aboutsummaryrefslogtreecommitdiffstats
path: root/arch/x86_64/kernel/kprobes.c
blob: 4e680f87a75f739729427a4b6027714e1c7f93f5 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
/*
 *  Kernel Probes (KProbes)
 *  arch/x86_64/kernel/kprobes.c
 *
 * 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.
 * 2004-Oct	Jim Keniston <kenistoj@us.ibm.com> and Prasanna S Panchamukhi
 *		<prasanna@in.ibm.com> adapted for x86_64
 * 2005-Mar	Roland McGrath <roland@redhat.com>
 *		Fixed to handle %rip-relative addressing mode correctly.
 * 2005-May     Rusty Lynch <rusty.lynch@intel.com>
 *              Added function return probes functionality
 */

#include <linux/config.h>
#include <linux/kprobes.h>
#include <linux/ptrace.h>
#include <linux/spinlock.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/preempt.h>
#include <linux/moduleloader.h>
#include <asm/cacheflush.h>
#include <asm/pgtable.h>
#include <asm/kdebug.h>

static DECLARE_MUTEX(kprobe_mutex);

static struct kprobe *current_kprobe;
static unsigned long kprobe_status, kprobe_old_rflags, kprobe_saved_rflags;
static struct kprobe *kprobe_prev;
static unsigned long kprobe_status_prev, kprobe_old_rflags_prev, kprobe_saved_rflags_prev;
static struct pt_regs jprobe_saved_regs;
static long *jprobe_saved_rsp;
static kprobe_opcode_t *get_insn_slot(void);
static void free_insn_slot(kprobe_opcode_t *slot);
void jprobe_return_end(void);

/* copy of the kernel stack at the probe fire time */
static kprobe_opcode_t jprobes_stack[MAX_STACK_SIZE];

/*
 * returns non-zero if opcode modifies the interrupt flag.
 */
static inline int 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;
	}

	if (*insn  >= 0x40 && *insn <= 0x4f && *++insn == 0xcf)
		return 1;
	return 0;
}

int arch_prepare_kprobe(struct kprobe *p)
{
	/* insn: must be on special executable page on x86_64. */
	up(&kprobe_mutex);
	p->ainsn.insn = get_insn_slot();
	down(&kprobe_mutex);
	if (!p->ainsn.insn) {
		return -ENOMEM;
	}
	return 0;
}

/*
 * Determine if the instruction uses the %rip-relative addressing mode.
 * If it does, return the address of the 32-bit displacement word.
 * If not, return null.
 */
static inline s32 *is_riprel(u8 *insn)
{
#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 % 64))
	static const u64 onebyte_has_modrm[256 / 64] = {
		/*      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 u64 twobyte_has_modrm[256 / 64] = {
		/*      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
	int need_modrm;

	/* Skip legacy instruction prefixes.  */
	while (1) {
		switch (*insn) {
		case 0x66:
		case 0x67:
		case 0x2e:
		case 0x3e:
		case 0x26:
		case 0x64:
		case 0x65:
		case 0x36:
		case 0xf0:
		case 0xf3:
		case 0xf2:
			++insn;
			continue;
		}
		break;
	}

	/* Skip REX instruction prefix.  */
	if ((*insn & 0xf0) == 0x40)
		++insn;

	if (*insn == 0x0f) {	/* Two-byte opcode.  */
		++insn;
		need_modrm = test_bit(*insn, twobyte_has_modrm);
	} else {		/* One-byte opcode.  */
		need_modrm = test_bit(*insn, onebyte_has_modrm);
	}

	if (need_modrm) {
		u8 modrm = *++insn;
		if ((modrm & 0xc7) == 0x05) { /* %rip+disp32 addressing mode */
			/* Displacement follows ModRM byte.  */
			return (s32 *) ++insn;
		}
	}

	/* No %rip-relative addressing mode here.  */
	return NULL;
}

void arch_copy_kprobe(struct kprobe *p)
{
	s32 *ripdisp;
	memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE);
	ripdisp = is_riprel(p->ainsn.insn);
	if (ripdisp) {
		/*
		 * The copied instruction uses the %rip-relative
		 * addressing mode.  Adjust the displacement for the
		 * difference between the original location of this
		 * instruction and the location of the copy that will
		 * actually be run.  The tricky bit here is making sure
		 * that the sign extension happens correctly in this
		 * calculation, since we need a signed 32-bit result to
		 * be sign-extended to 64 bits when it's added to the
		 * %rip value and yield the same 64-bit result that the
		 * sign-extension of the original signed 32-bit
		 * displacement would have given.
		 */
		s64 disp = (u8 *) p->addr + *ripdisp - (u8 *) p->ainsn.insn;
		BUG_ON((s64) (s32) disp != disp); /* Sanity check.  */
		*ripdisp = disp;
	}
	p->opcode = *p->addr;
}

