socket.c 81.6 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);
E
Eric Dumazet 已提交
159
static const struct net_proto_family __rcu *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
	struct socket_wq *wq;
244

245
	ei = kmem_cache_alloc(sock_inode_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
246 247
	if (!ei)
		return NULL;
248 249
	wq = kmalloc(sizeof(*wq), GFP_KERNEL);
	if (!wq) {
250 251 252
		kmem_cache_free(sock_inode_cachep, ei);
		return NULL;
	}
253 254 255
	init_waitqueue_head(&wq->wait);
	wq->fasync_list = NULL;
	RCU_INIT_POINTER(ei->socket.wq, wq);
256

L
Linus Torvalds 已提交
257 258 259 260 261 262 263 264 265 266 267
	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;
}

static void sock_destroy_inode(struct inode *inode)
{
268
	struct socket_alloc *ei;
269
	struct socket_wq *wq;
270 271

	ei = container_of(inode, struct socket_alloc, vfs_inode);
272
	wq = rcu_dereference_protected(ei->socket.wq, 1);
273
	kfree_rcu(wq, rcu);
274
	kmem_cache_free(sock_inode_cachep, ei);
L
Linus Torvalds 已提交
275 276
}

277
static void init_once(void *foo)
L
Linus Torvalds 已提交
278
{
279
	struct socket_alloc *ei = (struct socket_alloc *)foo;
L
Linus Torvalds 已提交
280

C
Christoph Lameter 已提交
281
	inode_init_once(&ei->vfs_inode);
L
Linus Torvalds 已提交
282
}
283

L
Linus Torvalds 已提交
284 285 286
static int init_inodecache(void)
{
	sock_inode_cachep = kmem_cache_create("sock_inode_cache",
287 288 289 290 291
					      sizeof(struct socket_alloc),
					      0,
					      (SLAB_HWCACHE_ALIGN |
					       SLAB_RECLAIM_ACCOUNT |
					       SLAB_MEM_SPREAD),
292
					      init_once);
L
Linus Torvalds 已提交
293 294 295 296 297
	if (sock_inode_cachep == NULL)
		return -ENOMEM;
	return 0;
}

298
static const struct super_operations sockfs_ops = {
299 300 301
	.alloc_inode	= sock_alloc_inode,
	.destroy_inode	= sock_destroy_inode,
	.statfs		= simple_statfs,
L
Linus Torvalds 已提交
302 303
};

304 305 306 307 308 309 310 311 312
/*
 * 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
313
static const struct dentry_operations sockfs_dentry_operations = {
314
	.d_dname  = sockfs_dname,
L
Linus Torvalds 已提交
315 316
};

317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
static struct dentry *sockfs_mount(struct file_system_type *fs_type,
			 int flags, const char *dev_name, void *data)
{
	return mount_pseudo(fs_type, "socket:", &sockfs_ops,
		&sockfs_dentry_operations, SOCKFS_MAGIC);
}

static struct vfsmount *sock_mnt __read_mostly;

static struct file_system_type sock_fs_type = {
	.name =		"sockfs",
	.mount =	sockfs_mount,
	.kill_sb =	kill_anon_super,
};

L
Linus Torvalds 已提交
332 333 334
/*
 *	Obtains the first available file descriptor and sets it up for use.
 *
335 336
 *	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
337 338 339 340 341 342 343 344 345 346 347 348
 *	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.
 */

349
static int sock_alloc_file(struct socket *sock, struct file **f, int flags)
L
Linus Torvalds 已提交
350
{
351
	struct qstr name = { .name = "" };
352
	struct path path;
353
	struct file *file;
L
Linus Torvalds 已提交
354 355
	int fd;

356
	fd = get_unused_fd_flags(flags);
357 358
	if (unlikely(fd < 0))
		return fd;
L
Linus Torvalds 已提交
359

360
	path.dentry = d_alloc_pseudo(sock_mnt->mnt_sb, &name);
361
	if (unlikely(!path.dentry)) {
362
		put_unused_fd(fd);
363
		return -ENOMEM;
364
	}
365
	path.mnt = mntget(sock_mnt);
366

367
	d_instantiate(path.dentry, SOCK_INODE(sock));
A
Al Viro 已提交
368
	SOCK_INODE(sock)->i_fop = &socket_file_ops;
369

370
	file = alloc_file(&path, FMODE_READ | FMODE_WRITE,
371
		  &socket_file_ops);
A
Al Viro 已提交
372 373
	if (unlikely(!file)) {
		/* drop dentry, keep inode */
A
Al Viro 已提交
374
		ihold(path.dentry->d_inode);
375
		path_put(&path);
A
Al Viro 已提交
376 377 378 379 380
		put_unused_fd(fd);
		return -ENFILE;
	}

	sock->file = file;
381
	file->f_flags = O_RDWR | (flags & O_NONBLOCK);
382 383
	file->f_pos = 0;
	file->private_data = sock;
L
Linus Torvalds 已提交
384

385 386
	*f = file;
	return fd;
387 388
}

389
int sock_map_fd(struct socket *sock, int flags)
390 391
{
	struct file *newfile;
392
	int fd = sock_alloc_file(sock, &newfile, flags);
393

394
	if (likely(fd >= 0))
395
		fd_install(fd, newfile);
396

L
Linus Torvalds 已提交
397 398
	return fd;
}
399
EXPORT_SYMBOL(sock_map_fd);
L
Linus Torvalds 已提交
400

401 402 403 404 405
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
406 407
	*err = -ENOTSOCK;
	return NULL;
408 409
}

L
Linus Torvalds 已提交
410
/**
411
 *	sockfd_lookup - Go from a file number to its socket slot
L
Linus Torvalds 已提交
412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427
 *	@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;

428 429
	file = fget(fd);
	if (!file) {
L
Linus Torvalds 已提交
430 431 432
		*err = -EBADF;
		return NULL;
	}
433

434 435
	sock = sock_from_file(file, err);
	if (!sock)
L
Linus Torvalds 已提交
436
		fput(file);
437 438
	return sock;
}
439
EXPORT_SYMBOL(sockfd_lookup);
L
Linus Torvalds 已提交
440

441 442 443 444 445
static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed)
{
	struct file *file;
	struct socket *sock;

446
	*err = -EBADF;
447 448 449 450 451 452
	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
453
	}
454
	return NULL;
L
Linus Torvalds 已提交
455 456 457 458
}

/**
 *	sock_alloc	-	allocate a socket
459
 *
L
Linus Torvalds 已提交
460 461 462 463 464 465 466
 *	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)
{
467 468
	struct inode *inode;
	struct socket *sock;
L
Linus Torvalds 已提交
469

470
	inode = new_inode_pseudo(sock_mnt->mnt_sb);
L
Linus Torvalds 已提交
471 472 473 474 475
	if (!inode)
		return NULL;

	sock = SOCKET_I(inode);

476
	kmemcheck_annotate_bitfield(sock, type);
477
	inode->i_ino = get_next_ino();
478
	inode->i_mode = S_IFSOCK | S_IRWXUGO;
479 480
	inode->i_uid = current_fsuid();
	inode->i_gid = current_fsgid();
L
Linus Torvalds 已提交
481

