1 // SPDX-License-Identifier: GPL-2.0-or-later
3 * Kernel Probes (KProbes)
5 * Copyright (C) IBM Corporation, 2002, 2004
7 * 2002-Oct Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
8 * Probes initial implementation ( includes contributions from
10 * 2004-July Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
11 * interface to access function arguments.
12 * 2004-Oct Jim Keniston <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
13 * <prasanna@in.ibm.com> adapted for x86_64 from i386.
14 * 2005-Mar Roland McGrath <roland@redhat.com>
15 * Fixed to handle %rip-relative addressing mode correctly.
16 * 2005-May Hien Nguyen <hien@us.ibm.com>, Jim Keniston
17 * <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
18 * <prasanna@in.ibm.com> added function-return probes.
19 * 2005-May Rusty Lynch <rusty.lynch@intel.com>
20 * Added function return probes functionality
21 * 2006-Feb Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp> added
22 * kprobe-booster and kretprobe-booster for i386.
23 * 2007-Dec Masami Hiramatsu <mhiramat@redhat.com> added kprobe-booster
24 * and kretprobe-booster for x86-64
25 * 2007-Dec Masami Hiramatsu <mhiramat@redhat.com>, Arjan van de Ven
26 * <arjan@infradead.org> and Jim Keniston <jkenisto@us.ibm.com>
27 * unified x86 kprobes code.
29 #include <linux/kprobes.h>
30 #include <linux/ptrace.h>
31 #include <linux/string.h>
32 #include <linux/slab.h>
33 #include <linux/hardirq.h>
34 #include <linux/preempt.h>
35 #include <linux/sched/debug.h>
36 #include <linux/perf_event.h>
37 #include <linux/extable.h>
38 #include <linux/kdebug.h>
39 #include <linux/kallsyms.h>
40 #include <linux/kgdb.h>
41 #include <linux/ftrace.h>
42 #include <linux/kasan.h>
43 #include <linux/moduleloader.h>
44 #include <linux/objtool.h>
45 #include <linux/vmalloc.h>
46 #include <linux/pgtable.h>
47 #include <linux/set_memory.h>
48 #include <linux/cfi.h>
50 #include <asm/text-patching.h>
51 #include <asm/cacheflush.h>
53 #include <linux/uaccess.h>
54 #include <asm/alternative.h>
56 #include <asm/debugreg.h>
61 DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
62 DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
64 #define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\
65 (((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) | \
66 (b4##UL << 0x4)|(b5##UL << 0x5)|(b6##UL << 0x6)|(b7##UL << 0x7) | \
67 (b8##UL << 0x8)|(b9##UL << 0x9)|(ba##UL << 0xa)|(bb##UL << 0xb) | \
68 (bc##UL << 0xc)|(bd##UL << 0xd)|(be##UL << 0xe)|(bf##UL << 0xf)) \
71 * Undefined/reserved opcodes, conditional jump, Opcode Extension
72 * Groups, and some special opcodes can not boost.
73 * This is non-const and volatile to keep gcc from statically
74 * optimizing it out, as variable_test_bit makes gcc think only
75 * *(unsigned long*) is used.
77 static volatile u32 twobyte_is_boostable[256 / 32] = {
78 /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
79 /* ---------------------------------------------- */
80 W(0x00, 0, 0, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0) | /* 00 */
81 W(0x10, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1) , /* 10 */
82 W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 20 */
83 W(0x30, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 30 */
84 W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 40 */
85 W(0x50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 50 */
86 W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1) | /* 60 */
87 W(0x70, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1) , /* 70 */
88 W(0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 80 */
89 W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */
90 W(0xa0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* a0 */
91 W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1) , /* b0 */
92 W(0xc0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* c0 */
93 W(0xd0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) , /* d0 */
94 W(0xe0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* e0 */
95 W(0xf0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 0) /* f0 */
96 /* ----------------------------------------------- */
97 /* 0 1 2 3 4 5 6 7 8 9 a b c d e f */
101 struct kretprobe_blackpoint kretprobe_blacklist[] = {
102 {"__switch_to", }, /* This function switches only current task, but
103 doesn't switch kernel stack.*/
104 {NULL, NULL} /* Terminator */
107 const int kretprobe_blacklist_size = ARRAY_SIZE(kretprobe_blacklist);
109 static nokprobe_inline void
110 __synthesize_relative_insn(void *dest, void *from, void *to, u8 op)
112 struct __arch_relative_insn {
117 insn = (struct __arch_relative_insn *)dest;
118 insn->raddr = (s32)((long)(to) - ((long)(from) + 5));
122 /* Insert a jump instruction at address 'from', which jumps to address 'to'.*/
123 void synthesize_reljump(void *dest, void *from, void *to)
125 __synthesize_relative_insn(dest, from, to, JMP32_INSN_OPCODE);
127 NOKPROBE_SYMBOL(synthesize_reljump);
129 /* Insert a call instruction at address 'from', which calls address 'to'.*/
130 void synthesize_relcall(void *dest, void *from, void *to)
132 __synthesize_relative_insn(dest, from, to, CALL_INSN_OPCODE);
134 NOKPROBE_SYMBOL(synthesize_relcall);
137 * Returns non-zero if INSN is boostable.
