Blame view
kernel/fork.c
55.8 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 |
/* * linux/kernel/fork.c * * Copyright (C) 1991, 1992 Linus Torvalds */ /* * 'fork.c' contains the help-routines for the 'fork' system call * (see also entry.S and others). * Fork is rather simple, once you get the hang of it, but the memory * management can be a bitch. See 'mm/memory.c': 'copy_page_range()' */ |
1da177e4c
|
13 14 15 |
#include <linux/slab.h> #include <linux/init.h> #include <linux/unistd.h> |
1da177e4c
|
16 17 18 |
#include <linux/module.h> #include <linux/vmalloc.h> #include <linux/completion.h> |
1da177e4c
|
19 20 21 22 |
#include <linux/personality.h> #include <linux/mempolicy.h> #include <linux/sem.h> #include <linux/file.h> |
9f3acc314
|
23 |
#include <linux/fdtable.h> |
da9cbc873
|
24 |
#include <linux/iocontext.h> |
1da177e4c
|
25 26 27 |
#include <linux/key.h> #include <linux/binfmts.h> #include <linux/mman.h> |
cddb8a5c1
|
28 |
#include <linux/mmu_notifier.h> |
1da177e4c
|
29 |
#include <linux/fs.h> |
615d6e875
|
30 31 |
#include <linux/mm.h> #include <linux/vmacache.h> |
ab516013a
|
32 |
#include <linux/nsproxy.h> |
c59ede7b7
|
33 |
#include <linux/capability.h> |
1da177e4c
|
34 |
#include <linux/cpu.h> |
b4f48b636
|
35 |
#include <linux/cgroup.h> |
1da177e4c
|
36 |
#include <linux/security.h> |
a1e78772d
|
37 |
#include <linux/hugetlb.h> |
e2cfabdfd
|
38 |
#include <linux/seccomp.h> |
1da177e4c
|
39 40 41 42 |
#include <linux/swap.h> #include <linux/syscalls.h> #include <linux/jiffies.h> #include <linux/futex.h> |
8141c7f3e
|
43 |
#include <linux/compat.h> |
207205a2b
|
44 |
#include <linux/kthread.h> |
7c3ab7381
|
45 |
#include <linux/task_io_accounting_ops.h> |
ab2af1f50
|
46 |
#include <linux/rcupdate.h> |
1da177e4c
|
47 48 49 |
#include <linux/ptrace.h> #include <linux/mount.h> #include <linux/audit.h> |
78fb74669
|
50 |
#include <linux/memcontrol.h> |
f201ae235
|
51 |
#include <linux/ftrace.h> |
5e2bf0142
|
52 |
#include <linux/proc_fs.h> |
1da177e4c
|
53 54 |
#include <linux/profile.h> #include <linux/rmap.h> |
f8af4da3b
|
55 |
#include <linux/ksm.h> |
1da177e4c
|
56 |
#include <linux/acct.h> |
8f0ab5147
|
57 |
#include <linux/tsacct_kern.h> |
9f46080c4
|
58 |
#include <linux/cn_proc.h> |
ba96a0c88
|
59 |
#include <linux/freezer.h> |
ca74e92b4
|
60 |
#include <linux/delayacct.h> |
ad4ecbcba
|
61 |
#include <linux/taskstats_kern.h> |
0a4254058
|
62 |
#include <linux/random.h> |
522ed7767
|
63 |
#include <linux/tty.h> |
fd0928df9
|
64 |
#include <linux/blkdev.h> |
5ad4e53bd
|
65 |
#include <linux/fs_struct.h> |
7c9f8861e
|
66 |
#include <linux/magic.h> |
cdd6c482c
|
67 |
#include <linux/perf_event.h> |
42c4ab41a
|
68 |
#include <linux/posix-timers.h> |
8e7cac798
|
69 |
#include <linux/user-return-notifier.h> |
3d5992d2a
|
70 |
#include <linux/oom.h> |
ba76149f4
|
71 |
#include <linux/khugepaged.h> |
d80e731ec
|
72 |
#include <linux/signalfd.h> |
0326f5a94
|
73 |
#include <linux/uprobes.h> |
a27bb332c
|
74 |
#include <linux/aio.h> |
52f5684c8
|
75 |
#include <linux/compiler.h> |
16db3d3f1
|
76 |
#include <linux/sysctl.h> |
5c9a8750a
|
77 |
#include <linux/kcov.h> |
1da177e4c
|
78 79 80 81 82 83 84 |
#include <asm/pgtable.h> #include <asm/pgalloc.h> #include <asm/uaccess.h> #include <asm/mmu_context.h> #include <asm/cacheflush.h> #include <asm/tlbflush.h> |
ad8d75fff
|
85 |
#include <trace/events/sched.h> |
43d2b1132
|
86 87 |
#define CREATE_TRACE_POINTS #include <trace/events/task.h> |
1da177e4c
|
88 |
/* |
ac1b398de
|
89 90 91 92 93 94 95 96 97 98 |
* Minimum number of threads to boot the kernel */ #define MIN_THREADS 20 /* * Maximum number of threads */ #define MAX_THREADS FUTEX_TID_MASK /* |
1da177e4c
|
99 100 101 |
* Protected counters by write_lock_irq(&tasklist_lock) */ unsigned long total_forks; /* Handle normal Linux uptimes. */ |
fb0a685cb
|
102 |
int nr_threads; /* The idle threads do not count.. */ |
1da177e4c
|
103 104 105 106 |
int max_threads; /* tunable limit on nr_threads */ DEFINE_PER_CPU(unsigned long, process_counts) = 0; |
c59923a15
|
107 |
__cacheline_aligned DEFINE_RWLOCK(tasklist_lock); /* outer */ |
db1466b3e
|
108 109 110 111 112 113 114 115 |
#ifdef CONFIG_PROVE_RCU int lockdep_tasklist_lock_is_held(void) { return lockdep_is_held(&tasklist_lock); } EXPORT_SYMBOL_GPL(lockdep_tasklist_lock_is_held); #endif /* #ifdef CONFIG_PROVE_RCU */ |
1da177e4c
|
116 117 118 119 120 |
int nr_processes(void) { int cpu; int total = 0; |
1d5107509
|
121 |
for_each_possible_cpu(cpu) |
1da177e4c
|
122 123 124 125 |
total += per_cpu(process_counts, cpu); return total; } |
f19b9f74b
|
126 127 128 |
void __weak arch_release_task_struct(struct task_struct *tsk) { } |
f5e102873
|
129 |
#ifndef CONFIG_ARCH_TASK_STRUCT_ALLOCATOR |
e18b890bb
|
130 |
static struct kmem_cache *task_struct_cachep; |
41101809a
|
131 132 133 134 135 |
static inline struct task_struct *alloc_task_struct_node(int node) { return kmem_cache_alloc_node(task_struct_cachep, GFP_KERNEL, node); } |
41101809a
|
136 137 |
static inline void free_task_struct(struct task_struct *tsk) { |
41101809a
|
138 139 |
kmem_cache_free(task_struct_cachep, tsk); } |
1da177e4c
|
140 |
#endif |
b235beea9
|
141 |
void __weak arch_release_thread_stack(unsigned long *stack) |
f19b9f74b
|
142 143 |
{ } |
b235beea9
|
144 |
#ifndef CONFIG_ARCH_THREAD_STACK_ALLOCATOR |
41101809a
|
145 |
|
0d15d74a1
|
146 147 148 149 |
/* * Allocate pages if THREAD_SIZE is >= PAGE_SIZE, otherwise use a * kmemcache based allocator. */ |
ba14a194a
|
150 |
# if THREAD_SIZE >= PAGE_SIZE || defined(CONFIG_VMAP_STACK) |
ac496bf48
|
151 152 153 154 155 156 157 158 159 |
#ifdef CONFIG_VMAP_STACK /* * vmalloc() is a bit slow, and calling vfree() enough times will force a TLB * flush. Try to minimize the number of calls by caching stacks. */ #define NR_CACHED_STACKS 2 static DEFINE_PER_CPU(struct vm_struct *, cached_stacks[NR_CACHED_STACKS]); #endif |
ba14a194a
|
160 |
static unsigned long *alloc_thread_stack_node(struct task_struct *tsk, int node) |
b69c49b78
|
161 |
{ |
ba14a194a
|
162 |
#ifdef CONFIG_VMAP_STACK |
ac496bf48
|
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 |
void *stack; int i; local_irq_disable(); for (i = 0; i < NR_CACHED_STACKS; i++) { struct vm_struct *s = this_cpu_read(cached_stacks[i]); if (!s) continue; this_cpu_write(cached_stacks[i], NULL); tsk->stack_vm_area = s; local_irq_enable(); return s->addr; } local_irq_enable(); stack = __vmalloc_node_range(THREAD_SIZE, THREAD_SIZE, VMALLOC_START, VMALLOC_END, THREADINFO_GFP | __GFP_HIGHMEM, PAGE_KERNEL, 0, node, __builtin_return_address(0)); |
ba14a194a
|
185 186 187 188 189 190 191 192 193 194 |
/* * We can't call find_vm_area() in interrupt context, and * free_thread_stack() can be called in interrupt context, * so cache the vm_struct. */ if (stack) tsk->stack_vm_area = find_vm_area(stack); return stack; #else |
4949148ad
|
195 196 |
struct page *page = alloc_pages_node(node, THREADINFO_GFP, THREAD_SIZE_ORDER); |
b6a84016b
|
197 198 |
return page ? page_address(page) : NULL; |
ba14a194a
|
199 |
#endif |
b69c49b78
|
200 |
} |
ba14a194a
|
201 |
static inline void free_thread_stack(struct task_struct *tsk) |
b69c49b78
|
202 |
{ |
ac496bf48
|
203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 |
#ifdef CONFIG_VMAP_STACK if (task_stack_vm_area(tsk)) { unsigned long flags; int i; local_irq_save(flags); for (i = 0; i < NR_CACHED_STACKS; i++) { if (this_cpu_read(cached_stacks[i])) continue; this_cpu_write(cached_stacks[i], tsk->stack_vm_area); local_irq_restore(flags); return; } local_irq_restore(flags); |
ba14a194a
|
218 |
vfree(tsk->stack); |
ac496bf48
|
219 220 221 222 223 |
return; } #endif __free_pages(virt_to_page(tsk->stack), THREAD_SIZE_ORDER); |
b69c49b78
|
224 |
} |
0d15d74a1
|
225 |
# else |
b235beea9
|
226 |
static struct kmem_cache *thread_stack_cache; |
0d15d74a1
|
227 |
|
9521d3997
|
228 |
static unsigned long *alloc_thread_stack_node(struct task_struct *tsk, |
0d15d74a1
|
229 230 |
int node) { |
b235beea9
|
231 |
return kmem_cache_alloc_node(thread_stack_cache, THREADINFO_GFP, node); |
0d15d74a1
|
232 |
} |
ba14a194a
|
233 |
static void free_thread_stack(struct task_struct *tsk) |
0d15d74a1
|
234 |
{ |
ba14a194a
|
235 |
kmem_cache_free(thread_stack_cache, tsk->stack); |
0d15d74a1
|
236 |
} |
b235beea9
|
237 |
void thread_stack_cache_init(void) |
0d15d74a1
|
238 |
{ |
b235beea9
|
239 |
thread_stack_cache = kmem_cache_create("thread_stack", THREAD_SIZE, |
0d15d74a1
|
240 |
THREAD_SIZE, 0, NULL); |
b235beea9
|
241 |
BUG_ON(thread_stack_cache == NULL); |
0d15d74a1
|
242 243 |
} # endif |
b69c49b78
|
244 |
#endif |
1da177e4c
|
245 |
/* SLAB cache for signal_struct structures (tsk->signal) */ |
e18b890bb
|
246 |
static struct kmem_cache *signal_cachep; |
1da177e4c
|
247 248 |
/* SLAB cache for sighand_struct structures (tsk->sighand) */ |
e18b890bb
|
249 |
struct kmem_cache *sighand_cachep; |
1da177e4c
|
250 251 |
/* SLAB cache for files_struct structures (tsk->files) */ |
e18b890bb
|
252 |
struct kmem_cache *files_cachep; |
1da177e4c
|
253 254 |
/* SLAB cache for fs_struct structures (tsk->fs) */ |
e18b890bb
|
255 |
struct kmem_cache *fs_cachep; |
1da177e4c
|
256 257 |
/* SLAB cache for vm_area_struct structures */ |
e18b890bb
|
258 |
struct kmem_cache *vm_area_cachep; |
1da177e4c
|
259 260 |
/* SLAB cache for mm_struct structures (tsk->mm) */ |
e18b890bb
|
261 |
static struct kmem_cache *mm_cachep; |
1da177e4c
|
262 |
|
ba14a194a
|
263 |
static void account_kernel_stack(struct task_struct *tsk, int account) |
c6a7f5728
|
264 |
{ |
ba14a194a
|
265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 |
