if (err) + goto out; + } + + /* Create the ucontext. */ + err |= __put_user(0, &frame->uc.uc_flags); + err |= __put_user(NULL, &frame->uc.uc_link); + err |= __save_altstack(&frame->uc.uc_stack, PT_REGS_SP(regs)); + + fp_to = (unsigned long)frame + sizeof(*frame); + + err |= copy_sc_to_user(&frame->uc.uc_mcontext, + (struct _xstate __user *)fp_to, + regs, set->sig[0]); + + err |= __put_user(fp_to, &frame->uc.uc_mcontext.fpstate); + if (sizeof(*set) == 16) { + err |= __put_user(set->sig[0], &frame->uc.uc_sigmask.sig[0]); + err |= __put_user(set->sig[1], &frame->uc.uc_sigmask.sig[1]); + } + else + err |= __copy_to_user(&frame->uc.uc_sigmask, set, + sizeof(*set)); + + /* + * Set up to return from userspace. If provided, use a stub + * already in userspace. + */ + /* x86-64 should always use SA_RESTORER. */ + if (ksig->ka.sa.sa_flags & SA_RESTORER) + err |= __put_user((void __user *)ksig->ka.sa.sa_restorer, + &frame->pretcode); + else + /* could use a vstub here */ + return err; + + if (err) + return err; + + PT_REGS_SP(regs) = (unsigned long) frame; + PT_REGS_DI(regs) = sig; + /* In case the signal handler was declared without prototypes */ + PT_REGS_AX(regs) = 0; + + /* + * This also works for non SA_SIGINFO handlers because they expect the + * next argument after the signal number on the stack. + */ + PT_REGS_SI(regs) = (unsigned long) &frame->info; + PT_REGS_DX(regs) = (unsigned long) &frame->uc; + PT_REGS_IP(regs) = (unsigned long) ksig->ka.sa.sa_handler; + out: + return err; +} +#endif + +SYSCALL_DEFINE0(rt_sigreturn) +{ + unsigned long sp = PT_REGS_SP(¤t->thread.regs); + struct rt_sigframe __user *frame = + (struct rt_sigframe __user *)(sp - sizeof(long)); + struct ucontext __user *uc = &frame->uc; + sigset_t set; + + if (copy_from_user(&set, &uc->uc_sigmask, sizeof(set))) + goto segfault; + + set_current_blocked(&set); + + if (copy_sc_from_user(¤t->thread.regs, &uc->uc_mcontext)) + goto segfault; + + /* Avoid ERESTART handling */ + PT_REGS_SYSCALL_NR(¤t->thread.regs) = -1; + return PT_REGS_SYSCALL_RET(¤t->thread.regs); + + segfault: + force_sig(SIGSEGV); + return 0; +} diff --git a/arch/x86/um/stub_segv.c b/arch/x86/um/stub_segv.c new file mode 100644 index 000000000..040668b98 --- /dev/null +++ b/arch/x86/um/stub_segv.c @@ -0,0 +1,20 @@ +/* + * Copyright (C) 2004 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Licensed under the GPL + */ + +#include <sysdep/stub.h> +#include <sysdep/faultinfo.h> +#include <sysdep/mcontext.h> +#include <sys/ucontext.h> + +void __attribute__ ((__section__ (".__syscall_stub"))) +stub_segv_handler(int sig, siginfo_t *info, void *p) +{ + struct faultinfo *f = get_stub_data(); + ucontext_t *uc = p; + + GET_FAULTINFO_FROM_MC(*f, &uc->uc_mcontext); + trap_myself(); +} + diff --git a/arch/x86/um/sys_call_table_32.c b/arch/x86/um/sys_call_table_32.c new file mode 100644 index 000000000..51655133e --- /dev/null +++ b/arch/x86/um/sys_call_table_32.c @@ -0,0 +1,40 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * System call table for UML/i386, copied from arch/x86/kernel/syscall_*.c + * with some changes for UML. + */ + +#include <linux/linkage.h> +#include <linux/sys.h> +#include <linux/cache.h> +#include <asm/syscall.h> + +extern asmlinkage long sys_ni_syscall(unsigned long, unsigned long, + unsigned long, unsigned long, + unsigned long, unsigned long); + +/* + * Below you can see, in terms of #define's, the differences between the x86-64 + * and the UML syscall table. + */ + +/* Not going to be implemented by UML, since we have no hardware. */ +#define sys_iopl sys_ni_syscall +#define sys_ioperm sys_ni_syscall + +#define sys_vm86old sys_ni_syscall +#define sys_vm86 sys_ni_syscall + +#define __SYSCALL_WITH_COMPAT(nr, native, compat) __SYSCALL(nr, native) +#define __SYSCALL_NORETURN __SYSCALL + +#define __SYSCALL(nr, sym) extern asmlinkage long