smatch_points_to_host_data.c

   1 /*
   2  * Copyright (C) 2020 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 /* The below code is an adjustment of the smatch_points_to_user_data.c to
  19  * track (untrusted and potentially malicious) data pointers received by a
  20  * guest kernel via various low-level port IO, MMIO, MSR, CPUID and PCI
  21  * config space access functions (see func_table for a full list) from an
  22  * untrusted host/VMM under confidential computing threat model
  23  * (where a guest VM does not trust the host/VMM).
  24  * We call this data as 'host' data here. Being able to track host data
  25  * allows creating new smatch patterns that can perform various checks
  26  * on such data, i.e. verify that that there are no spectre v1 gadgets
  27  * on this attack surface or even simply report locations where host data
  28  * is being processed in the kernel source code (useful for a targeted code
  29  * audit).
  30  * Similar as smatch_points_to_user_data.c it works with
  31  * smatch_kernel_host_data.c code.
  32  *
  33  *  Copyright (C) 2022 Elena Reshetova Intel Corporation.
  34  */
  35
  36 #include "smatch.h"
  37 #include "smatch_slist.h"
  38 #include "smatch_extra.h"
  39
  40 static int my_id;
  41 STATE(host_data);
  42 STATE(host_data_set);
  43
  44 static const char *returns_pointer_to_host_data[] = {
  45         "virtqueue_get_buf"
  46 };
  47
  48 struct host_fn_info {
  49         const char *name;
  50         int type;
  51         int param;
  52         const char *key;
  53         const sval_t *implies_start, *implies_end;
  54         func_hook *call_back;
  55 };
  56
  57 struct host_fn_info func_pointer_table[] = {
  58             /* Functions that return host data as argument 1 */
  59         { "memcpy_fromio", HOST_DATA, 0, "*$" },
  60         { "acpi_os_read_iomem", HOST_DATA, 1, "*$" },
  61         { "mmio_insb", HOST_DATA, 1, "*$" },
  62         { "mmio_insw", HOST_DATA, 1, "*$" },
  63         { "mmio_insl", HOST_DATA, 1, "*$" },
  64         { "acpi_read_bit_register", HOST_DATA, 1, "*$" },
  65           /* Functions that return host data as argument 3 */
  66         { "__virtio_cread_many", HOST_DATA, 2, "*$" },
  67         { "pci_user_read_config_word", HOST_DATA, 2, "*$" },
  68         { "pci_user_read_config_dword", HOST_DATA, 2, "*$" },
  69         { "pci_user_read_config_byte", HOST_DATA, 2, "*$" },
  70           /* Functions that return host data as argument 4 */
  71         { "pci_bus_read_config_byte", HOST_DATA, 3, "*$" },
  72         { "pci_bus_read_config_word", HOST_DATA, 3, "*$" },
  73         { "pci_bus_read_config_dword", HOST_DATA, 3, "*$" },
  74         /* Functions that return host data as argument 6 */
  75         { "raw_pci_read", HOST_DATA, 5, "*$" },
  76         /* Functions that return host data as arguments 2-5 */
  77         { "cpuid", HOST_DATA, 1, "*$" },
  78         { "cpuid", HOST_DATA, 2, "*$" },
  79         { "cpuid", HOST_DATA, 3, "*$" },
  80         { "cpuid", HOST_DATA, 4, "*$" },
  81         /* Functions that return host data as arguments 3-6 */
  82         { "cpuid_count", HOST_DATA, 2, "*$" },
  83         { "cpuid_count", HOST_DATA, 3, "*$" },
  84         { "cpuid_count", HOST_DATA, 4, "*$" },
  85         { "cpuid_count", HOST_DATA, 5, "*$" },
  86 };
  87
  88 bool is_host_data_fn(struct symbol *fn)
  89 {
  90         int i;
  91
  92         if (!fn || !fn->ident)
  93                 return false;
  94
  95         for (i = 0; i < ARRAY_SIZE(returns_pointer_to_host_data); i++) {
  96                 if (strcmp(fn->ident->name, returns_pointer_to_host_data[i]) == 0) {
  97 //                      func_gets_host_data = true;
  98                         return true;
  99                 }
 100         }
 101         return false;
 102 }
 103
 104 bool is_fn_points_to_host_data(const char *fn)
 105 {
 106         int i;
 107
 108         if (!fn)
 109                 return false;
 110
 111         for (i = 0; i < ARRAY_SIZE(returns_pointer_to_host_data); i++) {
 112                 if (strcmp(fn, returns_pointer_to_host_data[i]) == 0) {
 113                         return true;
 114                 }
 115         }
 116         return false;
 117 }
 118
 119 static bool is_points_to_host_data_fn(struct expression *expr)
 120 {
 121         expr = strip_expr(expr);
 122         if (!expr || expr->type != EXPR_CALL || expr->fn->type != EXPR_SYMBOL ||
 123             !expr->fn->symbol)
 124                 return false;
 125         return is_host_data_fn(expr->fn->symbol);