void *stack = task_stack_page(tsk); struct vm_struct *vm = task_stack_vm_area(tsk); BUILD_BUG_ON(IS_ENABLED(CONFIG_VMAP_STACK) && PAGE_SIZE % 1024 != 0); if (vm) { int i; BUG_ON(vm->nr_pages != THREAD_SIZE / PAGE_SIZE); for (i = 0; i < THREAD_SIZE / PAGE_SIZE; i++) { mod_zone_page_state(page_zone(vm->pages[i]), NR_KERNEL_STACK_KB, PAGE_SIZE / 1024 * account); } /* All stack pages belong to the same memcg. */ memcg_kmem_update_page_stat(vm->pages[0], MEMCG_KERNEL_STACK_KB, account * (THREAD_SIZE / 1024)); } else { /* * All stack pages are in the same zone and belong to the * same memcg. */ struct page *first_page = virt_to_page(stack); mod_zone_page_state(page_zone(first_page), NR_KERNEL_STACK_KB, THREAD_SIZE / 1024 * account); memcg_kmem_update_page_stat(first_page, MEMCG_KERNEL_STACK_KB, account * (THREAD_SIZE / 1024)); } |
c6a7f5728
|
297 |
} |
68f24b08e
|
298 |
static void release_task_stack(struct task_struct *tsk) |
1da177e4c
|
299 |
{ |
405c07597
|
300 301 |
if (WARN_ON(tsk->state != TASK_DEAD)) return; /* Better to leak the stack than to free prematurely */ |
ba14a194a
|
302 |
account_kernel_stack(tsk, -1); |
b235beea9
|
303 |
arch_release_thread_stack(tsk->stack); |
ba14a194a
|
304 |
free_thread_stack(tsk); |
68f24b08e
|
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 |
tsk->stack = NULL; #ifdef CONFIG_VMAP_STACK tsk->stack_vm_area = NULL; #endif } #ifdef CONFIG_THREAD_INFO_IN_TASK void put_task_stack(struct task_struct *tsk) { if (atomic_dec_and_test(&tsk->stack_refcount)) release_task_stack(tsk); } #endif void free_task(struct task_struct *tsk) { #ifndef CONFIG_THREAD_INFO_IN_TASK /* * The task is finally done with both the stack and thread_info, * so free both. */ release_task_stack(tsk); #else /* * If the task had a separate stack allocation, it should be gone * by now. */ WARN_ON_ONCE(atomic_read(&tsk->stack_refcount) != 0); #endif |
23f78d4a0
|
334 |
rt_mutex_debug_task_free(tsk); |
fb52607af
|
335 |
ftrace_graph_exit_task(tsk); |
e2cfabdfd
|
336 |
put_seccomp_filter(tsk); |
f19b9f74b
|
337 |
arch_release_task_struct(tsk); |
1da177e4c
|
338 339 340 |
free_task_struct(tsk); } EXPORT_SYMBOL(free_task); |
ea6d290ca
|
341 342 |
static inline void free_signal_struct(struct signal_struct *sig) { |
97101eb41
|
343 |
taskstats_tgid_free(sig); |
1c5354de9
|
344 |
sched_autogroup_exit(sig); |
7283094ec
|
345 346 347 348 |
/* * __mmdrop is not safe to call from softirq context on x86 due to * pgd_dtor so postpone it to the async context */ |
26db62f17
|
349 |
if (sig->oom_mm) |
7283094ec
|
350 |
mmdrop_async(sig->oom_mm); |
ea6d290ca
|
351 352 353 354 355 |
kmem_cache_free(signal_cachep, sig); } static inline void put_signal_struct(struct signal_struct *sig) { |
1c5354de9
|
356 |
if (atomic_dec_and_test(&sig->sigcnt)) |
ea6d290ca
|
357 358 |
free_signal_struct(sig); } |
158d9ebd1
|
359 |
void __put_task_struct(struct task_struct *tsk) |
1da177e4c
|
360 |
{ |
270f722d4
|
361 |
WARN_ON(!tsk->exit_state); |
1da177e4c
|
362 363 |
WARN_ON(atomic_read(&tsk->usage)); WARN_ON(tsk == current); |
2e91fa7f6
|
364 |
cgroup_free(tsk); |
156654f49
|
365 |
task_numa_free(tsk); |
1a2a4d06e
|
366 |
security_task_free(tsk); |
e0e817392
|
367 |
exit_creds(tsk); |
35df17c57
|
368 |
delayacct_tsk_free(tsk); |
ea6d290ca
|
369 |
put_signal_struct(tsk->signal); |
1da177e4c
|
370 371 372 373 |
if (!profile_handoff_task(tsk)) free_task(tsk); } |
77c100c83
|
374 |
EXPORT_SYMBOL_GPL(__put_task_struct); |
1da177e4c
|
375 |
|
6c0a9fa62
|
376 |
void __init __weak arch_task_cache_init(void) { } |
61c4628b5
|
377 |
|
ff691f6e0
|
378 379 380 |
/* * set_max_threads */ |
16db3d3f1
|
381 |
static void set_max_threads(unsigned int max_threads_suggested) |
ff691f6e0
|
382 |
{ |
ac1b398de
|
383 |
u64 threads; |
ff691f6e0
|
384 385 |
/* |
ac1b398de
|
386 387 |
* The number of threads shall be limited such that the thread * structures may only consume a small part of the available memory. |
ff691f6e0
|
388 |
*/ |
ac1b398de
|
389 390 391 392 393 |
if (fls64(totalram_pages) + fls64(PAGE_SIZE) > 64) threads = MAX_THREADS; else threads = div64_u64((u64) totalram_pages * (u64) PAGE_SIZE, (u64) THREAD_SIZE * 8UL); |
16db3d3f1
|
394 395 |
if (threads > max_threads_suggested) threads = max_threads_suggested; |
ac1b398de
|
396 |
max_threads = clamp_t(u64, threads, MIN_THREADS, MAX_THREADS); |
ff691f6e0
|
397 |
} |
5aaeb5c01
|
398 399 400 401 |
#ifdef CONFIG_ARCH_WANTS_DYNAMIC_TASK_STRUCT /* Initialized by the architecture: */ int arch_task_struct_size __read_mostly; #endif |
0c8c0f03e
|
402 |
|
ff691f6e0
|
403 |
void __init fork_init(void) |
1da177e4c
|
404 |
{ |
25f9c0817
|
405 |
int i; |
f5e102873
|
406 |
#ifndef CONFIG_ARCH_TASK_STRUCT_ALLOCATOR |
1da177e4c
|
407 408 409 410 |
#ifndef ARCH_MIN_TASKALIGN #define ARCH_MIN_TASKALIGN L1_CACHE_BYTES #endif /* create a slab on which task_structs can be allocated */ |
5d097056c
|
411 412 413 |
task_struct_cachep = kmem_cache_create("task_struct", arch_task_struct_size, ARCH_MIN_TASKALIGN, SLAB_PANIC|SLAB_NOTRACK|SLAB_ACCOUNT, NULL); |
1da177e4c
|
414 |
#endif |
61c4628b5
|
415 416 |
/* do the arch specific task caches init */ arch_task_cache_init(); |
16db3d3f1
|
417 |
set_max_threads(MAX_THREADS); |
1da177e4c
|
418 419 420 421 422 |
init_task.signal->rlim[RLIMIT_NPROC].rlim_cur = max_threads/2; init_task.signal->rlim[RLIMIT_NPROC].rlim_max = max_threads/2; init_task.signal->rlim[RLIMIT_SIGPENDING] = init_task.signal->rlim[RLIMIT_NPROC]; |
b376c3e1b
|
423 |
|
25f9c0817
|
424 425 426 |
for (i = 0; i < UCOUNT_COUNTS; i++) { init_user_ns.ucount_max[i] = max_threads/2; } |
1da177e4c
|
427 |
} |
52f5684c8
|
428 |
int __weak arch_dup_task_struct(struct task_struct *dst, |
61c4628b5
|
429 430 431 432 433 |
struct task_struct *src) { *dst = *src; return 0; } |
d4311ff1a
|
434 435 436 437 438 439 440 |
void set_task_stack_end_magic(struct task_struct *tsk) { unsigned long *stackend; stackend = end_of_stack(tsk); *stackend = STACK_END_MAGIC; /* for overflow detection */ } |
725fc629f
|
441 |
static struct task_struct *dup_task_struct(struct task_struct *orig, int node) |
1da177e4c
|
442 443 |
{ struct task_struct *tsk; |
b235beea9
|
444 |
unsigned long *stack; |
ba14a194a
|
445 |
struct vm_struct *stack_vm_area; |
3e26c149c
|
446 |
int err; |
1da177e4c
|
447 |
|
725fc629f
|
448 449 |
if (node == NUMA_NO_NODE) node = tsk_fork_get_node(orig); |
504f52b54
|
450 |
tsk = alloc_task_struct_node(node); |
1da177e4c
|
451 452 |
if (!tsk) return NULL; |
b235beea9
|
453 454 |
stack = alloc_thread_stack_node(tsk, node); if (!stack) |
f19b9f74b
|
455 |
goto free_tsk; |
1da177e4c
|
456 |
|
ba14a194a
|
457 |
stack_vm_area = task_stack_vm_area(tsk); |
fb0a685cb
|
458 |
err = arch_dup_task_struct(tsk, orig); |
ba14a194a
|
459 460 461 462 463 464 465 466 467 468 |
/* * arch_dup_task_struct() clobbers the stack-related fields. Make * sure they're properly initialized before using any stack-related * functions again. */ tsk->stack = stack; #ifdef CONFIG_VMAP_STACK tsk->stack_vm_area = stack_vm_area; #endif |
68f24b08e
|
469 470 471 |
#ifdef CONFIG_THREAD_INFO_IN_TASK atomic_set(&tsk->stack_refcount, 1); #endif |
ba14a194a
|
472 |
|
164c33c6a
|
473 |
if (err) |
b235beea9
|
474 |
goto free_stack; |
164c33c6a
|
475 |
|
dbd952127
|
476 477 478 479 480 481 482 483 484 |
#ifdef CONFIG_SECCOMP /* * We must handle setting up seccomp filters once we're under * the sighand lock in case orig has changed between now and * then. Until then, filter must be NULL to avoid messing up * the usage counts on the error path calling free_task. */ tsk->seccomp.filter = NULL; #endif |
87bec58a5
|
485 486 |
setup_thread_stack(tsk, orig); |
8e7cac798
|
487 |
clear_user_return_notifier(tsk); |
f26f9aff6
|
488 |
clear_tsk_need_resched(tsk); |
d4311ff1a
|
489 |
set_task_stack_end_magic(tsk); |
1da177e4c
|
490 |
|
0a4254058
|
491 492 493 |
#ifdef CONFIG_CC_STACKPROTECTOR tsk->stack_canary = get_random_int(); #endif |
fb0a685cb
|
494 495 496 497 498 |
/* * One for us, one for whoever does the "release_task()" (usually * parent) */ atomic_set(&tsk->usage, 2); |
6c5c93415
|
499 |
#ifdef CONFIG_BLK_DEV_IO_TRACE |
2056a782f
|
500 |
tsk->btrace_seq = 0; |
6c5c93415
|
501 |
#endif |
a0aa7f68a
|
502 |
tsk->splice_pipe = NULL; |
5640f7685
|
503 |
tsk->task_frag.page = NULL; |
093e5840a
|
504 |
tsk->wake_q.next = NULL; |
c6a7f5728
|
505 |
|
ba14a194a
|
506 |
account_kernel_stack(tsk, 1); |
c6a7f5728
|
507 |
|
5c9a8750a
|
508 |
kcov_task_init(tsk); |
1da177e4c
|
509 |
return tsk; |
61c4628b5
|
510 |
|
b235beea9
|
511 |
free_stack: |
ba14a194a
|
512 |
free_thread_stack(tsk); |
f19b9f74b
|
513 |
free_tsk: |
61c4628b5
|
514 515 |
free_task_struct(tsk); return NULL; |
1da177e4c
|
516 517 518 |
} #ifdef CONFIG_MMU |
0766f788e
|
519 520 |
static __latent_entropy int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) |
1da177e4c
|
521 |
{ |
297c5eee3
|
522 |
struct vm_area_struct *mpnt, *tmp, *prev, **pprev; |
1da177e4c
|
523 524 525 |
struct rb_node **rb_link, *rb_parent; int retval; unsigned long charge; |
1da177e4c
|
526 |
|
32cdba1e0
|
527 |
uprobe_start_dup_mmap(); |
7c0512679
|
528 529 530 531 |
if (down_write_killable(&oldmm->mmap_sem)) { retval = -EINTR; goto fail_uprobe_end; } |
ec8c0446b
|
532 |
flush_cache_dup_mm(oldmm); |
f8ac4ec9c
|
533 |
uprobe_dup_mmap(oldmm, mm); |
ad3394517
|
534 535 536 537 |
/* * Not linked in yet - no deadlock potential: */ down_write_nested(&mm->mmap_sem, SINGLE_DEPTH_NESTING); |
7ee782325
|
538 |
|
90f31d0ea
|
539 540 |
/* No ordering required: file already has been exposed. */ RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); |
4f7d46143
|
541 |
mm->total_vm = oldmm->total_vm; |
846383359
|
542 |
mm->data_vm = oldmm->data_vm; |
4f7d46143
|
543 544 |
mm->exec_vm = oldmm->exec_vm; mm->stack_vm = oldmm->stack_vm; |
1da177e4c
|
545 546 547 |
rb_link = &mm->mm_rb.rb_node; rb_parent = NULL; pprev = &mm->mmap; |
f8af4da3b
|
548 549 550 |
retval = ksm_fork(mm, oldmm); if (retval) goto out; |
ba76149f4
|
551 552 553 |
retval = khugepaged_fork(mm, oldmm); if (retval) goto out; |
1da177e4c
|
554 |
|
297c5eee3
|
555 |
prev = NULL; |
fd3e42fcc
|
556 |
for (mpnt = oldmm->mmap; mpnt; mpnt = mpnt->vm_next) { |
1da177e4c
|
557 558 559 |
struct file *file; if (mpnt->vm_flags & VM_DONTCOPY) { |
846383359
|
560 |
vm_stat_account(mm, mpnt->vm_flags, -vma_pages(mpnt)); |
