Blame view
fs/exec.c
52.7 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> |
a6f76f23d
|
61 |
#include "internal.h" |
1da177e4c
|
62 |
|
1da177e4c
|
63 |
int core_uses_pid; |
71ce92f3f
|
64 |
char core_pattern[CORENAME_MAX_SIZE] = "core"; |
a293980c2
|
65 |
unsigned int core_pipe_limit; |
d6e711448
|
66 |
int suid_dumpable = 0; |
1b0d300bd
|
67 68 69 70 71 |
struct core_name { char *corename; int used, size; }; static atomic_t call_count = ATOMIC_INIT(1); |
1da177e4c
|
72 |
/* The maximal length of core_pattern is also specified in sysctl.c */ |
e4dc1b14d
|
73 |
static LIST_HEAD(formats); |
1da177e4c
|
74 |
static DEFINE_RWLOCK(binfmt_lock); |
74641f584
|
75 |
int __register_binfmt(struct linux_binfmt * fmt, int insert) |
1da177e4c
|
76 |
{ |
1da177e4c
|
77 78 |
if (!fmt) return -EINVAL; |
1da177e4c
|
79 |
write_lock(&binfmt_lock); |
74641f584
|
80 81 |
insert ? list_add(&fmt->lh, &formats) : list_add_tail(&fmt->lh, &formats); |
1da177e4c
|
82 83 84 |
write_unlock(&binfmt_lock); return 0; } |
74641f584
|
85 |
EXPORT_SYMBOL(__register_binfmt); |
1da177e4c
|
86 |
|
f6b450d48
|
87 |
void unregister_binfmt(struct linux_binfmt * fmt) |
1da177e4c
|
88 |
{ |
1da177e4c
|
89 |
write_lock(&binfmt_lock); |
e4dc1b14d
|
90 |
list_del(&fmt->lh); |
1da177e4c
|
91 |
write_unlock(&binfmt_lock); |
1da177e4c
|
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 |
} 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
|
107 |
SYSCALL_DEFINE1(uselib, const char __user *, library) |
1da177e4c
|
108 |
{ |
964bd1836
|
109 |
struct file *file; |
964bd1836
|
110 111 |
char *tmp = getname(library); int error = PTR_ERR(tmp); |
47c805dc2
|
112 113 114 115 116 |
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
|
117 |
|
6e8341a11
|
118 119 |
if (IS_ERR(tmp)) goto out; |
47c805dc2
|
120 |
file = do_filp_open(AT_FDCWD, tmp, &uselib_flags, LOOKUP_FOLLOW); |
6e8341a11
|
121 122 123 |
putname(tmp); error = PTR_ERR(file); if (IS_ERR(file)) |
1da177e4c
|
124 125 126 |
goto out; error = -EINVAL; |
6e8341a11
|
127 |
if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) |
1da177e4c
|
128 |
goto exit; |
30524472c
|
129 |
error = -EACCES; |
6e8341a11
|
130 |
if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) |
6146f0d5e
|
131 |
goto exit; |
1da177e4c
|
132 |
|
2a12a9d78
|
133 |
fsnotify_open(file); |
6110e3abb
|
134 |
|
1da177e4c
|
135 136 137 138 139 |
error = -ENOEXEC; if(file->f_op) { struct linux_binfmt * fmt; read_lock(&binfmt_lock); |
e4dc1b14d
|
140 |
list_for_each_entry(fmt, &formats, lh) { |
1da177e4c
|
141 142 143 144 145 146 147 148 149 150 151 152 153 |
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
|
154 |
exit: |
1da177e4c
|
155 156 157 |
fput(file); out: return error; |
1da177e4c
|
158 |
} |
b6a2fea39
|
159 |
#ifdef CONFIG_MMU |
ae6b585ee
|
160 161 162 163 164 165 |
/* * 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
|
166 |
static void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) |
3c77f8457
|
167 168 169 170 171 172 173 174 |
{ struct mm_struct *mm = current->mm; long diff = (long)(pages - bprm->vma_pages); if (!mm || !diff) return; bprm->vma_pages = pages; |
3c77f8457
|
175 |
add_mm_counter(mm, MM_ANONPAGES, diff); |
3c77f8457
|
176 |
} |
0e028465d
|
177 |
static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, |
b6a2fea39
|
178 179 180 181 182 183 184 |
int write) { struct page *page; int ret; #ifdef CONFIG_STACK_GROWSUP if (write) { |
d05f3169c
|
185 |
ret = expand_downwards(bprm->vma, pos); |
b6a2fea39
|
186 187 188 189 190 191 192 193 194 195 |
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
|
196 |
unsigned long size = bprm->vma->vm_end - bprm->vma->vm_start; |
a64e715fc
|
197 |
struct rlimit *rlim; |
3c77f8457
|
198 |
acct_arg_size(bprm, size / PAGE_SIZE); |
a64e715fc
|
199 200 201 202 203 204 |
/* * We've historically supported up to 32 pages (ARG_MAX) * of argument strings even with small stacks */ if (size <= ARG_MAX) return page; |
b6a2fea39
|
205 206 207 208 209 210 211 212 |
/* * 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
|
213 |
rlim = current->signal->rlim; |
d554ed895
|
214 |
if (size > ACCESS_ONCE(rlim[RLIMIT_STACK].rlim_cur) / 4) { |
b6a2fea39
|
215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 |
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
|
244 |
int err; |
b6a2fea39
|
245 246 247 248 249 |
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
|
250 |
return -ENOMEM; |
b6a2fea39
|
251 252 253 254 255 256 257 258 259 260 |
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
|
261 |
BUILD_BUG_ON(VM_STACK_FLAGS & VM_STACK_INCOMPLETE_SETUP); |
b6a2fea39
|
262 263 |
vma->vm_end = STACK_TOP_MAX; vma->vm_start = vma->vm_end - PAGE_SIZE; |
a8bef8ff6
|
264 |
vma->vm_flags = VM_STACK_FLAGS | VM_STACK_INCOMPLETE_SETUP; |
3ed75eb8f
|
265 |
vma->vm_page_prot = vm_get_page_prot(vma->vm_flags); |
5beb49305
|
266 |
INIT_LIST_HEAD(&vma->anon_vma_chain); |
462e635e5
|
267 268 269 270 |
err = security_file_mmap(NULL, 0, 0, 0, vma->vm_start, 1); if (err) goto err; |
b6a2fea39
|
271 |
err = insert_vm_struct(mm, vma); |
eaccbfa56
|
272 |
if (err) |
b6a2fea39
|
273 |
goto err; |
b6a2fea39
|
274 275 276 |
mm->stack_vm = mm->total_vm = 1; up_write(&mm->mmap_sem); |
b6a2fea39
|
277 |
bprm->p = vma->vm_end - sizeof(void *); |
b6a2fea39
|
278 |
return 0; |
b6a2fea39
|
279 |
err: |
eaccbfa56
|
280 281 282 |
up_write(&mm->mmap_sem); bprm->vma = NULL; kmem_cache_free(vm_area_cachep, vma); |
b6a2fea39
|
283 284 285 286 287 288 289 290 291 |
return err; } static bool valid_arg_len(struct linux_binprm *bprm, long len) { return len <= MAX_ARG_STRLEN; } #else |
0e028465d
|
292 |
static inline void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) |
3c77f8457
|
293 294 |
{ } |
0e028465d
|
295 |
static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, |
b6a2fea39
|
296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 |
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
|
383 |
struct user_arg_ptr { |
0e028465d
|
384 385 386 387 388 389 390 391 392 |
#ifdef CONFIG_COMPAT bool is_compat; #endif union { const char __user *const __user *native; #ifdef CONFIG_COMPAT compat_uptr_t __user *compat; #endif } ptr; |
ba2d01629
|
393 394 395 |
}; static const char __user *get_user_arg_ptr(struct user_arg_ptr argv, int nr) |
1d1dbf813
|
396 |
{ |
0e028465d
|
397 398 399 400 401 402 403 404 |
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
|
405 |
|
0e028465d
|
406 407 408 409 410 |
return compat_ptr(compat); } #endif if (get_user(native, argv.ptr.native + nr)) |
1d1dbf813
|
411 |
return ERR_PTR(-EFAULT); |
0e028465d
|
412 |
return native; |
1d1dbf813
|
413 |
} |
1da177e4c
|
414 415 416 |
/* * count() counts the number of strings in array ARGV. */ |
ba2d01629
|
417 |
static int count(struct user_arg_ptr argv, int max) |
1da177e4c
|
418 419 |
{ int i = 0; |
0e028465d
|
420 |
if (argv.ptr.native != NULL) { |
1da177e4c
|
421 |
for (;;) { |
1d1dbf813
|
422 |
const char __user *p = get_user_arg_ptr(argv, i); |
1da177e4c
|
423 |
|
1da177e4c
|
424 425 |
if (!p) break; |
1d1dbf813
|
426 427 428 |
if (IS_ERR(p)) return -EFAULT; |
362e6663e
|
429 |
if (i++ >= max) |
1da177e4c
|
430 |
return -E2BIG; |
9aea5a65a
|
431 432 433 |
if (fatal_signal_pending(current)) return -ERESTARTNOHAND; |
1da177e4c
|
434 435 436 437 438 439 440 |
cond_resched(); } } return i; } /* |
b6a2fea39
|
441 442 443 |
* '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
|
444 |
*/ |
ba2d01629
|
445 |
static int copy_strings(int argc, struct user_arg_ptr argv, |
