smatch_struct_assignment.c

   1 /*
   2  * Copyright (C) 2014 Oracle.
   3  *
   4  * This program is free software; you can redistribute it and/or
   5  * modify it under the terms of the GNU General Public License
   6  * as published by the Free Software Foundation; either version 2
   7  * of the License, or (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, see http://www.gnu.org/copyleft/gpl.txt
  16  */
  17
  18 /*
  19  * This file started out by saying that if you have:
  20  *
  21  *      struct foo one, two;
  22  *      ...
  23  *      one = two;
  24  *
  25  * That's equivalent to saying:
  26  *
  27  *      one.x = two.x;
  28  *      one.y = two.y;
  29  *
  30  * Turning an assignment like that into a bunch of small fake assignments is
  31  * really useful.
  32  *
  33  * The call to memcpy(&one, &two, sizeof(foo)); is the same as "one = two;" so
  34  * we can re-use the code.  And we may as well use it for memset() too.
  35  * Assigning pointers is almost the same:
  36  *
  37  *      p1 = p2;
  38  *
  39  * Is the same as:
  40  *
  41  *      p1->x = p2->x;
  42  *      p1->y = p2->y;
  43  *
  44  * The problem is that you can go a bit crazy with pointers to pointers.
  45  *
  46  *      p1->x->y->z->one->two->three = p2->x->y->z->one->two->three;
  47  *
  48  * I don't have a proper solution for this problem right now.  I just copy one
  49  * level and don't nest.  It should handle limitted nesting but intelligently.
  50  *
  51  * The other thing is that you end up with a lot of garbage assignments where
  52  * we record "x could be anything. x->y could be anything. x->y->z->a->b->c
  53  * could *also* be anything!".  There should be a better way to filter this
  54  * useless information.
  55  *
  56  */
  57
  58 #include "scope.h"
  59 #include "smatch.h"
  60 #include "smatch_slist.h"
  61 #include "smatch_extra.h"
  62
  63 enum {
  64         COPY_NORMAL,
  65         COPY_MEMCPY,
  66         COPY_MEMSET,
  67 };
  68
  69 static struct symbol *get_struct_type(struct expression *expr)
  70 {
  71         struct symbol *type;
  72
  73         type = get_type(expr);
  74         if (!type)
  75                 return NULL;
  76         if (type->type == SYM_PTR)
  77                 type = get_real_base_type(type);
  78         if (type && type->type == SYM_STRUCT)
  79                 return type;
  80         return NULL;
  81 }
  82
  83 static struct expression *get_right_base_expr(struct symbol *left_type, struct expression *right)
  84 {
  85         struct symbol *struct_type;
  86
  87         if (!right)
  88                 return NULL;
  89
  90         struct_type = get_struct_type(right);
  91         if (!struct_type)
  92                 return NULL;
  93         if (struct_type != left_type)
  94                 return NULL;
  95
  96         if (right->type == EXPR_PREOP && right->op == '&')
  97                 right = strip_expr(right->unop);
  98
  99         if (right->type == EXPR_CALL)
 100                 return NULL;
 101
 102         if (is_pointer(right))
 103                 right = deref_expression(right);
 104
 105         return right;
 106 }
 107
 108 static struct expression *remove_addr(struct expression *expr)
 109 {
 110         expr = strip_expr(expr);
 111
 112         if (expr->type == EXPR_PREOP && expr->op == '&')
 113                 return strip_expr(expr->unop);
 114         return expr;
 115 }
 116
 117 static struct expression *faked_expression;
 118 struct expression *get_faked_expression(void)
 119 {
 120         if (!__in_fake_assign)
 121                 return NULL;
 122         return faked_expression;
 123 }
 124
 125 static void set_inner_struct_members(int mode, struct expression *faked, struct expression *left, struct expression *right, struct symbol *member)
 126 {
 127         struct expression *left_member, *right_member, *assign;
 128         struct symbol *base = get_real_base_type(member);
 129         struct symbol *tmp;
 130
 131         if (member->ident) {
 132                 left = member_expression(left, '.', member->ident);
 133                 if (mode != COPY_MEMSET && right)
 134                         right = member_expression(right, '.', member->ident);
 135         }
 136
 137         FOR_EACH_PTR(base->symbol_list, tmp) {
 138                 struct symbol *type;
 139
 140                 type = get_real_base_type(tmp);
 141                 if (!type)
 142                         continue;
 143
 144                 if (type->type == SYM_ARRAY)
 145                         continue;
 146                 if (type->type == SYM_UNION || type->type == SYM_STRUCT) {
 147                         set_inner_struct_members(mode, faked, left, right, tmp);
 148                         continue;