482
	percpu_add(sockets_in_use, 1);
L
Linus Torvalds 已提交
483 484 485 486 487 488 489 490
	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.
 */
491

L
Linus Torvalds 已提交
492 493 494 495 496
static int sock_no_open(struct inode *irrelevant, struct file *dontcare)
{
	return -ENXIO;
}

497
const struct file_operations bad_sock_fops = {
L
Linus Torvalds 已提交
498 499
	.owner = THIS_MODULE,
	.open = sock_no_open,
500
	.llseek = noop_llseek,
L
Linus Torvalds 已提交
501 502 503 504 505 506 507 508
};

/**
 *	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
509
 *	an inode not a file.
L
Linus Torvalds 已提交
510
 */
511

L
Linus Torvalds 已提交
512 513 514 515 516 517 518 519 520 521
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);
	}

522
	if (rcu_dereference_protected(sock->wq, 1)->fasync_list)
L
Linus Torvalds 已提交
523 524
		printk(KERN_ERR "sock_release: fasync list not empty!\n");

525
	percpu_sub(sockets_in_use, 1);
L
Linus Torvalds 已提交
526 527 528 529
	if (!sock->file) {
		iput(SOCK_INODE(sock));
		return;
	}
530
	sock->file = NULL;
L
Linus Torvalds 已提交
531
}
532
EXPORT_SYMBOL(sock_release);
L
Linus Torvalds 已提交
533

534
int sock_tx_timestamp(struct sock *sk, __u8 *tx_flags)
535
{
536
	*tx_flags = 0;
537
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_HARDWARE))
538
		*tx_flags |= SKBTX_HW_TSTAMP;
539
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_SOFTWARE))
540
		*tx_flags |= SKBTX_SW_TSTAMP;
541 542 543 544
	return 0;
}
EXPORT_SYMBOL(sock_tx_timestamp);

545 546
static inline int __sock_sendmsg_nosec(struct kiocb *iocb, struct socket *sock,
				       struct msghdr *msg, size_t size)
L
Linus Torvalds 已提交
547 548 549
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);

550 551
	sock_update_classid(sock->sk);

L
Linus Torvalds 已提交
552 553 554 555 556 557 558 559
	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;

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

560 561 562 563 564 565 566 567
static inline int __sock_sendmsg(struct kiocb *iocb, struct socket *sock,
				 struct msghdr *msg, size_t size)
{
	int err = security_socket_sendmsg(sock, msg, size);

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

L
Linus Torvalds 已提交
568 569 570 571 572 573 574 575 576 577 578 579 580
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;
}
581
EXPORT_SYMBOL(sock_sendmsg);
L
Linus Torvalds 已提交
582

583
static int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg, size_t size)
584 585 586 587 588 589 590 591 592 593 594 595 596
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

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

L
Linus Torvalds 已提交
597 598 599 600 601 602 603 604 605 606 607
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
	 */
608
	msg->msg_iov = (struct iovec *)vec;
L
Linus Torvalds 已提交
609 610 611 612 613
	msg->msg_iovlen = num;
	result = sock_sendmsg(sock, msg, size);
	set_fs(oldfs);
	return result;
}
614
EXPORT_SYMBOL(kernel_sendmsg);
L
Linus Torvalds 已提交
615

616 617 618 619 620 621 622 623 624 625
static int ktime2ts(ktime_t kt, struct timespec *ts)
{
	if (kt.tv64) {
		*ts = ktime_to_timespec(kt);
		return 1;
	} else {
		return 0;
	}
}

626 627 628 629 630 631
/*
 * 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)
{
632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649
	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 {
650
			skb_get_timestampns(skb, &ts[0]);
651
			put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPNS,
652
				 sizeof(ts[0]), &ts[0]);
653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669
		}
	}


	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;
670
	}
671 672 673
	if (!empty)
		put_cmsg(msg, SOL_SOCKET,
			 SCM_TIMESTAMPING, sizeof(ts), &ts);
674
}
675 676
EXPORT_SYMBOL_GPL(__sock_recv_timestamp);

S
stephen hemminger 已提交
677 678
static inline void sock_recv_drops(struct msghdr *msg, struct sock *sk,
				   struct sk_buff *skb)
679 680 681 682 683 684
{
	if (sock_flag(sk, SOCK_RXQ_OVFL) && skb && skb->dropcount)
		put_cmsg(msg, SOL_SOCKET, SO_RXQ_OVFL,
			sizeof(__u32), &skb->dropcount);
}

685
void __sock_recv_ts_and_drops(struct msghdr *msg, struct sock *sk,
686 687 688 689 690
	struct sk_buff *skb)
{
	sock_recv_timestamp(msg, sk, skb);
	sock_recv_drops(msg, sk, skb);
}
691
EXPORT_SYMBOL_GPL(__sock_recv_ts_and_drops);
692

693 694
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
695 696 697
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);

698 699
	sock_update_classid(sock->sk);

L
Linus Torvalds 已提交
700 701 702 703 704 705 706 707 708
	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;
	si->flags = flags;

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

709 710 711 712 713 714 715 716
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);
}

717
int sock_recvmsg(struct socket *sock, struct msghdr *msg,
L
Linus Torvalds 已提交
718 719 720 721 722 723
		 size_t size, int flags)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

724
	init_sync_kiocb(&iocb, NULL);
L
Linus Torvalds 已提交
725 726 727 728 729 730
	iocb.private = &siocb;
	ret = __sock_recvmsg(&iocb, sock, msg, size, flags);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}
731
EXPORT_SYMBOL(sock_recvmsg);
L
Linus Torvalds 已提交
732

733 734 735 736 737 738 739 740 741 742 743 744 745 746 747
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;
}

748 749 750 751 752 753 754 755 756 757 758 759 760 761 762
/**
 * kernel_recvmsg - Receive a message from a socket (kernel space)
 * @sock:       The socket to receive the message from
 * @msg:        Received message
 * @vec:        Input s/g array for message data
 * @num:        Size of input s/g array
 * @size:       Number of bytes to read
 * @flags:      Message flags (MSG_DONTWAIT, etc...)
 *
 * On return the msg structure contains the scatter/gather array passed in the
 * vec argument. The array is modified so that it consists of the unfilled
 * portion of the original array.
 *
 * The returned value is the total number of bytes received, or an error.
 */
763 764
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
765 766 767 768 769 770 771 772 773
{
	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
	 */
774
	msg->msg_iov = (struct iovec *)vec, msg->msg_iovlen = num;
L
Linus Torvalds 已提交
775 776 777 778
	result = sock_recvmsg(sock, msg, size, flags);
	set_fs(oldfs);
	return result;
}
779
EXPORT_SYMBOL(kernel_recvmsg);
L
Linus Torvalds 已提交
780 781 782 783 784 785

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

786 787
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
788 789 790 791
{
	struct socket *sock;
	int flags;

792 793 794 795 796 797
	sock = file->private_data;

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

798
	return kernel_sendpage(sock, page, offset, size, flags);
799
}
L
Linus Torvalds 已提交
800

J
Jens Axboe 已提交
801
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
802
				struct pipe_inode_info *pipe, size_t len,
J
Jens Axboe 已提交
803 804 805 806
				unsigned int flags)
{
	struct socket *sock = file->private_data;