1da177e4c
|
561 562 563 564 |
continue; } charge = 0; if (mpnt->vm_flags & VM_ACCOUNT) { |
b2412b7fa
|
565 |
unsigned long len = vma_pages(mpnt); |
191c54244
|
566 |
if (security_vm_enough_memory_mm(oldmm, len)) /* sic */ |
1da177e4c
|
567 568 569 |
goto fail_nomem; charge = len; } |
e94b17660
|
570 |
tmp = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL); |
1da177e4c
|
571 572 573 |
if (!tmp) goto fail_nomem; *tmp = *mpnt; |
5beb49305
|
574 |
INIT_LIST_HEAD(&tmp->anon_vma_chain); |
ef0855d33
|
575 576 |
retval = vma_dup_policy(mpnt, tmp); if (retval) |
1da177e4c
|
577 |
goto fail_nomem_policy; |
a247c3a97
|
578 |
tmp->vm_mm = mm; |
5beb49305
|
579 580 |
if (anon_vma_fork(tmp, mpnt)) goto fail_nomem_anon_vma_fork; |
de60f5f10
|
581 582 |
tmp->vm_flags &= ~(VM_LOCKED|VM_LOCKONFAULT|VM_UFFD_MISSING|VM_UFFD_WP); |
297c5eee3
|
583 |
tmp->vm_next = tmp->vm_prev = NULL; |
745f234be
|
584 |
tmp->vm_userfaultfd_ctx = NULL_VM_UFFD_CTX; |
1da177e4c
|
585 586 |
file = tmp->vm_file; if (file) { |
496ad9aa8
|
587 |
struct inode *inode = file_inode(file); |
b88ed2059
|
588 |
struct address_space *mapping = file->f_mapping; |
1da177e4c
|
589 590 591 |
get_file(file); if (tmp->vm_flags & VM_DENYWRITE) atomic_dec(&inode->i_writecount); |
83cde9e8b
|
592 |
i_mmap_lock_write(mapping); |
b88ed2059
|
593 |
if (tmp->vm_flags & VM_SHARED) |
4bb5f5d93
|
594 |
atomic_inc(&mapping->i_mmap_writable); |
b88ed2059
|
595 596 |
flush_dcache_mmap_lock(mapping); /* insert tmp into the share list, just after mpnt */ |
27ba0644e
|
597 598 |
vma_interval_tree_insert_after(tmp, mpnt, &mapping->i_mmap); |
b88ed2059
|
599 |
flush_dcache_mmap_unlock(mapping); |
83cde9e8b
|
600 |
i_mmap_unlock_write(mapping); |
1da177e4c
|
601 602 603 |
} /* |
a1e78772d
|
604 605 606 607 608 609 610 611 |
* Clear hugetlb-related page reserves for children. This only * affects MAP_PRIVATE mappings. Faults generated by the child * are not guaranteed to succeed, even if read-only */ if (is_vm_hugetlb_page(tmp)) reset_vma_resv_huge_pages(tmp); /* |
7ee782325
|
612 |
* Link in the new vma and copy the page table entries. |
1da177e4c
|
613 |
*/ |
1da177e4c
|
614 615 |
*pprev = tmp; pprev = &tmp->vm_next; |
297c5eee3
|
616 617 |
tmp->vm_prev = prev; prev = tmp; |
1da177e4c
|
618 619 620 621 622 623 |
__vma_link_rb(mm, tmp, rb_link, rb_parent); rb_link = &tmp->vm_rb.rb_right; rb_parent = &tmp->vm_rb; mm->map_count++; |
0b0db14c5
|
624 |
retval = copy_page_range(mm, oldmm, mpnt); |
1da177e4c
|
625 626 627 628 629 630 631 |
if (tmp->vm_ops && tmp->vm_ops->open) tmp->vm_ops->open(tmp); if (retval) goto out; } |
d6dd61c83
|
632 633 |
/* a new mm has just been created */ arch_dup_mmap(oldmm, mm); |
1da177e4c
|
634 |
retval = 0; |
1da177e4c
|
635 |
out: |
7ee782325
|
636 |
up_write(&mm->mmap_sem); |
fd3e42fcc
|
637 |
flush_tlb_mm(oldmm); |
1da177e4c
|
638 |
up_write(&oldmm->mmap_sem); |
7c0512679
|
639 |
fail_uprobe_end: |
32cdba1e0
|
640 |
uprobe_end_dup_mmap(); |
1da177e4c
|
641 |
return retval; |
5beb49305
|
642 |
fail_nomem_anon_vma_fork: |
ef0855d33
|
643 |
mpol_put(vma_policy(tmp)); |
1da177e4c
|
644 645 646 647 648 649 650 |
fail_nomem_policy: kmem_cache_free(vm_area_cachep, tmp); fail_nomem: retval = -ENOMEM; vm_unacct_memory(charge); goto out; } |
fb0a685cb
|
651 |
static inline int mm_alloc_pgd(struct mm_struct *mm) |
1da177e4c
|
652 653 654 655 656 657 |
{ mm->pgd = pgd_alloc(mm); if (unlikely(!mm->pgd)) return -ENOMEM; return 0; } |
fb0a685cb
|
658 |
static inline void mm_free_pgd(struct mm_struct *mm) |
1da177e4c
|
659 |
{ |
5e5419734
|
660 |
pgd_free(mm, mm->pgd); |
1da177e4c
|
661 662 |
} #else |
90f31d0ea
|
663 664 665 666 667 668 669 |
static int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm) { down_write(&oldmm->mmap_sem); RCU_INIT_POINTER(mm->exe_file, get_mm_exe_file(oldmm)); up_write(&oldmm->mmap_sem); return 0; } |
1da177e4c
|
670 671 672 |
#define mm_alloc_pgd(mm) (0) #define mm_free_pgd(mm) #endif /* CONFIG_MMU */ |
23ff44402
|
673 |
__cacheline_aligned_in_smp DEFINE_SPINLOCK(mmlist_lock); |
1da177e4c
|
674 |
|
e94b17660
|
675 |
#define allocate_mm() (kmem_cache_alloc(mm_cachep, GFP_KERNEL)) |
1da177e4c
|
676 |
#define free_mm(mm) (kmem_cache_free(mm_cachep, (mm))) |
4cb0e11b1
|
677 678 679 680 681 682 683 684 685 686 687 |
static unsigned long default_dump_filter = MMF_DUMP_FILTER_DEFAULT; static int __init coredump_filter_setup(char *s) { default_dump_filter = (simple_strtoul(s, NULL, 0) << MMF_DUMP_FILTER_SHIFT) & MMF_DUMP_FILTER_MASK; return 1; } __setup("coredump_filter=", coredump_filter_setup); |
1da177e4c
|
688 |
#include <linux/init_task.h> |
858f09930
|
689 690 691 692 |
static void mm_init_aio(struct mm_struct *mm) { #ifdef CONFIG_AIO spin_lock_init(&mm->ioctx_lock); |
db446a08c
|
693 |
mm->ioctx_table = NULL; |
858f09930
|
694 695 |
#endif } |
33144e842
|
696 697 698 699 700 701 |
static void mm_init_owner(struct mm_struct *mm, struct task_struct *p) { #ifdef CONFIG_MEMCG mm->owner = p; #endif } |
694a95fa6
|
702 703 |
static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p, struct user_namespace *user_ns) |
1da177e4c
|
704 |
{ |
41f727fde
|
705 706 707 |
mm->mmap = NULL; mm->mm_rb = RB_ROOT; mm->vmacache_seqnum = 0; |
1da177e4c
|
708 709 710 711 |
atomic_set(&mm->mm_users, 1); atomic_set(&mm->mm_count, 1); init_rwsem(&mm->mmap_sem); INIT_LIST_HEAD(&mm->mmlist); |
999d9fc16
|
712 |
mm->core_state = NULL; |
e1f56c89b
|
713 |
atomic_long_set(&mm->nr_ptes, 0); |
2d2f5119b
|
714 |
mm_nr_pmds_init(mm); |
41f727fde
|
715 716 |
mm->map_count = 0; mm->locked_vm = 0; |
ce65cefa5
|
717 |
mm->pinned_vm = 0; |
d559db086
|
718 |
memset(&mm->rss_stat, 0, sizeof(mm->rss_stat)); |
1da177e4c
|
719 |
spin_lock_init(&mm->page_table_lock); |
41f727fde
|
720 |
mm_init_cpumask(mm); |
858f09930
|
721 |
mm_init_aio(mm); |
cf475ad28
|
722 |
mm_init_owner(mm, p); |
41f727fde
|
723 |
mmu_notifier_mm_init(mm); |
208414059
|
724 |
clear_tlb_flush_pending(mm); |
41f727fde
|
725 726 727 |
#if defined(CONFIG_TRANSPARENT_HUGEPAGE) && !USE_SPLIT_PMD_PTLOCKS mm->pmd_huge_pte = NULL; #endif |
1da177e4c
|
728 |
|
a0715cc22
|
729 730 731 732 733 |
if (current->mm) { mm->flags = current->mm->flags & MMF_INIT_MASK; mm->def_flags = current->mm->def_flags & VM_INIT_DEF_MASK; } else { mm->flags = default_dump_filter; |
1da177e4c
|
734 |
mm->def_flags = 0; |
a0715cc22
|
735 |
} |
41f727fde
|
736 737 738 739 740 |
if (mm_alloc_pgd(mm)) goto fail_nopgd; if (init_new_context(p, mm)) goto fail_nocontext; |
78fb74669
|
741 |
|
694a95fa6
|
742 |
mm->user_ns = get_user_ns(user_ns); |
41f727fde
|
743 744 745 746 747 |
return mm; fail_nocontext: mm_free_pgd(mm); fail_nopgd: |
1da177e4c
|
748 749 750 |
free_mm(mm); return NULL; } |
c3f0327f8
|
751 752 753 754 755 756 757 758 759 760 761 762 |
static void check_mm(struct mm_struct *mm) { int i; for (i = 0; i < NR_MM_COUNTERS; i++) { long x = atomic_long_read(&mm->rss_stat.count[i]); if (unlikely(x)) printk(KERN_ALERT "BUG: Bad rss-counter state " "mm:%p idx:%d val:%ld ", mm, i, x); } |
b30fe6c7c
|
763 764 765 766 767 768 769 770 771 |
if (atomic_long_read(&mm->nr_ptes)) pr_alert("BUG: non-zero nr_ptes on freeing mm: %ld ", atomic_long_read(&mm->nr_ptes)); if (mm_nr_pmds(mm)) pr_alert("BUG: non-zero nr_pmds on freeing mm: %ld ", mm_nr_pmds(mm)); |
e009bb30c
|
772 |
#if defined(CONFIG_TRANSPARENT_HUGEPAGE) && !USE_SPLIT_PMD_PTLOCKS |
96dad67ff
|
773 |
VM_BUG_ON_MM(mm->pmd_huge_pte, mm); |
c3f0327f8
|
774 775 |
#endif } |
1da177e4c
|
776 777 778 |
/* * Allocate and initialize an mm_struct. */ |
fb0a685cb
|
779 |
struct mm_struct *mm_alloc(void) |
1da177e4c
|
780 |
{ |
fb0a685cb
|
781 |
struct mm_struct *mm; |
1da177e4c
|
782 783 |
mm = allocate_mm(); |
de03c72cf
|
784 785 786 787 |
if (!mm) return NULL; memset(mm, 0, sizeof(*mm)); |
694a95fa6
|
788 |
return mm_init(mm, current, current_user_ns()); |
1da177e4c
|
789 790 791 792 793 794 795 |
} /* * Called when the last reference to the mm * is dropped: either by a lazy thread or by * mmput. Free the page directory and the mm. */ |
7ad5b3a50
|
796 |
void __mmdrop(struct mm_struct *mm) |
1da177e4c
|
797 798 799 800 |
{ BUG_ON(mm == &init_mm); mm_free_pgd(mm); destroy_context(mm); |
cddb8a5c1
|
801 |
mmu_notifier_mm_destroy(mm); |
c3f0327f8
|
802 |
check_mm(mm); |
694a95fa6
|
803 |
put_user_ns(mm->user_ns); |
1da177e4c
|
804 805 |
free_mm(mm); } |
6d4e4c4fc
|
806 |
EXPORT_SYMBOL_GPL(__mmdrop); |
1da177e4c
|
807 |
|
ec8d7c14e
|
808 809 810 811 812 813 814 815 816 |
static inline void __mmput(struct mm_struct *mm) { VM_BUG_ON(atomic_read(&mm->mm_users)); uprobe_clear_state(mm); exit_aio(mm); ksm_exit(mm); khugepaged_exit(mm); /* must run before exit_mmap */ exit_mmap(mm); |
6fcb52a56
|
817 |
mm_put_huge_zero_page(mm); |
ec8d7c14e
|
818 819 820 821 822 823 824 825 |
set_mm_exe_file(mm, NULL); if (!list_empty(&mm->mmlist)) { spin_lock(&mmlist_lock); list_del(&mm->mmlist); spin_unlock(&mmlist_lock); } if (mm->binfmt) module_put(mm->binfmt->module); |
862e3073b
|
826 |
set_bit(MMF_OOM_SKIP, &mm->flags); |
ec8d7c14e
|
827 828 |
mmdrop(mm); } |
1da177e4c
|
829 830 831 832 833 |
/* * Decrement the use count and release all resources for an mm. */ void mmput(struct mm_struct *mm) { |
0ae26f1b3
|
834 |
might_sleep(); |
ec8d7c14e
|
835 836 837 838 |
if (atomic_dec_and_test(&mm->mm_users)) __mmput(mm); } EXPORT_SYMBOL_GPL(mmput); |
7ef949d77
|
839 |
#ifdef CONFIG_MMU |
ec8d7c14e
|
840 841 842 843 844 845 846 847 |
static void mmput_async_fn(struct work_struct *work) { struct mm_struct *mm = container_of(work, struct mm_struct, async_put_work); __mmput(mm); } void mmput_async(struct mm_struct *mm) { |
1da177e4c
|
848 |
if (atomic_dec_and_test(&mm->mm_users)) { |
ec8d7c14e
|
849 850 |
INIT_WORK(&mm->async_put_work, mmput_async_fn); schedule_work(&mm->async_put_work); |
1da177e4c
|
851 852 |
} } |
7ef949d77
|
853 |
#endif |
1da177e4c
|
854 |
|
90f31d0ea
|
855 856 857 858 859 |
