shm.c 27.4 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
/*
 * linux/ipc/shm.c
 * Copyright (C) 1992, 1993 Krishna Balasubramanian
 *	 Many improvements/fixes by Bruno Haible.
 * Replaced `struct shm_desc' by `struct vm_area_struct', July 1994.
 * Fixed the shm swap deallocation (shm_unuse()), August 1998 Andrea Arcangeli.
 *
 * /proc/sysvipc/shm support (c) 1999 Dragos Acostachioaie <dragos@iname.com>
 * BIGMEM support, Andrea Arcangeli <andrea@suse.de>
 * SMP thread shm, Jean-Luc Boyard <jean-luc.boyard@siemens.fr>
 * HIGHMEM support, Ingo Molnar <mingo@redhat.com>
 * Make shmmax, shmall, shmmni sysctl'able, Christoph Rohland <cr@sap.com>
 * Shared /dev/zero support, Kanoj Sarcar <kanoj@sgi.com>
 * Move the mm functionality over to mm/shmem.c, Christoph Rohland <cr@sap.com>
 *
S
Steve Grubb 已提交
16 17
 * support for audit of ipc object properties and permission changes
 * Dustin Kirkland <dustin.kirkland@us.ibm.com>
K
Kirill Korotaev 已提交
18 19 20 21
 *
 * namespaces support
 * OpenVZ, SWsoft Inc.
 * Pavel Emelianov <xemul@openvz.org>
L
Linus Torvalds 已提交
22 23 24 25 26 27 28 29 30 31 32 33 34
 */

#include <linux/slab.h>
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/shm.h>
#include <linux/init.h>
#include <linux/file.h>
#include <linux/mman.h>
#include <linux/shmem_fs.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/audit.h>
35
#include <linux/capability.h>
S
Stephen Rothwell 已提交
36
#include <linux/ptrace.h>
37
#include <linux/seq_file.h>
N
Nadia Derbey 已提交
38
#include <linux/rwsem.h>
K
Kirill Korotaev 已提交
39
#include <linux/nsproxy.h>
40
#include <linux/mount.h>
41
#include <linux/ipc_namespace.h>
S
Stephen Rothwell 已提交
42

L
Linus Torvalds 已提交
43 44 45 46
#include <asm/uaccess.h>

#include "util.h"

47 48 49 50 51 52 53 54 55
struct shm_file_data {
	int id;
	struct ipc_namespace *ns;
	struct file *file;
	const struct vm_operations_struct *vm_ops;
};

#define shm_file_data(file) (*((struct shm_file_data **)&(file)->private_data))

56
static const struct file_operations shm_file_operations;
L
Linus Torvalds 已提交
57 58
static struct vm_operations_struct shm_vm_ops;

59
#define shm_ids(ns)	((ns)->ids[IPC_SHM_IDS])
L
Linus Torvalds 已提交
60

K
Kirill Korotaev 已提交
61 62
#define shm_unlock(shp)			\
	ipc_unlock(&(shp)->shm_perm)
N
Nadia Derbey 已提交
63
#define shm_buildid(id, seq)	ipc_buildid(id, seq)
L
Linus Torvalds 已提交
64

N
Nadia Derbey 已提交
65
static int newseg(struct ipc_namespace *, struct ipc_params *);
66 67
static void shm_open(struct vm_area_struct *vma);
static void shm_close(struct vm_area_struct *vma);
K
Kirill Korotaev 已提交
68
static void shm_destroy (struct ipc_namespace *ns, struct shmid_kernel *shp);
L
Linus Torvalds 已提交
69
#ifdef CONFIG_PROC_FS
70
static int sysvipc_shm_proc_show(struct seq_file *s, void *it);
L
Linus Torvalds 已提交
71 72
#endif

73
void shm_init_ns(struct ipc_namespace *ns)
K
Kirill Korotaev 已提交
74 75 76 77 78
{
	ns->shm_ctlmax = SHMMAX;
	ns->shm_ctlall = SHMALL;
	ns->shm_ctlmni = SHMMNI;
	ns->shm_tot = 0;
79
	ipc_init_ids(&ns->ids[IPC_SHM_IDS]);
K
Kirill Korotaev 已提交
80 81
}

N
Nadia Derbey 已提交
82
/*
N
Nadia Derbey 已提交
83 84
 * Called with shm_ids.rw_mutex (writer) and the shp structure locked.
 * Only shm_ids.rw_mutex remains locked on exit.
N
Nadia Derbey 已提交
85
 */
K
Kirill Korotaev 已提交
86 87 88 89 90 91 92 93 94 95 96
static void do_shm_rmid(struct ipc_namespace *ns, struct shmid_kernel *shp)
{
	if (shp->shm_nattch){
		shp->shm_perm.mode |= SHM_DEST;
		/* Do not find it any more */
		shp->shm_perm.key = IPC_PRIVATE;
		shm_unlock(shp);
	} else
		shm_destroy(ns, shp);
}

97
#ifdef CONFIG_IPC_NS
K
Kirill Korotaev 已提交
98 99 100
void shm_exit_ns(struct ipc_namespace *ns)
{
	struct shmid_kernel *shp;
101
	struct kern_ipc_perm *perm;
N
Nadia Derbey 已提交
102 103
	int next_id;
	int total, in_use;
K
Kirill Korotaev 已提交
104

N
Nadia Derbey 已提交
105
	down_write(&shm_ids(ns).rw_mutex);
N
Nadia Derbey 已提交
106 107 108 109

	in_use = shm_ids(ns).in_use;

	for (total = 0, next_id = 0; total < in_use; next_id++) {
110 111
		perm = idr_find(&shm_ids(ns).ipcs_idr, next_id);
		if (perm == NULL)
K
Kirill Korotaev 已提交
112
			continue;
113 114
		ipc_lock_by_ptr(perm);
		shp = container_of(perm, struct shmid_kernel, shm_perm);
K
Kirill Korotaev 已提交
115
		do_shm_rmid(ns, shp);
N
Nadia Derbey 已提交
116
		total++;
K
Kirill Korotaev 已提交
117
	}
N
Nadia Derbey 已提交
118
	up_write(&shm_ids(ns).rw_mutex);
K
Kirill Korotaev 已提交
119
}
120
#endif
L
Linus Torvalds 已提交
121 122 123

void __init shm_init (void)
{
124
	shm_init_ns(&init_ipc_ns);
125 126
	ipc_init_proc_interface("sysvipc/shm",
				"       key      shmid perms       size  cpid  lpid nattch   uid   gid  cuid  cgid      atime      dtime      ctime\n",
K
Kirill Korotaev 已提交
127
				IPC_SHM_IDS, sysvipc_shm_proc_show);
L
Linus Torvalds 已提交
128 129
}

