Blame view
fs/exec.c
39.1 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 |
/* * linux/fs/exec.c * * Copyright (C) 1991, 1992 Linus Torvalds */ /* * #!-checking implemented by tytso. */ /* * Demand-loading implemented 01.12.91 - no need to read anything but * the header into memory. The inode of the executable is put into * "current->executable", and page faults do the actual loading. Clean. * * Once more I can proudly say that linux stood up to being changed: it * was less than 2 hours work to get demand-loading completely implemented. * * Demand loading changed July 1993 by Eric Youngdale. Use mmap instead, * current->executable is only used by the procfs. This allows a dispatch * table to check for several different types of binary formats. We keep * trying until we recognize the file or we run out of supported binary * formats. */ |
1da177e4c
|
24 25 |
#include <linux/slab.h> #include <linux/file.h> |
9f3acc314
|
26 |
#include <linux/fdtable.h> |
ba92a43db
|
27 |
#include <linux/mm.h> |
1da177e4c
|
28 29 |
#include <linux/stat.h> #include <linux/fcntl.h> |
ba92a43db
|
30 |
#include <linux/swap.h> |
74aadce98
|
31 |
#include <linux/string.h> |
1da177e4c
|
32 |
#include <linux/init.h> |
ca5b172bd
|
33 |
#include <linux/pagemap.h> |
cdd6c482c
|
34 |
#include <linux/perf_event.h> |
1da177e4c
|
35 36 37 38 39 |
#include <linux/highmem.h> #include <linux/spinlock.h> #include <linux/key.h> #include <linux/personality.h> #include <linux/binfmts.h> |
1da177e4c
|
40 |
#include <linux/utsname.h> |
84d737866
|
41 |
#include <linux/pid_namespace.h> |
1da177e4c
|
42 43 |
#include <linux/module.h> #include <linux/namei.h> |
1da177e4c
|
44 45 46 |
#include <linux/mount.h> #include <linux/security.h> #include <linux/syscalls.h> |
8f0ab5147
|
47 |
#include <linux/tsacct_kern.h> |
9f46080c4
|
48 |
#include <linux/cn_proc.h> |
473ae30bc
|
49 |
#include <linux/audit.h> |
6341c393f
|
50 |
#include <linux/tracehook.h> |
5f4123be3
|
51 |
#include <linux/kmod.h> |
6110e3abb
|
52 |
#include <linux/fsnotify.h> |
5ad4e53bd
|
53 |
#include <linux/fs_struct.h> |
61be228a0
|
54 |
#include <linux/pipe_fs_i.h> |
3d5992d2a
|
55 |
#include <linux/oom.h> |
0e028465d
|
56 |
#include <linux/compat.h> |
1da177e4c
|
57 58 59 |
#include <asm/uaccess.h> #include <asm/mmu_context.h> |
b6a2fea39
|
60 |
#include <asm/tlb.h> |
43d2b1132
|
61 62 |
#include <trace/events/task.h> |
a6f76f23d
|
63 |
#include "internal.h" |
179899fd5
|
64 |
#include "coredump.h" |
1da177e4c
|
65 |
|
4ff16c25e
|
66 |
#include <trace/events/sched.h> |
d6e711448
|
67 |
int suid_dumpable = 0; |
e4dc1b14d
|
68 |
static LIST_HEAD(formats); |
1da177e4c
|
69 |
static DEFINE_RWLOCK(binfmt_lock); |
8fc3dc5a3
|
70 |
void __register_binfmt(struct linux_binfmt * fmt, int insert) |
1da177e4c
|
71 |
{ |
8fc3dc5a3
|
72 |
BUG_ON(!fmt); |
1da177e4c
|
73 |
write_lock(&binfmt_lock); |
74641f584
|
74 75 |
insert ? list_add(&fmt->lh, &formats) : list_add_tail(&fmt->lh, &formats); |
1da177e4c
|
76 |
write_unlock(&binfmt_lock); |
1da177e4c
|
77 |
} |
74641f584
|
78 |
EXPORT_SYMBOL(__register_binfmt); |
1da177e4c
|
79 |
|
f6b450d48
|
80 |
void unregister_binfmt(struct linux_binfmt * fmt) |
1da177e4c
|
81 |
{ |
1da177e4c
|
82 |
write_lock(&binfmt_lock); |
e4dc1b14d
|
83 |
list_del(&fmt->lh); |
1da177e4c
|
84 |
write_unlock(&binfmt_lock); |
1da177e4c
|
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 |
} EXPORT_SYMBOL(unregister_binfmt); static inline void put_binfmt(struct linux_binfmt * fmt) { module_put(fmt->module); } /* * Note that a shared library must be both readable and executable due to * security reasons. * * Also note that we take the address to load from from the file itself. */ |
1e7bfb213
|
100 |
SYSCALL_DEFINE1(uselib, const char __user *, library) |
1da177e4c
|
101 |
{ |
964bd1836
|
102 |
struct file *file; |
91a27b2a7
|
103 |
struct filename *tmp = getname(library); |
964bd1836
|
104 |
int error = PTR_ERR(tmp); |
47c805dc2
|
105 106 107 108 109 |
static const struct open_flags uselib_flags = { .open_flag = O_LARGEFILE | O_RDONLY | __FMODE_EXEC, .acc_mode = MAY_READ | MAY_EXEC | MAY_OPEN, .intent = LOOKUP_OPEN }; |
964bd1836
|
110 |
|
6e8341a11
|
111 112 |
if (IS_ERR(tmp)) goto out; |
669abf4e5
|
113 |
file = do_filp_open(AT_FDCWD, tmp, &uselib_flags, LOOKUP_FOLLOW); |
6e8341a11
|
114 115 116 |
putname(tmp); error = PTR_ERR(file); if (IS_ERR(file)) |
1da177e4c
|
117 118 119 |
goto out; error = -EINVAL; |
6e8341a11
|
120 |
if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) |
1da177e4c
|
121 |
goto exit; |
30524472c
|
122 |
error = -EACCES; |
6e8341a11
|
123 |
if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) |
6146f0d5e
|
124 |
goto exit; |
1da177e4c
|
125 |
|
2a12a9d78
|
126 |
fsnotify_open(file); |
6110e3abb
|
127 |
|
1da177e4c
|
128 129 130 131 132 |
error = -ENOEXEC; if(file->f_op) { struct linux_binfmt * fmt; read_lock(&binfmt_lock); |
e4dc1b14d
|
133 |
list_for_each_entry(fmt, &formats, lh) { |
1da177e4c
|
134 135 136 137 138 139 140 141 142 143 144 145 146 |
if (!fmt->load_shlib) continue; if (!try_module_get(fmt->module)) continue; read_unlock(&binfmt_lock); error = fmt->load_shlib(file); read_lock(&binfmt_lock); put_binfmt(fmt); if (error != -ENOEXEC) break; } read_unlock(&binfmt_lock); } |
6e8341a11
|
147 |
exit: |
1da177e4c
|
148 149 150 |
fput(file); out: return error; |
1da177e4c
|
151 |
} |
b6a2fea39
|
152 |
#ifdef CONFIG_MMU |
ae6b585ee
|
153 154 155 156 157 158 |
/* * The nascent bprm->mm is not visible until exec_mmap() but it can * use a lot of memory, account these pages in current->mm temporary * for oom_badness()->get_mm_rss(). Once exec succeeds or fails, we * change the counter back via acct_arg_size(0). */ |
0e028465d
|
159 |
static void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) |
3c77f8457
|
160 161 162 163 164 165 166 167 |
{ struct mm_struct *mm = current->mm; long diff = (long)(pages - bprm->vma_pages); if (!mm || !diff) return; bprm->vma_pages = pages; |
3c77f8457
|
168 |
add_mm_counter(mm, MM_ANONPAGES, diff); |
3c77f8457
|
169 |
} |
0e028465d
|
170 |
static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, |
b6a2fea39
|
171 172 173 174 175 176 177 |
int write) { struct page *page; int ret; #ifdef CONFIG_STACK_GROWSUP if (write) { |
d05f3169c
|
178 |
ret = expand_downwards(bprm->vma, pos); |
b6a2fea39
|
179 180 181 182 183 184 185 186 187 188 |
if (ret < 0) return NULL; } #endif ret = get_user_pages(current, bprm->mm, pos, 1, write, 1, &page, NULL); if (ret <= 0) return NULL; if (write) { |
b6a2fea39
|
189 |
unsigned long size = bprm->vma->vm_end - bprm->vma->vm_start; |
a64e715fc
|
190 |
struct rlimit *rlim; |
3c77f8457
|
191 |
acct_arg_size(bprm, size / PAGE_SIZE); |
a64e715fc
|
192 193 194 195 196 197 |
/* * We've historically supported up to 32 pages (ARG_MAX) * of argument strings even with small stacks */ if (size <= ARG_MAX) return page; |