/** * set_mm_exe_file - change a reference to the mm's executable file * * This changes mm's executable file (shown as symlink /proc/[pid]/exe). * |
6e399cd14
|
860 861 862 863 864 |
* Main users are mmput() and sys_execve(). Callers prevent concurrent * invocations: in mmput() nobody alive left, in execve task is single * threaded. sys_prctl(PR_SET_MM_MAP/EXE_FILE) also needs to set the * mm->exe_file, but does so without using set_mm_exe_file() in order * to do avoid the need for any locks. |
90f31d0ea
|
865 |
*/ |
386460138
|
866 867 |
void set_mm_exe_file(struct mm_struct *mm, struct file *new_exe_file) { |
6e399cd14
|
868 869 870 871 872 873 874 875 |
struct file *old_exe_file; /* * It is safe to dereference the exe_file without RCU as * this function is only called if nobody else can access * this mm -- see comment above for justification. */ old_exe_file = rcu_dereference_raw(mm->exe_file); |
90f31d0ea
|
876 |
|
386460138
|
877 878 |
if (new_exe_file) get_file(new_exe_file); |
90f31d0ea
|
879 880 881 |
rcu_assign_pointer(mm->exe_file, new_exe_file); if (old_exe_file) fput(old_exe_file); |
386460138
|
882 |
} |
90f31d0ea
|
883 884 885 886 887 888 |
/** * get_mm_exe_file - acquire a reference to the mm's executable file * * Returns %NULL if mm has no associated executable file. * User must release file via fput(). */ |
386460138
|
889 890 891 |
struct file *get_mm_exe_file(struct mm_struct *mm) { struct file *exe_file; |
90f31d0ea
|
892 893 894 895 896 |
rcu_read_lock(); exe_file = rcu_dereference(mm->exe_file); if (exe_file && !get_file_rcu(exe_file)) exe_file = NULL; rcu_read_unlock(); |
386460138
|
897 898 |
return exe_file; } |
11163348a
|
899 |
EXPORT_SYMBOL(get_mm_exe_file); |
386460138
|
900 |
|
1da177e4c
|
901 |
/** |
cd81a9170
|
902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 |
* get_task_exe_file - acquire a reference to the task's executable file * * Returns %NULL if task's mm (if any) has no associated executable file or * this is a kernel thread with borrowed mm (see the comment above get_task_mm). * User must release file via fput(). */ struct file *get_task_exe_file(struct task_struct *task) { struct file *exe_file = NULL; struct mm_struct *mm; task_lock(task); mm = task->mm; if (mm) { if (!(task->flags & PF_KTHREAD)) exe_file = get_mm_exe_file(mm); } task_unlock(task); return exe_file; } EXPORT_SYMBOL(get_task_exe_file); |
386460138
|
923 |
|
1da177e4c
|
924 925 926 |
/** * get_task_mm - acquire a reference to the task's mm * |
246bb0b1d
|
927 |
* Returns %NULL if the task has no mm. Checks PF_KTHREAD (meaning |
1da177e4c
|
928 929 930 931 932 933 934 935 936 937 938 939 |
* this kernel workthread has transiently adopted a user mm with use_mm, * to do its AIO) is not set and if so returns a reference to it, after * bumping up the use count. User must release the mm via mmput() * after use. Typically used by /proc and ptrace. */ struct mm_struct *get_task_mm(struct task_struct *task) { struct mm_struct *mm; task_lock(task); mm = task->mm; if (mm) { |
246bb0b1d
|
940 |
if (task->flags & PF_KTHREAD) |
1da177e4c
|
941 942 943 944 945 946 947 948 |
mm = NULL; else atomic_inc(&mm->mm_users); } task_unlock(task); return mm; } EXPORT_SYMBOL_GPL(get_task_mm); |
8cdb878dc
|
949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 |
struct mm_struct *mm_access(struct task_struct *task, unsigned int mode) { struct mm_struct *mm; int err; err = mutex_lock_killable(&task->signal->cred_guard_mutex); if (err) return ERR_PTR(err); mm = get_task_mm(task); if (mm && mm != current->mm && !ptrace_may_access(task, mode)) { mmput(mm); mm = ERR_PTR(-EACCES); } mutex_unlock(&task->signal->cred_guard_mutex); return mm; } |
57b59c4a1
|
968 |
static void complete_vfork_done(struct task_struct *tsk) |
c415c3b47
|
969 |
{ |
d68b46fe1
|
970 |
struct completion *vfork; |
c415c3b47
|
971 |
|
d68b46fe1
|
972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 |
task_lock(tsk); vfork = tsk->vfork_done; if (likely(vfork)) { tsk->vfork_done = NULL; complete(vfork); } task_unlock(tsk); } static int wait_for_vfork_done(struct task_struct *child, struct completion *vfork) { int killed; freezer_do_not_count(); killed = wait_for_completion_killable(vfork); freezer_count(); if (killed) { task_lock(child); child->vfork_done = NULL; task_unlock(child); } put_task_struct(child); return killed; |
c415c3b47
|
998 |
} |
1da177e4c
|
999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 |
/* Please note the differences between mmput and mm_release. * mmput is called whenever we stop holding onto a mm_struct, * error success whatever. * * mm_release is called after a mm_struct has been removed * from the current process. * * This difference is important for error handling, when we * only half set up a mm_struct for a new process and need to restore * the old one. Because we mmput the new mm_struct before * restoring the old one. . . * Eric Biederman 10 January 1998 */ void mm_release(struct task_struct *tsk, struct mm_struct *mm) { |
8141c7f3e
|
1014 1015 |
/* Get rid of any futexes when releasing the mm */ #ifdef CONFIG_FUTEX |
fc6b177de
|
1016 |
if (unlikely(tsk->robust_list)) { |
8141c7f3e
|
1017 |
exit_robust_list(tsk); |
fc6b177de
|
1018 1019 |
tsk->robust_list = NULL; } |
8141c7f3e
|
1020 |
#ifdef CONFIG_COMPAT |
fc6b177de
|
1021 |
if (unlikely(tsk->compat_robust_list)) { |
8141c7f3e
|
1022 |
compat_exit_robust_list(tsk); |
fc6b177de
|
1023 1024 |
tsk->compat_robust_list = NULL; } |
8141c7f3e
|
1025 |
#endif |
322a2c100
|
1026 1027 |
if (unlikely(!list_empty(&tsk->pi_state_list))) exit_pi_state_list(tsk); |
8141c7f3e
|
1028 |
#endif |
0326f5a94
|
1029 |
uprobe_free_utask(tsk); |
1da177e4c
|
1030 1031 |
/* Get rid of any cached register state */ deactivate_mm(tsk, mm); |
fec1d0115
|
1032 |
/* |
735f2770a
|
1033 1034 1035 |
* Signal userspace if we're not exiting with a core dump * because we want to leave the value intact for debugging * purposes. |
fec1d0115
|
1036 |
*/ |
9c8a8228d
|
1037 |
if (tsk->clear_child_tid) { |
735f2770a
|
1038 |
if (!(tsk->signal->flags & SIGNAL_GROUP_COREDUMP) && |
9c8a8228d
|
1039 1040 1041 1042 1043 1044 1045 1046 1047 |
atomic_read(&mm->mm_users) > 1) { /* * We don't check the error code - if userspace has * not set up a proper pointer then tough luck. */ put_user(0, tsk->clear_child_tid); sys_futex(tsk->clear_child_tid, FUTEX_WAKE, 1, NULL, NULL, 0); } |
1da177e4c
|
1048 |
tsk->clear_child_tid = NULL; |
1da177e4c
|
1049 |
} |
f7505d64f
|
1050 1051 1052 1053 1054 1055 1056 |
/* * All done, finally we can wake up parent and return this mm to him. * Also kthread_stop() uses this completion for synchronization. */ if (tsk->vfork_done) complete_vfork_done(tsk); |
1da177e4c
|
1057 |
} |
a0a7ec308
|
1058 1059 1060 1061 |
/* * Allocate a new mm structure and copy contents from the * mm structure of the passed in task structure. */ |
ff252c1fc
|
1062 |
static struct mm_struct *dup_mm(struct task_struct *tsk) |
a0a7ec308
|
1063 1064 1065 |
{ struct mm_struct *mm, *oldmm = current->mm; int err; |
a0a7ec308
|
1066 1067 1068 1069 1070 |
mm = allocate_mm(); if (!mm) goto fail_nomem; memcpy(mm, oldmm, sizeof(*mm)); |
694a95fa6
|
1071 |
if (!mm_init(mm, tsk, mm->user_ns)) |
a0a7ec308
|
1072 |
goto fail_nomem; |
a0a7ec308
|
1073 1074 1075 1076 1077 1078 |
err = dup_mmap(mm, oldmm); if (err) goto free_pt; mm->hiwater_rss = get_mm_rss(mm); mm->hiwater_vm = mm->total_vm; |
801460d0c
|
1079 1080 |
if (mm->binfmt && !try_module_get(mm->binfmt->module)) goto free_pt; |
a0a7ec308
|
1081 1082 1083 |
return mm; free_pt: |
801460d0c
|
1084 1085 |
/* don't put binfmt in mmput, we haven't got module yet */ mm->binfmt = NULL; |
a0a7ec308
|
1086 1087 1088 1089 |
mmput(mm); fail_nomem: return NULL; |
a0a7ec308
|
1090 |
} |
fb0a685cb
|
1091 |
static int copy_mm(unsigned long clone_flags, struct task_struct *tsk) |
1da177e4c
|
1092 |
{ |
fb0a685cb
|
1093 |
struct mm_struct *mm, *oldmm; |
1da177e4c
|
1094 1095 1096 1097 |
int retval; tsk->min_flt = tsk->maj_flt = 0; tsk->nvcsw = tsk->nivcsw = 0; |
17406b82d
|
1098 1099 1100 |
#ifdef CONFIG_DETECT_HUNG_TASK tsk->last_switch_count = tsk->nvcsw + tsk->nivcsw; #endif |
1da177e4c
|
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 |
tsk->mm = NULL; tsk->active_mm = NULL; /* * Are we cloning a kernel thread? * * We need to steal a active VM for that.. */ oldmm = current->mm; if (!oldmm) return 0; |
615d6e875
|
1113 1114 |
/* initialize the new vmacache entries */ vmacache_flush(tsk); |
1da177e4c
|
1115 1116 1117 |
if (clone_flags & CLONE_VM) { atomic_inc(&oldmm->mm_users); mm = oldmm; |
1da177e4c
|
1118 1119 1120 1121 |
goto good_mm; } retval = -ENOMEM; |
a0a7ec308
|
1122 |
mm = dup_mm(tsk); |
1da177e4c
|
1123 1124 |
if (!mm) goto fail_nomem; |
1da177e4c
|
1125 1126 1127 1128 |
good_mm: tsk->mm = mm; tsk->active_mm = mm; return 0; |
1da177e4c
|
1129 1130 |
fail_nomem: return retval; |
1da177e4c
|
1131 |
} |
a39bc5169
|
1132 |
static int copy_fs(unsigned long clone_flags, struct task_struct *tsk) |
1da177e4c
|
1133 |
{ |
498052bba
|
1134 |
struct fs_struct *fs = current->fs; |
1da177e4c
|
1135 |
if (clone_flags & CLONE_FS) { |
498052bba
|
1136 |
/* tsk->fs is already what we want */ |
2a4419b5b
|
1137 |
spin_lock(&fs->lock); |
498052bba
|
1138 |
if (fs->in_exec) { |
2a4419b5b
|
1139 |
spin_unlock(&fs->lock); |
498052bba
|
1140 1141 1142 |
return -EAGAIN; } fs->users++; |
2a4419b5b
|
1143 |
spin_unlock(&fs->lock); |
1da177e4c
|
1144 1145 |
return 0; } |
498052bba
|
1146 |
tsk->fs = copy_fs_struct(fs); |
1da177e4c
|
1147 1148 1149 1150 |
if (!tsk->fs) return -ENOMEM; return 0; } |
fb0a685cb
|
1151 |
static int copy_files(unsigned long clone_flags, struct task_struct *tsk) |
a016f3389
|
1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 |
{ struct files_struct *oldf, *newf; int error = 0; /* * A background process may not have any files ... */ oldf = current->files; if (!oldf) goto out; if (clone_flags & CLONE_FILES) { atomic_inc(&oldf->count); goto out; } |
a016f3389
|