 149                 }
 150                 if (!tmp->ident)
 151                         continue;
 152
 153                 left_member = member_expression(left, '.', tmp->ident);
 154
 155                 switch (mode) {
 156                 case COPY_NORMAL:
 157                 case COPY_MEMCPY:
 158                         if (right)
 159                                 right_member = member_expression(right, '.', tmp->ident);
 160                         else
 161                                 right_member = unknown_value_expression(left_member);
 162                         break;
 163                 case COPY_MEMSET:
 164                         right_member = right;
 165                         break;
 166                 }
 167
 168                 assign = assign_expression(left_member, right_member);
 169                 __in_fake_assign++;
 170                 __split_expr(assign);
 171                 __in_fake_assign--;
 172         } END_FOR_EACH_PTR(tmp);
 173
 174
 175 }
 176
 177 static void __struct_members_copy(int mode, struct expression *faked,
 178                                   struct expression *left,
 179                                   struct expression *right)
 180 {
 181         struct symbol *struct_type, *tmp, *type;
 182         struct expression *left_member;
 183         struct expression *right_member;
 184         struct expression *assign;
 185         int op = '.';
 186
 187
 188         if (__in_fake_assign)
 189                 return;
 190         faked_expression = faked;
 191
 192         left = strip_expr(left);
 193         right = strip_expr(right);
 194
 195         struct_type = get_struct_type(left);
 196         if (!struct_type) {
 197                 /*
 198                  * This is not a struct assignment obviously.  But this is where
 199                  * memcpy() is handled so it feels like a good place to add this
 200                  * code.
 201                  */
 202
 203                 type = get_type(left);
 204                 if (!type || type->type != SYM_BASETYPE)
 205                         goto done;
 206
 207                 right = strip_expr(right);
 208                 if (right && right->type == EXPR_PREOP && right->op == '&')
 209                         right = remove_addr(right);
 210                 else
 211                         right = unknown_value_expression(left);
 212                 assign = assign_expression(left, right);
 213                 __in_fake_assign++;
 214                 __split_expr(assign);
 215                 __in_fake_assign--;
 216                 goto done;
 217         }
 218
 219         if (is_pointer(left)) {
 220                 left = deref_expression(left);
 221                 op = '*';
 222         }
 223         if (mode != COPY_MEMSET)
 224                 right = get_right_base_expr(struct_type, right);
 225
 226         FOR_EACH_PTR(struct_type->symbol_list, tmp) {
 227                 type = get_real_base_type(tmp);
 228                 if (!type)
 229                         continue;
 230                 if (type->type == SYM_ARRAY)
 231                         continue;
 232
 233                 if (type->type == SYM_UNION || type->type == SYM_STRUCT) {
 234                         set_inner_struct_members(mode, faked, left, right, tmp);
 235                         continue;
 236                 }
 237
 238                 if (!tmp->ident)
 239                         continue;
 240
 241                 left_member = member_expression(left, op, tmp->ident);
 242                 right_member = NULL;
 243
 244                 switch (mode) {
 245                 case COPY_NORMAL:
 246                 case COPY_MEMCPY:
 247                         if (right)
 248                                 right_member = member_expression(right, op, tmp->ident);
 249                         else
 250                                 right_member = unknown_value_expression(left_member);
 251                         break;
 252                 case COPY_MEMSET:
 253                         right_member = right;
 254                         break;
 255                 }
 256                 if (!right_member) {
 257                         sm_msg("internal.  No right member");
 258                         continue;
 259                 }
 260                 assign = assign_expression(left_member, right_member);
 261                 __in_fake_assign++;
 262                 __split_expr(assign);
 263                 __in_fake_assign--;
 264         } END_FOR_EACH_PTR(tmp);
 265
 266 done:
 267         faked_expression = NULL;
 268 }
 269
 270 static int returns_zeroed_mem(struct expression *expr)
 271 {
 272         char *fn;
 273
 274         if (expr->type != EXPR_CALL || expr->fn->type != EXPR_SYMBOL)
