arch/x86/kernel/kprobes.c

   1 /*
   2  *  Kernel Probes (KProbes)
   3  *
   4  * This program is free software; you can redistribute it and/or modify
   5  * it under the terms of the GNU General Public License as published by
   6  * the Free Software Foundation; either version 2 of the License, or
   7  * (at your option) any later version.
   8  *
   9  * This program is distributed in the hope that it will be useful,
  10  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  11  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  12  * GNU General Public License for more details.
  13  *
  14  * You should have received a copy of the GNU General Public License
  15  * along with this program; if not, write to the Free Software
  16  * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
  17  *
  18  * Copyright (C) IBM Corporation, 2002, 2004
  19  *
  20  * 2002-Oct     Created by Vamsi Krishna S <vamsi_krishna@in.ibm.com> Kernel
  21  *              Probes initial implementation ( includes contributions from
  22  *              Rusty Russell).
  23  * 2004-July    Suparna Bhattacharya <suparna@in.ibm.com> added jumper probes
  24  *              interface to access function arguments.
  25  * 2004-Oct     Jim Keniston <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
  26  *              <prasanna@in.ibm.com> adapted for x86_64 from i386.
  27  * 2005-Mar     Roland McGrath <roland@redhat.com>
  28  *              Fixed to handle %rip-relative addressing mode correctly.
  29  * 2005-May     Hien Nguyen <hien@us.ibm.com>, Jim Keniston
  30  *              <jkenisto@us.ibm.com> and Prasanna S Panchamukhi
  31  *              <prasanna@in.ibm.com> added function-return probes.
  32  * 2005-May     Rusty Lynch <rusty.lynch@intel.com>
  33  *              Added function return probes functionality
  34  * 2006-Feb     Masami Hiramatsu <hiramatu@sdl.hitachi.co.jp> added
  35  *              kprobe-booster and kretprobe-booster for i386.
  36  * 2007-Dec     Masami Hiramatsu <mhiramat@redhat.com> added kprobe-booster
  37  *              and kretprobe-booster for x86-64
  38  * 2007-Dec     Masami Hiramatsu <mhiramat@redhat.com>, Arjan van de Ven
  39  *              <arjan@infradead.org> and Jim Keniston <jkenisto@us.ibm.com>
  40  *              unified x86 kprobes code.
  41  */
  42
  43 #include <linux/kprobes.h>
  44 #include <linux/ptrace.h>
  45 #include <linux/string.h>
  46 #include <linux/slab.h>
  47 #include <linux/hardirq.h>
  48 #include <linux/preempt.h>
  49 #include <linux/module.h>
  50 #include <linux/kdebug.h>
  51
  52 #include <asm/cacheflush.h>
  53 #include <asm/desc.h>
  54 #include <asm/pgtable.h>
  55 #include <asm/uaccess.h>
  56 #include <asm/alternative.h>
  57
  58 void jprobe_return_end(void);
  59
  60 DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
  61 DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
  62
  63 #ifdef CONFIG_X86_64
  64 #define stack_addr(regs) ((unsigned long *)regs->sp)
  65 #else
  66 /*
  67  * "&regs->sp" looks wrong, but it's correct for x86_32.  x86_32 CPUs
  68  * don't save the ss and esp registers if the CPU is already in kernel
  69  * mode when it traps.  So for kprobes, regs->sp and regs->ss are not
  70  * the [nonexistent] saved stack pointer and ss register, but rather
  71  * the top 8 bytes of the pre-int3 stack.  So &regs->sp happens to
  72  * point to the top of the pre-int3 stack.
  73  */
  74 #define stack_addr(regs) ((unsigned long *)&regs->sp)
  75 #endif
  76
  77 #define W(row, b0, b1, b2, b3, b4, b5, b6, b7, b8, b9, ba, bb, bc, bd, be, bf)\
  78         (((b0##UL << 0x0)|(b1##UL << 0x1)|(b2##UL << 0x2)|(b3##UL << 0x3) |   \
  79           (b4##UL << 0x4)|(b5##UL << 0x5)|(b6##UL << 0x6)|(b7##UL << 0x7) |   \
  80           (b8##UL << 0x8)|(b9##UL << 0x9)|(ba##UL << 0xa)|(bb##UL << 0xb) |   \
  81           (bc##UL << 0xc)|(bd##UL << 0xd)|(be##UL << 0xe)|(bf##UL << 0xf))    \
  82          << (row % 32))
  83         /*
  84          * Undefined/reserved opcodes, conditional jump, Opcode Extension
  85          * Groups, and some special opcodes can not boost.
  86          */
  87 static const u32 twobyte_is_boostable[256 / 32] = {
  88         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
  89         /*      ----------------------------------------------          */
  90         W(0x00, 0, 0, 1, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0) | /* 00 */
  91         W(0x10, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 10 */
  92         W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 20 */
  93         W(0x30, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 30 */
  94         W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 40 */
  95         W(0x50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 50 */
  96         W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1) | /* 60 */
  97         W(0x70, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1) , /* 70 */
  98         W(0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 80 */
  99         W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 90 */
 100         W(0xa0, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* a0 */
 101         W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1) , /* b0 */
 102         W(0xc0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* c0 */
 103         W(0xd0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) , /* d0 */
 104         W(0xe0, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 1) | /* e0 */
 105         W(0xf0, 0, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 0)   /* f0 */
 106         /*      -----------------------------------------------         */
 107         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
 108 };
 109 static const u32 onebyte_has_modrm[256 / 32] = {
 110         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
 111         /*      -----------------------------------------------         */
 112         W(0x00, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 00 */
 113         W(0x10, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) , /* 10 */
 114         W(0x20, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) | /* 20 */
 115         W(0x30, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0) , /* 30 */
 116         W(0x40, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 40 */
 117         W(0x50, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 50 */
 118         W(0x60, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0) | /* 60 */
 119         W(0x70, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 70 */
 120         W(0x80, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 80 */
 121         W(0x90, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 90 */
 122         W(0xa0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* a0 */
 123         W(0xb0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* b0 */
 124         W(0xc0, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0) | /* c0 */
 125         W(0xd0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1) , /* d0 */
 126         W(0xe0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* e0 */
 127         W(0xf0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1)   /* f0 */
 128         /*      -----------------------------------------------         */
 129         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
 130 };
 131 static const u32 twobyte_has_modrm[256 / 32] = {
 132         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
 133         /*      -----------------------------------------------         */
 134         W(0x00, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1) | /* 0f */
 135         W(0x10, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0) , /* 1f */
 136         W(0x20, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1) | /* 2f */
 137         W(0x30, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) , /* 3f */
 138         W(0x40, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 4f */
 139         W(0x50, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 5f */
 140         W(0x60, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* 6f */
 141         W(0x70, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1) , /* 7f */
 142         W(0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0) | /* 8f */
 143         W(0x90, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* 9f */
 144         W(0xa0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1) | /* af */
 145         W(0xb0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1) , /* bf */
 146         W(0xc0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0) | /* cf */
 147         W(0xd0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) , /* df */
 148         W(0xe0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1) | /* ef */
 149         W(0xf0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0)   /* ff */
 150         /*      -----------------------------------------------         */
 151         /*      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f          */
 152 };
 153 #undef W
 154
 155 struct kretprobe_blackpoint kretprobe_blacklist[] = {
 156         {"__switch_to", }, /* This function switches only current task, but
 157                               doesn't switch kernel stack.*/
 158         {NULL, NULL}    /* Terminator */
 159 };
 160 const int kretprobe_blacklist_size = ARRAY_SIZE(kretprobe_blacklist);
 161
 162 /* Insert a jump instruction at address 'from', which jumps to address 'to'.*/
 163 static void __kprobes set_jmp_op(void *from, void *to)
 164 {
 165         struct __arch_jmp_op {
 166                 char op;
 167                 s32 raddr;
 168         } __attribute__((packed)) * jop;
 169         jop = (struct __arch_jmp_op *)from;
 170         jop->raddr = (s32)((long)(to) - ((long)(from) + 5));
 171         jop->op = RELATIVEJUMP_INSTRUCTION;
 172 }
 173
 174 /*
 175  * Check for the REX prefix which can only exist on X86_64
 176  * X86_32 always returns 0
 177  */
 178 static int __kprobes is_REX_prefix(kprobe_opcode_t *insn)
 179 {
 180 #ifdef CONFIG_X86_64
 181         if ((*insn & 0xf0) == 0x40)
 182                 return 1;
 183 #endif
 184         return 0;
 185 }
 186
 187 /*
 188  * Returns non-zero if opcode is boostable.
 189  * RIP relative instructions are adjusted at copying time in 64 bits mode
 190  */
 191 static int __kprobes can_boost(kprobe_opcode_t *opcodes)
 192 {
 193         kprobe_opcode_t opcode;
 194         kprobe_opcode_t *orig_opcodes = opcodes;
 195
 196 retry:
 197         if (opcodes - orig_opcodes > MAX_INSN_SIZE - 1)
 198                 return 0;
 199         opcode = *(opcodes++);
 200
 201         /* 2nd-byte opcode */
 202         if (opcode == 0x0f) {
 203                 if (opcodes - orig_opcodes > MAX_INSN_SIZE - 1)
 204                         return 0;
 205                 return test_bit(*opcodes,
 206                                 (unsigned long *)twobyte_is_boostable);
 207         }
 208
 209         switch (opcode & 0xf0) {
 210 #ifdef CONFIG_X86_64
 211         case 0x40:
 212                 goto retry; /* REX prefix is boostable */
 213 #endif
 214         case 0x60:
 215                 if (0x63 < opcode && opcode < 0x67)
 216                         goto retry; /* prefixes */
 217                 /* can't boost Address-size override and bound */
 218                 return (opcode != 0x62 && opcode != 0x67);
 219         case 0x70:
 220                 return 0; /* can't boost conditional jump */
 221         case 0xc0:
 222                 /* can't boost software-interruptions */
 223                 return (0xc1 < opcode && opcode < 0xcc) || opcode == 0xcf;
 224         case 0xd0:
 225                 /* can boost AA* and XLAT */
 226                 return (opcode == 0xd4 || opcode == 0xd5 || opcode == 0xd7);
 227         case 0xe0:
 228                 /* can boost in/out and absolute jmps */
 229                 return ((opcode & 0x04) || opcode == 0xea);
 230         case 0xf0:
 231                 if ((opcode & 0x0c) == 0 && opcode != 0xf1)
 232                         goto retry; /* lock/rep(ne) prefix */
 233                 /* clear and set flags are boostable */
 234                 return (opcode == 0xf5 || (0xf7 < opcode && opcode < 0xfe));
 235         default:
 236                 /* segment override prefixes are boostable */
 237                 if (opcode == 0x26 || opcode == 0x36 || opcode == 0x3e)
 238                         goto retry; /* prefixes */
 239                 /* CS override prefix and call are not boostable */
 240                 return (opcode != 0x2e && opcode != 0x9a);
 241         }
 242 }
 243
 244 /*
 245  * Returns non-zero if opcode modifies the interrupt flag.
 246  */
 247 static int __kprobes is_IF_modifier(kprobe_opcode_t *insn)
 248 {
 249         switch (*insn) {
 250         case 0xfa:              /* cli */
 251         case 0xfb:              /* sti */
 252         case 0xcf:              /* iret/iretd */
 253         case 0x9d:              /* popf/popfd */
 254                 return 1;
 255         }
 256
 257         /*
 258          * on X86_64, 0x40-0x4f are REX prefixes so we need to look
 259          * at the next byte instead.. but of course not recurse infinitely
 260          */
 261         if (is_REX_prefix(insn))
 262                 return is_IF_modifier(++insn);
 263
 264         return 0;
 265 }
 266
 267 /*
 268  * Adjust the displacement if the instruction uses the %rip-relative
 269  * addressing mode.
 270  * If it does, Return the address of the 32-bit displacement word.
 271  * If not, return null.
 272  * Only applicable to 64-bit x86.
 273  */
 274 static void __kprobes fix_riprel(struct kprobe *p)
 275 {
 276 #ifdef CONFIG_X86_64
 277         u8 *insn = p->ainsn.insn;
 278         s64 disp;
 279         int need_modrm;
 280
 281         /* Skip legacy instruction prefixes.  */
 282         while (1) {
 283                 switch (*insn) {
 284                 case 0x66:
 285                 case 0x67:
 286                 case 0x2e:
 287                 case 0x3e:
 288                 case 0x26:
 289                 case 0x64:
 290                 case 0x65:
 291                 case 0x36:
 292                 case 0xf0:
 293                 case 0xf3:
 294                 case 0xf2:
 295                         ++insn;
 296                         continue;
 297                 }
 298                 break;
 299         }
 300
 301         /* Skip REX instruction prefix.  */
 302         if (is_REX_prefix(insn))
 303                 ++insn;
 304
 305         if (*insn == 0x0f) {
 306                 /* Two-byte opcode.  */
 307                 ++insn;
 308                 need_modrm = test_bit(*insn,
 309                                       (unsigned long *)twobyte_has_modrm);
 310         } else
 311                 /* One-byte opcode.  */
 312                 need_modrm = test_bit(*insn,
 313                                       (unsigned long *)onebyte_has_modrm);
 314
 315         if (need_modrm) {
 316                 u8 modrm = *++insn;
 317                 if ((modrm & 0xc7) == 0x05) {
 318                         /* %rip+disp32 addressing mode */
 319                         /* Displacement follows ModRM byte.  */
 320                         ++insn;
 321                         /*
 322                          * The copied instruction uses the %rip-relative
 323                          * addressing mode.  Adjust the displacement for the
 324                          * difference between the original location of this
 325                          * instruction and the location of the copy that will
 326                          * actually be run.  The tricky bit here is making sure
 327                          * that the sign extension happens correctly in this
 328                          * calculation, since we need a signed 32-bit result to
 329                          * be sign-extended to 64 bits when it's added to the
 330                          * %rip value and yield the same 64-bit result that the
 331                          * sign-extension of the original signed 32-bit
 332                          * displacement would have given.
 333                          */
 334                         disp = (u8 *) p->addr + *((s32 *) insn) -
 335                                (u8 *) p->ainsn.insn;
 336                         BUG_ON((s64) (s32) disp != disp); /* Sanity check.  */
 337                         *(s32 *)insn = (s32) disp;
 338                 }
 339         }
 340 #endif
 341 }
 342
 343 static void __kprobes arch_copy_kprobe(struct kprobe *p)
 344 {
 345         memcpy(p->ainsn.insn, p->addr, MAX_INSN_SIZE * sizeof(kprobe_opcode_t));
 346
 347         fix_riprel(p);
 348
 349         if (can_boost(p->addr))
 350                 p->ainsn.boostable = 0;
 351         else
 352                 p->ainsn.boostable = -1;
 353
 354         p->opcode = *p->addr;
 355 }
 356
 357 int __kprobes arch_prepare_kprobe(struct kprobe *p)
 358 {
 359         /* insn: must be on special executable page on x86. */
 360         p->ainsn.insn = get_insn_slot();
 361         if (!p->ainsn.insn)
 362                 return -ENOMEM;
 363         arch_copy_kprobe(p);
 364         return 0;
 365 }
 366
 367 void __kprobes arch_arm_kprobe(struct kprobe *p)
 368 {
 369         text_poke(p->addr, ((unsigned char []){BREAKPOINT_INSTRUCTION}), 1);
 370 }
 371
 372 void __kprobes arch_disarm_kprobe(struct kprobe *p)
 373 {
 374         text_poke(p->addr, &p->opcode, 1);
 375 }
 376
 377 void __kprobes arch_remove_kprobe(struct kprobe *p)
 378 {
 379         mutex_lock(&kprobe_mutex);
 380         free_insn_slot(p->ainsn.insn, (p->ainsn.boostable == 1));
 381         mutex_unlock(&kprobe_mutex);
 382 }
 383
 384 static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
 385 {
 386         kcb->prev_kprobe.kp = kprobe_running();
 387         kcb->prev_kprobe.status = kcb->kprobe_status;
 388         kcb->prev_kprobe.old_flags = kcb->kprobe_old_flags;
 389         kcb->prev_kprobe.saved_flags = kcb->kprobe_saved_flags;
 390 }
 391
 392 static void __kprobes restore_previous_kprobe(struct kprobe_ctlblk *kcb)
 393 {
 394         __get_cpu_var(current_kprobe) = kcb->prev_kprobe.kp;
 395         kcb->kprobe_status = kcb->prev_kprobe.status;
 396         kcb->kprobe_old_flags = kcb->prev_kprobe.old_flags;
 397         kcb->kprobe_saved_flags = kcb->prev_kprobe.saved_flags;
 398 }
 399
 400 static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
 401                                 struct kprobe_ctlblk *kcb)
 402 {
 403         __get_cpu_var(current_kprobe) = p;
 404         kcb->kprobe_saved_flags = kcb->kprobe_old_flags
 405                 = (regs->flags & (X86_EFLAGS_TF | X86_EFLAGS_IF));
 406         if (is_IF_modifier(p->ainsn.insn))
 407                 kcb->kprobe_saved_flags &= ~X86_EFLAGS_IF;
 408 }
 409
 410 static void __kprobes clear_btf(void)
 411 {
 412         if (test_thread_flag(TIF_DEBUGCTLMSR))
 413                 update_debugctlmsr(0);
 414 }
 415
 416 static void __kprobes restore_btf(void)
 417 {
 418         if (test_thread_flag(TIF_DEBUGCTLMSR))
 419                 update_debugctlmsr(current->thread.debugctlmsr);
 420 }
 421
 422 static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
 423 {
 424         clear_btf();
 425         regs->flags |= X86_EFLAGS_TF;
 426         regs->flags &= ~X86_EFLAGS_IF;
 427         /* single step inline if the instruction is an int3 */
 428         if (p->opcode == BREAKPOINT_INSTRUCTION)
 429                 regs->ip = (unsigned long)p->addr;
 430         else
 431                 regs->ip = (unsigned long)p->ainsn.insn;
 432 }
 433
 434 /* Called with kretprobe_lock held */
 435 void __kprobes arch_prepare_kretprobe(struct kretprobe_instance *ri,
 436                                       struct pt_regs *regs)
 437 {
 438         unsigned long *sara = stack_addr(regs);
 439
 440         ri->ret_addr = (kprobe_opcode_t *) *sara;
 441
 442         /* Replace the return addr with trampoline addr */
 443         *sara = (unsigned long) &kretprobe_trampoline;
 444 }
 445
 446 static void __kprobes setup_singlestep(struct kprobe *p, struct pt_regs *regs,
 447                                        struct kprobe_ctlblk *kcb)
 448 {
 449 #if !defined(CONFIG_PREEMPT) || defined(CONFIG_PM)
 450         if (p->ainsn.boostable == 1 && !p->post_handler) {
 451                 /* Boost up -- we can execute copied instructions directly */
 452                 reset_current_kprobe();
 453                 regs->ip = (unsigned long)p->ainsn.insn;
 454                 preempt_enable_no_resched();
 455                 return;
 456         }
 457 #endif
 458         prepare_singlestep(p, regs);
 459         kcb->kprobe_status = KPROBE_HIT_SS;
 460 }
 461
 462 /*
 463  * We have reentered the kprobe_handler(), since another probe was hit while
 464  * within the handler. We save the original kprobes variables and just single
 465  * step on the instruction of the new probe without calling any user handlers.
 466  */
 467 static int __kprobes reenter_kprobe(struct kprobe *p, struct pt_regs *regs,
 468                                     struct kprobe_ctlblk *kcb)
 469 {
 470         switch (kcb->kprobe_status) {
 471         case KPROBE_HIT_SSDONE:
 472 #ifdef CONFIG_X86_64
 473                 /* TODO: Provide re-entrancy from post_kprobes_handler() and
 474                  * avoid exception stack corruption while single-stepping on
 475                  * the instruction of the new probe.
 476                  */
 477                 arch_disarm_kprobe(p);
 478                 regs->ip = (unsigned long)p->addr;
 479                 reset_current_kprobe();
 480                 preempt_enable_no_resched();
 481                 break;
 482 #endif
 483         case KPROBE_HIT_ACTIVE:
 484                 save_previous_kprobe(kcb);
 485                 set_current_kprobe(p, regs, kcb);
 486                 kprobes_inc_nmissed_count(p);
 487                 prepare_singlestep(p, regs);
 488                 kcb->kprobe_status = KPROBE_REENTER;
 489                 break;
 490         case KPROBE_HIT_SS:
 491                 if (p == kprobe_running()) {
 492                         regs->flags &= ~X86_EFLAGS_TF;
 493                         regs->flags |= kcb->kprobe_saved_flags;
 494                         return 0;
 495                 } else {
 496                         /* A probe has been hit in the codepath leading up
 497                          * to, or just after, single-stepping of a probed
 498                          * instruction. This entire codepath should strictly
 499                          * reside in .kprobes.text section. Raise a warning
 500                          * to highlight this peculiar case.
 501                          */
 502                 }
 503         default:
 504                 /* impossible cases */
 505                 WARN_ON(1);
 506                 return 0;
 507         }
 508
 509         return 1;
 510 }
 511
 512 /*
 513  * Interrupts are disabled on entry as trap3 is an interrupt gate and they
 514  * remain disabled thorough out this function.
 515  */
 516 static int __kprobes kprobe_handler(struct pt_regs *regs)
 517 {
 518         kprobe_opcode_t *addr;
 519         struct kprobe *p;
 520         struct kprobe_ctlblk *kcb;
 521
 522         addr = (kprobe_opcode_t *)(regs->ip - sizeof(kprobe_opcode_t));
 523         if (*addr != BREAKPOINT_INSTRUCTION) {
 524                 /*
 525                  * The breakpoint instruction was removed right
 526                  * after we hit it.  Another cpu has removed
 527                  * either a probepoint or a debugger breakpoint
 528                  * at this address.  In either case, no further
 529                  * handling of this interrupt is appropriate.
 530                  * Back up over the (now missing) int3 and run
 531                  * the original instruction.
 532                  */
 533                 regs->ip = (unsigned long)addr;
 534                 return 1;
 535         }
 536
 537         /*
 538          * We don't want to be preempted for the entire
 539          * duration of kprobe processing. We conditionally
 540          * re-enable preemption at the end of this function,
 541          * and also in reenter_kprobe() and setup_singlestep().
 542          */
 543         preempt_disable();
 544
 545         kcb = get_kprobe_ctlblk();
 546         p = get_kprobe(addr);
 547
 548         if (p) {
 549                 if (kprobe_running()) {
 550                         if (reenter_kprobe(p, regs, kcb))
 551                                 return 1;
 552                 } else {
 553                         set_current_kprobe(p, regs, kcb);
 554                         kcb->kprobe_status = KPROBE_HIT_ACTIVE;
 555
 556                         /*
 557                          * If we have no pre-handler or it returned 0, we
 558                          * continue with normal processing.  If we have a
 559                          * pre-handler and it returned non-zero, it prepped
 560                          * for calling the break_handler below on re-entry
 561                          * for jprobe processing, so get out doing nothing
 562                          * more here.
 563                          */
 564                         if (!p->pre_handler || !p->pre_handler(p, regs))
 565                                 setup_singlestep(p, regs, kcb);
 566                         return 1;
 567                 }
 568         } else if (kprobe_running()) {
 569                 p = __get_cpu_var(current_kprobe);
 570                 if (p->break_handler && p->break_handler(p, regs)) {
 571                         setup_singlestep(p, regs, kcb);
 572                         return 1;
 573                 }
 574         } /* else: not a kprobe fault; let the kernel handle it */
 575
 576         preempt_enable_no_resched();
 577         return 0;
 578 }
 579
 580 /*
 581  * When a retprobed function returns, this code saves registers and
 582  * calls trampoline_handler() runs, which calls the kretprobe's handler.
 583  */
 584 static void __used __kprobes kretprobe_trampoline_holder(void)
 585 {
 586         asm volatile (
 587                         ".global kretprobe_trampoline\n"
 588                         "kretprobe_trampoline: \n"
 589 #ifdef CONFIG_X86_64
 590                         /* We don't bother saving the ss register */
 591                         "       pushq %rsp\n"
 592                         "       pushfq\n"
 593                         /*
 594                          * Skip cs, ip, orig_ax.
 595                          * trampoline_handler() will plug in these values
 596                          */
 597                         "       subq $24, %rsp\n"
 598                         "       pushq %rdi\n"
 599                         "       pushq %rsi\n"
 600                         "       pushq %rdx\n"
 601                         "       pushq %rcx\n"
 602                         "       pushq %rax\n"
 603                         "       pushq %r8\n"
 604                         "       pushq %r9\n"
 605                         "       pushq %r10\n"
 606                         "       pushq %r11\n"
 607                         "       pushq %rbx\n"
 608                         "       pushq %rbp\n"
 609                         "       pushq %r12\n"
 610                         "       pushq %r13\n"
 611                         "       pushq %r14\n"
 612                         "       pushq %r15\n"
 613                         "       movq %rsp, %rdi\n"
 614                         "       call trampoline_handler\n"
 615                         /* Replace saved sp with true return address. */
 616                         "       movq %rax, 152(%rsp)\n"
 617                         "       popq %r15\n"
 618                         "       popq %r14\n"
 619                         "       popq %r13\n"
 620                         "       popq %r12\n"
 621                         "       popq %rbp\n"
 622                         "       popq %rbx\n"
 623                         "       popq %r11\n"
 624                         "       popq %r10\n"
 625                         "       popq %r9\n"
 626                         "       popq %r8\n"
 627                         "       popq %rax\n"
 628                         "       popq %rcx\n"
 629                         "       popq %rdx\n"
 630                         "       popq %rsi\n"
 631                         "       popq %rdi\n"
 632                         /* Skip orig_ax, ip, cs */
 633                         "       addq $24, %rsp\n"
 634                         "       popfq\n"
 635 #else
 636                         "       pushf\n"
 637                         /*
 638                          * Skip cs, ip, orig_ax.
 639                          * trampoline_handler() will plug in these values
 640                          */
 641                         "       subl $12, %esp\n"
 642                         "       pushl %fs\n"
 643                         "       pushl %ds\n"
 644                         "       pushl %es\n"
 645                         "       pushl %eax\n"
 646                         "       pushl %ebp\n"
 647                         "       pushl %edi\n"
 648                         "       pushl %esi\n"
 649                         "       pushl %edx\n"
 650                         "       pushl %ecx\n"
 651                         "       pushl %ebx\n"
 652                         "       movl %esp, %eax\n"
 653                         "       call trampoline_handler\n"
 654                         /* Move flags to cs */
 655                         "       movl 52(%esp), %edx\n"
 656                         "       movl %edx, 48(%esp)\n"
 657                         /* Replace saved flags with true return address. */
 658                         "       movl %eax, 52(%esp)\n"
 659                         "       popl %ebx\n"
 660                         "       popl %ecx\n"
 661                         "       popl %edx\n"
 662                         "       popl %esi\n"
 663                         "       popl %edi\n"
 664                         "       popl %ebp\n"
 665                         "       popl %eax\n"
 666                         /* Skip ip, orig_ax, es, ds, fs */
 667                         "       addl $20, %esp\n"
 668                         "       popf\n"
 669 #endif
 670                         "       ret\n");
 671 }
 672
 673 /*
 674  * Called from kretprobe_trampoline
 675  */
 676 static __used __kprobes void *trampoline_handler(struct pt_regs *regs)
 677 {
 678         struct kretprobe_instance *ri = NULL;
 679         struct hlist_head *head, empty_rp;
 680         struct hlist_node *node, *tmp;
 681         unsigned long flags, orig_ret_address = 0;
 682         unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
 683
 684         INIT_HLIST_HEAD(&empty_rp);
 685         spin_lock_irqsave(&kretprobe_lock, flags);
 686         head = kretprobe_inst_table_head(current);
 687         /* fixup registers */
 688 #ifdef CONFIG_X86_64
 689         regs->cs = __KERNEL_CS;
 690 #else
 691         regs->cs = __KERNEL_CS | get_kernel_rpl();
 692 #endif
 693         regs->ip = trampoline_address;
 694         regs->orig_ax = ~0UL;
 695
 696         /*
 697          * It is possible to have multiple instances associated with a given
 698          * task either because multiple functions in the call path have
 699          * return probes installed on them, and/or more then one
 700          * return probe was registered for a target function.
 701          *
 702          * We can handle this because:
 703          *     - instances are always pushed into the head of the list
 704          *     - when multiple return probes are registered for the same
 705          *       function, the (chronologically) first instance's ret_addr
 706          *       will be the real return address, and all the rest will
 707          *       point to kretprobe_trampoline.
 708          */
 709         hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
 710                 if (ri->task != current)
 711                         /* another task is sharing our hash bucket */
 712                         continue;
 713
 714                 if (ri->rp && ri->rp->handler) {
 715                         __get_cpu_var(current_kprobe) = &ri->rp->kp;
 716                         get_kprobe_ctlblk()->kprobe_status = KPROBE_HIT_ACTIVE;
 717                         ri->rp->handler(ri, regs);
 718                         __get_cpu_var(current_kprobe) = NULL;
 719                 }
 720
 721                 orig_ret_address = (unsigned long)ri->ret_addr;
 722                 recycle_rp_inst(ri, &empty_rp);
 723
 724                 if (orig_ret_address != trampoline_address)
 725                         /*
 726                          * This is the real return address. Any other
 727                          * instances associated with this task are for
 728                          * other calls deeper on the call stack
 729                          */
 730                         break;
 731         }
 732
 733         kretprobe_assert(ri, orig_ret_address, trampoline_address);
 734
 735         spin_unlock_irqrestore(&kretprobe_lock, flags);
 736
 737         hlist_for_each_entry_safe(ri, node, tmp, &empty_rp, hlist) {
 738                 hlist_del(&ri->hlist);
 739                 kfree(ri);
 740         }
 741         return (void *)orig_ret_address;
 742 }
 743
 744 /*
 745  * Called after single-stepping.  p->addr is the address of the
 746  * instruction whose first byte has been replaced by the "int 3"
 747  * instruction.  To avoid the SMP problems that can occur when we
 748  * temporarily put back the original opcode to single-step, we
 749  * single-stepped a copy of the instruction.  The address of this
 750  * copy is p->ainsn.insn.
 751  *
 752  * This function prepares to return from the post-single-step
 753  * interrupt.  We have to fix up the stack as follows:
 754  *
 755  * 0) Except in the case of absolute or indirect jump or call instructions,
 756  * the new ip is relative to the copied instruction.  We need to make
 757  * it relative to the original instruction.
 758  *
 759  * 1) If the single-stepped instruction was pushfl, then the TF and IF
 760  * flags are set in the just-pushed flags, and may need to be cleared.
 761  *
 762  * 2) If the single-stepped instruction was a call, the return address
 763  * that is atop the stack is the address following the copied instruction.
 764  * We need to make it the address following the original instruction.
 765  *
 766  * If this is the first time we've single-stepped the instruction at
 767  * this probepoint, and the instruction is boostable, boost it: add a
 768  * jump instruction after the copied instruction, that jumps to the next
 769  * instruction after the probepoint.
 770  */
 771 static void __kprobes resume_execution(struct kprobe *p,
 772                 struct pt_regs *regs, struct kprobe_ctlblk *kcb)
 773 {
 774         unsigned long *tos = stack_addr(regs);
 775         unsigned long copy_ip = (unsigned long)p->ainsn.insn;
 776         unsigned long orig_ip = (unsigned long)p->addr;
 777         kprobe_opcode_t *insn = p->ainsn.insn;
 778
 779         /*skip the REX prefix*/
 780         if (is_REX_prefix(insn))
 781                 insn++;
 782
 783         regs->flags &= ~X86_EFLAGS_TF;
 784         switch (*insn) {
 785         case 0x9c:      /* pushfl */
 786                 *tos &= ~(X86_EFLAGS_TF | X86_EFLAGS_IF);
 787                 *tos |= kcb->kprobe_old_flags;
 788                 break;
 789         case 0xc2:      /* iret/ret/lret */
 790         case 0xc3:
 791         case 0xca:
 792         case 0xcb:
 793         case 0xcf:
 794         case 0xea:      /* jmp absolute -- ip is correct */
 795                 /* ip is already adjusted, no more changes required */
 796                 p->ainsn.boostable = 1;
 797                 goto no_change;
 798         case 0xe8:      /* call relative - Fix return addr */
 799                 *tos = orig_ip + (*tos - copy_ip);
 800                 break;
 801 #ifdef CONFIG_X86_32
 802         case 0x9a:      /* call absolute -- same as call absolute, indirect */
 803                 *tos = orig_ip + (*tos - copy_ip);
 804                 goto no_change;
 805 #endif
 806         case 0xff:
 807                 if ((insn[1] & 0x30) == 0x10) {
 808                         /*
 809                          * call absolute, indirect
 810                          * Fix return addr; ip is correct.
 811                          * But this is not boostable
 812                          */
 813                         *tos = orig_ip + (*tos - copy_ip);
 814                         goto no_change;
 815                 } else if (((insn[1] & 0x31) == 0x20) ||
 816                            ((insn[1] & 0x31) == 0x21)) {
 817                         /*
 818                          * jmp near and far, absolute indirect
 819                          * ip is correct. And this is boostable
 820                          */
 821                         p->ainsn.boostable = 1;
 822                         goto no_change;
 823                 }
 824         default:
 825                 break;
 826         }
 827
 828         if (p->ainsn.boostable == 0) {
 829                 if ((regs->ip > copy_ip) &&
 830                     (regs->ip - copy_ip) + 5 < MAX_INSN_SIZE) {
 831                         /*
 832                          * These instructions can be executed directly if it
 833                          * jumps back to correct address.
 834                          */
 835                         set_jmp_op((void *)regs->ip,
 836                                    (void *)orig_ip + (regs->ip - copy_ip));
 837                         p->ainsn.boostable = 1;
 838                 } else {
 839                         p->ainsn.boostable = -1;
 840                 }
 841         }
 842
 843         regs->ip += orig_ip - copy_ip;
 844
 845 no_change:
 846         restore_btf();
 847 }
 848
 849 /*
 850  * Interrupts are disabled on entry as trap1 is an interrupt gate and they
 851  * remain disabled thoroughout this function.
 852  */
 853 static int __kprobes post_kprobe_handler(struct pt_regs *regs)
 854 {
 855         struct kprobe *cur = kprobe_running();
 856         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
 857
 858         if (!cur)
 859                 return 0;
 860
 861         resume_execution(cur, regs, kcb);
 862         regs->flags |= kcb->kprobe_saved_flags;
 863         trace_hardirqs_fixup_flags(regs->flags);
 864
 865         if ((kcb->kprobe_status != KPROBE_REENTER) && cur->post_handler) {
 866                 kcb->kprobe_status = KPROBE_HIT_SSDONE;
 867                 cur->post_handler(cur, regs, 0);
 868         }
 869
 870         /* Restore back the original saved kprobes variables and continue. */
 871         if (kcb->kprobe_status == KPROBE_REENTER) {
 872                 restore_previous_kprobe(kcb);
 873                 goto out;
 874         }
 875         reset_current_kprobe();
 876 out:
 877         preempt_enable_no_resched();
 878
 879         /*
 880          * if somebody else is singlestepping across a probe point, flags
 881          * will have TF set, in which case, continue the remaining processing
 882          * of do_debug, as if this is not a probe hit.
 883          */
 884         if (regs->flags & X86_EFLAGS_TF)
 885                 return 0;
 886
 887         return 1;
 888 }
 889
 890 int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
 891 {
 892         struct kprobe *cur = kprobe_running();
 893         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
 894
 895         switch (kcb->kprobe_status) {
 896         case KPROBE_HIT_SS:
 897         case KPROBE_REENTER:
 898                 /*
 899                  * We are here because the instruction being single
 900                  * stepped caused a page fault. We reset the current
 901                  * kprobe and the ip points back to the probe address
 902                  * and allow the page fault handler to continue as a
 903                  * normal page fault.
 904                  */
 905                 regs->ip = (unsigned long)cur->addr;
 906                 regs->flags |= kcb->kprobe_old_flags;
 907                 if (kcb->kprobe_status == KPROBE_REENTER)
 908                         restore_previous_kprobe(kcb);
 909                 else
 910                         reset_current_kprobe();
 911                 preempt_enable_no_resched();
 912                 break;
 913         case KPROBE_HIT_ACTIVE:
 914         case KPROBE_HIT_SSDONE:
 915                 /*
 916                  * We increment the nmissed count for accounting,
 917                  * we can also use npre/npostfault count for accounting
 918                  * these specific fault cases.
 919                  */
 920                 kprobes_inc_nmissed_count(cur);
 921
 922                 /*
 923                  * We come here because instructions in the pre/post
 924                  * handler caused the page_fault, this could happen
 925                  * if handler tries to access user space by
 926                  * copy_from_user(), get_user() etc. Let the
 927                  * user-specified handler try to fix it first.
 928                  */
 929                 if (cur->fault_handler && cur->fault_handler(cur, regs, trapnr))
 930                         return 1;
 931
 932                 /*
 933                  * In case the user-specified fault handler returned
 934                  * zero, try to fix up.
 935                  */
 936                 if (fixup_exception(regs))
 937                         return 1;
 938
 939                 /*
 940                  * fixup routine could not handle it,
 941                  * Let do_page_fault() fix it.
 942                  */
 943                 break;
 944         default:
 945                 break;
 946         }
 947         return 0;
 948 }
 949
 950 /*
 951  * Wrapper routine for handling exceptions.
 952  */
 953 int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
 954                                        unsigned long val, void *data)
 955 {
 956         struct die_args *args = data;
 957         int ret = NOTIFY_DONE;
 958
 959         if (args->regs && user_mode_vm(args->regs))
 960                 return ret;
 961
 962         switch (val) {
 963         case DIE_INT3:
 964                 if (kprobe_handler(args->regs))
 965                         ret = NOTIFY_STOP;
 966                 break;
 967         case DIE_DEBUG:
 968                 if (post_kprobe_handler(args->regs))
 969                         ret = NOTIFY_STOP;
 970                 break;
 971         case DIE_GPF:
 972                 /*
 973                  * To be potentially processing a kprobe fault and to
 974                  * trust the result from kprobe_running(), we have
 975                  * be non-preemptible.
 976                  */
 977                 if (!preemptible() && kprobe_running() &&
 978                     kprobe_fault_handler(args->regs, args->trapnr))
 979                         ret = NOTIFY_STOP;
 980                 break;
 981         default:
 982                 break;
 983         }
 984         return ret;
 985 }
 986
 987 int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
 988 {
 989         struct jprobe *jp = container_of(p, struct jprobe, kp);
 990         unsigned long addr;
 991         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
 992
 993         kcb->jprobe_saved_regs = *regs;
 994         kcb->jprobe_saved_sp = stack_addr(regs);
 995         addr = (unsigned long)(kcb->jprobe_saved_sp);
 996
 997         /*
 998          * As Linus pointed out, gcc assumes that the callee
 999          * owns the argument space and could overwrite it, e.g.
1000          * tailcall optimization. So, to be absolutely safe
1001          * we also save and restore enough stack bytes to cover
1002          * the argument area.
1003          */
1004         memcpy(kcb->jprobes_stack, (kprobe_opcode_t *)addr,
1005                MIN_STACK_SIZE(addr));
1006         regs->flags &= ~X86_EFLAGS_IF;
1007         trace_hardirqs_off();
1008         regs->ip = (unsigned long)(jp->entry);
1009         return 1;
1010 }
1011
1012 void __kprobes jprobe_return(void)
1013 {
1014         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
1015
1016         asm volatile (
1017 #ifdef CONFIG_X86_64
1018                         "       xchg   %%rbx,%%rsp      \n"
1019 #else
1020                         "       xchgl   %%ebx,%%esp     \n"
1021 #endif
1022                         "       int3                    \n"
1023                         "       .globl jprobe_return_end\n"
1024                         "       jprobe_return_end:      \n"
1025                         "       nop                     \n"::"b"
1026                         (kcb->jprobe_saved_sp):"memory");
1027 }
1028
1029 int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
1030 {
1031         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
1032         u8 *addr = (u8 *) (regs->ip - 1);
1033         struct jprobe *jp = container_of(p, struct jprobe, kp);
1034
1035         if ((addr > (u8 *) jprobe_return) &&
1036             (addr < (u8 *) jprobe_return_end)) {
1037                 if (stack_addr(regs) != kcb->jprobe_saved_sp) {
1038                         struct pt_regs *saved_regs = &kcb->jprobe_saved_regs;
1039                         printk(KERN_ERR
1040                                "current sp %p does not match saved sp %p\n",
1041                                stack_addr(regs), kcb->jprobe_saved_sp);
1042                         printk(KERN_ERR "Saved registers for jprobe %p\n", jp);
1043                         show_registers(saved_regs);
1044                         printk(KERN_ERR "Current registers\n");
1045                         show_registers(regs);
1046                         BUG();
1047                 }
1048                 *regs = kcb->jprobe_saved_regs;
1049                 memcpy((kprobe_opcode_t *)(kcb->jprobe_saved_sp),
1050                        kcb->jprobes_stack,
1051                        MIN_STACK_SIZE(kcb->jprobe_saved_sp));
1052                 preempt_enable_no_resched();
1053                 return 1;
1054         }
1055         return 0;
1056 }
1057
1058 int __init arch_init_kprobes(void)
1059 {
1060         return 0;
1061 }
1062
1063 int __kprobes arch_trampoline_kprobe(struct kprobe *p)
1064 {
1065         return 0;
1066 }