N
Nadia Derbey 已提交
130 131 132 133 134 135 136 137 138
/*
 * shm_lock_(check_)down routines are called in the paths where the rw_mutex
 * is held to protect access to the idr tree.
 */
static inline struct shmid_kernel *shm_lock_down(struct ipc_namespace *ns,
						int id)
{
	struct kern_ipc_perm *ipcp = ipc_lock_down(&shm_ids(ns), id);

139 140 141
	if (IS_ERR(ipcp))
		return (struct shmid_kernel *)ipcp;

N
Nadia Derbey 已提交
142 143 144 145 146 147 148 149 150
	return container_of(ipcp, struct shmid_kernel, shm_perm);
}

static inline struct shmid_kernel *shm_lock_check_down(
						struct ipc_namespace *ns,
						int id)
{
	struct kern_ipc_perm *ipcp = ipc_lock_check_down(&shm_ids(ns), id);

151 152 153
	if (IS_ERR(ipcp))
		return (struct shmid_kernel *)ipcp;

N
Nadia Derbey 已提交
154 155 156 157 158 159 160
	return container_of(ipcp, struct shmid_kernel, shm_perm);
}

/*
 * shm_lock_(check_) routines are called in the paths where the rw_mutex
 * is not held.
 */
161
static inline struct shmid_kernel *shm_lock(struct ipc_namespace *ns, int id)
L
Linus Torvalds 已提交
162
{
N
Nadia Derbey 已提交
163 164
	struct kern_ipc_perm *ipcp = ipc_lock(&shm_ids(ns), id);

165 166 167
	if (IS_ERR(ipcp))
		return (struct shmid_kernel *)ipcp;

N
Nadia Derbey 已提交
168
	return container_of(ipcp, struct shmid_kernel, shm_perm);
169 170 171 172 173
}

static inline struct shmid_kernel *shm_lock_check(struct ipc_namespace *ns,
						int id)
{
N
Nadia Derbey 已提交
174 175
	struct kern_ipc_perm *ipcp = ipc_lock_check(&shm_ids(ns), id);

176 177 178
	if (IS_ERR(ipcp))
		return (struct shmid_kernel *)ipcp;

N
Nadia Derbey 已提交
179
	return container_of(ipcp, struct shmid_kernel, shm_perm);
L
Linus Torvalds 已提交
180 181
}

N
Nadia Derbey 已提交
182
static inline void shm_rmid(struct ipc_namespace *ns, struct shmid_kernel *s)
L
Linus Torvalds 已提交
183
{
N
Nadia Derbey 已提交
184
	ipc_rmid(&shm_ids(ns), &s->shm_perm);
L
Linus Torvalds 已提交
185 186
}

K
Kirill Korotaev 已提交
187
static inline int shm_addid(struct ipc_namespace *ns, struct shmid_kernel *shp)
L
Linus Torvalds 已提交
188
{
K
Kirill Korotaev 已提交
189
	return ipc_addid(&shm_ids(ns), &shp->shm_perm, ns->shm_ctlmni);
L
Linus Torvalds 已提交
190 191 192 193
}



194 195
/* This is called by fork, once for every shm attach. */
static void shm_open(struct vm_area_struct *vma)
K
Kirill Korotaev 已提交
196
{
197 198
	struct file *file = vma->vm_file;
	struct shm_file_data *sfd = shm_file_data(file);
L
Linus Torvalds 已提交
199 200
	struct shmid_kernel *shp;

201
	shp = shm_lock(sfd->ns, sfd->id);
202
	BUG_ON(IS_ERR(shp));
L
Linus Torvalds 已提交
203
	shp->shm_atim = get_seconds();
204
	shp->shm_lprid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
205 206 207 208 209 210 211
	shp->shm_nattch++;
	shm_unlock(shp);
}

/*
 * shm_destroy - free the struct shmid_kernel
 *
N
Nadia Derbey 已提交
212
 * @ns: namespace
L
Linus Torvalds 已提交
213 214
 * @shp: struct to free
 *
N
Nadia Derbey 已提交
215
 * It has to be called with shp and shm_ids.rw_mutex (writer) locked,
L
Linus Torvalds 已提交
216 217
 * but returns with shp unlocked and freed.
 */
K
Kirill Korotaev 已提交
218
static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
L
Linus Torvalds 已提交
219
{
K
Kirill Korotaev 已提交
220
	ns->shm_tot -= (shp->shm_segsz + PAGE_SIZE - 1) >> PAGE_SHIFT;
N
Nadia Derbey 已提交
221
	shm_rmid(ns, shp);
L
Linus Torvalds 已提交
222 223 224 225
	shm_unlock(shp);
	if (!is_file_hugepages(shp->shm_file))
		shmem_lock(shp->shm_file, 0, shp->mlock_user);
	else
J
Josef Sipek 已提交
226
		user_shm_unlock(shp->shm_file->f_path.dentry->d_inode->i_size,
L
Linus Torvalds 已提交
227 228 229 230 231 232 233
						shp->mlock_user);
	fput (shp->shm_file);
	security_shm_free(shp);
	ipc_rcu_putref(shp);
}

/*
234
 * remove the attach descriptor vma.
L
Linus Torvalds 已提交
235 236 237 238
 * free memory for segment if it is marked destroyed.
 * The descriptor has already been removed from the current->mm->mmap list
 * and will later be kfree()d.
 */
239
static void shm_close(struct vm_area_struct *vma)
L
Linus Torvalds 已提交
240
{
241 242
	struct file * file = vma->vm_file;
	struct shm_file_data *sfd = shm_file_data(file);
L
Linus Torvalds 已提交
243
	struct shmid_kernel *shp;
244
	struct ipc_namespace *ns = sfd->ns;
K
Kirill Korotaev 已提交
245