75c96f858
|
446 |
struct linux_binprm *bprm) |
1da177e4c
|
447 448 449 |
{ struct page *kmapped_page = NULL; char *kaddr = NULL; |
b6a2fea39
|
450 |
unsigned long kpos = 0; |
1da177e4c
|
451 452 453 |
int ret; while (argc-- > 0) { |
d7627467b
|
454 |
const char __user *str; |
1da177e4c
|
455 456 |
int len; unsigned long pos; |
1d1dbf813
|
457 458 459 |
ret = -EFAULT; str = get_user_arg_ptr(argv, argc); if (IS_ERR(str)) |
1da177e4c
|
460 |
goto out; |
1da177e4c
|
461 |
|
1d1dbf813
|
462 463 464 465 466 467 |
len = strnlen_user(str, MAX_ARG_STRLEN); if (!len) goto out; ret = -E2BIG; if (!valid_arg_len(bprm, len)) |
1da177e4c
|
468 |
goto out; |
1da177e4c
|
469 |
|
b6a2fea39
|
470 |
/* We're going to work our way backwords. */ |
1da177e4c
|
471 |
pos = bprm->p; |
b6a2fea39
|
472 473 |
str += len; bprm->p -= len; |
1da177e4c
|
474 475 |
while (len > 0) { |
1da177e4c
|
476 |
int offset, bytes_to_copy; |
1da177e4c
|
477 |
|
9aea5a65a
|
478 479 480 481 |
if (fatal_signal_pending(current)) { ret = -ERESTARTNOHAND; goto out; } |
7993bc1f4
|
482 |
cond_resched(); |
1da177e4c
|
483 |
offset = pos % PAGE_SIZE; |
b6a2fea39
|
484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 |
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
|
500 |
if (!page) { |
b6a2fea39
|
501 |
ret = -E2BIG; |
1da177e4c
|
502 503 |
goto out; } |
1da177e4c
|
504 |
|
b6a2fea39
|
505 506 |
if (kmapped_page) { flush_kernel_dcache_page(kmapped_page); |
1da177e4c
|
507 |
kunmap(kmapped_page); |
b6a2fea39
|
508 509 |
put_arg_page(kmapped_page); } |
1da177e4c
|
510 511 |
kmapped_page = page; kaddr = kmap(kmapped_page); |
b6a2fea39
|
512 513 |
kpos = pos & PAGE_MASK; flush_arg_page(bprm, kpos, kmapped_page); |
1da177e4c
|
514 |
} |
b6a2fea39
|
515 |
if (copy_from_user(kaddr+offset, str, bytes_to_copy)) { |
1da177e4c
|
516 517 518 |
ret = -EFAULT; goto out; } |
1da177e4c
|
519 520 521 522 |
} } ret = 0; out: |
b6a2fea39
|
523 524 |
if (kmapped_page) { flush_kernel_dcache_page(kmapped_page); |
1da177e4c
|
525 |
kunmap(kmapped_page); |
b6a2fea39
|
526 527 |
put_arg_page(kmapped_page); } |
1da177e4c
|
528 529 530 531 532 533 |
return ret; } /* * Like copy_strings, but get argv and its values from kernel memory. */ |
ba2d01629
|
534 |
int copy_strings_kernel(int argc, const char *const *__argv, |
d7627467b
|
535 |
struct linux_binprm *bprm) |
1da177e4c
|
536 537 538 |
{ int r; mm_segment_t oldfs = get_fs(); |
ba2d01629
|
539 |
struct user_arg_ptr argv = { |
0e028465d
|
540 |
.ptr.native = (const char __user *const __user *)__argv, |
ba2d01629
|
541 |
}; |
1da177e4c
|
542 |
set_fs(KERNEL_DS); |
ba2d01629
|
543 |
r = copy_strings(argc, argv, bprm); |
1da177e4c
|
544 |
set_fs(oldfs); |
ba2d01629
|
545 |
|
1da177e4c
|
546 547 |
return r; } |
1da177e4c
|
548 549 550 |
EXPORT_SYMBOL(copy_strings_kernel); #ifdef CONFIG_MMU |
b6a2fea39
|
551 |
|
1da177e4c
|
552 |
/* |
b6a2fea39
|
553 554 555 |
* 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
|
556 |
* |
b6a2fea39
|
557 558 559 560 561 562 |
* 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
|
563 |
*/ |
b6a2fea39
|
564 |
static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift) |
1da177e4c
|
565 566 |
{ struct mm_struct *mm = vma->vm_mm; |
b6a2fea39
|
567 568 569 570 571 |
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
|
572 |
struct mmu_gather tlb; |
1da177e4c
|
573 |
|
b6a2fea39
|
574 |
BUG_ON(new_start > new_end); |
1da177e4c
|
575 |
|
b6a2fea39
|
576 577 578 579 580 581 582 583 584 585 |
/* * 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
|
586 587 |
if (vma_adjust(vma, new_start, old_end, vma->vm_pgoff, NULL)) return -ENOMEM; |
b6a2fea39
|
588 589 590 591 592 593 594 595 596 597 |
/* * 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, vma, new_start, length)) return -ENOMEM; lru_add_drain(); |
d16dfc550
|
598 |
tlb_gather_mmu(&tlb, mm, 0); |
b6a2fea39
|
599 600 601 602 |
if (new_end > old_start) { /* * when the old and new regions overlap clear from new_end. */ |
d16dfc550
|
603 |
free_pgd_range(&tlb, new_end, old_end, new_end, |
b6a2fea39
|
604 605 606 607 608 609 610 611 |
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
|
612 |
free_pgd_range(&tlb, old_start, old_end, new_end, |
b6a2fea39
|
613 |
vma->vm_next ? vma->vm_next->vm_start : 0); |
1da177e4c
|
614 |
} |
d16dfc550
|
615 |
tlb_finish_mmu(&tlb, new_end, old_end); |
b6a2fea39
|
616 617 |
/* |
5beb49305
|
618 |
* Shrink the vma to just the new range. Always succeeds. |
b6a2fea39
|
619 620 621 622 |
*/ vma_adjust(vma, new_start, new_end, vma->vm_pgoff, NULL); return 0; |
1da177e4c
|
623 |
} |
b6a2fea39
|
624 625 626 627 |
/* * Finalizes the stack vm_area_struct. The flags and permissions are updated, * the stack is optionally relocated, and some extra space is added. */ |
1da177e4c
|
628 629 630 631 |
int setup_arg_pages(struct linux_binprm *bprm, unsigned long stack_top, int executable_stack) { |
b6a2fea39
|
632 633 |
unsigned long ret; unsigned long stack_shift; |
1da177e4c
|
634 |
struct mm_struct *mm = current->mm; |
b6a2fea39
|
635 636 637 638 |
struct vm_area_struct *vma = bprm->vma; struct vm_area_struct *prev = NULL; unsigned long vm_flags; unsigned long stack_base; |
803bf5ec2
|
639 640 641 |
unsigned long stack_size; unsigned long stack_expand; unsigned long rlim_stack; |
1da177e4c
|
642 643 |
#ifdef CONFIG_STACK_GROWSUP |
1da177e4c
|
644 |
/* Limit stack size to 1GB */ |
d554ed895
|
645 |
stack_base = rlimit_max(RLIMIT_STACK); |
1da177e4c
|
646 647 |
if (stack_base > (1 << 30)) stack_base = 1 << 30; |
1da177e4c
|
648 |
|
b6a2fea39
|
649 650 651 |
/* Make sure we didn't let the argument array grow too large. */ if (vma->vm_end - vma->vm_start > stack_base) return -ENOMEM; |
1da177e4c
|
652 |
|
b6a2fea39
|
653 |
stack_base = PAGE_ALIGN(stack_top - stack_base); |
1da177e4c
|
654 |
|
b6a2fea39
|
655 656 657 |
stack_shift = vma->vm_start - stack_base; mm->arg_start = bprm->p - stack_shift; bprm->p = vma->vm_end - stack_shift; |
1da177e4c
|
658 |
#else |
b6a2fea39
|
659 660 |
stack_top = arch_align_stack(stack_top); stack_top = PAGE_ALIGN(stack_top); |
1b528181b
|
661 662 663 664 |
if (unlikely(stack_top < mmap_min_addr) || unlikely(vma->vm_end - vma->vm_start >= stack_top - mmap_min_addr)) return -ENOMEM; |
b6a2fea39
|
665 666 667 |
stack_shift = vma->vm_end - stack_top; bprm->p -= stack_shift; |
1da177e4c
|
668 |
mm->arg_start = bprm->p; |
1da177e4c
|
669 |
#endif |
1da177e4c
|
670 |
if (bprm->loader) |
b6a2fea39
|
671 672 |
bprm->loader -= stack_shift; bprm->exec -= stack_shift; |
1da177e4c
|
673 |
|
1da177e4c
|
674 |
down_write(&mm->mmap_sem); |
96a8e13ed
|
675 |
vm_flags = VM_STACK_FLAGS; |
b6a2fea39
|
676 677 678 679 680 681 682 683 684 685 686 |
/* * 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
|
687 |
vm_flags |= VM_STACK_INCOMPLETE_SETUP; |
b6a2fea39
|
688 689 690 691 692 693 694 695 696 697 |
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
|
698 699 |
if (ret) goto out_unlock; |
1da177e4c
|
700 |
} |
a8bef8ff6
|
701 702 |
/* mprotect_fixup is overkill to remove the temporary stack flags */ vma->vm_flags &= ~VM_STACK_INCOMPLETE_SETUP; |
5ef097dd7
|
703 |
stack_expand = 131072UL; /* randomly 32*4k (or 2*64k) pages */ |
803bf5ec2
|
704 705 706 707 708 709 |
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
|
710 |
#ifdef CONFIG_STACK_GROWSUP |
803bf5ec2
|
711 712 713 714 |
if (stack_size + stack_expand > rlim_stack) stack_base = vma->vm_start + rlim_stack; else stack_base = vma->vm_end + stack_expand; |
