Commit d6be29b8 authored by Masami Hiramatsu's avatar Masami Hiramatsu Committed by Ingo Molnar

x86: kprobes code for x86 unification

This patch unifies kprobes code.

- Unify kprobes_*.h to kprobes.h
- Unify kprobes_*.c to kprobes.c
  (Differences are separated by ifdefs)
 - Most differences are related to REX prefix and rip relatives.
 - Two inline assembly code are different.
 - One difference in kprobe_handlre()
 - One fixup exception code is different, but it will be unified
   if mm/extable_*.c are unified.
- Merge history logs into arch/x86/kernel/kprobes.c.
Signed-off-by: default avatarMasami Hiramatsu <mhiramat@redhat.com>
Signed-off-by: default avatarJim Keniston <jkenisto@us.ibm.com>
Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
Signed-off-by: default avatarThomas Gleixner <tglx@linutronix.de>
parent 8533bbe9
...@@ -35,7 +35,7 @@ obj-$(CONFIG_KEXEC) += machine_kexec_32.o relocate_kernel_32.o crash.o ...@@ -35,7 +35,7 @@ obj-$(CONFIG_KEXEC) += machine_kexec_32.o relocate_kernel_32.o crash.o
obj-$(CONFIG_CRASH_DUMP) += crash_dump_32.o obj-$(CONFIG_CRASH_DUMP) += crash_dump_32.o
obj-$(CONFIG_X86_NUMAQ) += numaq_32.o obj-$(CONFIG_X86_NUMAQ) += numaq_32.o
obj-$(CONFIG_X86_SUMMIT_NUMA) += summit_32.o obj-$(CONFIG_X86_SUMMIT_NUMA) += summit_32.o
obj-$(CONFIG_KPROBES) += kprobes_32.o obj-$(CONFIG_KPROBES) += kprobes.o
obj-$(CONFIG_MODULES) += module_32.o obj-$(CONFIG_MODULES) += module_32.o
obj-$(CONFIG_ACPI_SRAT) += srat_32.o obj-$(CONFIG_ACPI_SRAT) += srat_32.o
obj-$(CONFIG_EFI) += efi.o efi_32.o efi_stub_32.o obj-$(CONFIG_EFI) += efi.o efi_32.o efi_stub_32.o
......
...@@ -34,7 +34,7 @@ obj-$(CONFIG_EARLY_PRINTK) += early_printk.o ...@@ -34,7 +34,7 @@ obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
obj-$(CONFIG_GART_IOMMU) += pci-gart_64.o aperture_64.o obj-$(CONFIG_GART_IOMMU) += pci-gart_64.o aperture_64.o
obj-$(CONFIG_CALGARY_IOMMU) += pci-calgary_64.o tce_64.o obj-$(CONFIG_CALGARY_IOMMU) += pci-calgary_64.o tce_64.o
obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o obj-$(CONFIG_SWIOTLB) += pci-swiotlb_64.o
obj-$(CONFIG_KPROBES) += kprobes_64.o obj-$(CONFIG_KPROBES) += kprobes.o
obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o obj-$(CONFIG_X86_PM_TIMER) += pmtimer_64.o
obj-$(CONFIG_X86_VSMP) += vsmp_64.o obj-$(CONFIG_X86_VSMP) += vsmp_64.o
obj-$(CONFIG_K8_NB) += k8.o obj-$(CONFIG_K8_NB) += k8.o
......
...@@ -22,14 +22,22 @@ ...@@ -22,14 +22,22 @@
* Rusty Russell). * Rusty Russell).
* 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes * 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
* interface to access function arguments. * interface to access function arguments.
* 2004-Oct Jim Keniston <kenistoj@us.ibm.com> and Prasanna S Panchamukhi * 2004-Oct Jim Keniston <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
* <prasanna@in.ibm.com> adapted for x86_64 * <prasanna@in.ibm.com> adapted for x86_64 from i386.
* 2005-Mar Roland McGrath <roland@redhat.com> * 2005-Mar Roland McGrath <roland@redhat.com>
* Fixed to handle %rip-relative addressing mode correctly. * Fixed to handle %rip-relative addressing mode correctly.
* 2005-May Hien Nguyen <hien@us.ibm.com>, Jim Keniston
* <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
* <prasanna@in.ibm.com> added function-return probes.
* 2005-May Rusty Lynch <rusty.lynch@intel.com> * 2005-May Rusty Lynch <rusty.lynch@intel.com>
* Added function return probes functionality * Added function return probes functionality
* 2006-Feb Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp> added
* kprobe-booster and kretprobe-booster for i386.
* 2007-Dec Masami Hiramatsu <mhiramat@redhat.com> added kprobe-booster * 2007-Dec Masami Hiramatsu <mhiramat@redhat.com> added kprobe-booster
* and kretprobe-booster for x86-64 * and kretprobe-booster for x86-64
* 2007-Dec Masami Hiramatsu <mhiramat@redhat.com>, Arjan van de Ven
* <arjan@infradead.org> and Jim Keniston <jkenisto@us.ibm.com>
* unified x86 kprobes code.
*/ */
#include <linux/kprobes.h> #include <linux/kprobes.h>
...@@ -51,7 +59,19 @@ void jprobe_return_end(void); ...@@ -51,7 +59,19 @@ void jprobe_return_end(void);
DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL; DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk); DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
#ifdef CONFIG_X86_64
#define stack_addr(regs) ((unsigned long *)regs->sp) #define stack_addr(regs) ((unsigned long *)regs->sp)
#else
/*
* "&regs->sp" looks wrong, but it's correct for x86_32. x86_32 CPUs
* don't save the ss and esp registers if the CPU is already in kernel
* mode when it traps. So for kprobes, regs->sp and regs->ss are not
* the [nonexistent] saved stack pointer and ss register, but rather
* the top 8 bytes of the pre-int3 stack. So &regs->sp happens to
* point to the top of the pre-int3 stack.
*/
#define stack_addr(regs) ((unsigned long *)&regs->sp)
#endif
#define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\ #define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\
(((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) | \ (((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) | \
...@@ -151,8 +171,8 @@ static __always_inline void set_jmp_op(void *from, void *to) ...@@ -151,8 +171,8 @@ static __always_inline void set_jmp_op(void *from, void *to)
} }
/* /*
* returns non-zero if opcode is boostable. * Returns non-zero if opcode is boostable.