sym(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long); +#include <asm/syscalls_32.h> +#undef __SYSCALL + +#define __SYSCALL(nr, sym) sym, +const sys_call_ptr_t sys_call_table[] ____cacheline_aligned = { +#include <asm/syscalls_32.h> +}; + +int syscall_table_size = sizeof(sys_call_table); diff --git a/arch/x86/um/sys_call_table_64.c b/arch/x86/um/sys_call_table_64.c new file mode 100644 index 000000000..943d414f2 --- /dev/null +++ b/arch/x86/um/sys_call_table_64.c @@ -0,0 +1,36 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * System call table for UML/x86-64, copied from arch/x86/kernel/syscall_*.c + * with some changes for UML. + */ + +#include <linux/linkage.h> +#include <linux/sys.h> +#include <linux/cache.h> +#include <asm/syscall.h> + +extern asmlinkage long sys_ni_syscall(unsigned long, unsigned long, + unsigned long, unsigned long, + unsigned long, unsigned long); + +/* + * Below you can see, in terms of #define's, the differences between the x86-64 + * and the UML syscall table. + */ + +/* Not going to be implemented by UML, since we have no hardware. */ +#define sys_iopl sys_ni_syscall +#define sys_ioperm sys_ni_syscall + +#define __SYSCALL_NORETURN __SYSCALL + +#define __SYSCALL(nr, sym) extern asmlinkage long sym(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long, unsigned long); +#include <asm/syscalls_64.h> +#undef __SYSCALL + +#define __SYSCALL(nr, sym) sym, +const sys_call_ptr_t sys_call_table[] ____cacheline_aligned = { +#include <asm/syscalls_64.h> +}; + +int syscall_table_size = sizeof(sys_call_table); diff --git a/arch/x86/um/syscalls_32.c b/arch/x86/um/syscalls_32.c new file mode 100644 index 000000000..5c65254c8 --- /dev/null +++ b/arch/x86/um/syscalls_32.c @@ -0,0 +1,8 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/syscalls.h> +#include <os.h> + +SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2) +{ + return -EINVAL; +} diff --git a/arch/x86/um/syscalls_64.c b/arch/x86/um/syscalls_64.c new file mode 100644 index 000000000..6a00a28c9 --- /dev/null +++ b/arch/x86/um/syscalls_64.c @@ -0,0 +1,64 @@ +/* + * Copyright (C) 2003 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com) + * Copyright 2003 PathScale, Inc. + * + * Licensed under the GPL + */ + +#include <linux/sched.h> +#include <linux/sched/mm.h> +#include <linux/syscalls.h> +#include <linux/uaccess.h> +#include <asm/prctl.h> /* XXX This should get the constants from libc */ +#include <registers.h> +#include <os.h> + +long arch_prctl(struct task_struct *task, int option, + unsigned long __user *arg2) +{ + long ret = -EINVAL; + + switch (option) { + case ARCH_SET_FS: + current->thread.regs.regs.gp[FS_BASE / sizeof(unsigned long)] = + (unsigned long) arg2; + ret = 0; + break; + case ARCH_SET_GS: + current->thread.regs.regs.gp[GS_BASE / sizeof(unsigned long)] = + (unsigned long) arg2; + ret = 0; + break; + case ARCH_GET_FS: + ret = put_user(current->thread.regs.regs.gp[FS_BASE / sizeof(unsigned long)], arg2); + break; + case ARCH_GET_GS: + ret = put_user(current->thread.regs.regs.gp[GS_BASE / sizeof(unsigned long)], arg2); + break; + } + + return ret; +} + +SYSCALL_DEFINE2(arch_prctl, int, option, unsigned long, arg2) +{ + return arch_prctl(current, option, (unsigned long __user *) arg2); +} + +void arch_switch_to(struct task_struct *to) +{ + /* + * Nothing needs to be done on x86_64. + * The FS_BASE/GS_BASE registers are saved in the ptrace register set. + */ +} + +SYSCALL_DEFINE6(mmap, unsigned long, addr, unsigned long, len, + unsigned long, prot, unsigned long, flags, + unsigned long, fd, unsigned long, off) +{ + if (off & ~PAGE_MASK) + return -EINVAL; + + return ksys_mmap_pgoff(addr, len, prot, flags, fd, off >> PAGE_SHIFT); +} diff --git a/arch/x86/um/sysrq_32.c b/arch/x86/um/sysrq_32.c new file mode 100644 index 000000000..a1ee415c0 --- /dev/null +++ b/arch/x86/um/sysrq_32.c @@ -0,0 +1,33 @@ +/* + * Copyright (C) 2001 - 2003 Jeff Dike (jdike@addtoit.com) + * Licensed under the GPL + */ + +#include <linux/kernel.h> +#include <linux/smp.h> +#include <linux/sched.h> +#include <linux/sched/debug.h> +#include <linux/kallsyms.h> +#include <asm/ptrace.h> + +/* This is declared by <linux/sched.h> */ +void show_regs(struct pt_regs *regs) +{ + printk("\n"); + printk("EIP: %04lx:[<%08lx>] CPU: %d %s", + 0xffff & PT_REGS_CS(regs), PT_REGS_IP(regs), + smp_processor_id(), print_tainted()); + if (PT_REGS_CS(regs) & 3) + printk(" ESP: %04lx:%08lx", 0xffff & PT_REGS_SS(regs), + PT_REGS_SP(regs)); + printk(" EFLAGS: %08lx\n %s\n", PT_REGS_EFLAGS(regs), + print_tainted()); + printk("EAX: %08lx EBX: %08lx ECX: %08lx EDX: %08lx\n", + PT_REGS_AX(regs), PT_REGS_BX(regs), + PT_REGS_CX(regs), PT_REGS_DX(regs)); + printk("ESI: %08lx EDI: %08lx EBP: %08lx", + PT_REGS_SI(regs), PT_REGS_DI(regs), PT_REGS_BP(regs)); + printk(" DS: %04lx ES: %04lx\n", + 0xffff & PT_REGS_DS(regs), + 0xffff & PT_REGS_ES(regs)); +} diff --git a/arch/x86/um/sysrq_64.c b/arch/x86/um/sysrq_64.c new file mode 100644 index 000000000..340d8a243 --- /dev/null +++ b/arch/x86/um/sysrq_64.c @@ -0,0 +1,36 @@ +/* + * Copyright 2003 PathScale, Inc. + * + * Licensed under the GPL + */ + +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/pid.h> +#include <linux/sched.h> +#include <linux/sched/debug.h> +#include <linux/utsname.h> +#include <asm/current.h> +#include <asm/ptrace.h> + +void show_regs(struct pt_regs *regs) +{ + printk("\n"); + print_modules(); + printk(KERN_INFO "Pid: %d, comm: %.20s %s %s\n", task_pid_nr(current), + current->comm, print_tainted(), init_utsname()->release); + printk(KERN_INFO "RIP: %04lx:%pS\n", PT_REGS_CS(regs) & 0xffff, + (void *)PT_REGS_IP(regs)); + printk(KERN_INFO "RSP: %016lx EFLAGS: %08lx\n", PT_REGS_SP(regs), + PT_REGS_EFLAGS(regs)); + printk(KERN_INFO "RAX: %016lx RBX: %016lx RCX: %016lx\n", + PT_REGS_AX(regs), PT_REGS_BX(regs), PT_REGS_CX(regs)); + printk(KERN_INFO "RDX: %016lx RSI: %016lx RDI: %016lx\n", + PT_REGS_DX(regs), PT_REGS_SI(regs), PT_REGS_DI(regs)); + printk(KERN_INFO "RBP: %016lx R08: %016lx R09: %016lx\n", + PT_REGS_BP(regs), PT_REGS_R8(regs), PT_REGS_R9(regs)); + printk(KERN_INFO "R10: %016lx R11: %016lx R12: %016lx\n", + PT_REGS_R10(regs), PT_REGS_R11(regs), PT_REGS_R12(regs)); + printk(KERN_INFO "R13: %016lx R14: %016lx R15: %016lx\n", + PT_REGS_R13(regs), PT_REGS_R14(regs), PT_REGS_R15(regs)); +} diff --git a/arch/x86/um/tls_32.c b/arch/x86/um/tls_32.c new file mode 100644 index 000000000..1909c2e64 --- /dev/null +++ b/arch/x86/um/tls_32.c @@ -0,0 +1,386 @@ +/* + * Copyright (C) 2005 Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> + * Licensed under the GPL + */ + +#include <linux/percpu.h> +#include <linux/sched.h> +#include <linux/syscalls.h> +#include <linux/uaccess.h> +#include <asm/ptrace-abi.h> +#include <os.h> +#include <skas.h> +#include <sysdep/tls.h> +#include <asm/desc.h> +#include <stub-data.h> + +/* + * If needed we can detect when it's uninitialized. + * + * These are initialized in an initcall and unchanged thereafter. + */ +static int host_supports_tls = -1; +int host_gdt_entry_tls_min; + +static int do_set_thread_area(struct task_struct* task, struct user_desc *info) +{ + int ret; + + if (info->entry_number < host_gdt_entry_tls_min || + info->entry_number >= host_gdt_entry_tls_min + GDT_ENTRY_TLS_ENTRIES) + return -EINVAL; + + if (using_seccomp) { + int idx = info->entry_number - host_gdt_entry_tls_min; + struct stub_data *data = (void *)task->mm->context.id.stack; + + data->arch_data.tls[idx] = *info; + data->arch_data.sync |= BIT(idx); + + return 0; + } + + ret = os_set_thread_area(info, task->mm->context.id.pid); + + if (ret) + printk(KERN_ERR "PTRACE_SET_THREAD_AREA failed, err = %d, " + "index = %d\n", ret, info->entry_number); + + return ret; +} + +/* + * sys_get_thread_area: get a yet unused TLS descriptor index. + * XXX: Consider leaving one free slot for glibc usage at first place. This must + * be done here (and by changing GDT_ENTRY_TLS_* macros) and nowhere else. + * + * Also, this must be tested when compiling in SKAS mode with dynamic linking + * and running against NPTL. + */ +static int get_free_idx(struct task_struct* task) +{ + struct thread_struct *t = &task->thread; + int idx; + + for (idx = 0; idx < GDT_ENTRY_TLS_ENTRIES; idx++) + if (!t->arch.tls_array[idx].present) + return idx + GDT_ENTRY_TLS_MIN; + return -ESRCH; +} + +static inline void clear_user_desc(struct user_desc* info) +{ + /* Postcondition: LDT_empty(info) returns true. */ + memset(info, 0, sizeof(*info)); + + /* + * Check the LDT_empty or the i386 sys_get_thread_area code - we obtain + * indeed an empty user_desc. + */ + info->read_exec_only = 1; + info->seg_not_present = 1; +} + +#define O_FORCE 1 + +static int load_TLS(int flags, struct task_struct *to) +{ + int ret = 0; + int idx; + + for (idx = GDT_ENTRY_TLS_MIN; idx < GDT_ENTRY_TLS_MAX; idx++) { + struct uml_tls_struct* curr = + &to->thread.arch.tls_array[idx - GDT_ENTRY_TLS_MIN]; + + /* + * Actually, now if it wasn't flushed it gets cleared and + * flushed to the host, which will clear it. + */ + if (!curr->present) { + if (!curr->flushed) { + clear_user_desc(&curr->tls); + curr->tls.entry_number = idx; + } else { + WARN_ON(!LDT_empty(&curr->tls)); + continue; + } + } + + if (!(flags & O_FORCE) && curr->flushed) + continue; + + ret = do_set_thread_area(current, &curr->tls); + if (ret) + goto out; + + curr->flushed = 1; + } +out: + return ret; +} + +/* + * Verify if we need to do a flush for the new process, i.e. if there are any + * present desc's, only if they haven't been flushed. + */ +static inline int needs_TLS_update(struct task_struct *task) +{ + int i; + int ret = 0; + + for (i = GDT_ENTRY_TLS_MIN; i < GDT_ENTRY_TLS_MAX; i++) { + struct uml_tls_struct* curr = + &task->thread.arch.tls_array[i - GDT_ENTRY_TLS_MIN]; + + /* + * Can't test curr->present, we may need to clear a descriptor + * which had a value. + */ + if (curr->flushed) + continue; + ret = 1; + break; + } + return ret; +} + +/* + * On a newly forked process, the TLS descriptors haven't yet been flushed. So + * we mark them as such and the first switch_to will do the job. + */ +void clear_flushed_tls(struct task_struct *task) +{ + int i; + + for (i = GDT_ENTRY_TLS_MIN; i < GDT_ENTRY_TLS_MAX; i++) { + struct uml_tls_struct* curr = + &task->thread.arch.tls_array[i - GDT_ENTRY_TLS_MIN]; + + /* + * Still correct to do this, if it wasn't present on the host it + * will remain as flushed as it was. + */ + if (!curr->present) + continue; + + curr->flushed = 0; + } +} + +/* + * In SKAS0 mode, currently, multiple guest threads sharing the same ->mm have a + * common host process. So this is needed in SKAS0 too. + * + * However, if each thread had a different host process (and this was discussed + * for SMP support) this won't be needed. + * + * And this will not need be used when (and if) we'll add support to the host + * SKAS patch. + */ + +int arch_switch_tls(struct task_struct *to) +{ + if (!host_supports_tls) + return 0; + + /* + * We have no need whatsoever to switch TLS for kernel threads; beyond + * that, that would also result in us calling os_set_thread_area with + * task->mm == NULL, which would cause a crash. + */ + if (likely(to->mm)) + return load_TLS(O_FORCE, to); + + return 0; +} + +static int set_tls_entry(struct task_struct* task, struct user_desc *info, + int idx, int flushed) +{ + struct thread_struct *t = &task->thread; + + if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) + return -EINVAL; + + t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].tls = *info; + t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].present = 