138 * RIP relative instructions are adjusted at copying time in 64 bits mode
140 bool can_boost(struct insn *insn, void *addr)
142 kprobe_opcode_t opcode;
146 if (search_exception_tables((unsigned long)addr))
147 return false; /* Page fault may occur on this address. */
149 /* 2nd-byte opcode */
150 if (insn->opcode.nbytes == 2)
151 return test_bit(insn->opcode.bytes[1],
152 (unsigned long *)twobyte_is_boostable);
154 if (insn->opcode.nbytes != 1)
157 for_each_insn_prefix(insn, i, prefix) {
160 attr = inat_get_opcode_attribute(prefix);
161 /* Can't boost Address-size override prefix and CS override prefix */
162 if (prefix == 0x2e || inat_is_address_size_prefix(attr))
166 opcode = insn->opcode.bytes[0];
169 case 0x62: /* bound */
170 case 0x70 ... 0x7f: /* Conditional jumps */
171 case 0x9a: /* Call far */
172 case 0xcc ... 0xce: /* software exceptions */
173 case 0xd6: /* (UD) */
174 case 0xd8 ... 0xdf: /* ESC */
175 case 0xe0 ... 0xe3: /* LOOP*, JCXZ */
176 case 0xe8 ... 0xe9: /* near Call, JMP */
177 case 0xeb: /* Short JMP */
178 case 0xf0 ... 0xf4: /* LOCK/REP, HLT */
179 /* ... are not boostable */
181 case 0xc0 ... 0xc1: /* Grp2 */
182 case 0xd0 ... 0xd3: /* Grp2 */
184 * AMD uses nnn == 110 as SHL/SAL, but Intel makes it reserved.
186 return X86_MODRM_REG(insn->modrm.bytes[0]) != 0b110;
187 case 0xf6 ... 0xf7: /* Grp3 */
188 /* AMD uses nnn == 001 as TEST, but Intel makes it reserved. */
189 return X86_MODRM_REG(insn->modrm.bytes[0]) != 0b001;
190 case 0xfe: /* Grp4 */
191 /* Only INC and DEC are boostable */
192 return X86_MODRM_REG(insn->modrm.bytes[0]) == 0b000 ||
193 X86_MODRM_REG(insn->modrm.bytes[0]) == 0b001;
194 case 0xff: /* Grp5 */
195 /* Only INC, DEC, and indirect JMP are boostable */
196 return X86_MODRM_REG(insn->modrm.bytes[0]) == 0b000 ||
197 X86_MODRM_REG(insn->modrm.bytes[0]) == 0b001 ||
198 X86_MODRM_REG(insn->modrm.bytes[0]) == 0b100;
205 __recover_probed_insn(kprobe_opcode_t *buf, unsigned long addr)
210 kp = get_kprobe((void *)addr);
211 faddr = ftrace_location(addr) == addr;
213 * Use the current code if it is not modified by Kprobe
214 * and it cannot be modified by ftrace.
220 * Basically, kp->ainsn.insn has an original instruction.
221 * However, RIP-relative instruction can not do single-stepping
222 * at different place, __copy_instruction() tweaks the displacement of
223 * that instruction. In that case, we can't recover the instruction
224 * from the kp->ainsn.insn.
226 * On the other hand, in case on normal Kprobe, kp->opcode has a copy
227 * of the first byte of the probed instruction, which is overwritten
228 * by int3. And the instruction at kp->addr is not modified by kprobes
229 * except for the first byte, we can recover the original instruction
230 * from it and kp->opcode.
232 * In case of Kprobes using ftrace, we do not have a copy of
233 * the original instruction. In fact, the ftrace location might
234 * be modified at anytime and even could be in an inconsistent state.
235 * Fortunately, we know that the original code is the ideal 5-byte
238 if (copy_from_kernel_nofault(buf, (void *)addr,
239 MAX_INSN_SIZE * sizeof(kprobe_opcode_t)))
243 memcpy(buf, x86_nops[5], 5);
246 return (unsigned long)buf;
250 * Recover the probed instruction at addr for further analysis.