void arch_arm_kprobe(struct kprobe *p)
{
	*p->addr = BREAKPOINT_INSTRUCTION;
	flush_icache_range((unsigned long) p->addr,
			   (unsigned long) p->addr + sizeof(kprobe_opcode_t));
}

void arch_disarm_kprobe(struct kprobe *p)
{
	*p->addr = p->opcode;
	flush_icache_range((unsigned long) p->addr,
			   (unsigned long) p->addr + sizeof(kprobe_opcode_t));
}

void arch_remove_kprobe(struct kprobe *p)
{
	up(&kprobe_mutex);
	free_insn_slot(p->ainsn.insn);
	down(&kprobe_mutex);
}

static inline void save_previous_kprobe(void)
{
	kprobe_prev = current_kprobe;
	kprobe_status_prev = kprobe_status;
	kprobe_old_rflags_prev = kprobe_old_rflags;
	kprobe_saved_rflags_prev = kprobe_saved_rflags;
}

static inline void restore_previous_kprobe(void)
{
	current_kprobe = kprobe_prev;
	kprobe_status = kprobe_status_prev;
	kprobe_old_rflags = kprobe_old_rflags_prev;
	kprobe_saved_rflags = kprobe_saved_rflags_prev;
}

static inline void set_current_kprobe(struct kprobe *p, struct pt_regs *regs)
{
	current_kprobe = p;
	kprobe_saved_rflags = kprobe_old_rflags
		= (regs->eflags & (TF_MASK | IF_MASK));
	if (is_IF_modifier(p->ainsn.insn))
		kprobe_saved_rflags &= ~IF_MASK;
}

static void prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
{
	regs->eflags |= TF_MASK;
	regs->eflags &= ~IF_MASK;
	/*single step inline if the instruction is an int3*/
	if (p->opcode == BREAKPOINT_INSTRUCTION)
		regs->rip = (unsigned long)p->addr;
	else
		regs->rip = (unsigned long)p->ainsn.insn;
}

struct task_struct  *arch_get_kprobe_task(void *ptr)
{
	return ((struct thread_info *) (((unsigned long) ptr) &
					(~(THREAD_SIZE -1))))->task;
}

void arch_prepare_kretprobe(struct kretprobe *rp, struct pt_regs *regs)
{
	unsigned long *sara = (unsigned long *)regs->rsp;
	struct kretprobe_instance *ri;
	static void *orig_ret_addr;

	/*
	 * Save the return address when the return probe hits
	 * the first time, and use it to populate the (krprobe
	 * instance)->ret_addr for subsequent return probes at
	 * the same addrress since stack address would have
	 * the kretprobe_trampoline by then.
	 */
	if (((void*) *sara) != kretprobe_trampoline)
		orig_ret_addr = (void*) *sara;

	if ((ri = get_free_rp_inst(rp)) != NULL) {
		ri->rp = rp;
		ri->stack_addr = sara;
		ri->ret_addr = orig_ret_addr;
		add_rp_inst(ri);
		/* Replace the return addr with trampoline addr */
		*sara = (unsigned long) &kretprobe_trampoline;
	} else {
		rp->nmissed++;
	}
}

void arch_kprobe_flush_task(struct task_struct *tk)
{
	struct kretprobe_instance *ri;
	while ((ri = get_rp_inst_tsk(tk)) != NULL) {
		*((unsigned long *)(ri->stack_addr)) =
					(unsigned long) ri->ret_addr;
		recycle_rp_inst(ri);
	}
}

/*
 * Interrupts are disabled on entry as trap3 is an interrupt gate and they
 * remain disabled thorough out this function.
 */
int kprobe_handler(struct pt_regs *regs)
{
	struct kprobe *p;
	int ret = 0;
	kprobe_opcode_t *addr = (kprobe_opcode_t *)(regs->rip - sizeof(kprobe_opcode_t));

	/* We're in an interrupt, but this is clear and BUG()-safe. */
	preempt_disable();