807 808 809
	if (unlikely(!sock->ops->splice_read))
		return -EINVAL;

810 811
	sock_update_classid(sock->sk);

J
Jens Axboe 已提交
812 813 814
	return sock->ops->splice_read(sock, ppos, pipe, len, flags);
}

815
static struct sock_iocb *alloc_sock_iocb(struct kiocb *iocb,
816
					 struct sock_iocb *siocb)
817 818 819 820 821
{
	if (!is_sync_kiocb(iocb)) {
		siocb = kmalloc(sizeof(*siocb), GFP_KERNEL);
		if (!siocb)
			return NULL;
L
Linus Torvalds 已提交
822 823 824
		iocb->ki_dtor = sock_aio_dtor;
	}

825 826 827
	siocb->kiocb = iocb;
	iocb->private = siocb;
	return siocb;
L
Linus Torvalds 已提交
828 829
}

830
static ssize_t do_sock_read(struct msghdr *msg, struct kiocb *iocb,
831 832
		struct file *file, const struct iovec *iov,
		unsigned long nr_segs)
833 834 835 836
{
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
837

838 839
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
840

841 842 843 844
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
845
	msg->msg_iov = (struct iovec *)iov;
846 847 848 849 850 851
	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);
}

852 853
static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
				unsigned long nr_segs, loff_t pos)
854 855 856
{
	struct sock_iocb siocb, *x;

L
Linus Torvalds 已提交
857 858
	if (pos != 0)
		return -ESPIPE;
859 860

	if (iocb->ki_left == 0)	/* Match SYS5 behaviour */
L
Linus Torvalds 已提交
861 862
		return 0;

863 864

	x = alloc_sock_iocb(iocb, &siocb);
865 866
	if (!x)
		return -ENOMEM;
867
	return do_sock_read(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
868 869
}

870
static ssize_t do_sock_write(struct msghdr *msg, struct kiocb *iocb,
871 872
			struct file *file, const struct iovec *iov,
			unsigned long nr_segs)
L
Linus Torvalds 已提交
873
{
874 875 876
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
877

878 879
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
880

881 882 883 884
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
885
	msg->msg_iov = (struct iovec *)iov;
886 887 888 889
	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
890

891
	return __sock_sendmsg(iocb, sock, msg, size);
L
Linus Torvalds 已提交
892 893
}

894 895
static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov,
			  unsigned long nr_segs, loff_t pos)
896 897
{
	struct sock_iocb siocb, *x;
L
Linus Torvalds 已提交
898

899 900
	if (pos != 0)
		return -ESPIPE;
901 902

	x = alloc_sock_iocb(iocb, &siocb);
903 904
	if (!x)
		return -ENOMEM;
L
Linus Torvalds 已提交
905

906
	return do_sock_write(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
907 908 909 910 911 912 913
}

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

A
Arjan van de Ven 已提交
914
static DEFINE_MUTEX(br_ioctl_mutex);
915
static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg);
L
Linus Torvalds 已提交
916

917
void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *))
L
Linus Torvalds 已提交
918
{
A
Arjan van de Ven 已提交
919
	mutex_lock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
920
	br_ioctl_hook = hook;
A
Arjan van de Ven 已提交
921
	mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
922 923 924
}
EXPORT_SYMBOL(brioctl_set);

A
Arjan van de Ven 已提交
925
static DEFINE_MUTEX(vlan_ioctl_mutex);
926
static int (*vlan_ioctl_hook) (struct net *, void __user *arg);
L
Linus Torvalds 已提交
927

928
void vlan_ioctl_set(int (*hook) (struct net *, void __user *))
L
Linus Torvalds 已提交
929
{
A
Arjan van de Ven 已提交
930
	mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
931
	vlan_ioctl_hook = hook;
A
Arjan van de Ven 已提交
932
	mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
933 934 935
}
EXPORT_SYMBOL(vlan_ioctl_set);

A
Arjan van de Ven 已提交
936
static DEFINE_MUTEX(dlci_ioctl_mutex);
937
static int (*dlci_ioctl_hook) (unsigned int, void __user *);
L
Linus Torvalds 已提交
938

939
void dlci_ioctl_set(int (*hook) (unsigned int, void __user *))
L
Linus Torvalds 已提交
940
{
A
Arjan van de Ven 已提交
941
	mutex_lock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
942
	dlci_ioctl_hook = hook;
A
Arjan van de Ven 已提交
943
	mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
944 945 946
}
EXPORT_SYMBOL(dlci_ioctl_set);

947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964
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
965 966 967 968 969 970 971 972
/*
 *	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;
973
	struct sock *sk;
L
Linus Torvalds 已提交
974 975
	void __user *argp = (void __user *)arg;
	int pid, err;
976
	struct net *net;
L
Linus Torvalds 已提交
977

978
	sock = file->private_data;
979
	sk = sock->sk;
980
	net = sock_net(sk);
L
Linus Torvalds 已提交
981
	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
982
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
983
	} else
J
Johannes Berg 已提交
984
#ifdef CONFIG_WEXT_CORE
L
Linus Torvalds 已提交
985
	if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST) {
986
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
987
	} else
J
Johannes Berg 已提交
988
#endif
989
		switch (cmd) {
L
Linus Torvalds 已提交
990 991 992 993 994 995 996 997 998
		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:
999
			err = put_user(f_getown(sock->file),
1000
				       (int __user *)argp);
L
Linus Torvalds 已提交
1001 1002 1003 1004 1005 1006 1007 1008 1009
			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
1010
			mutex_lock(&br_ioctl_mutex);
1011
			if (br_ioctl_hook)
1012
				err = br_ioctl_hook(net, cmd, argp);
A
Arjan van de Ven 已提交
1013
			mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
1014 1015 1016 1017 1018 1019 1020
			break;
		case SIOCGIFVLAN:
		case SIOCSIFVLAN:
			err = -ENOPKG;
			if (!vlan_ioctl_hook)
				request_module("8021q");

A
Arjan van de Ven 已提交
1021
			mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1022
			if (vlan_ioctl_hook)
1023
				err = vlan_ioctl_hook(net, argp);
A
Arjan van de Ven 已提交
1024
			mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1025 1026 1027 1028 1029 1030 1031
			break;
		case SIOCADDDLCI:
		case SIOCDELDLCI:
			err = -ENOPKG;
			if (!dlci_ioctl_hook)
				request_module("dlci");

1032 1033
			mutex_lock(&dlci_ioctl_mutex);
			if (dlci_ioctl_hook)
L
Linus Torvalds 已提交
1034
				err = dlci_ioctl_hook(cmd, argp);
1035
			mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
1036 1037
			break;
		default:
1038
			err = sock_do_ioctl(net, sock, cmd, arg);
L
Linus Torvalds 已提交
1039
			break;
1040
		}
L
Linus Torvalds 已提交
1041 1042 1043 1044 1045 1046 1047
	return err;
}

int sock_create_lite(int family, int type, int protocol, struct socket **res)
{
	int err;
	struct socket *sock = NULL;
1048

L
Linus Torvalds 已提交
1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059
	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
1060 1061 1062 1063
	err = security_socket_post_create(sock, family, type, protocol, 1);
	if (err)
		goto out_release;

