socket.c 74.5 KB
Newer Older
L
Linus Torvalds 已提交
1 2 3 4 5 6
/*
 * NET		An implementation of the SOCKET network access protocol.
 *
 * Version:	@(#)socket.c	1.1.93	18/02/95
 *
 * Authors:	Orest Zborowski, <obz@Kodak.COM>
7
 *		Ross Biro
L
Linus Torvalds 已提交
8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44
 *		Fred N. van Kempen, <waltje@uWalt.NL.Mugnet.ORG>
 *
 * Fixes:
 *		Anonymous	:	NOTSOCK/BADF cleanup. Error fix in
 *					shutdown()
 *		Alan Cox	:	verify_area() fixes
 *		Alan Cox	:	Removed DDI
 *		Jonathan Kamens	:	SOCK_DGRAM reconnect bug
 *		Alan Cox	:	Moved a load of checks to the very
 *					top level.
 *		Alan Cox	:	Move address structures to/from user
 *					mode above the protocol layers.
 *		Rob Janssen	:	Allow 0 length sends.
 *		Alan Cox	:	Asynchronous I/O support (cribbed from the
 *					tty drivers).
 *		Niibe Yutaka	:	Asynchronous I/O for writes (4.4BSD style)
 *		Jeff Uphoff	:	Made max number of sockets command-line
 *					configurable.
 *		Matti Aarnio	:	Made the number of sockets dynamic,
 *					to be allocated when needed, and mr.
 *					Uphoff's max is used as max to be
 *					allowed to allocate.
 *		Linus		:	Argh. removed all the socket allocation
 *					altogether: it's in the inode now.
 *		Alan Cox	:	Made sock_alloc()/sock_release() public
 *					for NetROM and future kernel nfsd type
 *					stuff.
 *		Alan Cox	:	sendmsg/recvmsg basics.
 *		Tom Dyas	:	Export net symbols.
 *		Marcin Dalecki	:	Fixed problems with CONFIG_NET="n".
 *		Alan Cox	:	Added thread locking to sys_* calls
 *					for sockets. May have errors at the
 *					moment.
 *		Kevin Buhr	:	Fixed the dumb errors in the above.
 *		Andi Kleen	:	Some small cleanups, optimizations,
 *					and fixed a copy_from_user() bug.
 *		Tigran Aivazian	:	sys_send(args) calls sys_sendto(args, NULL, 0)
45
 *		Tigran Aivazian	:	Made listen(2) backlog sanity checks
L
Linus Torvalds 已提交
46 47 48 49 50 51 52 53 54 55
 *					protocol-independent
 *
 *
 *		This program is free software; you can redistribute it and/or
 *		modify it under the terms of the GNU General Public License
 *		as published by the Free Software Foundation; either version
 *		2 of the License, or (at your option) any later version.
 *
 *
 *	This module is effectively the top level interface to the BSD socket
56
 *	paradigm.
L
Linus Torvalds 已提交
57 58 59 60 61 62 63 64 65
 *
 *	Based upon Swansea University Computer Society NET3.039
 */

#include <linux/mm.h>
#include <linux/socket.h>
#include <linux/file.h>
#include <linux/net.h>
#include <linux/interrupt.h>
U
Ulrich Drepper 已提交
66
#include <linux/thread_info.h>
67
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
68 69 70
#include <linux/netdevice.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
A
Arjan van de Ven 已提交
71
#include <linux/mutex.h>
L
Linus Torvalds 已提交
72 73
#include <linux/wanrouter.h>
#include <linux/if_bridge.h>
74 75
#include <linux/if_frad.h>
#include <linux/if_vlan.h>
L
Linus Torvalds 已提交
76 77 78 79 80 81 82 83 84 85
#include <linux/init.h>
#include <linux/poll.h>
#include <linux/cache.h>
#include <linux/module.h>
#include <linux/highmem.h>
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/syscalls.h>
#include <linux/compat.h>
#include <linux/kmod.h>
86
#include <linux/audit.h>
87
#include <linux/wireless.h>
88
#include <linux/nsproxy.h>
N
Nick Black 已提交
89
#include <linux/magic.h>
90
#include <linux/slab.h>
L
Linus Torvalds 已提交
91 92 93 94 95

#include <asm/uaccess.h>
#include <asm/unistd.h>

#include <net/compat.h>
96
#include <net/wext.h>
97
#include <net/cls_cgroup.h>
L
Linus Torvalds 已提交
98 99 100 101

#include <net/sock.h>
#include <linux/netfilter.h>

102 103 104 105 106 107
#include <linux/if_tun.h>
#include <linux/ipv6_route.h>
#include <linux/route.h>
#include <linux/sockios.h>
#include <linux/atalk.h>

L
Linus Torvalds 已提交
108
static int sock_no_open(struct inode *irrelevant, struct file *dontcare);
109 110 111 112
static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
			 unsigned long nr_segs, loff_t pos);
static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov,
			  unsigned long nr_segs, loff_t pos);
113
static int sock_mmap(struct file *file, struct vm_area_struct *vma);
L
Linus Torvalds 已提交
114 115 116 117

static int sock_close(struct inode *inode, struct file *file);
static unsigned int sock_poll(struct file *file,
			      struct poll_table_struct *wait);
118
static long sock_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
119 120
#ifdef CONFIG_COMPAT
static long compat_sock_ioctl(struct file *file,
121
			      unsigned int cmd, unsigned long arg);
122
#endif
L
Linus Torvalds 已提交
123 124 125
static int sock_fasync(int fd, struct file *filp, int on);
static ssize_t sock_sendpage(struct file *file, struct page *page,
			     int offset, size_t size, loff_t *ppos, int more);
J
Jens Axboe 已提交
126
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
127
				struct pipe_inode_info *pipe, size_t len,
J
Jens Axboe 已提交
128
				unsigned int flags);
L
Linus Torvalds 已提交
129 130 131 132 133 134

/*
 *	Socket files have a set of 'special' operations as well as the generic file ones. These don't appear
 *	in the operation structures but are done directly via the socketcall() multiplexor.
 */

135
static const struct file_operations socket_file_ops = {
L
Linus Torvalds 已提交
136 137 138 139 140 141
	.owner =	THIS_MODULE,
	.llseek =	no_llseek,
	.aio_read =	sock_aio_read,
	.aio_write =	sock_aio_write,
	.poll =		sock_poll,
	.unlocked_ioctl = sock_ioctl,
142 143 144
#ifdef CONFIG_COMPAT
	.compat_ioctl = compat_sock_ioctl,
#endif
L
Linus Torvalds 已提交
145 146 147 148
	.mmap =		sock_mmap,
	.open =		sock_no_open,	/* special open code to disallow open via /proc */
	.release =	sock_close,
	.fasync =	sock_fasync,
149 150
	.sendpage =	sock_sendpage,
	.splice_write = generic_splice_sendpage,
J
Jens Axboe 已提交
151
	.splice_read =	sock_splice_read,
L
Linus Torvalds 已提交
152 153 154 155 156 157 158
};

/*
 *	The protocol list. Each protocol is registered in here.
 */

static DEFINE_SPINLOCK(net_family_lock);
159
static const struct net_proto_family *net_families[NPROTO] __read_mostly;
L
Linus Torvalds 已提交
160 161 162 163 164

/*
 *	Statistics counters of the socket lists
 */

165
static DEFINE_PER_CPU(int, sockets_in_use);
L
Linus Torvalds 已提交
166 167

/*
168 169 170
 * Support routines.
 * Move socket addresses back and forth across the kernel/user
 * divide and look after the messy bits.
L
Linus Torvalds 已提交
171 172 173 174 175 176 177 178 179 180 181 182 183
 */

/**
 *	move_addr_to_kernel	-	copy a socket address into kernel space
 *	@uaddr: Address in user space
 *	@kaddr: Address in kernel space
 *	@ulen: Length in user space
 *
 *	The address is copied into kernel space. If the provided address is
 *	too long an error code of -EINVAL is returned. If the copy gives
 *	invalid addresses -EFAULT is returned. On a success 0 is returned.
 */

184
int move_addr_to_kernel(void __user *uaddr, int ulen, struct sockaddr *kaddr)
L
Linus Torvalds 已提交
185
{
186
	if (ulen < 0 || ulen > sizeof(struct sockaddr_storage))
L
Linus Torvalds 已提交
187
		return -EINVAL;
188
	if (ulen == 0)
L
Linus Torvalds 已提交
189
		return 0;
190
	if (copy_from_user(kaddr, uaddr, ulen))
L
Linus Torvalds 已提交
191
		return -EFAULT;
192
	return audit_sockaddr(ulen, kaddr);
L
Linus Torvalds 已提交
193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210
}

/**
 *	move_addr_to_user	-	copy an address to user space
 *	@kaddr: kernel space address
 *	@klen: length of address in kernel
 *	@uaddr: user space address
 *	@ulen: pointer to user length field
 *
 *	The value pointed to by ulen on entry is the buffer length available.
 *	This is overwritten with the buffer space used. -EINVAL is returned
 *	if an overlong buffer is specified or a negative buffer size. -EFAULT
 *	is returned if either the buffer or the length field are not
 *	accessible.
 *	After copying the data up to the limit the user specifies, the true
 *	length of the data is written over the length limit the user
 *	specified. Zero is returned for a success.
 */
211

S
stephen hemminger 已提交
212 213
static int move_addr_to_user(struct sockaddr *kaddr, int klen,
			     void __user *uaddr, int __user *ulen)
L
Linus Torvalds 已提交
214 215 216 217
{
	int err;
	int len;

218 219
	err = get_user(len, ulen);
	if (err)
L
Linus Torvalds 已提交
220
		return err;
221 222
	if (len > klen)
		len = klen;
223
	if (len < 0 || len > sizeof(struct sockaddr_storage))
L
Linus Torvalds 已提交
224
		return -EINVAL;
225
	if (len) {
S
Steve Grubb 已提交
226 227
		if (audit_sockaddr(klen, kaddr))
			return -ENOMEM;
228
		if (copy_to_user(uaddr, kaddr, len))
L
Linus Torvalds 已提交
229 230 231
			return -EFAULT;
	}
	/*
232 233
	 *      "fromlen shall refer to the value before truncation.."
	 *                      1003.1g
L
Linus Torvalds 已提交
234 235 236 237
	 */
	return __put_user(klen, ulen);
}

238
static struct kmem_cache *sock_inode_cachep __read_mostly;
L
Linus Torvalds 已提交
239 240 241 242

static struct inode *sock_alloc_inode(struct super_block *sb)
{
	struct socket_alloc *ei;
243

244
	ei = kmem_cache_alloc(sock_inode_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
245 246
	if (!ei)
		return NULL;
247 248 249 250 251 252 253
	ei->socket.wq = kmalloc(sizeof(struct socket_wq), GFP_KERNEL);
	if (!ei->socket.wq) {
		kmem_cache_free(sock_inode_cachep, ei);
		return NULL;
	}
	init_waitqueue_head(&ei->socket.wq->wait);
	ei->socket.wq->fasync_list = NULL;
254

L
Linus Torvalds 已提交
255 256 257 258 259 260 261 262 263
	ei->socket.state = SS_UNCONNECTED;
	ei->socket.flags = 0;
	ei->socket.ops = NULL;
	ei->socket.sk = NULL;
	ei->socket.file = NULL;

	return &ei->vfs_inode;
}

264 265 266 267 268 269 270 271

static void wq_free_rcu(struct rcu_head *head)
{
	struct socket_wq *wq = container_of(head, struct socket_wq, rcu);

	kfree(wq);
}

L
Linus Torvalds 已提交
272 273
static void sock_destroy_inode(struct inode *inode)
{
274 275 276 277 278
	struct socket_alloc *ei;

	ei = container_of(inode, struct socket_alloc, vfs_inode);
	call_rcu(&ei->socket.wq->rcu, wq_free_rcu);
	kmem_cache_free(sock_inode_cachep, ei);
L
Linus Torvalds 已提交
279 280
}

281
static void init_once(void *foo)
L
Linus Torvalds 已提交
282
{
283
	struct socket_alloc *ei = (struct socket_alloc *)foo;
L
Linus Torvalds 已提交
284

C
Christoph Lameter 已提交
285
	inode_init_once(&ei->vfs_inode);
L
Linus Torvalds 已提交
286
}
287

L
Linus Torvalds 已提交
288 289 290
static int init_inodecache(void)
{
	sock_inode_cachep = kmem_cache_create("sock_inode_cache",
291 292 293 294 295
					      sizeof(struct socket_alloc),
					      0,
					      (SLAB_HWCACHE_ALIGN |
					       SLAB_RECLAIM_ACCOUNT |
					       SLAB_MEM_SPREAD),
296
					      init_once);
L
Linus Torvalds 已提交
297 298 299 300 301
	if (sock_inode_cachep == NULL)
		return -ENOMEM;
	return 0;
}

302
static const struct super_operations sockfs_ops = {
303 304 305
	.alloc_inode	= sock_alloc_inode,
	.destroy_inode	= sock_destroy_inode,
	.statfs		= simple_statfs,
L
Linus Torvalds 已提交
306 307
};

308
static int sockfs_get_sb(struct file_system_type *fs_type,
309 310
			 int flags, const char *dev_name, void *data,
			 struct vfsmount *mnt)
L
Linus Torvalds 已提交
311
{
312 313
	return get_sb_pseudo(fs_type, "socket:", &sockfs_ops, SOCKFS_MAGIC,
			     mnt);
L
Linus Torvalds 已提交
314 315
}

316
static struct vfsmount *sock_mnt __read_mostly;
L
Linus Torvalds 已提交
317 318 319 320 321 322

static struct file_system_type sock_fs_type = {
	.name =		"sockfs",
	.get_sb =	sockfs_get_sb,
	.kill_sb =	kill_anon_super,
};
323

324 325 326 327 328 329 330 331 332
/*
 * sockfs_dname() is called from d_path().
 */
static char *sockfs_dname(struct dentry *dentry, char *buffer, int buflen)
{
	return dynamic_dname(dentry, buffer, buflen, "socket:[%lu]",
				dentry->d_inode->i_ino);
}

A
Al Viro 已提交
333
static const struct dentry_operations sockfs_dentry_operations = {
334
	.d_dname  = sockfs_dname,
L
Linus Torvalds 已提交
335 336 337 338 339
};