	/* Check we're not actually recursing */
	if (kprobe_running()) {
		/* We *are* holding lock here, so this is safe.
		   Disarm the probe we just hit, and ignore it. */
		p = get_kprobe(addr);
		if (p) {
			if (kprobe_status == KPROBE_HIT_SS) {
				regs->eflags &= ~TF_MASK;
				regs->eflags |= kprobe_saved_rflags;
				unlock_kprobes();
				goto no_kprobe;
			} else if (kprobe_status == KPROBE_HIT_SSDONE) {
				/* TODO: Provide re-entrancy from
				 * post_kprobes_handler() and avoid exception
				 * stack corruption while single-stepping on
				 * the instruction of the new probe.
				 */
				arch_disarm_kprobe(p);
				regs->rip = (unsigned long)p->addr;
				ret = 1;
			} else {
				/* We have reentered the kprobe_handler(), since
				 * another probe was hit while within the
				 * handler. We here save the original kprobe
				 * variables and just single step on instruction
				 * of the new probe without calling any user
				 * handlers.
				 */
				save_previous_kprobe();
				set_current_kprobe(p, regs);
				p->nmissed++;
				prepare_singlestep(p, regs);
				kprobe_status = KPROBE_REENTER;
				return 1;
			}
		} else {
			p = current_kprobe;
			if (p->break_handler && p->break_handler(p, regs)) {
				goto ss_probe;
			}
		}
		/* If it's not ours, can't be delete race, (we hold lock). */
		goto no_kprobe;
	}

	lock_kprobes();
	p = get_kprobe(addr);
	if (!p) {
		unlock_kprobes();
		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.
			 */
			ret = 1;
		}
		/* Not one of ours: let kernel handle it */
		goto no_kprobe;
	}

	kprobe_status = KPROBE_HIT_ACTIVE;
	set_current_kprobe(p, regs);

	if (p->pre_handler && p->pre_handler(p, regs))
		/* handler has already set things up, so skip ss setup */
		return 1;

ss_probe:
	prepare_singlestep(p, regs);
	kprobe_status = KPROBE_HIT_SS;
	return 1;

no_kprobe:
	preempt_enable_no_resched();
	return ret;
}

/*
 * For function-return probes, init_kprobes() establishes a probepoint
 * here. When a retprobed function returns, this probe is hit and
 * trampoline_probe_handler() runs, calling the kretprobe's handler.
 */
 void kretprobe_trampoline_holder(void)
 {
 	asm volatile (  ".global kretprobe_trampoline\n"
 			"kretprobe_trampoline: \n"
 			"nop\n");
 }

/*
 * Called when we hit the probe point at kretprobe_trampoline
 */
int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs)
{
	struct task_struct *tsk;
	struct kretprobe_instance *ri;
	struct hlist_head *head;
	struct hlist_node *node;
	unsigned long *sara = (unsigned long *)regs->rsp - 1;

	tsk = arch_get_kprobe_task(sara);
	head = kretprobe_inst_table_head(tsk);

	hlist_for_each_entry(ri, node, head, hlist) {
		if (ri->stack_addr == sara && ri->rp) {
			if (ri->rp->handler)
				ri->rp->handler(ri, regs);
		}
	}
	return 0;
}

void trampoline_post_handler(struct kprobe *p, struct pt_regs *regs,
						unsigned long flags)
{
	struct kretprobe_instance *ri;
	/* RA already popped */
	unsigned long *sara = ((unsigned long *)regs->rsp) - 1;

	while ((ri = get_rp_inst(sara))) {
		regs->rip = (unsigned long)ri->ret_addr;
		recycle_rp_inst(ri);
	}
	regs->eflags &= ~TF_MASK;
}

/*
 * 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 rip 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 eflags, 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.
 */
static void resume_execution(struct kprobe *p, struct pt_regs *regs)
{
	unsigned long *tos = (unsigned long *)regs->rsp;
	unsigned long next_rip = 0;
	unsigned long copy_rip = (unsigned long)p->ainsn.insn;
	unsigned long orig_rip = (unsigned long)p->addr;
	kprobe_opcode_t *insn = p->ainsn.insn;

	/*skip the REX prefix*/
	if (*insn >= 0x40 && *insn <= 0x4f)
		insn++;

	switch (*insn) {
	case 0x9c:		/* pushfl */
		*tos &= ~(TF_MASK | IF_MASK);
		*tos |= kprobe_old_rflags;
		break;
	case 0xc3:		/* ret/lret */
	case 0xcb:
	case 0xc2:
	case 0xca:
		regs->eflags &= ~TF_MASK;
		/* rip is already adjusted, no more changes required*/
		return;
	case 0xe8:		/* call relative - Fix return addr */
		*tos = orig_rip + (*tos - copy_rip);
		break;
	case 0xff:
		if ((*insn & 0x30) == 0x10) {
			/* call absolute, indirect */
			/* Fix return addr; rip is correct. */
			next_rip = regs->rip;
			*tos = orig_rip + (*tos - copy_rip);
		} else if (((*insn & 0x31) == 0x20) ||	/* jmp near, absolute indirect */
			   ((*insn & 0x31) == 0x21)) {	/* jmp far, absolute indirect */
			/* rip is correct. */
			next_rip = regs->rip;
		}
		break;
	case 0xea:		/* jmp absolute -- rip is correct */
		next_rip = regs->rip;
		break;
	default:
		break;
	}

	regs->eflags &= ~TF_MASK;
	if (next_rip) {
		regs->rip = next_rip;
	} else {
		regs->rip = orig_rip + (regs->rip - copy_rip);
	}
}

