fork.c 41.6 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
/*
 *  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()'
 */

#include <linux/slab.h>
#include <linux/init.h>
#include <linux/unistd.h>
#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/completion.h>
#include <linux/personality.h>
#include <linux/mempolicy.h>
#include <linux/sem.h>
#include <linux/file.h>
A
Al Viro 已提交
24
#include <linux/fdtable.h>
25
#include <linux/iocontext.h>
L
Linus Torvalds 已提交
26 27 28
#include <linux/key.h>
#include <linux/binfmts.h>
#include <linux/mman.h>
A
Andrea Arcangeli 已提交
29
#include <linux/mmu_notifier.h>
L
Linus Torvalds 已提交
30
#include <linux/fs.h>
S
Serge E. Hallyn 已提交
31
#include <linux/nsproxy.h>
32
#include <linux/capability.h>
L
Linus Torvalds 已提交
33
#include <linux/cpu.h>
34
#include <linux/cgroup.h>
L
Linus Torvalds 已提交
35
#include <linux/security.h>
36
#include <linux/hugetlb.h>
L
Linus Torvalds 已提交
37 38 39
#include <linux/swap.h>
#include <linux/syscalls.h>
#include <linux/jiffies.h>
R
Roland McGrath 已提交
40
#include <linux/tracehook.h>
L
Linus Torvalds 已提交
41
#include <linux/futex.h>
42
#include <linux/compat.h>
43
#include <linux/task_io_accounting_ops.h>
44
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
45 46 47
#include <linux/ptrace.h>
#include <linux/mount.h>
#include <linux/audit.h>
48
#include <linux/memcontrol.h>
49
#include <linux/ftrace.h>
L
Linus Torvalds 已提交
50 51 52
#include <linux/profile.h>
#include <linux/rmap.h>
#include <linux/acct.h>
53
#include <linux/tsacct_kern.h>
M
Matt Helsley 已提交
54
#include <linux/cn_proc.h>
R
Rafael J. Wysocki 已提交
55
#include <linux/freezer.h>
56
#include <linux/delayacct.h>
57
#include <linux/taskstats_kern.h>
58
#include <linux/random.h>
M
Miloslav Trmac 已提交
59
#include <linux/tty.h>
60
#include <linux/proc_fs.h>
61
#include <linux/blkdev.h>
62
#include <linux/fs_struct.h>
63
#include <linux/magic.h>
64
#include <linux/perf_counter.h>
L
Linus Torvalds 已提交
65 66 67 68 69 70 71 72

#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>

73 74
#include <trace/events/sched.h>

L
Linus Torvalds 已提交
75 76 77 78 79 80 81 82 83 84
/*
 * Protected counters by write_lock_irq(&tasklist_lock)
 */
unsigned long total_forks;	/* Handle normal Linux uptimes. */
int nr_threads; 		/* The idle threads do not count.. */

int max_threads;		/* tunable limit on nr_threads */

DEFINE_PER_CPU(unsigned long, process_counts) = 0;

85
__cacheline_aligned DEFINE_RWLOCK(tasklist_lock);  /* outer */
L
Linus Torvalds 已提交
86 87 88 89 90 91 92 93 94 95 96 97 98 99 100

int nr_processes(void)
{
	int cpu;
	int total = 0;

	for_each_online_cpu(cpu)
		total += per_cpu(process_counts, cpu);

	return total;
}

#ifndef __HAVE_ARCH_TASK_STRUCT_ALLOCATOR
# define alloc_task_struct()	kmem_cache_alloc(task_struct_cachep, GFP_KERNEL)
# define free_task_struct(tsk)	kmem_cache_free(task_struct_cachep, (tsk))
101
static struct kmem_cache *task_struct_cachep;
L
Linus Torvalds 已提交
102 103
#endif

104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120
#ifndef __HAVE_ARCH_THREAD_INFO_ALLOCATOR
static inline struct thread_info *alloc_thread_info(struct task_struct *tsk)
{
#ifdef CONFIG_DEBUG_STACK_USAGE
	gfp_t mask = GFP_KERNEL | __GFP_ZERO;
#else
	gfp_t mask = GFP_KERNEL;
#endif
	return (struct thread_info *)__get_free_pages(mask, THREAD_SIZE_ORDER);
}

static inline void free_thread_info(struct thread_info *ti)
{
	free_pages((unsigned long)ti, THREAD_SIZE_ORDER);
}
#endif

L
Linus Torvalds 已提交
121
/* SLAB cache for signal_struct structures (tsk->signal) */
122
static struct kmem_cache *signal_cachep;
L
Linus Torvalds 已提交
123 124

/* SLAB cache for sighand_struct structures (tsk->sighand) */
125
struct kmem_cache *sighand_cachep;
L
Linus Torvalds 已提交
126 127

/* SLAB cache for files_struct structures (tsk->files) */
128
struct kmem_cache *files_cachep;
L
Linus Torvalds 已提交
129 130

/* SLAB cache for fs_struct structures (tsk->fs) */
131
struct kmem_cache *fs_cachep;
L
Linus Torvalds 已提交
132 133

/* SLAB cache for vm_area_struct structures */
134
struct kmem_cache *vm_area_cachep;
L
Linus Torvalds 已提交
135 136

/* SLAB cache for mm_struct structures (tsk->mm) */
137
static struct kmem_cache *mm_cachep;
L
Linus Torvalds 已提交
138 139 140

void free_task(struct task_struct *tsk)
{
P
Peter Zijlstra 已提交
141
	prop_local_destroy_single(&tsk->dirties);
R
Roman Zippel 已提交
142
	free_thread_info(tsk->stack);
I
Ingo Molnar 已提交
143
	rt_mutex_debug_task_free(tsk);
144
	ftrace_graph_exit_task(tsk);
L
Linus Torvalds 已提交
145 146 147 148
	free_task_struct(tsk);
}
EXPORT_SYMBOL(free_task);

149
void __put_task_struct(struct task_struct *tsk)
L
Linus Torvalds 已提交
150
{
E
Eugene Teo 已提交
151
	WARN_ON(!tsk->exit_state);
L
Linus Torvalds 已提交
152 153 154
	WARN_ON(atomic_read(&tsk->usage));
	WARN_ON(tsk == current);

155
	put_cred(tsk->real_cred);
156
	put_cred(tsk->cred);
157
	delayacct_tsk_free(tsk);
L
Linus Torvalds 已提交
158 159 160 161 162

	if (!profile_handoff_task(tsk))
		free_task(tsk);
}

S
Suresh Siddha 已提交
163 164 165 166 167 168 169
/*
 * macro override instead of weak attribute alias, to workaround
 * gcc 4.1.0 and 4.1.1 bugs with weak attribute and empty functions.
 */
#ifndef arch_task_cache_init
#define arch_task_cache_init()
#endif
170

L
Linus Torvalds 已提交
171 172 173 174 175 176 177 178 179
void __init fork_init(unsigned long mempages)
{
#ifndef __HAVE_ARCH_TASK_STRUCT_ALLOCATOR
#ifndef ARCH_MIN_TASKALIGN
#define ARCH_MIN_TASKALIGN	L1_CACHE_BYTES
#endif
	/* create a slab on which task_structs can be allocated */
	task_struct_cachep =
		kmem_cache_create("task_struct", sizeof(struct task_struct),
V
Vegard Nossum 已提交
180
			ARCH_MIN_TASKALIGN, SLAB_PANIC | SLAB_NOTRACK, NULL);
L
Linus Torvalds 已提交
181 182
#endif

183 184 185
	/* do the arch specific task caches init */
	arch_task_cache_init();

L
Linus Torvalds 已提交
186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204
	/*
	 * The default maximum number of threads is set to a safe
	 * value: the thread structures can take up at most half
	 * of memory.
	 */
	max_threads = mempages / (8 * THREAD_SIZE / PAGE_SIZE);

	/*
	 * we need to allow at least 20 threads to boot a system
	 */
	if(max_threads < 20)
		max_threads = 20;

	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];
}

205 206 207 208 209 210 211
int __attribute__((weak)) arch_dup_task_struct(struct task_struct *dst,
					       struct task_struct *src)
{
	*dst = *src;
	return 0;
}

L
Linus Torvalds 已提交
212 213 214 215
static struct task_struct *dup_task_struct(struct task_struct *orig)
{
	struct task_struct *tsk;
	struct thread_info *ti;
216 217
	unsigned long *stackend;