251 * Caller must lock kprobes by kprobe_mutex, or disable preemption
252 * for preventing to release referencing kprobes.
253 * Returns zero if the instruction can not get recovered (or access failed).
255 unsigned long recover_probed_instruction(kprobe_opcode_t *buf, unsigned long addr)
257 unsigned long __addr;
259 __addr = __recover_optprobed_insn(buf, addr);
263 return __recover_probed_insn(buf, addr);
266 /* Check if insn is INT or UD */
267 static inline bool is_exception_insn(struct insn *insn)
269 /* UD uses 0f escape */
270 if (insn->opcode.bytes[0] == 0x0f) {
271 /* UD0 / UD1 / UD2 */
272 return insn->opcode.bytes[1] == 0xff ||
273 insn->opcode.bytes[1] == 0xb9 ||
274 insn->opcode.bytes[1] == 0x0b;
277 /* INT3 / INT n / INTO / INT1 */
278 return insn->opcode.bytes[0] == 0xcc ||
279 insn->opcode.bytes[0] == 0xcd ||
280 insn->opcode.bytes[0] == 0xce ||
281 insn->opcode.bytes[0] == 0xf1;
285 * Check if paddr is at an instruction boundary and that instruction can
288 static bool can_probe(unsigned long paddr)
290 unsigned long addr, __addr, offset = 0;
292 kprobe_opcode_t buf[MAX_INSN_SIZE];
294 if (!kallsyms_lookup_size_offset(paddr, NULL, &offset))
297 /* Decode instructions */
298 addr = paddr - offset;
299 while (addr < paddr) {
301 * Check if the instruction has been modified by another
302 * kprobe, in which case we replace the breakpoint by the
303 * original instruction in our buffer.
304 * Also, jump optimization will change the breakpoint to
305 * relative-jump. Since the relative-jump itself is
306 * normally used, we just go through if there is no kprobe.
308 __addr = recover_probed_instruction(buf, addr);
312 if (insn_decode_kernel(&insn, (void *)__addr) < 0)
317 * If there is a dynamically installed kgdb sw breakpoint,
318 * this function should not be probed.
320 if (insn.opcode.bytes[0] == INT3_INSN_OPCODE &&
321 kgdb_has_hit_break(addr))
327 /* Check if paddr is at an instruction boundary */
331 __addr = recover_probed_instruction(buf, addr);
335 if (insn_decode_kernel(&insn, (void *)__addr) < 0)
338 /* INT and UD are special and should not be kprobed */
339 if (is_exception_insn(&insn))
342 if (IS_ENABLED(CONFIG_CFI_CLANG)) {
344 * The compiler generates the following instruction sequence
345 * for indirect call checks and cfi.c decodes this;
347 *Â movl -<id>, %r10d ; 6 bytes
348 * addl -4(%reg), %r10d ; 4 bytes
349 * je .Ltmp1 ; 2 bytes
353 * Also, these movl and addl are used for showing expected
354 * type. So those must not be touched.
356 if (insn.opcode.value == 0xBA)
358 else if (insn.opcode.value == 0x3)
363 /* This movl/addl is used for decoding CFI. */
364 if (is_cfi_trap(addr + offset))
372 /* If x86 supports IBT (ENDBR) it must be skipped. */
373 kprobe_opcode_t *arch_adjust_kprobe_addr(unsigned long addr, unsigned long offset,
376 if (is_endbr(*(u32 *)addr)) {
377 *on_func_entry = !offset || offset == 4;
382 *on_func_entry = !offset;
385 return (kprobe_opcode_t *)(addr + offset);
389 * Copy an instruction with recovering modified instruction by kprobes
390 * and adjust the displacement if the instruction uses the %rip-relative
391 * addressing mode. Note that since @real will be the final place of copied
392 * instruction, displacement must be adjust by @real, not @dest.
393 * This returns the length of copied instruction, or 0 if it has an error.