b6a2fea39
|
198 199 200 201 202 203 204 205 |
/* * Limit to 1/4-th the stack size for the argv+env strings. * This ensures that: * - the remaining binfmt code will not run out of stack space, * - the program will have a reasonable amount of stack left * to work from. */ |
a64e715fc
|
206 |
rlim = current->signal->rlim; |
d554ed895
|
207 |
if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) { |
b6a2fea39
|
208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 |
put_page(page); return NULL; } } return page; } static void put_arg_page(struct page *page) { put_page(page); } static void free_arg_page(struct linux_binprm *bprm, int i) { } static void free_arg_pages(struct linux_binprm *bprm) { } static void flush_arg_page(struct linux_binprm *bprm, unsigned long pos, struct page *page) { flush_cache_page(bprm->vma, pos, page_to_pfn(page)); } static int __bprm_mm_init(struct linux_binprm *bprm) { |
eaccbfa56
|
237 |
int err; |
b6a2fea39
|
238 239 240 241 242 |
struct vm_area_struct *vma = NULL; struct mm_struct *mm = bprm->mm; bprm->vma = vma = kmem_cache_zalloc(vm_area_cachep, GFP_KERNEL); if (!vma) |
eaccbfa56
|
243 |
return -ENOMEM; |
b6a2fea39
|
244 245 246 247 248 249 250 251 252 253 |
down_write(&mm->mmap_sem); vma->vm_mm = mm; /* * Place the stack at the largest stack address the architecture * supports. Later, we'll move this to an appropriate place. We don't * use STACK_TOP because that can depend on attributes which aren't * configured yet. */ |
aacb3d17a
|
254 |
BUILD_BUG_ON(VM_STACK_FLAGS & VM_STACK_INCOMPLETE_SETUP); |
b6a2fea39
|
255 256 |
vma->vm_end = STACK_TOP_MAX; vma->vm_start = vma->vm_end - PAGE_SIZE; |
a8bef8ff6
|
257 |
vma->vm_flags = VM_STACK_FLAGS | VM_STACK_INCOMPLETE_SETUP; |
3ed75eb8f
|
258 |
vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); |
5beb49305
|
259 |
INIT_LIST_HEAD(&vma->anon_vma_chain); |
462e635e5
|
260 |
|
b6a2fea39
|
261 |
err = insert_vm_struct(mm, vma); |
eaccbfa56
|
262 |
if (err) |
b6a2fea39
|
263 |
goto err; |
b6a2fea39
|
264 265 266 |
mm->stack_vm = mm->total_vm = 1; up_write(&mm->mmap_sem); |
b6a2fea39
|
267 |
bprm->p = vma->vm_end - sizeof(void *); |
b6a2fea39
|
268 |
return 0; |
b6a2fea39
|
269 |
err: |
eaccbfa56
|
270 271 272 |
up_write(&mm->mmap_sem); bprm->vma = NULL; kmem_cache_free(vm_area_cachep, vma); |
b6a2fea39
|
273 274 275 276 277 278 279 280 281 |
return err; } static bool valid_arg_len(struct linux_binprm *bprm, long len) { return len <= MAX_ARG_STRLEN; } #else |
0e028465d
|
282 |
static inline void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) |
3c77f8457
|
283 284 |
{ } |
0e028465d
|
285 |
static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, |
b6a2fea39
|
286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 |
int write) { struct page *page; page = bprm->page[pos / PAGE_SIZE]; if (!page && write) { page = alloc_page(GFP_HIGHUSER|__GFP_ZERO); if (!page) return NULL; bprm->page[pos / PAGE_SIZE] = page; } return page; } static void put_arg_page(struct page *page) { } static void free_arg_page(struct linux_binprm *bprm, int i) { if (bprm->page[i]) { __free_page(bprm->page[i]); bprm->page[i] = NULL; } } static void free_arg_pages(struct linux_binprm *bprm) { int i; for (i = 0; i < MAX_ARG_PAGES; i++) free_arg_page(bprm, i); } static void flush_arg_page(struct linux_binprm *bprm, unsigned long pos, struct page *page) { } static int __bprm_mm_init(struct linux_binprm *bprm) { bprm->p = PAGE_SIZE * MAX_ARG_PAGES - sizeof(void *); return 0; } static bool valid_arg_len(struct linux_binprm *bprm, long len) { return len <= bprm->p; } #endif /* CONFIG_MMU */ /* * Create a new mm_struct and populate it with a temporary stack * vm_area_struct. We don't have enough context at this point to set the stack * flags, permissions, and offset, so we use temporary values. We'll update * them later in setup_arg_pages(). */ int bprm_mm_init(struct linux_binprm *bprm) { int err; struct mm_struct *mm = NULL; bprm->mm = mm = mm_alloc(); err = -ENOMEM; if (!mm) goto err; err = init_new_context(current, mm); if (err) goto err; err = __bprm_mm_init(bprm); if (err) goto err; return 0; err: if (mm) { bprm->mm = NULL; mmdrop(mm); } return err; } |
ba2d01629
|
373 |
struct user_arg_ptr { |
0e028465d
|
374 375 376 377 378 379 |
#ifdef CONFIG_COMPAT bool is_compat; #endif union { const char __user *const __user *native; #ifdef CONFIG_COMPAT |
38b983b34
|
380 |
const compat_uptr_t __user *compat; |
0e028465d
|
381 382 |
#endif } ptr; |
ba2d01629
|
383 384 385 |
}; static const char __user *get_user_arg_ptr(struct user_arg_ptr argv, int nr) |
1d1dbf813
|
386 |
{ |
0e028465d
|
387 388 389 390 391 392 393 394 |
const char __user *native; #ifdef CONFIG_COMPAT if (unlikely(argv.is_compat)) { compat_uptr_t compat; if (get_user(compat, argv.ptr.compat + nr)) return ERR_PTR(-EFAULT); |
1d1dbf813
|
395 |
|
0e028465d
|
396 397 398 399 400 |
return compat_ptr(compat); } #endif if (get_user(native, argv.ptr.native + nr)) |
1d1dbf813
|
401 |
return ERR_PTR(-EFAULT); |
0e028465d
|
402 |
return native; |
1d1dbf813
|
403 |
} |
1da177e4c
|
404 405 406 |
/* * count() counts the number of strings in array ARGV. */ |
ba2d01629
|
407 |
static int count(struct user_arg_ptr argv, int max) |
1da177e4c
|
408 409 |
{ int i = 0; |
0e028465d
|
410 |
if (argv.ptr.native != NULL) { |
1da177e4c
|
411 |
for (;;) { |
1d1dbf813
|
412 |
const char __user *p = get_user_arg_ptr(argv, i); |
1da177e4c
|
413 |
|
1da177e4c
|
414 415 |
if (!p) break; |
1d1dbf813
|
416 417 418 |
if (IS_ERR(p)) return -EFAULT; |
362e6663e
|
419 |
if (i++ >= max) |
1da177e4c
|
420 |
return -E2BIG; |
9aea5a65a
|
421 422 423 |
if (fatal_signal_pending(current)) return -ERESTARTNOHAND; |
1da177e4c
|
424 425 426 427 428 429 430 |
cond_resched(); } } return i; } /* |
b6a2fea39
|
431 432 433 |
* 'copy_strings()' copies argument/environment strings from the old * processes's memory to the new process's stack. The call to get_user_pages() * ensures the destination page is created and not swapped out. |
1da177e4c
|
434 |
*/ |
ba2d01629
|
435 |
static int copy_strings(int argc, struct user_arg_ptr argv, |
75c96f858
|
436 |
struct linux_binprm *bprm) |
1da177e4c
|
437 438 439 |
{ struct page *kmapped_page = NULL; char *kaddr = NULL; |
b6a2fea39
|
440 |
unsigned long kpos = 0; |
1da177e4c
|
441 442 443 |
int ret; while (argc-- > 0) { |
d7627467b
|
444 |
const char __user *str; |
1da177e4c
|
445 446 |
int len; unsigned long pos; |
1d1dbf813
|
447 448 449 |
ret = -EFAULT; str = get_user_arg_ptr(argv, argc); if (IS_ERR(str)) |
1da177e4c
|
450 |
goto out; |
1da177e4c
|
451 |
|
1d1dbf813
|
452 453 454 455 456 457 |
len = strnlen_user(str, MAX_ARG_STRLEN); if (!len) goto out; ret = -E2BIG; if (!valid_arg_len(bprm, len)) |
1da177e4c
|
458 |
goto out; |
1da177e4c
|
459 |
|
b6a2fea39
|
460 |