P
Peter Zijlstra 已提交
218
	int err;
L
Linus Torvalds 已提交
219 220 221 222 223 224 225 226 227 228 229 230 231

	prepare_to_copy(orig);

	tsk = alloc_task_struct();
	if (!tsk)
		return NULL;

	ti = alloc_thread_info(tsk);
	if (!ti) {
		free_task_struct(tsk);
		return NULL;
	}

232 233 234 235
 	err = arch_dup_task_struct(tsk, orig);
	if (err)
		goto out;

R
Roman Zippel 已提交
236
	tsk->stack = ti;
P
Peter Zijlstra 已提交
237 238

	err = prop_local_init_single(&tsk->dirties);
239 240
	if (err)
		goto out;
P
Peter Zijlstra 已提交
241

242
	setup_thread_stack(tsk, orig);
243 244
	stackend = end_of_stack(tsk);
	*stackend = STACK_END_MAGIC;	/* for overflow detection */
L
Linus Torvalds 已提交
245

246 247 248 249
#ifdef CONFIG_CC_STACKPROTECTOR
	tsk->stack_canary = get_random_int();
#endif

L
Linus Torvalds 已提交
250 251
	/* One for us, one for whoever does the "release_task()" (usually parent) */
	atomic_set(&tsk->usage,2);
252
	atomic_set(&tsk->fs_excl, 0);
253
#ifdef CONFIG_BLK_DEV_IO_TRACE
254
	tsk->btrace_seq = 0;
255
#endif
256
	tsk->splice_pipe = NULL;
L
Linus Torvalds 已提交
257
	return tsk;
258 259 260 261 262

out:
	free_thread_info(ti);
	free_task_struct(tsk);
	return NULL;
L
Linus Torvalds 已提交
263 264 265
}

#ifdef CONFIG_MMU
A
Alexey Dobriyan 已提交
266
static int dup_mmap(struct mm_struct *mm, struct mm_struct *oldmm)
L
Linus Torvalds 已提交
267
{
H
Hugh Dickins 已提交
268
	struct vm_area_struct *mpnt, *tmp, **pprev;
L
Linus Torvalds 已提交
269 270 271 272 273 274
	struct rb_node **rb_link, *rb_parent;
	int retval;
	unsigned long charge;
	struct mempolicy *pol;

	down_write(&oldmm->mmap_sem);
275
	flush_cache_dup_mm(oldmm);
276 277 278 279
	/*
	 * Not linked in yet - no deadlock potential:
	 */
	down_write_nested(&mm->mmap_sem, SINGLE_DEPTH_NESTING);
280

L
Linus Torvalds 已提交
281 282 283 284
	mm->locked_vm = 0;
	mm->mmap = NULL;
	mm->mmap_cache = NULL;
	mm->free_area_cache = oldmm->mmap_base;
285
	mm->cached_hole_size = ~0UL;
L
Linus Torvalds 已提交
286
	mm->map_count = 0;
287
	cpumask_clear(mm_cpumask(mm));
L
Linus Torvalds 已提交
288 289 290 291 292
	mm->mm_rb = RB_ROOT;
	rb_link = &mm->mm_rb.rb_node;
	rb_parent = NULL;
	pprev = &mm->mmap;

H
Hugh Dickins 已提交
293
	for (mpnt = oldmm->mmap; mpnt; mpnt = mpnt->vm_next) {
L
Linus Torvalds 已提交
294 295 296
		struct file *file;

		if (mpnt->vm_flags & VM_DONTCOPY) {
297 298
			long pages = vma_pages(mpnt);
			mm->total_vm -= pages;
299
			vm_stat_account(mm, mpnt->vm_flags, mpnt->vm_file,
300
								-pages);
L
Linus Torvalds 已提交
301 302 303 304 305 306 307 308 309
			continue;
		}
		charge = 0;
		if (mpnt->vm_flags & VM_ACCOUNT) {
			unsigned int len = (mpnt->vm_end - mpnt->vm_start) >> PAGE_SHIFT;
			if (security_vm_enough_memory(len))
				goto fail_nomem;
			charge = len;
		}
310
		tmp = kmem_cache_alloc(vm_area_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
311 312 313
		if (!tmp)
			goto fail_nomem;
		*tmp = *mpnt;
314
		pol = mpol_dup(vma_policy(mpnt));
L
Linus Torvalds 已提交
315 316 317 318 319 320 321 322 323 324
		retval = PTR_ERR(pol);
		if (IS_ERR(pol))
			goto fail_nomem_policy;
		vma_set_policy(tmp, pol);
		tmp->vm_flags &= ~VM_LOCKED;
		tmp->vm_mm = mm;
		tmp->vm_next = NULL;
		anon_vma_link(tmp);
		file = tmp->vm_file;
		if (file) {
325
			struct inode *inode = file->f_path.dentry->d_inode;
H
Hugh Dickins 已提交
326 327
			struct address_space *mapping = file->f_mapping;

L
Linus Torvalds 已提交
328 329 330
			get_file(file);
			if (tmp->vm_flags & VM_DENYWRITE)
				atomic_dec(&inode->i_writecount);
H
Hugh Dickins 已提交
331 332 333
			spin_lock(&mapping->i_mmap_lock);
			if (tmp->vm_flags & VM_SHARED)
				mapping->i_mmap_writable++;
L
Linus Torvalds 已提交
334
			tmp->vm_truncate_count = mpnt->vm_truncate_count;
H
Hugh Dickins 已提交
335 336
			flush_dcache_mmap_lock(mapping);
			/* insert tmp into the share list, just after mpnt */
L
Linus Torvalds 已提交
337
			vma_prio_tree_add(tmp, mpnt);
H
Hugh Dickins 已提交
338 339
			flush_dcache_mmap_unlock(mapping);
			spin_unlock(&mapping->i_mmap_lock);
L
Linus Torvalds 已提交
340 341
		}

342 343 344 345 346 347 348 349
		/*
		 * 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);

L
Linus Torvalds 已提交
350
		/*
351
		 * Link in the new vma and copy the page table entries.
L
Linus Torvalds 已提交
352 353 354 355 356 357 358 359 360
		 */
		*pprev = tmp;
		pprev = &tmp->vm_next;

		__vma_link_rb(mm, tmp, rb_link, rb_parent);
		rb_link = &tmp->vm_rb.rb_right;
		rb_parent = &tmp->vm_rb;

		mm->map_count++;
361
		retval = copy_page_range(mm, oldmm, mpnt);
L
Linus Torvalds 已提交
362 363 364 365 366 367 368

		if (tmp->vm_ops && tmp->vm_ops->open)
			tmp->vm_ops->open(tmp);

		if (retval)
			goto out;
	}
369 370
	/* a new mm has just been created */
	arch_dup_mmap(oldmm, mm);
L
Linus Torvalds 已提交
371 372
	retval = 0;
out:
373
	up_write(&mm->mmap_sem);
H
Hugh Dickins 已提交
374
	flush_tlb_mm(oldmm);
L
Linus Torvalds 已提交
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
	up_write(&oldmm->mmap_sem);
	return retval;
fail_nomem_policy:
	kmem_cache_free(vm_area_cachep, tmp);
fail_nomem:
	retval = -ENOMEM;
	vm_unacct_memory(charge);
	goto out;
}

static inline int mm_alloc_pgd(struct mm_struct * mm)
{
	mm->pgd = pgd_alloc(mm);
	if (unlikely(!mm->pgd))
		return -ENOMEM;
	return 0;
}

static inline void mm_free_pgd(struct mm_struct * mm)
{
395
	pgd_free(mm, mm->pgd);
L
Linus Torvalds 已提交
396 397 398 399 400 401 402
}
#else
#define dup_mmap(mm, oldmm)	(0)
#define mm_alloc_pgd(mm)	(0)
#define mm_free_pgd(mm)
#endif /* CONFIG_MMU */

D
Daniel Walker 已提交
403
__cacheline_aligned_in_smp DEFINE_SPINLOCK(mmlist_lock);
L
Linus Torvalds 已提交
404

405
#define allocate_mm()	(kmem_cache_alloc(mm_cachep, GFP_KERNEL))
L
Linus Torvalds 已提交
406 407
#define free_mm(mm)	(kmem_cache_free(mm_cachep, (mm)))

408 409 410 411 412 413 414 415 416 417 418 419
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);