N
Nadia Derbey 已提交
246
	down_write(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
247
	/* remove from the list of attaches of the shm segment */
N
Nadia Derbey 已提交
248
	shp = shm_lock_down(ns, sfd->id);
249
	BUG_ON(IS_ERR(shp));
250
	shp->shm_lprid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
251 252 253
	shp->shm_dtim = get_seconds();
	shp->shm_nattch--;
	if(shp->shm_nattch == 0 &&
A
Andrew Morton 已提交
254
	   shp->shm_perm.mode & SHM_DEST)
K
Kirill Korotaev 已提交
255
		shm_destroy(ns, shp);
L
Linus Torvalds 已提交
256 257
	else
		shm_unlock(shp);
N
Nadia Derbey 已提交
258
	up_write(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
259 260
}

N
Nick Piggin 已提交
261
static int shm_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
262 263 264 265
{
	struct file *file = vma->vm_file;
	struct shm_file_data *sfd = shm_file_data(file);

N
Nick Piggin 已提交
266
	return sfd->vm_ops->fault(vma, vmf);
267 268 269
}

#ifdef CONFIG_NUMA
A
Adrian Bunk 已提交
270
static int shm_set_policy(struct vm_area_struct *vma, struct mempolicy *new)
271 272 273 274 275 276 277 278 279
{
	struct file *file = vma->vm_file;
	struct shm_file_data *sfd = shm_file_data(file);
	int err = 0;
	if (sfd->vm_ops->set_policy)
		err = sfd->vm_ops->set_policy(vma, new);
	return err;
}

A
Adrian Bunk 已提交
280 281
static struct mempolicy *shm_get_policy(struct vm_area_struct *vma,
					unsigned long addr)
282 283 284 285 286 287 288
{
	struct file *file = vma->vm_file;
	struct shm_file_data *sfd = shm_file_data(file);
	struct mempolicy *pol = NULL;

	if (sfd->vm_ops->get_policy)
		pol = sfd->vm_ops->get_policy(vma, addr);
289
	else if (vma->vm_policy)
290
		pol = vma->vm_policy;
291 292
	else
		pol = current->mempolicy;
293 294 295 296
	return pol;
}
#endif

L
Linus Torvalds 已提交
297 298
static int shm_mmap(struct file * file, struct vm_area_struct * vma)
{
299
	struct shm_file_data *sfd = shm_file_data(file);
300 301
	int ret;

302 303 304 305
	ret = sfd->file->f_op->mmap(sfd->file, vma);
	if (ret != 0)
		return ret;
	sfd->vm_ops = vma->vm_ops;
D
David Howells 已提交
306
#ifdef CONFIG_MMU
307
	BUG_ON(!sfd->vm_ops->fault);
D
David Howells 已提交
308
#endif
309 310
	vma->vm_ops = &shm_vm_ops;
	shm_open(vma);
311 312

	return ret;
L
Linus Torvalds 已提交
313 314
}

K
Kirill Korotaev 已提交
315 316
static int shm_release(struct inode *ino, struct file *file)
{
317
	struct shm_file_data *sfd = shm_file_data(file);
K
Kirill Korotaev 已提交
318

319 320 321
	put_ipc_ns(sfd->ns);
	shm_file_data(file) = NULL;
	kfree(sfd);
K
Kirill Korotaev 已提交
322 323 324
	return 0;
}

325 326 327 328 329 330 331 332 333 334 335 336
static int shm_fsync(struct file *file, struct dentry *dentry, int datasync)
{
	int (*fsync) (struct file *, struct dentry *, int datasync);
	struct shm_file_data *sfd = shm_file_data(file);
	int ret = -EINVAL;

	fsync = sfd->file->f_op->fsync;
	if (fsync)
		ret = fsync(sfd->file, sfd->file->f_path.dentry, datasync);
	return ret;
}

337 338 339 340 341
static unsigned long shm_get_unmapped_area(struct file *file,
	unsigned long addr, unsigned long len, unsigned long pgoff,
	unsigned long flags)
{
	struct shm_file_data *sfd = shm_file_data(file);
342 343 344 345 346 347 348 349 350 351 352 353 354
	return get_unmapped_area(sfd->file, addr, len, pgoff, flags);
}

int is_file_shm_hugepages(struct file *file)
{
	int ret = 0;

	if (file->f_op == &shm_file_operations) {
		struct shm_file_data *sfd;
		sfd = shm_file_data(file);
		ret = is_file_hugepages(sfd->file);
	}
	return ret;
355 356
}

357
static const struct file_operations shm_file_operations = {
K
Kirill Korotaev 已提交
358
	.mmap		= shm_mmap,
359
	.fsync		= shm_fsync,
K
Kirill Korotaev 已提交
360
	.release	= shm_release,
361
	.get_unmapped_area	= shm_get_unmapped_area,
L
Linus Torvalds 已提交
362 363 364 365 366
};

static struct vm_operations_struct shm_vm_ops = {
	.open	= shm_open,	/* callback for a new vm-area open */
	.close	= shm_close,	/* callback for when the vm-area is released */
367
	.fault	= shm_fault,
368 369 370
#if defined(CONFIG_NUMA)
	.set_policy = shm_set_policy,
	.get_policy = shm_get_policy,
L
Linus Torvalds 已提交
371 372 373
#endif
};

N
Nadia Derbey 已提交
374 375 376 377 378
/**
 * newseg - Create a new shared memory segment
 * @ns: namespace
 * @params: ptr to the structure that contains key, size and shmflg
 *
N
Nadia Derbey 已提交
379
 * Called with shm_ids.rw_mutex held as a writer.
N
Nadia Derbey 已提交
380 381
 */

N
Nadia Derbey 已提交
382
static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
L
Linus Torvalds 已提交
383
{
N
Nadia Derbey 已提交
384 385 386
	key_t key = params->key;
	int shmflg = params->flg;
	size_t size = params->u.size;
L
Linus Torvalds 已提交
387 388 389 390 391 392 393
	int error;
	struct shmid_kernel *shp;
	int numpages = (size + PAGE_SIZE -1) >> PAGE_SHIFT;
	struct file * file;
	char name[13];
	int id;

K
Kirill Korotaev 已提交
394
	if (size < SHMMIN || size > ns->shm_ctlmax)
L
Linus Torvalds 已提交
395 396
		return -EINVAL;

397
	if (ns->shm_tot + numpages > ns->shm_ctlall)
L
Linus Torvalds 已提交
398 399 400 401 402 403 404
		return -ENOSPC;

	shp = ipc_rcu_alloc(sizeof(*shp));
	if (!shp)
		return -ENOMEM;

	shp->shm_perm.key = key;
A
Andrew Morton 已提交
405
	shp->shm_perm.mode = (shmflg & S_IRWXUGO);
L
Linus Torvalds 已提交
406 407 408 409 410 411 412 413 414
	shp->mlock_user = NULL;

	shp->shm_perm.security = NULL;
	error = security_shm_alloc(shp);
	if (error) {
		ipc_rcu_putref(shp);
		return error;
	}