b6a2fea39
|
715 |
#else |
803bf5ec2
|
716 717 718 719 |
if (stack_size + stack_expand > rlim_stack) stack_base = vma->vm_end - rlim_stack; else stack_base = vma->vm_start - stack_expand; |
b6a2fea39
|
720 |
#endif |
3af9e8592
|
721 |
current->mm->start_stack = bprm->p; |
b6a2fea39
|
722 723 724 725 726 |
ret = expand_stack(vma, stack_base); if (ret) ret = -EFAULT; out_unlock: |
1da177e4c
|
727 |
up_write(&mm->mmap_sem); |
fc63cf237
|
728 |
return ret; |
1da177e4c
|
729 |
} |
1da177e4c
|
730 |
EXPORT_SYMBOL(setup_arg_pages); |
1da177e4c
|
731 732 733 734 |
#endif /* CONFIG_MMU */ struct file *open_exec(const char *name) { |
1da177e4c
|
735 |
struct file *file; |
e56b6a5dd
|
736 |
int err; |
47c805dc2
|
737 738 739 740 741 |
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
|
742 |
|
47c805dc2
|
743 |
file = do_filp_open(AT_FDCWD, name, &open_exec_flags, LOOKUP_FOLLOW); |
6e8341a11
|
744 |
if (IS_ERR(file)) |
e56b6a5dd
|
745 746 747 |
goto out; err = -EACCES; |
6e8341a11
|
748 749 |
if (!S_ISREG(file->f_path.dentry->d_inode->i_mode)) goto exit; |
e56b6a5dd
|
750 |
|
6e8341a11
|
751 752 |
if (file->f_path.mnt->mnt_flags & MNT_NOEXEC) goto exit; |
e56b6a5dd
|
753 |
|
2a12a9d78
|
754 |
fsnotify_open(file); |
6110e3abb
|
755 |
|
e56b6a5dd
|
756 |
err = deny_write_access(file); |
6e8341a11
|
757 758 |
if (err) goto exit; |
1da177e4c
|
759 |
|
6e8341a11
|
760 |
out: |
e56b6a5dd
|
761 |
return file; |
6e8341a11
|
762 763 |
exit: fput(file); |
e56b6a5dd
|
764 765 |
return ERR_PTR(err); } |
1da177e4c
|
766 |
EXPORT_SYMBOL(open_exec); |
6777d773a
|
767 768 |
int kernel_read(struct file *file, loff_t offset, char *addr, unsigned long count) |
1da177e4c
|
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 |
{ 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; |
34e55232e
|
792 |
sync_mm_rss(tsk, old_mm); |
1da177e4c
|
793 794 795 796 797 798 799 |
mm_release(tsk, old_mm); if (old_mm) { /* * 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
|
800 |
* checking core_state and changing tsk->mm. |
1da177e4c
|
801 802 |
*/ down_read(&old_mm->mmap_sem); |
999d9fc16
|
803 |
if (unlikely(old_mm->core_state)) { |
1da177e4c
|
804 805 806 807 808 809 810 811 812 |
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); |
3d5992d2a
|
813 814 815 816 |
if (old_mm && tsk->signal->oom_score_adj == OOM_SCORE_ADJ_MIN) { atomic_dec(&old_mm->oom_disable_count); atomic_inc(&tsk->mm->oom_disable_count); } |
1da177e4c
|
817 818 819 820 |
task_unlock(tsk); arch_pick_mmap_layout(mm); if (old_mm) { up_read(&old_mm->mmap_sem); |
7dddb12c6
|
821 |
BUG_ON(active_mm != old_mm); |
31a78f23b
|
822 |
mm_update_next_owner(old_mm); |
1da177e4c
|
823 824 825 826 827 828 829 830 831 832 833 834 835 |
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
|
836 |
static int de_thread(struct task_struct *tsk) |
1da177e4c
|
837 838 |
{ struct signal_struct *sig = tsk->signal; |
b2c903b87
|
839 |
struct sighand_struct *oldsighand = tsk->sighand; |
1da177e4c
|
840 |
spinlock_t *lock = &oldsighand->siglock; |
1da177e4c
|
841 |
|
aafe6c2a2
|
842 |
if (thread_group_empty(tsk)) |
1da177e4c
|
843 844 845 846 |
goto no_thread_group; /* * Kill all other threads in the thread group. |
1da177e4c
|
847 |
*/ |
1da177e4c
|
848 |
spin_lock_irq(lock); |
ed5d2cac1
|
849 |
if (signal_group_exit(sig)) { |
1da177e4c
|
850 851 852 853 854 |
/* * Another group action in progress, just * return so that the signal is processed. */ spin_unlock_irq(lock); |
1da177e4c
|
855 856 |
return -EAGAIN; } |
d344193a0
|
857 |
|
ed5d2cac1
|
858 |
sig->group_exit_task = tsk; |
d344193a0
|
859 860 861 |
sig->notify_count = zap_other_threads(tsk); if (!thread_group_leader(tsk)) sig->notify_count--; |
1da177e4c
|
862 |
|
d344193a0
|
863 |
while (sig->notify_count) { |
1da177e4c
|
864 865 866 867 868 |
__set_current_state(TASK_UNINTERRUPTIBLE); spin_unlock_irq(lock); schedule(); spin_lock_irq(lock); } |
1da177e4c
|
869 870 871 872 873 874 875 |
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
|
876 |
if (!thread_group_leader(tsk)) { |
8187926bd
|
877 |
struct task_struct *leader = tsk->group_leader; |
6db840fa7
|
878 |
|
2800d8d19
|
879 |
sig->notify_count = -1; /* for exit_notify() */ |
6db840fa7
|
880 881 882 883 884 885 886 887 |
for (;;) { write_lock_irq(&tasklist_lock); if (likely(leader->exit_state)) break; __set_current_state(TASK_UNINTERRUPTIBLE); write_unlock_irq(&tasklist_lock); schedule(); } |
1da177e4c
|
888 |
|
f5e902817
|
889 890 891 892 893 894 895 896 897 898 |
/* * 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
|
899 |
tsk->start_time = leader->start_time; |
f5e902817
|
900 |
|
bac0abd61
|
901 902 |
BUG_ON(!same_thread_group(leader, tsk)); BUG_ON(has_group_leader_pid(tsk)); |
1da177e4c
|
903 904 905 906 907 908 |
/* * 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
|
909 910 |
/* Become a process group leader with the old leader's pid. |
c18258c6f
|
911 912 |
* The old leader becomes a thread of the this thread group. * Note: The old leader also uses this pid until release_task |
d73d65293
|
913 914 |
* is called. Odd but simple and correct. */ |
aafe6c2a2
|
915 916 |
detach_pid(tsk, PIDTYPE_PID); tsk->pid = leader->pid; |
3743ca05f
|
917 |
attach_pid(tsk, PIDTYPE_PID, task_pid(leader)); |
aafe6c2a2
|
918 919 |
transfer_pid(leader, tsk, PIDTYPE_PGID); transfer_pid(leader, tsk, PIDTYPE_SID); |
9cd80bbb0
|
920 |
|
aafe6c2a2
|
921 |
list_replace_rcu(&leader->tasks, &tsk->tasks); |
9cd80bbb0
|
922 |
list_replace_init(&leader->sibling, &tsk->sibling); |
1da177e4c
|
923 |
|
aafe6c2a2
|
924 925 |
tsk->group_leader = tsk; leader->group_leader = tsk; |
de12a7878
|
926 |
|
aafe6c2a2
|
927 |
tsk->exit_signal = SIGCHLD; |
087806b12
|
928 |
leader->exit_signal = -1; |
962b564cf
|
929 930 931 |
BUG_ON(leader->exit_state != EXIT_ZOMBIE); leader->exit_state = EXIT_DEAD; |
eac1b5e57
|
932 933 934 935 936 937 938 939 |
/* * 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
|
940 |
write_unlock_irq(&tasklist_lock); |
8187926bd
|
941 942 |
release_task(leader); |
ed5d2cac1
|
943 |
} |
1da177e4c
|
944 |
|
6db840fa7
|
945 946 |
sig->group_exit_task = NULL; sig->notify_count = 0; |
1da177e4c
|
947 948 |
no_thread_group: |
1f10206cf
|
949 950 |
if (current->mm) setmax_mm_hiwater_rss(&sig->maxrss, current->mm); |
1da177e4c
|
951 |
exit_itimers(sig); |
cbaffba12
|
952 |
flush_itimer_signals(); |
329f7dba5
|
953 |
|
b2c903b87
|
954 955 |
if (atomic_read(&oldsighand->count) != 1) { struct sighand_struct *newsighand; |
1da177e4c
|
956 |
/* |
b2c903b87
|
957 958 |
* This ->sighand is shared with the CLONE_SIGHAND * but not CLONE_THREAD task, switch to the new one. |
1da177e4c
|
959 |
*/ |
b2c903b87
|
960 961 962 |
newsighand = kmem_cache_alloc(sighand_cachep, GFP_KERNEL); if (!newsighand) return -ENOMEM; |
1da177e4c
|
963 964 965 966 967 968 |
atomic_set(&newsighand->count, 1); memcpy(newsighand->action, oldsighand->action, sizeof(newsighand->action)); write_lock_irq(&tasklist_lock); spin_lock(&oldsighand->siglock); |
aafe6c2a2
|
969 |
rcu_assign_pointer(tsk->sighand, newsighand); |
1da177e4c
|
970 971 |
spin_unlock(&oldsighand->siglock); write_unlock_irq(&tasklist_lock); |
fba2afaae
|
972 |
__cleanup_sighand(oldsighand); |
1da177e4c
|
973 |
} |
aafe6c2a2
|
974 |
BUG_ON(!thread_group_leader(tsk)); |
1da177e4c
|
975 976 |
return 0; } |
0840a90d9
|
977 |
|
1da177e4c
|
978 979 980 981 |
/* * These functions flushes out all traces of the currently running executable * so that a new one can be started */ |
858119e15
|
982 |