L
Linus Torvalds 已提交
420 421
#include <linux/init_task.h>

422
static struct mm_struct * mm_init(struct mm_struct * mm, struct task_struct *p)
L
Linus Torvalds 已提交
423 424 425 426 427
{
	atomic_set(&mm->mm_users, 1);
	atomic_set(&mm->mm_count, 1);
	init_rwsem(&mm->mmap_sem);
	INIT_LIST_HEAD(&mm->mmlist);
428
	mm->flags = (current->mm) ? current->mm->flags : default_dump_filter;
429
	mm->oom_adj = (current->mm) ? current->mm->oom_adj : 0;
430
	mm->core_state = NULL;
L
Linus Torvalds 已提交
431
	mm->nr_ptes = 0;
432
	set_mm_counter(mm, file_rss, 0);
H
Hugh Dickins 已提交
433
	set_mm_counter(mm, anon_rss, 0);
L
Linus Torvalds 已提交
434
	spin_lock_init(&mm->page_table_lock);
J
Jens Axboe 已提交
435 436
	spin_lock_init(&mm->ioctx_lock);
	INIT_HLIST_HEAD(&mm->ioctx_list);
L
Linus Torvalds 已提交
437
	mm->free_area_cache = TASK_UNMAPPED_BASE;
438
	mm->cached_hole_size = ~0UL;
439
	mm_init_owner(mm, p);
L
Linus Torvalds 已提交
440 441 442

	if (likely(!mm_alloc_pgd(mm))) {
		mm->def_flags = 0;
A
Andrea Arcangeli 已提交
443
		mmu_notifier_mm_init(mm);
L
Linus Torvalds 已提交
444 445
		return mm;
	}
446

L
Linus Torvalds 已提交
447 448 449 450 451 452 453 454 455 456 457 458 459 460
	free_mm(mm);
	return NULL;
}

/*
 * Allocate and initialize an mm_struct.
 */
struct mm_struct * mm_alloc(void)
{
	struct mm_struct * mm;

	mm = allocate_mm();
	if (mm) {
		memset(mm, 0, sizeof(*mm));
461
		mm = mm_init(mm, current);
L
Linus Torvalds 已提交
462 463 464 465 466 467 468 469 470
	}
	return mm;
}

/*
 * 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.
 */
471
void __mmdrop(struct mm_struct *mm)
L
Linus Torvalds 已提交
472 473 474 475
{
	BUG_ON(mm == &init_mm);
	mm_free_pgd(mm);
	destroy_context(mm);
A
Andrea Arcangeli 已提交
476
	mmu_notifier_mm_destroy(mm);
L
Linus Torvalds 已提交
477 478
	free_mm(mm);
}
479
EXPORT_SYMBOL_GPL(__mmdrop);
L
Linus Torvalds 已提交
480 481 482 483 484 485

/*
 * Decrement the use count and release all resources for an mm.
 */
void mmput(struct mm_struct *mm)
{
A
Andrew Morton 已提交
486 487
	might_sleep();

L
Linus Torvalds 已提交
488 489 490
	if (atomic_dec_and_test(&mm->mm_users)) {
		exit_aio(mm);
		exit_mmap(mm);
M
Matt Helsley 已提交
491
		set_mm_exe_file(mm, NULL);
L
Linus Torvalds 已提交
492 493 494 495 496 497 498 499 500 501 502 503 504 505
		if (!list_empty(&mm->mmlist)) {
			spin_lock(&mmlist_lock);
			list_del(&mm->mmlist);
			spin_unlock(&mmlist_lock);
		}
		put_swap_token(mm);
		mmdrop(mm);
	}
}
EXPORT_SYMBOL_GPL(mmput);

/**
 * get_task_mm - acquire a reference to the task's mm
 *
506
 * Returns %NULL if the task has no mm.  Checks PF_KTHREAD (meaning
L
Linus Torvalds 已提交
507 508 509 510 511 512 513 514 515 516 517 518
 * 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) {
519
		if (task->flags & PF_KTHREAD)
L
Linus Torvalds 已提交
520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545
			mm = NULL;
		else
			atomic_inc(&mm->mm_users);
	}
	task_unlock(task);
	return mm;
}
EXPORT_SYMBOL_GPL(get_task_mm);

/* 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)
{
	struct completion *vfork_done = tsk->vfork_done;

546 547 548 549 550 551 552 553 554 555
	/* Get rid of any futexes when releasing the mm */
#ifdef CONFIG_FUTEX
	if (unlikely(tsk->robust_list))
		exit_robust_list(tsk);
#ifdef CONFIG_COMPAT
	if (unlikely(tsk->compat_robust_list))
		compat_exit_robust_list(tsk);
#endif
#endif

L
Linus Torvalds 已提交
556 557 558 559 560 561 562 563
	/* Get rid of any cached register state */
	deactivate_mm(tsk, mm);

	/* notify parent sleeping on vfork() */
	if (vfork_done) {
		tsk->vfork_done = NULL;
		complete(vfork_done);
	}
564 565 566 567 568 569 570 571 572 573

	/*
	 * If we're exiting normally, clear a user-space tid field if
	 * requested.  We leave this alone when dying by signal, to leave
	 * the value intact in a core dump, and to save the unnecessary
	 * trouble otherwise.  Userland only wants this done for a sys_exit.
	 */
	if (tsk->clear_child_tid
	    && !(tsk->flags & PF_SIGNALED)
	    && atomic_read(&mm->mm_users) > 1) {
L
Linus Torvalds 已提交
574 575 576 577 578 579 580 581 582 583 584 585
		u32 __user * tidptr = tsk->clear_child_tid;
		tsk->clear_child_tid = NULL;

		/*
		 * We don't check the error code - if userspace has
		 * not set up a proper pointer then tough luck.
		 */
		put_user(0, tidptr);
		sys_futex(tidptr, FUTEX_WAKE, 1, NULL, NULL, 0);
	}
}

586 587 588 589
/*
 * Allocate a new mm structure and copy contents from the
 * mm structure of the passed in task structure.
 */
590
struct mm_struct *dup_mm(struct task_struct *tsk)
591 592 593 594 595 596 597 598 599 600 601 602 603
{
	struct mm_struct *mm, *oldmm = current->mm;
	int err;

	if (!oldmm)
		return NULL;

	mm = allocate_mm();
	if (!mm)
		goto fail_nomem;

	memcpy(mm, oldmm, sizeof(*mm));

604 605 606 607
	/* Initializing for Swap token stuff */
	mm->token_priority = 0;
	mm->last_interval = 0;

608
	if (!mm_init(mm, tsk))
609 610 611 612 613
		goto fail_nomem;

	if (init_new_context(tsk, mm))
		goto fail_nocontext;

M
Matt Helsley 已提交
614 615
	dup_mm_exe_file(oldmm, mm);

616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
	err = dup_mmap(mm, oldmm);
	if (err)
		goto free_pt;

	mm->hiwater_rss = get_mm_rss(mm);
	mm->hiwater_vm = mm->total_vm;

	return mm;

free_pt:
	mmput(mm);

fail_nomem:
	return NULL;

fail_nocontext:
	/*
	 * If init_new_context() failed, we cannot use mmput() to free the mm
	 * because it calls destroy_context()
	 */
	mm_free_pgd(mm);
	free_mm(mm);
	return NULL;
}

L
Linus Torvalds 已提交
641 642 643 644 645 646 647
static int copy_mm(unsigned long clone_flags, struct task_struct * tsk)
{
	struct mm_struct * mm, *oldmm;
	int retval;

	tsk->min_flt = tsk->maj_flt = 0;
	tsk->nvcsw = tsk->nivcsw = 0;
648 649 650
#ifdef CONFIG_DETECT_HUNG_TASK
	tsk->last_switch_count = tsk->nvcsw + tsk->nivcsw;
#endif
L
Linus Torvalds 已提交
651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670

	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;

	if (clone_flags & CLONE_VM) {
		atomic_inc(&oldmm->mm_users);
		mm = oldmm;
		goto good_mm;
	}

	retval = -ENOMEM;
671
	mm = dup_mm(tsk);
L
Linus Torvalds 已提交
672 673 674 675
	if (!mm)
		goto fail_nomem;

good_mm:
676 677 678 679
	/* Initializing for Swap token stuff */
	mm->token_priority = 0;
	mm->last_interval = 0;