415
	sprintf (name, "SYSV%08x", key);
L
Linus Torvalds 已提交
416
	if (shmflg & SHM_HUGETLB) {
417 418
		/* hugetlb_file_setup takes care of mlock user accounting */
		file = hugetlb_file_setup(name, size);
L
Linus Torvalds 已提交
419 420
		shp->mlock_user = current->user;
	} else {
421 422 423 424 425 426 427 428 429
		int acctflag = VM_ACCOUNT;
		/*
		 * Do not allow no accounting for OVERCOMMIT_NEVER, even
	 	 * if it's asked for.
		 */
		if  ((shmflg & SHM_NORESERVE) &&
				sysctl_overcommit_memory != OVERCOMMIT_NEVER)
			acctflag = 0;
		file = shmem_file_setup(name, size, acctflag);
L
Linus Torvalds 已提交
430 431 432 433 434
	}
	error = PTR_ERR(file);
	if (IS_ERR(file))
		goto no_file;

K
Kirill Korotaev 已提交
435
	id = shm_addid(ns, shp);
436 437
	if (id < 0) {
		error = id;
L
Linus Torvalds 已提交
438
		goto no_id;
439
	}
L
Linus Torvalds 已提交
440

441
	shp->shm_cprid = task_tgid_vnr(current);
L
Linus Torvalds 已提交
442 443 444 445 446
	shp->shm_lprid = 0;
	shp->shm_atim = shp->shm_dtim = 0;
	shp->shm_ctim = get_seconds();
	shp->shm_segsz = size;
	shp->shm_nattch = 0;
N
Nadia Derbey 已提交
447
	shp->shm_perm.id = shm_buildid(id, shp->shm_perm.seq);
L
Linus Torvalds 已提交
448
	shp->shm_file = file;
449 450 451 452
	/*
	 * shmid gets reported as "inode#" in /proc/pid/maps.
	 * proc-ps tools use this. Changing this will break them.
	 */
N
Nadia Derbey 已提交
453
	file->f_dentry->d_inode->i_ino = shp->shm_perm.id;
454

K
Kirill Korotaev 已提交
455
	ns->shm_tot += numpages;
N
Nadia Derbey 已提交
456
	error = shp->shm_perm.id;
L
Linus Torvalds 已提交
457
	shm_unlock(shp);
N
Nadia Derbey 已提交
458
	return error;
L
Linus Torvalds 已提交
459 460 461 462 463 464 465 466 467

no_id:
	fput(file);
no_file:
	security_shm_free(shp);
	ipc_rcu_putref(shp);
	return error;
}

N
Nadia Derbey 已提交
468
/*
N
Nadia Derbey 已提交
469
 * Called with shm_ids.rw_mutex and ipcp locked.
N
Nadia Derbey 已提交
470
 */
N
Nadia Derbey 已提交
471
static inline int shm_security(struct kern_ipc_perm *ipcp, int shmflg)
N
Nadia Derbey 已提交
472
{
N
Nadia Derbey 已提交
473 474 475 476
	struct shmid_kernel *shp;

	shp = container_of(ipcp, struct shmid_kernel, shm_perm);
	return security_shm_associate(shp, shmflg);
N
Nadia Derbey 已提交
477 478
}

N
Nadia Derbey 已提交
479
/*
N
Nadia Derbey 已提交
480
 * Called with shm_ids.rw_mutex and ipcp locked.
N
Nadia Derbey 已提交
481
 */
N
Nadia Derbey 已提交
482 483
static inline int shm_more_checks(struct kern_ipc_perm *ipcp,
				struct ipc_params *params)
N
Nadia Derbey 已提交
484
{
N
Nadia Derbey 已提交
485 486 487 488
	struct shmid_kernel *shp;

	shp = container_of(ipcp, struct shmid_kernel, shm_perm);
	if (shp->shm_segsz < params->u.size)
N
Nadia Derbey 已提交
489 490 491 492 493
		return -EINVAL;

	return 0;
}

L
Linus Torvalds 已提交
494 495
asmlinkage long sys_shmget (key_t key, size_t size, int shmflg)
{
K
Kirill Korotaev 已提交
496
	struct ipc_namespace *ns;
N
Nadia Derbey 已提交
497 498
	struct ipc_ops shm_ops;
	struct ipc_params shm_params;
K
Kirill Korotaev 已提交
499 500

	ns = current->nsproxy->ipc_ns;
L
Linus Torvalds 已提交
501

N
Nadia Derbey 已提交
502 503 504
	shm_ops.getnew = newseg;
	shm_ops.associate = shm_security;
	shm_ops.more_checks = shm_more_checks;
N
Nadia Derbey 已提交
505

N
Nadia Derbey 已提交
506 507 508
	shm_params.key = key;
	shm_params.flg = shmflg;
	shm_params.u.size = size;
L
Linus Torvalds 已提交
509

N
Nadia Derbey 已提交
510
	return ipcget(ns, &shm_ids(ns), &shm_ops, &shm_params);
L
Linus Torvalds 已提交
511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555
}

static inline unsigned long copy_shmid_to_user(void __user *buf, struct shmid64_ds *in, int version)
{
	switch(version) {
	case IPC_64:
		return copy_to_user(buf, in, sizeof(*in));
	case IPC_OLD:
	    {
		struct shmid_ds out;

		ipc64_perm_to_ipc_perm(&in->shm_perm, &out.shm_perm);
		out.shm_segsz	= in->shm_segsz;
		out.shm_atime	= in->shm_atime;
		out.shm_dtime	= in->shm_dtime;
		out.shm_ctime	= in->shm_ctime;
		out.shm_cpid	= in->shm_cpid;
		out.shm_lpid	= in->shm_lpid;
		out.shm_nattch	= in->shm_nattch;

		return copy_to_user(buf, &out, sizeof(out));
	    }
	default:
		return -EINVAL;
	}
}

struct shm_setbuf {
	uid_t	uid;
	gid_t	gid;
	mode_t	mode;
};	

static inline unsigned long copy_shmid_from_user(struct shm_setbuf *out, void __user *buf, int version)
{
	switch(version) {
	case IPC_64:
	    {
		struct shmid64_ds tbuf;

