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_matching_member_expr(struct symbol *left_type, struct expression *right, struct symbol *left_member)
  84 {
  85         struct symbol *struct_type;
  86         int op = '.';
  87
  88         if (!left_member->ident)
  89                 return NULL;
  90
  91         struct_type = get_struct_type(right);
  92         if (!struct_type)
  93                 return NULL;
  94         if (struct_type != left_type)
  95                 return NULL;
  96
  97         if (right->type == EXPR_PREOP && right->op == '&')
  98                 right = strip_expr(right->unop);
  99
 100         if (is_pointer(right)) {
 101                 right = deref_expression(right);
 102                 op = '*';
 103         }
 104
 105         return member_expression(right, op, left_member->ident);
 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 __struct_members_copy(int mode, struct expression *faked,
 126                                   struct expression *left,
 127                                   struct expression *right)
 128 {
 129         struct symbol *struct_type, *tmp, *type;
 130         struct expression *left_member;
 131         struct expression *right_member = NULL;
 132         struct expression *assign;
 133         int op = '.';
 134
 135
 136         if (__in_fake_assign)
 137                 return;
 138         faked_expression = faked;
 139
 140         left = strip_expr(left);
 141         right = strip_expr(right);
 142
 143         struct_type = get_struct_type(left);
 144         if (!struct_type) {
 145                 /*
 146                  * This is not a struct assignment obviously.  But this is where
 147                  * memcpy() is handled so it feels like a good place to add this
 148                  * code.
 149                  */
 150
 151                 type = get_type(left);
 152                 if (!type || type->type != SYM_BASETYPE)
 153                         goto done;
 154
 155                 right = strip_expr(right);
 156                 if (right && right->type == EXPR_PREOP && right->op == '&')
 157                         right = remove_addr(right);
 158                 else
 159                         right = unknown_value_expression(left);
 160                 assign = assign_expression(left, right);
 161                 __in_fake_assign++;
 162                 __split_expr(assign);
 163                 __in_fake_assign--;
 164                 goto done;
 165         }
 166
 167         if (is_pointer(left)) {
 168                 left = deref_expression(left);
 169                 op = '*';
 170         }
 171
 172         FOR_EACH_PTR(struct_type->symbol_list, tmp) {
 173                 type = get_real_base_type(tmp);
 174                 if (type && type->type == SYM_ARRAY)
 175                         continue;
 176
 177                 left_member = member_expression(left, op, tmp->ident);
 178
 179                 switch (mode) {
 180                 case COPY_NORMAL:
 181                 case COPY_MEMCPY:
 182                         right_member = get_matching_member_expr(struct_type, right, tmp);
 183                         break;
 184                 case COPY_MEMSET:
 185                         right_member = right;
 186                         break;
 187                 }
 188                 if (!right_member)
 189                         right_member = unknown_value_expression(left_member);
 190                 assign = assign_expression(left_member, right_member);
 191                 __in_fake_assign++;
 192                 __split_expr(assign);
 193                 __in_fake_assign--;
 194         } END_FOR_EACH_PTR(tmp);
 195
 196 done:
 197         faked_expression = NULL;
 198 }
 199
 200 static int returns_zeroed_mem(struct expression *expr)
 201 {
 202         char *fn;
 203
 204         if (expr->type != EXPR_CALL || expr->fn->type != EXPR_SYMBOL)
 205                 return 0;
 206         fn = expr_to_var(expr->fn);
 207         if (!fn)
 208                 return 0;
 209         if (strcmp(fn, "kcalloc") == 0)
 210                 return 1;
 211         if (option_project == PROJ_KERNEL && strstr(fn, "zalloc"))
 212                 return 1;