/* We're going to work our way backwords. */ |
1da177e4c
|
461 |
pos = bprm->p; |
b6a2fea39
|
462 463 |
str += len; bprm->p -= len; |
1da177e4c
|
464 465 |
while (len > 0) { |
1da177e4c
|
466 |
int offset, bytes_to_copy; |
1da177e4c
|
467 |
|
9aea5a65a
|
468 469 470 471 |
if (fatal_signal_pending(current)) { ret = -ERESTARTNOHAND; goto out; } |
7993bc1f4
|
472 |
cond_resched(); |
1da177e4c
|
473 |
offset = pos % PAGE_SIZE; |
b6a2fea39
|
474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 |
if (offset == 0) offset = PAGE_SIZE; bytes_to_copy = offset; if (bytes_to_copy > len) bytes_to_copy = len; offset -= bytes_to_copy; pos -= bytes_to_copy; str -= bytes_to_copy; len -= bytes_to_copy; if (!kmapped_page || kpos != (pos & PAGE_MASK)) { struct page *page; page = get_arg_page(bprm, pos, 1); |
1da177e4c
|
490 |
if (!page) { |
b6a2fea39
|
491 |
ret = -E2BIG; |
1da177e4c
|
492 493 |
goto out; } |
1da177e4c
|
494 |
|
b6a2fea39
|
495 496 |
if (kmapped_page) { flush_kernel_dcache_page(kmapped_page); |
1da177e4c
|
497 |
kunmap(kmapped_page); |
b6a2fea39
|
498 499 |
put_arg_page(kmapped_page); } |
1da177e4c
|
500 501 |
kmapped_page = page; kaddr = kmap(kmapped_page); |
b6a2fea39
|
502 503 |
kpos = pos & PAGE_MASK; flush_arg_page(bprm, kpos, kmapped_page); |
1da177e4c
|
504 |
} |
b6a2fea39
|
505 |
if (copy_from_user(kaddr+offset, str, bytes_to_copy)) { |
1da177e4c
|
506 507 508 |
ret = -EFAULT; goto out; } |
1da177e4c
|
509 510 511 512 |
} } ret = 0; out: |
b6a2fea39
|
513 514 |
if (kmapped_page) { flush_kernel_dcache_page(kmapped_page); |
1da177e4c
|
515 |
kunmap(kmapped_page); |
b6a2fea39
|
516 517 |
put_arg_page(kmapped_page); } |
1da177e4c
|
518 519 520 521 522 523 |
return ret; } /* * Like copy_strings, but get argv and its values from kernel memory. */ |
ba2d01629
|
524 |
int copy_strings_kernel(int argc, const char *const *__argv, |
d7627467b
|
525 |
struct linux_binprm *bprm) |
1da177e4c
|
526 527 528 |
{ int r; mm_segment_t oldfs = get_fs(); |
ba2d01629
|
529 |
struct user_arg_ptr argv = { |
0e028465d
|
530 |
.ptr.native = (const char __user *const __user *)__argv, |
ba2d01629
|
531 |
}; |
1da177e4c
|
532 |
set_fs(KERNEL_DS); |
ba2d01629
|
533 |
r = copy_strings(argc, argv, bprm); |
1da177e4c
|
534 |
set_fs(oldfs); |
ba2d01629
|
535 |
|
1da177e4c
|
536 537 |
return r; } |
1da177e4c
|
538 539 540 |
EXPORT_SYMBOL(copy_strings_kernel); #ifdef CONFIG_MMU |
b6a2fea39
|
541 |
|
1da177e4c
|
542 |
/* |
b6a2fea39
|
543 544 545 |
* During bprm_mm_init(), we create a temporary stack at STACK_TOP_MAX. Once * the binfmt code determines where the new stack should reside, we shift it to * its final location. The process proceeds as follows: |
1da177e4c
|
546 |
* |
b6a2fea39
|
547 548 549 550 551 552 |
* 1) Use shift to calculate the new vma endpoints. * 2) Extend vma to cover both the old and new ranges. This ensures the * arguments passed to subsequent functions are consistent. * 3) Move vma's page tables to the new range. * 4) Free up any cleared pgd range. * 5) Shrink the vma to cover only the new range. |
1da177e4c
|
553 |
*/ |
b6a2fea39
|
554 |
static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift) |
1da177e4c
|
555 556 |
{ struct mm_struct *mm = vma->vm_mm; |
b6a2fea39
|
557 558 559 560 561 |
unsigned long old_start = vma->vm_start; unsigned long old_end = vma->vm_end; unsigned long length = old_end - old_start; unsigned long new_start = old_start - shift; unsigned long new_end = old_end - shift; |
d16dfc550
|
562 |
struct mmu_gather tlb; |
1da177e4c
|
563 |
|
b6a2fea39
|
564 |
BUG_ON(new_start > new_end); |
1da177e4c
|
565 |
|
b6a2fea39
|
566 567 568 569 570 571 572 573 574 575 |
/* * ensure there are no vmas between where we want to go * and where we are */ if (vma != find_vma(mm, new_start)) return -EFAULT; /* * cover the whole range: [new_start, old_end) */ |
5beb49305
|
576 577 |
if (vma_adjust(vma, new_start, old_end, vma->vm_pgoff, NULL)) return -ENOMEM; |
b6a2fea39
|
578 579 580 581 582 583 |
/* * move the page tables downwards, on failure we rely on * process cleanup to remove whatever mess we made. */ if (length != move_page_tables(vma, old_start, |
38a76013a
|
584 |
vma, new_start, length, false)) |
b6a2fea39
|
585 586 587 |
return -ENOMEM; lru_add_drain(); |
d16dfc550
|
588 |
tlb_gather_mmu(&tlb, mm, 0); |
b6a2fea39
|
589 590 591 592 |
if (new_end > old_start) { /* * when the old and new regions overlap clear from new_end. */ |
d16dfc550
|
593 |
free_pgd_range(&tlb, new_end, old_end, new_end, |
b6a2fea39
|
594 595 596 597 598 599 600 601 |
vma->vm_next ? vma->vm_next->vm_start : 0); } else { /* * otherwise, clean from old_start; this is done to not touch * the address space in [new_end, old_start) some architectures * have constraints on va-space that make this illegal (IA64) - * for the others its just a little faster. */ |
d16dfc550
|
602 |
free_pgd_range(&tlb, old_start, old_end, new_end, |
b6a2fea39
|
603 |
vma->vm_next ? vma->vm_next->vm_start : 0); |
1da177e4c
|
604 |
} |
d16dfc550
|
605 |
tlb_finish_mmu(&tlb, new_end, old_end); |
b6a2fea39
|
606 607 |
/* |
5beb49305
|
608 |
* Shrink the vma to just the new range. Always succeeds. |
b6a2fea39
|
609 610 611 612 |
*/ vma_adjust(vma, new_start, new_end, vma->vm_pgoff, NULL); return 0; |
1da177e4c
|
613 |
} |
b6a2fea39
|
614 615 616 617 |
/* * Finalizes the stack vm_area_struct. The flags and permissions are updated, * the stack is optionally relocated, and some extra space is added. */ |
1da177e4c
|
618 619 620 621 |
int setup_arg_pages(struct linux_binprm *bprm, unsigned long stack_top, int executable_stack) { |
b6a2fea39
|
622 623 |
unsigned long ret; unsigned long stack_shift; |
1da177e4c
|
624 |
struct mm_struct *mm = current->mm; |
b6a2fea39
|
625 626 627 628 |
struct vm_area_struct *vma = bprm->vma; struct vm_area_struct *prev = NULL; unsigned long vm_flags; unsigned long stack_base; |
803bf5ec2
|
629 630 631 |
unsigned long stack_size; unsigned long stack_expand; unsigned long rlim_stack; |
1da177e4c
|
632 633 |
#ifdef CONFIG_STACK_GROWSUP |
1da177e4c
|
634 |
/* Limit stack size to 1GB */ |
d554ed895
|
635 |
stack_base = rlimit_max(RLIMIT_STACK); |
1da177e4c
|
636 637 |
if (stack_base > (1 << 30)) stack_base = 1 << 30; |
1da177e4c
|
638 |
|
b6a2fea39
|
639 640 641 |
/* Make sure we didn't let the argument array grow too large. */ if (vma->vm_end - vma->vm_start > stack_base) return -ENOMEM; |
1da177e4c
|
642 |
|
b6a2fea39
|
643 |
stack_base = PAGE_ALIGN(stack_top - stack_base); |
1da177e4c
|
644 |
|
b6a2fea39
|
645 646 647 |
stack_shift = vma->vm_start - stack_base; mm->arg_start = bprm->p - stack_shift; bprm->p = vma->vm_end - stack_shift; |
1da177e4c
|
648 |
#else |
b6a2fea39
|
649 650 |
stack_top = arch_align_stack(stack_top); stack_top = PAGE_ALIGN(stack_top); |
1b528181b
|
651 652 653 654 |
if (unlikely(stack_top < mmap_min_addr) || unlikely(vma->vm_end - vma->vm_start >= stack_top - mmap_min_addr)) return -ENOMEM; |