/*
 *	Obtains the first available file descriptor and sets it up for use.
 *
340 341
 *	These functions create file structures and maps them to fd space
 *	of the current process. On success it returns file descriptor
L
Linus Torvalds 已提交
342 343 344 345 346 347 348 349 350 351 352 353
 *	and file struct implicitly stored in sock->file.
 *	Note that another thread may close file descriptor before we return
 *	from this function. We use the fact that now we do not refer
 *	to socket after mapping. If one day we will need it, this
 *	function will increment ref. count on file by 1.
 *
 *	In any case returned fd MAY BE not valid!
 *	This race condition is unavoidable
 *	with shared fd spaces, we cannot solve it inside kernel,
 *	but we take care of internal coherence yet.
 */

354
static int sock_alloc_file(struct socket *sock, struct file **f, int flags)
L
Linus Torvalds 已提交
355
{
356
	struct qstr name = { .name = "" };
357
	struct path path;
358
	struct file *file;
L
Linus Torvalds 已提交
359 360
	int fd;

361
	fd = get_unused_fd_flags(flags);
362 363
	if (unlikely(fd < 0))
		return fd;
L
Linus Torvalds 已提交
364

365 366
	path.dentry = d_alloc(sock_mnt->mnt_sb->s_root, &name);
	if (unlikely(!path.dentry)) {
367
		put_unused_fd(fd);
368
		return -ENOMEM;
369
	}
370
	path.mnt = mntget(sock_mnt);
371

372 373
	path.dentry->d_op = &sockfs_dentry_operations;
	d_instantiate(path.dentry, SOCK_INODE(sock));
A
Al Viro 已提交
374
	SOCK_INODE(sock)->i_fop = &socket_file_ops;
375

376
	file = alloc_file(&path, FMODE_READ | FMODE_WRITE,
377
		  &socket_file_ops);
A
Al Viro 已提交
378 379 380
	if (unlikely(!file)) {
		/* drop dentry, keep inode */
		atomic_inc(&path.dentry->d_inode->i_count);
381
		path_put(&path);
A
Al Viro 已提交
382 383 384 385 386
		put_unused_fd(fd);
		return -ENFILE;
	}

	sock->file = file;
387
	file->f_flags = O_RDWR | (flags & O_NONBLOCK);
388 389
	file->f_pos = 0;
	file->private_data = sock;
L
Linus Torvalds 已提交
390

391 392
	*f = file;
	return fd;
393 394
}

395
int sock_map_fd(struct socket *sock, int flags)
396 397
{
	struct file *newfile;
398
	int fd = sock_alloc_file(sock, &newfile, flags);
399

400
	if (likely(fd >= 0))
401
		fd_install(fd, newfile);
402

L
Linus Torvalds 已提交
403 404
	return fd;
}
405
EXPORT_SYMBOL(sock_map_fd);
L
Linus Torvalds 已提交
406

407 408 409 410 411
static struct socket *sock_from_file(struct file *file, int *err)
{
	if (file->f_op == &socket_file_ops)
		return file->private_data;	/* set in sock_map_fd */

E
Eric Dumazet 已提交
412 413
	*err = -ENOTSOCK;
	return NULL;
414 415
}

L
Linus Torvalds 已提交
416
/**
417
 *	sockfd_lookup - Go from a file number to its socket slot
L
Linus Torvalds 已提交
418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433
 *	@fd: file handle
 *	@err: pointer to an error code return
 *
 *	The file handle passed in is locked and the socket it is bound
 *	too is returned. If an error occurs the err pointer is overwritten
 *	with a negative errno code and NULL is returned. The function checks
 *	for both invalid handles and passing a handle which is not a socket.
 *
 *	On a success the socket object pointer is returned.
 */

struct socket *sockfd_lookup(int fd, int *err)
{
	struct file *file;
	struct socket *sock;

434 435
	file = fget(fd);
	if (!file) {
L
Linus Torvalds 已提交
436 437 438
		*err = -EBADF;
		return NULL;
	}
439

440 441
	sock = sock_from_file(file, err);
	if (!sock)
L
Linus Torvalds 已提交
442
		fput(file);
443 444
	return sock;
}
445
EXPORT_SYMBOL(sockfd_lookup);
L
Linus Torvalds 已提交
446

447 448 449 450 451
static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed)
{
	struct file *file;
	struct socket *sock;

452
	*err = -EBADF;
453 454 455 456 457 458
	file = fget_light(fd, fput_needed);
	if (file) {
		sock = sock_from_file(file, err);
		if (sock)
			return sock;
		fput_light(file, *fput_needed);
L
Linus Torvalds 已提交
459
	}
460
	return NULL;
L
Linus Torvalds 已提交
461 462 463 464
}

/**
 *	sock_alloc	-	allocate a socket
465
 *
L
Linus Torvalds 已提交
466 467 468 469 470 471 472
 *	Allocate a new inode and socket object. The two are bound together
 *	and initialised. The socket is then returned. If we are out of inodes
 *	NULL is returned.
 */

static struct socket *sock_alloc(void)
{
473 474
	struct inode *inode;
	struct socket *sock;
L
Linus Torvalds 已提交
475 476 477 478 479 480 481

	inode = new_inode(sock_mnt->mnt_sb);
	if (!inode)
		return NULL;

	sock = SOCKET_I(inode);

482
	kmemcheck_annotate_bitfield(sock, type);
483
	inode->i_mode = S_IFSOCK | S_IRWXUGO;
484 485
	inode->i_uid = current_fsuid();
	inode->i_gid = current_fsgid();
L
Linus Torvalds 已提交
486

487
	percpu_add(sockets_in_use, 1);
L
Linus Torvalds 已提交
488 489 490 491 492 493 494 495
	return sock;
}

/*
 *	In theory you can't get an open on this inode, but /proc provides
 *	a back door. Remember to keep it shut otherwise you'll let the
 *	creepy crawlies in.
 */
496

L
Linus Torvalds 已提交
497 498 499 500 501
static int sock_no_open(struct inode *irrelevant, struct file *dontcare)
{
	return -ENXIO;
}

502
const struct file_operations bad_sock_fops = {
L
Linus Torvalds 已提交
503 504
	.owner = THIS_MODULE,
	.open = sock_no_open,
505
	.llseek = noop_llseek,
L
Linus Torvalds 已提交
506 507 508 509 510 511 512 513
};

/**
 *	sock_release	-	close a socket
 *	@sock: socket to close
 *
 *	The socket is released from the protocol stack if it has a release
 *	callback, and the inode is then released if the socket is bound to
514
 *	an inode not a file.
L
Linus Torvalds 已提交
515
 */
516

L
Linus Torvalds 已提交
517 518 519 520 521 522 523 524 525 526
void sock_release(struct socket *sock)
{
	if (sock->ops) {
		struct module *owner = sock->ops->owner;

		sock->ops->release(sock);
		sock->ops = NULL;
		module_put(owner);
	}

527
	if (sock->wq->fasync_list)
L
Linus Torvalds 已提交
528 529
		printk(KERN_ERR "sock_release: fasync list not empty!\n");

530
	percpu_sub(sockets_in_use, 1);
L
Linus Torvalds 已提交
531 532 533 534
	if (!sock->file) {
		iput(SOCK_INODE(sock));
		return;
	}
535
	sock->file = NULL;
L
Linus Torvalds 已提交
536
}
537
EXPORT_SYMBOL(sock_release);
L
Linus Torvalds 已提交
538

539
int sock_tx_timestamp(struct sock *sk, __u8 *tx_flags)
540
{
541
	*tx_flags = 0;
542
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_HARDWARE))
543
		*tx_flags |= SKBTX_HW_TSTAMP;
544
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_SOFTWARE))
545
		*tx_flags |= SKBTX_SW_TSTAMP;
546 547 548 549
	return 0;
}
EXPORT_SYMBOL(sock_tx_timestamp);

550
static inline int __sock_sendmsg(struct kiocb *iocb, struct socket *sock,
L
Linus Torvalds 已提交
551 552 553 554 555
				 struct msghdr *msg, size_t size)
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);
	int err;

556 557
	sock_update_classid(sock->sk);

L
Linus Torvalds 已提交
558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582
	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;

	err = security_socket_sendmsg(sock, msg, size);
	if (err)
		return err;

	return sock->ops->sendmsg(iocb, sock, msg, size);
}

int sock_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

	init_sync_kiocb(&iocb, NULL);
	iocb.private = &siocb;
	ret = __sock_sendmsg(&iocb, sock, msg, size);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}
583
EXPORT_SYMBOL(sock_sendmsg);
L
Linus Torvalds 已提交
584 585 586 587 588 589 590 591 592 593 594 595

int kernel_sendmsg(struct socket *sock, struct msghdr *msg,
		   struct kvec *vec, size_t num, size_t size)
{
	mm_segment_t oldfs = get_fs();
	int result;

	set_fs(KERNEL_DS);
	/*
	 * the following is safe, since for compiler definitions of kvec and
	 * iovec are identical, yielding the same in-core layout and alignment
	 */
596
	msg->msg_iov = (struct iovec *)vec;
L
Linus Torvalds 已提交
597 598 599 600 601
	msg->msg_iovlen = num;
	result = sock_sendmsg(sock, msg, size);
	set_fs(oldfs);
	return result;
}
602
EXPORT_SYMBOL(kernel_sendmsg);
L
Linus Torvalds 已提交
603

604 605 606 607 608 609 610 611 612 613
static int ktime2ts(ktime_t kt, struct timespec *ts)
{
	if (kt.tv64) {
		*ts = ktime_to_timespec(kt);
		return 1;
	} else {
		return 0;
	}
}

614 615 616 617 618 619
/*
 * called from sock_recv_timestamp() if sock_flag(sk, SOCK_RCVTSTAMP)
 */
void __sock_recv_timestamp(struct msghdr *msg, struct sock *sk,
	struct sk_buff *skb)
{
620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637
	int need_software_tstamp = sock_flag(sk, SOCK_RCVTSTAMP);
	struct timespec ts[3];
	int empty = 1;
	struct skb_shared_hwtstamps *shhwtstamps =
		skb_hwtstamps(skb);

	/* Race occurred between timestamp enabling and packet
	   receiving.  Fill in the current time for now. */
	if (need_software_tstamp && skb->tstamp.tv64 == 0)
		__net_timestamp(skb);

	if (need_software_tstamp) {
		if (!sock_flag(sk, SOCK_RCVTSTAMPNS)) {
			struct timeval tv;
			skb_get_timestamp(skb, &tv);
			put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMP,
				 sizeof(tv), &tv);
		} else {
638
			skb_get_timestampns(skb, &ts[0]);
639
			put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPNS,
640
				 sizeof(ts[0]), &ts[0]);
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657
		}
	}


	memset(ts, 0, sizeof(ts));
	if (skb->tstamp.tv64 &&
	    sock_flag(sk, SOCK_TIMESTAMPING_SOFTWARE)) {
		skb_get_timestampns(skb, ts + 0);
		empty = 0;
	}
	if (shhwtstamps) {
		if (sock_flag(sk, SOCK_TIMESTAMPING_SYS_HARDWARE) &&
		    ktime2ts(shhwtstamps->syststamp, ts + 1))
			empty = 0;
		if (sock_flag(sk, SOCK_TIMESTAMPING_RAW_HARDWARE) &&
		    ktime2ts(shhwtstamps->hwtstamp, ts + 2))
			empty = 0;
658
	}
659 660 661
	if (!empty)
		put_cmsg(msg, SOL_SOCKET,
			 SCM_TIMESTAMPING, sizeof(ts), &ts);
662
}
663 664
EXPORT_SYMBOL_GPL(__sock_recv_timestamp);

S
stephen hemminger 已提交
665 666
static inline void sock_recv_drops(struct msghdr *msg, struct sock *sk,
				   struct sk_buff *skb)
667 668 669 670 671 672
{
	if (sock_flag(sk, SOCK_RXQ_OVFL) && skb && skb->dropcount)
		put_cmsg(msg, SOL_SOCKET, SO_RXQ_OVFL,
			sizeof(__u32), &skb->dropcount);
}

673
void __sock_recv_ts_and_drops(struct msghdr *msg, struct sock *sk,
674 675 676 677 678
	struct sk_buff *skb)
{
	sock_recv_timestamp(msg, sk, skb);
	sock_recv_drops(msg, sk, skb);
}
679
EXPORT_SYMBOL_GPL(__sock_recv_ts_and_drops);
680

681 682
static inline int __sock_recvmsg_nosec(struct kiocb *iocb, struct socket *sock,
				       struct msghdr *msg, size_t size, int flags)
L
Linus Torvalds 已提交
683 684 685
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);

686 687
	sock_update_classid(sock->sk);

L
Linus Torvalds 已提交
688 689 690 691 692 693 694 695 696
	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;
	si->flags = flags;

	return sock->ops->recvmsg(iocb, sock, msg, size, flags);
}

697 698 699 700 701 702 703 704
static inline int __sock_recvmsg(struct kiocb *iocb, struct socket *sock,
				 struct msghdr *msg, size_t size, int flags)
{
	int err = security_socket_recvmsg(sock, msg, size, flags);

	return err ?: __sock_recvmsg_nosec(iocb, sock, msg, size, flags);
}

705
int sock_recvmsg(struct socket *sock, struct msghdr *msg,
L
Linus Torvalds 已提交
706 707 708 709 710 711
		 size_t size, int flags)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

712
	init_sync_kiocb(&iocb, NULL);
L
Linus Torvalds 已提交
713 714 715 716 717 718
	iocb.private = &siocb;
	ret = __sock_recvmsg(&iocb, sock, msg, size, flags);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}
719
EXPORT_SYMBOL(sock_recvmsg);
L
Linus Torvalds 已提交
720

721 722 723 724 725 726 727 728 729 730 731 732 733 734 735
static int sock_recvmsg_nosec(struct socket *sock, struct msghdr *msg,
			      size_t size, int flags)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

	init_sync_kiocb(&iocb, NULL);
	iocb.private = &siocb;
	ret = __sock_recvmsg_nosec(&iocb, sock, msg, size, flags);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}

736 737
int kernel_recvmsg(struct socket *sock, struct msghdr *msg,
		   struct kvec *vec, size_t num, size_t size, int flags)
L
Linus Torvalds 已提交
738 739 740 741 742 743 744 745 746
{
	mm_segment_t oldfs = get_fs();
	int result;

	set_fs(KERNEL_DS);
	/*
	 * the following is safe, since for compiler definitions of kvec and
	 * iovec are identical, yielding the same in-core layout and alignment
	 */
747
	msg->msg_iov = (struct iovec *)vec, msg->msg_iovlen = num;
L
Linus Torvalds 已提交
748 749 750 751
	result = sock_recvmsg(sock, msg, size, flags);
	set_fs(oldfs);
	return result;
}
752
EXPORT_SYMBOL(kernel_recvmsg);
L
Linus Torvalds 已提交
753 754 755 756 757 758

static void sock_aio_dtor(struct kiocb *iocb)
{
	kfree(iocb->private);
}

759 760
static ssize_t sock_sendpage(struct file *file, struct page *page,
			     int offset, size_t size, loff_t *ppos, int more)
L
Linus Torvalds 已提交
761 762 763 764
{
	struct socket *sock;
	int flags;