L
Linus Torvalds 已提交
1064 1065 1066
out:
	*res = sock;
	return err;
V
Venkat Yekkirala 已提交
1067 1068 1069 1070
out_release:
	sock_release(sock);
	sock = NULL;
	goto out;
L
Linus Torvalds 已提交
1071
}
1072
EXPORT_SYMBOL(sock_create_lite);
L
Linus Torvalds 已提交
1073 1074

/* No kernel lock held - perfect */
1075
static unsigned int sock_poll(struct file *file, poll_table *wait)
L
Linus Torvalds 已提交
1076 1077 1078 1079
{
	struct socket *sock;

	/*
1080
	 *      We can't return errors to poll, so it's either yes or no.
L
Linus Torvalds 已提交
1081
	 */
1082
	sock = file->private_data;
L
Linus Torvalds 已提交
1083 1084 1085
	return sock->ops->poll(file, sock, wait);
}

1086
static int sock_mmap(struct file *file, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
1087
{
1088
	struct socket *sock = file->private_data;
L
Linus Torvalds 已提交
1089 1090 1091 1092

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

1093
static int sock_close(struct inode *inode, struct file *filp)
L
Linus Torvalds 已提交
1094 1095
{
	/*
1096 1097
	 *      It was possible the inode is NULL we were
	 *      closing an unfinished socket.
L
Linus Torvalds 已提交
1098 1099
	 */

1100
	if (!inode) {
L
Linus Torvalds 已提交
1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115
		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)
1116
 *	   or under socket lock
L
Linus Torvalds 已提交
1117 1118 1119 1120
 */

static int sock_fasync(int fd, struct file *filp, int on)
{
1121 1122
	struct socket *sock = filp->private_data;
	struct sock *sk = sock->sk;
1123
	struct socket_wq *wq;
L
Linus Torvalds 已提交
1124

1125
	if (sk == NULL)
L
Linus Torvalds 已提交
1126 1127 1128
		return -EINVAL;

	lock_sock(sk);
1129 1130
	wq = rcu_dereference_protected(sock->wq, sock_owned_by_user(sk));
	fasync_helper(fd, filp, on, &wq->fasync_list);
L
Linus Torvalds 已提交
1131

1132
	if (!wq->fasync_list)
1133 1134
		sock_reset_flag(sk, SOCK_FASYNC);
	else
E
Eric Dumazet 已提交
1135
		sock_set_flag(sk, SOCK_FASYNC);
L
Linus Torvalds 已提交
1136

1137
	release_sock(sk);
L
Linus Torvalds 已提交
1138 1139 1140
	return 0;
}

1141
/* This function may be called only under socket lock or callback_lock or rcu_lock */
L
Linus Torvalds 已提交
1142 1143 1144

int sock_wake_async(struct socket *sock, int how, int band)
{
1145 1146 1147 1148 1149 1150 1151 1152
	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
1153
		return -1;
1154
	}
1155
	switch (how) {
1156
	case SOCK_WAKE_WAITD:
L
Linus Torvalds 已提交
1157 1158 1159
		if (test_bit(SOCK_ASYNC_WAITDATA, &sock->flags))
			break;
		goto call_kill;
1160
	case SOCK_WAKE_SPACE:
L
Linus Torvalds 已提交
1161 1162 1163
		if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags))
			break;
		/* fall through */
1164
	case SOCK_WAKE_IO:
1165
call_kill:
1166
		kill_fasync(&wq->fasync_list, SIGIO, band);
L
Linus Torvalds 已提交
1167
		break;
1168
	case SOCK_WAKE_URG:
1169
		kill_fasync(&wq->fasync_list, SIGURG, band);
L
Linus Torvalds 已提交
1170
	}
1171
	rcu_read_unlock();
L
Linus Torvalds 已提交
1172 1173
	return 0;
}
1174
EXPORT_SYMBOL(sock_wake_async);
L
Linus Torvalds 已提交
1175

P
Pavel Emelyanov 已提交
1176
int __sock_create(struct net *net, int family, int type, int protocol,
1177
			 struct socket **res, int kern)
L
Linus Torvalds 已提交
1178 1179 1180
{
	int err;
	struct socket *sock;
1181
	const struct net_proto_family *pf;
L
Linus Torvalds 已提交
1182 1183

	/*
1184
	 *      Check protocol is in range
L
Linus Torvalds 已提交
1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196
	 */
	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) {
1197
		static int warned;
L
Linus Torvalds 已提交
1198 1199
		if (!warned) {
			warned = 1;
1200 1201
			printk(KERN_INFO "%s uses obsolete (PF_INET,SOCK_PACKET)\n",
			       current->comm);
L
Linus Torvalds 已提交
1202 1203 1204 1205 1206 1207 1208
		}
		family = PF_PACKET;
	}

	err = security_socket_create(family, type, protocol, kern);
	if (err)
		return err;
1209

1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224
	/*
	 *	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;

1225
#ifdef CONFIG_MODULES
1226 1227 1228
	/* 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
1229 1230 1231
	 * requested real, full-featured networking support upon configuration.
	 * Otherwise module support will break!
	 */
E
Eric Dumazet 已提交
1232
	if (rcu_access_pointer(net_families[family]) == NULL)
1233
		request_module("net-pf-%d", family);
L
Linus Torvalds 已提交
1234 1235
#endif

1236 1237 1238 1239 1240
	rcu_read_lock();
	pf = rcu_dereference(net_families[family]);
	err = -EAFNOSUPPORT;
	if (!pf)
		goto out_release;
L
Linus Torvalds 已提交
1241 1242 1243 1244 1245

	/*
	 * We will call the ->create function, that possibly is in a loadable
	 * module, so we have to bump that loadable module refcnt first.
	 */
1246
	if (!try_module_get(pf->owner))
L
Linus Torvalds 已提交
1247 1248
		goto out_release;

1249 1250 1251
	/* Now protected by module ref count */
	rcu_read_unlock();

1252
	err = pf->create(net, sock, protocol, kern);
1253
	if (err < 0)
L
Linus Torvalds 已提交
1254
		goto out_module_put;
1255

L
Linus Torvalds 已提交
1256 1257 1258 1259
	/*
	 * Now to bump the refcnt of the [loadable] module that owns this
	 * socket at sock_release time we decrement its refcnt.
	 */
1260 1261 1262
	if (!try_module_get(sock->ops->owner))
		goto out_module_busy;

L
Linus Torvalds 已提交
1263 1264 1265 1266
	/*
	 * Now that we're done with the ->create function, the [loadable]
	 * module can have its refcnt decremented
	 */
1267
	module_put(pf->owner);
V
Venkat Yekkirala 已提交
1268 1269
	err = security_socket_post_create(sock, family, type, protocol, kern);
	if (err)
1270
		goto out_sock_release;
1271
	*res = sock;
L
Linus Torvalds 已提交
1272