static void flush_old_files(struct files_struct * files) |
1da177e4c
|
983 984 |
{ long j = -1; |
badf16621
|
985 |
struct fdtable *fdt; |
1da177e4c
|
986 987 988 989 990 991 992 |
spin_lock(&files->file_lock); for (;;) { unsigned long set, i; j++; i = j * __NFDBITS; |
badf16621
|
993 |
fdt = files_fdtable(files); |
bbea9f696
|
994 |
if (i >= fdt->max_fds) |
1da177e4c
|
995 |
break; |
badf16621
|
996 |
set = fdt->close_on_exec->fds_bits[j]; |
1da177e4c
|
997 998 |
if (!set) continue; |
badf16621
|
999 |
fdt->close_on_exec->fds_bits[j] = 0; |
1da177e4c
|
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 |
spin_unlock(&files->file_lock); for ( ; set ; i++,set >>= 1) { if (set & 1) { sys_close(i); } } spin_lock(&files->file_lock); } spin_unlock(&files->file_lock); } |
59714d65d
|
1011 |
char *get_task_comm(char *buf, struct task_struct *tsk) |
1da177e4c
|
1012 1013 1014 1015 1016 |
{ /* buf must be at least sizeof(tsk->comm) in size */ task_lock(tsk); strncpy(buf, tsk->comm, sizeof(tsk->comm)); task_unlock(tsk); |
59714d65d
|
1017 |
return buf; |
1da177e4c
|
1018 |
} |
7d74f492e
|
1019 |
EXPORT_SYMBOL_GPL(get_task_comm); |
1da177e4c
|
1020 1021 1022 1023 |
void set_task_comm(struct task_struct *tsk, char *buf) { task_lock(tsk); |
4614a696b
|
1024 1025 1026 1027 1028 1029 1030 1031 1032 |
/* * 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
|
1033 1034 |
strlcpy(tsk->comm, buf, sizeof(tsk->comm)); task_unlock(tsk); |
cdd6c482c
|
1035 |
perf_event_comm(tsk); |
1da177e4c
|
1036 1037 1038 1039 |
} int flush_old_exec(struct linux_binprm * bprm) { |
221af7f87
|
1040 |
int retval; |
1da177e4c
|
1041 1042 1043 1044 1045 1046 1047 1048 |
/* * 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
|
1049 |
set_mm_exe_file(bprm->mm, bprm->file); |
1da177e4c
|
1050 |
/* |
1da177e4c
|
1051 1052 |
* Release all of the old mmap stuff */ |
3c77f8457
|
1053 |
acct_arg_size(bprm, 0); |
1da177e4c
|
1054 1055 |
retval = exec_mmap(bprm->mm); if (retval) |
fd8328be8
|
1056 |
goto out; |
1da177e4c
|
1057 1058 |
bprm->mm = NULL; /* We're using it now */ |
7ab02af42
|
1059 |
|
dac853ae8
|
1060 |
set_fs(USER_DS); |
98391cf4d
|
1061 |
current->flags &= ~(PF_RANDOMIZE | PF_KTHREAD); |
7ab02af42
|
1062 1063 |
flush_thread(); current->personality &= ~bprm->per_clear; |
221af7f87
|
1064 1065 1066 1067 1068 1069 |
return 0; out: return retval; } EXPORT_SYMBOL(flush_old_exec); |
1b5d783c9
|
1070 1071 1072 1073 1074 1075 |
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
|
1076 1077 1078 |
void setup_new_exec(struct linux_binprm * bprm) { int i, ch; |
d7627467b
|
1079 |
const char *name; |
221af7f87
|
1080 1081 1082 |
char tcomm[sizeof(current->comm)]; arch_pick_mmap_layout(current->mm); |
1da177e4c
|
1083 1084 |
/* This is the point of no return */ |
1da177e4c
|
1085 |
current->sas_ss_sp = current->sas_ss_size = 0; |
da9592ede
|
1086 |
if (current_euid() == current_uid() && current_egid() == current_gid()) |
6c5d52382
|
1087 |
set_dumpable(current->mm, 1); |
d6e711448
|
1088 |
else |
6c5d52382
|
1089 |
set_dumpable(current->mm, suid_dumpable); |
d6e711448
|
1090 |
|
1da177e4c
|
1091 |
name = bprm->filename; |
367720923
|
1092 1093 |
/* Copies the binary name from after last slash */ |
1da177e4c
|
1094 1095 |
for (i=0; (ch = *(name++)) != '\0';) { if (ch == '/') |
367720923
|
1096 |
i = 0; /* overwrite what we wrote */ |
1da177e4c
|
1097 1098 1099 1100 1101 1102 |
else if (i < (sizeof(tcomm) - 1)) tcomm[i++] = ch; } tcomm[i] = '\0'; set_task_comm(current, tcomm); |
0551fbd29
|
1103 1104 1105 1106 1107 |
/* 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
|
1108 1109 1110 |
/* install the new credentials */ if (bprm->cred->uid != current_euid() || bprm->cred->gid != current_egid()) { |
d2d56c5f5
|
1111 |
current->pdeath_signal = 0; |
1b5d783c9
|
1112 1113 1114 1115 |
} else { would_dump(bprm, bprm->file); if (bprm->interp_flags & BINPRM_FLAGS_ENFORCE_NONDUMP) set_dumpable(current->mm, suid_dumpable); |
1da177e4c
|
1116 |
} |
f65cb45cb
|
1117 1118 1119 1120 1121 |
/* * Flush performance counters when crossing a * security domain: */ if (!get_dumpable(current->mm)) |
cdd6c482c
|
1122 |
perf_event_exit_task(current); |
f65cb45cb
|
1123 |
|
1da177e4c
|
1124 1125 1126 1127 1128 1129 1130 |
/* An exec changes our domain. We are no longer part of the thread group */ current->self_exec_id++; flush_signal_handlers(current, 0); flush_old_files(current->files); |
1da177e4c
|
1131 |
} |
221af7f87
|
1132 |
EXPORT_SYMBOL(setup_new_exec); |
1da177e4c
|
1133 |
|
a6f76f23d
|
1134 |
/* |
a2a8474c3
|
1135 1136 1137 1138 1139 1140 1141 |
* 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
|
1142 |
if (mutex_lock_interruptible(¤t->signal->cred_guard_mutex)) |
a2a8474c3
|
1143 1144 1145 1146 1147 |
return -ERESTARTNOINTR; bprm->cred = prepare_exec_creds(); if (likely(bprm->cred)) return 0; |
9b1bf12d5
|
1148 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a2a8474c3
|
1149 1150 1151 1152 1153 1154 1155 |
return -ENOMEM; } void free_bprm(struct linux_binprm *bprm) { free_arg_pages(bprm); if (bprm->cred) { |
9b1bf12d5
|
1156 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a2a8474c3
|
1157 1158 1159 1160 1161 1162 |
abort_creds(bprm->cred); } kfree(bprm); } /* |
a6f76f23d
|
1163 1164 1165 1166 1167 1168 1169 1170 |
* 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
|
1171 1172 |
/* * cred_guard_mutex must be held at least to this point to prevent |
a6f76f23d
|
1173 |
* ptrace_attach() from altering our determination of the task's |
a2a8474c3
|
1174 1175 |
* credentials; any time after this it may be unlocked. */ |
a6f76f23d
|
1176 |
security_bprm_committed_creds(bprm); |
9b1bf12d5
|
1177 |
mutex_unlock(¤t->signal->cred_guard_mutex); |
a6f76f23d
|
1178 1179 1180 1181 1182 |
} EXPORT_SYMBOL(install_exec_creds); /* * determine how safe it is to execute the proposed program |
9b1bf12d5
|
1183 |
* - the caller must hold ->cred_guard_mutex to protect against |
a6f76f23d
|
1184 1185 |
* PTRACE_ATTACH */ |
498052bba
|
1186 |
int check_unsafe_exec(struct linux_binprm *bprm) |
a6f76f23d
|
1187 |
{ |
0bf2f3aec
|
1188 |
struct task_struct *p = current, *t; |
f1191b50e
|
1189 |
unsigned n_fs; |
498052bba
|
1190 |
int res = 0; |
a6f76f23d
|
1191 |
|
4b9d33e6d
|
1192 1193 1194 1195 1196 1197 |
if (p->ptrace) { if (p->ptrace & PT_PTRACE_CAP) bprm->unsafe |= LSM_UNSAFE_PTRACE_CAP; else bprm->unsafe |= LSM_UNSAFE_PTRACE; } |
a6f76f23d
|
1198 |
|
0bf2f3aec
|
1199 |
n_fs = 1; |
2a4419b5b
|
1200 |
spin_lock(&p->fs->lock); |
437f7fdb6
|
1201 |
rcu_read_lock(); |
0bf2f3aec
|
1202 1203 1204 |
for (t = next_thread(p); t != p; t = next_thread(t)) { if (t->fs == p->fs) n_fs++; |
0bf2f3aec
|
1205 |
} |
437f7fdb6
|
1206 |
rcu_read_unlock(); |
0bf2f3aec
|
1207 |
|
f1191b50e
|
1208 |
if (p->fs->users > n_fs) { |
a6f76f23d
|
1209 |
bprm->unsafe |= LSM_UNSAFE_SHARE; |
498052bba
|
1210 |
} else { |
8c652f96d
|
1211 1212 1213 1214 1215 |
res = -EAGAIN; if (!p->fs->in_exec) { p->fs->in_exec = 1; res = 1; } |
498052bba
|
1216 |
} |
2a4419b5b
|
1217 |
spin_unlock(&p->fs->lock); |
498052bba
|
1218 1219 |
return res; |
a6f76f23d
|
1220 |
} |
1da177e4c
|
1221 1222 1223 |
/* * Fill the binprm structure from the inode. * Check permissions, then read the first 128 (BINPRM_BUF_SIZE) bytes |
a6f76f23d
|
1224 1225 |
* * This may be called multiple times for binary chains (scripts for example). |
1da177e4c
|
1226 1227 1228 |
*/ int prepare_binprm(struct linux_binprm *bprm) { |
a6f76f23d
|
1229 |
umode_t mode; |
0f7fc9e4d