		if (copy_from_user(&tbuf, buf, sizeof(tbuf)))
			return -EFAULT;

		out->uid	= tbuf.shm_perm.uid;
		out->gid	= tbuf.shm_perm.gid;
A
Andrew Morton 已提交
556
		out->mode	= tbuf.shm_perm.mode;
L
Linus Torvalds 已提交
557 558 559 560 561 562 563 564 565 566 567 568

		return 0;
	    }
	case IPC_OLD:
	    {
		struct shmid_ds tbuf_old;

		if (copy_from_user(&tbuf_old, buf, sizeof(tbuf_old)))
			return -EFAULT;

		out->uid	= tbuf_old.shm_perm.uid;
		out->gid	= tbuf_old.shm_perm.gid;
A
Andrew Morton 已提交
569
		out->mode	= tbuf_old.shm_perm.mode;
L
Linus Torvalds 已提交
570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603

		return 0;
	    }
	default:
		return -EINVAL;
	}
}

static inline unsigned long copy_shminfo_to_user(void __user *buf, struct shminfo64 *in, int version)
{
	switch(version) {
	case IPC_64:
		return copy_to_user(buf, in, sizeof(*in));
	case IPC_OLD:
	    {
		struct shminfo out;

		if(in->shmmax > INT_MAX)
			out.shmmax = INT_MAX;
		else
			out.shmmax = (int)in->shmmax;

		out.shmmin	= in->shmmin;
		out.shmmni	= in->shmmni;
		out.shmseg	= in->shmseg;
		out.shmall	= in->shmall; 

		return copy_to_user(buf, &out, sizeof(out));
	    }
	default:
		return -EINVAL;
	}
}

N
Nadia Derbey 已提交
604
/*
N
Nadia Derbey 已提交
605
 * Called with shm_ids.rw_mutex held as a reader
N
Nadia Derbey 已提交
606
 */
K
Kirill Korotaev 已提交
607 608
static void shm_get_stat(struct ipc_namespace *ns, unsigned long *rss,
		unsigned long *swp)
L
Linus Torvalds 已提交
609
{
N
Nadia Derbey 已提交
610 611
	int next_id;
	int total, in_use;
L
Linus Torvalds 已提交
612 613 614 615

	*rss = 0;
	*swp = 0;

N
Nadia Derbey 已提交
616 617 618
	in_use = shm_ids(ns).in_use;

	for (total = 0, next_id = 0; total < in_use; next_id++) {
L
Linus Torvalds 已提交
619 620 621
		struct shmid_kernel *shp;
		struct inode *inode;

N
Nadia Derbey 已提交
622
		shp = idr_find(&shm_ids(ns).ipcs_idr, next_id);
N
Nadia Derbey 已提交
623
		if (shp == NULL)
L
Linus Torvalds 已提交
624 625
			continue;

J
Josef Sipek 已提交
626
		inode = shp->shm_file->f_path.dentry->d_inode;
L
Linus Torvalds 已提交
627 628 629 630 631 632 633 634 635 636 637

		if (is_file_hugepages(shp->shm_file)) {
			struct address_space *mapping = inode->i_mapping;
			*rss += (HPAGE_SIZE/PAGE_SIZE)*mapping->nrpages;
		} else {
			struct shmem_inode_info *info = SHMEM_I(inode);
			spin_lock(&info->lock);
			*rss += inode->i_mapping->nrpages;
			*swp += info->swapped;
			spin_unlock(&info->lock);
		}
N
Nadia Derbey 已提交
638 639

		total++;
L
Linus Torvalds 已提交
640 641 642 643 644 645 646 647
	}
}

asmlinkage long sys_shmctl (int shmid, int cmd, struct shmid_ds __user *buf)
{
	struct shm_setbuf setbuf;
	struct shmid_kernel *shp;
	int err, version;
K
Kirill Korotaev 已提交
648
	struct ipc_namespace *ns;
L
Linus Torvalds 已提交
649 650 651 652 653 654 655

	if (cmd < 0 || shmid < 0) {
		err = -EINVAL;
		goto out;
	}

	version = ipc_parse_version(&cmd);
K
Kirill Korotaev 已提交
656
	ns = current->nsproxy->ipc_ns;
L
Linus Torvalds 已提交
657 658 659 660 661 662 663 664 665 666 667

	switch (cmd) { /* replace with proc interface ? */
	case IPC_INFO:
	{
		struct shminfo64 shminfo;

		err = security_shm_shmctl(NULL, cmd);
		if (err)
			return err;

		memset(&shminfo,0,sizeof(shminfo));
K
Kirill Korotaev 已提交
668 669 670
		shminfo.shmmni = shminfo.shmseg = ns->shm_ctlmni;
		shminfo.shmmax = ns->shm_ctlmax;
		shminfo.shmall = ns->shm_ctlall;
L
Linus Torvalds 已提交
671 672 673 674

		shminfo.shmmin = SHMMIN;
		if(copy_shminfo_to_user (buf, &shminfo, version))
			return -EFAULT;
N
Nadia Derbey 已提交
675

N
Nadia Derbey 已提交
676
		down_read(&shm_ids(ns).rw_mutex);
N
Nadia Derbey 已提交
677
		err = ipc_get_maxid(&shm_ids(ns));
N
Nadia Derbey 已提交
678
		up_read(&shm_ids(ns).rw_mutex);
N
Nadia Derbey 已提交
679

L
Linus Torvalds 已提交
680 681 682 683 684 685 686 687 688 689 690 691 692
		if(err<0)
			err = 0;
		goto out;
	}
	case SHM_INFO:
	{
		struct shm_info shm_info;

		err = security_shm_shmctl(NULL, cmd);
		if (err)
			return err;

		memset(&shm_info,0,sizeof(shm_info));
N
Nadia Derbey 已提交
693
		down_read(&shm_ids(ns).rw_mutex);
K
Kirill Korotaev 已提交
694 695 696
		shm_info.used_ids = shm_ids(ns).in_use;
		shm_get_stat (ns, &shm_info.shm_rss, &shm_info.shm_swp);
		shm_info.shm_tot = ns->shm_tot;
L
Linus Torvalds 已提交
697 698
		shm_info.swap_attempts = 0;
		shm_info.swap_successes = 0;
N
Nadia Derbey 已提交
699
		err = ipc_get_maxid(&shm_ids(ns));
N
Nadia Derbey 已提交
700
		up_read(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
701 702 703 704 705 706 707 708 709 710 711 712 713
		if(copy_to_user (buf, &shm_info, sizeof(shm_info))) {
			err = -EFAULT;
			goto out;
		}