765 766 767 768 769 770
	sock = file->private_data;

	flags = !(file->f_flags & O_NONBLOCK) ? 0 : MSG_DONTWAIT;
	if (more)
		flags |= MSG_MORE;

771
	return kernel_sendpage(sock, page, offset, size, flags);
772
}
L
Linus Torvalds 已提交
773

J
Jens Axboe 已提交
774
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
775
				struct pipe_inode_info *pipe, size_t len,
J
Jens Axboe 已提交
776 777 778 779
				unsigned int flags)
{
	struct socket *sock = file->private_data;

780 781 782
	if (unlikely(!sock->ops->splice_read))
		return -EINVAL;

783 784
	sock_update_classid(sock->sk);

J
Jens Axboe 已提交
785 786 787
	return sock->ops->splice_read(sock, ppos, pipe, len, flags);
}

788
static struct sock_iocb *alloc_sock_iocb(struct kiocb *iocb,
789
					 struct sock_iocb *siocb)
790 791 792 793 794
{
	if (!is_sync_kiocb(iocb)) {
		siocb = kmalloc(sizeof(*siocb), GFP_KERNEL);
		if (!siocb)
			return NULL;
L
Linus Torvalds 已提交
795 796 797
		iocb->ki_dtor = sock_aio_dtor;
	}

798 799 800
	siocb->kiocb = iocb;
	iocb->private = siocb;
	return siocb;
L
Linus Torvalds 已提交
801 802
}

803
static ssize_t do_sock_read(struct msghdr *msg, struct kiocb *iocb,
804 805
		struct file *file, const struct iovec *iov,
		unsigned long nr_segs)
806 807 808 809
{
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
810

811 812
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
813

814 815 816 817
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
818
	msg->msg_iov = (struct iovec *)iov;
819 820 821 822 823 824
	msg->msg_iovlen = nr_segs;
	msg->msg_flags = (file->f_flags & O_NONBLOCK) ? MSG_DONTWAIT : 0;

	return __sock_recvmsg(iocb, sock, msg, size, msg->msg_flags);
}

825 826
static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
				unsigned long nr_segs, loff_t pos)
827 828 829
{
	struct sock_iocb siocb, *x;

L
Linus Torvalds 已提交
830 831
	if (pos != 0)
		return -ESPIPE;
832 833

	if (iocb->ki_left == 0)	/* Match SYS5 behaviour */
L
Linus Torvalds 已提交
834 835
		return 0;

836 837

	x = alloc_sock_iocb(iocb, &siocb);
838 839
	if (!x)
		return -ENOMEM;
840
	return do_sock_read(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
841 842
}

843
static ssize_t do_sock_write(struct msghdr *msg, struct kiocb *iocb,
844 845
			struct file *file, const struct iovec *iov,
			unsigned long nr_segs)
L
Linus Torvalds 已提交
846
{
847 848 849
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
850

851 852
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
853

854 855 856 857
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
858
	msg->msg_iov = (struct iovec *)iov;
859 860 861 862
	msg->msg_iovlen = nr_segs;
	msg->msg_flags = (file->f_flags & O_NONBLOCK) ? MSG_DONTWAIT : 0;
	if (sock->type == SOCK_SEQPACKET)
		msg->msg_flags |= MSG_EOR;
L
Linus Torvalds 已提交
863

864
	return __sock_sendmsg(iocb, sock, msg, size);
L
Linus Torvalds 已提交
865 866
}

867 868
static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov,
			  unsigned long nr_segs, loff_t pos)
869 870
{
	struct sock_iocb siocb, *x;
L
Linus Torvalds 已提交
871

872 873
	if (pos != 0)
		return -ESPIPE;
874 875

	x = alloc_sock_iocb(iocb, &siocb);
876 877
	if (!x)
		return -ENOMEM;
L
Linus Torvalds 已提交
878

879
	return do_sock_write(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
880 881 882 883 884 885 886
}

/*
 * Atomic setting of ioctl hooks to avoid race
 * with module unload.
 */

A
Arjan van de Ven 已提交
887
static DEFINE_MUTEX(br_ioctl_mutex);
888
static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg);
L
Linus Torvalds 已提交
889

890
void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *))
L
Linus Torvalds 已提交
891
{
A
Arjan van de Ven 已提交
892
	mutex_lock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
893
	br_ioctl_hook = hook;
A
Arjan van de Ven 已提交
894
	mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
895 896 897
}
EXPORT_SYMBOL(brioctl_set);

A
Arjan van de Ven 已提交
898
static DEFINE_MUTEX(vlan_ioctl_mutex);
899
static int (*vlan_ioctl_hook) (struct net *, void __user *arg);
L
Linus Torvalds 已提交
900

901
void vlan_ioctl_set(int (*hook) (struct net *, void __user *))
L
Linus Torvalds 已提交
902
{
A
Arjan van de Ven 已提交
903
	mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
904
	vlan_ioctl_hook = hook;
A
Arjan van de Ven 已提交
905
	mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
906 907 908
}
EXPORT_SYMBOL(vlan_ioctl_set);

A
Arjan van de Ven 已提交
909
static DEFINE_MUTEX(dlci_ioctl_mutex);
910
static int (*dlci_ioctl_hook) (unsigned int, void __user *);
L
Linus Torvalds 已提交
911

912
void dlci_ioctl_set(int (*hook) (unsigned int, void __user *))
L
Linus Torvalds 已提交
913
{
A
Arjan van de Ven 已提交
914
	mutex_lock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
915
	dlci_ioctl_hook = hook;
A
Arjan van de Ven 已提交
916
	mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
917 918 919
}
EXPORT_SYMBOL(dlci_ioctl_set);

920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937
static long sock_do_ioctl(struct net *net, struct socket *sock,
				 unsigned int cmd, unsigned long arg)
{
	int err;
	void __user *argp = (void __user *)arg;

	err = sock->ops->ioctl(sock, cmd, arg);

	/*
	 * If this ioctl is unknown try to hand it down
	 * to the NIC driver.
	 */
	if (err == -ENOIOCTLCMD)
		err = dev_ioctl(net, cmd, argp);

	return err;
}

L
Linus Torvalds 已提交
938 939 940 941 942 943 944 945
/*
 *	With an ioctl, arg may well be a user mode pointer, but we don't know
 *	what to do with it - that's up to the protocol still.
 */

static long sock_ioctl(struct file *file, unsigned cmd, unsigned long arg)
{
	struct socket *sock;
946
	struct sock *sk;
L
Linus Torvalds 已提交
947 948
	void __user *argp = (void __user *)arg;
	int pid, err;
949
	struct net *net;
L
Linus Torvalds 已提交
950

951
	sock = file->private_data;
952
	sk = sock->sk;
953
	net = sock_net(sk);
L
Linus Torvalds 已提交
954
	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
955
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
956
	} else
J
Johannes Berg 已提交
957
#ifdef CONFIG_WEXT_CORE
L
Linus Torvalds 已提交
958
	if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST) {
959
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
960
	} else
J
Johannes Berg 已提交
961
#endif
962
		switch (cmd) {
L
Linus Torvalds 已提交
963 964 965 966 967 968 969 970 971
		case FIOSETOWN:
		case SIOCSPGRP:
			err = -EFAULT;
			if (get_user(pid, (int __user *)argp))
				break;
			err = f_setown(sock->file, pid, 1);
			break;
		case FIOGETOWN:
		case SIOCGPGRP:
972
			err = put_user(f_getown(sock->file),
973
				       (int __user *)argp);
L
Linus Torvalds 已提交
974 975 976 977 978 979 980 981 982
			break;
		case SIOCGIFBR:
		case SIOCSIFBR:
		case SIOCBRADDBR:
		case SIOCBRDELBR:
			err = -ENOPKG;
			if (!br_ioctl_hook)
				request_module("bridge");

A
Arjan van de Ven 已提交
983
			mutex_lock(&br_ioctl_mutex);
984
			if (br_ioctl_hook)
985
				err = br_ioctl_hook(net, cmd, argp);
A
Arjan van de Ven 已提交
986
			mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
987 988 989 990 991 992 993
			break;
		case SIOCGIFVLAN:
		case SIOCSIFVLAN:
			err = -ENOPKG;
			if (!vlan_ioctl_hook)
				request_module("8021q");

A
Arjan van de Ven 已提交
994
			mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
995
			if (vlan_ioctl_hook)
996
				err = vlan_ioctl_hook(net, argp);
A
Arjan van de Ven 已提交
997
			mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
998 999 1000 1001 1002 1003 1004
			break;
		case SIOCADDDLCI:
		case SIOCDELDLCI:
			err = -ENOPKG;
			if (!dlci_ioctl_hook)
				request_module("dlci");

1005 1006
			mutex_lock(&dlci_ioctl_mutex);
			if (dlci_ioctl_hook)
L
Linus Torvalds 已提交
1007
				err = dlci_ioctl_hook(cmd, argp);
1008
			mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
1009 1010
			break;
		default:
1011
			err = sock_do_ioctl(net, sock, cmd, arg);
L
Linus Torvalds 已提交
1012
			break;
1013
		}
L
Linus Torvalds 已提交
1014 1015 1016 1017 1018 1019 1020
	return err;
}

int sock_create_lite(int family, int type, int protocol, struct socket **res)
{
	int err;
	struct socket *sock = NULL;
1021

L
Linus Torvalds 已提交
1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
	err = security_socket_create(family, type, protocol, 1);
	if (err)
		goto out;

	sock = sock_alloc();
	if (!sock) {
		err = -ENOMEM;
		goto out;
	}

	sock->type = type;
V
Venkat Yekkirala 已提交
1033 1034 1035 1036
	err = security_socket_post_create(sock, family, type, protocol, 1);
	if (err)
		goto out_release;

L
Linus Torvalds 已提交
1037 1038 1039
out:
	*res = sock;
	return err;
V
Venkat Yekkirala 已提交
1040 1041 1042 1043
out_release:
	sock_release(sock);
	sock = NULL;
	goto out;
L
Linus Torvalds 已提交
1044
}
1045
EXPORT_SYMBOL(sock_create_lite);
L
Linus Torvalds 已提交
1046 1047

/* No kernel lock held - perfect */
1048
static unsigned int sock_poll(struct file *file, poll_table *wait)
L
Linus Torvalds 已提交
1049 1050 1051 1052
{
	struct socket *sock;

	/*
1053
	 *      We can't return errors to poll, so it's either yes or no.
L
Linus Torvalds 已提交
1054
	 */
1055
	sock = file->private_data;
L
Linus Torvalds 已提交
1056 1057 1058
	return sock->ops->poll(file, sock, wait);
}

1059
static int sock_mmap(struct file *file, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
1060
{
1061
	struct socket *sock = file->private_data;
L
Linus Torvalds 已提交
1062 1063 1064 1065

	return sock->ops->mmap(file, sock, vma);
}

1066
static int sock_close(struct inode *inode, struct file *filp)
L
Linus Torvalds 已提交
1067 1068
{
	/*
1069 1070
	 *      It was possible the inode is NULL we were
	 *      closing an unfinished socket.
L
Linus Torvalds 已提交
1071 1072
	 */

1073
	if (!inode) {
L
Linus Torvalds 已提交
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088
		printk(KERN_DEBUG "sock_close: NULL inode\n");
		return 0;
	}
	sock_release(SOCKET_I(inode));
	return 0;
}

/*
 *	Update the socket async list
 *
 *	Fasync_list locking strategy.
 *
 *	1. fasync_list is modified only under process context socket lock
 *	   i.e. under semaphore.
 *	2. fasync_list is used under read_lock(&sk->sk_callback_lock)
1089
 *	   or under socket lock
L
Linus Torvalds 已提交
1090 1091 1092 1093
 */

static int sock_fasync(int fd, struct file *filp, int on)
{
1094 1095
	struct socket *sock = filp->private_data;
	struct sock *sk = sock->sk;
L
Linus Torvalds 已提交
1096

1097
	if (sk == NULL)
L
Linus Torvalds 已提交
1098 1099 1100 1101
		return -EINVAL;

	lock_sock(sk);

1102
	fasync_helper(fd, filp, on, &sock->wq->fasync_list);
L
Linus Torvalds 已提交
1103

1104
	if (!sock->wq->fasync_list)
1105 1106
		sock_reset_flag(sk, SOCK_FASYNC);
	else
E
Eric Dumazet 已提交
1107
		sock_set_flag(sk, SOCK_FASYNC);
L
Linus Torvalds 已提交
1108

1109
	release_sock(sk);
L
Linus Torvalds 已提交
1110 1111 1112
	return 0;
}

1113
/* This function may be called only under socket lock or callback_lock or rcu_lock */
L
Linus Torvalds 已提交
1114 1115 1116

int sock_wake_async(struct socket *sock, int how, int band)
{
1117 1118 1119 1120 1121 1122 1123 1124
	struct socket_wq *wq;

	if (!sock)
		return -1;
	rcu_read_lock();
	wq = rcu_dereference(sock->wq);
	if (!wq || !wq->fasync_list) {
		rcu_read_unlock();
L
Linus Torvalds 已提交
1125
		return -1;
1126
	}
1127
	switch (how) {
1128
	case SOCK_WAKE_WAITD:
L
Linus Torvalds 已提交
1129 1130 1131
		if (test_bit(SOCK_ASYNC_WAITDATA, &sock->flags))
			break;
		goto call_kill;
1132
	case SOCK_WAKE_SPACE:
L
Linus Torvalds 已提交
1133 1134 1135
		if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags))
			break;
		/* fall through */
1136
	case SOCK_WAKE_IO:
1137
call_kill:
1138
		kill_fasync(&wq->fasync_list, SIGIO, band);
L
Linus Torvalds 已提交
1139
		break;
1140
	case SOCK_WAKE_URG:
1141
		kill_fasync(&wq->fasync_list, SIGURG, band);
L
Linus Torvalds 已提交
1142
	}
1143
	rcu_read_unlock();
L
Linus Torvalds 已提交
1144 1145
	return 0;
}
1146
EXPORT_SYMBOL(sock_wake_async);
L
Linus Torvalds 已提交
1147

1148
static int __sock_create(struct net *net, int family, int type, int protocol,
1149
			 struct socket **res, int kern)
L
Linus Torvalds 已提交
1150 1151 1152
{
	int err;
	struct socket *sock;
1153
	const struct net_proto_family *pf;
L
Linus Torvalds 已提交
1154 1155

	/*
1156
	 *      Check protocol is in range
L
Linus Torvalds 已提交
1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168
	 */
	if (family < 0 || family >= NPROTO)
		return -EAFNOSUPPORT;
	if (type < 0 || type >= SOCK_MAX)
		return -EINVAL;

	/* Compatibility.