L
Linus Torvalds 已提交
680 681 682 683 684 685 686 687
	tsk->mm = mm;
	tsk->active_mm = mm;
	return 0;

fail_nomem:
	return retval;
}

A
Alexey Dobriyan 已提交
688
static int copy_fs(unsigned long clone_flags, struct task_struct *tsk)
L
Linus Torvalds 已提交
689
{
A
Al Viro 已提交
690
	struct fs_struct *fs = current->fs;
L
Linus Torvalds 已提交
691
	if (clone_flags & CLONE_FS) {
A
Al Viro 已提交
692 693 694 695 696 697 698 699
		/* tsk->fs is already what we want */
		write_lock(&fs->lock);
		if (fs->in_exec) {
			write_unlock(&fs->lock);
			return -EAGAIN;
		}
		fs->users++;
		write_unlock(&fs->lock);
L
Linus Torvalds 已提交
700 701
		return 0;
	}
A
Al Viro 已提交
702
	tsk->fs = copy_fs_struct(fs);
L
Linus Torvalds 已提交
703 704 705 706 707
	if (!tsk->fs)
		return -ENOMEM;
	return 0;
}

708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734
static int copy_files(unsigned long clone_flags, struct task_struct * tsk)
{
	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;
	}

	newf = dup_fd(oldf, &error);
	if (!newf)
		goto out;

	tsk->files = newf;
	error = 0;
out:
	return error;
}

735
static int copy_io(unsigned long clone_flags, struct task_struct *tsk)
736 737 738 739 740 741
{
#ifdef CONFIG_BLOCK
	struct io_context *ioc = current->io_context;

	if (!ioc)
		return 0;
742 743 744 745 746 747 748 749
	/*
	 * Share io context with parent, if CLONE_IO is set
	 */
	if (clone_flags & CLONE_IO) {
		tsk->io_context = ioc_task_link(ioc);
		if (unlikely(!tsk->io_context))
			return -ENOMEM;
	} else if (ioprio_valid(ioc->ioprio)) {
750 751 752 753 754 755 756 757 758 759
		tsk->io_context = alloc_io_context(GFP_KERNEL, -1);
		if (unlikely(!tsk->io_context))
			return -ENOMEM;

		tsk->io_context->ioprio = ioc->ioprio;
	}
#endif
	return 0;
}

A
Alexey Dobriyan 已提交
760
static int copy_sighand(unsigned long clone_flags, struct task_struct *tsk)
L
Linus Torvalds 已提交
761 762 763
{
	struct sighand_struct *sig;

Z
Zhaolei 已提交
764
	if (clone_flags & CLONE_SIGHAND) {
L
Linus Torvalds 已提交
765 766 767 768
		atomic_inc(&current->sighand->count);
		return 0;
	}
	sig = kmem_cache_alloc(sighand_cachep, GFP_KERNEL);
I
Ingo Molnar 已提交
769
	rcu_assign_pointer(tsk->sighand, sig);
L
Linus Torvalds 已提交
770 771 772 773 774 775 776
	if (!sig)
		return -ENOMEM;
	atomic_set(&sig->count, 1);
	memcpy(sig->action, current->sighand->action, sizeof(sig->action));
	return 0;
}

777
void __cleanup_sighand(struct sighand_struct *sighand)
778 779 780 781 782
{
	if (atomic_dec_and_test(&sighand->count))
		kmem_cache_free(sighand_cachep, sighand);
}

783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802

/*
 * Initialize POSIX timer handling for a thread group.
 */
static void posix_cpu_timers_init_group(struct signal_struct *sig)
{
	/* Thread group counters. */
	thread_group_cputime_init(sig);

	/* Expiration times and increments. */
	sig->it_virt_expires = cputime_zero;
	sig->it_virt_incr = cputime_zero;
	sig->it_prof_expires = cputime_zero;
	sig->it_prof_incr = cputime_zero;

	/* Cached expiration times. */
	sig->cputime_expires.prof_exp = cputime_zero;
	sig->cputime_expires.virt_exp = cputime_zero;
	sig->cputime_expires.sched_exp = 0;

803 804 805 806 807 808
	if (sig->rlim[RLIMIT_CPU].rlim_cur != RLIM_INFINITY) {
		sig->cputime_expires.prof_exp =
			secs_to_cputime(sig->rlim[RLIMIT_CPU].rlim_cur);
		sig->cputimer.running = 1;
	}

809 810 811 812 813 814
	/* The timer lists. */
	INIT_LIST_HEAD(&sig->cpu_timers[0]);
	INIT_LIST_HEAD(&sig->cpu_timers[1]);
	INIT_LIST_HEAD(&sig->cpu_timers[2]);
}

A
Alexey Dobriyan 已提交
815
static int copy_signal(unsigned long clone_flags, struct task_struct *tsk)
L
Linus Torvalds 已提交
816 817 818 819
{
	struct signal_struct *sig;

	if (clone_flags & CLONE_THREAD) {
820 821 822
		atomic_inc(&current->signal->count);
		atomic_inc(&current->signal->live);
		return 0;
L
Linus Torvalds 已提交
823
	}
824

825
	sig = kmem_cache_alloc(signal_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
826 827 828 829 830 831 832 833
	tsk->signal = sig;
	if (!sig)
		return -ENOMEM;

	atomic_set(&sig->count, 1);
	atomic_set(&sig->live, 1);
	init_waitqueue_head(&sig->wait_chldexit);
	sig->flags = 0;
834 835
	if (clone_flags & CLONE_NEWPID)
		sig->flags |= SIGNAL_UNKILLABLE;
L
Linus Torvalds 已提交
836 837 838
	sig->group_exit_code = 0;
	sig->group_exit_task = NULL;
	sig->group_stop_count = 0;
839
	sig->curr_target = tsk;
L
Linus Torvalds 已提交
840 841 842
	init_sigpending(&sig->shared_pending);
	INIT_LIST_HEAD(&sig->posix_timers);

843
	hrtimer_init(&sig->real_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL);
844
	sig->it_real_incr.tv64 = 0;
L
Linus Torvalds 已提交
845 846 847
	sig->real_timer.function = it_real_fn;

	sig->leader = 0;	/* session leadership doesn't inherit */
848
	sig->tty_old_pgrp = NULL;
A
Alan Cox 已提交
849
	sig->tty = NULL;
L
Linus Torvalds 已提交
850

851
	sig->utime = sig->stime = sig->cutime = sig->cstime = cputime_zero;
852 853
	sig->gtime = cputime_zero;
	sig->cgtime = cputime_zero;
L
Linus Torvalds 已提交
854 855
	sig->nvcsw = sig->nivcsw = sig->cnvcsw = sig->cnivcsw = 0;
	sig->min_flt = sig->maj_flt = sig->cmin_flt = sig->cmaj_flt = 0;
856
	sig->inblock = sig->oublock = sig->cinblock = sig->coublock = 0;
857
	task_io_accounting_init(&sig->ioac);
858
	sig->sum_sched_runtime = 0;
859
	taskstats_tgid_init(sig);
L
Linus Torvalds 已提交
860 861 862 863 864

	task_lock(current->group_leader);
	memcpy(sig->rlim, current->signal->rlim, sizeof sig->rlim);
	task_unlock(current->group_leader);

865 866
	posix_cpu_timers_init_group(sig);

867
	acct_init_pacct(&sig->pacct);
L
Linus Torvalds 已提交
868

M
Miloslav Trmac 已提交
869 870
	tty_audit_fork(sig);

L
Linus Torvalds 已提交
871 872 873
	return 0;
}

874 875
void __cleanup_signal(struct signal_struct *sig)
{
876
	thread_group_cputime_free(sig);
A
Alan Cox 已提交
877
	tty_kref_put(sig->tty);
878 879 880
	kmem_cache_free(signal_cachep, sig);
}

A
Alexey Dobriyan 已提交
881
static void cleanup_signal(struct task_struct *tsk)
882 883 884 885 886 887 888 889 890
{
	struct signal_struct *sig = tsk->signal;

	atomic_dec(&sig->live);

	if (atomic_dec_and_test(&sig->count))
		__cleanup_signal(sig);
}

A
Alexey Dobriyan 已提交
891
static void copy_flags(unsigned long clone_flags, struct task_struct *p)
L
Linus Torvalds 已提交
892 893 894
{
	unsigned long new_flags = p->flags;