1167 1168 1169 1170 1171 1172 1173 1174 1175 |
newf = dup_fd(oldf, &error); if (!newf) goto out; tsk->files = newf; error = 0; out: return error; } |
fadad878c
|
1176 |
static int copy_io(unsigned long clone_flags, struct task_struct *tsk) |
fd0928df9
|
1177 1178 1179 |
{ #ifdef CONFIG_BLOCK struct io_context *ioc = current->io_context; |
6e736be7f
|
1180 |
struct io_context *new_ioc; |
fd0928df9
|
1181 1182 1183 |
if (!ioc) return 0; |
fadad878c
|
1184 1185 1186 1187 |
/* * Share io context with parent, if CLONE_IO is set */ if (clone_flags & CLONE_IO) { |
3d48749d9
|
1188 1189 |
ioc_task_link(ioc); tsk->io_context = ioc; |
fadad878c
|
1190 |
} else if (ioprio_valid(ioc->ioprio)) { |
6e736be7f
|
1191 1192 |
new_ioc = get_task_io_context(tsk, GFP_KERNEL, NUMA_NO_NODE); if (unlikely(!new_ioc)) |
fd0928df9
|
1193 |
return -ENOMEM; |
6e736be7f
|
1194 |
new_ioc->ioprio = ioc->ioprio; |
11a3122f6
|
1195 |
put_io_context(new_ioc); |
fd0928df9
|
1196 1197 1198 1199 |
} #endif return 0; } |
a39bc5169
|
1200 |
static int copy_sighand(unsigned long clone_flags, struct task_struct *tsk) |
1da177e4c
|
1201 1202 |
{ struct sighand_struct *sig; |
60348802e
|
1203 |
if (clone_flags & CLONE_SIGHAND) { |
1da177e4c
|
1204 1205 1206 1207 |
atomic_inc(¤t->sighand->count); return 0; } sig = kmem_cache_alloc(sighand_cachep, GFP_KERNEL); |
e56d09031
|
1208 |
rcu_assign_pointer(tsk->sighand, sig); |
1da177e4c
|
1209 1210 |
if (!sig) return -ENOMEM; |
9d7fb0427
|
1211 |
|
1da177e4c
|
1212 1213 1214 1215 |
atomic_set(&sig->count, 1); memcpy(sig->action, current->sighand->action, sizeof(sig->action)); return 0; } |
a7e5328a0
|
1216 |
void __cleanup_sighand(struct sighand_struct *sighand) |
c81addc9d
|
1217 |
{ |
d80e731ec
|
1218 1219 |
if (atomic_dec_and_test(&sighand->count)) { signalfd_cleanup(sighand); |
392809b25
|
1220 1221 1222 1223 |
/* * sighand_cachep is SLAB_DESTROY_BY_RCU so we can free it * without an RCU grace period, see __lock_task_sighand(). */ |
c81addc9d
|
1224 |
kmem_cache_free(sighand_cachep, sighand); |
d80e731ec
|
1225 |
} |
c81addc9d
|
1226 |
} |
f06febc96
|
1227 1228 1229 1230 1231 |
/* * Initialize POSIX timer handling for a thread group. */ static void posix_cpu_timers_init_group(struct signal_struct *sig) { |
78d7d407b
|
1232 |
unsigned long cpu_limit; |
316c1608d
|
1233 |
cpu_limit = READ_ONCE(sig->rlim[RLIMIT_CPU].rlim_cur); |
78d7d407b
|
1234 1235 |
if (cpu_limit != RLIM_INFINITY) { sig->cputime_expires.prof_exp = secs_to_cputime(cpu_limit); |
d5c373eb5
|
1236 |
sig->cputimer.running = true; |
6279a751f
|
1237 |
} |
f06febc96
|
1238 1239 1240 1241 1242 |
/* The timer lists. */ INIT_LIST_HEAD(&sig->cpu_timers[0]); INIT_LIST_HEAD(&sig->cpu_timers[1]); INIT_LIST_HEAD(&sig->cpu_timers[2]); } |
a39bc5169
|
1243 |
static int copy_signal(unsigned long clone_flags, struct task_struct *tsk) |
1da177e4c
|
1244 1245 |
{ struct signal_struct *sig; |
1da177e4c
|
1246 |
|
4ab6c0833
|
1247 |
if (clone_flags & CLONE_THREAD) |
490dea45d
|
1248 |
return 0; |
490dea45d
|
1249 |
|
a56704ef6
|
1250 |
sig = kmem_cache_zalloc(signal_cachep, GFP_KERNEL); |
1da177e4c
|
1251 1252 1253 |
tsk->signal = sig; if (!sig) return -ENOMEM; |
b3ac022cb
|
1254 |
sig->nr_threads = 1; |
1da177e4c
|
1255 |
atomic_set(&sig->live, 1); |
b3ac022cb
|
1256 |
atomic_set(&sig->sigcnt, 1); |
0c740d0af
|
1257 1258 1259 1260 |
/* list_add(thread_node, thread_head) without INIT_LIST_HEAD() */ sig->thread_head = (struct list_head)LIST_HEAD_INIT(tsk->thread_node); tsk->thread_node = (struct list_head)LIST_HEAD_INIT(sig->thread_head); |
1da177e4c
|
1261 |
init_waitqueue_head(&sig->wait_chldexit); |
db51aeccd
|
1262 |
sig->curr_target = tsk; |
1da177e4c
|
1263 1264 |
init_sigpending(&sig->shared_pending); INIT_LIST_HEAD(&sig->posix_timers); |
e78c34967
|
1265 |
seqlock_init(&sig->stats_lock); |
9d7fb0427
|
1266 |
prev_cputime_init(&sig->prev_cputime); |
1da177e4c
|
1267 |
|
c9cb2e3d7
|
1268 |
hrtimer_init(&sig->real_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); |
1da177e4c
|
1269 |
sig->real_timer.function = it_real_fn; |
1da177e4c
|
1270 |
|
1da177e4c
|
1271 1272 1273 |
task_lock(current->group_leader); memcpy(sig->rlim, current->signal->rlim, sizeof sig->rlim); task_unlock(current->group_leader); |
6279a751f
|
1274 |
posix_cpu_timers_init_group(sig); |
522ed7767
|
1275 |
tty_audit_fork(sig); |
5091faa44
|
1276 |
sched_autogroup_fork(sig); |
522ed7767
|
1277 |
|
a63d83f42
|
1278 |
sig->oom_score_adj = current->signal->oom_score_adj; |
dabb16f63
|
1279 |
sig->oom_score_adj_min = current->signal->oom_score_adj_min; |
28b83c519
|
1280 |
|
ebec18a6d
|
1281 1282 |
sig->has_child_subreaper = current->signal->has_child_subreaper || current->signal->is_child_subreaper; |
9b1bf12d5
|
1283 |
mutex_init(&sig->cred_guard_mutex); |
1da177e4c
|
1284 1285 |
return 0; } |
dbd952127
|
1286 1287 1288 1289 1290 1291 1292 1293 1294 |
static void copy_seccomp(struct task_struct *p) { #ifdef CONFIG_SECCOMP /* * Must be called with sighand->lock held, which is common to * all threads in the group. Holding cred_guard_mutex is not * needed because this new task is not yet running and cannot * be racing exec. */ |
69f6a34bd
|
1295 |
assert_spin_locked(¤t->sighand->siglock); |
dbd952127
|
1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 |
/* Ref-count the new filter user, and assign it. */ get_seccomp_filter(current); p->seccomp = current->seccomp; /* * Explicitly enable no_new_privs here in case it got set * between the task_struct being duplicated and holding the * sighand lock. The seccomp state and nnp must be in sync. */ if (task_no_new_privs(current)) task_set_no_new_privs(p); /* * If the parent gained a seccomp mode after copying thread * flags and between before we held the sighand lock, we have * to manually enable the seccomp thread flag here. */ if (p->seccomp.mode != SECCOMP_MODE_DISABLED) set_tsk_thread_flag(p, TIF_SECCOMP); #endif } |
17da2bd90
|
1318 |
SYSCALL_DEFINE1(set_tid_address, int __user *, tidptr) |
1da177e4c
|
1319 1320 |
{ current->clear_child_tid = tidptr; |
b488893a3
|
1321 |
return task_pid_vnr(current); |
1da177e4c
|
1322 |
} |
a39bc5169
|
1323 |
static void rt_mutex_init_task(struct task_struct *p) |
23f78d4a0
|
1324 |
{ |
1d6154825
|
1325 |
raw_spin_lock_init(&p->pi_lock); |
e29e175b0
|
1326 |
#ifdef CONFIG_RT_MUTEXES |
fb00aca47
|
1327 1328 |
p->pi_waiters = RB_ROOT; p->pi_waiters_leftmost = NULL; |
23f78d4a0
|
1329 |
p->pi_blocked_on = NULL; |
23f78d4a0
|
1330 1331 |
#endif } |
1da177e4c
|
1332 |
/* |
f06febc96
|
1333 1334 1335 1336 |
* Initialize POSIX timer handling for a single task. */ static void posix_cpu_timers_init(struct task_struct *tsk) { |
648616343
|
1337 1338 |
tsk->cputime_expires.prof_exp = 0; tsk->cputime_expires.virt_exp = 0; |
f06febc96
|
1339 1340 1341 1342 1343 |
tsk->cputime_expires.sched_exp = 0; INIT_LIST_HEAD(&tsk->cpu_timers[0]); INIT_LIST_HEAD(&tsk->cpu_timers[1]); INIT_LIST_HEAD(&tsk->cpu_timers[2]); } |
819077398
|
1344 1345 1346 1347 1348 |
static inline void init_task_pid(struct task_struct *task, enum pid_type type, struct pid *pid) { task->pids[type].pid = pid; } |
f06febc96
|
1349 |
/* |
1da177e4c
|
1350 1351 1352 1353 1354 1355 1356 |
* This creates a new process as a copy of the old one, * but does not actually start it yet. * * It copies the registers, and all the appropriate * parts of the process environment (as per the clone * flags). The actual kick-off is left to the caller. */ |
0766f788e
|
1357 1358 |
static __latent_entropy struct task_struct *copy_process( unsigned long clone_flags, |
36c8b5868
|
1359 |
unsigned long stack_start, |
36c8b5868
|
1360 |
unsigned long stack_size, |
36c8b5868
|
1361 |
int __user *child_tidptr, |
09a05394f
|
1362 |
struct pid *pid, |
3033f14ab
|
1363 |
int trace, |
725fc629f
|
1364 1365 |
unsigned long tls, int node) |
1da177e4c
|
1366 1367 |
{ int retval; |
a24efe62d
|
1368 |
struct task_struct *p; |
1da177e4c
|
1369 1370 1371 |
if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS)) return ERR_PTR(-EINVAL); |
e66eded83
|
1372 1373 |
if ((clone_flags & (CLONE_NEWUSER|CLONE_FS)) == (CLONE_NEWUSER|CLONE_FS)) return ERR_PTR(-EINVAL); |
1da177e4c
|
1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 |
/* * Thread groups must share signals as well, and detached threads * can only be started up within the thread group. */ if ((clone_flags & CLONE_THREAD) && !(clone_flags & CLONE_SIGHAND)) return ERR_PTR(-EINVAL); /* * Shared signal handlers imply shared VM. By way of the above, * thread groups also imply shared VM. Blocking this case allows * for various simplifications in other code. */ if ((clone_flags & CLONE_SIGHAND) && !(clone_flags & CLONE_VM)) return ERR_PTR(-EINVAL); |
123be07b0
|
1388 1389 1390 1391 1392 1393 1394 1395 1396 |
/* * Siblings of global init remain as zombies on exit since they are * not reaped by their parent (swapper). To solve this and to avoid * multi-rooted process trees, prevent global and container-inits * from creating siblings. */ if ((clone_flags & CLONE_PARENT) && current->signal->flags & SIGNAL_UNKILLABLE) return ERR_PTR(-EINVAL); |
8382fcac1
|
1397 |
/* |
40a0d32d1
|
1398 |
* If the new process will be in a different pid or user namespace |
faf00da54
|
1399 |
* do not allow it to share a thread group with the forking task. |
8382fcac1
|
1400 |
*/ |
faf00da54
|
1401 |
if (clone_flags & CLONE_THREAD) { |
40a0d32d1
|
1402 1403 1404 1405 1406 |
if ((clone_flags & (CLONE_NEWUSER | CLONE_NEWPID)) || (task_active_pid_ns(current) != current->nsproxy->pid_ns_for_children)) return ERR_PTR(-EINVAL); } |
8382fcac1
|
1407 |
|
1da177e4c
|
1408 1409 1410 1411 1412 |
retval = security_task_create(clone_flags); if (retval) goto fork_out; retval = -ENOMEM; |
725fc629f
|
1413 |
p = dup_task_struct(current, node); |
1da177e4c
|
1414 1415 |
if (!p) goto fork_out; |
f7e8b616e
|
1416 |
ftrace_graph_init_task(p); |
bea493a03
|
1417 |
rt_mutex_init_task(p); |
d12c1a379
|
1418 |
#ifdef CONFIG_PROVE_LOCKING |
de30a2b35
|
1419 1420 1421 |