		err = err < 0 ? 0 : err;
		goto out;
	}
	case SHM_STAT:
	case IPC_STAT:
	{
		struct shmid64_ds tbuf;
		int result;
714 715 716

		if (!buf) {
			err = -EFAULT;
L
Linus Torvalds 已提交
717
			goto out;
718 719 720 721 722 723 724 725
		}

		if (cmd == SHM_STAT) {
			shp = shm_lock(ns, shmid);
			if (IS_ERR(shp)) {
				err = PTR_ERR(shp);
				goto out;
			}
N
Nadia Derbey 已提交
726
			result = shp->shm_perm.id;
L
Linus Torvalds 已提交
727
		} else {
728 729 730 731 732
			shp = shm_lock_check(ns, shmid);
			if (IS_ERR(shp)) {
				err = PTR_ERR(shp);
				goto out;
			}
L
Linus Torvalds 已提交
733 734 735 736 737 738 739 740
			result = 0;
		}
		err=-EACCES;
		if (ipcperms (&shp->shm_perm, S_IRUGO))
			goto out_unlock;
		err = security_shm_shmctl(shp, cmd);
		if (err)
			goto out_unlock;
741
		memset(&tbuf, 0, sizeof(tbuf));
L
Linus Torvalds 已提交
742 743 744 745 746 747 748
		kernel_to_ipc64_perm(&shp->shm_perm, &tbuf.shm_perm);
		tbuf.shm_segsz	= shp->shm_segsz;
		tbuf.shm_atime	= shp->shm_atim;
		tbuf.shm_dtime	= shp->shm_dtim;
		tbuf.shm_ctime	= shp->shm_ctim;
		tbuf.shm_cpid	= shp->shm_cprid;
		tbuf.shm_lpid	= shp->shm_lprid;
749
		tbuf.shm_nattch	= shp->shm_nattch;
L
Linus Torvalds 已提交
750 751 752 753 754 755 756 757 758 759
		shm_unlock(shp);
		if(copy_shmid_to_user (buf, &tbuf, version))
			err = -EFAULT;
		else
			err = result;
		goto out;
	}
	case SHM_LOCK:
	case SHM_UNLOCK:
	{
760 761 762
		shp = shm_lock_check(ns, shmid);
		if (IS_ERR(shp)) {
			err = PTR_ERR(shp);
L
Linus Torvalds 已提交
763 764 765
			goto out;
		}

S
Steve Grubb 已提交
766 767 768 769
		err = audit_ipc_obj(&(shp->shm_perm));
		if (err)
			goto out_unlock;

L
Linus Torvalds 已提交
770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
		if (!capable(CAP_IPC_LOCK)) {
			err = -EPERM;
			if (current->euid != shp->shm_perm.uid &&
			    current->euid != shp->shm_perm.cuid)
				goto out_unlock;
			if (cmd == SHM_LOCK &&
			    !current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur)
				goto out_unlock;
		}

		err = security_shm_shmctl(shp, cmd);
		if (err)
			goto out_unlock;
		
		if(cmd==SHM_LOCK) {
			struct user_struct * user = current->user;
			if (!is_file_hugepages(shp->shm_file)) {
				err = shmem_lock(shp->shm_file, 1, user);
788
				if (!err && !(shp->shm_perm.mode & SHM_LOCKED)){
A
Andrew Morton 已提交
789
					shp->shm_perm.mode |= SHM_LOCKED;
L
Linus Torvalds 已提交
790 791 792 793 794
					shp->mlock_user = user;
				}
			}
		} else if (!is_file_hugepages(shp->shm_file)) {
			shmem_lock(shp->shm_file, 0, shp->mlock_user);
A
Andrew Morton 已提交
795
			shp->shm_perm.mode &= ~SHM_LOCKED;
L
Linus Torvalds 已提交
796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812
			shp->mlock_user = NULL;
		}
		shm_unlock(shp);
		goto out;
	}
	case IPC_RMID:
	{
		/*
		 *	We cannot simply remove the file. The SVID states
		 *	that the block remains until the last person
		 *	detaches from it, then is deleted. A shmat() on
		 *	an RMID segment is legal in older Linux and if 
		 *	we change it apps break...
		 *
		 *	Instead we set a destroyed flag, and then blow
		 *	the name away when the usage hits zero.
		 */
N
Nadia Derbey 已提交
813 814
		down_write(&shm_ids(ns).rw_mutex);
		shp = shm_lock_check_down(ns, shmid);
815 816
		if (IS_ERR(shp)) {
			err = PTR_ERR(shp);
L
Linus Torvalds 已提交
817
			goto out_up;
818
		}
L
Linus Torvalds 已提交
819

S
Steve Grubb 已提交
820 821 822 823
		err = audit_ipc_obj(&(shp->shm_perm));
		if (err)
			goto out_unlock_up;

L
Linus Torvalds 已提交
824 825 826 827 828 829 830 831 832 833 834
		if (current->euid != shp->shm_perm.uid &&
		    current->euid != shp->shm_perm.cuid && 
		    !capable(CAP_SYS_ADMIN)) {
			err=-EPERM;
			goto out_unlock_up;
		}

		err = security_shm_shmctl(shp, cmd);
		if (err)
			goto out_unlock_up;

K
Kirill Korotaev 已提交
835
		do_shm_rmid(ns, shp);
N
Nadia Derbey 已提交
836
		up_write(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
837 838 839 840 841
		goto out;
	}

	case IPC_SET:
	{
842 843 844 845 846
		if (!buf) {
			err = -EFAULT;
			goto out;
		}