* RIP relative instructions are adjusted at copying time * RIP relative instructions are adjusted at copying time in 64 bits mode
*/ */
static __always_inline int can_boost(kprobe_opcode_t *opcodes) static __always_inline int can_boost(kprobe_opcode_t *opcodes)
{ {
...@@ -173,8 +193,10 @@ static __always_inline int can_boost(kprobe_opcode_t *opcodes) ...@@ -173,8 +193,10 @@ static __always_inline int can_boost(kprobe_opcode_t *opcodes)
} }
switch (opcode & 0xf0) { switch (opcode & 0xf0) {
#ifdef CONFIG_X86_64
case 0x40: case 0x40:
goto retry; /* REX prefix is boostable */ goto retry; /* REX prefix is boostable */
#endif
case 0x60: case 0x60:
if (0x63 < opcode && opcode < 0x67) if (0x63 < opcode && opcode < 0x67)
goto retry; /* prefixes */ goto retry; /* prefixes */
...@@ -206,7 +228,7 @@ static __always_inline int can_boost(kprobe_opcode_t *opcodes) ...@@ -206,7 +228,7 @@ static __always_inline int can_boost(kprobe_opcode_t *opcodes)
} }
/* /*
* returns non-zero if opcode modifies the interrupt flag. * Returns non-zero if opcode modifies the interrupt flag.
*/ */
static int __kprobes is_IF_modifier(kprobe_opcode_t *insn) static int __kprobes is_IF_modifier(kprobe_opcode_t *insn)
{ {
...@@ -217,16 +239,18 @@ static int __kprobes is_IF_modifier(kprobe_opcode_t *insn) ...@@ -217,16 +239,18 @@ static int __kprobes is_IF_modifier(kprobe_opcode_t *insn)
case 0x9d: /* popf/popfd */ case 0x9d: /* popf/popfd */
return 1; return 1;
} }
#ifdef CONFIG_X86_64
/* /*
* on 64 bit x86, 0x40-0x4f are prefixes so we need to look * on 64 bit x86, 0x40-0x4f are prefixes so we need to look
* at the next byte instead.. but of course not recurse infinitely * at the next byte instead.. but of course not recurse infinitely
*/ */
if (*insn >= 0x40 && *insn <= 0x4f) if (*insn >= 0x40 && *insn <= 0x4f)
return is_IF_modifier(++insn); return is_IF_modifier(++insn);
#endif
return 0; return 0;
} }
#ifdef CONFIG_X86_64
/* /*
* Adjust the displacement if the instruction uses the %rip-relative * Adjust the displacement if the instruction uses the %rip-relative
* addressing mode. * addressing mode.
...@@ -263,17 +287,20 @@ static void __kprobes fix_riprel(struct kprobe *p) ...@@ -263,17 +287,20 @@ static void __kprobes fix_riprel(struct kprobe *p)
if ((*insn & 0xf0) == 0x40) if ((*insn & 0xf0) == 0x40)
++insn; ++insn;
if (*insn == 0x0f) { /* Two-byte opcode. */ if (*insn == 0x0f) {
/* Two-byte opcode. */
++insn; ++insn;
need_modrm = test_bit(*insn, need_modrm = test_bit(*insn,
(unsigned long *)twobyte_has_modrm); (unsigned long *)twobyte_has_modrm);
} else /* One-byte opcode. */ } else
/* One-byte opcode. */
need_modrm = test_bit(*insn, need_modrm = test_bit(*insn,
(unsigned long *)onebyte_has_modrm); (unsigned long *)onebyte_has_modrm);
if (need_modrm) { if (need_modrm) {
u8 modrm = *++insn; u8 modrm = *++insn;
if ((modrm & 0xc7) == 0x05) { /* %rip+disp32 addressing mode */ if ((modrm & 0xc7) == 0x05) {
/* %rip+disp32 addressing mode */
/* Displacement follows ModRM byte. */ /* Displacement follows ModRM byte. */
++insn; ++insn;
/* /*
...@@ -296,11 +323,14 @@ static void __kprobes fix_riprel(struct kprobe *p) ...@@ -296,11 +323,14 @@ static void __kprobes fix_riprel(struct kprobe *p)
} }
} }
} }
#endif
static void __kprobes arch_copy_kprobe(struct kprobe *p) static void __kprobes arch_copy_kprobe(struct kprobe *p)
{ {
memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t)); memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
#ifdef CONFIG_X86_64
fix_riprel(p); fix_riprel(p);
#endif
if (can_boost(p->addr)) if (can_boost(p->addr))
p->ainsn.boostable = 0; p->ainsn.boostable = 0;
else else
...@@ -365,13 +395,13 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs, ...@@ -365,13 +395,13 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
static __always_inline void clear_btf(void) static __always_inline void clear_btf(void)
{ {
if (test_thread_flag(TIF_DEBUGCTLMSR)) if (test_thread_flag(TIF_DEBUGCTLMSR))
wrmsrl(MSR_IA32_DEBUGCTLMSR, 0); wrmsr(MSR_IA32_DEBUGCTLMSR, 0, 0);
} }
static __always_inline void restore_btf(void) static __always_inline void restore_btf(void)
{ {
if (test_thread_flag(TIF_DEBUGCTLMSR)) if (test_thread_flag(TIF_DEBUGCTLMSR))
wrmsrl(MSR_IA32_DEBUGCTLMSR, current->thread.debugctlmsr); wrmsr(MSR_IA32_DEBUGCTLMSR, current->thread.debugctlmsr, 0);
} }
static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs) static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
...@@ -427,6 +457,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) ...@@ -427,6 +457,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
regs->flags &= ~TF_MASK; regs->flags &= ~TF_MASK;
regs->flags |= kcb->kprobe_saved_flags; regs->flags |= kcb->kprobe_saved_flags;
goto no_kprobe; goto no_kprobe;
#ifdef CONFIG_X86_64
} else if (kcb->kprobe_status == KPROBE_HIT_SSDONE) { } else if (kcb->kprobe_status == KPROBE_HIT_SSDONE) {
/* TODO: Provide re-entrancy from /* TODO: Provide re-entrancy from
* post_kprobes_handler() and avoid exception * post_kprobes_handler() and avoid exception
...@@ -437,6 +468,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) ...@@ -437,6 +468,7 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
regs->ip = (unsigned long)p->addr; regs->ip = (unsigned long)p->addr;
reset_current_kprobe(); reset_current_kprobe();
return 1; return 1;
#endif
} }
/* We have reentered the kprobe_handler(), since /* We have reentered the kprobe_handler(), since
* another probe was hit while within the handler. * another probe was hit while within the handler.