	   This uglymoron is moved from INET layer to here to avoid
	   deadlock in module load.
	 */
	if (family == PF_INET && type == SOCK_PACKET) {
1169
		static int warned;
L
Linus Torvalds 已提交
1170 1171
		if (!warned) {
			warned = 1;
1172 1173
			printk(KERN_INFO "%s uses obsolete (PF_INET,SOCK_PACKET)\n",
			       current->comm);
L
Linus Torvalds 已提交
1174 1175 1176 1177 1178 1179 1180
		}
		family = PF_PACKET;
	}

	err = security_socket_create(family, type, protocol, kern);
	if (err)
		return err;
1181

1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
	/*
	 *	Allocate the socket and allow the family to set things up. if
	 *	the protocol is 0, the family is instructed to select an appropriate
	 *	default.
	 */
	sock = sock_alloc();
	if (!sock) {
		if (net_ratelimit())
			printk(KERN_WARNING "socket: no more sockets\n");
		return -ENFILE;	/* Not exactly a match, but its the
				   closest posix thing */
	}

	sock->type = type;

1197
#ifdef CONFIG_MODULES
1198 1199 1200
	/* Attempt to load a protocol module if the find failed.
	 *
	 * 12/09/1996 Marcin: But! this makes REALLY only sense, if the user
L
Linus Torvalds 已提交
1201 1202 1203
	 * requested real, full-featured networking support upon configuration.
	 * Otherwise module support will break!
	 */
1204
	if (net_families[family] == NULL)
1205
		request_module("net-pf-%d", family);
L
Linus Torvalds 已提交
1206 1207
#endif

1208 1209 1210 1211 1212
	rcu_read_lock();
	pf = rcu_dereference(net_families[family]);
	err = -EAFNOSUPPORT;
	if (!pf)
		goto out_release;
L
Linus Torvalds 已提交
1213 1214 1215 1216 1217

	/*
	 * We will call the ->create function, that possibly is in a loadable
	 * module, so we have to bump that loadable module refcnt first.
	 */
1218
	if (!try_module_get(pf->owner))
L
Linus Torvalds 已提交
1219 1220
		goto out_release;

1221 1222 1223
	/* Now protected by module ref count */
	rcu_read_unlock();

1224
	err = pf->create(net, sock, protocol, kern);
1225
	if (err < 0)
L
Linus Torvalds 已提交
1226
		goto out_module_put;
1227

L
Linus Torvalds 已提交
1228 1229 1230 1231
	/*
	 * Now to bump the refcnt of the [loadable] module that owns this
	 * socket at sock_release time we decrement its refcnt.
	 */
1232 1233 1234
	if (!try_module_get(sock->ops->owner))
		goto out_module_busy;

L
Linus Torvalds 已提交
1235 1236 1237 1238
	/*
	 * Now that we're done with the ->create function, the [loadable]
	 * module can have its refcnt decremented
	 */
1239
	module_put(pf->owner);
V
Venkat Yekkirala 已提交
1240 1241
	err = security_socket_post_create(sock, family, type, protocol, kern);
	if (err)
1242
		goto out_sock_release;
1243
	*res = sock;
L
Linus Torvalds 已提交
1244

1245 1246 1247 1248
	return 0;

out_module_busy:
	err = -EAFNOSUPPORT;
L
Linus Torvalds 已提交
1249
out_module_put:
1250 1251 1252
	sock->ops = NULL;
	module_put(pf->owner);
out_sock_release:
L
Linus Torvalds 已提交
1253
	sock_release(sock);
1254 1255 1256 1257 1258
	return err;

out_release:
	rcu_read_unlock();
	goto out_sock_release;
L
Linus Torvalds 已提交
1259 1260 1261 1262
}

int sock_create(int family, int type, int protocol, struct socket **res)
{
1263
	return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0);
L
Linus Torvalds 已提交
1264
}
1265
EXPORT_SYMBOL(sock_create);
L
Linus Torvalds 已提交
1266 1267 1268

int sock_create_kern(int family, int type, int protocol, struct socket **res)
{
1269
	return __sock_create(&init_net, family, type, protocol, res, 1);
L
Linus Torvalds 已提交
1270
}
1271
EXPORT_SYMBOL(sock_create_kern);
L
Linus Torvalds 已提交
1272

1273
SYSCALL_DEFINE3(socket, int, family, int, type, int, protocol)
L
Linus Torvalds 已提交
1274 1275 1276
{
	int retval;
	struct socket *sock;
1277 1278
	int flags;

1279 1280 1281 1282 1283 1284
	/* Check the SOCK_* constants for consistency.  */
	BUILD_BUG_ON(SOCK_CLOEXEC != O_CLOEXEC);
	BUILD_BUG_ON((SOCK_MAX | SOCK_TYPE_MASK) != SOCK_TYPE_MASK);
	BUILD_BUG_ON(SOCK_CLOEXEC & SOCK_TYPE_MASK);
	BUILD_BUG_ON(SOCK_NONBLOCK & SOCK_TYPE_MASK);

1285
	flags = type & ~SOCK_TYPE_MASK;
1286
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1287 1288
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1289

U
Ulrich Drepper 已提交
1290 1291 1292
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1293 1294 1295 1296
	retval = sock_create(family, type, protocol, &sock);
	if (retval < 0)
		goto out;

1297
	retval = sock_map_fd(sock, flags & (O_CLOEXEC | O_NONBLOCK));
L
Linus Torvalds 已提交
1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313
	if (retval < 0)
		goto out_release;

out:
	/* It may be already another descriptor 8) Not kernel problem. */
	return retval;

out_release:
	sock_release(sock);
	return retval;
}

/*
 *	Create a pair of connected sockets.
 */

1314 1315
SYSCALL_DEFINE4(socketpair, int, family, int, type, int, protocol,
		int __user *, usockvec)
L
Linus Torvalds 已提交
1316 1317 1318
{
	struct socket *sock1, *sock2;
	int fd1, fd2, err;
A
Al Viro 已提交
1319
	struct file *newfile1, *newfile2;
1320 1321 1322
	int flags;

	flags = type & ~SOCK_TYPE_MASK;
1323
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1324 1325
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1326

U
Ulrich Drepper 已提交
1327 1328 1329
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343
	/*
	 * Obtain the first socket and check if the underlying protocol
	 * supports the socketpair call.
	 */

	err = sock_create(family, type, protocol, &sock1);
	if (err < 0)
		goto out;

	err = sock_create(family, type, protocol, &sock2);
	if (err < 0)
		goto out_release_1;

	err = sock1->ops->socketpair(sock1, sock2);
1344
	if (err < 0)
L
Linus Torvalds 已提交
1345 1346
		goto out_release_both;

1347
	fd1 = sock_alloc_file(sock1, &newfile1, flags);
1348 1349
	if (unlikely(fd1 < 0)) {
		err = fd1;
A
Al Viro 已提交
1350
		goto out_release_both;
1351
	}
L
Linus Torvalds 已提交
1352

1353
	fd2 = sock_alloc_file(sock2, &newfile2, flags);
1354 1355 1356 1357 1358 1359
	if (unlikely(fd2 < 0)) {
		err = fd2;
		fput(newfile1);
		put_unused_fd(fd1);
		sock_release(sock2);
		goto out;
A
Al Viro 已提交
1360 1361
	}

A
Al Viro 已提交
1362
	audit_fd_pair(fd1, fd2);
A
Al Viro 已提交
1363 1364
	fd_install(fd1, newfile1);
	fd_install(fd2, newfile2);
L
Linus Torvalds 已提交
1365 1366 1367 1368
	/* fd1 and fd2 may be already another descriptors.
	 * Not kernel problem.
	 */

1369
	err = put_user(fd1, &usockvec[0]);
L
Linus Torvalds 已提交
1370 1371 1372 1373 1374 1375 1376 1377 1378 1379
	if (!err)
		err = put_user(fd2, &usockvec[1]);
	if (!err)
		return 0;

	sys_close(fd2);
	sys_close(fd1);
	return err;

out_release_both:
1380
	sock_release(sock2);
L
Linus Torvalds 已提交
1381
out_release_1:
1382
	sock_release(sock1);
L
Linus Torvalds 已提交
1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394
out:
	return err;
}

/*
 *	Bind a name to a socket. Nothing much to do here since it's
 *	the protocol's responsibility to handle the local address.
 *
 *	We move the socket address to kernel space before we call
 *	the protocol layer (having also checked the address is ok).
 */

1395
SYSCALL_DEFINE3(bind, int, fd, struct sockaddr __user *, umyaddr, int, addrlen)
L
Linus Torvalds 已提交
1396 1397
{
	struct socket *sock;
1398
	struct sockaddr_storage address;
1399
	int err, fput_needed;
L
Linus Torvalds 已提交
1400

1401
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1402
	if (sock) {
1403
		err = move_addr_to_kernel(umyaddr, addrlen, (struct sockaddr *)&address);
1404 1405
		if (err >= 0) {
			err = security_socket_bind(sock,
1406
						   (struct sockaddr *)&address,
1407
						   addrlen);
1408 1409
			if (!err)
				err = sock->ops->bind(sock,
1410
						      (struct sockaddr *)
1411
						      &address, addrlen);
L
Linus Torvalds 已提交
1412
		}
1413
		fput_light(sock->file, fput_needed);
1414
	}
L
Linus Torvalds 已提交
1415 1416 1417 1418 1419 1420 1421 1422 1423
	return err;
}

/*
 *	Perform a listen. Basically, we allow the protocol to do anything
 *	necessary for a listen, and if that works, we mark the socket as
 *	ready for listening.
 */

1424
SYSCALL_DEFINE2(listen, int, fd, int, backlog)
L
Linus Torvalds 已提交
1425 1426
{
	struct socket *sock;
1427
	int err, fput_needed;
1428
	int somaxconn;
1429 1430 1431

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock) {
1432
		somaxconn = sock_net(sock->sk)->core.sysctl_somaxconn;
1433 1434
		if ((unsigned)backlog > somaxconn)
			backlog = somaxconn;
L
Linus Torvalds 已提交
1435 1436

		err = security_socket_listen(sock, backlog);
1437 1438
		if (!err)
			err = sock->ops->listen(sock, backlog);
L
Linus Torvalds 已提交
1439

1440
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456
	}
	return err;
}

/*
 *	For accept, we attempt to create a new socket, set up the link
 *	with the client, wake up the client, then return the new
 *	connected fd. We collect the address of the connector in kernel
 *	space and move it to user at the very end. This is unclean because
 *	we open the socket then return an error.
 *
 *	1003.1g adds the ability to recvmsg() to query connection pending
 *	status to recvmsg. We need to add that support in a way thats
 *	clean when we restucture accept also.
 */

1457 1458
SYSCALL_DEFINE4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen, int, flags)
L
Linus Torvalds 已提交
1459 1460
{
	struct socket *sock, *newsock;
1461
	struct file *newfile;
1462
	int err, len, newfd, fput_needed;
1463
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1464

1465
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
U
Ulrich Drepper 已提交
1466 1467 1468 1469 1470
		return -EINVAL;

	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

1471
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1472 1473 1474 1475
	if (!sock)
		goto out;

	err = -ENFILE;
1476 1477
	newsock = sock_alloc();
	if (!newsock)
L
Linus Torvalds 已提交
1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488
		goto out_put;

	newsock->type = sock->type;
	newsock->ops = sock->ops;

	/*
	 * We don't need try_module_get here, as the listening socket (sock)
	 * has the protocol module (sock->ops->owner) held.
	 */
	__module_get(newsock->ops->owner);

1489
	newfd = sock_alloc_file(newsock, &newfile, flags);
1490 1491
	if (unlikely(newfd < 0)) {
		err = newfd;
1492 1493
		sock_release(newsock);
		goto out_put;
1494 1495
	}

1496 1497
	err = security_socket_accept(sock, newsock);
	if (err)
1498
		goto out_fd;
1499

L
Linus Torvalds 已提交
1500 1501
	err = sock->ops->accept(sock, newsock, sock->file->f_flags);
	if (err < 0)
1502
		goto out_fd;
L
Linus Torvalds 已提交
1503 1504

	if (upeer_sockaddr) {
1505
		if (newsock->ops->getname(newsock, (struct sockaddr *)&address,
1506
					  &len, 2) < 0) {
L
Linus Torvalds 已提交
1507
			err = -ECONNABORTED;
1508
			goto out_fd;
L
Linus Torvalds 已提交
1509
		}
1510 1511
		err = move_addr_to_user((struct sockaddr *)&address,
					len, upeer_sockaddr, upeer_addrlen);
L
Linus Torvalds 已提交
1512
		if (err < 0)
1513
			goto out_fd;
L
Linus Torvalds 已提交
1514 1515 1516 1517
	}

	/* File flags are not inherited via accept() unlike another OSes. */

1518 1519
	fd_install(newfd, newfile);
	err = newfd;
L
Linus Torvalds 已提交
1520 1521

out_put:
1522
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1523 1524
out:
	return err;
1525
out_fd:
1526
	fput(newfile);
1527
	put_unused_fd(newfd);
L
Linus Torvalds 已提交
1528 1529 1530
	goto out_put;
}

1531 1532
SYSCALL_DEFINE3(accept, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen)
U
Ulrich Drepper 已提交
1533
{
U
Ulrich Drepper 已提交
1534
	return sys_accept4(fd, upeer_sockaddr, upeer_addrlen, 0);
U
Ulrich Drepper 已提交
1535 1536
}

L
Linus Torvalds 已提交
1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548
/*
 *	Attempt to connect to a socket with the server address.  The address
 *	is in user space so we verify it is OK and move it to kernel space.
 *
 *	For 1003.1g we need to add clean support for a bind to AF_UNSPEC to
 *	break bindings
 *
 *	NOTE: 1003.1g draft 6.3 is broken with respect to AX.25/NetROM and
 *	other SEQPACKET protocols that take time to connect() as it doesn't
 *	include the -EINPROGRESS status for such sockets.
 */