DEBUG_LOCKS_WARN_ON(!p->hardirqs_enabled); DEBUG_LOCKS_WARN_ON(!p->softirqs_enabled); #endif |
1da177e4c
|
1422 |
retval = -EAGAIN; |
3b11a1dec
|
1423 |
if (atomic_read(&p->real_cred->user->processes) >= |
78d7d407b
|
1424 |
task_rlimit(p, RLIMIT_NPROC)) { |
b57922b6c
|
1425 1426 |
if (p->real_cred->user != INIT_USER && !capable(CAP_SYS_RESOURCE) && !capable(CAP_SYS_ADMIN)) |
1da177e4c
|
1427 1428 |
goto bad_fork_free; } |
72fa59970
|
1429 |
current->flags &= ~PF_NPROC_EXCEEDED; |
1da177e4c
|
1430 |
|
f1752eec6
|
1431 1432 1433 |
retval = copy_creds(p, clone_flags); if (retval < 0) goto bad_fork_free; |
1da177e4c
|
1434 1435 1436 1437 1438 1439 |
/* * If multiple threads are within copy_process(), then this check * triggers too late. This doesn't hurt, the check is only there * to stop root fork bombs. */ |
04ec93fe9
|
1440 |
retval = -EAGAIN; |
1da177e4c
|
1441 1442 |
if (nr_threads >= max_threads) goto bad_fork_cleanup_count; |
ca74e92b4
|
1443 |
delayacct_tsk_init(p); /* Must remain after dup_task_struct() */ |
514ddb446
|
1444 1445 |
p->flags &= ~(PF_SUPERPRIV | PF_WQ_WORKER); p->flags |= PF_FORKNOEXEC; |
1da177e4c
|
1446 1447 |
INIT_LIST_HEAD(&p->children); INIT_LIST_HEAD(&p->sibling); |
f41d911f8
|
1448 |
rcu_copy_process(p); |
1da177e4c
|
1449 1450 |
p->vfork_done = NULL; spin_lock_init(&p->alloc_lock); |
1da177e4c
|
1451 |
|
1da177e4c
|
1452 |
init_sigpending(&p->pending); |
648616343
|
1453 1454 |
p->utime = p->stime = p->gtime = 0; p->utimescaled = p->stimescaled = 0; |
9d7fb0427
|
1455 |
prev_cputime_init(&p->prev_cputime); |
6a61671bb
|
1456 |
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN |
b7ce2277f
|
1457 |
seqcount_init(&p->vtime_seqcount); |
6a61671bb
|
1458 |
p->vtime_snap = 0; |
7098c1eac
|
1459 |
p->vtime_snap_whence = VTIME_INACTIVE; |
6a61671bb
|
1460 |
#endif |
a3a2e76c7
|
1461 1462 1463 |
#if defined(SPLIT_RSS_COUNTING) memset(&p->rss_stat, 0, sizeof(p->rss_stat)); #endif |
172ba844a
|
1464 |
|
6976675d9
|
1465 |
p->default_timer_slack_ns = current->timer_slack_ns; |
5995477ab
|
1466 |
task_io_accounting_init(&p->ioac); |
1da177e4c
|
1467 |
acct_clear_integrals(p); |
f06febc96
|
1468 |
posix_cpu_timers_init(p); |
1da177e4c
|
1469 |
|
ccbf62d8a
|
1470 |
p->start_time = ktime_get_ns(); |
57e0be041
|
1471 |
p->real_start_time = ktime_get_boot_ns(); |
1da177e4c
|
1472 |
p->io_context = NULL; |
1da177e4c
|
1473 |
p->audit_context = NULL; |
b4f48b636
|
1474 |
cgroup_fork(p); |
1da177e4c
|
1475 |
#ifdef CONFIG_NUMA |
846a16bf0
|
1476 |
p->mempolicy = mpol_dup(p->mempolicy); |
fb0a685cb
|
1477 1478 1479 |
if (IS_ERR(p->mempolicy)) { retval = PTR_ERR(p->mempolicy); p->mempolicy = NULL; |
e8604cb43
|
1480 |
goto bad_fork_cleanup_threadgroup_lock; |
fb0a685cb
|
1481 |
} |
1da177e4c
|
1482 |
#endif |
778d3b0ff
|
1483 1484 1485 |
#ifdef CONFIG_CPUSETS p->cpuset_mem_spread_rotor = NUMA_NO_NODE; p->cpuset_slab_spread_rotor = NUMA_NO_NODE; |
cc9a6c877
|
1486 |
seqcount_init(&p->mems_allowed_seq); |
778d3b0ff
|
1487 |
#endif |
de30a2b35
|
1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 |
#ifdef CONFIG_TRACE_IRQFLAGS p->irq_events = 0; p->hardirqs_enabled = 0; p->hardirq_enable_ip = 0; p->hardirq_enable_event = 0; p->hardirq_disable_ip = _THIS_IP_; p->hardirq_disable_event = 0; p->softirqs_enabled = 1; p->softirq_enable_ip = _THIS_IP_; p->softirq_enable_event = 0; p->softirq_disable_ip = 0; p->softirq_disable_event = 0; p->hardirq_context = 0; p->softirq_context = 0; #endif |
8bcbde548
|
1503 1504 |
p->pagefault_disabled = 0; |
fbb9ce953
|
1505 1506 1507 1508 1509 |
#ifdef CONFIG_LOCKDEP p->lockdep_depth = 0; /* no locks held yet */ p->curr_chain_key = 0; p->lockdep_recursion = 0; #endif |
1da177e4c
|
1510 |
|
408894ee4
|
1511 1512 1513 |
#ifdef CONFIG_DEBUG_MUTEXES p->blocked_on = NULL; /* not blocked yet */ #endif |
cafe56359
|
1514 1515 1516 1517 |
#ifdef CONFIG_BCACHE p->sequential_io = 0; p->sequential_io_avg = 0; #endif |
0f4814065
|
1518 |
|
3c90e6e99
|
1519 |
/* Perform scheduler related setup. Assign this task to a CPU. */ |
aab03e05e
|
1520 1521 1522 |
retval = sched_fork(clone_flags, p); if (retval) goto bad_fork_cleanup_policy; |
6ab423e0e
|
1523 |
|
cdd6c482c
|
1524 |
retval = perf_event_init_task(p); |
6ab423e0e
|
1525 1526 |
if (retval) goto bad_fork_cleanup_policy; |
fb0a685cb
|
1527 1528 |
retval = audit_alloc(p); if (retval) |
6c72e3501
|
1529 |
goto bad_fork_cleanup_perf; |
1da177e4c
|
1530 |
/* copy all the process information */ |
ab602f799
|
1531 |
shm_init_task(p); |
fb0a685cb
|
1532 1533 |
retval = copy_semundo(clone_flags, p); if (retval) |
1da177e4c
|
1534 |
goto bad_fork_cleanup_audit; |
fb0a685cb
|
1535 1536 |
retval = copy_files(clone_flags, p); if (retval) |
1da177e4c
|
1537 |
goto bad_fork_cleanup_semundo; |
fb0a685cb
|
1538 1539 |
retval = copy_fs(clone_flags, p); if (retval) |
1da177e4c
|
1540 |
goto bad_fork_cleanup_files; |
fb0a685cb
|
1541 1542 |
retval = copy_sighand(clone_flags, p); if (retval) |
1da177e4c
|
1543 |
goto bad_fork_cleanup_fs; |
fb0a685cb
|
1544 1545 |
retval = copy_signal(clone_flags, p); if (retval) |
1da177e4c
|
1546 |
goto bad_fork_cleanup_sighand; |
fb0a685cb
|
1547 1548 |
retval = copy_mm(clone_flags, p); if (retval) |
1da177e4c
|
1549 |
goto bad_fork_cleanup_signal; |
fb0a685cb
|
1550 1551 |
retval = copy_namespaces(clone_flags, p); if (retval) |
d84f4f992
|
1552 |
goto bad_fork_cleanup_mm; |
fb0a685cb
|
1553 1554 |
retval = copy_io(clone_flags, p); if (retval) |
fd0928df9
|
1555 |
goto bad_fork_cleanup_namespaces; |
3033f14ab
|
1556 |
retval = copy_thread_tls(clone_flags, stack_start, stack_size, p, tls); |
1da177e4c
|
1557 |
if (retval) |
fd0928df9
|
1558 |
goto bad_fork_cleanup_io; |
1da177e4c
|
1559 |
|
425fb2b4b
|
1560 |
if (pid != &init_struct_pid) { |
c2b1df2eb
|
1561 |
pid = alloc_pid(p->nsproxy->pid_ns_for_children); |
35f71bc0a
|
1562 1563 |
if (IS_ERR(pid)) { retval = PTR_ERR(pid); |
0740aa5f6
|
1564 |
goto bad_fork_cleanup_thread; |
35f71bc0a
|
1565 |
} |
425fb2b4b
|
1566 |
} |
1da177e4c
|
1567 1568 1569 1570 |
p->set_child_tid = (clone_flags & CLONE_CHILD_SETTID) ? child_tidptr : NULL; /* * Clear TID on mm_release()? */ |
fb0a685cb
|
1571 |
p->clear_child_tid = (clone_flags & CLONE_CHILD_CLEARTID) ? child_tidptr : NULL; |
73c101011
|
1572 1573 1574 |
#ifdef CONFIG_BLOCK p->plug = NULL; #endif |
42b2dd0a0
|
1575 |
#ifdef CONFIG_FUTEX |
8f17d3a50
|
1576 1577 1578 1579 |
p->robust_list = NULL; #ifdef CONFIG_COMPAT p->compat_robust_list = NULL; #endif |
c87e2837b
|
1580 1581 |
INIT_LIST_HEAD(&p->pi_state_list); p->pi_state_cache = NULL; |
42b2dd0a0
|
1582 |
#endif |
1da177e4c
|
1583 |
/* |
f9a3879ab
|
1584 1585 1586 |
* sigaltstack should be cleared when sharing the same VM */ if ((clone_flags & (CLONE_VM|CLONE_VFORK)) == CLONE_VM) |
2a7421383
|
1587 |
sas_ss_reset(p); |
f9a3879ab
|
1588 1589 |
/* |
6580807da
|
1590 1591 |
* Syscall tracing and stepping should be turned off in the * child regardless of CLONE_PTRACE. |
1da177e4c
|
1592 |
*/ |
6580807da
|
1593 |
user_disable_single_step(p); |
1da177e4c
|
1594 |
clear_tsk_thread_flag(p, TIF_SYSCALL_TRACE); |
ed75e8d58
|
1595 1596 1597 |
#ifdef TIF_SYSCALL_EMU clear_tsk_thread_flag(p, TIF_SYSCALL_EMU); #endif |
9745512ce
|
1598 |
clear_all_latency_tracing(p); |
1da177e4c
|
1599 |
|
1da177e4c
|
1600 |
/* ok, now we should be set up.. */ |
18c830df7
|
1601 1602 |
p->pid = pid_nr(pid); if (clone_flags & CLONE_THREAD) { |
5f8aadd8b
|
1603 |
p->exit_signal = -1; |
18c830df7
|
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 |
p->group_leader = current->group_leader; p->tgid = current->tgid; } else { if (clone_flags & CLONE_PARENT) p->exit_signal = current->group_leader->exit_signal; else p->exit_signal = (clone_flags & CSIGNAL); p->group_leader = p; p->tgid = p->pid; } |
5f8aadd8b
|
1614 |
|
9d823e8f6
|
1615 1616 |
p->nr_dirtied = 0; p->nr_dirtied_pause = 128 >> (PAGE_SHIFT - 10); |
83712358b
|
1617 |
p->dirty_paused_when = 0; |
9d823e8f6
|
1618 |
|
bb8cbbfee
|
1619 |
p->pdeath_signal = 0; |
47e65328a
|
1620 |
INIT_LIST_HEAD(&p->thread_group); |
158e1645e
|
1621 |
p->task_works = NULL; |
1da177e4c
|
1622 |
|
568ac8882
|
1623 |
threadgroup_change_begin(current); |
18c830df7
|
1624 |
/* |
7e47682ea
|
1625 1626 1627 1628 1629 |
* Ensure that the cgroup subsystem policies allow the new process to be * forked. It should be noted the the new process's css_set can be changed * between here and cgroup_post_fork() if an organisation operation is in * progress. */ |
b53202e63
|
1630 |
retval = cgroup_can_fork(p); |
7e47682ea
|
1631 1632 1633 1634 |
if (retval) goto bad_fork_free_pid; /* |
18c830df7
|
1635 1636 1637 |
* Make it visible to the rest of the system, but dont wake it up yet. * Need tasklist lock for parent etc handling! */ |
1da177e4c
|
1638 |
write_lock_irq(&tasklist_lock); |
1da177e4c
|
1639 |
/* CLONE_PARENT re-uses the old parent */ |
2d5516cbb
|
1640 |
if (clone_flags & (CLONE_PARENT|CLONE_THREAD)) { |
1da177e4c
|
1641 |
p->real_parent = current->real_parent; |
2d5516cbb
|
1642 1643 |
p->parent_exec_id = current->parent_exec_id; } else { |
1da177e4c
|
1644 |
p->real_parent = current; |
2d5516cbb
|
1645 1646 |
p->parent_exec_id = current->self_exec_id; } |
1da177e4c
|
1647 |
|
3f17da699
|
1648 |
spin_lock(¤t->sighand->siglock); |
4a2c7a783
|
1649 1650 |
/* |
dbd952127
|
1651 1652 1653 1654 1655 1656 |
* Copy seccomp details explicitly here, in case they were changed * before holding sighand lock. */ copy_seccomp(p); /* |
4a2c7a783
|
1657 1658 1659 1660 1661 1662 |
* Process group and session signals need to be delivered to just the * parent before the fork or both the parent and the child after the * fork. Restart if a signal comes in before we add the new process to * it's process group. * A fatal signal pending means that current will exit, so the new * thread can't slip out of an OOM kill (or normal SIGKILL). |
fb0a685cb
|
1663 |
*/ |
23ff44402
|
1664 |
recalc_sigpending(); |
4a2c7a783
|
1665 1666 1667 1668 |