...@@ -461,10 +493,9 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) ...@@ -461,10 +493,9 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
goto no_kprobe; goto no_kprobe;
} }
p = __get_cpu_var(current_kprobe); p = __get_cpu_var(current_kprobe);
if (p->break_handler && p->break_handler(p, regs)) { if (p->break_handler && p->break_handler(p, regs))
goto ss_probe; goto ss_probe;
} }
}
goto no_kprobe; goto no_kprobe;
} }
...@@ -519,8 +550,10 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) ...@@ -519,8 +550,10 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
*/ */
void __kprobes kretprobe_trampoline_holder(void) void __kprobes kretprobe_trampoline_holder(void)
{ {
asm volatile ( ".global kretprobe_trampoline\n" asm volatile (
".global kretprobe_trampoline\n"
"kretprobe_trampoline: \n" "kretprobe_trampoline: \n"
#ifdef CONFIG_X86_64
/* We don't bother saving the ss register */ /* We don't bother saving the ss register */
" pushq %rsp\n" " pushq %rsp\n"
" pushfq\n" " pushfq\n"
...@@ -566,25 +599,64 @@ static int __kprobes kprobe_handler(struct pt_regs *regs) ...@@ -566,25 +599,64 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
/* Skip orig_ax, ip, cs */ /* Skip orig_ax, ip, cs */
" addq $24, %rsp\n" " addq $24, %rsp\n"
" popfq\n" " popfq\n"
#else
" pushf\n"
/*
* Skip cs, ip, orig_ax.
* trampoline_handler() will plug in these values
*/
" subl $12, %esp\n"
" pushl %fs\n"
" pushl %ds\n"
" pushl %es\n"
" pushl %eax\n"
" pushl %ebp\n"
" pushl %edi\n"
" pushl %esi\n"
" pushl %edx\n"
" pushl %ecx\n"
" pushl %ebx\n"
" movl %esp, %eax\n"
" call trampoline_handler\n"
/* Move flags to cs */
" movl 52(%esp), %edx\n"
" movl %edx, 48(%esp)\n"
/* Replace saved flags with true return address. */
" movl %eax, 52(%esp)\n"
" popl %ebx\n"
" popl %ecx\n"
" popl %edx\n"
" popl %esi\n"
" popl %edi\n"
" popl %ebp\n"
" popl %eax\n"
/* Skip ip, orig_ax, es, ds, fs */
" addl $20, %esp\n"
" popf\n"
#endif
" ret\n"); " ret\n");
} }
/* /*
* Called from kretprobe_trampoline * Called from kretprobe_trampoline
*/ */
fastcall void * __kprobes trampoline_handler(struct pt_regs *regs) void * __kprobes trampoline_handler(struct pt_regs *regs)
{ {
struct kretprobe_instance *ri = NULL; struct kretprobe_instance *ri = NULL;
struct hlist_head *head, empty_rp; struct hlist_head *head, empty_rp;
struct hlist_node *node, *tmp; struct hlist_node *node, *tmp;
unsigned long flags, orig_ret_address = 0; unsigned long flags, orig_ret_address = 0;
unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline; unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
INIT_HLIST_HEAD(&empty_rp); INIT_HLIST_HEAD(&empty_rp);
spin_lock_irqsave(&kretprobe_lock, flags); spin_lock_irqsave(&kretprobe_lock, flags);
head = kretprobe_inst_table_head(current); head = kretprobe_inst_table_head(current);
/* fixup registers */ /* fixup registers */
#ifdef CONFIG_X86_64
regs->cs = __KERNEL_CS; regs->cs = __KERNEL_CS;
#else
regs->cs = __KERNEL_CS | get_kernel_rpl();
#endif
regs->ip = trampoline_address; regs->ip = trampoline_address;
regs->orig_ax = ~0UL; regs->orig_ax = ~0UL;
...@@ -671,9 +743,11 @@ static void __kprobes resume_execution(struct kprobe *p, ...@@ -671,9 +743,11 @@ static void __kprobes resume_execution(struct kprobe *p,
unsigned long orig_ip = (unsigned long)p->addr; unsigned long orig_ip = (unsigned long)p->addr;
kprobe_opcode_t *insn = p->ainsn.insn; kprobe_opcode_t *insn = p->ainsn.insn;
#ifdef CONFIG_X86_64
/*skip the REX prefix*/ /*skip the REX prefix*/
if (*insn >= 0x40 && *insn <= 0x4f) if (*insn >= 0x40 && *insn <= 0x4f)
insn++; insn++;
#endif
regs->flags &= ~TF_MASK; regs->flags &= ~TF_MASK;
switch (*insn) { switch (*insn) {
...@@ -693,6 +767,11 @@ static void __kprobes resume_execution(struct kprobe *p, ...@@ -693,6 +767,11 @@ static void __kprobes resume_execution(struct kprobe *p,
case 0xe8: /* call relative - Fix return addr */ case 0xe8: /* call relative - Fix return addr */
*tos = orig_ip + (*tos - copy_ip); *tos = orig_ip + (*tos - copy_ip);
break; break;
#ifndef CONFIG_X86_64
case 0x9a: /* call absolute -- same as call absolute, indirect */
*tos = orig_ip + (*tos - copy_ip);
goto no_change;
#endif
case 0xff: case 0xff:
if ((insn[1] & 0x30) == 0x10) { if ((insn[1] & 0x30) == 0x10) {
/* /*
...@@ -783,9 +862,8 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) ...@@ -783,9 +862,8 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
{ {
struct kprobe *cur = kprobe_running(); struct kprobe *cur = kprobe_running();
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
const struct exception_table_entry *fixup;
switch(kcb->kprobe_status) { switch (kcb->kprobe_status) {
case KPROBE_HIT_SS: case KPROBE_HIT_SS:
case KPROBE_REENTER: case KPROBE_REENTER:
/* /*
...@@ -826,12 +904,19 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr) ...@@ -826,12 +904,19 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
* In case the user-specified fault handler returned * In case the user-specified fault handler returned
* zero, try to fix up. * zero, try to fix up.
*/ */
#ifdef CONFIG_X86_64
{
const struct exception_table_entry *fixup;
fixup = search_exception_tables(regs->ip); fixup = search_exception_tables(regs->ip);
if (fixup) { if (fixup) {
regs->ip = fixup->fixup; regs->ip = fixup->fixup;
return 1; return 1;
} }
}
#else
if (fixup_exception(regs))
return 1;
#endif
/* /*
* fixup routine could not handle it, * fixup routine could not handle it,
* Let do_page_fault() fix it. * Let do_page_fault() fix it.