 213         return 0;
 214 }
 215
 216 void __fake_struct_member_assignments(struct expression *expr)
 217 {
 218         struct symbol *struct_type;
 219
 220         if (is_zero(expr->right))
 221                 return;
 222
 223         struct_type = get_struct_type(expr->left);
 224         if (!struct_type)
 225                 return;
 226
 227         if (returns_zeroed_mem(expr->right))
 228                 __struct_members_copy(COPY_MEMSET, expr, expr->left, zero_expr());
 229         else
 230                 __struct_members_copy(COPY_NORMAL, expr, expr->left, expr->right);
 231 }
 232
 233 static void match_memset(const char *fn, struct expression *expr, void *_size_arg)
 234 {
 235         struct expression *buf;
 236         struct expression *val;
 237
 238         buf = get_argument_from_call_expr(expr->args, 0);
 239         val = get_argument_from_call_expr(expr->args, 1);
 240
 241         buf = strip_expr(buf);
 242         __struct_members_copy(COPY_MEMSET, expr, remove_addr(buf), val);
 243 }
 244
 245 static void match_memcpy(const char *fn, struct expression *expr, void *_arg)
 246 {
 247         struct expression *dest;
 248         struct expression *src;
 249
 250         dest = get_argument_from_call_expr(expr->args, 0);
 251         src = get_argument_from_call_expr(expr->args, 1);
 252
 253         __struct_members_copy(COPY_MEMCPY, expr, remove_addr(dest), remove_addr(src));
 254 }
 255
 256 static void match_memcpy_unknown(const char *fn, struct expression *expr, void *_arg)
 257 {
 258         struct expression *dest;
 259
 260         dest = get_argument_from_call_expr(expr->args, 0);
 261         __struct_members_copy(COPY_MEMCPY, expr, remove_addr(dest), NULL);
 262 }
 263
 264 static void register_clears_param(void)
 265 {
 266         struct token *token;
 267         char name[256];
 268         const char *function;
 269         int param;
 270
 271         if (option_project == PROJ_NONE)
 272                 return;
 273
 274         snprintf(name, 256, "%s.clears_argument", option_project_str);
 275
 276         token = get_tokens_file(name);
 277         if (!token)
 278                 return;
 279         if (token_type(token) != TOKEN_STREAMBEGIN)
 280                 return;
 281         token = token->next;
 282         while (token_type(token) != TOKEN_STREAMEND) {
 283                 if (token_type(token) != TOKEN_IDENT)
 284                         return;
 285                 function = show_ident(token->ident);
 286                 token = token->next;
 287                 if (token_type(token) != TOKEN_NUMBER)
 288                         return;
 289                 param = atoi(token->number);
 290                 add_function_hook(function, &match_memcpy_unknown, INT_PTR(param));
 291                 token = token->next;
 292         }
 293         clear_token_alloc();
 294 }
 295
 296 static void db_param_cleared(struct expression *expr, int param, char *key, char *value)
 297 {
 298         struct expression *arg;
 299
 300         while (expr->type == EXPR_ASSIGNMENT)
 301                 expr = strip_expr(expr->right);
 302         if (expr->type != EXPR_CALL)
 303                 return;
 304
 305         /*
 306          * FIXME:  __struct_members_copy() requires an expression but
 307          * get_variable_from_key() returns a name/sym pair so that doesn't
 308          * work here.
 309          */
 310         if (strcmp(key, "$") != 0)
 311                 return;
 312
 313         arg = get_argument_from_call_expr(expr->args, param);
 314         if (!arg)
 315                 return;
 316
 317         if (strcmp(value, "0") == 0)
 318                 __struct_members_copy(COPY_MEMSET, expr, remove_addr(arg), zero_expr());
 319         else
 320                 __struct_members_copy(COPY_MEMCPY, expr, remove_addr(arg), NULL);
 321 }
 322
 323 void register_struct_assignment(int id)
 324 {
 325         add_function_hook("memset", &match_memset, NULL);
 326
 327         add_function_hook("memcpy", &match_memcpy, INT_PTR(0));
 328         add_function_hook("memmove", &match_memcpy, INT_PTR(0));
 329
 330         register_clears_param();
 331         select_return_states_hook(PARAM_CLEARED, &db_param_cleared);
 332 }