1549 1550
SYSCALL_DEFINE3(connect, int, fd, struct sockaddr __user *, uservaddr,
		int, addrlen)
L
Linus Torvalds 已提交
1551 1552
{
	struct socket *sock;
1553
	struct sockaddr_storage address;
1554
	int err, fput_needed;
L
Linus Torvalds 已提交
1555

1556
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1557 1558
	if (!sock)
		goto out;
1559
	err = move_addr_to_kernel(uservaddr, addrlen, (struct sockaddr *)&address);
L
Linus Torvalds 已提交
1560 1561 1562
	if (err < 0)
		goto out_put;

1563
	err =
1564
	    security_socket_connect(sock, (struct sockaddr *)&address, addrlen);
L
Linus Torvalds 已提交
1565 1566 1567
	if (err)
		goto out_put;

1568
	err = sock->ops->connect(sock, (struct sockaddr *)&address, addrlen,
L
Linus Torvalds 已提交
1569 1570
				 sock->file->f_flags);
out_put:
1571
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1572 1573 1574 1575 1576 1577 1578 1579 1580
out:
	return err;
}

/*
 *	Get the local address ('name') of a socket object. Move the obtained
 *	name to user space.
 */

1581 1582
SYSCALL_DEFINE3(getsockname, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1583 1584
{
	struct socket *sock;
1585
	struct sockaddr_storage address;
1586
	int len, err, fput_needed;
1587

1588
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1589 1590 1591 1592 1593 1594 1595
	if (!sock)
		goto out;

	err = security_socket_getsockname(sock);
	if (err)
		goto out_put;

1596
	err = sock->ops->getname(sock, (struct sockaddr *)&address, &len, 0);
L
Linus Torvalds 已提交
1597 1598
	if (err)
		goto out_put;
1599
	err = move_addr_to_user((struct sockaddr *)&address, len, usockaddr, usockaddr_len);
L
Linus Torvalds 已提交
1600 1601

out_put:
1602
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1603 1604 1605 1606 1607 1608 1609 1610 1611
out:
	return err;
}

/*
 *	Get the remote address ('name') of a socket object. Move the obtained
 *	name to user space.
 */

1612 1613
SYSCALL_DEFINE3(getpeername, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1614 1615
{
	struct socket *sock;
1616
	struct sockaddr_storage address;
1617
	int len, err, fput_needed;
L
Linus Torvalds 已提交
1618

1619 1620
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1621 1622
		err = security_socket_getpeername(sock);
		if (err) {
1623
			fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1624 1625 1626
			return err;
		}

1627
		err =
1628
		    sock->ops->getname(sock, (struct sockaddr *)&address, &len,
1629
				       1);
L
Linus Torvalds 已提交
1630
		if (!err)
1631
			err = move_addr_to_user((struct sockaddr *)&address, len, usockaddr,
1632
						usockaddr_len);
1633
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1634 1635 1636 1637 1638 1639 1640 1641 1642 1643
	}
	return err;
}

/*
 *	Send a datagram to a given address. We move the address into kernel
 *	space and check the user space data area is readable before invoking
 *	the protocol.
 */

1644 1645 1646
SYSCALL_DEFINE6(sendto, int, fd, void __user *, buff, size_t, len,
		unsigned, flags, struct sockaddr __user *, addr,
		int, addr_len)
L
Linus Torvalds 已提交
1647 1648
{
	struct socket *sock;
1649
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1650 1651 1652
	int err;
	struct msghdr msg;
	struct iovec iov;
1653 1654
	int fput_needed;

1655 1656
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (!sock)
1657
		goto out;
1658

1659 1660 1661 1662 1663 1664 1665 1666
	iov.iov_base = buff;
	iov.iov_len = len;
	msg.msg_name = NULL;
	msg.msg_iov = &iov;
	msg.msg_iovlen = 1;
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_namelen = 0;
1667
	if (addr) {
1668
		err = move_addr_to_kernel(addr, addr_len, (struct sockaddr *)&address);
L
Linus Torvalds 已提交
1669 1670
		if (err < 0)
			goto out_put;
1671
		msg.msg_name = (struct sockaddr *)&address;
1672
		msg.msg_namelen = addr_len;
L
Linus Torvalds 已提交
1673 1674 1675 1676 1677 1678
	}
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
	msg.msg_flags = flags;
	err = sock_sendmsg(sock, &msg, len);

1679
out_put:
1680
	fput_light(sock->file, fput_needed);
1681
out:
L
Linus Torvalds 已提交
1682 1683 1684 1685
	return err;
}

/*
1686
 *	Send a datagram down a socket.
L
Linus Torvalds 已提交
1687 1688
 */

1689 1690
SYSCALL_DEFINE4(send, int, fd, void __user *, buff, size_t, len,
		unsigned, flags)
L
Linus Torvalds 已提交
1691 1692 1693 1694 1695
{
	return sys_sendto(fd, buff, len, flags, NULL, 0);
}

/*
1696
 *	Receive a frame from the socket and optionally record the address of the
L
Linus Torvalds 已提交
1697 1698 1699 1700
 *	sender. We verify the buffers are writable and if needed move the
 *	sender address from kernel to user space.
 */

1701 1702 1703
SYSCALL_DEFINE6(recvfrom, int, fd, void __user *, ubuf, size_t, size,
		unsigned, flags, struct sockaddr __user *, addr,
		int __user *, addr_len)
L
Linus Torvalds 已提交
1704 1705 1706 1707
{
	struct socket *sock;
	struct iovec iov;
	struct msghdr msg;
1708
	struct sockaddr_storage address;
1709
	int err, err2;
1710 1711
	int fput_needed;

1712
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1713
	if (!sock)
1714
		goto out;
L
Linus Torvalds 已提交
1715

1716 1717 1718 1719 1720 1721
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_iovlen = 1;
	msg.msg_iov = &iov;
	iov.iov_len = size;
	iov.iov_base = ubuf;
1722 1723
	msg.msg_name = (struct sockaddr *)&address;
	msg.msg_namelen = sizeof(address);
L
Linus Torvalds 已提交
1724 1725
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
1726
	err = sock_recvmsg(sock, &msg, size, flags);
L
Linus Torvalds 已提交
1727

1728
	if (err >= 0 && addr != NULL) {
1729 1730
		err2 = move_addr_to_user((struct sockaddr *)&address,
					 msg.msg_namelen, addr, addr_len);
1731 1732
		if (err2 < 0)
			err = err2;
L
Linus Torvalds 已提交
1733
	}
1734 1735

	fput_light(sock->file, fput_needed);
1736
out:
L
Linus Torvalds 已提交
1737 1738 1739 1740
	return err;
}

/*
1741
 *	Receive a datagram from a socket.
L
Linus Torvalds 已提交
1742 1743
 */

1744 1745
asmlinkage long sys_recv(int fd, void __user *ubuf, size_t size,
			 unsigned flags)
L
Linus Torvalds 已提交
1746 1747 1748 1749 1750 1751 1752 1753 1754
{
	return sys_recvfrom(fd, ubuf, size, flags, NULL, NULL);
}

/*
 *	Set a socket option. Because we don't know the option lengths we have
 *	to pass the user mode parameter for the protocols to sort out.
 */

1755 1756
SYSCALL_DEFINE5(setsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int, optlen)
L
Linus Torvalds 已提交
1757
{
1758
	int err, fput_needed;
L
Linus Torvalds 已提交
1759 1760 1761 1762
	struct socket *sock;

	if (optlen < 0)
		return -EINVAL;
1763 1764 1765 1766

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
		err = security_socket_setsockopt(sock, level, optname);
1767 1768
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1769 1770

		if (level == SOL_SOCKET)
1771 1772 1773
			err =
			    sock_setsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1774
		else
1775 1776 1777
			err =
			    sock->ops->setsockopt(sock, level, optname, optval,
						  optlen);
1778 1779
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1780 1781 1782 1783 1784 1785 1786 1787 1788
	}
	return err;
}

/*
 *	Get a socket option. Because we don't know the option lengths we have
 *	to pass a user mode parameter for the protocols to sort out.
 */

1789 1790
SYSCALL_DEFINE5(getsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int __user *, optlen)
L
Linus Torvalds 已提交
1791
{
1792
	int err, fput_needed;
L
Linus Torvalds 已提交
1793 1794
	struct socket *sock;

1795 1796
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
1797 1798 1799
		err = security_socket_getsockopt(sock, level, optname);
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1800 1801

		if (level == SOL_SOCKET)
1802 1803 1804
			err =
			    sock_getsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1805
		else
1806 1807 1808
			err =
			    sock->ops->getsockopt(sock, level, optname, optval,
						  optlen);
1809 1810
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1811 1812 1813 1814 1815 1816 1817 1818
	}
	return err;
}

/*
 *	Shutdown a socket.
 */

1819
SYSCALL_DEFINE2(shutdown, int, fd, int, how)
L
Linus Torvalds 已提交
1820
{
1821
	int err, fput_needed;
L
Linus Torvalds 已提交
1822 1823
	struct socket *sock;

1824 1825
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1826
		err = security_socket_shutdown(sock, how);
1827 1828 1829
		if (!err)
			err = sock->ops->shutdown(sock, how);
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1830 1831 1832 1833
	}
	return err;
}

1834
/* A couple of helpful macros for getting the address of the 32/64 bit
L
Linus Torvalds 已提交
1835 1836 1837 1838 1839 1840 1841 1842 1843 1844
 * fields which are the same type (int / unsigned) on our platforms.
 */
#define COMPAT_MSG(msg, member)	((MSG_CMSG_COMPAT & flags) ? &msg##_compat->member : &msg->member)
#define COMPAT_NAMELEN(msg)	COMPAT_MSG(msg, msg_namelen)
#define COMPAT_FLAGS(msg)	COMPAT_MSG(msg, msg_flags)

/*
 *	BSD sendmsg interface
 */

1845
SYSCALL_DEFINE3(sendmsg, int, fd, struct msghdr __user *, msg, unsigned, flags)
L
Linus Torvalds 已提交
1846
{
1847 1848
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
1849
	struct socket *sock;
1850
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1851
	struct iovec iovstack[UIO_FASTIOV], *iov = iovstack;
1852
	unsigned char ctl[sizeof(struct cmsghdr) + 20]
1853 1854
	    __attribute__ ((aligned(sizeof(__kernel_size_t))));
	/* 20 is size of ipv6_pktinfo */
L
Linus Torvalds 已提交
1855 1856 1857
	unsigned char *ctl_buf = ctl;
	struct msghdr msg_sys;
	int err, ctl_len, iov_size, total_len;
1858
	int fput_needed;
1859

L
Linus Torvalds 已提交
1860 1861 1862 1863
	err = -EFAULT;
	if (MSG_CMSG_COMPAT & flags) {
		if (get_compat_msghdr(&msg_sys, msg_compat))
			return -EFAULT;
1864
	} else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr)))
L
Linus Torvalds 已提交
1865 1866
		return -EFAULT;

1867
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1868
	if (!sock)
L
Linus Torvalds 已提交
1869 1870 1871 1872 1873 1874 1875
		goto out;

	/* do not move before msg_sys is valid */
	err = -EMSGSIZE;
	if (msg_sys.msg_iovlen > UIO_MAXIOV)
		goto out_put;

1876
	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
1877 1878 1879 1880 1881 1882 1883 1884 1885 1886
	err = -ENOMEM;
	iov_size = msg_sys.msg_iovlen * sizeof(struct iovec);
	if (msg_sys.msg_iovlen > UIO_FASTIOV) {
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
			goto out_put;
	}

	/* This will also move the address data into kernel space */
	if (MSG_CMSG_COMPAT & flags) {
1887 1888 1889
		err = verify_compat_iovec(&msg_sys, iov,
					  (struct sockaddr *)&address,
					  VERIFY_READ);
L
Linus Torvalds 已提交
1890
	} else
1891 1892 1893
		err = verify_iovec(&msg_sys, iov,
				   (struct sockaddr *)&address,
				   VERIFY_READ);
1894
	if (err < 0)
L
Linus Torvalds 已提交
1895 1896 1897 1898 1899 1900 1901
		goto out_freeiov;
	total_len = err;

	err = -ENOBUFS;

	if (msg_sys.msg_controllen > INT_MAX)
		goto out_freeiov;
1902
	ctl_len = msg_sys.msg_controllen;
L
Linus Torvalds 已提交
1903
	if ((MSG_CMSG_COMPAT & flags) && ctl_len) {
1904 1905 1906
		err =
		    cmsghdr_from_user_compat_to_kern(&msg_sys, sock->sk, ctl,
						     sizeof(ctl));
L
Linus Torvalds 已提交
1907 1908 1909
		if (err)
			goto out_freeiov;
		ctl_buf = msg_sys.msg_control;
A
Al Viro 已提交
1910
		ctl_len = msg_sys.msg_controllen;
L
Linus Torvalds 已提交
1911
	} else if (ctl_len) {
1912
		if (ctl_len > sizeof(ctl)) {
L
Linus Torvalds 已提交
1913
			ctl_buf = sock_kmalloc(sock->sk, ctl_len, GFP_KERNEL);
1914
			if (ctl_buf == NULL)
L
Linus Torvalds 已提交
1915 1916 1917 1918 1919 1920 1921 1922
				goto out_freeiov;
		}
		err = -EFAULT;
		/*
		 * Careful! Before this, msg_sys.msg_control contains a user pointer.
		 * Afterwards, it will be a kernel pointer. Thus the compiler-assisted
		 * checking falls down on this.
		 */
1923 1924
		if (copy_from_user(ctl_buf,
				   (void __user __force *)msg_sys.msg_control,
1925
				   ctl_len))
L
Linus Torvalds 已提交
1926 1927 1928 1929 1930 1931 1932 1933 1934 1935
			goto out_freectl;
		msg_sys.msg_control = ctl_buf;
	}
	msg_sys.msg_flags = flags;

	if (sock->file->f_flags & O_NONBLOCK)
		msg_sys.msg_flags |= MSG_DONTWAIT;
	err = sock_sendmsg(sock, &msg_sys, total_len);

out_freectl:
1936
	if (ctl_buf != ctl)
L
Linus Torvalds 已提交
1937 1938 1939 1940 1941
		sock_kfree_s(sock->sk, ctl_buf, ctl_len);
out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
out_put:
1942
	fput_light(sock->file, fput_needed);
1943
out:
L
Linus Torvalds 已提交
1944 1945 1946
	return err;
}