 126 }
 127
 128 static bool is_array_of_host_data(struct expression *expr)
 129 {
 130         struct expression *deref;
 131         struct symbol *type;
 132
 133         if (expr->type == EXPR_PREOP && expr->op == '&') {
 134                 expr = strip_expr(expr->unop);
 135                 if (expr->type == EXPR_PREOP && expr->op == '*')
 136                         expr = strip_expr(expr->unop);
 137         }
 138
 139         /* This is for array elements &foo->data[4] */
 140         if (expr->type == EXPR_BINOP && expr->op == '+') {
 141                 if (points_to_host_data(expr->left))
 142                         return true;
 143                 if (points_to_host_data(expr->right))
 144                         return true;
 145         }
 146
 147         /* This is for if you have: foo = skb->data; frob(foo->array); */
 148         type = get_type(expr);
 149         if (!type || type->type != SYM_ARRAY)
 150                 return false;
 151
 152         if (expr->type != EXPR_DEREF)
 153                 return false;
 154         deref = strip_expr(expr->deref);
 155         if (deref->type != EXPR_PREOP || deref->op != '*')
 156                 return false;
 157         deref = strip_expr(deref->unop);
 158         return points_to_host_data(deref);
 159 }
 160
 161 bool points_to_host_data(struct expression *expr)
 162 {
 163         struct sm_state *sm;
 164
 165         if (!expr)
 166                 return false;
 167
 168         expr = strip_expr(expr);
 169         if (!expr)
 170                 return false;
 171
 172         if (is_fake_call(expr))
 173                 return false;
 174
 175         if (expr->type == EXPR_ASSIGNMENT)
 176                 return points_to_host_data(expr->left);
 177
 178         if (is_array_of_host_data(expr))
 179                 return true;
 180
 181         if (expr->type == EXPR_BINOP && expr->op == '+')
 182                 expr = strip_expr(expr->left);
 183
 184         if (is_points_to_host_data_fn(expr))
 185                 return true;
 186
 187         sm = get_sm_state_expr(my_id, expr);
 188         if (!sm)
 189                 return false;
 190         if (slist_has_state(sm->possible, &host_data) ||
 191             slist_has_state(sm->possible, &host_data_set))
 192                 return true;
 193         return false;
 194 }
 195
 196 void set_points_to_host_data(struct expression *expr, bool is_new)
 197 {
 198         struct expression *tmp;
 199
 200         tmp = get_assigned_expr(expr);
 201         if (tmp)
 202                 set_state_expr(my_id, tmp, is_new ? &host_data_set : &host_data);
 203         set_state_expr(my_id, expr, is_new ? &host_data_set : &host_data);
 204 }
 205
 206 static void match_assign_host(struct expression *expr)
 207 {
 208
 209         if (is_fake_call(expr->right))
 210                 return;
 211
 212         if (!is_ptr_type(get_type(expr->left))){
 213                 return;
 214         }
 215
 216         if (points_to_host_data(expr->right)) {
 217                 set_points_to_host_data(expr->left, false);
 218                 return;
 219         }
 220
 221         if (get_state_expr(my_id, expr->left)){
 222                 set_state_expr(my_id, expr->left, &undefined);
 223         }
 224 }
 225
 226 static void match_memcpy_host(const char *fn, struct expression *expr, void *_unused)
 227 {
 228         struct expression *dest, *src;
 229
 230         dest = get_argument_from_call_expr(expr->args, 0);
 231         src = get_argument_from_call_expr(expr->args, 1);
 232
 233         if (points_to_host_data(src)) {
 234                 set_points_to_host_data(dest, false);
 235                 return;
 236         }
 237
 238         if (get_state_expr(my_id, dest))
 239                 set_state_expr(my_id, dest, &undefined);
 240 }
 241
 242 static void match_host_function(const char *fn, struct expression *expr, void *_unused)
 243 {
 244         struct expression *dest;
 245
 246         for (int i = 0; i < ARRAY_SIZE(func_pointer_table); i++)
 247                 if (sym_name_is(func_pointer_table[i].name, expr->fn)){
 248                         dest = get_argument_from_call_expr(expr->args, func_pointer_table[i].param);
 249                         dest = strip_expr(dest);
 250                         if (!dest)
 251                                 return;
 252                         set_state_expr(my_id, dest, &host_data);
 253                 }
 254 }
 255
 256 static void return_info_callback_host(int return_id, char *return_ranges,