895
	new_flags &= ~PF_SUPERPRIV;
L
Linus Torvalds 已提交
896
	new_flags |= PF_FORKNOEXEC;
R
Roland McGrath 已提交
897
	new_flags |= PF_STARTING;
L
Linus Torvalds 已提交
898
	p->flags = new_flags;
899
	clear_freeze_flag(p);
L
Linus Torvalds 已提交
900 901
}

902
SYSCALL_DEFINE1(set_tid_address, int __user *, tidptr)
L
Linus Torvalds 已提交
903 904 905
{
	current->clear_child_tid = tidptr;

906
	return task_pid_vnr(current);
L
Linus Torvalds 已提交
907 908
}

A
Alexey Dobriyan 已提交
909
static void rt_mutex_init_task(struct task_struct *p)
I
Ingo Molnar 已提交
910 911
{
	spin_lock_init(&p->pi_lock);
912
#ifdef CONFIG_RT_MUTEXES
I
Ingo Molnar 已提交
913 914 915 916 917
	plist_head_init(&p->pi_waiters, &p->pi_lock);
	p->pi_blocked_on = NULL;
#endif
}

918 919 920 921 922 923 924
#ifdef CONFIG_MM_OWNER
void mm_init_owner(struct mm_struct *mm, struct task_struct *p)
{
	mm->owner = p;
}
#endif /* CONFIG_MM_OWNER */

925 926 927 928 929 930 931 932 933 934 935 936 937
/*
 * Initialize POSIX timer handling for a single task.
 */
static void posix_cpu_timers_init(struct task_struct *tsk)
{
	tsk->cputime_expires.prof_exp = cputime_zero;
	tsk->cputime_expires.virt_exp = cputime_zero;
	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]);
}

L
Linus Torvalds 已提交
938 939 940 941 942 943 944 945
/*
 * 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.
 */
946 947 948 949 950
static struct task_struct *copy_process(unsigned long clone_flags,
					unsigned long stack_start,
					struct pt_regs *regs,
					unsigned long stack_size,
					int __user *child_tidptr,
R
Roland McGrath 已提交
951 952
					struct pid *pid,
					int trace)
L
Linus Torvalds 已提交
953 954
{
	int retval;
955
	struct task_struct *p;
956
	int cgroup_callbacks_done = 0;
L
Linus Torvalds 已提交
957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984

	if ((clone_flags & (CLONE_NEWNS|CLONE_FS)) == (CLONE_NEWNS|CLONE_FS))
		return ERR_PTR(-EINVAL);

	/*
	 * 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);

	retval = security_task_create(clone_flags);
	if (retval)
		goto fork_out;

	retval = -ENOMEM;
	p = dup_task_struct(current);
	if (!p)
		goto fork_out;

985 986
	ftrace_graph_init_task(p);

987 988
	rt_mutex_init_task(p);

I
Ingo Molnar 已提交
989
#ifdef CONFIG_PROVE_LOCKING
990 991 992
	DEBUG_LOCKS_WARN_ON(!p->hardirqs_enabled);
	DEBUG_LOCKS_WARN_ON(!p->softirqs_enabled);
#endif
L
Linus Torvalds 已提交
993
	retval = -EAGAIN;
994
	if (atomic_read(&p->real_cred->user->processes) >=
L
Linus Torvalds 已提交
995 996
			p->signal->rlim[RLIMIT_NPROC].rlim_cur) {
		if (!capable(CAP_SYS_ADMIN) && !capable(CAP_SYS_RESOURCE) &&
997
		    p->real_cred->user != INIT_USER)
L
Linus Torvalds 已提交
998 999 1000
			goto bad_fork_free;
	}

1001 1002 1003
	retval = copy_creds(p, clone_flags);
	if (retval < 0)
		goto bad_fork_free;
L
Linus Torvalds 已提交
1004 1005 1006 1007 1008 1009

	/*
	 * 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.
	 */
1010
	retval = -EAGAIN;
L
Linus Torvalds 已提交
1011 1012 1013
	if (nr_threads >= max_threads)
		goto bad_fork_cleanup_count;

A
Al Viro 已提交
1014
	if (!try_module_get(task_thread_info(p)->exec_domain->module))
L
Linus Torvalds 已提交
1015 1016 1017 1018 1019 1020
		goto bad_fork_cleanup_count;

	if (p->binfmt && !try_module_get(p->binfmt->module))
		goto bad_fork_cleanup_put_domain;

	p->did_exec = 0;
1021
	delayacct_tsk_init(p);	/* Must remain after dup_task_struct() */
L
Linus Torvalds 已提交
1022 1023 1024
	copy_flags(clone_flags, p);
	INIT_LIST_HEAD(&p->children);
	INIT_LIST_HEAD(&p->sibling);
P
Paul E. McKenney 已提交
1025 1026 1027 1028
#ifdef CONFIG_PREEMPT_RCU
	p->rcu_read_lock_nesting = 0;
	p->rcu_flipctr_idx = 0;
#endif /* #ifdef CONFIG_PREEMPT_RCU */
L
Linus Torvalds 已提交
1029 1030 1031 1032 1033 1034 1035
	p->vfork_done = NULL;
	spin_lock_init(&p->alloc_lock);

	init_sigpending(&p->pending);

	p->utime = cputime_zero;
	p->stime = cputime_zero;
1036
	p->gtime = cputime_zero;
1037 1038
	p->utimescaled = cputime_zero;
	p->stimescaled = cputime_zero;
P
Peter Zijlstra 已提交
1039
	p->prev_utime = cputime_zero;
1040
	p->prev_stime = cputime_zero;
1041

1042 1043
	p->default_timer_slack_ns = current->timer_slack_ns;

1044
	task_io_accounting_init(&p->ioac);
L
Linus Torvalds 已提交
1045 1046
	acct_clear_integrals(p);

1047
	posix_cpu_timers_init(p);
L
Linus Torvalds 已提交
1048 1049 1050

	p->lock_depth = -1;		/* -1 = no lock */
	do_posix_clock_monotonic_gettime(&p->start_time);
1051 1052
	p->real_start_time = p->start_time;
	monotonic_to_bootbased(&p->real_start_time);
L
Linus Torvalds 已提交
1053 1054
	p->io_context = NULL;
	p->audit_context = NULL;
1055
	cgroup_fork(p);
L
Linus Torvalds 已提交
1056
#ifdef CONFIG_NUMA
1057
	p->mempolicy = mpol_dup(p->mempolicy);
L
Linus Torvalds 已提交
1058 1059 1060
 	if (IS_ERR(p->mempolicy)) {
 		retval = PTR_ERR(p->mempolicy);
 		p->mempolicy = NULL;
1061
 		goto bad_fork_cleanup_cgroup;
L
Linus Torvalds 已提交
1062
 	}
1063
	mpol_fix_fork_child_flag(p);
L
Linus Torvalds 已提交
1064
#endif
1065 1066
#ifdef CONFIG_TRACE_IRQFLAGS
	p->irq_events = 0;
1067 1068 1069
#ifdef __ARCH_WANT_INTERRUPTS_ON_CTXSW
	p->hardirqs_enabled = 1;
#else
1070
	p->hardirqs_enabled = 0;
1071
#endif
1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083
	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
I
Ingo Molnar 已提交
1084 1085 1086 1087 1088
#ifdef CONFIG_LOCKDEP
	p->lockdep_depth = 0; /* no locks held yet */
	p->curr_chain_key = 0;
	p->lockdep_recursion = 0;
#endif
L
Linus Torvalds 已提交
1089

1090 1091 1092
#ifdef CONFIG_DEBUG_MUTEXES
	p->blocked_on = NULL; /* not blocked yet */
#endif
1093 1094

	p->bts = NULL;
1095

1096 1097
	/* Perform scheduler related setup. Assign this task to a CPU. */
	sched_fork(p, clone_flags);
1098 1099 1100 1101

	retval = perf_counter_init_task(p);
	if (retval)
		goto bad_fork_cleanup_policy;
1102