1947 1948
static int __sys_recvmsg(struct socket *sock, struct msghdr __user *msg,
			 struct msghdr *msg_sys, unsigned flags, int nosec)
L
Linus Torvalds 已提交
1949
{
1950 1951
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
1952
	struct iovec iovstack[UIO_FASTIOV];
1953
	struct iovec *iov = iovstack;
L
Linus Torvalds 已提交
1954 1955 1956 1957
	unsigned long cmsg_ptr;
	int err, iov_size, total_len, len;

	/* kernel mode address */
1958
	struct sockaddr_storage addr;
L
Linus Torvalds 已提交
1959 1960 1961 1962

	/* user mode address pointers */
	struct sockaddr __user *uaddr;
	int __user *uaddr_len;
1963

L
Linus Torvalds 已提交
1964
	if (MSG_CMSG_COMPAT & flags) {
1965
		if (get_compat_msghdr(msg_sys, msg_compat))
L
Linus Torvalds 已提交
1966
			return -EFAULT;
1967
	} else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr)))
1968
		return -EFAULT;
L
Linus Torvalds 已提交
1969 1970

	err = -EMSGSIZE;
1971 1972
	if (msg_sys->msg_iovlen > UIO_MAXIOV)
		goto out;
1973 1974

	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
1975
	err = -ENOMEM;
1976 1977
	iov_size = msg_sys->msg_iovlen * sizeof(struct iovec);
	if (msg_sys->msg_iovlen > UIO_FASTIOV) {
L
Linus Torvalds 已提交
1978 1979
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
1980
			goto out;
L
Linus Torvalds 已提交
1981 1982 1983
	}

	/*
1984 1985
	 *      Save the user-mode address (verify_iovec will change the
	 *      kernel msghdr to use the kernel address space)
L
Linus Torvalds 已提交
1986
	 */
1987

1988
	uaddr = (__force void __user *)msg_sys->msg_name;
L
Linus Torvalds 已提交
1989 1990
	uaddr_len = COMPAT_NAMELEN(msg);
	if (MSG_CMSG_COMPAT & flags) {
1991
		err = verify_compat_iovec(msg_sys, iov,
1992 1993
					  (struct sockaddr *)&addr,
					  VERIFY_WRITE);
L
Linus Torvalds 已提交
1994
	} else
1995
		err = verify_iovec(msg_sys, iov,
1996 1997
				   (struct sockaddr *)&addr,
				   VERIFY_WRITE);
L
Linus Torvalds 已提交
1998 1999
	if (err < 0)
		goto out_freeiov;
2000
	total_len = err;
L
Linus Torvalds 已提交
2001

2002 2003
	cmsg_ptr = (unsigned long)msg_sys->msg_control;
	msg_sys->msg_flags = flags & (MSG_CMSG_CLOEXEC|MSG_CMSG_COMPAT);
2004

L
Linus Torvalds 已提交
2005 2006
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
2007 2008
	err = (nosec ? sock_recvmsg_nosec : sock_recvmsg)(sock, msg_sys,
							  total_len, flags);
L
Linus Torvalds 已提交
2009 2010 2011 2012 2013
	if (err < 0)
		goto out_freeiov;
	len = err;

	if (uaddr != NULL) {
2014
		err = move_addr_to_user((struct sockaddr *)&addr,
2015
					msg_sys->msg_namelen, uaddr,
2016
					uaddr_len);
L
Linus Torvalds 已提交
2017 2018 2019
		if (err < 0)
			goto out_freeiov;
	}
2020
	err = __put_user((msg_sys->msg_flags & ~MSG_CMSG_COMPAT),
2021
			 COMPAT_FLAGS(msg));
L
Linus Torvalds 已提交
2022 2023 2024
	if (err)
		goto out_freeiov;
	if (MSG_CMSG_COMPAT & flags)
2025
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2026 2027
				 &msg_compat->msg_controllen);
	else
2028
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2029 2030 2031 2032 2033 2034 2035 2036
				 &msg->msg_controllen);
	if (err)
		goto out_freeiov;
	err = len;

out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056
out:
	return err;
}

/*
 *	BSD recvmsg interface
 */

SYSCALL_DEFINE3(recvmsg, int, fd, struct msghdr __user *, msg,
		unsigned int, flags)
{
	int fput_needed, err;
	struct msghdr msg_sys;
	struct socket *sock = sockfd_lookup_light(fd, &err, &fput_needed);

	if (!sock)
		goto out;

	err = __sys_recvmsg(sock, msg, &msg_sys, flags, 0);

2057
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
2058 2059 2060 2061
out:
	return err;
}

2062 2063 2064 2065 2066 2067 2068 2069 2070 2071
/*
 *     Linux recvmmsg interface
 */

int __sys_recvmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen,
		   unsigned int flags, struct timespec *timeout)
{
	int fput_needed, err, datagrams;
	struct socket *sock;
	struct mmsghdr __user *entry;
2072
	struct compat_mmsghdr __user *compat_entry;
2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091
	struct msghdr msg_sys;
	struct timespec end_time;

	if (timeout &&
	    poll_select_set_timeout(&end_time, timeout->tv_sec,
				    timeout->tv_nsec))
		return -EINVAL;

	datagrams = 0;

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (!sock)
		return err;

	err = sock_error(sock->sk);
	if (err)
		goto out_put;

	entry = mmsg;
2092
	compat_entry = (struct compat_mmsghdr __user *)mmsg;
2093 2094 2095 2096 2097

	while (datagrams < vlen) {
		/*
		 * No need to ask LSM for more than the first datagram.
		 */
2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113
		if (MSG_CMSG_COMPAT & flags) {
			err = __sys_recvmsg(sock, (struct msghdr __user *)compat_entry,
					    &msg_sys, flags, datagrams);
			if (err < 0)
				break;
			err = __put_user(err, &compat_entry->msg_len);
			++compat_entry;
		} else {
			err = __sys_recvmsg(sock, (struct msghdr __user *)entry,
					    &msg_sys, flags, datagrams);
			if (err < 0)
				break;
			err = put_user(err, &entry->msg_len);
			++entry;
		}

2114 2115 2116 2117
		if (err)
			break;
		++datagrams;

2118 2119 2120 2121
		/* MSG_WAITFORONE turns on MSG_DONTWAIT after one packet */
		if (flags & MSG_WAITFORONE)
			flags |= MSG_DONTWAIT;

2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141
		if (timeout) {
			ktime_get_ts(timeout);
			*timeout = timespec_sub(end_time, *timeout);
			if (timeout->tv_sec < 0) {
				timeout->tv_sec = timeout->tv_nsec = 0;
				break;
			}

			/* Timeout, return less than vlen datagrams */
			if (timeout->tv_nsec == 0 && timeout->tv_sec == 0)
				break;
		}

		/* Out of band data, return right away */
		if (msg_sys.msg_flags & MSG_OOB)
			break;
	}

out_put:
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
2142

2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189
	if (err == 0)
		return datagrams;

	if (datagrams != 0) {
		/*
		 * We may return less entries than requested (vlen) if the
		 * sock is non block and there aren't enough datagrams...
		 */
		if (err != -EAGAIN) {
			/*
			 * ... or  if recvmsg returns an error after we
			 * received some datagrams, where we record the
			 * error to return on the next call or if the
			 * app asks about it using getsockopt(SO_ERROR).
			 */
			sock->sk->sk_err = -err;
		}

		return datagrams;
	}

	return err;
}

SYSCALL_DEFINE5(recvmmsg, int, fd, struct mmsghdr __user *, mmsg,
		unsigned int, vlen, unsigned int, flags,
		struct timespec __user *, timeout)
{
	int datagrams;
	struct timespec timeout_sys;

	if (!timeout)
		return __sys_recvmmsg(fd, mmsg, vlen, flags, NULL);

	if (copy_from_user(&timeout_sys, timeout, sizeof(timeout_sys)))
		return -EFAULT;

	datagrams = __sys_recvmmsg(fd, mmsg, vlen, flags, &timeout_sys);

	if (datagrams > 0 &&
	    copy_to_user(timeout, &timeout_sys, sizeof(timeout_sys)))
		datagrams = -EFAULT;

	return datagrams;
}

#ifdef __ARCH_WANT_SYS_SOCKETCALL
L
Linus Torvalds 已提交
2190 2191
/* Argument list sizes for sys_socketcall */
#define AL(x) ((x) * sizeof(unsigned long))
2192
static const unsigned char nargs[20] = {
2193 2194 2195 2196
	AL(0), AL(3), AL(3), AL(3), AL(2), AL(3),
	AL(3), AL(3), AL(4), AL(4), AL(4), AL(6),
	AL(6), AL(2), AL(5), AL(5), AL(3), AL(3),
	AL(4), AL(5)
2197 2198
};

L
Linus Torvalds 已提交
2199 2200 2201
#undef AL

/*
2202
 *	System call vectors.
L
Linus Torvalds 已提交
2203 2204 2205
 *
 *	Argument checking cleaned up. Saved 20% in size.
 *  This function doesn't need to set the kernel lock because
2206
 *  it is set by the callees.
L
Linus Torvalds 已提交
2207 2208
 */

2209
SYSCALL_DEFINE2(socketcall, int, call, unsigned long __user *, args)
L
Linus Torvalds 已提交
2210 2211
{
	unsigned long a[6];
2212
	unsigned long a0, a1;
L
Linus Torvalds 已提交
2213
	int err;
2214
	unsigned int len;
L
Linus Torvalds 已提交
2215

2216
	if (call < 1 || call > SYS_RECVMMSG)
L
Linus Torvalds 已提交
2217 2218
		return -EINVAL;

2219 2220 2221 2222
	len = nargs[call];
	if (len > sizeof(a))
		return -EINVAL;

L
Linus Torvalds 已提交
2223
	/* copy_from_user should be SMP safe. */
2224
	if (copy_from_user(a, args, len))
L
Linus Torvalds 已提交
2225
		return -EFAULT;
2226

A
Al Viro 已提交
2227
	audit_socketcall(nargs[call] / sizeof(unsigned long), a);
2228

2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245
	a0 = a[0];
	a1 = a[1];

	switch (call) {
	case SYS_SOCKET:
		err = sys_socket(a0, a1, a[2]);
		break;
	case SYS_BIND:
		err = sys_bind(a0, (struct sockaddr __user *)a1, a[2]);
		break;
	case SYS_CONNECT:
		err = sys_connect(a0, (struct sockaddr __user *)a1, a[2]);
		break;
	case SYS_LISTEN:
		err = sys_listen(a0, a1);
		break;
	case SYS_ACCEPT:
U
Ulrich Drepper 已提交
2246 2247
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], 0);
2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293
		break;
	case SYS_GETSOCKNAME:
		err =
		    sys_getsockname(a0, (struct sockaddr __user *)a1,
				    (int __user *)a[2]);
		break;
	case SYS_GETPEERNAME:
		err =
		    sys_getpeername(a0, (struct sockaddr __user *)a1,
				    (int __user *)a[2]);
		break;
	case SYS_SOCKETPAIR:
		err = sys_socketpair(a0, a1, a[2], (int __user *)a[3]);
		break;
	case SYS_SEND:
		err = sys_send(a0, (void __user *)a1, a[2], a[3]);
		break;
	case SYS_SENDTO:
		err = sys_sendto(a0, (void __user *)a1, a[2], a[3],
				 (struct sockaddr __user *)a[4], a[5]);
		break;
	case SYS_RECV:
		err = sys_recv(a0, (void __user *)a1, a[2], a[3]);
		break;
	case SYS_RECVFROM:
		err = sys_recvfrom(a0, (void __user *)a1, a[2], a[3],
				   (struct sockaddr __user *)a[4],
				   (int __user *)a[5]);
		break;
	case SYS_SHUTDOWN:
		err = sys_shutdown(a0, a1);
		break;
	case SYS_SETSOCKOPT:
		err = sys_setsockopt(a0, a1, a[2], (char __user *)a[3], a[4]);
		break;
	case SYS_GETSOCKOPT:
		err =
		    sys_getsockopt(a0, a1, a[2], (char __user *)a[3],
				   (int __user *)a[4]);
		break;
	case SYS_SENDMSG:
		err = sys_sendmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
	case SYS_RECVMSG:
		err = sys_recvmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
2294 2295 2296 2297
	case SYS_RECVMMSG:
		err = sys_recvmmsg(a0, (struct mmsghdr __user *)a1, a[2], a[3],
				   (struct timespec __user *)a[4]);
		break;
U
Ulrich Drepper 已提交
2298 2299 2300
	case SYS_ACCEPT4:
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], a[3]);
U
Ulrich Drepper 已提交
2301
		break;
2302 2303 2304
	default:
		err = -EINVAL;
		break;
L
Linus Torvalds 已提交
2305 2306 2307 2308
	}
	return err;
}

2309
#endif				/* __ARCH_WANT_SYS_SOCKETCALL */
L
Linus Torvalds 已提交
2310

2311 2312 2313 2314
/**
 *	sock_register - add a socket protocol handler
 *	@ops: description of protocol
 *
L
Linus Torvalds 已提交
2315 2316
 *	This function is called by a protocol handler that wants to
 *	advertise its address family, and have it linked into the
2317 2318
 *	socket interface. The value ops->family coresponds to the
 *	socket system call protocol family.
L
Linus Torvalds 已提交
2319
 */
2320
int sock_register(const struct net_proto_family *ops)
L
Linus Torvalds 已提交
2321 2322 2323 2324
{
	int err;

	if (ops->family >= NPROTO) {
2325 2326
		printk(KERN_CRIT "protocol %d >= NPROTO(%d)\n", ops->family,
		       NPROTO);
L
Linus Torvalds 已提交
2327 2328
		return -ENOBUFS;
	}
2329 2330 2331 2332 2333

	spin_lock(&net_family_lock);
	if (net_families[ops->family])
		err = -EEXIST;
	else {
2334
		net_families[ops->family] = ops;
L
Linus Torvalds 已提交
2335 2336
		err = 0;
	}
2337 2338
	spin_unlock(&net_family_lock);

2339
	printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family);
L
Linus Torvalds 已提交
2340 2341
	return err;
}
2342
EXPORT_SYMBOL(sock_register);
L
Linus Torvalds 已提交
2343

2344 2345 2346 2347
/**
 *	sock_unregister - remove a protocol handler
 *	@family: protocol family to remove
 *
L
Linus Torvalds 已提交
2348 2349
 *	This function is called by a protocol handler that wants to
 *	remove its address family, and have it unlinked from the
2350 2351 2352 2353 2354 2355
 *	new socket creation.
 *
 *	If protocol handler is a module, then it can use module reference
 *	counts to protect against new references. If protocol handler is not
 *	a module then it needs to provide its own protection in
 *	the ops->create routine.
L
Linus Torvalds 已提交
2356
 */
2357
void sock_unregister(int family)
L
Linus Torvalds 已提交
2358
{
2359
	BUG_ON(family < 0 || family >= NPROTO);
L
Linus Torvalds 已提交
2360