395 int __copy_instruction(u8 *dest, u8 *src, u8 *real, struct insn *insn)
397 kprobe_opcode_t buf[MAX_INSN_SIZE];
398 unsigned long recovered_insn = recover_probed_instruction(buf, (unsigned long)src);
401 if (!recovered_insn || !insn)
404 /* This can access kernel text if given address is not recovered */
405 if (copy_from_kernel_nofault(dest, (void *)recovered_insn,
409 ret = insn_decode_kernel(insn, dest);
413 /* We can not probe force emulate prefixed instruction */
414 if (insn_has_emulate_prefix(insn))
417 /* Another subsystem puts a breakpoint, failed to recover */
418 if (insn->opcode.bytes[0] == INT3_INSN_OPCODE)
421 /* We should not singlestep on the exception masking instructions */
422 if (insn_masking_exception(insn))
426 /* Only x86_64 has RIP relative instructions */
427 if (insn_rip_relative(insn)) {
431 * The copied instruction uses the %rip-relative addressing
432 * mode. Adjust the displacement for the difference between
433 * the original location of this instruction and the location
434 * of the copy that will actually be run. The tricky bit here
435 * is making sure that the sign extension happens correctly in
436 * this calculation, since we need a signed 32-bit result to
437 * be sign-extended to 64 bits when it's added to the %rip
438 * value and yield the same 64-bit result that the sign-
439 * extension of the original signed 32-bit displacement would
442 newdisp = (u8 *) src + (s64) insn->displacement.value
444 if ((s64) (s32) newdisp != newdisp) {
445 pr_err("Kprobes error: new displacement does not fit into s32 (%llx)\n", newdisp);
448 disp = (u8 *) dest + insn_offset_displacement(insn);
449 *(s32 *) disp = (s32) newdisp;
455 /* Prepare reljump or int3 right after instruction */
456 static int prepare_singlestep(kprobe_opcode_t *buf, struct kprobe *p,
459 int len = insn->length;
461 if (!IS_ENABLED(CONFIG_PREEMPTION) &&
462 !p->post_handler && can_boost(insn, p->addr) &&
463 MAX_INSN_SIZE - len >= JMP32_INSN_SIZE) {
465 * These instructions can be executed directly if it
466 * jumps back to correct address.
468 synthesize_reljump(buf + len, p->ainsn.insn + len,
469 p->addr + insn->length);
470 len += JMP32_INSN_SIZE;
471 p->ainsn.boostable = 1;
473 /* Otherwise, put an int3 for trapping singlestep */
474 if (MAX_INSN_SIZE - len < INT3_INSN_SIZE)
477 buf[len] = INT3_INSN_OPCODE;
478 len += INT3_INSN_SIZE;
484 /* Make page to RO mode when allocate it */
485 void *alloc_insn_page(void)
489 page = module_alloc(PAGE_SIZE);
494 * TODO: Once additional kernel code protection mechanisms are set, ensure
495 * that the page was not maliciously altered and it is still zeroed.
497 set_memory_rox((unsigned long)page, 1);
502 /* Kprobe x86 instruction emulation - only regs->ip or IF flag modifiers */
504 static void kprobe_emulate_ifmodifiers(struct kprobe *p, struct pt_regs *regs)
506 switch (p->ainsn.opcode) {
508 regs->flags &= ~(X86_EFLAGS_IF);
511 regs->flags |= X86_EFLAGS_IF;
513 case 0x9c: /* pushf */
514 int3_emulate_push(regs, regs->flags);
516 case 0x9d: /* popf */
517 regs->flags = int3_emulate_pop(regs);
520 regs->ip = regs->ip - INT3_INSN_SIZE + p->ainsn.size;
522 NOKPROBE_SYMBOL(kprobe_emulate_ifmodifiers);
524 static void kprobe_emulate_ret(struct kprobe *p, struct pt_regs *regs)
526 int3_emulate_ret(regs);
528 NOKPROBE_SYMBOL(kprobe_emulate_ret);
530 static void kprobe_emulate_call(struct kprobe *p, struct pt_regs *regs)
532 unsigned long func = regs->ip - INT3_INSN_SIZE + p->ainsn.size;
534 func += p->ainsn.rel32;
535 int3_emulate_call(regs, func);
537 NOKPROBE_SYMBOL(kprobe_emulate_call);
539 static void kprobe_emulate_jmp(struct kprobe *p, struct pt_regs *regs)
541 unsigned long ip = regs->ip - INT3_INSN_SIZE + p->ainsn.size;
543 ip += p->ainsn.rel32;