if (signal_pending(current)) { spin_unlock(¤t->sighand->siglock); write_unlock_irq(&tasklist_lock); retval = -ERESTARTNOINTR; |
7e47682ea
|
1669 |
goto bad_fork_cancel_cgroup; |
4a2c7a783
|
1670 |
} |
73b9ebfe1
|
1671 |
if (likely(p->pid)) { |
4b9d33e6d
|
1672 |
ptrace_init_task(p, (clone_flags & CLONE_PTRACE) || trace); |
73b9ebfe1
|
1673 |
|
819077398
|
1674 |
init_task_pid(p, PIDTYPE_PID, pid); |
73b9ebfe1
|
1675 |
if (thread_group_leader(p)) { |
819077398
|
1676 1677 |
init_task_pid(p, PIDTYPE_PGID, task_pgrp(current)); init_task_pid(p, PIDTYPE_SID, task_session(current)); |
1c4042c29
|
1678 |
if (is_child_reaper(pid)) { |
17cf22c33
|
1679 |
ns_of_pid(pid)->child_reaper = p; |
1c4042c29
|
1680 1681 |
p->signal->flags |= SIGNAL_UNKILLABLE; } |
73b9ebfe1
|
1682 |
|
fea9d1755
|
1683 |
p->signal->leader_pid = pid; |
9c9f4ded9
|
1684 |
p->signal->tty = tty_kref_get(current->signal->tty); |
9cd80bbb0
|
1685 |
list_add_tail(&p->sibling, &p->real_parent->children); |
5e85d4abe
|
1686 |
list_add_tail_rcu(&p->tasks, &init_task.tasks); |
819077398
|
1687 1688 |
attach_pid(p, PIDTYPE_PGID); attach_pid(p, PIDTYPE_SID); |
909ea9646
|
1689 |
__this_cpu_inc(process_counts); |
80628ca06
|
1690 1691 1692 1693 |
} else { current->signal->nr_threads++; atomic_inc(¤t->signal->live); atomic_inc(¤t->signal->sigcnt); |
80628ca06
|
1694 1695 |
list_add_tail_rcu(&p->thread_group, &p->group_leader->thread_group); |
0c740d0af
|
1696 1697 |
list_add_tail_rcu(&p->thread_node, &p->signal->thread_head); |
73b9ebfe1
|
1698 |
} |
819077398
|
1699 |
attach_pid(p, PIDTYPE_PID); |
73b9ebfe1
|
1700 |
nr_threads++; |
1da177e4c
|
1701 |
} |
1da177e4c
|
1702 |
total_forks++; |
3f17da699
|
1703 |
spin_unlock(¤t->sighand->siglock); |
4af4206be
|
1704 |
syscall_tracepoint_update(p); |
1da177e4c
|
1705 |
write_unlock_irq(&tasklist_lock); |
4af4206be
|
1706 |
|
c13cf856c
|
1707 |
proc_fork_connector(p); |
b53202e63
|
1708 |
cgroup_post_fork(p); |
c9e75f049
|
1709 |
threadgroup_change_end(current); |
cdd6c482c
|
1710 |
perf_event_fork(p); |
43d2b1132
|
1711 1712 |
trace_task_newtask(p, clone_flags); |
3ab679661
|
1713 |
uprobe_copy_process(p, clone_flags); |
43d2b1132
|
1714 |
|
1da177e4c
|
1715 |
return p; |
7e47682ea
|
1716 |
bad_fork_cancel_cgroup: |
b53202e63
|
1717 |
cgroup_cancel_fork(p); |
425fb2b4b
|
1718 |
bad_fork_free_pid: |
568ac8882
|
1719 |
threadgroup_change_end(current); |
425fb2b4b
|
1720 1721 |
if (pid != &init_struct_pid) free_pid(pid); |
0740aa5f6
|
1722 1723 |
bad_fork_cleanup_thread: exit_thread(p); |
fd0928df9
|
1724 |
bad_fork_cleanup_io: |
b69f22920
|
1725 1726 |
if (p->io_context) exit_io_context(p); |
ab516013a
|
1727 |
bad_fork_cleanup_namespaces: |
444f378b2
|
1728 |
exit_task_namespaces(p); |
1da177e4c
|
1729 |
bad_fork_cleanup_mm: |
c9f01245b
|
1730 |
if (p->mm) |
1da177e4c
|
1731 1732 |
mmput(p->mm); bad_fork_cleanup_signal: |
4ab6c0833
|
1733 |
if (!(clone_flags & CLONE_THREAD)) |
1c5354de9
|
1734 |
free_signal_struct(p->signal); |
1da177e4c
|
1735 |
bad_fork_cleanup_sighand: |
a7e5328a0
|
1736 |
__cleanup_sighand(p->sighand); |
1da177e4c
|
1737 1738 1739 1740 1741 1742 1743 1744 |
bad_fork_cleanup_fs: exit_fs(p); /* blocking */ bad_fork_cleanup_files: exit_files(p); /* blocking */ bad_fork_cleanup_semundo: exit_sem(p); bad_fork_cleanup_audit: audit_free(p); |
6c72e3501
|
1745 |
bad_fork_cleanup_perf: |
cdd6c482c
|
1746 |
perf_event_free_task(p); |
6c72e3501
|
1747 |
bad_fork_cleanup_policy: |
1da177e4c
|
1748 |
#ifdef CONFIG_NUMA |
f0be3d32b
|
1749 |
mpol_put(p->mempolicy); |
e8604cb43
|
1750 |
bad_fork_cleanup_threadgroup_lock: |
1da177e4c
|
1751 |
#endif |
35df17c57
|
1752 |
delayacct_tsk_free(p); |
1da177e4c
|
1753 |
bad_fork_cleanup_count: |
d84f4f992
|
1754 |
atomic_dec(&p->cred->user->processes); |
e0e817392
|
1755 |
exit_creds(p); |
1da177e4c
|
1756 |
bad_fork_free: |
405c07597
|
1757 |
p->state = TASK_DEAD; |
68f24b08e
|
1758 |
put_task_stack(p); |
1da177e4c
|
1759 |
free_task(p); |
fe7d37d1f
|
1760 1761 |
fork_out: return ERR_PTR(retval); |
1da177e4c
|
1762 |
} |
f106eee10
|
1763 1764 1765 1766 1767 1768 1769 1770 1771 |
static inline void init_idle_pids(struct pid_link *links) { enum pid_type type; for (type = PIDTYPE_PID; type < PIDTYPE_MAX; ++type) { INIT_HLIST_NODE(&links[type].node); /* not really needed */ links[type].pid = &init_struct_pid; } } |
0db0628d9
|
1772 |
struct task_struct *fork_idle(int cpu) |
1da177e4c
|
1773 |
{ |
36c8b5868
|
1774 |
struct task_struct *task; |
725fc629f
|
1775 1776 |
task = copy_process(CLONE_VM, 0, 0, NULL, &init_struct_pid, 0, 0, cpu_to_node(cpu)); |
f106eee10
|
1777 1778 |
if (!IS_ERR(task)) { init_idle_pids(task->pids); |
753ca4f31
|
1779 |
init_idle(task, cpu); |
f106eee10
|
1780 |
} |
73b9ebfe1
|
1781 |
|
1da177e4c
|
1782 1783 |
return task; } |
1da177e4c
|
1784 1785 1786 1787 1788 1789 |
/* * Ok, this is the main fork-routine. * * It copies the process, and if successful kick-starts * it and waits for it to finish using the VM if required. */ |
3033f14ab
|
1790 |
long _do_fork(unsigned long clone_flags, |
1da177e4c
|
1791 |
unsigned long stack_start, |
1da177e4c
|
1792 1793 |
unsigned long stack_size, int __user *parent_tidptr, |
3033f14ab
|
1794 1795 |
int __user *child_tidptr, unsigned long tls) |
1da177e4c
|
1796 1797 1798 |
{ struct task_struct *p; int trace = 0; |
92476d7fc
|
1799 |
long nr; |
1da177e4c
|
1800 |
|
bdff746a3
|
1801 |
/* |
4b9d33e6d
|
1802 1803 1804 1805 |
* Determine whether and which event to report to ptracer. When * called from kernel_thread or CLONE_UNTRACED is explicitly * requested, no event is reported; otherwise, report if the event * for the type of forking is enabled. |
09a05394f
|
1806 |
*/ |
e80d6661c
|
1807 |
if (!(clone_flags & CLONE_UNTRACED)) { |
4b9d33e6d
|
1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 |
if (clone_flags & CLONE_VFORK) trace = PTRACE_EVENT_VFORK; else if ((clone_flags & CSIGNAL) != SIGCHLD) trace = PTRACE_EVENT_CLONE; else trace = PTRACE_EVENT_FORK; if (likely(!ptrace_event_enabled(current, trace))) trace = 0; } |
1da177e4c
|
1818 |
|
62e791c1b
|
1819 |
p = copy_process(clone_flags, stack_start, stack_size, |
725fc629f
|
1820 |
child_tidptr, NULL, trace, tls, NUMA_NO_NODE); |
38addce8b
|
1821 |
add_latent_entropy(); |
1da177e4c
|
1822 1823 1824 1825 1826 1827 |
/* * Do this prior waking up the new thread - the thread pointer * might get invalid after that point, if the thread exits quickly. */ if (!IS_ERR(p)) { struct completion vfork; |
4e52365f2
|
1828 |
struct pid *pid; |
1da177e4c
|
1829 |
|
0a16b6075
|
1830 |
trace_sched_process_fork(current, p); |
4e52365f2
|
1831 1832 |
pid = get_task_pid(p, PIDTYPE_PID); nr = pid_vnr(pid); |
30e49c263
|
1833 1834 1835 |
if (clone_flags & CLONE_PARENT_SETTID) put_user(nr, parent_tidptr); |
a6f5e0637
|
1836 |
|
1da177e4c
|
1837 1838 1839 |
if (clone_flags & CLONE_VFORK) { p->vfork_done = &vfork; init_completion(&vfork); |
d68b46fe1
|
1840 |
get_task_struct(p); |
1da177e4c
|
1841 |
} |
3e51e3edf
|
1842 |
wake_up_new_task(p); |
1da177e4c
|
1843 |
|
4b9d33e6d
|
1844 1845 |
/* forking complete and child started to run, tell ptracer */ if (unlikely(trace)) |
4e52365f2
|
1846 |
ptrace_event_pid(trace, pid); |
09a05394f
|
1847 |
|
1da177e4c
|
1848 |
if (clone_flags & CLONE_VFORK) { |
d68b46fe1
|
1849 |
if (!wait_for_vfork_done(p, &vfork)) |
4e52365f2
|
1850 |
ptrace_event_pid(PTRACE_EVENT_VFORK_DONE, pid); |
1da177e4c
|
1851 |
} |
4e52365f2
|
1852 1853 |
put_pid(pid); |
1da177e4c
|
1854 |
} else { |
92476d7fc
|
1855 |
nr = PTR_ERR(p); |
1da177e4c
|
1856 |
} |
92476d7fc
|
1857 |
return nr; |
1da177e4c
|
1858 |
} |
3033f14ab
|
1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 |
#ifndef CONFIG_HAVE_COPY_THREAD_TLS /* For compatibility with architectures that call do_fork directly rather than * using the syscall entry points below. */ long do_fork(unsigned long clone_flags, unsigned long stack_start, unsigned long stack_size, int __user *parent_tidptr, int __user *child_tidptr) { return _do_fork(clone_flags, stack_start, stack_size, parent_tidptr, child_tidptr, 0); } #endif |
2aa3a7f86
|
1872 1873 1874 1875 1876 |
/* * Create a kernel thread. */ pid_t kernel_thread(int (*fn)(void *), void *arg, unsigned long flags) { |
3033f14ab
|
1877 1878 |
return _do_fork(flags|CLONE_VM|CLONE_UNTRACED, (unsigned long)fn, (unsigned long)arg, NULL, NULL, 0); |
2aa3a7f86
|
1879 |
} |
2aa3a7f86
|
1880 |
|
d2125043a
|
1881 1882 1883 1884 |
#ifdef __ARCH_WANT_SYS_FORK SYSCALL_DEFINE0(fork) { #ifdef CONFIG_MMU |
3033f14ab
|
1885 |
return _do_fork(SIGCHLD, 0, 0, NULL, NULL, 0); |
d2125043a
|
1886 1887 |
#else /* can not support in nommu mode */ |
5d59e1827
|
1888 |
return -EINVAL; |
d2125043a
|
1889 1890 1891 1892 1893 1894 1895 |
#endif } #endif #ifdef __ARCH_WANT_SYS_VFORK SYSCALL_DEFINE0(vfork) { |
3033f14ab
|
1896 1897 |
return _do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, 0, 0, NULL, NULL, 0); |
d2125043a
|
1898 1899 1900 1901 1902 1903 1904 |
} #endif #ifdef __ARCH_WANT_SYS_CLONE #ifdef CONFIG_CLONE_BACKWARDS SYSCALL_DEFINE5(clone, unsigned long, clone_flags, unsigned long, newsp, int __user *, parent_tidptr, |
3033f14ab
|
1905 |
unsigned long, tls, |
d2125043a
|
1906 1907 1908 1909 1910 |
int __user *, child_tidptr) #elif defined(CONFIG_CLONE_BACKWARDS2) SYSCALL_DEFINE5(clone, unsigned long, newsp, unsigned long, clone_flags, int __user *, parent_tidptr, int __user *, child_tidptr, |
3033f14ab
|
1911 |
unsigned long, tls) |
dfa9771a7
|
1912 1913 1914 1915 1916 |
#elif defined(CONFIG_CLONE_BACKWARDS3) SYSCALL_DEFINE6(clone, unsigned long, clone_flags, unsigned long, newsp, int, stack_size, int __user *, parent_tidptr, int __user *, child_tidptr, |
3033f14ab
|
1917 |
unsigned long, tls) |
d2125043a
|
1918 1919 1920 1921 |
#else SYSCALL_DEFINE5(clone, unsigned long, clone_flags, unsigned long, newsp, int __user *, parent_tidptr, int __user *, child_tidptr, |
3033f14ab
|
1922 |
unsigned long, tls) |
d2125043a
|
1923 1924 |
#endif { |
3033f14ab