L
Linus Torvalds 已提交
847 848 849 850
		if (copy_shmid_from_user (&setbuf, buf, version)) {
			err = -EFAULT;
			goto out;
		}
N
Nadia Derbey 已提交
851 852
		down_write(&shm_ids(ns).rw_mutex);
		shp = shm_lock_check_down(ns, shmid);
853 854
		if (IS_ERR(shp)) {
			err = PTR_ERR(shp);
L
Linus Torvalds 已提交
855
			goto out_up;
856
		}
S
Steve Grubb 已提交
857 858 859
		err = audit_ipc_obj(&(shp->shm_perm));
		if (err)
			goto out_unlock_up;
860
		err = audit_ipc_set_perm(0, setbuf.uid, setbuf.gid, setbuf.mode);
S
Steve Grubb 已提交
861 862
		if (err)
			goto out_unlock_up;
L
Linus Torvalds 已提交
863 864 865 866 867 868 869 870 871 872 873 874 875
		err=-EPERM;
		if (current->euid != shp->shm_perm.uid &&
		    current->euid != shp->shm_perm.cuid && 
		    !capable(CAP_SYS_ADMIN)) {
			goto out_unlock_up;
		}

		err = security_shm_shmctl(shp, cmd);
		if (err)
			goto out_unlock_up;
		
		shp->shm_perm.uid = setbuf.uid;
		shp->shm_perm.gid = setbuf.gid;
A
Andrew Morton 已提交
876
		shp->shm_perm.mode = (shp->shm_perm.mode & ~S_IRWXUGO)
L
Linus Torvalds 已提交
877 878 879 880 881 882 883 884 885 886 887 888 889 890
			| (setbuf.mode & S_IRWXUGO);
		shp->shm_ctim = get_seconds();
		break;
	}

	default:
		err = -EINVAL;
		goto out;
	}

	err = 0;
out_unlock_up:
	shm_unlock(shp);
out_up:
N
Nadia Derbey 已提交
891
	up_write(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915
	goto out;
out_unlock:
	shm_unlock(shp);
out:
	return err;
}

/*
 * Fix shmaddr, allocate descriptor, map shm, add attach descriptor to lists.
 *
 * NOTE! Despite the name, this is NOT a direct system call entrypoint. The
 * "raddr" thing points to kernel space, and there has to be a wrapper around
 * this.
 */
long do_shmat(int shmid, char __user *shmaddr, int shmflg, ulong *raddr)
{
	struct shmid_kernel *shp;
	unsigned long addr;
	unsigned long size;
	struct file * file;
	int    err;
	unsigned long flags;
	unsigned long prot;
	int acc_mode;
916
	unsigned long user_addr;
K
Kirill Korotaev 已提交
917
	struct ipc_namespace *ns;
918 919 920
	struct shm_file_data *sfd;
	struct path path;
	mode_t f_mode;
L
Linus Torvalds 已提交
921

922 923
	err = -EINVAL;
	if (shmid < 0)
L
Linus Torvalds 已提交
924
		goto out;
925
	else if ((addr = (ulong)shmaddr)) {
L
Linus Torvalds 已提交
926 927 928 929 930 931 932
		if (addr & (SHMLBA-1)) {
			if (shmflg & SHM_RND)
				addr &= ~(SHMLBA-1);	   /* round down */
			else
#ifndef __ARCH_FORCE_SHMLBA
				if (addr & ~PAGE_MASK)
#endif
933
					goto out;
L
Linus Torvalds 已提交
934 935 936 937
		}
		flags = MAP_SHARED | MAP_FIXED;
	} else {
		if ((shmflg & SHM_REMAP))
938
			goto out;
L
Linus Torvalds 已提交
939 940 941 942 943 944 945

		flags = MAP_SHARED;
	}

	if (shmflg & SHM_RDONLY) {
		prot = PROT_READ;
		acc_mode = S_IRUGO;
946
		f_mode = FMODE_READ;
L
Linus Torvalds 已提交
947 948 949
	} else {
		prot = PROT_READ | PROT_WRITE;
		acc_mode = S_IRUGO | S_IWUGO;
950
		f_mode = FMODE_READ | FMODE_WRITE;
L
Linus Torvalds 已提交
951 952 953 954 955 956 957 958 959 960
	}
	if (shmflg & SHM_EXEC) {
		prot |= PROT_EXEC;
		acc_mode |= S_IXUGO;
	}

	/*
	 * We cannot rely on the fs check since SYSV IPC does have an
	 * additional creator id...
	 */
K
Kirill Korotaev 已提交
961
	ns = current->nsproxy->ipc_ns;
962 963 964
	shp = shm_lock_check(ns, shmid);
	if (IS_ERR(shp)) {
		err = PTR_ERR(shp);
L
Linus Torvalds 已提交
965
		goto out;
966
	}
967 968 969 970

	err = -EACCES;
	if (ipcperms(&shp->shm_perm, acc_mode))
		goto out_unlock;
L
Linus Torvalds 已提交
971 972

	err = security_shm_shmat(shp, shmaddr, shmflg);
973 974 975 976
	if (err)
		goto out_unlock;

	path.dentry = dget(shp->shm_file->f_path.dentry);
977
	path.mnt    = shp->shm_file->f_path.mnt;
L
Linus Torvalds 已提交
978
	shp->shm_nattch++;
979
	size = i_size_read(path.dentry->d_inode);
L
Linus Torvalds 已提交
980 981
	shm_unlock(shp);

982 983 984
	err = -ENOMEM;
	sfd = kzalloc(sizeof(*sfd), GFP_KERNEL);
	if (!sfd)
985
		goto out_put_dentry;
986 987

	err = -ENOMEM;
988 989

	file = alloc_file(path.mnt, path.dentry, f_mode, &shm_file_operations);
990 991 992 993 994
	if (!file)
		goto out_free;

	file->private_data = sfd;
	file->f_mapping = shp->shm_file->f_mapping;
N
Nadia Derbey 已提交
995
	sfd->id = shp->shm_perm.id;
996 997 998 999
	sfd->ns = get_ipc_ns(ns);
	sfd->file = shp->shm_file;
	sfd->vm_ops = NULL;

L
Linus Torvalds 已提交
1000 1001
	down_write(&current->mm->mmap_sem);
	if (addr && !(shmflg & SHM_REMAP)) {
1002
		err = -EINVAL;
L
Linus Torvalds 已提交
1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013
		if (find_vma_intersection(current->mm, addr, addr + size))
			goto invalid;
		/*
		 * If shm segment goes below stack, make sure there is some
		 * space left for the stack to grow (at least 4 pages).
		 */
		if (addr < current->mm->start_stack &&
		    addr > current->mm->start_stack - size - PAGE_SIZE * 5)
			goto invalid;
	}
		
1014 1015 1016 1017 1018
	user_addr = do_mmap (file, addr, size, prot, flags, 0);
	*raddr = user_addr;
	err = 0;
	if (IS_ERR_VALUE(user_addr))
		err = (long)user_addr;
L
Linus Torvalds 已提交
1019 1020 1021
invalid:
	up_write(&current->mm->mmap_sem);