b6a2fea39
|
655 656 657 |
stack_shift = vma->vm_end - stack_top; bprm->p -= stack_shift; |
1da177e4c
|
658 |
mm->arg_start = bprm->p; |
1da177e4c
|
659 |
#endif |
1da177e4c
|
660 |
if (bprm->loader) |
b6a2fea39
|
661 662 |
bprm->loader -= stack_shift; bprm->exec -= stack_shift; |
1da177e4c
|
663 |
|
1da177e4c
|
664 |
down_write(&mm->mmap_sem); |
96a8e13ed
|
665 |
vm_flags = VM_STACK_FLAGS; |
b6a2fea39
|
666 667 668 669 670 671 672 673 674 675 676 |
/* * Adjust stack execute permissions; explicitly enable for * EXSTACK_ENABLE_X, disable for EXSTACK_DISABLE_X and leave alone * (arch default) otherwise. */ if (unlikely(executable_stack == EXSTACK_ENABLE_X)) vm_flags |= VM_EXEC; else if (executable_stack == EXSTACK_DISABLE_X) vm_flags &= ~VM_EXEC; vm_flags |= mm->def_flags; |
a8bef8ff6
|
677 |
vm_flags |= VM_STACK_INCOMPLETE_SETUP; |
b6a2fea39
|
678 679 680 681 682 683 684 685 686 687 |
ret = mprotect_fixup(vma, &prev, vma->vm_start, vma->vm_end, vm_flags); if (ret) goto out_unlock; BUG_ON(prev != vma); /* Move stack pages down in memory. */ if (stack_shift) { ret = shift_arg_pages(vma, stack_shift); |
fc63cf237
|
688 689 |
if (ret) goto out_unlock; |
1da177e4c
|
690 |
} |
a8bef8ff6
|
691 692 |
/* mprotect_fixup is overkill to remove the temporary stack flags */ vma->vm_flags &= ~VM_STACK_INCOMPLETE_SETUP; |
5ef097dd7
|
693 |
stack_expand = 131072UL; /* randomly 32*4k (or 2*64k) pages */ |
803bf5ec2
|
694 695 696 697 698 699 |
stack_size = vma->vm_end - vma->vm_start; /* * Align this down to a page boundary as expand_stack * will align it up. */ rlim_stack = rlimit(RLIMIT_STACK) & PAGE_MASK; |
b6a2fea39
|
700 |
#ifdef CONFIG_STACK_GROWSUP |
803bf5ec2
|
701 702 703 704 |
if (stack_size + stack_expand > rlim_stack) stack_base = vma->vm_start + rlim_stack; else stack_base = vma->vm_end + stack_expand; |
b6a2fea39
|
705 |
#else |
803bf5ec2
|
706 707 708 709 |
if (stack_size + stack_expand > rlim_stack) stack_base = vma->vm_end - rlim_stack; else stack_base = vma->vm_start - stack_expand; |
b6a2fea39
|
710 |
#endif |
3af9e8592
|
711 |
current->mm->start_stack = bprm->p; |
b6a2fea39
|
712 713 714 715 716 |
ret = expand_stack(vma, stack_base); if (ret) ret = -EFAULT; out_unlock: |
1da177e4c
|
717 |
up_write(&mm->mmap_sem); |
fc63cf237
|
718 |
return ret; |
1da177e4c
|
719 |
} |
1da177e4c
|
720 |
EXPORT_SYMBOL(setup_arg_pages); |
1da177e4c
|
721 722 723 724 |
#endif /* CONFIG_MMU */ struct file *open_exec(const char *name) { |
1da177e4c
|
725 |
struct file *file; |
e56b6a5dd
|
726 |
int err; |
669abf4e5
|
727 |
struct filename tmp = { .name = name }; |
47c805dc2
|
728 729 730 731 732 |
static const struct open_flags open_exec_flags = { .open_flag = O_LARGEFILE | O_RDONLY | __FMODE_EXEC, .acc_mode = MAY_EXEC | MAY_OPEN, .intent = LOOKUP_OPEN }; |
1da177e4c
|
733 |
|
669abf4e5
|
734 |
file = do_filp_open(AT_FDCWD, &tmp, &open_exec_flags, LOOKUP_FOLLOW); |
6e8341a11
|
735 |
if (IS_ERR(file)) |
e56b6a5dd
|
736 737 738 |
goto out; err = -EACCES; |
6e8341a11
|
739 740 |
if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) goto exit; |
e56b6a5dd
|
741 |
|
6e8341a11
|
742 743 |
if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) goto exit; |
e56b6a5dd
|
744 |
|
2a12a9d78
|
745 |
fsnotify_open(file); |
6110e3abb
|
746 |
|
e56b6a5dd
|
747 |
err = deny_write_access(file); |
6e8341a11
|
748 749 |
if (err) goto exit; |
1da177e4c
|
750 |
|
6e8341a11
|
751 |
out: |
e56b6a5dd
|
752 |
return file; |
6e8341a11
|
753 754 |
exit: fput(file); |
e56b6a5dd
|
755 756 |
return ERR_PTR(err); } |
1da177e4c
|
757 |
EXPORT_SYMBOL(open_exec); |
6777d773a
|
758 759 |
int kernel_read(struct file *file, loff_t offset, char *addr, unsigned long count) |
1da177e4c
|
760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 |
{ mm_segment_t old_fs; loff_t pos = offset; int result; old_fs = get_fs(); set_fs(get_ds()); /* The cast to a user pointer is valid due to the set_fs() */ result = vfs_read(file, (void __user *)addr, count, &pos); set_fs(old_fs); return result; } EXPORT_SYMBOL(kernel_read); static int exec_mmap(struct mm_struct *mm) { struct task_struct *tsk; struct mm_struct * old_mm, *active_mm; /* Notify parent that we're no longer interested in the old VM */ tsk = current; old_mm = current->mm; mm_release(tsk, old_mm); if (old_mm) { |
4fe7efdbd
|
786 |
sync_mm_rss(old_mm); |
1da177e4c
|
787 788 789 790 |
/* * Make sure that if there is a core dump in progress * for the old mm, we get out and die instead of going * through with the exec. We must hold mmap_sem around |
999d9fc16
|
791 |
* checking core_state and changing tsk->mm. |
1da177e4c
|
792 793 |
*/ down_read(&old_mm->mmap_sem); |
999d9fc16
|
794 |
if (unlikely(old_mm->core_state)) { |
1da177e4c
|
795 796 797 798 799 800 801 802 803 804 805 806 807 |
up_read(&old_mm->mmap_sem); return -EINTR; } } task_lock(tsk); active_mm = tsk->active_mm; tsk->mm = mm; tsk->active_mm = mm; activate_mm(active_mm, mm); task_unlock(tsk); arch_pick_mmap_layout(mm); if (old_mm) { up_read(&old_mm->mmap_sem); |
7dddb12c6
|
808 |
BUG_ON(active_mm != old_mm); |
701085b21
|
809 |
setmax_mm_hiwater_rss(&tsk->signal->maxrss, old_mm); |
31a78f23b
|
810 |
mm_update_next_owner(old_mm); |
1da177e4c
|
811 812 813 814 815 816 817 818 819 820 821 822 823 |
mmput(old_mm); return 0; } mmdrop(active_mm); return 0; } /* * This function makes sure the current process has its own signal table, * so that flush_signal_handlers can later reset the handlers without * disturbing other processes. (Other processes might share the signal * table via the CLONE_SIGHAND option to clone().) */ |
858119e15
|
824 |
static int de_thread(struct task_struct *tsk) |
1da177e4c
|
825 826 |
{ struct signal_struct *sig = tsk->signal; |
b2c903b87
|
827 |
struct sighand_struct *oldsighand = tsk->sighand; |
1da177e4c
|
828 |
spinlock_t *lock = &oldsighand->siglock; |
1da177e4c
|
829 |
|
aafe6c2a2
|
830 |
if (thread_group_empty(tsk)) |
1da177e4c
|
831 832 833 834 |
goto no_thread_group; /* * Kill all other threads in the thread group. |
1da177e4c
|
835 |
*/ |
1da177e4c
|
836 |
spin_lock_irq(lock); |
ed5d2cac1
|
837 |
if (signal_group_exit(sig)) { |
1da177e4c
|
838 839 840 841 842 |
/* * Another group action in progress, just * return so that the signal is processed. */ spin_unlock_irq(lock); |
1da177e4c
|
843 844 |
return -EAGAIN; } |
d344193a0
|
845 |
|
ed5d2cac1
|
846 |
sig->group_exit_task = tsk; |
d344193a0
|
847 848 849 |
sig->notify_count = zap_other_threads(tsk); if (!thread_group_leader(tsk)) sig->notify_count--; |
1da177e4c
|
850 |
|
d344193a0
|
851 |
while (sig->notify_count) { |
d5bbd43d5
|
852 |
__set_current_state(TASK_KILLABLE); |
1da177e4c
|
853 854 |
spin_unlock_irq(lock); schedule(); |
d5bbd43d5
|
855 856 |
if (unlikely(__fatal_signal_pending(tsk))) goto killed; |
1da177e4c