1273 1274 1275 1276
	return 0;

out_module_busy:
	err = -EAFNOSUPPORT;
L
Linus Torvalds 已提交
1277
out_module_put:
1278 1279 1280
	sock->ops = NULL;
	module_put(pf->owner);
out_sock_release:
L
Linus Torvalds 已提交
1281
	sock_release(sock);
1282 1283 1284 1285 1286
	return err;

out_release:
	rcu_read_unlock();
	goto out_sock_release;
L
Linus Torvalds 已提交
1287
}
P
Pavel Emelyanov 已提交
1288
EXPORT_SYMBOL(__sock_create);
L
Linus Torvalds 已提交
1289 1290 1291

int sock_create(int family, int type, int protocol, struct socket **res)
{
1292
	return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0);
L
Linus Torvalds 已提交
1293
}
1294
EXPORT_SYMBOL(sock_create);
L
Linus Torvalds 已提交
1295 1296 1297

int sock_create_kern(int family, int type, int protocol, struct socket **res)
{
1298
	return __sock_create(&init_net, family, type, protocol, res, 1);
L
Linus Torvalds 已提交
1299
}
1300
EXPORT_SYMBOL(sock_create_kern);
L
Linus Torvalds 已提交
1301

1302
SYSCALL_DEFINE3(socket, int, family, int, type, int, protocol)
L
Linus Torvalds 已提交
1303 1304 1305
{
	int retval;
	struct socket *sock;
1306 1307
	int flags;

1308 1309 1310 1311 1312 1313
	/* 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);

1314
	flags = type & ~SOCK_TYPE_MASK;
1315
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1316 1317
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1318

U
Ulrich Drepper 已提交
1319 1320 1321
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1322 1323 1324 1325
	retval = sock_create(family, type, protocol, &sock);
	if (retval < 0)
		goto out;

1326
	retval = sock_map_fd(sock, flags & (O_CLOEXEC | O_NONBLOCK));
L
Linus Torvalds 已提交
1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342
	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.
 */

1343 1344
SYSCALL_DEFINE4(socketpair, int, family, int, type, int, protocol,
		int __user *, usockvec)
L
Linus Torvalds 已提交
1345 1346 1347
{
	struct socket *sock1, *sock2;
	int fd1, fd2, err;
A
Al Viro 已提交
1348
	struct file *newfile1, *newfile2;
1349 1350 1351
	int flags;

	flags = type & ~SOCK_TYPE_MASK;
1352
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1353 1354
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1355

U
Ulrich Drepper 已提交
1356 1357 1358
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
	/*
	 * 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);
1373
	if (err < 0)
L
Linus Torvalds 已提交
1374 1375
		goto out_release_both;

1376
	fd1 = sock_alloc_file(sock1, &newfile1, flags);
1377 1378
	if (unlikely(fd1 < 0)) {
		err = fd1;
A
Al Viro 已提交
1379
		goto out_release_both;
1380
	}
L
Linus Torvalds 已提交
1381

1382
	fd2 = sock_alloc_file(sock2, &newfile2, flags);
1383 1384 1385 1386 1387 1388
	if (unlikely(fd2 < 0)) {
		err = fd2;
		fput(newfile1);
		put_unused_fd(fd1);
		sock_release(sock2);
		goto out;
A
Al Viro 已提交
1389 1390
	}

A
Al Viro 已提交
1391
	audit_fd_pair(fd1, fd2);
A
Al Viro 已提交
1392 1393
	fd_install(fd1, newfile1);
	fd_install(fd2, newfile2);
L
Linus Torvalds 已提交
1394 1395 1396 1397
	/* fd1 and fd2 may be already another descriptors.
	 * Not kernel problem.
	 */

1398
	err = put_user(fd1, &usockvec[0]);
L
Linus Torvalds 已提交
1399 1400 1401 1402 1403 1404 1405 1406 1407 1408
	if (!err)
		err = put_user(fd2, &usockvec[1]);
	if (!err)
		return 0;

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

out_release_both:
1409
	sock_release(sock2);
L
Linus Torvalds 已提交
1410
out_release_1:
1411
	sock_release(sock1);
L
Linus Torvalds 已提交
1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423
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).
 */

1424
SYSCALL_DEFINE3(bind, int, fd, struct sockaddr __user *, umyaddr, int, addrlen)
L
Linus Torvalds 已提交
1425 1426
{
	struct socket *sock;
1427
	struct sockaddr_storage address;
1428
	int err, fput_needed;
L
Linus Torvalds 已提交
1429

1430
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1431
	if (sock) {
1432
		err = move_addr_to_kernel(umyaddr, addrlen, (struct sockaddr *)&address);
1433 1434
		if (err >= 0) {
			err = security_socket_bind(sock,
1435
						   (struct sockaddr *)&address,
1436
						   addrlen);
1437 1438
			if (!err)
				err = sock->ops->bind(sock,
1439
						      (struct sockaddr *)
1440
						      &address, addrlen);
L
Linus Torvalds 已提交
1441
		}
1442
		fput_light(sock->file, fput_needed);
1443
	}
L
Linus Torvalds 已提交
1444 1445 1446 1447 1448 1449 1450 1451 1452
	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.
 */

1453
SYSCALL_DEFINE2(listen, int, fd, int, backlog)
L
Linus Torvalds 已提交
1454 1455
{
	struct socket *sock;
1456
	int err, fput_needed;
1457
	int somaxconn;
1458 1459 1460

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock) {
1461
		somaxconn = sock_net(sock->sk)->core.sysctl_somaxconn;
1462 1463
		if ((unsigned)backlog > somaxconn)
			backlog = somaxconn;
L
Linus Torvalds 已提交
1464 1465

		err = security_socket_listen(sock, backlog);
1466 1467
		if (!err)
			err = sock->ops->listen(sock, backlog);
L
Linus Torvalds 已提交
1468

1469
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485
	}
	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.
 */

1486 1487
SYSCALL_DEFINE4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen, int, flags)
L
Linus Torvalds 已提交
1488 1489
{
	struct socket *sock, *newsock;
1490
	struct file *newfile;
1491
	int err, len, newfd, fput_needed;
1492
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1493

1494
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
U
Ulrich Drepper 已提交
1495 1496 1497 1498 1499
		return -EINVAL;

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

1500
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1501 1502 1503 1504
	if (!sock)
		goto out;

	err = -ENFILE;
1505 1506
	newsock = sock_alloc();
	if (!newsock)
L
Linus Torvalds 已提交
1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517
		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);

1518
	newfd = sock_alloc_file(newsock, &newfile, flags);
1519 1520
	if (unlikely(newfd < 0)) {
		err = newfd;
1521 1522
		sock_release(newsock);
		goto out_put;
1523 1524
	}

1525 1526
	err = security_socket_accept(sock, newsock);
	if (err)
1527
		goto out_fd;
1528

L
Linus Torvalds 已提交
1529 1530
	err = sock->ops->accept(sock, newsock, sock->file->f_flags);
	if (err < 0)
1531
		goto out_fd;
L
Linus Torvalds 已提交
1532 1533

	if (upeer_sockaddr) {
1534
		if (newsock->ops->getname(newsock, (struct sockaddr *)&address,
1535
					  &len, 2) < 0) {
L
Linus Torvalds 已提交
1536
			err = -ECONNABORTED;
1537
			goto out_fd;
L
Linus Torvalds 已提交
1538
		}
1539 1540
		err = move_addr_to_user((struct sockaddr *)&address,
					len, upeer_sockaddr, upeer_addrlen);
L
Linus Torvalds 已提交
1541
		if (err < 0)
1542
			goto out_fd;
L
Linus Torvalds 已提交
1543 1544 1545 1546
	}