1022 1023 1024
	fput(file);

out_nattch:
N
Nadia Derbey 已提交
1025 1026
	down_write(&shm_ids(ns).rw_mutex);
	shp = shm_lock_down(ns, shmid);
1027
	BUG_ON(IS_ERR(shp));
L
Linus Torvalds 已提交
1028 1029
	shp->shm_nattch--;
	if(shp->shm_nattch == 0 &&
A
Andrew Morton 已提交
1030
	   shp->shm_perm.mode & SHM_DEST)
K
Kirill Korotaev 已提交
1031
		shm_destroy(ns, shp);
L
Linus Torvalds 已提交
1032 1033
	else
		shm_unlock(shp);
N
Nadia Derbey 已提交
1034
	up_write(&shm_ids(ns).rw_mutex);
L
Linus Torvalds 已提交
1035 1036 1037

out:
	return err;
1038 1039 1040 1041 1042 1043 1044

out_unlock:
	shm_unlock(shp);
	goto out;

out_free:
	kfree(sfd);
1045
out_put_dentry:
1046 1047
	dput(path.dentry);
	goto out_nattch;
L
Linus Torvalds 已提交
1048 1049
}

S
Stephen Rothwell 已提交
1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061
asmlinkage long sys_shmat(int shmid, char __user *shmaddr, int shmflg)
{
	unsigned long ret;
	long err;

	err = do_shmat(shmid, shmaddr, shmflg, &ret);
	if (err)
		return err;
	force_successful_syscall_return();
	return (long)ret;
}

L
Linus Torvalds 已提交
1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073
/*
 * detach and kill segment if marked destroyed.
 * The work is done in shm_close.
 */
asmlinkage long sys_shmdt(char __user *shmaddr)
{
	struct mm_struct *mm = current->mm;
	struct vm_area_struct *vma, *next;
	unsigned long addr = (unsigned long)shmaddr;
	loff_t size = 0;
	int retval = -EINVAL;

1074 1075 1076
	if (addr & ~PAGE_MASK)
		return retval;

L
Linus Torvalds 已提交
1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
	down_write(&mm->mmap_sem);

	/*
	 * This function tries to be smart and unmap shm segments that
	 * were modified by partial mlock or munmap calls:
	 * - It first determines the size of the shm segment that should be
	 *   unmapped: It searches for a vma that is backed by shm and that
	 *   started at address shmaddr. It records it's size and then unmaps
	 *   it.
	 * - Then it unmaps all shm vmas that started at shmaddr and that
	 *   are within the initially determined size.
	 * Errors from do_munmap are ignored: the function only fails if
	 * it's called with invalid parameters or if it's called to unmap
	 * a part of a vma. Both calls in this function are for full vmas,
	 * the parameters are directly copied from the vma itself and always
	 * valid - therefore do_munmap cannot fail. (famous last words?)
	 */
	/*
	 * If it had been mremap()'d, the starting address would not
	 * match the usual checks anyway. So assume all vma's are
	 * above the starting address given.
	 */
	vma = find_vma(mm, addr);

	while (vma) {
		next = vma->vm_next;

		/*
		 * Check if the starting address would match, i.e. it's
		 * a fragment created by mprotect() and/or munmap(), or it
		 * otherwise it starts at this address with no hassles.
		 */
1109
		if ((vma->vm_ops == &shm_vm_ops) &&
L
Linus Torvalds 已提交
1110 1111 1112
			(vma->vm_start - addr)/PAGE_SIZE == vma->vm_pgoff) {


J
Josef Sipek 已提交
1113
			size = vma->vm_file->f_path.dentry->d_inode->i_size;
L
Linus Torvalds 已提交
1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132
			do_munmap(mm, vma->vm_start, vma->vm_end - vma->vm_start);
			/*
			 * We discovered the size of the shm segment, so
			 * break out of here and fall through to the next
			 * loop that uses the size information to stop
			 * searching for matching vma's.
			 */
			retval = 0;
			vma = next;
			break;
		}
		vma = next;
	}

	/*
	 * We need look no further than the maximum address a fragment
	 * could possibly have landed at. Also cast things to loff_t to
	 * prevent overflows and make comparisions vs. equal-width types.
	 */
1133
	size = PAGE_ALIGN(size);
L
Linus Torvalds 已提交
1134 1135 1136 1137
	while (vma && (loff_t)(vma->vm_end - addr) <= size) {
		next = vma->vm_next;

		/* finding a matching vma now does not alter retval */
1138
		if ((vma->vm_ops == &shm_vm_ops) &&
L
Linus Torvalds 已提交
1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149
			(vma->vm_start - addr)/PAGE_SIZE == vma->vm_pgoff)

			do_munmap(mm, vma->vm_start, vma->vm_end - vma->vm_start);
		vma = next;
	}

	up_write(&mm->mmap_sem);
	return retval;
}

#ifdef CONFIG_PROC_FS
1150
static int sysvipc_shm_proc_show(struct seq_file *s, void *it)
L
Linus Torvalds 已提交
1151
{
1152 1153
	struct shmid_kernel *shp = it;
	char *format;
L
Linus Torvalds 已提交
1154 1155 1156 1157

#define SMALL_STRING "%10d %10d  %4o %10u %5u %5u  %5d %5u %5u %5u %5u %10lu %10lu %10lu\n"
#define BIG_STRING   "%10d %10d  %4o %21u %5u %5u  %5d %5u %5u %5u %5u %10lu %10lu %10lu\n"

1158 1159 1160 1161 1162 1163
	if (sizeof(size_t) <= sizeof(int))
		format = SMALL_STRING;
	else
		format = BIG_STRING;
	return seq_printf(s, format,
			  shp->shm_perm.key,
N
Nadia Derbey 已提交
1164
			  shp->shm_perm.id,
A
Andrew Morton 已提交
1165
			  shp->shm_perm.mode,
1166 1167 1168
			  shp->shm_segsz,
			  shp->shm_cprid,
			  shp->shm_lprid,
1169
			  shp->shm_nattch,
1170 1171 1172 1173 1174 1175 1176
			  shp->shm_perm.uid,
			  shp->shm_perm.gid,
			  shp->shm_perm.cuid,
			  shp->shm_perm.cgid,
			  shp->shm_atim,
			  shp->shm_dtim,
			  shp->shm_ctim);
L
Linus Torvalds 已提交
1177 1178
}
#endif