...@@ -907,9 +992,14 @@ void __kprobes jprobe_return(void) ...@@ -907,9 +992,14 @@ void __kprobes jprobe_return(void)
{ {
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk(); struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
asm volatile (" xchg %%rbx,%%rsp \n" asm volatile (
#ifdef CONFIG_X86_64
" xchg %%rbx,%%rsp \n"
#else
" xchgl %%ebx,%%esp \n"
#endif
" int3 \n" " int3 \n"
" .globl jprobe_return_end \n" " .globl jprobe_return_end\n"
" jprobe_return_end: \n" " jprobe_return_end: \n"
" nop \n"::"b" " nop \n"::"b"
(kcb->jprobe_saved_sp):"memory"); (kcb->jprobe_saved_sp):"memory");
...@@ -921,14 +1011,16 @@ int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs) ...@@ -921,14 +1011,16 @@ int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
u8 *addr = (u8 *) (regs->ip - 1); u8 *addr = (u8 *) (regs->ip - 1);
struct jprobe *jp = container_of(p, struct jprobe, kp); struct jprobe *jp = container_of(p, struct jprobe, kp);
if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) { if ((addr > (u8 *) jprobe_return) &&
(addr < (u8 *) jprobe_return_end)) {
if (stack_addr(regs) != kcb->jprobe_saved_sp) { if (stack_addr(regs) != kcb->jprobe_saved_sp) {
struct pt_regs *saved_regs = &kcb->jprobe_saved_regs; struct pt_regs *saved_regs = &kcb->jprobe_saved_regs;
printk("current sp %p does not match saved sp %p\n", printk(KERN_ERR
"current sp %p does not match saved sp %p\n",
stack_addr(regs), kcb->jprobe_saved_sp); stack_addr(regs), kcb->jprobe_saved_sp);
printk("Saved registers for jprobe %p\n", jp); printk(KERN_ERR "Saved registers for jprobe %p\n", jp);
show_registers(saved_regs); show_registers(saved_regs);
printk("Current registers\n"); printk(KERN_ERR "Current registers\n");
show_registers(regs); show_registers(regs);
BUG(); BUG();
} }
......
/*
* Kernel Probes (KProbes)
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
* Copyright (C) IBM Corporation, 2002, 2004
*
* 2002-Oct Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
* Probes initial implementation ( includes contributions from
* Rusty Russell).
* 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
* interface to access function arguments.
* 2005-May Hien Nguyen <hien@us.ibm.com>, Jim Keniston
* <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
* <prasanna@in.ibm.com> added function-return probes.
*/
#include <linux/kprobes.h>
#include <linux/ptrace.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/preempt.h>
#include <linux/module.h>
#include <linux/kdebug.h>
#include <asm/cacheflush.h>
#include <asm/desc.h>
#include <asm/pgtable.h>
#include <asm/uaccess.h>
#include <asm/alternative.h>
void jprobe_return_end(void);
DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
/*
* "&regs->sp" looks wrong, but it's correct for x86_32. x86_32 CPUs
* don't save the ss and esp registers if the CPU is already in kernel
* mode when it traps. So for kprobes, regs->sp and regs->ss are not
* the [nonexistent] saved stack pointer and ss register, but rather
* the top 8 bytes of the pre-int3 stack. So &regs->sp happens to
* point to the top of the pre-int3 stack.
*/
#define stack_addr(regs) ((unsigned long *)&regs->sp)
#define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\
(((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) | \
(b4##UL << 0x4)|(b5##UL << 0x5)|(b6##UL << 0x6)|(b7##UL << 0x7) | \
(b8##UL << 0x8)|(b9##UL << 0x9)|(ba##UL << 0xa)|(bb##UL << 0xb) | \
(bc##UL << 0xc)|(bd##UL << 0xd)|(be##UL << 0xe)|(bf##UL << 0xf)) \
<< (row % 32))
/*
* Undefined/reserved opcodes, conditional jump, Opcode Extension
* Groups, and some special opcodes can not boost.
*/
static const u32 twobyte_is_boostable[256 / 32] = {
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
/* ---------------------------------------------- */
W(0x00, 0, 0, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0) | /* 00 */
W(0x10, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 10 */
W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 20 */
W(0x30, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 30 */
W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 40 */
W(0x50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 50 */
W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1) | /* 60 */
W(0x70, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1) , /* 70 */
W(0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 80 */
W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */
W(0xa0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* a0 */
W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1) , /* b0 */
W(0xc0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* c0 */
W(0xd0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) , /* d0 */
W(0xe0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* e0 */
W(0xf0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 0) /* f0 */
/* ----------------------------------------------- */
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
};
static const u32 onebyte_has_modrm[256 / 32] = {
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
/* ----------------------------------------------- */
W(0x00, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 00 */
W(0x10, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) , /* 10 */
W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 20 */
W(0x30, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) , /* 30 */
W(0x40, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 40 */
W(0x50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 50 */
W(0x60, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0) | /* 60 */
W(0x70, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 70 */
W(0x80, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 80 */
W(0x90, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 90 */
W(0xa0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* a0 */
W(0xb0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* b0 */
W(0xc0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0) | /* c0 */
W(0xd0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) , /* d0 */
W(0xe0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* e0 */
W(0xf0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1) /* f0 */
/* ----------------------------------------------- */
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
};
static const u32 twobyte_has_modrm[256 / 32] = {
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
/* ----------------------------------------------- */
W(0x00, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1) | /* 0f */
W(0x10, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0) , /* 1f */
W(0x20, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* 2f */
W(0x30, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 3f */
W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 4f */
W(0x50, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 5f */
W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 6f */
W(0x70, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1) , /* 7f */
W(0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 8f */
W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 9f */
W(0xa0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1) | /* af */
W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1) , /* bf */
W(0xc0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0) | /* cf */
W(0xd0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* df */
W(0xe0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* ef */
W(0xf0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0) /* ff */
/* ----------------------------------------------- */
/* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
};
#undef W
struct kretprobe_blackpoint kretprobe_blacklist[] = {
{"__switch_to", }, /* This function switches only current task, but
doesn't switch kernel stack.*/
{NULL, NULL} /* Terminator */
};
const int kretprobe_blacklist_size = ARRAY_SIZE(kretprobe_blacklist);
/* Insert a jump instruction at address 'from', which jumps to address 'to'.*/
static __always_inline void set_jmp_op(void *from, void *to)
{
struct __arch_jmp_op {
char op;
s32 raddr;
} __attribute__((packed)) * jop;
jop = (struct __arch_jmp_op *)from;
jop->raddr = (s32)((long)(to) - ((long)(from) + 5));
jop->op = RELATIVEJUMP_INSTRUCTION;
}
/*
* returns non-zero if opcode is boostable.
*/
static __always_inline int can_boost(kprobe_opcode_t *opcodes)
{
kprobe_opcode_t opcode;
kprobe_opcode_t *orig_opcodes = opcodes;
retry:
if (opcodes - orig_opcodes > MAX_INSN_SIZE - 1)
return 0;
opcode = *(opcodes++);
/* 2nd-byte opcode */
if (opcode == 0x0f) {
if (opcodes - orig_opcodes > MAX_INSN_SIZE - 1)
return 0;
return test_bit(*opcodes,
(unsigned long *)twobyte_is_boostable);
}
switch (opcode & 0xf0) {
case 0x60:
if (0x63 < opcode && opcode < 0x67)
goto retry; /* prefixes */
/* can't boost Address-size override and bound */
return (opcode != 0x62 && opcode != 0x67);
case 0x70:
return 0; /* can't boost conditional jump */
case 0xc0:
/* can't boost software-interruptions */
return (0xc1 < opcode && opcode < 0xcc) || opcode == 0xcf;
case 0xd0:
/* can boost AA* and XLAT */
return (opcode == 0xd4 || opcode == 0xd5 || opcode == 0xd7);
case 0xe0:
/* can boost in/out and absolute jmps */
return ((opcode & 0x04) || opcode == 0xea);
case 0xf0:
if ((opcode & 0x0c) == 0 && opcode != 0xf1)
goto retry; /* lock/rep(ne) prefix */
/* clear and set flags are boostable */
return (opcode == 0xf5 || (0xf7 < opcode && opcode < 0xfe));
default:
/* segment override prefixes are boostable */
if (opcode == 0x26 || opcode == 0x36 || opcode == 0x3e)
goto retry; /* prefixes */
/* CS override prefix and call are not boostable */
return (opcode != 0x2e && opcode != 0x9a);
}
}
/*
* returns non-zero if opcode modifies the interrupt flag.