L
Linus Torvalds 已提交
1103
	if ((retval = audit_alloc(p)))
1104
		goto bad_fork_cleanup_policy;
L
Linus Torvalds 已提交
1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117
	/* copy all the process information */
	if ((retval = copy_semundo(clone_flags, p)))
		goto bad_fork_cleanup_audit;
	if ((retval = copy_files(clone_flags, p)))
		goto bad_fork_cleanup_semundo;
	if ((retval = copy_fs(clone_flags, p)))
		goto bad_fork_cleanup_files;
	if ((retval = copy_sighand(clone_flags, p)))
		goto bad_fork_cleanup_fs;
	if ((retval = copy_signal(clone_flags, p)))
		goto bad_fork_cleanup_sighand;
	if ((retval = copy_mm(clone_flags, p)))
		goto bad_fork_cleanup_signal;
S
Serge E. Hallyn 已提交
1118
	if ((retval = copy_namespaces(clone_flags, p)))
D
David Howells 已提交
1119
		goto bad_fork_cleanup_mm;
1120
	if ((retval = copy_io(clone_flags, p)))
1121
		goto bad_fork_cleanup_namespaces;
A
Alexey Dobriyan 已提交
1122
	retval = copy_thread(clone_flags, stack_start, stack_size, p, regs);
L
Linus Torvalds 已提交
1123
	if (retval)
1124
		goto bad_fork_cleanup_io;
L
Linus Torvalds 已提交
1125

1126 1127
	if (pid != &init_struct_pid) {
		retval = -ENOMEM;
1128
		pid = alloc_pid(p->nsproxy->pid_ns);
1129
		if (!pid)
1130
			goto bad_fork_cleanup_io;
1131 1132

		if (clone_flags & CLONE_NEWPID) {
1133
			retval = pid_ns_prepare_proc(p->nsproxy->pid_ns);
1134 1135 1136
			if (retval < 0)
				goto bad_fork_free_pid;
		}
1137 1138 1139 1140 1141 1142 1143
	}

	p->pid = pid_nr(pid);
	p->tgid = p->pid;
	if (clone_flags & CLONE_THREAD)
		p->tgid = current->tgid;

1144 1145 1146
	if (current->nsproxy != p->nsproxy) {
		retval = ns_cgroup_clone(p, pid);
		if (retval)
1147
			goto bad_fork_free_pid;
1148 1149
	}

L
Linus Torvalds 已提交
1150 1151 1152 1153 1154
	p->set_child_tid = (clone_flags & CLONE_CHILD_SETTID) ? child_tidptr : NULL;
	/*
	 * Clear TID on mm_release()?
	 */
	p->clear_child_tid = (clone_flags & CLONE_CHILD_CLEARTID) ? child_tidptr: NULL;
1155
#ifdef CONFIG_FUTEX
1156 1157 1158 1159
	p->robust_list = NULL;
#ifdef CONFIG_COMPAT
	p->compat_robust_list = NULL;
#endif
1160 1161
	INIT_LIST_HEAD(&p->pi_state_list);
	p->pi_state_cache = NULL;
1162
#endif
1163 1164 1165 1166 1167 1168
	/*
	 * sigaltstack should be cleared when sharing the same VM
	 */
	if ((clone_flags & (CLONE_VM|CLONE_VFORK)) == CLONE_VM)
		p->sas_ss_sp = p->sas_ss_size = 0;

L
Linus Torvalds 已提交
1169 1170 1171 1172 1173
	/*
	 * Syscall tracing should be turned off in the child regardless
	 * of CLONE_PTRACE.
	 */
	clear_tsk_thread_flag(p, TIF_SYSCALL_TRACE);
1174 1175 1176
#ifdef TIF_SYSCALL_EMU
	clear_tsk_thread_flag(p, TIF_SYSCALL_EMU);
#endif
A
Arjan van de Ven 已提交
1177
	clear_all_latency_tracing(p);
L
Linus Torvalds 已提交
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188

	/* ok, now we should be set up.. */
	p->exit_signal = (clone_flags & CLONE_THREAD) ? -1 : (clone_flags & CSIGNAL);
	p->pdeath_signal = 0;
	p->exit_state = 0;

	/*
	 * Ok, make it visible to the rest of the system.
	 * We dont wake it up yet.
	 */
	p->group_leader = p;
O
Oleg Nesterov 已提交
1189
	INIT_LIST_HEAD(&p->thread_group);
L
Linus Torvalds 已提交
1190

1191 1192 1193 1194 1195 1196
	/* Now that the task is set up, run cgroup callbacks if
	 * necessary. We need to run them before the task is visible
	 * on the tasklist. */
	cgroup_fork_callbacks(p);
	cgroup_callbacks_done = 1;

L
Linus Torvalds 已提交
1197 1198 1199 1200
	/* Need tasklist lock for parent etc handling! */
	write_lock_irq(&tasklist_lock);

	/*
N
Nick Piggin 已提交
1201 1202 1203 1204 1205 1206 1207
	 * The task hasn't been attached yet, so its cpus_allowed mask will
	 * not be changed, nor will its assigned CPU.
	 *
	 * The cpus_allowed mask of the parent may have changed after it was
	 * copied first time - so re-copy it here, then check the child's CPU
	 * to ensure it is on a valid CPU (and if not, just force it back to
	 * parent's CPU). This avoids alot of nasty races.
L
Linus Torvalds 已提交
1208 1209
	 */
	p->cpus_allowed = current->cpus_allowed;
P
Peter Zijlstra 已提交
1210
	p->rt.nr_cpus_allowed = current->rt.nr_cpus_allowed;
1211 1212
	if (unlikely(!cpu_isset(task_cpu(p), p->cpus_allowed) ||
			!cpu_online(task_cpu(p))))
N
Nick Piggin 已提交
1213
		set_task_cpu(p, smp_processor_id());
L
Linus Torvalds 已提交
1214 1215

	/* CLONE_PARENT re-uses the old parent */
1216
	if (clone_flags & (CLONE_PARENT|CLONE_THREAD)) {
L
Linus Torvalds 已提交
1217
		p->real_parent = current->real_parent;
1218 1219
		p->parent_exec_id = current->parent_exec_id;
	} else {
L
Linus Torvalds 已提交
1220
		p->real_parent = current;
1221 1222
		p->parent_exec_id = current->self_exec_id;
	}
L
Linus Torvalds 已提交
1223

1224
	spin_lock(&current->sighand->siglock);
1225 1226 1227 1228 1229 1230 1231 1232 1233

	/*
	 * 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).
 	 */
D
Daniel Walker 已提交
1234
	recalc_sigpending();
1235 1236 1237 1238
	if (signal_pending(current)) {
		spin_unlock(&current->sighand->siglock);
		write_unlock_irq(&tasklist_lock);
		retval = -ERESTARTNOINTR;
1239
		goto bad_fork_free_pid;
1240 1241
	}

L
Linus Torvalds 已提交
1242 1243
	if (clone_flags & CLONE_THREAD) {
		p->group_leader = current->group_leader;
O
Oleg Nesterov 已提交
1244
		list_add_tail_rcu(&p->thread_group, &p->group_leader->thread_group);
L
Linus Torvalds 已提交
1245 1246
	}

1247
	if (likely(p->pid)) {
R
Roland McGrath 已提交
1248
		list_add_tail(&p->sibling, &p->real_parent->children);
R
Roland McGrath 已提交
1249
		tracehook_finish_clone(p, clone_flags, trace);
1250 1251

		if (thread_group_leader(p)) {
E
Eric W. Biederman 已提交
1252
			if (clone_flags & CLONE_NEWPID)
1253
				p->nsproxy->pid_ns->child_reaper = p;
1254

1255
			p->signal->leader_pid = pid;
A
Alan Cox 已提交
1256 1257
			tty_kref_put(p->signal->tty);
			p->signal->tty = tty_kref_get(current->signal->tty);
E
Eric W. Biederman 已提交
1258 1259
			attach_pid(p, PIDTYPE_PGID, task_pgrp(current));
			attach_pid(p, PIDTYPE_SID, task_session(current));
1260
			list_add_tail_rcu(&p->tasks, &init_task.tasks);
L
Linus Torvalds 已提交
1261
			__get_cpu_var(process_counts)++;
1262
		}
1263
		attach_pid(p, PIDTYPE_PID, pid);
1264
		nr_threads++;
L
Linus Torvalds 已提交
1265 1266 1267
	}

	total_forks++;
1268
	spin_unlock(&current->sighand->siglock);
L
Linus Torvalds 已提交
1269
	write_unlock_irq(&tasklist_lock);
1270
	proc_fork_connector(p);
1271
	cgroup_post_fork(p);
L
Linus Torvalds 已提交
1272 1273
	return p;