|
1230 |
struct inode * inode = bprm->file->f_path.dentry->d_inode; |
1da177e4c
|
1231 1232 1233 |
int retval; mode = inode->i_mode; |
1da177e4c
|
1234 1235 |
if (bprm->file->f_op == NULL) return -EACCES; |
a6f76f23d
|
1236 1237 1238 |
/* clear any previous set[ug]id data from a previous binary */ bprm->cred->euid = current_euid(); bprm->cred->egid = current_egid(); |
1da177e4c
|
1239 |
|
a6f76f23d
|
1240 |
if (!(bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)) { |
1da177e4c
|
1241 1242 |
/* Set-uid? */ if (mode & S_ISUID) { |
a6f76f23d
|
1243 1244 |
bprm->per_clear |= PER_CLEAR_ON_SETID; bprm->cred->euid = inode->i_uid; |
1da177e4c
|
1245 1246 1247 1248 1249 1250 1251 1252 1253 |
} /* 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
|
1254 1255 |
bprm->per_clear |= PER_CLEAR_ON_SETID; bprm->cred->egid = inode->i_gid; |
1da177e4c
|
1256 1257 1258 1259 |
} } /* fill in binprm security blob */ |
a6f76f23d
|
1260 |
retval = security_bprm_set_creds(bprm); |
1da177e4c
|
1261 1262 |
if (retval) return retval; |
a6f76f23d
|
1263 |
bprm->cred_prepared = 1; |
1da177e4c
|
1264 |
|
a6f76f23d
|
1265 1266 |
memset(bprm->buf, 0, BINPRM_BUF_SIZE); return kernel_read(bprm->file, 0, bprm->buf, BINPRM_BUF_SIZE); |
1da177e4c
|
1267 1268 1269 |
} EXPORT_SYMBOL(prepare_binprm); |
4fc75ff48
|
1270 1271 1272 1273 1274 |
/* * 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
|
1275 |
int remove_arg_zero(struct linux_binprm *bprm) |
1da177e4c
|
1276 |
{ |
b6a2fea39
|
1277 1278 1279 1280 |
int ret = 0; unsigned long offset; char *kaddr; struct page *page; |
4fc75ff48
|
1281 |
|
b6a2fea39
|
1282 1283 |
if (!bprm->argc) return 0; |
1da177e4c
|
1284 |
|
b6a2fea39
|
1285 1286 1287 1288 1289 1290 1291 1292 |
do { offset = bprm->p & ~PAGE_MASK; page = get_arg_page(bprm, bprm->p, 0); if (!page) { ret = -EFAULT; goto out; } kaddr = kmap_atomic(page, KM_USER0); |
4fc75ff48
|
1293 |
|
b6a2fea39
|
1294 1295 1296 |
for (; offset < PAGE_SIZE && kaddr[offset]; offset++, bprm->p++) ; |
4fc75ff48
|
1297 |
|
b6a2fea39
|
1298 1299 |
kunmap_atomic(kaddr, KM_USER0); put_arg_page(page); |
4fc75ff48
|
1300 |
|
b6a2fea39
|
1301 1302 1303 |
if (offset == PAGE_SIZE) free_arg_page(bprm, (bprm->p >> PAGE_SHIFT) - 1); } while (offset == PAGE_SIZE); |
4fc75ff48
|
1304 |
|
b6a2fea39
|
1305 1306 1307 |
bprm->p++; bprm->argc--; ret = 0; |
4fc75ff48
|
1308 |
|
b6a2fea39
|
1309 1310 |
out: return ret; |
1da177e4c
|
1311 |
} |
1da177e4c
|
1312 1313 1314 1315 1316 1317 1318 |
EXPORT_SYMBOL(remove_arg_zero); /* * cycle the list of binary formats handler, until one recognizes the image */ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) { |
85f334666
|
1319 |
unsigned int depth = bprm->recursion_depth; |
1da177e4c
|
1320 1321 |
int try,retval; struct linux_binfmt *fmt; |
bb188d7e6
|
1322 |
pid_t old_pid; |
1da177e4c
|
1323 |
|
1da177e4c
|
1324 1325 1326 |
retval = security_bprm_check(bprm); if (retval) return retval; |
473ae30bc
|
1327 1328 1329 |
retval = audit_bprm(bprm); if (retval) return retval; |
bb188d7e6
|
1330 1331 1332 1333 |
/* Need to fetch pid before load_binary changes it */ rcu_read_lock(); old_pid = task_pid_nr_ns(current, task_active_pid_ns(current->parent)); rcu_read_unlock(); |
1da177e4c
|
1334 1335 1336 |
retval = -ENOENT; for (try=0; try<2; try++) { read_lock(&binfmt_lock); |
e4dc1b14d
|
1337 |
list_for_each_entry(fmt, &formats, lh) { |
1da177e4c
|
1338 1339 1340 1341 1342 1343 1344 |
int (*fn)(struct linux_binprm *, struct pt_regs *) = fmt->load_binary; if (!fn) continue; if (!try_module_get(fmt->module)) continue; read_unlock(&binfmt_lock); retval = fn(bprm, regs); |
85f334666
|
1345 1346 1347 1348 1349 1350 |
/* * Restore the depth counter to its starting value * in this call, so we don't have to rely on every * load_binary function to restore it on return. */ bprm->recursion_depth = depth; |
1da177e4c
|
1351 |
if (retval >= 0) { |
85f334666
|
1352 |
if (depth == 0) |
bb188d7e6
|
1353 1354 |
ptrace_event(PTRACE_EVENT_EXEC, old_pid); |
1da177e4c
|
1355 1356 1357 1358 1359 1360 |
put_binfmt(fmt); allow_write_access(bprm->file); if (bprm->file) fput(bprm->file); bprm->file = NULL; current->did_exec = 1; |
9f46080c4
|
1361 |
proc_exec_connector(current); |
1da177e4c
|
1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 |
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
|
1374 |
#ifdef CONFIG_MODULES |
1da177e4c
|
1375 1376 |
if (retval != -ENOEXEC || bprm->mm == NULL) { break; |
5f4123be3
|
1377 |
} else { |
1da177e4c
|
1378 1379 1380 1381 1382 1383 1384 |
#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
|
1385 1386 |
if (try) break; /* -ENOEXEC */ |
1da177e4c
|
1387 |
request_module("binfmt-%04x", *(unsigned short *)(&bprm->buf[2])); |
1da177e4c
|
1388 |
} |
b4edf8bd0
|
1389 1390 1391 |
#else break; #endif |
1da177e4c
|
1392 1393 1394 1395 1396 1397 1398 1399 1400 |
} return retval; } EXPORT_SYMBOL(search_binary_handler); /* * sys_execve() executes a new program. */ |
ba2d01629
|
1401 1402 1403 1404 |
static int do_execve_common(const char *filename, struct user_arg_ptr argv, struct user_arg_ptr envp, struct pt_regs *regs) |
1da177e4c
|
1405 1406 1407 |
{ struct linux_binprm *bprm; struct file *file; |
3b1253880
|
1408 |
struct files_struct *displaced; |
8c652f96d
|
1409 |
bool clear_in_exec; |
1da177e4c
|
1410 |
int retval; |
72fa59970
|
1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 |
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
|
1428 |
|
3b1253880
|
1429 |
retval = unshare_files(&displaced); |
fd8328be8
|
1430 1431 |
if (retval) goto out_ret; |
1da177e4c
|
1432 |
retval = -ENOMEM; |
11b0b5abb
|
1433 |
bprm = kzalloc(sizeof(*bprm), GFP_KERNEL); |
1da177e4c
|
1434 |
if (!bprm) |
fd8328be8
|
1435 |
goto out_files; |
1da177e4c
|
1436 |
|
a2a8474c3
|
1437 1438 |
retval = prepare_bprm_creds(bprm); if (retval) |
a6f76f23d
|
1439 |
goto out_free; |
498052bba
|
1440 1441 |
retval = check_unsafe_exec(bprm); |
8c652f96d
|
1442 |
if (retval < 0) |
a2a8474c3
|
1443 |
goto out_free; |
8c652f96d
|
1444 |
clear_in_exec = retval; |
a2a8474c3
|
1445 |
current->in_execve = 1; |
a6f76f23d
|
1446 |
|
1da177e4c
|
1447 1448 1449 |
file = open_exec(filename); retval = PTR_ERR(file); if (IS_ERR(file)) |
498052bba
|
1450 |
goto out_unmark; |
1da177e4c
|
1451 1452 |
sched_exec(); |
1da177e4c
|
1453 1454 1455 |
bprm->file = file; bprm->filename = filename; bprm->interp = filename; |
1da177e4c
|
1456 |
|
b6a2fea39
|
1457 1458 1459 |
retval = bprm_mm_init(bprm); if (retval) goto out_file; |
1da177e4c
|
1460 |
|
b6a2fea39
|
1461 |
bprm->argc = count(argv, MAX_ARG_STRINGS); |
1da177e4c
|
1462 |
if ((retval = bprm->argc) < 0) |
a6f76f23d
|
1463 |
goto out; |
1da177e4c
|
1464 |
|
b6a2fea39
|
1465 |
bprm->envc = count(envp, MAX_ARG_STRINGS); |
1da177e4c
|
1466 |
if ((retval = bprm->envc) < 0) |
1da177e4c
|
1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 |
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; retval = search_binary_handler(bprm,regs); |
a6f76f23d
|
1487 1488 |
if (retval < 0) goto out; |
1da177e4c
|
1489 |
|
a6f76f23d
|
1490 |
/* execve succeeded */ |
498052bba
|
1491 |
current->fs->in_exec = 0; |
f9ce1f1cd
|
1492 |
current->in_execve = 0; |
a6f76f23d
|
1493 1494 1495 1496 1497 |
acct_update_integrals(current); free_bprm(bprm); if (displaced) put_files_struct(displaced); return retval; |
1da177e4c
|
1498 |
|
a6f76f23d
|
1499 |
out: |
3c77f8457
|
1500 1501 1502 1503 |
if (bprm->mm) { acct_arg_size(bprm, 0); mmput(bprm->mm); } |
1da177e4c
|