|
857 858 |
spin_lock_irq(lock); } |
1da177e4c
|
859 860 861 862 863 864 865 |
spin_unlock_irq(lock); /* * At this point all other threads have exited, all we have to * do is to wait for the thread group leader to become inactive, * and to assume its PID: */ |
aafe6c2a2
|
866 |
if (!thread_group_leader(tsk)) { |
8187926bd
|
867 |
struct task_struct *leader = tsk->group_leader; |
6db840fa7
|
868 |
|
2800d8d19
|
869 |
sig->notify_count = -1; /* for exit_notify() */ |
6db840fa7
|
870 871 872 873 |
for (;;) { write_lock_irq(&tasklist_lock); if (likely(leader->exit_state)) break; |
d5bbd43d5
|
874 |
__set_current_state(TASK_KILLABLE); |
6db840fa7
|
875 876 |
write_unlock_irq(&tasklist_lock); schedule(); |
d5bbd43d5
|
877 878 |
if (unlikely(__fatal_signal_pending(tsk))) goto killed; |
6db840fa7
|
879 |
} |
1da177e4c
|
880 |
|
f5e902817
|
881 882 883 884 885 886 887 888 889 890 |
/* * The only record we have of the real-time age of a * process, regardless of execs it's done, is start_time. * All the past CPU time is accumulated in signal_struct * from sister threads now dead. But in this non-leader * exec, nothing survives from the original leader thread, * whose birth marks the true age of this process now. * When we take on its identity by switching to its PID, we * also take its birthdate (always earlier than our own). */ |
aafe6c2a2
|
891 |
tsk->start_time = leader->start_time; |
f5e902817
|
892 |
|
bac0abd61
|
893 894 |
BUG_ON(!same_thread_group(leader, tsk)); BUG_ON(has_group_leader_pid(tsk)); |
1da177e4c
|
895 896 897 898 899 900 |
/* * An exec() starts a new thread group with the * TGID of the previous thread group. Rehash the * two threads with a switched PID, and release * the former thread group leader: */ |
d73d65293
|
901 902 |
/* Become a process group leader with the old leader's pid. |
c18258c6f
|
903 904 |
* The old leader becomes a thread of the this thread group. * Note: The old leader also uses this pid until release_task |
d73d65293
|
905 906 |
* is called. Odd but simple and correct. */ |
aafe6c2a2
|
907 908 |
detach_pid(tsk, PIDTYPE_PID); tsk->pid = leader->pid; |
3743ca05f
|
909 |
attach_pid(tsk, PIDTYPE_PID, task_pid(leader)); |
aafe6c2a2
|
910 911 |
transfer_pid(leader, tsk, PIDTYPE_PGID); transfer_pid(leader, tsk, PIDTYPE_SID); |
9cd80bbb0
|
912 |
|
aafe6c2a2
|
913 |
list_replace_rcu(&leader->tasks, &tsk->tasks); |
9cd80bbb0
|
914 |
list_replace_init(&leader->sibling, &tsk->sibling); |
1da177e4c
|
915 |
|
aafe6c2a2
|
916 917 |
tsk->group_leader = tsk; leader->group_leader = tsk; |
de12a7878
|
918 |
|
aafe6c2a2
|
919 |
tsk->exit_signal = SIGCHLD; |
087806b12
|
920 |
leader->exit_signal = -1; |
962b564cf
|
921 922 923 |
BUG_ON(leader->exit_state != EXIT_ZOMBIE); leader->exit_state = EXIT_DEAD; |
eac1b5e57
|
924 925 926 927 928 929 930 931 |
/* * We are going to release_task()->ptrace_unlink() silently, * the tracer can sleep in do_wait(). EXIT_DEAD guarantees * the tracer wont't block again waiting for this thread. */ if (unlikely(leader->ptrace)) __wake_up_parent(leader, leader->parent); |
1da177e4c
|
932 |
write_unlock_irq(&tasklist_lock); |
8187926bd
|
933 934 |
release_task(leader); |
ed5d2cac1
|
935 |
} |
1da177e4c
|
936 |
|
6db840fa7
|
937 938 |
sig->group_exit_task = NULL; sig->notify_count = 0; |
1da177e4c
|
939 940 |
no_thread_group: |
e63682534
|
941 942 |
/* we have changed execution domain */ tsk->exit_signal = SIGCHLD; |
1da177e4c
|
943 |
exit_itimers(sig); |
cbaffba12
|
944 |
flush_itimer_signals(); |
329f7dba5
|
945 |
|
b2c903b87
|
946 947 |
if (atomic_read(&oldsighand->count) != 1) { struct sighand_struct *newsighand; |
1da177e4c
|
948 |
/* |
b2c903b87
|
949 950 |
* This ->sighand is shared with the CLONE_SIGHAND * but not CLONE_THREAD task, switch to the new one. |
1da177e4c
|
951 |
*/ |
b2c903b87
|
952 953 954 |
newsighand = kmem_cache_alloc(sighand_cachep, GFP_KERNEL); if (!newsighand) return -ENOMEM; |
1da177e4c
|
955 956 957 958 959 960 |
atomic_set(&newsighand->count, 1); memcpy(newsighand->action, oldsighand->action, sizeof(newsighand->action)); write_lock_irq(&tasklist_lock); spin_lock(&oldsighand->siglock); |
aafe6c2a2
|
961 |
rcu_assign_pointer(tsk->sighand, newsighand); |
1da177e4c
|
962 963 |
spin_unlock(&oldsighand->siglock); write_unlock_irq(&tasklist_lock); |
fba2afaae
|
964 |
__cleanup_sighand(oldsighand); |
1da177e4c
|
965 |
} |
aafe6c2a2
|
966 |
BUG_ON(!thread_group_leader(tsk)); |
1da177e4c
|
967 |
return 0; |
d5bbd43d5
|
968 969 970 971 972 973 974 975 |
killed: /* protects against exit_notify() and __exit_signal() */ read_lock(&tasklist_lock); sig->group_exit_task = NULL; sig->notify_count = 0; read_unlock(&tasklist_lock); return -EAGAIN; |
1da177e4c
|
976 |
} |
0840a90d9
|
977 |
|
59714d65d
|
978 |
char *get_task_comm(char *buf, struct task_struct *tsk) |
1da177e4c
|
979 980 981 982 983 |
{ /* buf must be at least sizeof(tsk->comm) in size */ task_lock(tsk); strncpy(buf, tsk->comm, sizeof(tsk->comm)); task_unlock(tsk); |
59714d65d
|
984 |
return buf; |
1da177e4c
|
985 |
} |
7d74f492e
|
986 |
EXPORT_SYMBOL_GPL(get_task_comm); |
1da177e4c
|
987 |
|
6a6d27de3
|
988 989 990 991 |
/* * These functions flushes out all traces of the currently running executable * so that a new one can be started */ |
1da177e4c
|
992 993 994 |
void set_task_comm(struct task_struct *tsk, char *buf) { task_lock(tsk); |
4614a696b
|
995 |
|
43d2b1132
|
996 |
trace_task_rename(tsk, buf); |
4614a696b
|
997 998 999 1000 1001 1002 1003 1004 |
/* * Threads may access current->comm without holding * the task lock, so write the string carefully. * Readers without a lock may see incomplete new * names but are safe from non-terminating string reads. */ memset(tsk->comm, 0, TASK_COMM_LEN); wmb(); |
1da177e4c
|
1005 1006 |
strlcpy(tsk->comm, buf, sizeof(tsk->comm)); task_unlock(tsk); |
cdd6c482c
|
1007 |
perf_event_comm(tsk); |
1da177e4c
|
1008 |
} |
96e02d158
|
1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 |
static void filename_to_taskname(char *tcomm, const char *fn, unsigned int len) { int i, ch; /* Copies the binary name from after last slash */ for (i = 0; (ch = *(fn++)) != '\0';) { if (ch == '/') i = 0; /* overwrite what we wrote */ else if (i < len - 1) tcomm[i++] = ch; } tcomm[i] = '\0'; } |
1da177e4c
|
1023 1024 |
int flush_old_exec(struct linux_binprm * bprm) { |
221af7f87
|
1025 |
int retval; |
1da177e4c
|
1026 1027 1028 1029 1030 1031 1032 1033 |
/* * Make sure we have a private signal table and that * we are unassociated from the previous thread group. */ retval = de_thread(current); if (retval) goto out; |
925d1c401
|
1034 |
set_mm_exe_file(bprm->mm, bprm->file); |
96e02d158
|
1035 |
filename_to_taskname(bprm->tcomm, bprm->filename, sizeof(bprm->tcomm)); |
1da177e4c
|
1036 |
/* |
1da177e4c
|
1037 1038 |
* Release all of the old mmap stuff */ |
3c77f8457
|
1039 |