*/
static int __kprobes is_IF_modifier(kprobe_opcode_t *insn)
{
switch (*insn) {
case 0xfa: /* cli */
case 0xfb: /* sti */
case 0xcf: /* iret/iretd */
case 0x9d: /* popf/popfd */
return 1;
}
return 0;
}
static void __kprobes arch_copy_kprobe(struct kprobe *p)
{
memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
if (can_boost(p->addr))
p->ainsn.boostable = 0;
else
p->ainsn.boostable = -1;
p->opcode = *p->addr;
}
int __kprobes arch_prepare_kprobe(struct kprobe *p)
{
/* insn: must be on special executable page on x86. */
p->ainsn.insn = get_insn_slot();
if (!p->ainsn.insn)
return -ENOMEM;
arch_copy_kprobe(p);
return 0;
}
void __kprobes arch_arm_kprobe(struct kprobe *p)
{
text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1);
}
void __kprobes arch_disarm_kprobe(struct kprobe *p)
{
text_poke(p->addr, &p->opcode, 1);
}
void __kprobes arch_remove_kprobe(struct kprobe *p)
{
mutex_lock(&kprobe_mutex);
free_insn_slot(p->ainsn.insn, (p->ainsn.boostable == 1));
mutex_unlock(&kprobe_mutex);
}
static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
{
kcb->prev_kprobe.kp = kprobe_running();
kcb->prev_kprobe.status = kcb->kprobe_status;
kcb->prev_kprobe.old_flags = kcb->kprobe_old_flags;
kcb->prev_kprobe.saved_flags = kcb->kprobe_saved_flags;
}
static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
{
__get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
kcb->kprobe_status = kcb->prev_kprobe.status;
kcb->kprobe_old_flags = kcb->prev_kprobe.old_flags;
kcb->kprobe_saved_flags = kcb->prev_kprobe.saved_flags;
}
static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
struct kprobe_ctlblk *kcb)
{
__get_cpu_var(current_kprobe) = p;
kcb->kprobe_saved_flags = kcb->kprobe_old_flags
= (regs->flags & (TF_MASK | IF_MASK));
if (is_IF_modifier(p->ainsn.insn))
kcb->kprobe_saved_flags &= ~IF_MASK;
}
static __always_inline void clear_btf(void)
{
if (test_thread_flag(TIF_DEBUGCTLMSR))
wrmsr(MSR_IA32_DEBUGCTLMSR, 0, 0);
}
static __always_inline void restore_btf(void)
{
if (test_thread_flag(TIF_DEBUGCTLMSR))
wrmsr(MSR_IA32_DEBUGCTLMSR, current->thread.debugctlmsr, 0);
}
static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
{
clear_btf();
regs->flags |= TF_MASK;
regs->flags &= ~IF_MASK;
/*single step inline if the instruction is an int3*/
if (p->opcode == BREAKPOINT_INSTRUCTION)
regs->ip = (unsigned long)p->addr;
else
regs->ip = (unsigned long)p->ainsn.insn;
}
/* Called with kretprobe_lock held */
void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
struct pt_regs *regs)
{
unsigned long *sara = stack_addr(regs);
ri->ret_addr = (kprobe_opcode_t *) *sara;
/* Replace the return addr with trampoline addr */
*sara = (unsigned long) &kretprobe_trampoline;
}
/*
* Interrupts are disabled on entry as trap3 is an interrupt gate and they
* remain disabled thorough out this function.
*/
static int __kprobes kprobe_handler(struct pt_regs *regs)
{
struct kprobe *p;
int ret = 0;
kprobe_opcode_t *addr;
struct kprobe_ctlblk *kcb;
addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t));
/*
* We don't want to be preempted for the entire
* duration of kprobe processing
*/
preempt_disable();
kcb = get_kprobe_ctlblk();
/* Check we're not actually recursing */
if (kprobe_running()) {
p = get_kprobe(addr);
if (p) {
if (kcb->kprobe_status == KPROBE_HIT_SS &&
*p->ainsn.insn == BREAKPOINT_INSTRUCTION) {
regs->flags &= ~TF_MASK;
regs->flags |= kcb->kprobe_saved_flags;
goto no_kprobe;
}
/* We have reentered the kprobe_handler(), since
* another probe was hit while within the handler.
* We here save the original kprobes variables and
* just single step on the instruction of the new probe
* without calling any user handlers.
*/
save_previous_kprobe(kcb);
set_current_kprobe(p, regs, kcb);
kprobes_inc_nmissed_count(p);
prepare_singlestep(p, regs);
kcb->kprobe_status = KPROBE_REENTER;
return 1;
} else {
if (*addr != BREAKPOINT_INSTRUCTION) {
/* The breakpoint instruction was removed by
* another cpu right after we hit, no further
* handling of this interrupt is appropriate
*/
regs->ip = (unsigned long)addr;
ret = 1;
goto no_kprobe;
}
p = __get_cpu_var(current_kprobe);
if (p->break_handler && p->break_handler(p, regs)) {
goto ss_probe;
}
}
goto no_kprobe;
}
p = get_kprobe(addr);
if (!p) {
if (*addr != BREAKPOINT_INSTRUCTION) {
/*
* The breakpoint instruction was removed right
* after we hit it. Another cpu has removed
* either a probepoint or a debugger breakpoint
* at this address. In either case, no further
* handling of this interrupt is appropriate.
* Back up over the (now missing) int3 and run
* the original instruction.