1504 1505 1506 1507 1508 1509 |
out_file: if (bprm->file) { allow_write_access(bprm->file); fput(bprm->file); } |
a6f76f23d
|
1510 |
|
498052bba
|
1511 |
out_unmark: |
8c652f96d
|
1512 1513 |
if (clear_in_exec) current->fs->in_exec = 0; |
f9ce1f1cd
|
1514 |
current->in_execve = 0; |
a6f76f23d
|
1515 1516 |
out_free: |
08a6fac1c
|
1517 |
free_bprm(bprm); |
1da177e4c
|
1518 |
|
fd8328be8
|
1519 |
out_files: |
3b1253880
|
1520 1521 |
if (displaced) reset_files_struct(displaced); |
1da177e4c
|
1522 1523 1524 |
out_ret: return retval; } |
ba2d01629
|
1525 1526 1527 1528 1529 |
int do_execve(const char *filename, const char __user *const __user *__argv, const char __user *const __user *__envp, struct pt_regs *regs) { |
0e028465d
|
1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 |
struct user_arg_ptr argv = { .ptr.native = __argv }; struct user_arg_ptr envp = { .ptr.native = __envp }; return do_execve_common(filename, argv, envp, regs); } #ifdef CONFIG_COMPAT int compat_do_execve(char *filename, compat_uptr_t __user *__argv, compat_uptr_t __user *__envp, struct pt_regs *regs) { struct user_arg_ptr argv = { .is_compat = true, .ptr.compat = __argv, }; struct user_arg_ptr envp = { .is_compat = true, .ptr.compat = __envp, }; |
ba2d01629
|
1549 1550 |
return do_execve_common(filename, argv, envp, regs); } |
0e028465d
|
1551 |
#endif |
ba2d01629
|
1552 |
|
964ee7df9
|
1553 |
void set_binfmt(struct linux_binfmt *new) |
1da177e4c
|
1554 |
{ |
801460d0c
|
1555 1556 1557 1558 |
struct mm_struct *mm = current->mm; if (mm->binfmt) module_put(mm->binfmt->module); |
1da177e4c
|
1559 |
|
801460d0c
|
1560 |
mm->binfmt = new; |
964ee7df9
|
1561 1562 |
if (new) __module_get(new->module); |
1da177e4c
|
1563 1564 1565 |
} EXPORT_SYMBOL(set_binfmt); |
1b0d300bd
|
1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 |
static int expand_corename(struct core_name *cn) { char *old_corename = cn->corename; cn->size = CORENAME_MAX_SIZE * atomic_inc_return(&call_count); cn->corename = krealloc(old_corename, cn->size, GFP_KERNEL); if (!cn->corename) { kfree(old_corename); return -ENOMEM; } return 0; } static int cn_printf(struct core_name *cn, const char *fmt, ...) { char *cur; int need; int ret; va_list arg; va_start(arg, fmt); need = vsnprintf(NULL, 0, fmt, arg); va_end(arg); if (likely(need < cn->size - cn->used - 1)) goto out_printf; ret = expand_corename(cn); if (ret) goto expand_fail; out_printf: cur = cn->corename + cn->used; va_start(arg, fmt); vsnprintf(cur, need + 1, fmt, arg); va_end(arg); cn->used += need; return 0; expand_fail: return ret; } |
2c563731f
|
1610 1611 1612 1613 1614 1615 |
static void cn_escape(char *str) { for (; *str; str++) if (*str == '/') *str = '!'; } |
57cc083ad
|
1616 1617 1618 |
static int cn_print_exe_file(struct core_name *cn) { struct file *exe_file; |
2c563731f
|
1619 |
char *pathbuf, *path; |
57cc083ad
|
1620 1621 1622 |
int ret; exe_file = get_mm_exe_file(current->mm); |
2c563731f
|
1623 1624 1625 1626 1627 1628 |
if (!exe_file) { char *commstart = cn->corename + cn->used; ret = cn_printf(cn, "%s (path unknown)", current->comm); cn_escape(commstart); return ret; } |
57cc083ad
|
1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 |
pathbuf = kmalloc(PATH_MAX, GFP_TEMPORARY); if (!pathbuf) { ret = -ENOMEM; goto put_exe_file; } path = d_path(&exe_file->f_path, pathbuf, PATH_MAX); if (IS_ERR(path)) { ret = PTR_ERR(path); goto free_buf; } |
2c563731f
|
1641 |
cn_escape(path); |
57cc083ad
|
1642 1643 1644 1645 1646 1647 1648 1649 1650 |
ret = cn_printf(cn, "%s", path); free_buf: kfree(pathbuf); put_exe_file: fput(exe_file); return ret; } |
1da177e4c
|
1651 1652 1653 1654 |
/* format_corename will inspect the pattern parameter, and output a * name into corename, which must have space for at least * CORENAME_MAX_SIZE bytes plus one byte for the zero terminator. */ |
1b0d300bd
|
1655 |
static int format_corename(struct core_name *cn, long signr) |
1da177e4c
|
1656 |
{ |
86a264abe
|
1657 |
const struct cred *cred = current_cred(); |
565b9b14e
|
1658 1659 |
const char *pat_ptr = core_pattern; int ispipe = (*pat_ptr == '|'); |
1da177e4c
|
1660 |
int pid_in_pattern = 0; |
1b0d300bd
|
1661 1662 1663 1664 1665 1666 1667 1668 |
int err = 0; cn->size = CORENAME_MAX_SIZE * atomic_read(&call_count); cn->corename = kmalloc(cn->size, GFP_KERNEL); cn->used = 0; if (!cn->corename) return -ENOMEM; |
1da177e4c
|
1669 1670 1671 1672 1673 |
/* Repeat as long as we have more pattern to process and more output space */ while (*pat_ptr) { if (*pat_ptr != '%') { |
1b0d300bd
|
1674 |
if (*pat_ptr == 0) |
1da177e4c
|
1675 |
goto out; |
1b0d300bd
|
1676 |
err = cn_printf(cn, "%c", *pat_ptr++); |
1da177e4c
|
1677 1678 |
} else { switch (*++pat_ptr) { |
1b0d300bd
|
1679 |
/* single % at the end, drop that */ |
1da177e4c
|
1680 1681 1682 1683 |
case 0: goto out; /* Double percent, output one percent */ case '%': |
1b0d300bd
|
1684 |
err = cn_printf(cn, "%c", '%'); |
1da177e4c
|
1685 1686 1687 1688 |
break; /* pid */ case 'p': pid_in_pattern = 1; |
1b0d300bd
|
1689 1690 |
err = cn_printf(cn, "%d", task_tgid_vnr(current)); |
1da177e4c
|
1691 1692 1693 |
break; /* uid */ case 'u': |
1b0d300bd
|
1694 |
err = cn_printf(cn, "%d", cred->uid); |
1da177e4c
|
1695 1696 1697 |
break; /* gid */ case 'g': |
1b0d300bd
|
1698 |
err = cn_printf(cn, "%d", cred->gid); |
1da177e4c
|
1699 1700 1701 |
break; /* signal that caused the coredump */ case 's': |
1b0d300bd
|
1702 |
err = cn_printf(cn, "%ld", signr); |
1da177e4c
|
1703 1704 1705 1706 1707 |
break; /* UNIX time of coredump */ case 't': { struct timeval tv; do_gettimeofday(&tv); |
1b0d300bd
|
1708 |
err = cn_printf(cn, "%lu", tv.tv_sec); |
1da177e4c
|
1709 1710 1711 |
break; } /* hostname */ |
2c563731f
|
1712 1713 |
case 'h': { char *namestart = cn->corename + cn->used; |
1da177e4c
|
1714 |
down_read(&uts_sem); |
1b0d300bd
|
1715 1716 |
err = cn_printf(cn, "%s", utsname()->nodename); |
1da177e4c
|
1717 |
up_read(&uts_sem); |
2c563731f
|
1718 |
cn_escape(namestart); |
1da177e4c
|
1719 |
break; |
2c563731f
|
1720 |
} |
1da177e4c
|
1721 |
/* executable */ |
2c563731f
|
1722 1723 |
case 'e': { char *commstart = cn->corename + cn->used; |
1b0d300bd
|
1724 |
err = cn_printf(cn, "%s", current->comm); |
2c563731f
|
1725 |
cn_escape(commstart); |
1da177e4c
|
1726 |
break; |
2c563731f
|
1727 |
} |
57cc083ad
|
1728 1729 1730 |
case 'E': err = cn_print_exe_file(cn); break; |
74aadce98
|
1731 1732 |
/* core limit size */ case 'c': |
1b0d300bd
|
1733 1734 |
err = cn_printf(cn, "%lu", rlimit(RLIMIT_CORE)); |
74aadce98
|
1735 |
break; |
1da177e4c
|
1736 1737 1738 1739 1740 |
default: break; } ++pat_ptr; } |
1b0d300bd
|
1741 1742 1743 |
if (err) return err; |
1da177e4c
|
1744 |
} |
1b0d300bd
|
1745 |
|
1da177e4c
|
1746 1747 1748 1749 |
/* Backward compatibility with core_uses_pid: * * If core_pattern does not include a %p (as is the default) * and core_uses_pid is set, then .%pid will be appended to |
c4bbafda7
|
1750 |
* the filename. Do not do this for piped commands. */ |
6409324b3
|
1751 |
if (!ispipe && !pid_in_pattern && core_uses_pid) { |
1b0d300bd
|
1752 1753 1754 |
err = cn_printf(cn, ".%d", task_tgid_vnr(current)); if (err) return err; |
1da177e4c
|
1755 |
} |
c4bbafda7
|
1756 |
out: |
c4bbafda7
|
1757 |
return ispipe; |
1da177e4c
|
1758 |
} |
5c99cbf49
|
1759 |
static int zap_process(struct task_struct *start, int exit_code) |
aceecc041
|
1760 1761 |
{ struct task_struct *t; |
8cd9c2491
|
1762 |
int nr = 0; |
281de339c
|
1763 |
|
d5f70c00a
|
1764 |