1274 1275 1276
bad_fork_free_pid:
	if (pid != &init_struct_pid)
		free_pid(pid);
1277 1278
bad_fork_cleanup_io:
	put_io_context(p->io_context);
S
Serge E. Hallyn 已提交
1279
bad_fork_cleanup_namespaces:
1280
	exit_task_namespaces(p);
L
Linus Torvalds 已提交
1281 1282 1283 1284
bad_fork_cleanup_mm:
	if (p->mm)
		mmput(p->mm);
bad_fork_cleanup_signal:
1285
	cleanup_signal(p);
L
Linus Torvalds 已提交
1286
bad_fork_cleanup_sighand:
1287
	__cleanup_sighand(p->sighand);
L
Linus Torvalds 已提交
1288 1289 1290 1291 1292 1293 1294 1295 1296
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);
bad_fork_cleanup_policy:
1297
	perf_counter_free_task(p);
L
Linus Torvalds 已提交
1298
#ifdef CONFIG_NUMA
1299
	mpol_put(p->mempolicy);
1300
bad_fork_cleanup_cgroup:
L
Linus Torvalds 已提交
1301
#endif
1302
	cgroup_exit(p, cgroup_callbacks_done);
1303
	delayacct_tsk_free(p);
L
Linus Torvalds 已提交
1304 1305 1306
	if (p->binfmt)
		module_put(p->binfmt->module);
bad_fork_cleanup_put_domain:
A
Al Viro 已提交
1307
	module_put(task_thread_info(p)->exec_domain->module);
L
Linus Torvalds 已提交
1308
bad_fork_cleanup_count:
D
David Howells 已提交
1309
	atomic_dec(&p->cred->user->processes);
1310
	put_cred(p->real_cred);
1311
	put_cred(p->cred);
L
Linus Torvalds 已提交
1312 1313
bad_fork_free:
	free_task(p);
1314 1315
fork_out:
	return ERR_PTR(retval);
L
Linus Torvalds 已提交
1316 1317
}

A
Adrian Bunk 已提交
1318
noinline struct pt_regs * __cpuinit __attribute__((weak)) idle_regs(struct pt_regs *regs)
L
Linus Torvalds 已提交
1319 1320 1321 1322 1323
{
	memset(regs, 0, sizeof(struct pt_regs));
	return regs;
}

1324
struct task_struct * __cpuinit fork_idle(int cpu)
L
Linus Torvalds 已提交
1325
{
1326
	struct task_struct *task;
L
Linus Torvalds 已提交
1327 1328
	struct pt_regs regs;

1329
	task = copy_process(CLONE_VM, 0, idle_regs(&regs), 0, NULL,
R
Roland McGrath 已提交
1330
			    &init_struct_pid, 0);
1331 1332
	if (!IS_ERR(task))
		init_idle(task, cpu);
1333

L
Linus Torvalds 已提交
1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
	return task;
}

/*
 *  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.
 */
long do_fork(unsigned long clone_flags,
	      unsigned long stack_start,
	      struct pt_regs *regs,
	      unsigned long stack_size,
	      int __user *parent_tidptr,
	      int __user *child_tidptr)
{
	struct task_struct *p;
	int trace = 0;
1352
	long nr;
L
Linus Torvalds 已提交
1353

1354 1355 1356 1357 1358 1359 1360 1361 1362 1363
	/*
	 * Do some preliminary argument and permissions checking before we
	 * actually start allocating stuff
	 */
	if (clone_flags & CLONE_NEWUSER) {
		if (clone_flags & CLONE_THREAD)
			return -EINVAL;
		/* hopefully this check will go away when userns support is
		 * complete
		 */
1364 1365
		if (!capable(CAP_SYS_ADMIN) || !capable(CAP_SETUID) ||
				!capable(CAP_SETGID))
1366 1367 1368
			return -EPERM;
	}

1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385
	/*
	 * We hope to recycle these flags after 2.6.26
	 */
	if (unlikely(clone_flags & CLONE_STOPPED)) {
		static int __read_mostly count = 100;

		if (count > 0 && printk_ratelimit()) {
			char comm[TASK_COMM_LEN];

			count--;
			printk(KERN_INFO "fork(): process `%s' used deprecated "
					"clone flags 0x%lx\n",
				get_task_comm(comm, current),
				clone_flags & CLONE_STOPPED);
		}
	}

R
Roland McGrath 已提交
1386 1387 1388 1389 1390
	/*
	 * When called from kernel_thread, don't do user tracing stuff.
	 */
	if (likely(user_mode(regs)))
		trace = tracehook_prepare_clone(clone_flags);
L
Linus Torvalds 已提交
1391

1392
	p = copy_process(clone_flags, stack_start, regs, stack_size,
R
Roland McGrath 已提交
1393
			 child_tidptr, NULL, trace);
L
Linus Torvalds 已提交
1394 1395 1396 1397 1398 1399 1400
	/*
	 * 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;

1401 1402
		trace_sched_process_fork(current, p);

1403
		nr = task_pid_vnr(p);
1404 1405 1406

		if (clone_flags & CLONE_PARENT_SETTID)
			put_user(nr, parent_tidptr);
1407

L
Linus Torvalds 已提交
1408 1409 1410 1411 1412
		if (clone_flags & CLONE_VFORK) {
			p->vfork_done = &vfork;
			init_completion(&vfork);
		}

1413 1414 1415
		if (!(clone_flags & CLONE_THREAD))
			perf_counter_fork(p);

1416
		audit_finish_fork(p);
1417
		tracehook_report_clone(regs, clone_flags, nr, p);
R
Roland McGrath 已提交
1418 1419 1420 1421 1422 1423 1424 1425 1426 1427

		/*
		 * We set PF_STARTING at creation in case tracing wants to
		 * use this to distinguish a fully live task from one that
		 * hasn't gotten to tracehook_report_clone() yet.  Now we
		 * clear it and set the child going.
		 */
		p->flags &= ~PF_STARTING;

		if (unlikely(clone_flags & CLONE_STOPPED)) {
L
Linus Torvalds 已提交
1428 1429 1430 1431 1432
			/*
			 * We'll start up with an immediate SIGSTOP.
			 */
			sigaddset(&p->pending.signal, SIGSTOP);
			set_tsk_thread_flag(p, TIF_SIGPENDING);
1433
			__set_task_state(p, TASK_STOPPED);
R
Roland McGrath 已提交
1434 1435
		} else {
			wake_up_new_task(p, clone_flags);
L
Linus Torvalds 已提交
1436 1437
		}

R
Roland McGrath 已提交
1438 1439 1440
		tracehook_report_clone_complete(trace, regs,
						clone_flags, nr, p);

L
Linus Torvalds 已提交
1441
		if (clone_flags & CLONE_VFORK) {
R
Rafael J. Wysocki 已提交
1442
			freezer_do_not_count();
L
Linus Torvalds 已提交
1443
			wait_for_completion(&vfork);
R
Rafael J. Wysocki 已提交
1444
			freezer_count();
R
Roland McGrath 已提交
1445
			tracehook_report_vfork_done(p, nr);
L
Linus Torvalds 已提交
1446 1447
		}
	} else {
1448
		nr = PTR_ERR(p);
L
Linus Torvalds 已提交
1449
	}
1450
	return nr;
L
Linus Torvalds 已提交
1451 1452
}

1453 1454 1455 1456
#ifndef ARCH_MIN_MMSTRUCT_ALIGN
#define ARCH_MIN_MMSTRUCT_ALIGN 0
#endif

1457
static void sighand_ctor(void *data)
1458 1459 1460
{
	struct sighand_struct *sighand = data;

C
Christoph Lameter 已提交
1461
	spin_lock_init(&sighand->siglock);
D
Davide Libenzi 已提交
1462
	init_waitqueue_head(&sighand->signalfd_wqh);
1463 1464
}

L
Linus Torvalds 已提交
1465 1466 1467 1468
void __init proc_caches_init(void)
{
	sighand_cachep = kmem_cache_create("sighand_cache",
			sizeof(struct sighand_struct), 0,
V
Vegard Nossum 已提交
1469 1470
			SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_DESTROY_BY_RCU|
			SLAB_NOTRACK, sighand_ctor);
L
Linus Torvalds 已提交
1471 1472
	signal_cachep = kmem_cache_create("signal_cache",
			sizeof(struct signal_struct), 0,
V
Vegard Nossum 已提交
1473
			SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK, NULL);
1474
	files_cachep = kmem_cache_create("files_cache",
L
Linus Torvalds 已提交
1475
			sizeof(struct files_struct), 0,
V
Vegard Nossum 已提交
1476
			SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK, NULL);
1477
	fs_cachep = kmem_cache_create("fs_cache",
L
Linus Torvalds 已提交
1478
			sizeof(struct fs_struct), 0,
V
Vegard Nossum 已提交
1479
			SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK, NULL);
L
Linus Torvalds 已提交
1480
	mm_cachep = kmem_cache_create("mm_struct",
1481
			sizeof(struct mm_struct), ARCH_MIN_MMSTRUCT_ALIGN,
V
Vegard Nossum 已提交
1482
			SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK, NULL);
1483
	vm_area_cachep = KMEM_CACHE(vm_area_struct, SLAB_PANIC);
1484
	mmap_init();
L
Linus Torvalds 已提交
1485
}
1486 1487 1488 1489 1490