 257                                  struct expression *returned_expr,
 258                                  int param,
 259                                  const char *printed_name,
 260                                  struct sm_state *sm)
 261 {
 262         int type;
 263
 264         if (strncmp(printed_name, "&$", 2) == 0)
 265                 return;
 266
 267         if (param >= 0) {
 268                 if (!slist_has_state(sm->possible, &host_data_set))
 269                         return;
 270                 type = HOST_PTR_SET;
 271         } else {
 272                 if (slist_has_state(sm->possible, &host_data_set))
 273                         type = HOST_PTR_SET;
 274                 else if (slist_has_state(sm->possible, &host_data))
 275                         type = HOST_PTR;
 276                 else
 277                         return;
 278         }
 279         if (parent_is_gone_var_sym(sm->name, sm->sym))
 280                 return;
 281
 282         sql_insert_return_states(return_id, return_ranges, type,
 283                                  param, printed_name, "");
 284 }
 285
 286 static void returns_host_ptr_helper(struct expression *expr, int param, char *key, char *value, bool set)
 287 {
 288         struct expression *arg;
 289         struct expression *call;
 290         char *name;
 291         struct symbol *sym;
 292
 293         call = expr;
 294         while (call->type == EXPR_ASSIGNMENT)
 295                 call = strip_expr(call->right);
 296         if (call->type != EXPR_CALL)
 297                 return;
 298
 299         if (!set && !we_pass_host_data(call))
 300                 return;
 301
 302         if (param == -1) {
 303                 if (expr->type != EXPR_ASSIGNMENT) {
 304                         /* Nothing to do.  Fake assignments should handle it */
 305                         return;
 306                 }
 307                 arg = expr->left;
 308                 goto set_host;
 309         }
 310
 311         arg = get_argument_from_call_expr(call->args, param);
 312         if (!arg)
 313                 return;
 314 set_host:
 315         name = get_variable_from_key(arg, key, &sym);
 316         if (!name || !sym)
 317                 goto free;
 318         if (set)
 319                 set_state(my_id, name, sym, &host_data_set);
 320         else
 321                 set_state(my_id, name, sym, &host_data);
 322 free:
 323         free_string(name);
 324 }
 325
 326 static void returns_host_ptr(struct expression *expr, int param, char *key, char *value)
 327 {
 328         returns_host_ptr_helper(expr, param, key, value, false);
 329 }
 330
 331 static void returns_host_ptr_set(struct expression *expr, int param, char *key, char *value)
 332 {
 333         returns_host_ptr_helper(expr, param, key, value, true);
 334 }
 335
 336 static void set_param_host_ptr(const char *name, struct symbol *sym, char *key, char *value)
 337 {
 338         struct expression *expr;
 339         char *fullname;
 340
 341         expr = symbol_expression(sym);
 342         fullname = get_variable_from_key(expr, key, NULL);
 343         if (!fullname)
 344                 return;
 345         set_state(my_id, fullname, sym, &host_data);
 346 }
 347
 348 static void caller_info_callback_host(struct expression *call, int param, char *printed_name, struct sm_state *sm)
 349 {
 350         if (!slist_has_state(sm->possible, &host_data) &&
 351             !slist_has_state(sm->possible, &host_data_set))
 352                 return;
 353         sql_insert_caller_info(call, HOST_PTR, param, printed_name, "");
 354 }
 355
 356 void register_points_to_host_data(int id)
 357 {
 358         my_id = id;
 359
 360         if (option_project != PROJ_KERNEL)
 361                 return;
 362
 363         add_hook(&match_assign_host, ASSIGNMENT_HOOK);
 364
 365         for (int i = 0; i < ARRAY_SIZE(func_pointer_table); i++)
 366                 add_function_hook(func_pointer_table[i].name, &match_host_function, NULL);
 367
 368         add_function_hook("memcpy", &match_memcpy_host, NULL);
 369         add_function_hook("__memcpy", &match_memcpy_host, NULL);
 370
 371         add_caller_info_callback(my_id, caller_info_callback_host);
 372         add_return_info_callback(my_id, return_info_callback_host);
 373
 374         select_caller_info_hook(set_param_host_ptr, HOST_PTR);
 375         select_return_states_hook(HOST_PTR, &returns_host_ptr);
 376         select_return_states_hook(HOST_PTR_SET, &returns_host_ptr_set);
 377 }