acct_arg_size(bprm, 0); |
1da177e4c
|
1040 1041 |
retval = exec_mmap(bprm->mm); if (retval) |
fd8328be8
|
1042 |
goto out; |
1da177e4c
|
1043 1044 |
bprm->mm = NULL; /* We're using it now */ |
7ab02af42
|
1045 |
|
dac853ae8
|
1046 |
set_fs(USER_DS); |
b40a79591
|
1047 1048 |
current->flags &= ~(PF_RANDOMIZE | PF_FORKNOEXEC | PF_KTHREAD | PF_NOFREEZE); |
7ab02af42
|
1049 1050 |
flush_thread(); current->personality &= ~bprm->per_clear; |
221af7f87
|
1051 1052 1053 1054 1055 1056 |
return 0; out: return retval; } EXPORT_SYMBOL(flush_old_exec); |
1b5d783c9
|
1057 1058 1059 1060 1061 1062 |
void would_dump(struct linux_binprm *bprm, struct file *file) { if (inode_permission(file->f_path.dentry->d_inode, MAY_READ) < 0) bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP; } EXPORT_SYMBOL(would_dump); |
221af7f87
|
1063 1064 |
void setup_new_exec(struct linux_binprm * bprm) { |
221af7f87
|
1065 |
arch_pick_mmap_layout(current->mm); |
1da177e4c
|
1066 1067 |
/* This is the point of no return */ |
1da177e4c
|
1068 |
current->sas_ss_sp = current->sas_ss_size = 0; |
8e96e3b7b
|
1069 |
if (uid_eq(current_euid(), current_uid()) && gid_eq(current_egid(), current_gid())) |
0f4cfb2e4
|
1070 |
set_dumpable(current->mm, SUID_DUMPABLE_ENABLED); |
d6e711448
|
1071 |
else |
6c5d52382
|
1072 |
set_dumpable(current->mm, suid_dumpable); |
d6e711448
|
1073 |
|
96e02d158
|
1074 |
set_task_comm(current, bprm->tcomm); |
1da177e4c
|
1075 |
|
0551fbd29
|
1076 1077 1078 1079 1080 |
/* Set the new mm task size. We have to do that late because it may * depend on TIF_32BIT which is only updated in flush_thread() on * some architectures like powerpc */ current->mm->task_size = TASK_SIZE; |
a6f76f23d
|
1081 |
/* install the new credentials */ |
8e96e3b7b
|
1082 1083 |
if (!uid_eq(bprm->cred->uid, current_euid()) || !gid_eq(bprm->cred->gid, current_egid())) { |
d2d56c5f5
|
1084 |
current->pdeath_signal = 0; |
1b5d783c9
|
1085 1086 1087 1088 |
} else { would_dump(bprm, bprm->file); if (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP) set_dumpable(current->mm, suid_dumpable); |
1da177e4c
|
1089 |
} |
f65cb45cb
|
1090 1091 1092 1093 1094 |
/* * Flush performance counters when crossing a * security domain: */ if (!get_dumpable(current->mm)) |
cdd6c482c
|
1095 |
perf_event_exit_task(current); |
f65cb45cb
|
1096 |
|
1da177e4c
|
1097 1098 1099 1100 1101 1102 |
/* An exec changes our domain. We are no longer part of the thread group */ current->self_exec_id++; flush_signal_handlers(current, 0); |
6a6d27de3
|
1103 |
do_close_on_exec(current->files); |
1da177e4c
|
1104 |
} |
221af7f87
|
1105 |
EXPORT_SYMBOL(setup_new_exec); |
1da177e4c
|
1106 |
|
a6f76f23d
|
1107 |
/* |
a2a8474c3
|
1108 1109 1110 1111 1112 1113 1114 |
* Prepare credentials and lock ->cred_guard_mutex. * install_exec_creds() commits the new creds and drops the lock. * Or, if exec fails before, free_bprm() should release ->cred and * and unlock. */ int prepare_bprm_creds(struct linux_binprm *bprm) { |
9b1bf12d5
|
1115 |
if (mutex_lock_interruptible(¤t->signal->cred_guard_mutex)) |
a2a8474c3
|
1116 1117 1118 1119 1120 |
return -ERESTARTNOINTR; bprm->cred = prepare_exec_creds(); if (likely(bprm->cred)) return 0; |
9b1bf12d5
|
1121 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a2a8474c3
|
1122 1123 1124 1125 1126 1127 1128 |
return -ENOMEM; } void free_bprm(struct linux_binprm *bprm) { free_arg_pages(bprm); if (bprm->cred) { |
9b1bf12d5
|
1129 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a2a8474c3
|
1130 1131 |
abort_creds(bprm->cred); } |
b66c59840
|
1132 1133 1134 |
/* If a binfmt changed the interp, free it. */ if (bprm->interp != bprm->filename) kfree(bprm->interp); |
a2a8474c3
|
1135 1136 |
kfree(bprm); } |
b66c59840
|
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 |
int bprm_change_interp(char *interp, struct linux_binprm *bprm) { /* If a binfmt changed the interp, free it first. */ if (bprm->interp != bprm->filename) kfree(bprm->interp); bprm->interp = kstrdup(interp, GFP_KERNEL); if (!bprm->interp) return -ENOMEM; return 0; } EXPORT_SYMBOL(bprm_change_interp); |
a2a8474c3
|
1148 |
/* |
a6f76f23d
|
1149 1150 1151 1152 1153 1154 1155 1156 |
* install the new credentials for this executable */ void install_exec_creds(struct linux_binprm *bprm) { security_bprm_committing_creds(bprm); commit_creds(bprm->cred); bprm->cred = NULL; |
a2a8474c3
|
1157 1158 |
/* * cred_guard_mutex must be held at least to this point to prevent |
a6f76f23d
|
1159 |
* ptrace_attach() from altering our determination of the task's |
a2a8474c3
|
1160 1161 |
* credentials; any time after this it may be unlocked. */ |
a6f76f23d
|
1162 |
security_bprm_committed_creds(bprm); |
9b1bf12d5
|
1163 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a6f76f23d
|
1164 1165 1166 1167 1168 |
} EXPORT_SYMBOL(install_exec_creds); /* * determine how safe it is to execute the proposed program |
9b1bf12d5
|
1169 |
* - the caller must hold ->cred_guard_mutex to protect against |
a6f76f23d
|
1170 1171 |
* PTRACE_ATTACH */ |
f47ec3f28
|
1172 |
static int check_unsafe_exec(struct linux_binprm *bprm) |
a6f76f23d
|
1173 |
{ |
0bf2f3aec
|
1174 |
struct task_struct *p = current, *t; |
f1191b50e
|
1175 |
unsigned n_fs; |
498052bba
|
1176 |
int res = 0; |
a6f76f23d
|
1177 |
|
4b9d33e6d
|
1178 1179 1180 1181 1182 1183 |
if (p->ptrace) { if (p->ptrace & PT_PTRACE_CAP) bprm->unsafe |= LSM_UNSAFE_PTRACE_CAP; else bprm->unsafe |= LSM_UNSAFE_PTRACE; } |
a6f76f23d
|
1184 |
|
259e5e6c7
|
1185 1186 1187 1188 1189 1190 |
/* * This isn't strictly necessary, but it makes it harder for LSMs to * mess up. */ if (current->no_new_privs) bprm->unsafe |= LSM_UNSAFE_NO_NEW_PRIVS; |
0bf2f3aec
|
1191 |
n_fs = 1; |
2a4419b5b
|
1192 |
spin_lock(&p->fs->lock); |
437f7fdb6
|
1193 |
rcu_read_lock(); |
0bf2f3aec
|
1194 1195 1196 |
for (t = next_thread(p); t != p; t = next_thread(t)) { if (t->fs == p->fs) n_fs++; |
0bf2f3aec
|
1197 |
} |
437f7fdb6
|
1198 |
rcu_read_unlock(); |
0bf2f3aec
|
1199 |
|
f1191b50e
|
1200 |
if (p->fs->users > n_fs) { |
a6f76f23d
|
1201 |
bprm->unsafe |= LSM_UNSAFE_SHARE; |
498052bba
|
1202 |
} else { |
8c652f96d
|
1203 1204 1205 1206 1207 |
res = -EAGAIN; if (!p->fs->in_exec) { p->fs->in_exec = 1; res = 1; } |
498052bba
|
1208 |
} |
2a4419b5b
|
1209 |
spin_unlock(&p->fs->lock); |
498052bba
|
1210 1211 |
return res; |
a6f76f23d
|
1212 |
} |
1da177e4c
|
1213 1214 1215 |
/* * Fill the binprm structure from the inode. * Check permissions, then read the first 128 (BINPRM_BUF_SIZE) bytes |
a6f76f23d
|
1216 1217 |
* * This may be called multiple times for binary chains (scripts for example). |
1da177e4c
|
1218 1219 1220 |
*/ int prepare_binprm(struct linux_binprm *bprm) { |
a6f76f23d
|
1221 |
umode_t mode; |
0f7fc9e4d
|
1222 |
struct inode * inode = bprm->file->f_path.dentry->d_inode; |
1da177e4c
|
1223 1224 1225 |
int retval; mode = inode->i_mode; |