544 int3_emulate_jmp(regs, ip);
546 NOKPROBE_SYMBOL(kprobe_emulate_jmp);
548 static void kprobe_emulate_jcc(struct kprobe *p, struct pt_regs *regs)
550 unsigned long ip = regs->ip - INT3_INSN_SIZE + p->ainsn.size;
552 int3_emulate_jcc(regs, p->ainsn.jcc.type, ip, p->ainsn.rel32);
554 NOKPROBE_SYMBOL(kprobe_emulate_jcc);
556 static void kprobe_emulate_loop(struct kprobe *p, struct pt_regs *regs)
558 unsigned long ip = regs->ip - INT3_INSN_SIZE + p->ainsn.size;
561 if (p->ainsn.loop.type != 3) { /* LOOP* */
562 if (p->ainsn.loop.asize == 32)
563 match = ((*(u32 *)®s->cx)--) != 0;
565 else if (p->ainsn.loop.asize == 64)
566 match = ((*(u64 *)®s->cx)--) != 0;
569 match = ((*(u16 *)®s->cx)--) != 0;
571 if (p->ainsn.loop.asize == 32)
572 match = *(u32 *)(®s->cx) == 0;
574 else if (p->ainsn.loop.asize == 64)
575 match = *(u64 *)(®s->cx) == 0;
578 match = *(u16 *)(®s->cx) == 0;
581 if (p->ainsn.loop.type == 0) /* LOOPNE */
582 match = match && !(regs->flags & X86_EFLAGS_ZF);
583 else if (p->ainsn.loop.type == 1) /* LOOPE */
584 match = match && (regs->flags & X86_EFLAGS_ZF);
587 ip += p->ainsn.rel32;
588 int3_emulate_jmp(regs, ip);
590 NOKPROBE_SYMBOL(kprobe_emulate_loop);
592 static const int addrmode_regoffs[] = {
593 offsetof(struct pt_regs, ax),
594 offsetof(struct pt_regs, cx),
595 offsetof(struct pt_regs, dx),
596 offsetof(struct pt_regs, bx),
597 offsetof(struct pt_regs, sp),
598 offsetof(struct pt_regs, bp),
599 offsetof(struct pt_regs, si),
600 offsetof(struct pt_regs, di),
602 offsetof(struct pt_regs, r8),
603 offsetof(struct pt_regs, r9),
604 offsetof(struct pt_regs, r10),
605 offsetof(struct pt_regs, r11),
606 offsetof(struct pt_regs, r12),
607 offsetof(struct pt_regs, r13),
608 offsetof(struct pt_regs, r14),
609 offsetof(struct pt_regs, r15),
613 static void kprobe_emulate_call_indirect(struct kprobe *p, struct pt_regs *regs)
615 unsigned long offs = addrmode_regoffs[p->ainsn.indirect.reg];
617 int3_emulate_push(regs, regs->ip - INT3_INSN_SIZE + p->ainsn.size);
618 int3_emulate_jmp(regs, regs_get_register(regs, offs));
620 NOKPROBE_SYMBOL(kprobe_emulate_call_indirect);
622 static void kprobe_emulate_jmp_indirect(struct kprobe *p, struct pt_regs *regs)
624 unsigned long offs = addrmode_regoffs[p->ainsn.indirect.reg];
626 int3_emulate_jmp(regs, regs_get_register(regs, offs));
628 NOKPROBE_SYMBOL(kprobe_emulate_jmp_indirect);
630 static int prepare_emulation(struct kprobe *p, struct insn *insn)
632 insn_byte_t opcode = insn->opcode.bytes[0];
637 case 0x9c: /* pushfl */
638 case 0x9d: /* popf/popfd */
640 * IF modifiers must be emulated since it will enable interrupt while
641 * int3 single stepping.
643 p->ainsn.emulate_op = kprobe_emulate_ifmodifiers;
644 p->ainsn.opcode = opcode;
646 case 0xc2: /* ret/lret */
650 p->ainsn.emulate_op = kprobe_emulate_ret;
652 case 0x9a: /* far call absolute -- segment is not supported */
653 case 0xea: /* far jmp absolute -- segment is not supported */
654 case 0xcc: /* int3 */
655 case 0xcf: /* iret -- in-kernel IRET is not supported */
658 case 0xe8: /* near call relative */
659 p->ainsn.emulate_op = kprobe_emulate_call;
660 if (insn->immediate.nbytes == 2)
661 p->ainsn.rel32 = *(s16 *)&insn->immediate.value;
663 p->ainsn.rel32 = *(s32 *)&insn->immediate.value;
665 case 0xeb: /* short jump relative */
666 case 0xe9: /* near jump relative */
667 p->ainsn.emulate_op = kprobe_emulate_jmp;
668 if (insn->immediate.nbytes == 1)
669 p->ainsn.rel32 = *(s8 *)&insn->immediate.value;
670 else if (insn->immediate.nbytes == 2)
671 p->ainsn.rel32 = *(s16 *)&insn->immediate.value;
673 p->ainsn.rel32 = *(s32 *)&insn->immediate.value;
676 /* 1 byte conditional jump */
677 p->ainsn.emulate_op = kprobe_emulate_jcc;
678 p->ainsn.jcc.type = opcode & 0xf;