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

1547 1548
	fd_install(newfd, newfile);
	err = newfd;
L
Linus Torvalds 已提交
1549 1550

out_put:
1551
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1552 1553
out:
	return err;
1554
out_fd:
1555
	fput(newfile);
1556
	put_unused_fd(newfd);
L
Linus Torvalds 已提交
1557 1558 1559
	goto out_put;
}

1560 1561
SYSCALL_DEFINE3(accept, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen)
U
Ulrich Drepper 已提交
1562
{
U
Ulrich Drepper 已提交
1563
	return sys_accept4(fd, upeer_sockaddr, upeer_addrlen, 0);
U
Ulrich Drepper 已提交
1564 1565
}

L
Linus Torvalds 已提交
1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577
/*
 *	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.
 */

1578 1579
SYSCALL_DEFINE3(connect, int, fd, struct sockaddr __user *, uservaddr,
		int, addrlen)
L
Linus Torvalds 已提交
1580 1581
{
	struct socket *sock;
1582
	struct sockaddr_storage address;
1583
	int err, fput_needed;
L
Linus Torvalds 已提交
1584

1585
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1586 1587
	if (!sock)
		goto out;
1588
	err = move_addr_to_kernel(uservaddr, addrlen, (struct sockaddr *)&address);
L
Linus Torvalds 已提交
1589 1590 1591
	if (err < 0)
		goto out_put;

1592
	err =
1593
	    security_socket_connect(sock, (struct sockaddr *)&address, addrlen);
L
Linus Torvalds 已提交
1594 1595 1596
	if (err)
		goto out_put;

1597
	err = sock->ops->connect(sock, (struct sockaddr *)&address, addrlen,
L
Linus Torvalds 已提交
1598 1599
				 sock->file->f_flags);
out_put:
1600
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1601 1602 1603 1604 1605 1606 1607 1608 1609
out:
	return err;
}

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

1610 1611
SYSCALL_DEFINE3(getsockname, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1612 1613
{
	struct socket *sock;
1614
	struct sockaddr_storage address;
1615
	int len, err, fput_needed;
1616

1617
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1618 1619 1620 1621 1622 1623 1624
	if (!sock)
		goto out;

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

1625
	err = sock->ops->getname(sock, (struct sockaddr *)&address, &len, 0);
L
Linus Torvalds 已提交
1626 1627
	if (err)
		goto out_put;
1628
	err = move_addr_to_user((struct sockaddr *)&address, len, usockaddr, usockaddr_len);
L
Linus Torvalds 已提交
1629 1630

out_put:
1631
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1632 1633 1634 1635 1636 1637 1638 1639 1640
out:
	return err;
}

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

1641 1642
SYSCALL_DEFINE3(getpeername, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1643 1644
{
	struct socket *sock;
1645
	struct sockaddr_storage address;
1646
	int len, err, fput_needed;
L
Linus Torvalds 已提交
1647

1648 1649
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1650 1651
		err = security_socket_getpeername(sock);
		if (err) {
1652
			fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1653 1654 1655
			return err;
		}

1656
		err =
1657
		    sock->ops->getname(sock, (struct sockaddr *)&address, &len,
1658
				       1);
L
Linus Torvalds 已提交
1659
		if (!err)
1660
			err = move_addr_to_user((struct sockaddr *)&address, len, usockaddr,
1661
						usockaddr_len);
1662
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1663 1664 1665 1666 1667 1668 1669 1670 1671 1672
	}
	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.
 */

1673 1674 1675
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
1676 1677
{
	struct socket *sock;
1678
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1679 1680 1681
	int err;
	struct msghdr msg;
	struct iovec iov;
1682 1683
	int fput_needed;

1684 1685
	if (len > INT_MAX)
		len = INT_MAX;
1686 1687
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (!sock)
1688
		goto out;
1689

1690 1691 1692 1693 1694 1695 1696 1697
	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;
1698
	if (addr) {
1699
		err = move_addr_to_kernel(addr, addr_len, (struct sockaddr *)&address);
L
Linus Torvalds 已提交
1700 1701
		if (err < 0)
			goto out_put;
1702
		msg.msg_name = (struct sockaddr *)&address;
1703
		msg.msg_namelen = addr_len;
L
Linus Torvalds 已提交
1704 1705 1706 1707 1708 1709
	}
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
	msg.msg_flags = flags;
	err = sock_sendmsg(sock, &msg, len);

1710
out_put:
1711
	fput_light(sock->file, fput_needed);
1712
out:
L
Linus Torvalds 已提交
1713 1714 1715 1716
	return err;
}

/*
1717
 *	Send a datagram down a socket.
L
Linus Torvalds 已提交
1718 1719
 */

1720 1721
SYSCALL_DEFINE4(send, int, fd, void __user *, buff, size_t, len,
		unsigned, flags)
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726
{
	return sys_sendto(fd, buff, len, flags, NULL, 0);
}

/*
1727
 *	Receive a frame from the socket and optionally record the address of the
L
Linus Torvalds 已提交
1728 1729 1730 1731
 *	sender. We verify the buffers are writable and if needed move the
 *	sender address from kernel to user space.
 */

1732 1733 1734
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
1735 1736 1737 1738
{
	struct socket *sock;
	struct iovec iov;
	struct msghdr msg;
1739
	struct sockaddr_storage address;
1740
	int err, err2;
1741 1742
	int fput_needed;

1743 1744
	if (size > INT_MAX)
		size = INT_MAX;
1745
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1746
	if (!sock)
1747
		goto out;
L
Linus Torvalds 已提交
1748

1749 1750 1751 1752 1753 1754
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_iovlen = 1;
	msg.msg_iov = &iov;
	iov.iov_len = size;
	iov.iov_base = ubuf;
1755 1756
	msg.msg_name = (struct sockaddr *)&address;
	msg.msg_namelen = sizeof(address);
L
Linus Torvalds 已提交
1757 1758
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
1759
	err = sock_recvmsg(sock, &msg, size, flags);
L
Linus Torvalds 已提交
1760

1761
	if (err >= 0 && addr != NULL) {
1762 1763
		err2 = move_addr_to_user((struct sockaddr *)&address,
					 msg.msg_namelen, addr, addr_len);
1764 1765
		if (err2 < 0)
			err = err2;
L
Linus Torvalds 已提交
1766
	}
1767 1768

	fput_light(sock->file, fput_needed);
1769
out:
L
Linus Torvalds 已提交
1770 1771 1772 1773
	return err;
}

/*
1774
 *	Receive a datagram from a socket.
L
Linus Torvalds 已提交
1775 1776
 */

1777 1778
asmlinkage long sys_recv(int fd, void __user *ubuf, size_t size,
			 unsigned flags)
L
Linus Torvalds 已提交
1779 1780 1781 1782 1783 1784 1785 1786 1787
{
	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.
 */