 275                 return 0;
 276         fn = expr_to_var(expr->fn);
 277         if (!fn)
 278                 return 0;
 279         if (strcmp(fn, "kcalloc") == 0)
 280                 return 1;
 281         if (option_project == PROJ_KERNEL && strstr(fn, "zalloc"))
 282                 return 1;
 283         return 0;
 284 }
 285
 286 void __fake_struct_member_assignments(struct expression *expr)
 287 {
 288         struct symbol *struct_type;
 289
 290         if (is_zero(expr->right))
 291                 return;
 292
 293         struct_type = get_struct_type(expr->left);
 294         if (!struct_type)
 295                 return;
 296
 297         if (returns_zeroed_mem(expr->right))
 298                 __struct_members_copy(COPY_MEMSET, expr, expr->left, zero_expr());
 299         else
 300                 __struct_members_copy(COPY_NORMAL, expr, expr->left, expr->right);
 301 }
 302
 303 static void match_memset(const char *fn, struct expression *expr, void *_size_arg)
 304 {
 305         struct expression *buf;
 306         struct expression *val;
 307
 308         buf = get_argument_from_call_expr(expr->args, 0);
 309         val = get_argument_from_call_expr(expr->args, 1);
 310
 311         buf = strip_expr(buf);
 312         __struct_members_copy(COPY_MEMSET, expr, remove_addr(buf), val);
 313 }
 314
 315 static void match_memcpy(const char *fn, struct expression *expr, void *_arg)
 316 {
 317         struct expression *dest;
 318         struct expression *src;
 319
 320         dest = get_argument_from_call_expr(expr->args, 0);
 321         src = get_argument_from_call_expr(expr->args, 1);
 322
 323         __struct_members_copy(COPY_MEMCPY, expr, remove_addr(dest), remove_addr(src));
 324 }
 325
 326 static void match_memcpy_unknown(const char *fn, struct expression *expr, void *_arg)
 327 {
 328         struct expression *dest;
 329
 330         dest = get_argument_from_call_expr(expr->args, 0);
 331         __struct_members_copy(COPY_MEMCPY, expr, remove_addr(dest), NULL);
 332 }
 333
 334 static void register_clears_param(void)
 335 {
 336         struct token *token;
 337         char name[256];
 338         const char *function;
 339         int param;
 340
 341         if (option_project == PROJ_NONE)
 342                 return;
 343
 344         snprintf(name, 256, "%s.clears_argument", option_project_str);
 345
 346         token = get_tokens_file(name);
 347         if (!token)
 348                 return;
 349         if (token_type(token) != TOKEN_STREAMBEGIN)
 350                 return;
 351         token = token->next;
 352         while (token_type(token) != TOKEN_STREAMEND) {
 353                 if (token_type(token) != TOKEN_IDENT)
 354                         return;
 355                 function = show_ident(token->ident);
 356                 token = token->next;
 357                 if (token_type(token) != TOKEN_NUMBER)
 358                         return;
 359                 param = atoi(token->number);
 360                 add_function_hook(function, &match_memcpy_unknown, INT_PTR(param));
 361                 token = token->next;
 362         }
 363         clear_token_alloc();
 364 }
 365
 366 static void db_param_cleared(struct expression *expr, int param, char *key, char *value)
 367 {
 368         struct expression *arg;
 369
 370         while (expr->type == EXPR_ASSIGNMENT)
 371                 expr = strip_expr(expr->right);
 372         if (expr->type != EXPR_CALL)
 373                 return;
 374
 375         /*
 376          * FIXME:  __struct_members_copy() requires an expression but
 377          * get_variable_from_key() returns a name/sym pair so that doesn't
 378          * work here.
 379          */
 380         if (strcmp(key, "$") != 0)
 381                 return;
 382
 383         arg = get_argument_from_call_expr(expr->args, param);
 384         if (!arg)
 385                 return;
 386
 387         if (strcmp(value, "0") == 0)
 388                 __struct_members_copy(COPY_MEMSET, expr, remove_addr(arg), zero_expr());
 389         else
 390                 __struct_members_copy(COPY_MEMCPY, expr, remove_addr(arg), NULL);
 391 }
 392
 393 void register_struct_assignment(int id)
 394 {
 395         add_function_hook("memset", &match_memset, NULL);
 396         add_function_hook("__memset", &match_memset, NULL);
 397
 398         add_function_hook("memcpy", &match_memcpy, INT_PTR(0));
 399         add_function_hook("memmove", &match_memcpy, INT_PTR(0));
 400         add_function_hook("__memcpy", &match_memcpy, INT_PTR(0));
 401         add_function_hook("__memmove", &match_memcpy, INT_PTR(0));
 402
 403         register_clears_param();
 404         select_return_states_hook(PARAM_CLEARED, &db_param_cleared);
 405 }