*/
regs->ip = (unsigned long)addr;
ret = 1;
}
/* Not one of ours: let kernel handle it */
goto no_kprobe;
}
set_current_kprobe(p, regs, kcb);
kcb->kprobe_status = KPROBE_HIT_ACTIVE;
if (p->pre_handler && p->pre_handler(p, regs))
/* handler has already set things up, so skip ss setup */
return 1;
ss_probe:
#if !defined(CONFIG_PREEMPT) || defined(CONFIG_PM)
if (p->ainsn.boostable == 1 && !p->post_handler) {
/* Boost up -- we can execute copied instructions directly */
reset_current_kprobe();
regs->ip = (unsigned long)p->ainsn.insn;
preempt_enable_no_resched();
return 1;
}
#endif
prepare_singlestep(p, regs);
kcb->kprobe_status = KPROBE_HIT_SS;
return 1;
no_kprobe:
preempt_enable_no_resched();
return ret;
}
/*
* When a retprobed function returns, this code saves registers and
* calls trampoline_handler() runs, which calls the kretprobe's handler.
*/
void __kprobes kretprobe_trampoline_holder(void)
{
asm volatile ( ".global kretprobe_trampoline\n"
"kretprobe_trampoline: \n"
" pushf\n"
/*
* Skip cs, ip, orig_ax.
* trampoline_handler() will plug in these values
*/
" subl $12, %esp\n"
" pushl %fs\n"
" pushl %ds\n"
" pushl %es\n"
" pushl %eax\n"
" pushl %ebp\n"
" pushl %edi\n"
" pushl %esi\n"
" pushl %edx\n"
" pushl %ecx\n"
" pushl %ebx\n"
" movl %esp, %eax\n"
" call trampoline_handler\n"
/* Move flags to cs */
" movl 52(%esp), %edx\n"
" movl %edx, 48(%esp)\n"
/* Replace saved flags with true return address. */
" movl %eax, 52(%esp)\n"
" popl %ebx\n"
" popl %ecx\n"
" popl %edx\n"
" popl %esi\n"
" popl %edi\n"
" popl %ebp\n"
" popl %eax\n"
/* Skip ip, orig_ax, es, ds, fs */
" addl $20, %esp\n"
" popf\n"
" ret\n");
}
/*
* Called from kretprobe_trampoline
*/
void * __kprobes trampoline_handler(struct pt_regs *regs)
{
struct kretprobe_instance *ri = NULL;
struct hlist_head *head, empty_rp;
struct hlist_node *node, *tmp;
unsigned long flags, orig_ret_address = 0;
unsigned long trampoline_address =(unsigned long)&kretprobe_trampoline;
INIT_HLIST_HEAD(&empty_rp);
spin_lock_irqsave(&kretprobe_lock, flags);
head = kretprobe_inst_table_head(current);
/* fixup registers */
regs->cs = __KERNEL_CS | get_kernel_rpl();
regs->ip = trampoline_address;
regs->orig_ax = ~0UL;
/*
* It is possible to have multiple instances associated with a given
* task either because multiple functions in the call path have
* return probes installed on them, and/or more then one
* return probe was registered for a target function.
*
* We can handle this because:
* - instances are always pushed into the head of the list
* - when multiple return probes are registered for the same
* function, the (chronologically) first instance's ret_addr
* will be the real return address, and all the rest will
* point to kretprobe_trampoline.
*/
hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
if (ri->task != current)
/* another task is sharing our hash bucket */
continue;
if (ri->rp && ri->rp->handler) {
__get_cpu_var(current_kprobe) = &ri->rp->kp;
get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
ri->rp->handler(ri, regs);
__get_cpu_var(current_kprobe) = NULL;
}
orig_ret_address = (unsigned long)ri->ret_addr;
recycle_rp_inst(ri, &empty_rp);
if (orig_ret_address != trampoline_address)
/*
* This is the real return address. Any other
* instances associated with this task are for
* other calls deeper on the call stack
*/
break;
}
kretprobe_assert(ri, orig_ret_address, trampoline_address);
spin_unlock_irqrestore(&kretprobe_lock, flags);
hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
hlist_del(&ri->hlist);
kfree(ri);
}
return (void *)orig_ret_address;
}
/*
* Called after single-stepping. p->addr is the address of the
* instruction whose first byte has been replaced by the "int 3"
* instruction. To avoid the SMP problems that can occur when we
* temporarily put back the original opcode to single-step, we
* single-stepped a copy of the instruction. The address of this
* copy is p->ainsn.insn.
*
* This function prepares to return from the post-single-step
* interrupt. We have to fix up the stack as follows:
*
* 0) Except in the case of absolute or indirect jump or call instructions,
* the new ip is relative to the copied instruction. We need to make
* it relative to the original instruction.
*
* 1) If the single-stepped instruction was pushfl, then the TF and IF
* flags are set in the just-pushed flags, and may need to be cleared.
*
* 2) If the single-stepped instruction was a call, the return address
* that is atop the stack is the address following the copied instruction.
* We need to make it the address following the original instruction.
*
* If this is the first time we've single-stepped the instruction at
* this probepoint, and the instruction is boostable, boost it: add a
* jump instruction after the copied instruction, that jumps to the next
* instruction after the probepoint.
*/
static void __kprobes resume_execution(struct kprobe *p,
struct pt_regs *regs, struct kprobe_ctlblk *kcb)
{
unsigned long *tos = stack_addr(regs);
unsigned long copy_ip = (unsigned long)p->ainsn.insn;
unsigned long orig_ip = (unsigned long)p->addr;
kprobe_opcode_t *insn = p->ainsn.insn;
regs->flags &= ~TF_MASK;
switch (*insn) {
case 0x9c: /* pushfl */
*tos &= ~(TF_MASK | IF_MASK);
*tos |= kcb->kprobe_old_flags;
break;
case 0xc2: /* iret/ret/lret */
case 0xc3:
case 0xca:
case 0xcb:
case 0xcf:
case 0xea: /* jmp absolute -- ip is correct */
/* ip is already adjusted, no more changes required */
p->ainsn.boostable = 1;
goto no_change;
case 0xe8: /* call relative - Fix return addr */
*tos = orig_ip + (*tos - copy_ip);
break;
case 0x9a: /* call absolute -- same as call absolute, indirect */
*tos = orig_ip + (*tos - copy_ip);
goto no_change;
case 0xff:
if ((insn[1] & 0x30) == 0x10) {
/*
* call absolute, indirect
* Fix return addr; ip is correct.
* But this is not boostable
*/
*tos = orig_ip + (*tos - copy_ip);
goto no_change;
} else if (((insn[1] & 0x31) == 0x20) ||
((insn[1] & 0x31) == 0x21)) {
/*
* jmp near and far, absolute indirect
* ip is correct. And this is boostable
*/
p->ainsn.boostable = 1;
goto no_change;
}
default:
break;
}
if (p->ainsn.boostable == 0) {
if ((regs->ip > copy_ip) &&
(regs->ip - copy_ip) + 5 < MAX_INSN_SIZE) {
/*
* These instructions can be executed directly if it
* jumps back to correct address.