679 p->ainsn.rel32 = insn->immediate.value;
682 opcode = insn->opcode.bytes[1];
683 if ((opcode & 0xf0) == 0x80) {
684 /* 2 bytes Conditional Jump */
685 p->ainsn.emulate_op = kprobe_emulate_jcc;
686 p->ainsn.jcc.type = opcode & 0xf;
687 if (insn->immediate.nbytes == 2)
688 p->ainsn.rel32 = *(s16 *)&insn->immediate.value;
690 p->ainsn.rel32 = *(s32 *)&insn->immediate.value;
691 } else if (opcode == 0x01 &&
692 X86_MODRM_REG(insn->modrm.bytes[0]) == 0 &&
693 X86_MODRM_MOD(insn->modrm.bytes[0]) == 3) {
694 /* VM extensions - not supported */
698 case 0xe0: /* Loop NZ */
699 case 0xe1: /* Loop */
700 case 0xe2: /* Loop */
701 case 0xe3: /* J*CXZ */
702 p->ainsn.emulate_op = kprobe_emulate_loop;
703 p->ainsn.loop.type = opcode & 0x3;
704 p->ainsn.loop.asize = insn->addr_bytes * 8;
705 p->ainsn.rel32 = *(s8 *)&insn->immediate.value;
709 * Since the 0xff is an extended group opcode, the instruction
710 * is determined by the MOD/RM byte.
712 opcode = insn->modrm.bytes[0];
713 switch (X86_MODRM_REG(opcode)) {
714 case 0b010: /* FF /2, call near, absolute indirect */
715 p->ainsn.emulate_op = kprobe_emulate_call_indirect;
717 case 0b100: /* FF /4, jmp near, absolute indirect */
718 p->ainsn.emulate_op = kprobe_emulate_jmp_indirect;
720 case 0b011: /* FF /3, call far, absolute indirect */
721 case 0b101: /* FF /5, jmp far, absolute indirect */
725 if (!p->ainsn.emulate_op)
728 if (insn->addr_bytes != sizeof(unsigned long))
729 return -EOPNOTSUPP; /* Don't support different size */
730 if (X86_MODRM_MOD(opcode) != 3)
731 return -EOPNOTSUPP; /* TODO: support memory addressing */
733 p->ainsn.indirect.reg = X86_MODRM_RM(opcode);
735 if (X86_REX_B(insn->rex_prefix.value))
736 p->ainsn.indirect.reg += 8;
742 p->ainsn.size = insn->length;
747 static int arch_copy_kprobe(struct kprobe *p)
750 kprobe_opcode_t buf[MAX_INSN_SIZE];
753 /* Copy an instruction with recovering if other optprobe modifies it.*/
754 len = __copy_instruction(buf, p->addr, p->ainsn.insn, &insn);
758 /* Analyze the opcode and setup emulate functions */
759 ret = prepare_emulation(p, &insn);
763 /* Add int3 for single-step or booster jmp */
764 len = prepare_singlestep(buf, p, &insn);
768 /* Also, displacement change doesn't affect the first byte */
771 p->ainsn.tp_len = len;
772 perf_event_text_poke(p->ainsn.insn, NULL, 0, buf, len);
774 /* OK, write back the instruction(s) into ROX insn buffer */
775 text_poke(p->ainsn.insn, buf, len);
780 int arch_prepare_kprobe(struct kprobe *p)
784 if (alternatives_text_reserved(p->addr, p->addr))
787 if (!can_probe((unsigned long)p->addr))
790 memset(&p->ainsn, 0, sizeof(p->ainsn));
792 /* insn: must be on special executable page on x86. */
793 p->ainsn.insn = get_insn_slot();
797 ret = arch_copy_kprobe(p);
799 free_insn_slot(p->ainsn.insn, 0);
800 p->ainsn.insn = NULL;
806 void arch_arm_kprobe(struct kprobe *p)
808 u8 int3 = INT3_INSN_OPCODE;
810 text_poke(p->addr, &int3, 1);
812 perf_event_text_poke(p->addr, &p->opcode, 1, &int3, 1);
815 void arch_disarm_kprobe(struct kprobe *p)
817 u8 int3 = INT3_INSN_OPCODE;
819 perf_event_text_poke(p->addr, &int3, 1, &p->opcode, 1);
820 text_poke(p->addr, &p->opcode, 1);
824 void arch_remove_kprobe(struct kprobe *p)
827 /* Record the perf event before freeing the slot */
828 perf_event_text_poke(p->ainsn.insn, p->ainsn.insn,
829 p->ainsn.tp_len, NULL, 0);
830 free_insn_slot(p->ainsn.insn, p->ainsn.boostable);
831 p->ainsn.insn = NULL;
835 static nokprobe_inline void
836 save_previous_kprobe(struct kprobe_ctlblk *kcb)
838 kcb->prev_kprobe.kp = kprobe_running();
839 kcb->prev_kprobe.status = kcb->kprobe_status;
840 kcb->prev_kprobe.old_flags = kcb->kprobe_old_flags;
841 kcb->prev_kprobe.saved_flags = kcb->kprobe_saved_flags;