/*
 * Interrupts are disabled on entry as trap1 is an interrupt gate and they
 * remain disabled thoroughout this function.  And we hold kprobe lock.
 */
int post_kprobe_handler(struct pt_regs *regs)
{
	if (!kprobe_running())
		return 0;

	if ((kprobe_status != KPROBE_REENTER) && current_kprobe->post_handler) {
		kprobe_status = KPROBE_HIT_SSDONE;
		current_kprobe->post_handler(current_kprobe, regs, 0);
	}

	if (current_kprobe->post_handler != trampoline_post_handler)
		resume_execution(current_kprobe, regs);
	regs->eflags |= kprobe_saved_rflags;

	/* Restore the original saved kprobes variables and continue. */
	if (kprobe_status == KPROBE_REENTER) {
		restore_previous_kprobe();
		goto out;
	} else {
		unlock_kprobes();
	}
out:
	preempt_enable_no_resched();

	/*
	 * if somebody else is singlestepping across a probe point, eflags
	 * will have TF set, in which case, continue the remaining processing
	 * of do_debug, as if this is not a probe hit.
	 */
	if (regs->eflags & TF_MASK)
		return 0;

	return 1;
}

/* Interrupts disabled, kprobe_lock held. */
int kprobe_fault_handler(struct pt_regs *regs, int trapnr)
{
	if (current_kprobe->fault_handler
	    && current_kprobe->fault_handler(current_kprobe, regs, trapnr))
		return 1;

	if (kprobe_status & KPROBE_HIT_SS) {
		resume_execution(current_kprobe, regs);
		regs->eflags |= kprobe_old_rflags;

		unlock_kprobes();
		preempt_enable_no_resched();
	}
	return 0;
}

/*
 * Wrapper routine for handling exceptions.
 */
int kprobe_exceptions_notify(struct notifier_block *self, unsigned long val,
			     void *data)
{
	struct die_args *args = (struct die_args *)data;
	switch (val) {
	case DIE_INT3:
		if (kprobe_handler(args->regs))
			return NOTIFY_STOP;
		break;
	case DIE_DEBUG:
		if (post_kprobe_handler(args->regs))
			return NOTIFY_STOP;
		break;
	case DIE_GPF:
		if (kprobe_running() &&
		    kprobe_fault_handler(args->regs, args->trapnr))
			return NOTIFY_STOP;
		break;
	case DIE_PAGE_FAULT:
		if (kprobe_running() &&
		    kprobe_fault_handler(args->regs, args->trapnr))
			return NOTIFY_STOP;
		break;
	default:
		break;
	}
	return NOTIFY_DONE;
}

int setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
{
	struct jprobe *jp = container_of(p, struct jprobe, kp);
	unsigned long addr;

	jprobe_saved_regs = *regs;
	jprobe_saved_rsp = (long *) regs->rsp;
	addr = (unsigned long)jprobe_saved_rsp;
	/*
	 * 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(jprobes_stack, (kprobe_opcode_t *) addr, MIN_STACK_SIZE(addr));
	regs->eflags &= ~IF_MASK;
	regs->rip = (unsigned long)(jp->entry);
	return 1;
}

void jprobe_return(void)
{
	preempt_enable_no_resched();
	asm volatile ("       xchg   %%rbx,%%rsp     \n"
		      "       int3			\n"
		      "       .globl jprobe_return_end	\n"
		      "       jprobe_return_end:	\n"
		      "       nop			\n"::"b"
		      (jprobe_saved_rsp):"memory");
}

int longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
{
	u8 *addr = (u8 *) (regs->rip - 1);
	unsigned long stack_addr = (unsigned long)jprobe_saved_rsp;
	struct jprobe *jp = container_of(p, struct jprobe, kp);

	if ((addr > (u8 *) jprobe_return) && (addr < (u8 *) jprobe_return_end)) {
		if ((long *)regs->rsp != jprobe_saved_rsp) {
			struct pt_regs *saved_regs =
			    container_of(jprobe_saved_rsp, struct pt_regs, rsp);
			printk("current rsp %p does not match saved rsp %p\n",
			       (long *)regs->rsp, jprobe_saved_rsp);
			printk("Saved registers for jprobe %p\n", jp);
			show_registers(saved_regs);
			printk("Current registers\n");
			show_registers(regs);
			BUG();
		}
		*regs = jprobe_saved_regs;
		memcpy((kprobe_opcode_t *) stack_addr, jprobes_stack,
		       MIN_STACK_SIZE(stack_addr));
		return 1;
	}
	return 0;
}

/*
 * kprobe->ainsn.insn points to the copy of the instruction to be single-stepped.
 * By default on x86_64, pages we get from kmalloc or vmalloc are not
 * executable.  Single-stepping an instruction on such a page yields an
 * oops.  So instead of storing the instruction copies in their respective
 * kprobe objects, we allocate a page, map it executable, and store all the
 * instruction copies there.  (We can allocate additional pages if somebody
 * inserts a huge number of probes.)  Each page can hold up to INSNS_PER_PAGE
 * instruction slots, each of which is MAX_INSN_SIZE*sizeof(kprobe_opcode_t)
 * bytes.
 */
#define INSNS_PER_PAGE (PAGE_SIZE/(MAX_INSN_SIZE*sizeof(kprobe_opcode_t)))
struct kprobe_insn_page {
	struct hlist_node hlist;
	kprobe_opcode_t *insns;		/* page of instruction slots */
	char slot_used[INSNS_PER_PAGE];
	int nused;
};

static struct hlist_head kprobe_insn_pages;

/**
 * get_insn_slot() - Find a slot on an executable page for an instruction.
 * We allocate an executable page if there's no room on existing ones.
 */
static kprobe_opcode_t *get_insn_slot(void)
{
	struct kprobe_insn_page *kip;
	struct hlist_node *pos;

	hlist_for_each(pos, &kprobe_insn_pages) {
		kip = hlist_entry(pos, struct kprobe_insn_page, hlist);
		if (kip->nused < INSNS_PER_PAGE) {
			int i;
			for (i = 0; i < INSNS_PER_PAGE; i++) {
				if (!kip->slot_used[i]) {
					kip->slot_used[i] = 1;
					kip->nused++;
					return kip->insns + (i*MAX_INSN_SIZE);
				}
			}
			/* Surprise!  No unused slots.  Fix kip->nused. */
			kip->nused = INSNS_PER_PAGE;
		}
	}

	/* All out of space.  Need to allocate a new page. Use slot 0.*/
	kip = kmalloc(sizeof(struct kprobe_insn_page), GFP_KERNEL);
	if (!kip) {
		return NULL;
	}

	/*
	 * For the %rip-relative displacement fixups to be doable, we
	 * need our instruction copy to be within +/- 2GB of any data it
	 * might access via %rip.  That is, within 2GB of where the
	 * kernel image and loaded module images reside.  So we allocate
	 * a page in the module loading area.
	 */
	kip->insns = module_alloc(PAGE_SIZE);
	if (!kip->insns) {
		kfree(kip);
		return NULL;
	}
	INIT_HLIST_NODE(&kip->hlist);
	hlist_add_head(&kip->hlist, &kprobe_insn_pages);
	memset(kip->slot_used, 0, INSNS_PER_PAGE);
	kip->slot_used[0] = 1;
	kip->nused = 1;
	return kip->insns;
}

/**
 * free_insn_slot() - Free instruction slot obtained from get_insn_slot().
 */
static void free_insn_slot(kprobe_opcode_t *slot)
{
	struct kprobe_insn_page *kip;
	struct hlist_node *pos;

	hlist_for_each(pos, &kprobe_insn_pages) {
		kip = hlist_entry(pos, struct kprobe_insn_page, hlist);
		if (kip->insns <= slot
		    && slot < kip->insns+(INSNS_PER_PAGE*MAX_INSN_SIZE)) {
			int i = (slot - kip->insns) / MAX_INSN_SIZE;
			kip->slot_used[i] = 0;
			kip->nused--;
			if (kip->nused == 0) {
				/*
				 * Page is no longer in use.  Free it unless
				 * it's the last one.  We keep the last one
				 * so as not to have to set it up again the
				 * next time somebody inserts a probe.
				 */
				hlist_del(&kip->hlist);
				if (hlist_empty(&kprobe_insn_pages)) {
					INIT_HLIST_NODE(&kip->hlist);
					hlist_add_head(&kip->hlist,
						&kprobe_insn_pages);
				} else {
					module_free(NULL, kip->insns);
					kfree(kip);
				}
			}
			return;
		}
	}
}