1da177e4c
|
1226 1227 |
if (bprm->file->f_op == NULL) return -EACCES; |
a6f76f23d
|
1228 1229 1230 |
/* clear any previous set[ug]id data from a previous binary */ bprm->cred->euid = current_euid(); bprm->cred->egid = current_egid(); |
1da177e4c
|
1231 |
|
259e5e6c7
|
1232 |
if (!(bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID) && |
3cdf5b45f
|
1233 1234 1235 |
!current->no_new_privs && kuid_has_mapping(bprm->cred->user_ns, inode->i_uid) && kgid_has_mapping(bprm->cred->user_ns, inode->i_gid)) { |
1da177e4c
|
1236 1237 |
/* Set-uid? */ if (mode & S_ISUID) { |
a6f76f23d
|
1238 1239 |
bprm->per_clear |= PER_CLEAR_ON_SETID; bprm->cred->euid = inode->i_uid; |
1da177e4c
|
1240 1241 1242 1243 1244 1245 1246 1247 1248 |
} /* Set-gid? */ /* * If setgid is set but no group execute bit then this * is a candidate for mandatory locking, not a setgid * executable. */ if ((mode & (S_ISGID | S_IXGRP)) == (S_ISGID | S_IXGRP)) { |
a6f76f23d
|
1249 1250 |
bprm->per_clear |= PER_CLEAR_ON_SETID; bprm->cred->egid = inode->i_gid; |
1da177e4c
|
1251 1252 1253 1254 |
} } /* fill in binprm security blob */ |
a6f76f23d
|
1255 |
retval = security_bprm_set_creds(bprm); |
1da177e4c
|
1256 1257 |
if (retval) return retval; |
a6f76f23d
|
1258 |
bprm->cred_prepared = 1; |
1da177e4c
|
1259 |
|
a6f76f23d
|
1260 1261 |
memset(bprm->buf, 0, BINPRM_BUF_SIZE); return kernel_read(bprm->file, 0, bprm->buf, BINPRM_BUF_SIZE); |
1da177e4c
|
1262 1263 1264 |
} EXPORT_SYMBOL(prepare_binprm); |
4fc75ff48
|
1265 1266 1267 1268 1269 |
/* * Arguments are '\0' separated strings found at the location bprm->p * points to; chop off the first by relocating brpm->p to right after * the first '\0' encountered. */ |
b6a2fea39
|
1270 |
int remove_arg_zero(struct linux_binprm *bprm) |
1da177e4c
|
1271 |
{ |
b6a2fea39
|
1272 1273 1274 1275 |
int ret = 0; unsigned long offset; char *kaddr; struct page *page; |
4fc75ff48
|
1276 |
|
b6a2fea39
|
1277 1278 |
if (!bprm->argc) return 0; |
1da177e4c
|
1279 |
|
b6a2fea39
|
1280 1281 1282 1283 1284 1285 1286 |
do { offset = bprm->p & ~PAGE_MASK; page = get_arg_page(bprm, bprm->p, 0); if (!page) { ret = -EFAULT; goto out; } |
e8e3c3d66
|
1287 |
kaddr = kmap_atomic(page); |
4fc75ff48
|
1288 |
|
b6a2fea39
|
1289 1290 1291 |
for (; offset < PAGE_SIZE && kaddr[offset]; offset++, bprm->p++) ; |
4fc75ff48
|
1292 |
|
e8e3c3d66
|
1293 |
kunmap_atomic(kaddr); |
b6a2fea39
|
1294 |
put_arg_page(page); |
4fc75ff48
|
1295 |
|
b6a2fea39
|
1296 1297 1298 |
if (offset == PAGE_SIZE) free_arg_page(bprm, (bprm->p >> PAGE_SHIFT) - 1); } while (offset == PAGE_SIZE); |
4fc75ff48
|
1299 |
|
b6a2fea39
|
1300 1301 1302 |
bprm->p++; bprm->argc--; ret = 0; |
4fc75ff48
|
1303 |
|
b6a2fea39
|
1304 1305 |
out: return ret; |
1da177e4c
|
1306 |
} |
1da177e4c
|
1307 1308 1309 1310 1311 |
EXPORT_SYMBOL(remove_arg_zero); /* * cycle the list of binary formats handler, until one recognizes the image */ |
3c456bfc4
|
1312 |
int search_binary_handler(struct linux_binprm *bprm) |
1da177e4c
|
1313 |
{ |
85f334666
|
1314 |
unsigned int depth = bprm->recursion_depth; |
1da177e4c
|
1315 1316 |
int try,retval; struct linux_binfmt *fmt; |
6308191f6
|
1317 |
pid_t old_pid, old_vpid; |
1da177e4c
|
1318 |
|
d74026986
|
1319 1320 1321 |
/* This allows 4 levels of binfmt rewrites before failing hard. */ if (depth > 5) return -ELOOP; |
1da177e4c
|
1322 1323 1324 |
retval = security_bprm_check(bprm); if (retval) return retval; |
473ae30bc
|
1325 1326 1327 |
retval = audit_bprm(bprm); if (retval) return retval; |
bb188d7e6
|
1328 |
/* Need to fetch pid before load_binary changes it */ |
6308191f6
|
1329 |
old_pid = current->pid; |
bb188d7e6
|
1330 |
rcu_read_lock(); |
6308191f6
|
1331 |
old_vpid = task_pid_nr_ns(current, task_active_pid_ns(current->parent)); |
bb188d7e6
|
1332 |
rcu_read_unlock(); |
1da177e4c
|
1333 1334 1335 |
retval = -ENOENT; for (try=0; try<2; try++) { read_lock(&binfmt_lock); |
e4dc1b14d
|
1336 |
list_for_each_entry(fmt, &formats, lh) { |
71613c3b8
|
1337 |
int (*fn)(struct linux_binprm *) = fmt->load_binary; |
1da177e4c
|
1338 1339 1340 1341 1342 |
if (!fn) continue; if (!try_module_get(fmt->module)) continue; read_unlock(&binfmt_lock); |
d74026986
|
1343 |
bprm->recursion_depth = depth + 1; |
71613c3b8
|
1344 |
retval = fn(bprm); |
85f334666
|
1345 |
bprm->recursion_depth = depth; |
1da177e4c
|
1346 |
if (retval >= 0) { |
4ff16c25e
|
1347 1348 |
if (depth == 0) { trace_sched_process_exec(current, old_pid, bprm); |
6308191f6
|
1349 |
ptrace_event(PTRACE_EVENT_EXEC, old_vpid); |
4ff16c25e
|
1350 |
} |
1da177e4c
|
1351 1352 1353 1354 1355 1356 |
put_binfmt(fmt); allow_write_access(bprm->file); if (bprm->file) fput(bprm->file); bprm->file = NULL; current->did_exec = 1; |
9f46080c4
|
1357 |
proc_exec_connector(current); |
1da177e4c
|
1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 |
return retval; } read_lock(&binfmt_lock); put_binfmt(fmt); if (retval != -ENOEXEC || bprm->mm == NULL) break; if (!bprm->file) { read_unlock(&binfmt_lock); return retval; } } read_unlock(&binfmt_lock); |
b4edf8bd0
|
1370 |
#ifdef CONFIG_MODULES |
1da177e4c
|
1371 1372 |
if (retval != -ENOEXEC || bprm->mm == NULL) { break; |
5f4123be3
|
1373 |
} else { |
1da177e4c
|
1374 1375 1376 1377 1378 1379 1380 |
#define printable(c) (((c)=='\t') || ((c)==' ') || (0x20<=(c) && (c)<=0x7e)) if (printable(bprm->buf[0]) && printable(bprm->buf[1]) && printable(bprm->buf[2]) && printable(bprm->buf[3])) break; /* -ENOEXEC */ |
912193521
|
1381 1382 |
if (try) break; /* -ENOEXEC */ |
1da177e4c
|
1383 |
request_module("binfmt-%04x", *(unsigned short *)(&bprm->buf[2])); |
1da177e4c
|
1384 |
} |
b4edf8bd0
|
1385 1386 1387 |
#else break; #endif |
1da177e4c
|
1388 1389 1390 1391 1392 1393 1394 1395 1396 |
} return retval; } EXPORT_SYMBOL(search_binary_handler); /* * sys_execve() executes a new program. */ |
ba2d01629
|
1397 1398 |
static int do_execve_common(const char *filename, struct user_arg_ptr argv, |
835ab32df
|
1399 |
struct user_arg_ptr envp) |
1da177e4c
|
1400 1401 1402 |
{ struct linux_binprm *bprm; struct file *file; |
3b1253880
|
1403 |
struct files_struct *displaced; |
8c652f96d
|
1404 |
bool clear_in_exec; |
1da177e4c
|
1405 |
int retval; |
72fa59970
|
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 |
const struct cred *cred = current_cred(); /* * We move the actual failure in case of RLIMIT_NPROC excess from * set*uid() to execve() because too many poorly written programs * don't check setuid() return code. Here we additionally recheck * whether NPROC limit is still exceeded. */ if ((current->flags & PF_NPROC_EXCEEDED) && atomic_read(&cred->user->processes) > rlimit(RLIMIT_NPROC)) { retval = -EAGAIN; goto out_ret; } /* We're below the limit (still or again), so we don't want to make * further execve() calls fail. */ current->flags &= ~PF_NPROC_EXCEEDED; |