2361
	spin_lock(&net_family_lock);
2362
	net_families[family] = NULL;
2363 2364 2365 2366
	spin_unlock(&net_family_lock);

	synchronize_rcu();

2367
	printk(KERN_INFO "NET: Unregistered protocol family %d\n", family);
L
Linus Torvalds 已提交
2368
}
2369
EXPORT_SYMBOL(sock_unregister);
L
Linus Torvalds 已提交
2370

2371
static int __init sock_init(void)
L
Linus Torvalds 已提交
2372 2373
{
	/*
2374
	 *      Initialize sock SLAB cache.
L
Linus Torvalds 已提交
2375
	 */
2376

L
Linus Torvalds 已提交
2377 2378 2379
	sk_init();

	/*
2380
	 *      Initialize skbuff SLAB cache
L
Linus Torvalds 已提交
2381 2382 2383 2384
	 */
	skb_init();

	/*
2385
	 *      Initialize the protocols module.
L
Linus Torvalds 已提交
2386 2387 2388 2389 2390
	 */

	init_inodecache();
	register_filesystem(&sock_fs_type);
	sock_mnt = kern_mount(&sock_fs_type);
2391 2392

	/* The real protocol initialization is performed in later initcalls.
L
Linus Torvalds 已提交
2393 2394 2395 2396 2397
	 */

#ifdef CONFIG_NETFILTER
	netfilter_init();
#endif
2398

2399 2400 2401 2402
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING
	skb_timestamping_init();
#endif

2403
	return 0;
L
Linus Torvalds 已提交
2404 2405
}

2406 2407
core_initcall(sock_init);	/* early initcall */

L
Linus Torvalds 已提交
2408 2409 2410 2411 2412 2413
#ifdef CONFIG_PROC_FS
void socket_seq_show(struct seq_file *seq)
{
	int cpu;
	int counter = 0;

2414
	for_each_possible_cpu(cpu)
2415
	    counter += per_cpu(sockets_in_use, cpu);
L
Linus Torvalds 已提交
2416 2417 2418 2419 2420 2421 2422

	/* It can be negative, by the way. 8) */
	if (counter < 0)
		counter = 0;

	seq_printf(seq, "sockets: used %d\n", counter);
}
2423
#endif				/* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
2424

2425
#ifdef CONFIG_COMPAT
2426 2427
static int do_siocgstamp(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_timeval __user *up)
2428 2429 2430 2431 2432 2433
{
	mm_segment_t old_fs = get_fs();
	struct timeval ktv;
	int err;

	set_fs(KERNEL_DS);
2434
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&ktv);
2435 2436 2437 2438 2439 2440 2441 2442
	set_fs(old_fs);
	if (!err) {
		err = put_user(ktv.tv_sec, &up->tv_sec);
		err |= __put_user(ktv.tv_usec, &up->tv_usec);
	}
	return err;
}

2443 2444
static int do_siocgstampns(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_timespec __user *up)
2445 2446 2447 2448 2449 2450
{
	mm_segment_t old_fs = get_fs();
	struct timespec kts;
	int err;

	set_fs(KERNEL_DS);
2451
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&kts);
2452 2453 2454 2455 2456 2457 2458 2459
	set_fs(old_fs);
	if (!err) {
		err = put_user(kts.tv_sec, &up->tv_sec);
		err |= __put_user(kts.tv_nsec, &up->tv_nsec);
	}
	return err;
}

2460
static int dev_ifname32(struct net *net, struct compat_ifreq __user *uifr32)
2461 2462 2463 2464 2465
{
	struct ifreq __user *uifr;
	int err;

	uifr = compat_alloc_user_space(sizeof(struct ifreq));
2466
	if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq)))
2467 2468
		return -EFAULT;

2469
	err = dev_ioctl(net, SIOCGIFNAME, uifr);
2470 2471 2472
	if (err)
		return err;

2473
	if (copy_in_user(uifr32, uifr, sizeof(struct compat_ifreq)))
2474 2475 2476 2477 2478
		return -EFAULT;

	return 0;
}

2479
static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32)
2480
{
2481
	struct compat_ifconf ifc32;
2482 2483
	struct ifconf ifc;
	struct ifconf __user *uifc;
2484
	struct compat_ifreq __user *ifr32;
2485 2486 2487 2488
	struct ifreq __user *ifr;
	unsigned int i, j;
	int err;

2489
	if (copy_from_user(&ifc32, uifc32, sizeof(struct compat_ifconf)))
2490 2491 2492 2493 2494 2495 2496 2497
		return -EFAULT;

	if (ifc32.ifcbuf == 0) {
		ifc32.ifc_len = 0;
		ifc.ifc_len = 0;
		ifc.ifc_req = NULL;
		uifc = compat_alloc_user_space(sizeof(struct ifconf));
	} else {
2498 2499
		size_t len = ((ifc32.ifc_len / sizeof(struct compat_ifreq)) + 1) *
			sizeof(struct ifreq);
2500 2501 2502 2503
		uifc = compat_alloc_user_space(sizeof(struct ifconf) + len);
		ifc.ifc_len = len;
		ifr = ifc.ifc_req = (void __user *)(uifc + 1);
		ifr32 = compat_ptr(ifc32.ifcbuf);
2504
		for (i = 0; i < ifc32.ifc_len; i += sizeof(struct compat_ifreq)) {
2505
			if (copy_in_user(ifr, ifr32, sizeof(struct compat_ifreq)))
2506 2507 2508 2509 2510 2511 2512 2513
				return -EFAULT;
			ifr++;
			ifr32++;
		}
	}
	if (copy_to_user(uifc, &ifc, sizeof(struct ifconf)))
		return -EFAULT;

2514
	err = dev_ioctl(net, SIOCGIFCONF, uifc);
2515 2516 2517 2518 2519 2520 2521 2522 2523
	if (err)
		return err;

	if (copy_from_user(&ifc, uifc, sizeof(struct ifconf)))
		return -EFAULT;

	ifr = ifc.ifc_req;
	ifr32 = compat_ptr(ifc32.ifcbuf);
	for (i = 0, j = 0;
2524 2525 2526
	     i + sizeof(struct compat_ifreq) <= ifc32.ifc_len && j < ifc.ifc_len;
	     i += sizeof(struct compat_ifreq), j += sizeof(struct ifreq)) {
		if (copy_in_user(ifr32, ifr, sizeof(struct compat_ifreq)))
2527 2528 2529 2530 2531 2532 2533 2534 2535 2536
			return -EFAULT;
		ifr32++;
		ifr++;
	}

	if (ifc32.ifcbuf == 0) {
		/* Translate from 64-bit structure multiple to
		 * a 32-bit one.
		 */
		i = ifc.ifc_len;
2537
		i = ((i / sizeof(struct ifreq)) * sizeof(struct compat_ifreq));
2538 2539 2540 2541
		ifc32.ifc_len = i;
	} else {
		ifc32.ifc_len = i;
	}
2542
	if (copy_to_user(uifc32, &ifc32, sizeof(struct compat_ifconf)))
2543 2544 2545 2546 2547
		return -EFAULT;

	return 0;
}

2548
static int ethtool_ioctl(struct net *net, struct compat_ifreq __user *ifr32)
2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565
{
	struct ifreq __user *ifr;
	u32 data;
	void __user *datap;

	ifr = compat_alloc_user_space(sizeof(*ifr));

	if (copy_in_user(&ifr->ifr_name, &ifr32->ifr_name, IFNAMSIZ))
		return -EFAULT;

	if (get_user(data, &ifr32->ifr_ifru.ifru_data))
		return -EFAULT;

	datap = compat_ptr(data);
	if (put_user(datap, &ifr->ifr_ifru.ifru_data))
		return -EFAULT;

2566
	return dev_ioctl(net, SIOCETHTOOL, ifr);
2567 2568
}

2569 2570 2571 2572 2573 2574
static int compat_siocwandev(struct net *net, struct compat_ifreq __user *uifr32)
{
	void __user *uptr;
	compat_uptr_t uptr32;
	struct ifreq __user *uifr;

2575
	uifr = compat_alloc_user_space(sizeof(*uifr));
2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589
	if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq)))
		return -EFAULT;

	if (get_user(uptr32, &uifr32->ifr_settings.ifs_ifsu))
		return -EFAULT;

	uptr = compat_ptr(uptr32);

	if (put_user(uptr, &uifr->ifr_settings.ifs_ifsu.raw_hdlc))
		return -EFAULT;

	return dev_ioctl(net, SIOCWANDEV, uifr);
}

2590 2591
static int bond_ioctl(struct net *net, unsigned int cmd,
			 struct compat_ifreq __user *ifr32)
2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604
{
	struct ifreq kifr;
	struct ifreq __user *uifr;
	mm_segment_t old_fs;
	int err;
	u32 data;
	void __user *datap;

	switch (cmd) {
	case SIOCBONDENSLAVE:
	case SIOCBONDRELEASE:
	case SIOCBONDSETHWADDR:
	case SIOCBONDCHANGEACTIVE:
2605
		if (copy_from_user(&kifr, ifr32, sizeof(struct compat_ifreq)))
2606 2607 2608
			return -EFAULT;

		old_fs = get_fs();
2609
		set_fs(KERNEL_DS);
2610
		err = dev_ioctl(net, cmd, &kifr);
2611
		set_fs(old_fs);
2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626

		return err;
	case SIOCBONDSLAVEINFOQUERY:
	case SIOCBONDINFOQUERY:
		uifr = compat_alloc_user_space(sizeof(*uifr));
		if (copy_in_user(&uifr->ifr_name, &ifr32->ifr_name, IFNAMSIZ))
			return -EFAULT;

		if (get_user(data, &ifr32->ifr_ifru.ifru_data))
			return -EFAULT;

		datap = compat_ptr(data);
		if (put_user(datap, &uifr->ifr_ifru.ifru_data))
			return -EFAULT;

2627
		return dev_ioctl(net, cmd, uifr);
2628 2629
	default:
		return -EINVAL;
2630
	}
2631 2632
}

2633 2634
static int siocdevprivate_ioctl(struct net *net, unsigned int cmd,
				 struct compat_ifreq __user *u_ifreq32)
2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658
{
	struct ifreq __user *u_ifreq64;
	char tmp_buf[IFNAMSIZ];
	void __user *data64;
	u32 data32;

	if (copy_from_user(&tmp_buf[0], &(u_ifreq32->ifr_ifrn.ifrn_name[0]),
			   IFNAMSIZ))
		return -EFAULT;
	if (__get_user(data32, &u_ifreq32->ifr_ifru.ifru_data))
		return -EFAULT;
	data64 = compat_ptr(data32);

	u_ifreq64 = compat_alloc_user_space(sizeof(*u_ifreq64));

	/* Don't check these user accesses, just let that get trapped
	 * in the ioctl handler instead.
	 */
	if (copy_to_user(&u_ifreq64->ifr_ifrn.ifrn_name[0], &tmp_buf[0],
			 IFNAMSIZ))
		return -EFAULT;
	if (__put_user(data64, &u_ifreq64->ifr_ifru.ifru_data))
		return -EFAULT;

2659
	return dev_ioctl(net, cmd, u_ifreq64);
2660 2661
}

2662 2663
static int dev_ifsioc(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_ifreq __user *uifr32)
2664
{
2665
	struct ifreq __user *uifr;
2666 2667
	int err;

2668 2669 2670 2671 2672 2673
	uifr = compat_alloc_user_space(sizeof(*uifr));
	if (copy_in_user(uifr, uifr32, sizeof(*uifr32)))
		return -EFAULT;

	err = sock_do_ioctl(net, sock, cmd, (unsigned long)uifr);

2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685
	if (!err) {
		switch (cmd) {
		case SIOCGIFFLAGS:
		case SIOCGIFMETRIC:
		case SIOCGIFMTU:
		case SIOCGIFMEM:
		case SIOCGIFHWADDR:
		case SIOCGIFINDEX:
		case SIOCGIFADDR:
		case SIOCGIFBRDADDR:
		case SIOCGIFDSTADDR:
		case SIOCGIFNETMASK:
2686
		case SIOCGIFPFLAGS:
2687
		case SIOCGIFTXQLEN:
2688 2689
		case SIOCGMIIPHY:
		case SIOCGMIIREG:
2690
			if (copy_in_user(uifr32, uifr, sizeof(*uifr32)))
2691 2692 2693 2694 2695 2696 2697
				err = -EFAULT;
			break;
		}
	}
	return err;
}

2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717
static int compat_sioc_ifmap(struct net *net, unsigned int cmd,
			struct compat_ifreq __user *uifr32)
{
	struct ifreq ifr;
	struct compat_ifmap __user *uifmap32;
	mm_segment_t old_fs;
	int err;

	uifmap32 = &uifr32->ifr_ifru.ifru_map;
	err = copy_from_user(&ifr, uifr32, sizeof(ifr.ifr_name));
	err |= __get_user(ifr.ifr_map.mem_start, &uifmap32->mem_start);
	err |= __get_user(ifr.ifr_map.mem_end, &uifmap32->mem_end);
	err |= __get_user(ifr.ifr_map.base_addr, &uifmap32->base_addr);
	err |= __get_user(ifr.ifr_map.irq, &uifmap32->irq);
	err |= __get_user(ifr.ifr_map.dma, &uifmap32->dma);
	err |= __get_user(ifr.ifr_map.port, &uifmap32->port);
	if (err)
		return -EFAULT;

	old_fs = get_fs();
2718
	set_fs(KERNEL_DS);
2719
	err = dev_ioctl(net, cmd, (void __user *)&ifr);
2720
	set_fs(old_fs);
2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741

	if (cmd == SIOCGIFMAP && !err) {
		err = copy_to_user(uifr32, &ifr, sizeof(ifr.ifr_name));
		err |= __put_user(ifr.ifr_map.mem_start, &uifmap32->mem_start);
		err |= __put_user(ifr.ifr_map.mem_end, &uifmap32->mem_end);
		err |= __put_user(ifr.ifr_map.base_addr, &uifmap32->base_addr);
		err |= __put_user(ifr.ifr_map.irq, &uifmap32->irq);
		err |= __put_user(ifr.ifr_map.dma, &uifmap32->dma);
		err |= __put_user(ifr.ifr_map.port, &uifmap32->port);
		if (err)
			err = -EFAULT;
	}
	return err;
}

static int compat_siocshwtstamp(struct net *net, struct compat_ifreq __user *uifr32)
{
	void __user *uptr;
	compat_uptr_t uptr32;
	struct ifreq __user *uifr;