|
1925 |
return _do_fork(clone_flags, newsp, 0, parent_tidptr, child_tidptr, tls); |
d2125043a
|
1926 1927 |
} #endif |
5fd63b308
|
1928 1929 1930 |
#ifndef ARCH_MIN_MMSTRUCT_ALIGN #define ARCH_MIN_MMSTRUCT_ALIGN 0 #endif |
51cc50685
|
1931 |
static void sighand_ctor(void *data) |
aa1757f90
|
1932 1933 |
{ struct sighand_struct *sighand = data; |
a35afb830
|
1934 |
spin_lock_init(&sighand->siglock); |
b8fceee17
|
1935 |
init_waitqueue_head(&sighand->signalfd_wqh); |
aa1757f90
|
1936 |
} |
1da177e4c
|
1937 1938 1939 1940 |
void __init proc_caches_init(void) { sighand_cachep = kmem_cache_create("sighand_cache", sizeof(struct sighand_struct), 0, |
2dff44052
|
1941 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_DESTROY_BY_RCU| |
5d097056c
|
1942 |
SLAB_NOTRACK|SLAB_ACCOUNT, sighand_ctor); |
1da177e4c
|
1943 1944 |
signal_cachep = kmem_cache_create("signal_cache", sizeof(struct signal_struct), 0, |
5d097056c
|
1945 1946 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK|SLAB_ACCOUNT, NULL); |
20c2df83d
|
1947 |
files_cachep = kmem_cache_create("files_cache", |
1da177e4c
|
1948 |
sizeof(struct files_struct), 0, |
5d097056c
|
1949 1950 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK|SLAB_ACCOUNT, NULL); |
20c2df83d
|
1951 |
fs_cachep = kmem_cache_create("fs_cache", |
1da177e4c
|
1952 |
sizeof(struct fs_struct), 0, |
5d097056c
|
1953 1954 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK|SLAB_ACCOUNT, NULL); |
6345d24da
|
1955 1956 1957 1958 1959 1960 1961 |
/* * FIXME! The "sizeof(struct mm_struct)" currently includes the * whole struct cpumask for the OFFSTACK case. We could change * this to *only* allocate as much of it as required by the * maximum number of CPU's we can ever have. The cpumask_allocation * is at the end of the structure, exactly for that reason. */ |
1da177e4c
|
1962 |
mm_cachep = kmem_cache_create("mm_struct", |
5fd63b308
|
1963 |
sizeof(struct mm_struct), ARCH_MIN_MMSTRUCT_ALIGN, |
5d097056c
|
1964 1965 1966 |
SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK|SLAB_ACCOUNT, NULL); vm_area_cachep = KMEM_CACHE(vm_area_struct, SLAB_PANIC|SLAB_ACCOUNT); |
8feae1311
|
1967 |
mmap_init(); |
665771939
|
1968 |
nsproxy_cache_init(); |
1da177e4c
|
1969 |
} |
cf2e340f4
|
1970 |
|
cf2e340f4
|
1971 |
/* |
9bfb23fc4
|
1972 |
* Check constraints on flags passed to the unshare system call. |
cf2e340f4
|
1973 |
*/ |
9bfb23fc4
|
1974 |
static int check_unshare_flags(unsigned long unshare_flags) |
cf2e340f4
|
1975 |
{ |
9bfb23fc4
|
1976 1977 |
if (unshare_flags & ~(CLONE_THREAD|CLONE_FS|CLONE_NEWNS|CLONE_SIGHAND| CLONE_VM|CLONE_FILES|CLONE_SYSVSEM| |
50804fe37
|
1978 |
CLONE_NEWUTS|CLONE_NEWIPC|CLONE_NEWNET| |
a79a908fd
|
1979 |
CLONE_NEWUSER|CLONE_NEWPID|CLONE_NEWCGROUP)) |
9bfb23fc4
|
1980 |
return -EINVAL; |
cf2e340f4
|
1981 |
/* |
12c641ab8
|
1982 1983 1984 1985 |
* Not implemented, but pretend it works if there is nothing * to unshare. Note that unsharing the address space or the * signal handlers also need to unshare the signal queues (aka * CLONE_THREAD). |
cf2e340f4
|
1986 |
*/ |
9bfb23fc4
|
1987 |
if (unshare_flags & (CLONE_THREAD | CLONE_SIGHAND | CLONE_VM)) { |
12c641ab8
|
1988 1989 1990 1991 1992 1993 1994 1995 1996 |
if (!thread_group_empty(current)) return -EINVAL; } if (unshare_flags & (CLONE_SIGHAND | CLONE_VM)) { if (atomic_read(¤t->sighand->count) > 1) return -EINVAL; } if (unshare_flags & CLONE_VM) { if (!current_is_single_threaded()) |
9bfb23fc4
|
1997 1998 |
return -EINVAL; } |
cf2e340f4
|
1999 2000 2001 2002 2003 |
return 0; } /* |
99d1419d9
|
2004 |
* Unshare the filesystem structure if it is being shared |
cf2e340f4
|
2005 2006 2007 2008 |
*/ static int unshare_fs(unsigned long unshare_flags, struct fs_struct **new_fsp) { struct fs_struct *fs = current->fs; |
498052bba
|
2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 |
if (!(unshare_flags & CLONE_FS) || !fs) return 0; /* don't need lock here; in the worst case we'll do useless copy */ if (fs->users == 1) return 0; *new_fsp = copy_fs_struct(fs); if (!*new_fsp) return -ENOMEM; |
cf2e340f4
|
2019 2020 2021 2022 2023 |
return 0; } /* |
a016f3389
|
2024 |
* Unshare file descriptor table if it is being shared |
cf2e340f4
|
2025 2026 2027 2028 |
*/ static int unshare_fd(unsigned long unshare_flags, struct files_struct **new_fdp) { struct files_struct *fd = current->files; |
a016f3389
|
2029 |
int error = 0; |
cf2e340f4
|
2030 2031 |
if ((unshare_flags & CLONE_FILES) && |
a016f3389
|
2032 2033 2034 2035 2036 |
(fd && atomic_read(&fd->count) > 1)) { *new_fdp = dup_fd(fd, &error); if (!*new_fdp) return error; } |
cf2e340f4
|
2037 2038 2039 2040 2041 |
return 0; } /* |
cf2e340f4
|
2042 2043 2044 2045 2046 2047 2048 |
* unshare allows a process to 'unshare' part of the process * context which was originally shared using clone. copy_* * functions used by do_fork() cannot be used here directly * because they modify an inactive task_struct that is being * constructed. Here we are modifying the current, active, * task_struct. */ |
6559eed8c
|
2049 |
SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags) |
cf2e340f4
|
2050 |
{ |
cf2e340f4
|
2051 |
struct fs_struct *fs, *new_fs = NULL; |
cf2e340f4
|
2052 |
struct files_struct *fd, *new_fd = NULL; |
b2e0d9870
|
2053 |
struct cred *new_cred = NULL; |
cf7b708c8
|
2054 |
struct nsproxy *new_nsproxy = NULL; |
9edff4ab1
|
2055 |
int do_sysvsem = 0; |
9bfb23fc4
|
2056 |
int err; |
cf2e340f4
|
2057 |
|
50804fe37
|
2058 |
/* |
faf00da54
|
2059 2060 |
* If unsharing a user namespace must also unshare the thread group * and unshare the filesystem root and working directories. |
b2e0d9870
|
2061 2062 |
*/ if (unshare_flags & CLONE_NEWUSER) |
e66eded83
|
2063 |
unshare_flags |= CLONE_THREAD | CLONE_FS; |
b2e0d9870
|
2064 |
/* |
50804fe37
|
2065 2066 2067 2068 |
* If unsharing vm, must also unshare signal handlers. */ if (unshare_flags & CLONE_VM) unshare_flags |= CLONE_SIGHAND; |
6013f67fc
|
2069 |
/* |
12c641ab8
|
2070 2071 2072 2073 2074 |
* If unsharing a signal handlers, must also unshare the signal queues. */ if (unshare_flags & CLONE_SIGHAND) unshare_flags |= CLONE_THREAD; /* |
9bfb23fc4
|
2075 2076 2077 2078 |
* If unsharing namespace, must also unshare filesystem information. */ if (unshare_flags & CLONE_NEWNS) unshare_flags |= CLONE_FS; |
50804fe37
|
2079 2080 2081 2082 |
err = check_unshare_flags(unshare_flags); if (err) goto bad_unshare_out; |
9bfb23fc4
|
2083 |
/* |
6013f67fc
|
2084 2085 2086 2087 2088 |
* CLONE_NEWIPC must also detach from the undolist: after switching * to a new ipc namespace, the semaphore arrays from the old * namespace are unreachable. */ if (unshare_flags & (CLONE_NEWIPC|CLONE_SYSVSEM)) |
9edff4ab1
|
2089 |
do_sysvsem = 1; |
fb0a685cb
|
2090 2091 |
err = unshare_fs(unshare_flags, &new_fs); if (err) |
9bfb23fc4
|
2092 |
goto bad_unshare_out; |
fb0a685cb
|
2093 2094 |
err = unshare_fd(unshare_flags, &new_fd); if (err) |
9bfb23fc4
|
2095 |
goto bad_unshare_cleanup_fs; |
b2e0d9870
|
2096 |
err = unshare_userns(unshare_flags, &new_cred); |
fb0a685cb
|
2097 |
if (err) |
9edff4ab1
|
2098 |
goto bad_unshare_cleanup_fd; |
b2e0d9870
|
2099 2100 2101 2102 |
err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy, new_cred, new_fs); if (err) goto bad_unshare_cleanup_cred; |
c0b2fc316
|
2103 |
|
b2e0d9870
|
2104 |
if (new_fs || new_fd || do_sysvsem || new_cred || new_nsproxy) { |
9edff4ab1
|
2105 2106 2107 2108 2109 2110 |
if (do_sysvsem) { /* * CLONE_SYSVSEM is equivalent to sys_exit(). */ exit_sem(current); } |
ab602f799
|
2111 2112 2113 2114 2115 |
if (unshare_flags & CLONE_NEWIPC) { /* Orphan segments in old ns (see sem above). */ exit_shm(current); shm_init_task(current); } |
ab516013a
|
2116 |
|
6f977e6b2
|
2117 |
if (new_nsproxy) |
cf7b708c8
|
2118 |
switch_task_namespaces(current, new_nsproxy); |
cf2e340f4
|
2119 |
|
cf7b708c8
|
2120 |
task_lock(current); |
cf2e340f4
|
2121 2122 |
if (new_fs) { fs = current->fs; |
2a4419b5b
|
2123 |
spin_lock(&fs->lock); |
cf2e340f4
|
2124 |
current->fs = new_fs; |
498052bba
|
2125 2126 2127 2128 |
if (--fs->users) new_fs = NULL; else new_fs = fs; |
2a4419b5b
|
2129 |
spin_unlock(&fs->lock); |
cf2e340f4
|
2130 |
} |
cf2e340f4
|
2131 2132 2133 2134 2135 2136 2137 |
if (new_fd) { fd = current->files; current->files = new_fd; new_fd = fd; } task_unlock(current); |
b2e0d9870
|
2138 2139 2140 2141 2142 2143 |
if (new_cred) { /* Install the new user namespace */ commit_creds(new_cred); new_cred = NULL; } |
cf2e340f4
|
2144 |
} |
b2e0d9870
|
2145 2146 2147 |
bad_unshare_cleanup_cred: if (new_cred) put_cred(new_cred); |
cf2e340f4
|
2148 2149 2150 |
bad_unshare_cleanup_fd: if (new_fd) put_files_struct(new_fd); |
cf2e340f4
|
2151 2152 |
bad_unshare_cleanup_fs: if (new_fs) |
498052bba
|
2153 |
free_fs_struct(new_fs); |
cf2e340f4
|
2154 |
|
cf2e340f4
|
2155 2156 2157 |
bad_unshare_out: return err; } |
3b1253880
|
2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 |
/* * Helper to unshare the files of the current task. * We don't want to expose copy_files internals to * the exec layer of the kernel. */ int unshare_files(struct files_struct **displaced) { struct task_struct *task = current; |
50704516f
|
2168 |
struct files_struct *copy = NULL; |
3b1253880
|
2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 |
int error; error = unshare_fd(CLONE_FILES, ©); if (error || !copy) { *displaced = NULL; return error; } *displaced = task->files; task_lock(task); task->files = copy; task_unlock(task); return 0; } |
16db3d3f1
|
2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 |
int sysctl_max_threads(struct ctl_table *table, int write, void __user *buffer, size_t *lenp, loff_t *ppos) { struct ctl_table t; int ret; int threads = max_threads; int min = MIN_THREADS; int max = MAX_THREADS; t = *table; t.data = &threads; t.extra1 = &min; t.extra2 = &max; ret = proc_dointvec_minmax(&t, write, buffer, lenp, ppos); if (ret || !write) return ret; set_max_threads(threads); return 0; } |