844 static nokprobe_inline void
845 restore_previous_kprobe(struct kprobe_ctlblk *kcb)
847 __this_cpu_write(current_kprobe, kcb->prev_kprobe.kp);
848 kcb->kprobe_status = kcb->prev_kprobe.status;
849 kcb->kprobe_old_flags = kcb->prev_kprobe.old_flags;
850 kcb->kprobe_saved_flags = kcb->prev_kprobe.saved_flags;
853 static nokprobe_inline void
854 set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
855 struct kprobe_ctlblk *kcb)
857 __this_cpu_write(current_kprobe, p);
858 kcb->kprobe_saved_flags = kcb->kprobe_old_flags
859 = (regs->flags & X86_EFLAGS_IF);
862 static void kprobe_post_process(struct kprobe *cur, struct pt_regs *regs,
863 struct kprobe_ctlblk *kcb)
865 /* Restore back the original saved kprobes variables and continue. */
866 if (kcb->kprobe_status == KPROBE_REENTER) {
867 /* This will restore both kcb and current_kprobe */
868 restore_previous_kprobe(kcb);
871 * Always update the kcb status because
872 * reset_curent_kprobe() doesn't update kcb.
874 kcb->kprobe_status = KPROBE_HIT_SSDONE;
875 if (cur->post_handler)
876 cur->post_handler(cur, regs, 0);
877 reset_current_kprobe();
880 NOKPROBE_SYMBOL(kprobe_post_process);
882 static void setup_singlestep(struct kprobe *p, struct pt_regs *regs,
883 struct kprobe_ctlblk *kcb, int reenter)
885 if (setup_detour_execution(p, regs, reenter))
888 #if !defined(CONFIG_PREEMPTION)
889 if (p->ainsn.boostable) {
890 /* Boost up -- we can execute copied instructions directly */
892 reset_current_kprobe();
894 * Reentering boosted probe doesn't reset current_kprobe,
895 * nor set current_kprobe, because it doesn't use single
898 regs->ip = (unsigned long)p->ainsn.insn;
903 save_previous_kprobe(kcb);
904 set_current_kprobe(p, regs, kcb);
905 kcb->kprobe_status = KPROBE_REENTER;
907 kcb->kprobe_status = KPROBE_HIT_SS;
909 if (p->ainsn.emulate_op) {
910 p->ainsn.emulate_op(p, regs);
911 kprobe_post_process(p, regs, kcb);
915 /* Disable interrupt, and set ip register on trampoline */
916 regs->flags &= ~X86_EFLAGS_IF;
917 regs->ip = (unsigned long)p->ainsn.insn;
919 NOKPROBE_SYMBOL(setup_singlestep);
922 * Called after single-stepping. p->addr is the address of the
923 * instruction whose first byte has been replaced by the "int3"
924 * instruction. To avoid the SMP problems that can occur when we
925 * temporarily put back the original opcode to single-step, we
926 * single-stepped a copy of the instruction. The address of this
927 * copy is p->ainsn.insn. We also doesn't use trap, but "int3" again
928 * right after the copied instruction.
929 * Different from the trap single-step, "int3" single-step can not
930 * handle the instruction which changes the ip register, e.g. jmp,
931 * call, conditional jmp, and the instructions which changes the IF
932 * flags because interrupt must be disabled around the single-stepping.
933 * Such instructions are software emulated, but others are single-stepped
936 * When the 2nd "int3" handled, the regs->ip and regs->flags needs to
937 * be adjusted, so that we can resume execution on correct code.
939 static void resume_singlestep(struct kprobe *p, struct pt_regs *regs,
940 struct kprobe_ctlblk *kcb)
942 unsigned long copy_ip = (unsigned long)p->ainsn.insn;
943 unsigned long orig_ip = (unsigned long)p->addr;
945 /* Restore saved interrupt flag and ip register */
946 regs->flags |= kcb->kprobe_saved_flags;
947 /* Note that regs->ip is executed int3 so must be a step back */
948 regs->ip += (orig_ip - copy_ip) - INT3_INSN_SIZE;
950 NOKPROBE_SYMBOL(resume_singlestep);
953 * We have reentered the kprobe_handler(), since another probe was hit while
954 * within the handler. We save the original kprobes variables and just single
955 * step on the instruction of the new probe without calling any user handlers.