*/
set_jmp_op((void *)regs->ip,
(void *)orig_ip + (regs->ip - copy_ip));
p->ainsn.boostable = 1;
} else {
p->ainsn.boostable = -1;
}
}
regs->ip += orig_ip - copy_ip;
no_change:
restore_btf();
return;
}
/*
* Interrupts are disabled on entry as trap1 is an interrupt gate and they
* remain disabled thoroughout this function.
*/
static int __kprobes post_kprobe_handler(struct pt_regs *regs)
{
struct kprobe *cur = kprobe_running();
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
if (!cur)
return 0;
if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
kcb->kprobe_status = KPROBE_HIT_SSDONE;
cur->post_handler(cur, regs, 0);
}
resume_execution(cur, regs, kcb);
regs->flags |= kcb->kprobe_saved_flags;
trace_hardirqs_fixup_flags(regs->flags);
/* Restore back the original saved kprobes variables and continue. */
if (kcb->kprobe_status == KPROBE_REENTER) {
restore_previous_kprobe(kcb);
goto out;
}
reset_current_kprobe();
out:
preempt_enable_no_resched();
/*
* if somebody else is singlestepping across a probe point, flags
* will have TF set, in which case, continue the remaining processing
* of do_debug, as if this is not a probe hit.
*/
if (regs->flags & TF_MASK)
return 0;
return 1;
}
int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
{
struct kprobe *cur = kprobe_running();
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
switch(kcb->kprobe_status) {
case KPROBE_HIT_SS:
case KPROBE_REENTER:
/*
* We are here because the instruction being single
* stepped caused a page fault. We reset the current
* kprobe and the ip points back to the probe address
* and allow the page fault handler to continue as a
* normal page fault.
*/
regs->ip = (unsigned long)cur->addr;
regs->flags |= kcb->kprobe_old_flags;
if (kcb->kprobe_status == KPROBE_REENTER)
restore_previous_kprobe(kcb);
else
reset_current_kprobe();
preempt_enable_no_resched();
break;
case KPROBE_HIT_ACTIVE:
case KPROBE_HIT_SSDONE:
/*
* We increment the nmissed count for accounting,
* we can also use npre/npostfault count for accounting
* these specific fault cases.
*/
kprobes_inc_nmissed_count(cur);
/*
* We come here because instructions in the pre/post
* handler caused the page_fault, this could happen
* if handler tries to access user space by
* copy_from_user(), get_user() etc. Let the
* user-specified handler try to fix it first.
*/
if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
return 1;
/*
* In case the user-specified fault handler returned
* zero, try to fix up.
*/
if (fixup_exception(regs))
return 1;
/*
* fixup routine could not handle it,
* Let do_page_fault() fix it.
*/
break;
default:
break;
}
return 0;
}
/*
* Wrapper routine for handling exceptions.
*/
int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data)
{
struct die_args *args = (struct die_args *)data;
int ret = NOTIFY_DONE;
if (args->regs && user_mode_vm(args->regs))
return ret;
switch (val) {
case DIE_INT3:
if (kprobe_handler(args->regs))
ret = NOTIFY_STOP;
break;
case DIE_DEBUG:
if (post_kprobe_handler(args->regs))
ret = NOTIFY_STOP;
break;
case DIE_GPF:
/* kprobe_running() needs smp_processor_id() */
preempt_disable();
if (kprobe_running() &&
kprobe_fault_handler(args->regs, args->trapnr))
ret = NOTIFY_STOP;
preempt_enable();
break;
default:
break;
}
return ret;
}
int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
{
struct jprobe *jp = container_of(p, struct jprobe, kp);
unsigned long addr;
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
kcb->jprobe_saved_regs = *regs;
kcb->jprobe_saved_sp = stack_addr(regs);
addr = (unsigned long)(kcb->jprobe_saved_sp);
/*
* As Linus pointed out, gcc assumes that the callee
* owns the argument space and could overwrite it, e.g.
* tailcall optimization. So, to be absolutely safe
* we also save and restore enough stack bytes to cover
* the argument area.
*/
memcpy(kcb->jprobes_stack, (kprobe_opcode_t *)addr,
MIN_STACK_SIZE(addr));
regs->flags &= ~IF_MASK;
trace_hardirqs_off();
regs->ip = (unsigned long)(jp->entry);
return 1;
}
void __kprobes jprobe_return(void)
{
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
asm volatile (" xchgl %%ebx,%%esp \n"
" int3 \n"
" .globl jprobe_return_end \n"
" jprobe_return_end: \n"
" nop \n"::"b"
(kcb->jprobe_saved_sp):"memory");
}
int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
{
struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
u8 *addr = (u8 *) (regs->ip - 1);
struct jprobe *jp = container_of(p, struct jprobe, kp);
if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) {
if (stack_addr(regs) != kcb->jprobe_saved_sp) {
struct pt_regs *saved_regs = &kcb->jprobe_saved_regs;
printk("current sp %p does not match saved sp %p\n",
stack_addr(regs), kcb->jprobe_saved_sp);
printk("Saved registers for jprobe %p\n", jp);
show_registers(saved_regs);
printk("Current registers\n");
show_registers(regs);
BUG();
}
*regs = kcb->jprobe_saved_regs;
memcpy((kprobe_opcode_t *)(kcb->jprobe_saved_sp),
kcb->jprobes_stack,
MIN_STACK_SIZE(kcb->jprobe_saved_sp));
preempt_enable_no_resched();
return 1;
}
return 0;
}
int __init arch_init_kprobes(void)
{
return 0;
}
int __kprobes arch_trampoline_kprobe(struct kprobe *p)
{
return 0;
}
#ifdef CONFIG_X86_32 #ifndef _ASM_KPROBES_H
# include "kprobes_32.h" #define _ASM_KPROBES_H
#else /*
# include "kprobes_64.h" * Kernel Probes (KProbes)
#endif *
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
* Copyright (C) IBM Corporation, 2002, 2004
*
* See arch/x86/kernel/kprobes.c for x86 kprobes history.
*/
#include <linux/types.h>
#include <linux/ptrace.h>
#include <linux/percpu.h>
#define __ARCH_WANT_KPROBES_INSN_SLOT
struct pt_regs;
struct kprobe;
typedef u8 kprobe_opcode_t;
#define BREAKPOINT_INSTRUCTION 0xcc
#define RELATIVEJUMP_INSTRUCTION 0xe9
#define MAX_INSN_SIZE 16
#define MAX_STACK_SIZE 64
#define MIN_STACK_SIZE(ADDR) (((MAX_STACK_SIZE) < \
(((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR))) \
? (MAX_STACK_SIZE) \
: (((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR)))
#define ARCH_SUPPORTS_KRETPROBES
#define flush_insn_slot(p) do { } while (0)
extern const int kretprobe_blacklist_size;
void arch_remove_kprobe(struct kprobe *p);
void kretprobe_trampoline(void);
/* Architecture specific copy of original instruction*/
struct arch_specific_insn {
/* copy of the original instruction */
kprobe_opcode_t *insn;
/*
* boostable = -1: This instruction type is not boostable.