1da177e4c
|
1423 |
|
3b1253880
|
1424 |
retval = unshare_files(&displaced); |
fd8328be8
|
1425 1426 |
if (retval) goto out_ret; |
1da177e4c
|
1427 |
retval = -ENOMEM; |
11b0b5abb
|
1428 |
bprm = kzalloc(sizeof(*bprm), GFP_KERNEL); |
1da177e4c
|
1429 |
if (!bprm) |
fd8328be8
|
1430 |
goto out_files; |
1da177e4c
|
1431 |
|
a2a8474c3
|
1432 1433 |
retval = prepare_bprm_creds(bprm); if (retval) |
a6f76f23d
|
1434 |
goto out_free; |
498052bba
|
1435 1436 |
retval = check_unsafe_exec(bprm); |
8c652f96d
|
1437 |
if (retval < 0) |
a2a8474c3
|
1438 |
goto out_free; |
8c652f96d
|
1439 |
clear_in_exec = retval; |
a2a8474c3
|
1440 |
current->in_execve = 1; |
a6f76f23d
|
1441 |
|
1da177e4c
|
1442 1443 1444 |
file = open_exec(filename); retval = PTR_ERR(file); if (IS_ERR(file)) |
498052bba
|
1445 |
goto out_unmark; |
1da177e4c
|
1446 1447 |
sched_exec(); |
1da177e4c
|
1448 1449 1450 |
bprm->file = file; bprm->filename = filename; bprm->interp = filename; |
1da177e4c
|
1451 |
|
b6a2fea39
|
1452 1453 1454 |
retval = bprm_mm_init(bprm); if (retval) goto out_file; |
1da177e4c
|
1455 |
|
b6a2fea39
|
1456 |
bprm->argc = count(argv, MAX_ARG_STRINGS); |
1da177e4c
|
1457 |
if ((retval = bprm->argc) < 0) |
a6f76f23d
|
1458 |
goto out; |
1da177e4c
|
1459 |
|
b6a2fea39
|
1460 |
bprm->envc = count(envp, MAX_ARG_STRINGS); |
1da177e4c
|
1461 |
if ((retval = bprm->envc) < 0) |
1da177e4c
|
1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 |
goto out; retval = prepare_binprm(bprm); if (retval < 0) goto out; retval = copy_strings_kernel(1, &bprm->filename, bprm); if (retval < 0) goto out; bprm->exec = bprm->p; retval = copy_strings(bprm->envc, envp, bprm); if (retval < 0) goto out; retval = copy_strings(bprm->argc, argv, bprm); if (retval < 0) goto out; |
3c456bfc4
|
1480 |
retval = search_binary_handler(bprm); |
a6f76f23d
|
1481 1482 |
if (retval < 0) goto out; |
1da177e4c
|
1483 |
|
a6f76f23d
|
1484 |
/* execve succeeded */ |
498052bba
|
1485 |
current->fs->in_exec = 0; |
f9ce1f1cd
|
1486 |
current->in_execve = 0; |
a6f76f23d
|
1487 1488 1489 1490 1491 |
acct_update_integrals(current); free_bprm(bprm); if (displaced) put_files_struct(displaced); return retval; |
1da177e4c
|
1492 |
|
a6f76f23d
|
1493 |
out: |
3c77f8457
|
1494 1495 1496 1497 |
if (bprm->mm) { acct_arg_size(bprm, 0); mmput(bprm->mm); } |
1da177e4c
|
1498 1499 1500 1501 1502 1503 |
out_file: if (bprm->file) { allow_write_access(bprm->file); fput(bprm->file); } |
a6f76f23d
|
1504 |
|
498052bba
|
1505 |
out_unmark: |
8c652f96d
|
1506 1507 |
if (clear_in_exec) current->fs->in_exec = 0; |
f9ce1f1cd
|
1508 |
current->in_execve = 0; |
a6f76f23d
|
1509 1510 |
out_free: |
08a6fac1c
|
1511 |
free_bprm(bprm); |
1da177e4c
|
1512 |
|
fd8328be8
|
1513 |
out_files: |
3b1253880
|
1514 1515 |
if (displaced) reset_files_struct(displaced); |
1da177e4c
|
1516 1517 1518 |
out_ret: return retval; } |
ba2d01629
|
1519 1520 |
int do_execve(const char *filename, const char __user *const __user *__argv, |
da3d4c5fa
|
1521 |
const char __user *const __user *__envp) |
ba2d01629
|
1522 |
{ |
0e028465d
|
1523 1524 |
struct user_arg_ptr argv = { .ptr.native = __argv }; struct user_arg_ptr envp = { .ptr.native = __envp }; |
835ab32df
|
1525 |
return do_execve_common(filename, argv, envp); |
0e028465d
|
1526 1527 1528 |
} #ifdef CONFIG_COMPAT |
d03d26e58
|
1529 |
static int compat_do_execve(const char *filename, |
38b983b34
|
1530 |
const compat_uptr_t __user *__argv, |
d03d26e58
|
1531 |
const compat_uptr_t __user *__envp) |
0e028465d
|
1532 1533 1534 1535 1536 1537 1538 1539 1540 |
{ struct user_arg_ptr argv = { .is_compat = true, .ptr.compat = __argv, }; struct user_arg_ptr envp = { .is_compat = true, .ptr.compat = __envp, }; |
835ab32df
|
1541 |
return do_execve_common(filename, argv, envp); |
ba2d01629
|
1542 |
} |
0e028465d
|
1543 |
#endif |
ba2d01629
|
1544 |
|
964ee7df9
|
1545 |
void set_binfmt(struct linux_binfmt *new) |
1da177e4c
|
1546 |
{ |
801460d0c
|
1547 1548 1549 1550 |
struct mm_struct *mm = current->mm; if (mm->binfmt) module_put(mm->binfmt->module); |
1da177e4c
|
1551 |
|
801460d0c
|
1552 |
mm->binfmt = new; |
964ee7df9
|
1553 1554 |
if (new) __module_get(new->module); |
1da177e4c
|
1555 1556 1557 |
} EXPORT_SYMBOL(set_binfmt); |
6c5d52382
|
1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 |
/* * set_dumpable converts traditional three-value dumpable to two flags and * stores them into mm->flags. It modifies lower two bits of mm->flags, but * these bits are not changed atomically. So get_dumpable can observe the * intermediate state. To avoid doing unexpected behavior, get get_dumpable * return either old dumpable or new one by paying attention to the order of * modifying the bits. * * dumpable | mm->flags (binary) * old new | initial interim final * ---------+----------------------- * 0 1 | 00 01 01 * 0 2 | 00 10(*) 11 * 1 0 | 01 00 00 * 1 2 | 01 11 11 * 2 0 | 11 10(*) 00 * 2 1 | 11 11 01 * * (*) get_dumpable regards interim value of 10 as 11. */ void set_dumpable(struct mm_struct *mm, int value) { switch (value) { |
54b501992
|
1581 |
case SUID_DUMPABLE_DISABLED: |
6c5d52382
|
1582 1583 1584 1585 |
clear_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; |
54b501992
|
1586 |
case SUID_DUMPABLE_ENABLED: |
6c5d52382
|
1587 1588 1589 1590 |
set_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; |
54b501992
|
1591 |
case SUID_DUMPABLE_SAFE: |
6c5d52382
|
1592 1593 1594 1595 1596 1597 |
set_bit(MMF_DUMP_SECURELY, &mm->flags); smp_wmb(); set_bit(MMF_DUMPABLE, &mm->flags); break; } } |
6c5d52382
|
1598 |
|
10c28d937
|
1599 |
int __get_dumpable(unsigned long mm_flags) |
6c5d52382
|
1600 1601 |
{ int ret; |
30736a4d4
|
1602 |
ret = mm_flags & MMF_DUMPABLE_MASK; |
54b501992
|
1603 |
return (ret > SUID_DUMPABLE_ENABLED) ? SUID_DUMPABLE_SAFE : ret; |
6c5d52382
|
1604 |
} |
30736a4d4
|
1605 1606 1607 1608 |
int get_dumpable(struct mm_struct *mm) { return __get_dumpable(mm->flags); } |
38b983b34
|
1609 1610 1611 1612 1613 |
SYSCALL_DEFINE3(execve, const char __user *, filename, const char __user *const __user *, argv, const char __user *const __user *, envp) { |
91a27b2a7
|
1614 |
struct filename *path = getname(filename); |
38b983b34
|
1615 1616 |
int error = PTR_ERR(path); if (!IS_ERR(path)) { |
da3d4c5fa
|
1617 |
error = do_execve(path->name, argv, envp); |
38b983b34
|
1618 1619 1620 1621 1622 1623 1624 1625 1626 |
putname(path); } return error; } #ifdef CONFIG_COMPAT asmlinkage long compat_sys_execve(const char __user * filename, const compat_uptr_t __user * argv, const compat_uptr_t __user * envp) { |
91a27b2a7
|
1627 |
struct filename *path = getname(filename); |
38b983b34
|
1628 1629 |
int error = PTR_ERR(path); if (!IS_ERR(path)) { |
d03d26e58
|
1630 |
error = compat_do_execve(path->name, argv, envp); |
38b983b34
|
1631 1632 1633 1634 1635 |
putname(path); } return error; } #endif |