/*
 * Check constraints on flags passed to the unshare system call and
 * force unsharing of additional process context as appropriate.
 */
A
Alexey Dobriyan 已提交
1491
static void check_unshare_flags(unsigned long *flags_ptr)
1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532
{
	/*
	 * If unsharing a thread from a thread group, must also
	 * unshare vm.
	 */
	if (*flags_ptr & CLONE_THREAD)
		*flags_ptr |= CLONE_VM;

	/*
	 * If unsharing vm, must also unshare signal handlers.
	 */
	if (*flags_ptr & CLONE_VM)
		*flags_ptr |= CLONE_SIGHAND;

	/*
	 * If unsharing signal handlers and the task was created
	 * using CLONE_THREAD, then must unshare the thread
	 */
	if ((*flags_ptr & CLONE_SIGHAND) &&
	    (atomic_read(&current->signal->count) > 1))
		*flags_ptr |= CLONE_THREAD;

	/*
	 * If unsharing namespace, must also unshare filesystem information.
	 */
	if (*flags_ptr & CLONE_NEWNS)
		*flags_ptr |= CLONE_FS;
}

/*
 * Unsharing of tasks created with CLONE_THREAD is not supported yet
 */
static int unshare_thread(unsigned long unshare_flags)
{
	if (unshare_flags & CLONE_THREAD)
		return -EINVAL;

	return 0;
}

/*
1533
 * Unshare the filesystem structure if it is being shared
1534 1535 1536 1537 1538
 */
static int unshare_fs(unsigned long unshare_flags, struct fs_struct **new_fsp)
{
	struct fs_struct *fs = current->fs;

A
Al Viro 已提交
1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
	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;
1549 1550 1551 1552 1553

	return 0;
}

/*
1554
 * Unsharing of sighand is not supported yet
1555 1556 1557 1558 1559
 */
static int unshare_sighand(unsigned long unshare_flags, struct sighand_struct **new_sighp)
{
	struct sighand_struct *sigh = current->sighand;

1560
	if ((unshare_flags & CLONE_SIGHAND) && atomic_read(&sigh->count) > 1)
1561 1562 1563 1564 1565 1566
		return -EINVAL;
	else
		return 0;
}

/*
1567
 * Unshare vm if it is being shared
1568 1569 1570 1571 1572 1573
 */
static int unshare_vm(unsigned long unshare_flags, struct mm_struct **new_mmp)
{
	struct mm_struct *mm = current->mm;

	if ((unshare_flags & CLONE_VM) &&
1574
	    (mm && atomic_read(&mm->mm_users) > 1)) {
1575
		return -EINVAL;
1576
	}
1577 1578 1579 1580 1581

	return 0;
}

/*
1582
 * Unshare file descriptor table if it is being shared
1583 1584 1585 1586
 */
static int unshare_fd(unsigned long unshare_flags, struct files_struct **new_fdp)
{
	struct files_struct *fd = current->files;
1587
	int error = 0;
1588 1589

	if ((unshare_flags & CLONE_FILES) &&
1590 1591 1592 1593 1594
	    (fd && atomic_read(&fd->count) > 1)) {
		*new_fdp = dup_fd(fd, &error);
		if (!*new_fdp)
			return error;
	}
1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606

	return 0;
}

/*
 * 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.
 */
1607
SYSCALL_DEFINE1(unshare, unsigned long, unshare_flags)
1608 1609 1610
{
	int err = 0;
	struct fs_struct *fs, *new_fs = NULL;
1611
	struct sighand_struct *new_sigh = NULL;
1612 1613
	struct mm_struct *mm, *new_mm = NULL, *active_mm = NULL;
	struct files_struct *fd, *new_fd = NULL;
1614
	struct nsproxy *new_nsproxy = NULL;
1615
	int do_sysvsem = 0;
1616 1617 1618

	check_unshare_flags(&unshare_flags);

1619 1620 1621
	/* Return -EINVAL for all unsupported flags */
	err = -EINVAL;
	if (unshare_flags & ~(CLONE_THREAD|CLONE_FS|CLONE_NEWNS|CLONE_SIGHAND|
K
Kirill Korotaev 已提交
1622
				CLONE_VM|CLONE_FILES|CLONE_SYSVSEM|
1623
				CLONE_NEWUTS|CLONE_NEWIPC|CLONE_NEWNET))
1624 1625
		goto bad_unshare_out;

1626 1627 1628 1629 1630 1631
	/*
	 * 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))
1632
		do_sysvsem = 1;
1633 1634 1635 1636 1637
	if ((err = unshare_thread(unshare_flags)))
		goto bad_unshare_out;
	if ((err = unshare_fs(unshare_flags, &new_fs)))
		goto bad_unshare_cleanup_thread;
	if ((err = unshare_sighand(unshare_flags, &new_sigh)))
1638
		goto bad_unshare_cleanup_fs;
1639 1640 1641 1642
	if ((err = unshare_vm(unshare_flags, &new_mm)))
		goto bad_unshare_cleanup_sigh;
	if ((err = unshare_fd(unshare_flags, &new_fd)))
		goto bad_unshare_cleanup_vm;
1643 1644
	if ((err = unshare_nsproxy_namespaces(unshare_flags, &new_nsproxy,
			new_fs)))
1645
		goto bad_unshare_cleanup_fd;
1646

1647 1648 1649 1650 1651 1652 1653
	if (new_fs ||  new_mm || new_fd || do_sysvsem || new_nsproxy) {
		if (do_sysvsem) {
			/*
			 * CLONE_SYSVSEM is equivalent to sys_exit().
			 */
			exit_sem(current);
		}
S
Serge E. Hallyn 已提交
1654

1655
		if (new_nsproxy) {
1656 1657
			switch_task_namespaces(current, new_nsproxy);
			new_nsproxy = NULL;
1658
		}
1659

1660 1661
		task_lock(current);

1662 1663
		if (new_fs) {
			fs = current->fs;
A
Al Viro 已提交
1664
			write_lock(&fs->lock);
1665
			current->fs = new_fs;
A
Al Viro 已提交
1666 1667 1668 1669 1670
			if (--fs->users)
				new_fs = NULL;
			else
				new_fs = fs;
			write_unlock(&fs->lock);
1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690
		}

		if (new_mm) {
			mm = current->mm;
			active_mm = current->active_mm;
			current->mm = new_mm;
			current->active_mm = new_mm;
			activate_mm(active_mm, new_mm);
			new_mm = mm;
		}

		if (new_fd) {
			fd = current->files;
			current->files = new_fd;
			new_fd = fd;
		}

		task_unlock(current);
	}

1691
	if (new_nsproxy)
1692
		put_nsproxy(new_nsproxy);
1693

1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708
bad_unshare_cleanup_fd:
	if (new_fd)
		put_files_struct(new_fd);

bad_unshare_cleanup_vm:
	if (new_mm)
		mmput(new_mm);

bad_unshare_cleanup_sigh:
	if (new_sigh)
		if (atomic_dec_and_test(&new_sigh->count))
			kmem_cache_free(sighand_cachep, new_sigh);

bad_unshare_cleanup_fs:
	if (new_fs)
A
Al Viro 已提交
1709
		free_fs_struct(new_fs);
1710 1711 1712 1713 1714

bad_unshare_cleanup_thread:
bad_unshare_out:
	return err;
}
1715 1716 1717 1718 1719 1720 1721 1722 1723 1724

/*
 *	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;
1725
	struct files_struct *copy = NULL;
1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738
	int error;

	error = unshare_fd(CLONE_FILES, &copy);
	if (error || !copy) {
		*displaced = NULL;
		return error;
	}
	*displaced = task->files;
	task_lock(task);
	task->files = copy;
	task_unlock(task);
	return 0;
}