1788 1789
SYSCALL_DEFINE5(setsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int, optlen)
L
Linus Torvalds 已提交
1790
{
1791
	int err, fput_needed;
L
Linus Torvalds 已提交
1792 1793 1794 1795
	struct socket *sock;

	if (optlen < 0)
		return -EINVAL;
1796 1797 1798 1799

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
		err = security_socket_setsockopt(sock, level, optname);
1800 1801
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1802 1803

		if (level == SOL_SOCKET)
1804 1805 1806
			err =
			    sock_setsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1807
		else
1808 1809 1810
			err =
			    sock->ops->setsockopt(sock, level, optname, optval,
						  optlen);
1811 1812
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1813 1814 1815 1816 1817 1818 1819 1820 1821
	}
	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.
 */

1822 1823
SYSCALL_DEFINE5(getsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int __user *, optlen)
L
Linus Torvalds 已提交
1824
{
1825
	int err, fput_needed;
L
Linus Torvalds 已提交
1826 1827
	struct socket *sock;

1828 1829
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
1830 1831 1832
		err = security_socket_getsockopt(sock, level, optname);
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1833 1834

		if (level == SOL_SOCKET)
1835 1836 1837
			err =
			    sock_getsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1838
		else
1839 1840 1841
			err =
			    sock->ops->getsockopt(sock, level, optname, optval,
						  optlen);
1842 1843
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1844 1845 1846 1847 1848 1849 1850 1851
	}
	return err;
}

/*
 *	Shutdown a socket.
 */

1852
SYSCALL_DEFINE2(shutdown, int, fd, int, how)
L
Linus Torvalds 已提交
1853
{
1854
	int err, fput_needed;
L
Linus Torvalds 已提交
1855 1856
	struct socket *sock;

1857 1858
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1859
		err = security_socket_shutdown(sock, how);
1860 1861 1862
		if (!err)
			err = sock->ops->shutdown(sock, how);
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1863 1864 1865 1866
	}
	return err;
}

1867
/* A couple of helpful macros for getting the address of the 32/64 bit
L
Linus Torvalds 已提交
1868 1869 1870 1871 1872 1873
 * 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)

1874 1875 1876 1877 1878
struct used_address {
	struct sockaddr_storage name;
	unsigned int name_len;
};

1879
static int __sys_sendmsg(struct socket *sock, struct msghdr __user *msg,
1880 1881
			 struct msghdr *msg_sys, unsigned flags,
			 struct used_address *used_address)
L
Linus Torvalds 已提交
1882
{
1883 1884
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
1885
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1886
	struct iovec iovstack[UIO_FASTIOV], *iov = iovstack;
1887
	unsigned char ctl[sizeof(struct cmsghdr) + 20]
1888 1889
	    __attribute__ ((aligned(sizeof(__kernel_size_t))));
	/* 20 is size of ipv6_pktinfo */
L
Linus Torvalds 已提交
1890 1891
	unsigned char *ctl_buf = ctl;
	int err, ctl_len, iov_size, total_len;
1892

L
Linus Torvalds 已提交
1893 1894
	err = -EFAULT;
	if (MSG_CMSG_COMPAT & flags) {
1895
		if (get_compat_msghdr(msg_sys, msg_compat))
L
Linus Torvalds 已提交
1896
			return -EFAULT;
1897
	} else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr)))
L
Linus Torvalds 已提交
1898 1899 1900 1901
		return -EFAULT;

	/* do not move before msg_sys is valid */
	err = -EMSGSIZE;
1902 1903
	if (msg_sys->msg_iovlen > UIO_MAXIOV)
		goto out;
L
Linus Torvalds 已提交
1904

1905
	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
1906
	err = -ENOMEM;
1907 1908
	iov_size = msg_sys->msg_iovlen * sizeof(struct iovec);
	if (msg_sys->msg_iovlen > UIO_FASTIOV) {
L
Linus Torvalds 已提交
1909 1910
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
1911
			goto out;
L
Linus Torvalds 已提交
1912 1913 1914 1915
	}

	/* This will also move the address data into kernel space */
	if (MSG_CMSG_COMPAT & flags) {
1916
		err = verify_compat_iovec(msg_sys, iov,
1917 1918
					  (struct sockaddr *)&address,
					  VERIFY_READ);
L
Linus Torvalds 已提交
1919
	} else
1920
		err = verify_iovec(msg_sys, iov,
1921 1922
				   (struct sockaddr *)&address,
				   VERIFY_READ);
1923
	if (err < 0)
L
Linus Torvalds 已提交
1924 1925 1926 1927 1928
		goto out_freeiov;
	total_len = err;

	err = -ENOBUFS;

1929
	if (msg_sys->msg_controllen > INT_MAX)
L
Linus Torvalds 已提交
1930
		goto out_freeiov;
1931
	ctl_len = msg_sys->msg_controllen;
L
Linus Torvalds 已提交
1932
	if ((MSG_CMSG_COMPAT & flags) && ctl_len) {
1933
		err =
1934
		    cmsghdr_from_user_compat_to_kern(msg_sys, sock->sk, ctl,
1935
						     sizeof(ctl));
L
Linus Torvalds 已提交
1936 1937
		if (err)
			goto out_freeiov;
1938 1939
		ctl_buf = msg_sys->msg_control;
		ctl_len = msg_sys->msg_controllen;
L
Linus Torvalds 已提交
1940
	} else if (ctl_len) {
1941
		if (ctl_len > sizeof(ctl)) {
L
Linus Torvalds 已提交
1942
			ctl_buf = sock_kmalloc(sock->sk, ctl_len, GFP_KERNEL);
1943
			if (ctl_buf == NULL)
L
Linus Torvalds 已提交
1944 1945 1946 1947
				goto out_freeiov;
		}
		err = -EFAULT;
		/*
1948
		 * Careful! Before this, msg_sys->msg_control contains a user pointer.
L
Linus Torvalds 已提交
1949 1950 1951
		 * Afterwards, it will be a kernel pointer. Thus the compiler-assisted
		 * checking falls down on this.
		 */
1952
		if (copy_from_user(ctl_buf,
1953
				   (void __user __force *)msg_sys->msg_control,
1954
				   ctl_len))
L
Linus Torvalds 已提交
1955
			goto out_freectl;
1956
		msg_sys->msg_control = ctl_buf;
L
Linus Torvalds 已提交
1957
	}
1958
	msg_sys->msg_flags = flags;
L
Linus Torvalds 已提交
1959 1960

	if (sock->file->f_flags & O_NONBLOCK)
1961
		msg_sys->msg_flags |= MSG_DONTWAIT;
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983
	/*
	 * If this is sendmmsg() and current destination address is same as
	 * previously succeeded address, omit asking LSM's decision.
	 * used_address->name_len is initialized to UINT_MAX so that the first
	 * destination address never matches.
	 */
	if (used_address && used_address->name_len == msg_sys->msg_namelen &&
	    !memcmp(&used_address->name, msg->msg_name,
		    used_address->name_len)) {
		err = sock_sendmsg_nosec(sock, msg_sys, total_len);
		goto out_freectl;
	}
	err = sock_sendmsg(sock, msg_sys, total_len);
	/*
	 * If this is sendmmsg() and sending to current destination address was
	 * successful, remember it.
	 */
	if (used_address && err >= 0) {
		used_address->name_len = msg_sys->msg_namelen;
		memcpy(&used_address->name, msg->msg_name,
		       used_address->name_len);
	}
L
Linus Torvalds 已提交
1984 1985

out_freectl:
1986
	if (ctl_buf != ctl)
L
Linus Torvalds 已提交
1987 1988 1989 1990
		sock_kfree_s(sock->sk, ctl_buf, ctl_len);
out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
out:
	return err;
}