1; + t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].flushed = flushed; + + return 0; +} + +int arch_set_tls(struct task_struct *new, unsigned long tls) +{ + struct user_desc info; + int idx, ret = -EFAULT; + + if (copy_from_user(&info, (void __user *) tls, sizeof(info))) + goto out; + + ret = -EINVAL; + if (LDT_empty(&info)) + goto out; + + idx = info.entry_number; + + ret = set_tls_entry(new, &info, idx, 0); +out: + return ret; +} + +static int get_tls_entry(struct task_struct *task, struct user_desc *info, + int idx) +{ + struct thread_struct *t = &task->thread; + + if (idx < GDT_ENTRY_TLS_MIN || idx > GDT_ENTRY_TLS_MAX) + return -EINVAL; + + if (!t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].present) + goto clear; + + *info = t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].tls; + +out: + /* + * Temporary debugging check, to make sure that things have been + * flushed. This could be triggered if load_TLS() failed. + */ + if (unlikely(task == current && + !t->arch.tls_array[idx - GDT_ENTRY_TLS_MIN].flushed)) { + printk(KERN_ERR "get_tls_entry: task with pid %d got here " + "without flushed TLS.", current->pid); + } + + return 0; +clear: + /* + * When the TLS entry has not been set, the values read to user in the + * tls_array are 0 (because it's cleared at boot, see + * arch/i386/kernel/head.S:cpu_gdt_table). Emulate that. + */ + clear_user_desc(info); + info->entry_number = idx; + goto out; +} + +SYSCALL_DEFINE1(set_thread_area, struct user_desc __user *, user_desc) +{ + struct user_desc info; + int idx, ret; + + if (!host_supports_tls) + return -ENOSYS; + + if (copy_from_user(&info, user_desc, sizeof(info))) + return -EFAULT; + + idx = info.entry_number; + + if (idx == -1) { + idx = get_free_idx(current); + if (idx < 0) + return idx; + info.entry_number = idx; + /* Tell the user which slot we chose for him.*/ + if (put_user(idx, &user_desc->entry_number)) + return -EFAULT; + } + + ret = do_set_thread_area(current, &info); + if (ret) + return ret; + return set_tls_entry(current, &info, idx, 1); +} + +/* + * Perform set_thread_area on behalf of the traced child. + * Note: error handling is not done on the deferred load, and this differ from + * i386. However the only possible error are caused by bugs. + */ +int ptrace_set_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc) +{ + struct user_desc info; + + if (!host_supports_tls) + return -EIO; + + if (copy_from_user(&info, user_desc, sizeof(info))) + return -EFAULT; + + return set_tls_entry(child, &info, idx, 0); +} + +SYSCALL_DEFINE1(get_thread_area, struct user_desc __user *, user_desc) +{ + struct user_desc info; + int idx, ret; + + if (!host_supports_tls) + return -ENOSYS; + + if (get_user(idx, &user_desc->entry_number)) + return -EFAULT; + + ret = get_tls_entry(current, &info, idx); + if (ret < 0) + goto out; + + if (copy_to_user(user_desc, &info, sizeof(info))) + ret = -EFAULT; + +out: + return ret; +} + +/* + * Perform get_thread_area on behalf of the traced child. + */ +int ptrace_get_thread_area(struct task_struct *child, int idx, + struct user_desc __user *user_desc) +{ + struct user_desc info; + int ret; + + if (!host_supports_tls) + return -EIO; + + ret = get_tls_entry(child, &info, idx); + if (ret < 0) + goto out; + + if (copy_to_user(user_desc, &info, sizeof(info))) + ret = -EFAULT; +out: + return ret; +} + +/* + * This code is really i386-only, but it detects and logs x86_64 GDT indexes + * if a 32-bit UML is running on a 64-bit host. + */ +static int __init __setup_host_supports_tls(void) +{ + check_host_supports_tls(&host_supports_tls, &host_gdt_entry_tls_min); + if (host_supports_tls) { + printk(KERN_INFO "Host TLS support detected\n"); + printk(KERN_INFO "Detected host type: "); + switch (host_gdt_entry_tls_min) { + case GDT_ENTRY_TLS_MIN_I386: + printk(KERN_CONT "i386"); + break; + case GDT_ENTRY_TLS_MIN_X86_64: + printk(KERN_CONT "x86_64"); + break; + } + printk(KERN_CONT " (GDT indexes %d to %d)\n", + host_gdt_entry_tls_min, + host_gdt_entry_tls_min + GDT_ENTRY_TLS_ENTRIES); + } else + printk(KERN_ERR " Host TLS support NOT detected! " + "TLS support inside UML will not work\n"); + return 0; +} + +__initcall(__setup_host_supports_tls); diff --git a/arch/x86/um/tls_64.c b/arch/x86/um/tls_64.c new file mode 100644 index 000000000..c51a613f6 --- /dev/null +++ b/arch/x86/um/tls_64.c @@ -0,0 +1,18 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <linux/sched.h> +#include <asm/ptrace-abi.h> + +void clear_flushed_tls(struct task_struct *task) +{ +} + +int arch_set_tls(struct task_struct *t, unsigned long tls) +{ + /* + * If CLONE_SETTLS is set, we need to save the thread id + * so it can be set during context switches. + */ + t->thread.regs.regs.gp[FS_BASE / sizeof(unsigned long)] = tls; + + return 0; +} diff --git a/arch/x86/um/user-offsets.c b/arch/x86/um/user-offsets.c new file mode 100644 index 000000000..d6e1cd995 --- /dev/null +++ b/arch/x86/um/user-offsets.c @@ -0,0 +1,79 @@ +// SPDX-License-Identifier: GPL-2.0 +#include <stdio.h> +#include <stddef.h> +#include <signal.h> +#include <poll.h> +#include <sys/mman.h> +#include <sys/user.h> +#define __FRAME_OFFSETS +#include <linux/ptrace.h> +#include <asm/types.h> +#include <linux/kbuild.h> + +#define DEFINE_LONGS(sym, val) \ + COMMENT(#val " / sizeof(unsigned long)"); \ + DEFINE(sym, val / sizeof(unsigned long)) + +/* workaround for a warning with -Wmissing-prototypes */ +void foo(void); + +void foo(void) +{ +#ifdef __i386__ + DEFINE(HOST_IP, EIP); + DEFINE(HOST_SP, UESP); + DEFINE(HOST_EFLAGS, EFL); + DEFINE(HOST_AX, EAX); + DEFINE(HOST_BX, EBX); + DEFINE(HOST_CX, ECX); + DEFINE(HOST_DX, EDX); + DEFINE(HOST_SI, ESI); + DEFINE(HOST_DI, EDI); + DEFINE(HOST_BP, EBP); + DEFINE(HOST_CS, CS); + DEFINE(HOST_SS, SS); + DEFINE(HOST_DS, DS); + DEFINE(HOST_FS, FS); + DEFINE(HOST_ES, ES); + DEFINE(HOST_GS, GS); + DEFINE(HOST_ORIG_AX, ORIG_EAX); +#else + DEFINE_LONGS(HOST_BX, RBX); + DEFINE_LONGS(HOST_CX, RCX); + DEFINE_LONGS(HOST_DI, RDI); + DEFINE_LONGS(HOST_SI, RSI); + DEFINE_LONGS(HOST_DX, RDX); + DEFINE_LONGS(HOST_BP, RBP); + DEFINE_LONGS(HOST_AX, RAX); + DEFINE_LONGS(HOST_R8, R8); + DEFINE_LONGS(HOST_R9, R9); + DEFINE_LONGS(HOST_R10, R10); + DEFINE_LONGS(HOST_R11, R11); + DEFINE_LONGS(HOST_R12, R12); + DEFINE_LONGS(HOST_R13, R13); + DEFINE_LONGS(HOST_R14, R14); + DEFINE_LONGS(HOST_R15, R15); + DEFINE_LONGS(HOST_ORIG_AX, ORIG_RAX); + DEFINE_LONGS(HOST_CS, CS); + DEFINE_LONGS(HOST_SS, SS); + DEFINE_LONGS(HOST_EFLAGS, EFLAGS); +#if 0 + DEFINE_LONGS(HOST_FS, FS); + DEFINE_LONGS(HOST_GS, GS); + DEFINE_LONGS(HOST_DS, DS); + DEFINE_LONGS(HOST_ES, ES); +#endif + + DEFINE_LONGS(HOST_IP, RIP); + DEFINE_LONGS(HOST_SP, RSP); +#endif + + DEFINE(UM_FRAME_SIZE, sizeof(struct user_regs_struct)); + DEFINE(UM_POLLIN, POLLIN); + DEFINE(UM_POLLPRI, POLLPRI); + DEFINE(UM_POLLOUT, POLLOUT); + + DEFINE(UM_PROT_READ, PROT_READ); + DEFINE(UM_PROT_WRITE, PROT_WRITE); + DEFINE(UM_PROT_EXEC, PROT_EXEC); +} diff --git a/arch/x86/um/vdso/.gitignore b/arch/x86/um/vdso/.gitignore new file mode 100644 index 000000000..652e31d82 --- /dev/null +++ b/arch/x86/um/vdso/.gitignore @@ -0,0 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only +vdso.lds diff --git a/arch/x86/um/vdso/Makefile b/arch/x86/um/vdso/Makefile new file mode 100644 index 000000000..7664cbedb --- /dev/null +++ b/arch/x86/um/vdso/Makefile @@ -0,0 +1,56 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# Building vDSO images for x86. +# + +# files to link into the vdso +vobjs-y := vdso-note.o um_vdso.o + +# files to link into kernel +obj-y += vdso.o vma.o + +vobjs := $(foreach F,$(vobjs-y),$(obj)/$F) + +$(obj)/vdso.o: $(obj)/vdso.so + +targets += vdso.so vdso.so.dbg vdso.lds $(vobjs-y) + +CPPFLAGS_vdso.lds += -P -C + +VDSO_LDFLAGS_vdso.lds = -m64 -Wl,-soname=linux-vdso.so.1 \ + -Wl,-z,max-page-size=4096 + +$(obj)/vdso.o: $(src)/vdso.S $(obj)/vdso.so + +$(obj)/vdso.so.dbg: $(obj)/vdso.lds $(vobjs) FORCE + $(call if_changed,vdso) + +$(obj)/%.so: OBJCOPYFLAGS := -S +$(obj)/%.so: $(obj)/%.so.dbg FORCE + $(call if_changed,objcopy) + +# +# Don't omit frame pointers for ease of userspace debugging, but do +# optimize sibling calls. +# +CFL := $(PROFILING) -mcmodel=small -fPIC -O2 -fasynchronous-unwind-tables -m64 \ + $(filter -g%,$(KBUILD_CFLAGS)) -fno-stack-protector \ + -fno-omit-frame-pointer -foptimize-sibling-calls + +$(vobjs): KBUILD_CFLAGS += $(CFL) + +# +# vDSO code runs in userspace and -pg doesn't help with profiling anyway. +# +CFLAGS_REMOVE_vdso-note.o = -pg -fprofile-arcs -ftest-coverage +CFLAGS_REMOVE_um_vdso.o = -pg -fprofile-arcs -ftest-coverage + +# +# The DSO images are built using a special linker script. +# +quiet_cmd_vdso = VDSO $@ + cmd_vdso = $(CC) -nostdlib -o $@ \ + $(CC_FLAGS_LTO) $(VDSO_LDFLAGS) $(VDSO_LDFLAGS_$(filter %.lds,$(^F))) \ + -Wl,-T,$(filter %.lds,$^) $(filter %.o,$^) + +VDSO_LDFLAGS = -fPIC -shared -Wl,--hash-style=sysv -z noexecstack -Wl,--no-undefined diff --git a/arch/x86/um/vdso/um_vdso.c b/arch/x86/um/vdso/um_vdso.c new file mode 100644 index 000000000..ca1468865 --- /dev/null +++ b/arch/x86/um/vdso/um_vdso.c @@ -0,0 +1,54 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2011 Richard Weinberger <richrd@nod.at> + * + * This vDSO turns all calls into a syscall so that UML can trap them. + */ + + +/* Disable profiling for userspace code */ +#define DISABLE_BRANCH_PROFILING + +#include <vdso/gettime.h> +#include <linux/time.h> +#include <asm/unistd.h> + +int __vdso_clock_gettime(clockid_t clock, struct __kernel_timespec *ts) +{ + long ret; + + asm("syscall" + : "=a" (ret) + : "0" (__NR_clock_gettime), "D" (clock), "S" (ts) + : "rcx", "r11", "memory"); + + return ret; +} +int clock_gettime(clockid_t, struct __kernel_timespec *) + __attribute__((weak, alias("__vdso_clock_gettime"))); + +int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz) +{ + long ret; + + asm("syscall" + : "=a" (ret) + : "0" (__NR_gettimeofday), "D" (tv), "S" (tz) + : "rcx", "r11", "memory"); + + return ret; +} +int gettimeofday(struct __kernel_old_timeval *, struct timezone *) + __attribute__((weak, alias("__vdso_gettimeofday"))); + +__kernel_old_time_t __vdso_time(__kernel_old_time_t *t) +{ + long secs; + + asm volatile("syscall" + : "=a" (secs) + : "0" (__NR_time), "D" (t) : "cc", "r11", "cx", "memory"); + + return secs; +} +__kernel_old_time_t time(__kernel_old_time_t *t) __attribute__((weak, alias("__vdso_time"))); diff --git a/arch/x86/um/vdso/vdso-layout.lds.S b/arch/x86/um/vdso/vdso-layout.lds.S new file mode 100644 index 000000000..439b790df --- /dev/null +++ b/arch/x86/um/vdso/vdso-layout.lds.S @@ -0,0 +1,65 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Linker script for vDSO. This is an ELF shared object prelinked to + * its virtual address, and with only one read-only segment. + * This script controls its layout. + */ + +SECTIONS +{ + . = VDSO_PRELINK + SIZEOF_HEADERS; + + .hash : { *(.hash) } :text + .gnu.hash : { *(.gnu.hash) } + .dynsym : { *(.dynsym) } + .dynstr : { *(.dynstr) } + .gnu.version : { *(.gnu.version) } + .gnu.version_d : { *(.gnu.version_d) } + .gnu.version_r : { *(.gnu.version_r) } + + .note : { *(.note.*) } :text :note + + .eh_frame_hdr : { *(.eh_frame_hdr) } :text :eh_frame_hdr + .eh_frame : { KEEP (*(.eh_frame)) } :text + + .dynamic : { *(.dynamic) } :text :dynamic + + .rodata : { *(.rodata*) } :text + .data : { + *(.data*) + *(.sdata*) + *(.got.plt) *(.got) + *(.gnu.linkonce.d.*) + *(.bss*) + *(.dynbss*) + *(.gnu.linkonce.b.