start->signal->flags = SIGNAL_GROUP_EXIT; |
5c99cbf49
|
1765 |
start->signal->group_exit_code = exit_code; |
d5f70c00a
|
1766 |
start->signal->group_stop_count = 0; |
aceecc041
|
1767 1768 1769 |
t = start; do { |
6dfca3298
|
1770 |
task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK); |
aceecc041
|
1771 |
if (t != current && t->mm) { |
281de339c
|
1772 1773 |
sigaddset(&t->pending.signal, SIGKILL); signal_wake_up(t, 1); |
8cd9c2491
|
1774 |
nr++; |
aceecc041
|
1775 |
} |
e4901f92a
|
1776 |
} while_each_thread(start, t); |
8cd9c2491
|
1777 1778 |
return nr; |
aceecc041
|
1779 |
} |
dcf560c59
|
1780 |
static inline int zap_threads(struct task_struct *tsk, struct mm_struct *mm, |
8cd9c2491
|
1781 |
struct core_state *core_state, int exit_code) |
1da177e4c
|
1782 1783 |
{ struct task_struct *g, *p; |
5debfa6da
|
1784 |
unsigned long flags; |
8cd9c2491
|
1785 |
int nr = -EAGAIN; |
dcf560c59
|
1786 1787 |
spin_lock_irq(&tsk->sighand->siglock); |
ed5d2cac1
|
1788 |
if (!signal_group_exit(tsk->signal)) { |
8cd9c2491
|
1789 |
mm->core_state = core_state; |
5c99cbf49
|
1790 |
nr = zap_process(tsk, exit_code); |
1da177e4c
|
1791 |
} |
dcf560c59
|
1792 |
spin_unlock_irq(&tsk->sighand->siglock); |
8cd9c2491
|
1793 1794 |
if (unlikely(nr < 0)) return nr; |
1da177e4c
|
1795 |
|
8cd9c2491
|
1796 |
if (atomic_read(&mm->mm_users) == nr + 1) |
5debfa6da
|
1797 |
goto done; |
e4901f92a
|
1798 1799 |
/* * We should find and kill all tasks which use this mm, and we should |
999d9fc16
|
1800 |
* count them correctly into ->nr_threads. We don't take tasklist |
e4901f92a
|
1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 |
* lock, but this is safe wrt: * * fork: * None of sub-threads can fork after zap_process(leader). All * processes which were created before this point should be * visible to zap_threads() because copy_process() adds the new * process to the tail of init_task.tasks list, and lock/unlock * of ->siglock provides a memory barrier. * * do_exit: * The caller holds mm->mmap_sem. This means that the task which * uses this mm can't pass exit_mm(), so it can't exit or clear * its ->mm. * * de_thread: * It does list_replace_rcu(&leader->tasks, ¤t->tasks), * we must see either old or new leader, this does not matter. * However, it can change p->sighand, so lock_task_sighand(p) * must be used. Since p->mm != NULL and we hold ->mmap_sem * it can't fail. * * Note also that "g" can be the old leader with ->mm == NULL * and already unhashed and thus removed from ->thread_group. * This is OK, __unhash_process()->list_del_rcu() does not * clear the ->next pointer, we will find the new leader via * next_thread(). */ |
7b1c6154f
|
1828 |
rcu_read_lock(); |
aceecc041
|
1829 |
for_each_process(g) { |
5debfa6da
|
1830 1831 |
if (g == tsk->group_leader) continue; |
15b9f360c
|
1832 1833 |
if (g->flags & PF_KTHREAD) continue; |
aceecc041
|
1834 1835 1836 |
p = g; do { if (p->mm) { |
15b9f360c
|
1837 |
if (unlikely(p->mm == mm)) { |
5debfa6da
|
1838 |
lock_task_sighand(p, &flags); |
5c99cbf49
|
1839 |
nr += zap_process(p, exit_code); |
5debfa6da
|
1840 1841 |
unlock_task_sighand(p, &flags); } |
aceecc041
|
1842 1843 |
break; } |
e4901f92a
|
1844 |
} while_each_thread(g, p); |
aceecc041
|
1845 |
} |
7b1c6154f
|
1846 |
rcu_read_unlock(); |
5debfa6da
|
1847 |
done: |
c5f1cc8c1
|
1848 |
atomic_set(&core_state->nr_threads, nr); |
8cd9c2491
|
1849 |
return nr; |
1da177e4c
|
1850 |
} |
9d5b327bf
|
1851 |
static int coredump_wait(int exit_code, struct core_state *core_state) |
1da177e4c
|
1852 |
{ |
dcf560c59
|
1853 1854 |
struct task_struct *tsk = current; struct mm_struct *mm = tsk->mm; |
dcf560c59
|
1855 |
struct completion *vfork_done; |
269b005a2
|
1856 |
int core_waiters = -EBUSY; |
1da177e4c
|
1857 |
|
9d5b327bf
|
1858 |
init_completion(&core_state->startup); |
b564daf80
|
1859 1860 |
core_state->dumper.task = tsk; core_state->dumper.next = NULL; |
269b005a2
|
1861 1862 1863 1864 |
down_write(&mm->mmap_sem); if (!mm->core_state) core_waiters = zap_threads(tsk, mm, core_state, exit_code); |
2384f55f8
|
1865 |
up_write(&mm->mmap_sem); |
dcf560c59
|
1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 |
if (unlikely(core_waiters < 0)) goto fail; /* * Make sure nobody is waiting for us to release the VM, * otherwise we can deadlock when we wait on each other */ vfork_done = tsk->vfork_done; if (vfork_done) { tsk->vfork_done = NULL; complete(vfork_done); } |
2384f55f8
|
1878 |
if (core_waiters) |
9d5b327bf
|
1879 |
wait_for_completion(&core_state->startup); |
dcf560c59
|
1880 |
fail: |
dcf560c59
|
1881 |
return core_waiters; |
1da177e4c
|
1882 |
} |
a94e2d408
|
1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 |
static void coredump_finish(struct mm_struct *mm) { struct core_thread *curr, *next; struct task_struct *task; next = mm->core_state->dumper.next; while ((curr = next) != NULL) { next = curr->next; task = curr->task; /* * see exit_mm(), curr->task must not see * ->task == NULL before we read ->next. */ smp_mb(); curr->task = NULL; wake_up_process(task); } mm->core_state = NULL; } |
6c5d52382
|
1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 |
/* * 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) { case 0: clear_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; case 1: set_bit(MMF_DUMPABLE, &mm->flags); smp_wmb(); clear_bit(MMF_DUMP_SECURELY, &mm->flags); break; case 2: set_bit(MMF_DUMP_SECURELY, &mm->flags); smp_wmb(); set_bit(MMF_DUMPABLE, &mm->flags); break; } } |
6c5d52382
|
1943 |
|
30736a4d4
|
1944 |
static int __get_dumpable(unsigned long mm_flags) |
6c5d52382
|
1945 1946 |
{ int ret; |
30736a4d4
|
1947 |
ret = mm_flags & MMF_DUMPABLE_MASK; |
6c5d52382
|
1948 1949 |
return (ret >= 2) ? 2 : ret; } |
30736a4d4
|
1950 1951 1952 1953 |
int get_dumpable(struct mm_struct *mm) { return __get_dumpable(mm->flags); } |
61be228a0
|
1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 |
static void wait_for_dump_helpers(struct file *file) { struct pipe_inode_info *pipe; pipe = file->f_path.dentry->d_inode->i_pipe; pipe_lock(pipe); pipe->readers++; pipe->writers--; while ((pipe->readers > 1) && (!signal_pending(current))) { wake_up_interruptible_sync(&pipe->wait); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); pipe_wait(pipe); } pipe->readers--; pipe->writers++; pipe_unlock(pipe); } |
898b374af
|
1975 |
/* |
1bef82917
|
1976 |
* umh_pipe_setup |
898b374af
|
1977 1978 1979 1980 1981 1982 1983 1984 1985 |
* helper function to customize the process used * to collect the core in userspace. Specifically * it sets up a pipe and installs it as fd 0 (stdin) * for the process. Returns 0 on success, or * PTR_ERR on failure. * Note that it also sets the core limit to 1. This * is a special value that we use to trap recursive * core dumps */ |
879669961
|
1986 |
static int umh_pipe_setup(struct subprocess_info *info, struct cred *new) |
898b374af
|
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 |
{ struct file *rp, *wp; struct fdtable *fdt; struct coredump_params *cp = (struct coredump_params *)info->data; struct files_struct *cf = current->files; wp = create_write_pipe(0); if (IS_ERR(wp)) return PTR_ERR(wp); rp = create_read_pipe(wp, 0); if (IS_ERR(rp)) { free_write_pipe(wp); return PTR_ERR(rp); } cp->file = wp; sys_close(0); fd_install(0, rp); spin_lock(&cf->file_lock); fdt = files_fdtable(cf); FD_SET(0, fdt->open_fds); FD_CLR(0, fdt->close_on_exec); spin_unlock(&cf->file_lock); /* and disallow core files too */ current->signal->rlim[RLIMIT_CORE] = (struct rlimit){1, 1}; return 0; } |