957 static int reenter_kprobe(struct kprobe *p, struct pt_regs *regs,
958 struct kprobe_ctlblk *kcb)
960 switch (kcb->kprobe_status) {
961 case KPROBE_HIT_SSDONE:
962 case KPROBE_HIT_ACTIVE:
964 kprobes_inc_nmissed_count(p);
965 setup_singlestep(p, regs, kcb, 1);
968 /* A probe has been hit in the codepath leading up to, or just
969 * after, single-stepping of a probed instruction. This entire
970 * codepath should strictly reside in .kprobes.text section.
971 * Raise a BUG or we'll continue in an endless reentering loop
972 * and eventually a stack overflow.
974 pr_err("Unrecoverable kprobe detected.\n");
978 /* impossible cases */
985 NOKPROBE_SYMBOL(reenter_kprobe);
987 static nokprobe_inline int kprobe_is_ss(struct kprobe_ctlblk *kcb)
989 return (kcb->kprobe_status == KPROBE_HIT_SS ||
990 kcb->kprobe_status == KPROBE_REENTER);
994 * Interrupts are disabled on entry as trap3 is an interrupt gate and they
995 * remain disabled throughout this function.
997 int kprobe_int3_handler(struct pt_regs *regs)
999 kprobe_opcode_t *addr;
1001 struct kprobe_ctlblk *kcb;
1003 if (user_mode(regs))
1006 addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t));
1008 * We don't want to be preempted for the entire duration of kprobe
1009 * processing. Since int3 and debug trap disables irqs and we clear
1010 * IF while singlestepping, it must be no preemptible.
1013 kcb = get_kprobe_ctlblk();
1014 p = get_kprobe(addr);
1017 if (kprobe_running()) {
1018 if (reenter_kprobe(p, regs, kcb))
1021 set_current_kprobe(p, regs, kcb);
1022 kcb->kprobe_status = KPROBE_HIT_ACTIVE;
1025 * If we have no pre-handler or it returned 0, we
1026 * continue with normal processing. If we have a
1027 * pre-handler and it returned non-zero, that means
1028 * user handler setup registers to exit to another
1029 * instruction, we must skip the single stepping.
1031 if (!p->pre_handler || !p->pre_handler(p, regs))
1032 setup_singlestep(p, regs, kcb, 0);
1034 reset_current_kprobe();
1037 } else if (kprobe_is_ss(kcb)) {
1038 p = kprobe_running();
1039 if ((unsigned long)p->ainsn.insn < regs->ip &&
1040 (unsigned long)p->ainsn.insn + MAX_INSN_SIZE > regs->ip) {
1041 /* Most provably this is the second int3 for singlestep */
1042 resume_singlestep(p, regs, kcb);
1043 kprobe_post_process(p, regs, kcb);
1046 } /* else: not a kprobe fault; let the kernel handle it */
1050 NOKPROBE_SYMBOL(kprobe_int3_handler);
1052 int kprobe_fault_handler(struct pt_regs *regs, int trapnr)
1054 struct kprobe *cur = kprobe_running();
1055 struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
1057 if (unlikely(regs->ip == (unsigned long)cur->ainsn.insn)) {
1058 /* This must happen on single-stepping */
1059 WARN_ON(kcb->kprobe_status != KPROBE_HIT_SS &&
1060 kcb->kprobe_status != KPROBE_REENTER);
1062 * We are here because the instruction being single
1063 * stepped caused a page fault. We reset the current
1064 * kprobe and the ip points back to the probe address
1065 * and allow the page fault handler to continue as a
1066 * normal page fault.
1068 regs->ip = (unsigned long)cur->addr;
1071 * If the IF flag was set before the kprobe hit,
1074 regs->flags |= kcb->kprobe_old_flags;
1076 if (kcb->kprobe_status == KPROBE_REENTER)
1077 restore_previous_kprobe(kcb);
1079 reset_current_kprobe();
1084 NOKPROBE_SYMBOL(kprobe_fault_handler);
1086 int __init arch_populate_kprobe_blacklist(void)
1088 return kprobe_add_area_blacklist((unsigned long)__entry_text_start,
1089 (unsigned long)__entry_text_end);
1092 int __init arch_init_kprobes(void)
1097 int arch_trampoline_kprobe(struct kprobe *p)