2742
	uifr = compat_alloc_user_space(sizeof(*uifr));
2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756
	if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq)))
		return -EFAULT;

	if (get_user(uptr32, &uifr32->ifr_data))
		return -EFAULT;

	uptr = compat_ptr(uptr32);

	if (put_user(uptr, &uifr->ifr_data))
		return -EFAULT;

	return dev_ioctl(net, SIOCSHWTSTAMP, uifr);
}

2757
struct rtentry32 {
2758
	u32		rt_pad1;
2759 2760 2761
	struct sockaddr rt_dst;         /* target address               */
	struct sockaddr rt_gateway;     /* gateway addr (RTF_GATEWAY)   */
	struct sockaddr rt_genmask;     /* target network mask (IP)     */
2762 2763 2764 2765 2766 2767 2768
	unsigned short	rt_flags;
	short		rt_pad2;
	u32		rt_pad3;
	unsigned char	rt_tos;
	unsigned char	rt_class;
	short		rt_pad4;
	short		rt_metric;      /* +1 for binary compatibility! */
2769
	/* char * */ u32 rt_dev;        /* forcing the device at add    */
2770 2771
	u32		rt_mtu;         /* per route MTU/Window         */
	u32		rt_window;      /* Window clamping              */
2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787
	unsigned short  rt_irtt;        /* Initial RTT                  */
};

struct in6_rtmsg32 {
	struct in6_addr		rtmsg_dst;
	struct in6_addr		rtmsg_src;
	struct in6_addr		rtmsg_gateway;
	u32			rtmsg_type;
	u16			rtmsg_dst_len;
	u16			rtmsg_src_len;
	u32			rtmsg_metric;
	u32			rtmsg_info;
	u32			rtmsg_flags;
	s32			rtmsg_ifindex;
};

2788 2789
static int routing_ioctl(struct net *net, struct socket *sock,
			 unsigned int cmd, void __user *argp)
2790 2791 2792 2793 2794 2795 2796 2797 2798
{
	int ret;
	void *r = NULL;
	struct in6_rtmsg r6;
	struct rtentry r4;
	char devname[16];
	u32 rtdev;
	mm_segment_t old_fs = get_fs();

2799 2800
	if (sock && sock->sk && sock->sk->sk_family == AF_INET6) { /* ipv6 */
		struct in6_rtmsg32 __user *ur6 = argp;
2801
		ret = copy_from_user(&r6.rtmsg_dst, &(ur6->rtmsg_dst),
2802
			3 * sizeof(struct in6_addr));
2803 2804 2805 2806 2807 2808 2809
		ret |= __get_user(r6.rtmsg_type, &(ur6->rtmsg_type));
		ret |= __get_user(r6.rtmsg_dst_len, &(ur6->rtmsg_dst_len));
		ret |= __get_user(r6.rtmsg_src_len, &(ur6->rtmsg_src_len));
		ret |= __get_user(r6.rtmsg_metric, &(ur6->rtmsg_metric));
		ret |= __get_user(r6.rtmsg_info, &(ur6->rtmsg_info));
		ret |= __get_user(r6.rtmsg_flags, &(ur6->rtmsg_flags));
		ret |= __get_user(r6.rtmsg_ifindex, &(ur6->rtmsg_ifindex));
2810 2811 2812

		r = (void *) &r6;
	} else { /* ipv4 */
2813
		struct rtentry32 __user *ur4 = argp;
2814
		ret = copy_from_user(&r4.rt_dst, &(ur4->rt_dst),
2815
					3 * sizeof(struct sockaddr));
2816 2817 2818 2819 2820 2821
		ret |= __get_user(r4.rt_flags, &(ur4->rt_flags));
		ret |= __get_user(r4.rt_metric, &(ur4->rt_metric));
		ret |= __get_user(r4.rt_mtu, &(ur4->rt_mtu));
		ret |= __get_user(r4.rt_window, &(ur4->rt_window));
		ret |= __get_user(r4.rt_irtt, &(ur4->rt_irtt));
		ret |= __get_user(rtdev, &(ur4->rt_dev));
2822
		if (rtdev) {
2823
			ret |= copy_from_user(devname, compat_ptr(rtdev), 15);
2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835
			r4.rt_dev = devname; devname[15] = 0;
		} else
			r4.rt_dev = NULL;

		r = (void *) &r4;
	}

	if (ret) {
		ret = -EFAULT;
		goto out;
	}

2836
	set_fs(KERNEL_DS);
2837
	ret = sock_do_ioctl(net, sock, cmd, (unsigned long) r);
2838
	set_fs(old_fs);
2839 2840 2841 2842 2843 2844 2845 2846 2847

out:
	return ret;
}

/* Since old style bridge ioctl's endup using SIOCDEVPRIVATE
 * for some operations; this forces use of the newer bridge-utils that
 * use compatiable ioctls
 */
2848
static int old_bridge_ioctl(compat_ulong_t __user *argp)
2849
{
2850
	compat_ulong_t tmp;
2851

2852
	if (get_user(tmp, argp))
2853 2854 2855 2856 2857 2858
		return -EFAULT;
	if (tmp == BRCTL_GET_VERSION)
		return BRCTL_VERSION + 1;
	return -EINVAL;
}

2859 2860 2861 2862 2863 2864
static int compat_sock_ioctl_trans(struct file *file, struct socket *sock,
			 unsigned int cmd, unsigned long arg)
{
	void __user *argp = compat_ptr(arg);
	struct sock *sk = sock->sk;
	struct net *net = sock_net(sk);
2865

2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878
	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15))
		return siocdevprivate_ioctl(net, cmd, argp);

	switch (cmd) {
	case SIOCSIFBR:
	case SIOCGIFBR:
		return old_bridge_ioctl(argp);
	case SIOCGIFNAME:
		return dev_ifname32(net, argp);
	case SIOCGIFCONF:
		return dev_ifconf(net, argp);
	case SIOCETHTOOL:
		return ethtool_ioctl(net, argp);
2879 2880
	case SIOCWANDEV:
		return compat_siocwandev(net, argp);
2881 2882 2883
	case SIOCGIFMAP:
	case SIOCSIFMAP:
		return compat_sioc_ifmap(net, cmd, argp);
2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897
	case SIOCBONDENSLAVE:
	case SIOCBONDRELEASE:
	case SIOCBONDSETHWADDR:
	case SIOCBONDSLAVEINFOQUERY:
	case SIOCBONDINFOQUERY:
	case SIOCBONDCHANGEACTIVE:
		return bond_ioctl(net, cmd, argp);
	case SIOCADDRT:
	case SIOCDELRT:
		return routing_ioctl(net, sock, cmd, argp);
	case SIOCGSTAMP:
		return do_siocgstamp(net, sock, cmd, argp);
	case SIOCGSTAMPNS:
		return do_siocgstampns(net, sock, cmd, argp);
2898 2899
	case SIOCSHWTSTAMP:
		return compat_siocshwtstamp(net, argp);
2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941

	case FIOSETOWN:
	case SIOCSPGRP:
	case FIOGETOWN:
	case SIOCGPGRP:
	case SIOCBRADDBR:
	case SIOCBRDELBR:
	case SIOCGIFVLAN:
	case SIOCSIFVLAN:
	case SIOCADDDLCI:
	case SIOCDELDLCI:
		return sock_ioctl(file, cmd, arg);

	case SIOCGIFFLAGS:
	case SIOCSIFFLAGS:
	case SIOCGIFMETRIC:
	case SIOCSIFMETRIC:
	case SIOCGIFMTU:
	case SIOCSIFMTU:
	case SIOCGIFMEM:
	case SIOCSIFMEM:
	case SIOCGIFHWADDR:
	case SIOCSIFHWADDR:
	case SIOCADDMULTI:
	case SIOCDELMULTI:
	case SIOCGIFINDEX:
	case SIOCGIFADDR:
	case SIOCSIFADDR:
	case SIOCSIFHWBROADCAST:
	case SIOCDIFADDR:
	case SIOCGIFBRDADDR:
	case SIOCSIFBRDADDR:
	case SIOCGIFDSTADDR:
	case SIOCSIFDSTADDR:
	case SIOCGIFNETMASK:
	case SIOCSIFNETMASK:
	case SIOCSIFPFLAGS:
	case SIOCGIFPFLAGS:
	case SIOCGIFTXQLEN:
	case SIOCSIFTXQLEN:
	case SIOCBRADDIF:
	case SIOCBRDELIF:
2942 2943 2944 2945
	case SIOCSIFNAME:
	case SIOCGMIIPHY:
	case SIOCGMIIREG:
	case SIOCSMIIREG:
2946
		return dev_ifsioc(net, sock, cmd, argp);
2947

2948 2949 2950 2951
	case SIOCSARP:
	case SIOCGARP:
	case SIOCDARP:
	case SIOCATMARK:
2952 2953 2954 2955 2956 2957 2958 2959
		return sock_do_ioctl(net, sock, cmd, arg);
	}

	/* Prevent warning from compat_sys_ioctl, these always
	 * result in -EINVAL in the native case anyway. */
	switch (cmd) {
	case SIOCRTMSG:
	case SIOCGIFCOUNT:
2960 2961 2962
	case SIOCSRARP:
	case SIOCGRARP:
	case SIOCDRARP:
2963 2964 2965 2966
	case SIOCSIFLINK:
	case SIOCGIFSLAVE:
	case SIOCSIFSLAVE:
		return -EINVAL;
2967 2968 2969 2970
	}

	return -ENOIOCTLCMD;
}
2971

2972
static long compat_sock_ioctl(struct file *file, unsigned cmd,
2973
			      unsigned long arg)
2974 2975 2976
{
	struct socket *sock = file->private_data;
	int ret = -ENOIOCTLCMD;
2977 2978 2979 2980 2981
	struct sock *sk;
	struct net *net;

	sk = sock->sk;
	net = sock_net(sk);
2982 2983 2984 2985

	if (sock->ops->compat_ioctl)
		ret = sock->ops->compat_ioctl(sock, cmd, arg);

2986 2987 2988 2989
	if (ret == -ENOIOCTLCMD &&
	    (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST))
		ret = compat_wext_handle_ioctl(net, cmd, arg);

2990 2991 2992
	if (ret == -ENOIOCTLCMD)
		ret = compat_sock_ioctl_trans(file, sock, cmd, arg);

2993 2994 2995 2996
	return ret;
}
#endif

2997 2998 2999 3000
int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen)
{
	return sock->ops->bind(sock, addr, addrlen);
}
3001
EXPORT_SYMBOL(kernel_bind);
3002 3003 3004 3005 3006

int kernel_listen(struct socket *sock, int backlog)
{
	return sock->ops->listen(sock, backlog);
}
3007
EXPORT_SYMBOL(kernel_listen);
3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021

int kernel_accept(struct socket *sock, struct socket **newsock, int flags)
{
	struct sock *sk = sock->sk;
	int err;

	err = sock_create_lite(sk->sk_family, sk->sk_type, sk->sk_protocol,
			       newsock);
	if (err < 0)
		goto done;

	err = sock->ops->accept(sock, *newsock, flags);
	if (err < 0) {
		sock_release(*newsock);
3022
		*newsock = NULL;
3023 3024 3025 3026
		goto done;
	}

	(*newsock)->ops = sock->ops;
3027
	__module_get((*newsock)->ops->owner);
3028 3029 3030 3031

done:
	return err;
}
3032
EXPORT_SYMBOL(kernel_accept);
3033 3034

int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen,
3035
		   int flags)
3036 3037 3038
{
	return sock->ops->connect(sock, addr, addrlen, flags);
}
3039
EXPORT_SYMBOL(kernel_connect);
3040 3041 3042 3043 3044 3045

int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 0);
}
3046
EXPORT_SYMBOL(kernel_getsockname);
3047 3048 3049 3050 3051 3052

int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 1);
}
3053
EXPORT_SYMBOL(kernel_getpeername);
3054 3055 3056 3057 3058

int kernel_getsockopt(struct socket *sock, int level, int optname,
			char *optval, int *optlen)
{
	mm_segment_t oldfs = get_fs();
3059 3060
	char __user *uoptval;
	int __user *uoptlen;
3061 3062
	int err;

3063 3064 3065
	uoptval = (char __user __force *) optval;
	uoptlen = (int __user __force *) optlen;

3066 3067
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3068
		err = sock_getsockopt(sock, level, optname, uoptval, uoptlen);
3069
	else
3070 3071
		err = sock->ops->getsockopt(sock, level, optname, uoptval,
					    uoptlen);
3072 3073 3074
	set_fs(oldfs);
	return err;
}
3075
EXPORT_SYMBOL(kernel_getsockopt);
3076 3077

int kernel_setsockopt(struct socket *sock, int level, int optname,
3078
			char *optval, unsigned int optlen)
3079 3080
{
	mm_segment_t oldfs = get_fs();
3081
	char __user *uoptval;
3082 3083
	int err;

3084 3085
	uoptval = (char __user __force *) optval;

3086 3087
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3088
		err = sock_setsockopt(sock, level, optname, uoptval, optlen);
3089
	else
3090
		err = sock->ops->setsockopt(sock, level, optname, uoptval,
3091 3092 3093 3094
					    optlen);
	set_fs(oldfs);
	return err;
}
3095
EXPORT_SYMBOL(kernel_setsockopt);
3096 3097 3098 3099

int kernel_sendpage(struct socket *sock, struct page *page, int offset,
		    size_t size, int flags)
{
3100 3101
	sock_update_classid(sock->sk);

3102 3103 3104 3105 3106
	if (sock->ops->sendpage)
		return sock->ops->sendpage(sock, page, offset, size, flags);

	return sock_no_sendpage(sock, page, offset, size, flags);
}
3107
EXPORT_SYMBOL(kernel_sendpage);
3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119

int kernel_sock_ioctl(struct socket *sock, int cmd, unsigned long arg)
{
	mm_segment_t oldfs = get_fs();
	int err;

	set_fs(KERNEL_DS);
	err = sock->ops->ioctl(sock, cmd, arg);
	set_fs(oldfs);

	return err;
}
3120
EXPORT_SYMBOL(kernel_sock_ioctl);
3121

3122 3123 3124 3125 3126
int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how)
{
	return sock->ops->shutdown(sock, how);
}
EXPORT_SYMBOL(kernel_sock_shutdown);