* boostable = 0: This instruction type is boostable.
* boostable = 1: This instruction has been boosted: we have
* added a relative jump after the instruction copy in insn,
* so no single-step and fixup are needed (unless there's
* a post_handler or break_handler).
*/
int boostable;
};
struct prev_kprobe {
struct kprobe *kp;
unsigned long status;
unsigned long old_flags;
unsigned long saved_flags;
};
/* per-cpu kprobe control block */
struct kprobe_ctlblk {
unsigned long kprobe_status;
unsigned long kprobe_old_flags;
unsigned long kprobe_saved_flags;
unsigned long *jprobe_saved_sp;
struct pt_regs jprobe_saved_regs;
kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
struct prev_kprobe prev_kprobe;
};
/* trap3/1 are intr gates for kprobes. So, restore the status of IF,
* if necessary, before executing the original int3/1 (trap) handler.
*/
static inline void restore_interrupts(struct pt_regs *regs)
{
if (regs->flags & IF_MASK)
local_irq_enable();
}
extern int kprobe_fault_handler(struct pt_regs *regs, int trapnr);
extern int kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data);
#endif /* _ASM_KPROBES_H */
#ifndef _ASM_KPROBES_H
#define _ASM_KPROBES_H
/*
* Kernel Probes (KProbes)
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
* Copyright (C) IBM Corporation, 2002, 2004
*
* 2002-Oct Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
* Probes initial implementation ( includes suggestions from
* Rusty Russell).
* 2004-Oct Prasanna S Panchamukhi <prasanna@in.ibm.com> and Jim Keniston
* kenistoj@us.ibm.com adopted from i386.
*/
#include <linux/types.h>
#include <linux/ptrace.h>
#include <linux/percpu.h>
#define __ARCH_WANT_KPROBES_INSN_SLOT
struct pt_regs;
struct kprobe;
typedef u8 kprobe_opcode_t;
#define BREAKPOINT_INSTRUCTION 0xcc
#define RELATIVEJUMP_INSTRUCTION 0xe9
#define MAX_INSN_SIZE 16
#define MAX_STACK_SIZE 64
#define MIN_STACK_SIZE(ADDR) (((MAX_STACK_SIZE) < \
(((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR))) \
? (MAX_STACK_SIZE) \
: (((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR)))
#define ARCH_SUPPORTS_KRETPROBES
#define flush_insn_slot(p) do { } while (0)
extern const int kretprobe_blacklist_size;
void arch_remove_kprobe(struct kprobe *p);
void kretprobe_trampoline(void);
/* Architecture specific copy of original instruction*/
struct arch_specific_insn {
/* copy of the original instruction */
kprobe_opcode_t *insn;
/*
* boostable = -1: This instruction type is not boostable.
* boostable = 0: This instruction type is boostable.
* boostable = 1: This instruction has been boosted: we have
* added a relative jump after the instruction copy in insn,
* so no single-step and fixup are needed (unless there's
* a post_handler or break_handler).
*/
int boostable;
};
struct prev_kprobe {
struct kprobe *kp;
unsigned long status;
unsigned long old_flags;
unsigned long saved_flags;
};
/* per-cpu kprobe control block */
struct kprobe_ctlblk {
unsigned long kprobe_status;
unsigned long kprobe_old_flags;
unsigned long kprobe_saved_flags;
unsigned long *jprobe_saved_sp;
struct pt_regs jprobe_saved_regs;
kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
struct prev_kprobe prev_kprobe;
};
/* trap3/1 are intr gates for kprobes. So, restore the status of IF,
* if necessary, before executing the original int3/1 (trap) handler.
*/
static inline void restore_interrupts(struct pt_regs *regs)
{
if (regs->flags & IF_MASK)
local_irq_enable();
}
extern int kprobe_fault_handler(struct pt_regs *regs, int trapnr);
extern int kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data);
#endif /* _ASM_KPROBES_H */
#ifndef _ASM_KPROBES_H
#define _ASM_KPROBES_H
/*
* Kernel Probes (KProbes)
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
*
* Copyright (C) IBM Corporation, 2002, 2004
*
* 2002-Oct Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
* Probes initial implementation ( includes suggestions from
* Rusty Russell).
* 2004-Oct Prasanna S Panchamukhi <prasanna@in.ibm.com> and Jim Keniston
* kenistoj@us.ibm.com adopted from i386.
*/
#include <linux/types.h>
#include <linux/ptrace.h>
#include <linux/percpu.h>
#define __ARCH_WANT_KPROBES_INSN_SLOT
struct pt_regs;
struct kprobe;
typedef u8 kprobe_opcode_t;
#define BREAKPOINT_INSTRUCTION 0xcc
#define RELATIVEJUMP_INSTRUCTION 0xe9
#define MAX_INSN_SIZE 16
#define MAX_STACK_SIZE 64
#define MIN_STACK_SIZE(ADDR) (((MAX_STACK_SIZE) < \
(((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR))) \
? (MAX_STACK_SIZE) \
: (((unsigned long)current_thread_info()) + THREAD_SIZE \
- (unsigned long)(ADDR)))
#define ARCH_SUPPORTS_KRETPROBES
#define flush_insn_slot(p) do { } while (0)
extern const int kretprobe_blacklist_size;
void arch_remove_kprobe(struct kprobe *p);
void kretprobe_trampoline(void);
/* Architecture specific copy of original instruction*/
struct arch_specific_insn {
/* copy of the original instruction */
kprobe_opcode_t *insn;
/*
* boostable = -1: This instruction type is not boostable.
* boostable = 0: This instruction type is boostable.
* boostable = 1: This instruction has been boosted: we have
* added a relative jump after the instruction copy in insn,
* so no single-step and fixup are needed (unless there's
* a post_handler or break_handler).
*/
int boostable;
};
struct prev_kprobe {
struct kprobe *kp;
unsigned long status;
unsigned long old_flags;
unsigned long saved_flags;
};
/* per-cpu kprobe control block */
struct kprobe_ctlblk {
unsigned long kprobe_status;
unsigned long kprobe_old_flags;
unsigned long kprobe_saved_flags;
unsigned long *jprobe_saved_sp;
struct pt_regs jprobe_saved_regs;
kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];
struct prev_kprobe prev_kprobe;
};
/* trap3/1 are intr gates for kprobes. So, restore the status of IF,
* if necessary, before executing the original int3/1 (trap) handler.
*/
static inline void restore_interrupts(struct pt_regs *regs)
{
if (regs->flags & IF_MASK)
local_irq_enable();
}
extern int kprobe_fault_handler(struct pt_regs *regs, int trapnr);
extern int kprobe_exceptions_notify(struct notifier_block *self,
unsigned long val, void *data);
#endif /* _ASM_KPROBES_H */
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment