gcc/c-family/c-ubsan.c

   1 /* UndefinedBehaviorSanitizer, undefined behavior detector.
   2    Copyright (C) 2013-2017 Free Software Foundation, Inc.
   3    Contributed by Marek Polacek <polacek@redhat.com>
   4
   5 This file is part of GCC.
   6
   7 GCC is free software; you can redistribute it and/or modify it under
   8 the terms of the GNU General Public License as published by the Free
   9 Software Foundation; either version 3, or (at your option) any later
  10 version.
  11
  12 GCC is distributed in the hope that it will be useful, but WITHOUT ANY
  13 WARRANTY; without even the implied warranty of MERCHANTABILITY or
  14 FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
  15 for more details.
  16
  17 You should have received a copy of the GNU General Public License
  18 along with GCC; see the file COPYING3.  If not see
  19 <http://www.gnu.org/licenses/>.  */
  20
  21 #include "config.h"
  22 #include "system.h"
  23 #include "coretypes.h"
  24 #include "tm.h"
  25 #include "c-family/c-common.h"
  26 #include "ubsan.h"
  27 #include "c-family/c-ubsan.h"
  28 #include "asan.h"
  29 #include "stor-layout.h"
  30 #include "builtins.h"
  31 #include "gimplify.h"
  32
  33 /* Instrument division by zero and INT_MIN / -1.  If not instrumenting,
  34    return NULL_TREE.  */
  35
  36 tree
  37 ubsan_instrument_division (location_t loc, tree op0, tree op1)
  38 {
  39   tree t, tt;
  40   tree type = TREE_TYPE (op0);
  41
  42   /* At this point both operands should have the same type,
  43      because they are already converted to RESULT_TYPE.
  44      Use TYPE_MAIN_VARIANT since typedefs can confuse us.  */
  45   gcc_assert (TYPE_MAIN_VARIANT (TREE_TYPE (op0))
  46               == TYPE_MAIN_VARIANT (TREE_TYPE (op1)));
  47
  48   op0 = unshare_expr (op0);
  49   op1 = unshare_expr (op1);
  50
  51   if (TREE_CODE (type) == INTEGER_TYPE
  52       && (flag_sanitize & SANITIZE_DIVIDE))
  53     t = fold_build2 (EQ_EXPR, boolean_type_node,
  54                      op1, build_int_cst (type, 0));
  55   else if (TREE_CODE (type) == REAL_TYPE
  56            && (flag_sanitize & SANITIZE_FLOAT_DIVIDE))
  57     t = fold_build2 (EQ_EXPR, boolean_type_node,
  58                      op1, build_real (type, dconst0));
  59   else
  60     return NULL_TREE;
  61
  62   /* We check INT_MIN / -1 only for signed types.  */
  63   if (TREE_CODE (type) == INTEGER_TYPE
  64       && (flag_sanitize & SANITIZE_DIVIDE)
  65       && !TYPE_UNSIGNED (type))
  66     {
  67       tree x;
  68       tt = fold_build2 (EQ_EXPR, boolean_type_node, unshare_expr (op1),
  69                         build_int_cst (type, -1));
  70       x = fold_build2 (EQ_EXPR, boolean_type_node, op0,
  71                        TYPE_MIN_VALUE (type));
  72       x = fold_build2 (TRUTH_AND_EXPR, boolean_type_node, x, tt);
  73       t = fold_build2 (TRUTH_OR_EXPR, boolean_type_node, t, x);
  74     }
  75
  76   /* If the condition was folded to 0, no need to instrument
  77      this expression.  */
  78   if (integer_zerop (t))
  79     return NULL_TREE;
  80
  81   /* In case we have a SAVE_EXPR in a conditional context, we need to
  82      make sure it gets evaluated before the condition.  */
  83   t = fold_build2 (COMPOUND_EXPR, TREE_TYPE (t), unshare_expr (op0), t);
  84   t = fold_build2 (COMPOUND_EXPR, TREE_TYPE (t), unshare_expr (op1), t);
  85   if (flag_sanitize_undefined_trap_on_error)
  86     tt = build_call_expr_loc (loc, builtin_decl_explicit (BUILT_IN_TRAP), 0);
  87   else
  88     {
  89       tree data = ubsan_create_data ("__ubsan_overflow_data", 1, &loc,
  90                                      ubsan_type_descriptor (type), NULL_TREE,
  91                                      NULL_TREE);
  92       data = build_fold_addr_expr_loc (loc, data);
  93       enum built_in_function bcode
  94         = (flag_sanitize_recover & SANITIZE_DIVIDE)
  95           ? BUILT_IN_UBSAN_HANDLE_DIVREM_OVERFLOW
  96           : BUILT_IN_UBSAN_HANDLE_DIVREM_OVERFLOW_ABORT;
  97       tt = builtin_decl_explicit (bcode);
  98       op0 = unshare_expr (op0);
  99       op1 = unshare_expr (op1);
 100       tt = build_call_expr_loc (loc, tt, 3, data, ubsan_encode_value (op0),
 101                                 ubsan_encode_value (op1));
 102     }
 103   t = fold_build3 (COND_EXPR, void_type_node, t, tt, void_node);
 104
 105   return t;
 106 }
 107
 108 /* Instrument left and right shifts.  */
 109
 110 tree
 111 ubsan_instrument_shift (location_t loc, enum tree_code code,
 112                         tree op0, tree op1)
 113 {
 114   tree t, tt = NULL_TREE;
 115   tree type0 = TREE_TYPE (op0);
 116   tree type1 = TREE_TYPE (op1);
 117   if (!INTEGRAL_TYPE_P (type0))
 118     return NULL_TREE;
 119
 120   tree op1_utype = unsigned_type_for (type1);
 121   HOST_WIDE_INT op0_prec = TYPE_PRECISION (type0);
 122   tree uprecm1 = build_int_cst (op1_utype, op0_prec - 1);
 123
 124   op0 = unshare_expr (op0);
 125   op1 = unshare_expr (op1);
 126
 127   t = fold_convert_loc (loc, op1_utype, op1);
 128   t = fold_build2 (GT_EXPR, boolean_type_node, t, uprecm1);
 129
 130   /* If this is not a signed operation, don't perform overflow checks.
 131      Also punt on bit-fields.  */
 132   if (TYPE_OVERFLOW_WRAPS (type0)
 133       || GET_MODE_BITSIZE (TYPE_MODE (type0)) != TYPE_PRECISION (type0)
 134       || (flag_sanitize & SANITIZE_SHIFT_BASE) == 0)
 135     ;
 136
 137   /* For signed x << y, in C99/C11, the following:
 138      (unsigned) x >> (uprecm1 - y)
 139      if non-zero, is undefined.  */
 140   else if (code == LSHIFT_EXPR && flag_isoc99 && cxx_dialect < cxx11)
 141     {
 142       tree x = fold_build2 (MINUS_EXPR, op1_utype, uprecm1,
 143                             fold_convert (op1_utype, unshare_expr (op1)));
 144       tt = fold_convert_loc (loc, unsigned_type_for (type0), op0);
 145       tt = fold_build2 (RSHIFT_EXPR, TREE_TYPE (tt), tt, x);
 146       tt = fold_build2 (NE_EXPR, boolean_type_node, tt,
 147                         build_int_cst (TREE_TYPE (tt), 0));
 148     }
 149
 150   /* For signed x << y, in C++11 and later, the following:
 151      x < 0 || ((unsigned) x >> (uprecm1 - y))
 152      if > 1, is undefined.  */
 153   else if (code == LSHIFT_EXPR && cxx_dialect >= cxx11)
 154     {
 155       tree x = fold_build2 (MINUS_EXPR, op1_utype, uprecm1,
 156                             fold_convert (op1_utype, unshare_expr (op1)));
 157       tt = fold_convert_loc (loc, unsigned_type_for (type0),
 158                              unshare_expr (op0));
 159       tt = fold_build2 (RSHIFT_EXPR, TREE_TYPE (tt), tt, x);
 160       tt = fold_build2 (GT_EXPR, boolean_type_node, tt,
 161                         build_int_cst (TREE_TYPE (tt), 1));
 162       x = fold_build2 (LT_EXPR, boolean_type_node, unshare_expr (op0),
 163                        build_int_cst (type0, 0));
 164       tt = fold_build2 (TRUTH_OR_EXPR, boolean_type_node, x, tt);
 165     }
 166
 167   /* If the condition was folded to 0, no need to instrument
 168      this expression.  */
 169   if (integer_zerop (t) && (tt == NULL_TREE || integer_zerop (tt)))
 170     return NULL_TREE;
 171
 172   /* In case we have a SAVE_EXPR in a conditional context, we need to
 173      make sure it gets evaluated before the condition.  */
 174   t = fold_build2 (COMPOUND_EXPR, TREE_TYPE (t), unshare_expr (op0), t);
 175
 176   enum sanitize_code recover_kind = SANITIZE_SHIFT_EXPONENT;
 177   tree else_t = void_node;
 178   if (tt)
 179     {
 180       if ((flag_sanitize & SANITIZE_SHIFT_EXPONENT) == 0)
 181         {
 182           t = fold_build1 (TRUTH_NOT_EXPR, boolean_type_node, t);
 183           t = fold_build2 (TRUTH_AND_EXPR, boolean_type_node, t, tt);
 184           recover_kind = SANITIZE_SHIFT_BASE;
 185         }
 186       else
 187         {
 188           if (flag_sanitize_undefined_trap_on_error
 189               || ((!(flag_sanitize_recover & SANITIZE_SHIFT_EXPONENT))
 190                   == (!(flag_sanitize_recover & SANITIZE_SHIFT_BASE))))
 191             t = fold_build2 (TRUTH_OR_EXPR, boolean_type_node, t, tt);
 192           else
 193             else_t = tt;
 194         }
 195     }
 196
 197   if (flag_sanitize_undefined_trap_on_error)
 198     tt = build_call_expr_loc (loc, builtin_decl_explicit (BUILT_IN_TRAP), 0);
 199   else
 200     {
 201       tree data = ubsan_create_data ("__ubsan_shift_data", 1, &loc,
 202                                      ubsan_type_descriptor (type0),
 203                                      ubsan_type_descriptor (type1), NULL_TREE,
 204                                      NULL_TREE);
 205       data = build_fold_addr_expr_loc (loc, data);
 206
 207       enum built_in_function bcode
 208         = (flag_sanitize_recover & recover_kind)
 209           ? BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS
 210           : BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS_ABORT;
 211       tt = builtin_decl_explicit (bcode);
 212       op0 = unshare_expr (op0);
 213       op1 = unshare_expr (op1);
 214       tt = build_call_expr_loc (loc, tt, 3, data, ubsan_encode_value (op0),
 215                                 ubsan_encode_value (op1));
 216       if (else_t != void_node)
 217         {
 218           bcode = (flag_sanitize_recover & SANITIZE_SHIFT_BASE)
 219                   ? BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS
 220                   : BUILT_IN_UBSAN_HANDLE_SHIFT_OUT_OF_BOUNDS_ABORT;
 221           tree else_tt = builtin_decl_explicit (bcode);
 222           op0 = unshare_expr (op0);
 223           op1 = unshare_expr (op1);
 224           else_tt = build_call_expr_loc (loc, else_tt, 3, data,
 225                                          ubsan_encode_value (op0),
 226                                          ubsan_encode_value (op1));
 227           else_t = fold_build3 (COND_EXPR, void_type_node, else_t,
 228                                 else_tt, void_node);
 229         }
 230     }
 231   t = fold_build3 (COND_EXPR, void_type_node, t, tt, else_t);
 232
 233   return t;
 234 }
 235
 236 /* Instrument variable length array bound.  */
 237
 238 tree
 239 ubsan_instrument_vla (location_t loc, tree size)
 240 {
 241   tree type = TREE_TYPE (size);
 242   tree t, tt;
 243
 244   t = fold_build2 (LE_EXPR, boolean_type_node, size, build_int_cst (type, 0));
 245   if (flag_sanitize_undefined_trap_on_error)
 246     tt = build_call_expr_loc (loc, builtin_decl_explicit (BUILT_IN_TRAP), 0);
 247   else
 248     {
 249       tree data = ubsan_create_data ("__ubsan_vla_data", 1, &loc,
 250                                      ubsan_type_descriptor (type), NULL_TREE,
 251                                      NULL_TREE);
 252       data = build_fold_addr_expr_loc (loc, data);
 253       enum built_in_function bcode
 254         = (flag_sanitize_recover & SANITIZE_VLA)
 255           ? BUILT_IN_UBSAN_HANDLE_VLA_BOUND_NOT_POSITIVE
 256           : BUILT_IN_UBSAN_HANDLE_VLA_BOUND_NOT_POSITIVE_ABORT;
 257       tt = builtin_decl_explicit (bcode);
 258       tt = build_call_expr_loc (loc, tt, 2, data, ubsan_encode_value (size));
 259     }
 260   t = fold_build3 (COND_EXPR, void_type_node, t, tt, void_node);
 261
 262   return t;
 263 }
 264
 265 /* Instrument missing return in C++ functions returning non-void.  */
 266
 267 tree
 268 ubsan_instrument_return (location_t loc)
 269 {
 270   if (flag_sanitize_undefined_trap_on_error)
 271     return build_call_expr_loc (loc, builtin_decl_explicit (BUILT_IN_TRAP), 0);
 272
 273   tree data = ubsan_create_data ("__ubsan_missing_return_data", 1, &loc,
 274                                  NULL_TREE, NULL_TREE);
 275   tree t = builtin_decl_explicit (BUILT_IN_UBSAN_HANDLE_MISSING_RETURN);
 276   return build_call_expr_loc (loc, t, 1, build_fold_addr_expr_loc (loc, data));
 277 }
 278
 279 /* Instrument array bounds for ARRAY_REFs.  We create special builtin,
 280    that gets expanded in the sanopt pass, and make an array dimension
 281    of it.  ARRAY is the array, *INDEX is an index to the array.
 282    Return NULL_TREE if no instrumentation is emitted.
 283    IGNORE_OFF_BY_ONE is true if the ARRAY_REF is inside a ADDR_EXPR.  */
 284
 285 tree
 286 ubsan_instrument_bounds (location_t loc, tree array, tree *index,
 287                          bool ignore_off_by_one)
 288 {
 289   tree type = TREE_TYPE (array);
 290   tree domain = TYPE_DOMAIN (type);
 291
 292   if (domain == NULL_TREE || TYPE_MAX_VALUE (domain) == NULL_TREE)
 293     return NULL_TREE;
 294
 295   tree bound = TYPE_MAX_VALUE (domain);
 296   if (ignore_off_by_one)
 297     bound = fold_build2 (PLUS_EXPR, TREE_TYPE (bound), bound,
 298                          build_int_cst (TREE_TYPE (bound), 1));
 299
 300   /* Detect flexible array members and suchlike, unless
 301      -fsanitize=bounds-strict.  */
 302   tree base = get_base_address (array);
 303   if ((flag_sanitize & SANITIZE_BOUNDS_STRICT) == 0
 304       && TREE_CODE (array) == COMPONENT_REF
 305       && base && (INDIRECT_REF_P (base) || TREE_CODE (base) == MEM_REF))
 306     {
 307       tree next = NULL_TREE;
 308       tree cref = array;
 309
 310       /* Walk all structs/unions.  */
 311       while (TREE_CODE (cref) == COMPONENT_REF)
 312         {
 313           if (TREE_CODE (TREE_TYPE (TREE_OPERAND (cref, 0))) == RECORD_TYPE)
 314             for (next = DECL_CHAIN (TREE_OPERAND (cref, 1));
 315                  next && TREE_CODE (next) != FIELD_DECL;
 316                  next = DECL_CHAIN (next))
 317               ;
 318           if (next)
 319             /* Not a last element.  Instrument it.  */
 320             break;
 321           /* Ok, this is the last field of the structure/union.  But the
 322              aggregate containing the field must be the last field too,
 323              recursively.  */
 324           cref = TREE_OPERAND (cref, 0);
 325         }
 326       if (!next)
 327         /* Don't instrument this flexible array member-like array in non-strict
 328            -fsanitize=bounds mode.  */
 329         return NULL_TREE;
 330     }
 331
 332   /* Don't emit instrumentation in the most common cases.  */
 333   tree idx = NULL_TREE;
 334   if (TREE_CODE (*index) == INTEGER_CST)
 335     idx = *index;
 336   else if (TREE_CODE (*index) == BIT_AND_EXPR
 337            && TREE_CODE (TREE_OPERAND (*index, 1)) == INTEGER_CST)
 338     idx = TREE_OPERAND (*index, 1);
 339   if (idx
 340       && TREE_CODE (bound) == INTEGER_CST
 341       && tree_int_cst_sgn (idx) >= 0
 342       && tree_int_cst_le (idx, bound))
 343     return NULL_TREE;
 344
 345   *index = save_expr (*index);
 346   /* Create a "(T *) 0" tree node to describe the array type.  */
 347   tree zero_with_type = build_int_cst (build_pointer_type (type), 0);
 348   return build_call_expr_internal_loc (loc, IFN_UBSAN_BOUNDS,
 349                                        void_type_node, 3, zero_with_type,
 350                                        *index, bound);
 351 }
 352
 353 /* Return true iff T is an array that was instrumented by SANITIZE_BOUNDS.  */
 354
 355 bool
 356 ubsan_array_ref_instrumented_p (const_tree t)
 357 {
 358   if (TREE_CODE (t) != ARRAY_REF)
 359     return false;
 360
 361   tree op1 = TREE_OPERAND (t, 1);
 362   return TREE_CODE (op1) == COMPOUND_EXPR
 363          && TREE_CODE (TREE_OPERAND (op1, 0)) == CALL_EXPR
 364          && CALL_EXPR_FN (TREE_OPERAND (op1, 0)) == NULL_TREE
 365          && CALL_EXPR_IFN (TREE_OPERAND (op1, 0)) == IFN_UBSAN_BOUNDS;
 366 }
 367
 368 /* Instrument an ARRAY_REF, if it hasn't already been instrumented.
 369    IGNORE_OFF_BY_ONE is true if the ARRAY_REF is inside a ADDR_EXPR.  */
 370
 371 void
 372 ubsan_maybe_instrument_array_ref (tree *expr_p, bool ignore_off_by_one)
 373 {
 374   if (!ubsan_array_ref_instrumented_p (*expr_p)
 375       && do_ubsan_in_current_function ())
 376     {
 377       tree op0 = TREE_OPERAND (*expr_p, 0);
 378       tree op1 = TREE_OPERAND (*expr_p, 1);
 379       tree e = ubsan_instrument_bounds (EXPR_LOCATION (*expr_p), op0, &op1,
 380                                         ignore_off_by_one);
 381       if (e != NULL_TREE)
 382         {
 383           tree t = copy_node (*expr_p);
 384           TREE_OPERAND (t, 1) = build2 (COMPOUND_EXPR, TREE_TYPE (op1),
 385                                         e, op1);
 386           *expr_p = t;
 387         }
 388     }
 389 }
 390
 391 static tree
 392 ubsan_maybe_instrument_reference_or_call (location_t loc, tree op, tree ptype,
 393                                           enum ubsan_null_ckind ckind)
 394 {
 395   if (!do_ubsan_in_current_function ())
 396     return NULL_TREE;
 397
 398   tree type = TREE_TYPE (ptype);
 399   tree orig_op = op;
 400   bool instrument = false;
 401   unsigned int mina = 0;
 402
 403   if (flag_sanitize & SANITIZE_ALIGNMENT)
 404     {
 405       mina = min_align_of_type (type);
 406       if (mina <= 1)
 407         mina = 0;
 408     }
 409   while ((TREE_CODE (op) == NOP_EXPR
 410           || TREE_CODE (op) == NON_LVALUE_EXPR)
 411          && TREE_CODE (TREE_TYPE (op)) == POINTER_TYPE)
 412     op = TREE_OPERAND (op, 0);
 413   if (TREE_CODE (op) == NOP_EXPR
 414       && TREE_CODE (TREE_TYPE (op)) == REFERENCE_TYPE)
 415     {
 416       if (mina && mina > min_align_of_type (TREE_TYPE (TREE_TYPE (op))))
 417         instrument = true;
 418     }
 419   else
 420     {
 421       if ((flag_sanitize & SANITIZE_NULL) && TREE_CODE (op) == ADDR_EXPR)
 422         {
 423           bool strict_overflow_p = false;
 424           /* tree_single_nonzero_warnv_p will not return true for non-weak
 425              non-automatic decls with -fno-delete-null-pointer-checks,
 426              which is disabled during -fsanitize=null.  We don't want to
 427              instrument those, just weak vars though.  */
 428           int save_flag_delete_null_pointer_checks
 429             = flag_delete_null_pointer_checks;
 430           flag_delete_null_pointer_checks = 1;
 431           if (!tree_single_nonzero_warnv_p (op, &strict_overflow_p)
 432               || strict_overflow_p)
 433             instrument = true;
 434           flag_delete_null_pointer_checks
 435             = save_flag_delete_null_pointer_checks;
 436         }
 437       else if (flag_sanitize & SANITIZE_NULL)
 438         instrument = true;
 439       if (mina && mina > 1)
 440         {
 441           if (!POINTER_TYPE_P (TREE_TYPE (op))
 442               || mina > get_pointer_alignment (op) / BITS_PER_UNIT)
 443             instrument = true;
 444         }
 445     }
 446   if (!instrument)
 447     return NULL_TREE;
 448   op = save_expr (orig_op);
 449   gcc_assert (POINTER_TYPE_P (ptype));
 450   if (TREE_CODE (ptype) == REFERENCE_TYPE)
 451     ptype = build_pointer_type (TREE_TYPE (ptype));
 452   tree kind = build_int_cst (ptype, ckind);
 453   tree align = build_int_cst (pointer_sized_int_node, mina);
 454   tree call
 455     = build_call_expr_internal_loc (loc, IFN_UBSAN_NULL, void_type_node,
 456                                     3, op, kind, align);
 457   TREE_SIDE_EFFECTS (call) = 1;
 458   return fold_build2 (COMPOUND_EXPR, TREE_TYPE (op), call, op);
 459 }
 460
 461 /* Instrument a NOP_EXPR to REFERENCE_TYPE if needed.  */
 462
 463 void
 464 ubsan_maybe_instrument_reference (tree stmt)
 465 {
 466   tree op = TREE_OPERAND (stmt, 0);
 467   op = ubsan_maybe_instrument_reference_or_call (EXPR_LOCATION (stmt), op,
 468                                                  TREE_TYPE (stmt),
 469                                                  UBSAN_REF_BINDING);
 470   if (op)
 471     TREE_OPERAND (stmt, 0) = op;
 472 }
 473
 474 /* Instrument a CALL_EXPR to a method if needed.  */
 475
 476 void
 477 ubsan_maybe_instrument_member_call (tree stmt, bool is_ctor)
 478 {
 479   if (call_expr_nargs (stmt) == 0)
 480     return;
 481   tree op = CALL_EXPR_ARG (stmt, 0);
 482   if (op == error_mark_node
 483       || !POINTER_TYPE_P (TREE_TYPE (op)))
 484     return;
 485   op = ubsan_maybe_instrument_reference_or_call (EXPR_LOCATION (stmt), op,
 486                                                  TREE_TYPE (op),
 487                                                  is_ctor ? UBSAN_CTOR_CALL
 488                                                  : UBSAN_MEMBER_CALL);
 489   if (op)
 490     CALL_EXPR_ARG (stmt, 0) = op;
 491 }