*) + } + + .altinstructions : { *(.altinstructions) } + .altinstr_replacement : { *(.altinstr_replacement) } + + /* + * Align the actual code well away from the non-instruction data. + * This is the best thing for the I-cache. + */ + . = ALIGN(0x100); + + .text : { *(.text*) } :text =0x90909090 +} + +/* + * Very old versions of ld do not recognize this name token; use the constant. + */ +#define PT_GNU_EH_FRAME 0x6474e550 + +/* + * We must supply the ELF program headers explicitly to get just one + * PT_LOAD segment, and set the flags explicitly to make segments read-only. + */ +PHDRS +{ + text PT_LOAD FLAGS(5) FILEHDR PHDRS; /* PF_R|PF_X */ + dynamic PT_DYNAMIC FLAGS(4); /* PF_R */ + note PT_NOTE FLAGS(4); /* PF_R */ + eh_frame_hdr PT_GNU_EH_FRAME; +} diff --git a/arch/x86/um/vdso/vdso-note.S b/arch/x86/um/vdso/vdso-note.S new file mode 100644 index 000000000..79a071e43 --- /dev/null +++ b/arch/x86/um/vdso/vdso-note.S @@ -0,0 +1,12 @@ +/* + * This supplies .note.* sections to go into the PT_NOTE inside the vDSO text. + * Here we can supply some information useful to userland. + */ + +#include <linux/uts.h> +#include <linux/version.h> +#include <linux/elfnote.h> + +ELFNOTE_START(Linux, 0, "a") + .long LINUX_VERSION_CODE +ELFNOTE_END diff --git a/arch/x86/um/vdso/vdso.S b/arch/x86/um/vdso/vdso.S new file mode 100644 index 000000000..a6eaf293a --- /dev/null +++ b/arch/x86/um/vdso/vdso.S @@ -0,0 +1,11 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#include <linux/init.h> +#include <linux/linkage.h> + +__INITDATA + +SYM_DATA_START(vdso_start) + .incbin "arch/x86/um/vdso/vdso.so" +SYM_DATA_END_LABEL(vdso_start, SYM_L_GLOBAL, vdso_end) + +__FINIT diff --git a/arch/x86/um/vdso/vdso.lds.S b/arch/x86/um/vdso/vdso.lds.S new file mode 100644 index 000000000..401600eff --- /dev/null +++ b/arch/x86/um/vdso/vdso.lds.S @@ -0,0 +1,31 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Linker script for 64-bit vDSO. + * We #include the file to define the layout details. + * Here we only choose the prelinked virtual address. + * + * This file defines the version script giving the user-exported symbols in + * the DSO. We can define local symbols here called VDSO* to make their + * values visible using the asm-x86/vdso.h macros from the kernel proper. + */ + +#define VDSO_PRELINK 0xffffffffff700000 +#include "vdso-layout.lds.S" + +/* + * This controls what userland symbols we export from the vDSO. + */ +VERSION { + LINUX_2.6 { + global: + clock_gettime; + __vdso_clock_gettime; + gettimeofday; + __vdso_gettimeofday; + time; + __vdso_time; + local: *; + }; +} + +VDSO64_PRELINK = VDSO_PRELINK; diff --git a/arch/x86/um/vdso/vma.c b/arch/x86/um/vdso/vma.c new file mode 100644 index 000000000..51a2b9f2e --- /dev/null +++ b/arch/x86/um/vdso/vma.c @@ -0,0 +1,55 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2011 Richard Weinberger <richrd@nod.at> + */ + +#include <linux/slab.h> +#include <linux/sched.h> +#include <linux/mm.h> +#include <asm/page.h> +#include <asm/elf.h> +#include <linux/init.h> + +unsigned long um_vdso_addr; +static struct page *um_vdso; + +extern unsigned long task_size; +extern char vdso_start[], vdso_end[]; + +static int __init init_vdso(void) +{ + BUG_ON(vdso_end - vdso_start > PAGE_SIZE); + + um_vdso_addr = task_size - PAGE_SIZE; + + um_vdso = alloc_page(GFP_KERNEL); + if (!um_vdso) + panic("Cannot allocate vdso\n"); + + copy_page(page_address(um_vdso), vdso_start); + + return 0; +} +subsys_initcall(init_vdso); + +int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp) +{ + struct vm_area_struct *vma; + struct mm_struct *mm = current->mm; + static struct vm_special_mapping vdso_mapping = { + .name = "[vdso]", + .pages = &um_vdso, + }; + + if (mmap_write_lock_killable(mm)) + return -EINTR; + + vma = _install_special_mapping(mm, um_vdso_addr, PAGE_SIZE, + VM_READ|VM_EXEC| + VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC, + &vdso_mapping); + + mmap_write_unlock(mm); + + return IS_ERR(vma) ? PTR_ERR(vma) : 0; +} |