8cd3ac3ac
|
2018 |
void do_coredump(long signr, int exit_code, struct pt_regs *regs) |
1da177e4c
|
2019 |
{ |
9d5b327bf
|
2020 |
struct core_state core_state; |
1b0d300bd
|
2021 |
struct core_name cn; |
1da177e4c
|
2022 2023 |
struct mm_struct *mm = current->mm; struct linux_binfmt * binfmt; |
d84f4f992
|
2024 2025 |
const struct cred *old_cred; struct cred *cred; |
1da177e4c
|
2026 |
int retval = 0; |
d6e711448
|
2027 |
int flag = 0; |
d5bf4c4f5
|
2028 |
int ispipe; |
a293980c2
|
2029 |
static atomic_t core_dump_count = ATOMIC_INIT(0); |
f6151dfea
|
2030 2031 2032 |
struct coredump_params cprm = { .signr = signr, .regs = regs, |
d554ed895
|
2033 |
.limit = rlimit(RLIMIT_CORE), |
30736a4d4
|
2034 2035 2036 2037 2038 2039 |
/* * We must use the same mm->flags while dumping core to avoid * inconsistency of bit flags, since this flag is not protected * by any locks. */ .mm_flags = mm->flags, |
f6151dfea
|
2040 |
}; |
1da177e4c
|
2041 |
|
0a4ff8c25
|
2042 |
audit_core_dumps(signr); |
801460d0c
|
2043 |
binfmt = mm->binfmt; |
1da177e4c
|
2044 2045 |
if (!binfmt || !binfmt->core_dump) goto fail; |
269b005a2
|
2046 2047 |
if (!__get_dumpable(cprm.mm_flags)) goto fail; |
d84f4f992
|
2048 2049 |
cred = prepare_creds(); |
5e43aef53
|
2050 |
if (!cred) |
d84f4f992
|
2051 |
goto fail; |
d6e711448
|
2052 2053 2054 2055 2056 |
/* * We cannot trust fsuid as being the "true" uid of the * process nor do we know its entire history. We only know it * was tainted so we dump it as root in mode 2. */ |
30736a4d4
|
2057 2058 |
if (__get_dumpable(cprm.mm_flags) == 2) { /* Setuid core dump mode */ |
d6e711448
|
2059 |
flag = O_EXCL; /* Stop rewrite attacks */ |
d84f4f992
|
2060 |
cred->fsuid = 0; /* Dump root private */ |
d6e711448
|
2061 |
} |
1291cf416
|
2062 |
|
9d5b327bf
|
2063 |
retval = coredump_wait(exit_code, &core_state); |
5e43aef53
|
2064 2065 |
if (retval < 0) goto fail_creds; |
d84f4f992
|
2066 2067 |
old_cred = override_creds(cred); |
1da177e4c
|
2068 2069 2070 2071 2072 |
/* * Clear any false indication of pending signals that might * be seen by the filesystem code called to write the core file. */ |
1da177e4c
|
2073 |
clear_thread_flag(TIF_SIGPENDING); |
1b0d300bd
|
2074 |
ispipe = format_corename(&cn, signr); |
c4bbafda7
|
2075 |
if (ispipe) { |
d5bf4c4f5
|
2076 2077 |
int dump_count; char **helper_argv; |
99b645674
|
2078 2079 2080 2081 2082 2083 2084 |
if (ispipe < 0) { printk(KERN_WARNING "format_corename failed "); printk(KERN_WARNING "Aborting core "); goto fail_corename; } |
898b374af
|
2085 |
if (cprm.limit == 1) { |
725eae32d
|
2086 2087 2088 |
/* * Normally core limits are irrelevant to pipes, since * we're not writing to the file system, but we use |
898b374af
|
2089 2090 |
* cprm.limit of 1 here as a speacial value. Any * non-1 limit gets set to RLIM_INFINITY below, but |
725eae32d
|
2091 2092 |
* a limit of 0 skips the dump. This is a consistent * way to catch recursive crashes. We can still crash |
898b374af
|
2093 |
* if the core_pattern binary sets RLIM_CORE = !1 |
725eae32d
|
2094 2095 2096 2097 2098 2099 2100 |
* but it runs as root, and can do lots of stupid things * Note that we use task_tgid_vnr here to grab the pid * of the process group leader. That way we get the * right pid if a thread in a multi-threaded * core_pattern process dies. */ printk(KERN_WARNING |
898b374af
|
2101 2102 |
"Process %d(%s) has RLIMIT_CORE set to 1 ", |
725eae32d
|
2103 2104 2105 2106 2107 |
task_tgid_vnr(current), current->comm); printk(KERN_WARNING "Aborting core "); goto fail_unlock; } |
d5bf4c4f5
|
2108 |
cprm.limit = RLIM_INFINITY; |
725eae32d
|
2109 |
|
a293980c2
|
2110 2111 2112 2113 2114 2115 2116 2117 2118 |
dump_count = atomic_inc_return(&core_dump_count); if (core_pipe_limit && (core_pipe_limit < dump_count)) { printk(KERN_WARNING "Pid %d(%s) over core_pipe_limit ", task_tgid_vnr(current), current->comm); printk(KERN_WARNING "Skipping core dump "); goto fail_dropcount; } |
1b0d300bd
|
2119 |
helper_argv = argv_split(GFP_KERNEL, cn.corename+1, NULL); |
350eaf791
|
2120 2121 2122 2123 |
if (!helper_argv) { printk(KERN_WARNING "%s failed to allocate memory ", __func__); |
a293980c2
|
2124 |
goto fail_dropcount; |
350eaf791
|
2125 |
} |
323211371
|
2126 |
|
d5bf4c4f5
|
2127 2128 2129 2130 2131 |
retval = call_usermodehelper_fns(helper_argv[0], helper_argv, NULL, UMH_WAIT_EXEC, umh_pipe_setup, NULL, &cprm); argv_free(helper_argv); if (retval) { |
d025c9db7
|
2132 2133 |
printk(KERN_INFO "Core dump to %s pipe failed ", |
1b0d300bd
|
2134 |
cn.corename); |
d5bf4c4f5
|
2135 |
goto close_fail; |
d025c9db7
|
2136 |
} |
c71354112
|
2137 2138 2139 2140 2141 |
} else { struct inode *inode; if (cprm.limit < binfmt->min_coredump) goto fail_unlock; |
1b0d300bd
|
2142 |
cprm.file = filp_open(cn.corename, |
6d4df677f
|
2143 2144 |
O_CREAT | 2 | O_NOFOLLOW | O_LARGEFILE | flag, 0600); |
c71354112
|
2145 2146 |
if (IS_ERR(cprm.file)) goto fail_unlock; |
1da177e4c
|
2147 |
|
c71354112
|
2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 |
inode = cprm.file->f_path.dentry->d_inode; if (inode->i_nlink > 1) goto close_fail; if (d_unhashed(cprm.file->f_path.dentry)) goto close_fail; /* * AK: actually i see no reason to not allow this for named * pipes etc, but keep the previous behaviour for now. */ if (!S_ISREG(inode->i_mode)) goto close_fail; /* * Dont allow local users get cute and trick others to coredump * into their pre-created files. */ if (inode->i_uid != current_fsuid()) goto close_fail; if (!cprm.file->f_op || !cprm.file->f_op->write) goto close_fail; if (do_truncate(cprm.file->f_path.dentry, 0, 0, cprm.file)) goto close_fail; } |
1da177e4c
|
2170 |
|
c71354112
|
2171 |
retval = binfmt->core_dump(&cprm); |
1da177e4c
|
2172 2173 |
if (retval) current->signal->group_exit_code |= 0x80; |
d5bf4c4f5
|
2174 |
|
61be228a0
|
2175 |
if (ispipe && core_pipe_limit) |
f6151dfea
|
2176 |
wait_for_dump_helpers(cprm.file); |
d5bf4c4f5
|
2177 2178 2179 |
close_fail: if (cprm.file) filp_close(cprm.file, NULL); |
a293980c2
|
2180 |
fail_dropcount: |
d5bf4c4f5
|
2181 |
if (ispipe) |
a293980c2
|
2182 |
atomic_dec(&core_dump_count); |
1da177e4c
|
2183 |
fail_unlock: |
1b0d300bd
|
2184 2185 |
kfree(cn.corename); fail_corename: |
5e43aef53
|
2186 |
coredump_finish(mm); |
d84f4f992
|
2187 |
revert_creds(old_cred); |
5e43aef53
|
2188 |
fail_creds: |
d84f4f992
|
2189 |
put_cred(cred); |
1da177e4c
|
2190 |
fail: |
8cd3ac3ac
|
2191 |
return; |
1da177e4c
|
2192 |
} |
3aa0ce825
|
2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 |
/* * Core dumping helper functions. These are the only things you should * do on a core-file: use only these functions to write out all the * necessary info. */ int dump_write(struct file *file, const void *addr, int nr) { return access_ok(VERIFY_READ, addr, nr) && file->f_op->write(file, addr, nr, &file->f_pos) == nr; } |
8fd01d6cf
|
2203 |
EXPORT_SYMBOL(dump_write); |
3aa0ce825
|
2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 |
int dump_seek(struct file *file, loff_t off) { int ret = 1; if (file->f_op->llseek && file->f_op->llseek != no_llseek) { if (file->f_op->llseek(file, off, SEEK_CUR) < 0) return 0; } else { char *buf = (char *)get_zeroed_page(GFP_KERNEL); if (!buf) return 0; while (off > 0) { unsigned long n = off; if (n > PAGE_SIZE) n = PAGE_SIZE; if (!dump_write(file, buf, n)) { ret = 0; break; } off -= n; } free_page((unsigned long)buf); } return ret; } |
8fd01d6cf
|
2232 |
EXPORT_SYMBOL(dump_seek); |