/*
 *	BSD sendmsg interface
 */

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

	if (!sock)
		goto out;

2008
	err = __sys_sendmsg(sock, msg, &msg_sys, flags, NULL);
2009

2010
	fput_light(sock->file, fput_needed);
2011
out:
L
Linus Torvalds 已提交
2012 2013 2014
	return err;
}

2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026
/*
 *	Linux sendmmsg interface
 */

int __sys_sendmmsg(int fd, struct mmsghdr __user *mmsg, unsigned int vlen,
		   unsigned int flags)
{
	int fput_needed, err, datagrams;
	struct socket *sock;
	struct mmsghdr __user *entry;
	struct compat_mmsghdr __user *compat_entry;
	struct msghdr msg_sys;
2027
	struct used_address used_address;
2028

2029 2030 2031
	if (vlen > UIO_MAXIOV)
		vlen = UIO_MAXIOV;

2032 2033 2034 2035 2036 2037
	datagrams = 0;

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

2038
	used_address.name_len = UINT_MAX;
2039 2040
	entry = mmsg;
	compat_entry = (struct compat_mmsghdr __user *)mmsg;
2041
	err = 0;
2042 2043 2044 2045

	while (datagrams < vlen) {
		if (MSG_CMSG_COMPAT & flags) {
			err = __sys_sendmsg(sock, (struct msghdr __user *)compat_entry,
2046
					    &msg_sys, flags, &used_address);
2047 2048 2049 2050 2051 2052
			if (err < 0)
				break;
			err = __put_user(err, &compat_entry->msg_len);
			++compat_entry;
		} else {
			err = __sys_sendmsg(sock, (struct msghdr __user *)entry,
2053
					    &msg_sys, flags, &used_address);
2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066
			if (err < 0)
				break;
			err = put_user(err, &entry->msg_len);
			++entry;
		}

		if (err)
			break;
		++datagrams;
	}

	fput_light(sock->file, fput_needed);

2067 2068
	/* We only return an error if no datagrams were able to be sent */
	if (datagrams != 0)
2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
		return datagrams;

	return err;
}

SYSCALL_DEFINE4(sendmmsg, int, fd, struct mmsghdr __user *, mmsg,
		unsigned int, vlen, unsigned int, flags)
{
	return __sys_sendmmsg(fd, mmsg, vlen, flags);
}

2080 2081
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
2082
{
2083 2084
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
2085
	struct iovec iovstack[UIO_FASTIOV];
2086
	struct iovec *iov = iovstack;
L
Linus Torvalds 已提交
2087 2088 2089 2090
	unsigned long cmsg_ptr;
	int err, iov_size, total_len, len;

	/* kernel mode address */
2091
	struct sockaddr_storage addr;
L
Linus Torvalds 已提交
2092 2093 2094 2095

	/* user mode address pointers */
	struct sockaddr __user *uaddr;
	int __user *uaddr_len;
2096

L
Linus Torvalds 已提交
2097
	if (MSG_CMSG_COMPAT & flags) {
2098
		if (get_compat_msghdr(msg_sys, msg_compat))
L
Linus Torvalds 已提交
2099
			return -EFAULT;
2100
	} else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr)))
2101
		return -EFAULT;
L
Linus Torvalds 已提交
2102 2103

	err = -EMSGSIZE;
2104 2105
	if (msg_sys->msg_iovlen > UIO_MAXIOV)
		goto out;
2106 2107

	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
2108
	err = -ENOMEM;
2109 2110
	iov_size = msg_sys->msg_iovlen * sizeof(struct iovec);
	if (msg_sys->msg_iovlen > UIO_FASTIOV) {
L
Linus Torvalds 已提交
2111 2112
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
2113
			goto out;
L
Linus Torvalds 已提交
2114 2115 2116
	}

	/*
2117 2118
	 *      Save the user-mode address (verify_iovec will change the
	 *      kernel msghdr to use the kernel address space)
L
Linus Torvalds 已提交
2119
	 */
2120

2121
	uaddr = (__force void __user *)msg_sys->msg_name;
L
Linus Torvalds 已提交
2122 2123
	uaddr_len = COMPAT_NAMELEN(msg);
	if (MSG_CMSG_COMPAT & flags) {
2124
		err = verify_compat_iovec(msg_sys, iov,
2125 2126
					  (struct sockaddr *)&addr,
					  VERIFY_WRITE);
L
Linus Torvalds 已提交
2127
	} else
2128
		err = verify_iovec(msg_sys, iov,
2129 2130
				   (struct sockaddr *)&addr,
				   VERIFY_WRITE);
L
Linus Torvalds 已提交
2131 2132
	if (err < 0)
		goto out_freeiov;
2133
	total_len = err;
L
Linus Torvalds 已提交
2134

2135 2136
	cmsg_ptr = (unsigned long)msg_sys->msg_control;
	msg_sys->msg_flags = flags & (MSG_CMSG_CLOEXEC|MSG_CMSG_COMPAT);
2137

L
Linus Torvalds 已提交
2138 2139
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
2140 2141
	err = (nosec ? sock_recvmsg_nosec : sock_recvmsg)(sock, msg_sys,
							  total_len, flags);
L
Linus Torvalds 已提交
2142 2143 2144 2145 2146
	if (err < 0)
		goto out_freeiov;
	len = err;

	if (uaddr != NULL) {
2147
		err = move_addr_to_user((struct sockaddr *)&addr,
2148
					msg_sys->msg_namelen, uaddr,
2149
					uaddr_len);
L
Linus Torvalds 已提交
2150 2151 2152
		if (err < 0)
			goto out_freeiov;
	}
2153
	err = __put_user((msg_sys->msg_flags & ~MSG_CMSG_COMPAT),
2154
			 COMPAT_FLAGS(msg));
L
Linus Torvalds 已提交
2155 2156 2157
	if (err)
		goto out_freeiov;
	if (MSG_CMSG_COMPAT & flags)
2158
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2159 2160
				 &msg_compat->msg_controllen);
	else
2161
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2162 2163 2164 2165 2166 2167 2168 2169
				 &msg->msg_controllen);
	if (err)
		goto out_freeiov;
	err = len;

out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189
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);

2190
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
2191 2192 2193 2194
out:
	return err;
}

2195 2196 2197 2198 2199 2200 2201 2202 2203 2204
/*
 *     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;
2205
	struct compat_mmsghdr __user *compat_entry;
2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224
	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;
2225
	compat_entry = (struct compat_mmsghdr __user *)mmsg;
2226 2227 2228 2229 2230

	while (datagrams < vlen) {
		/*
		 * No need to ask LSM for more than the first datagram.
		 */
2231 2232
		if (MSG_CMSG_COMPAT & flags) {
			err = __sys_recvmsg(sock, (struct msghdr __user *)compat_entry,
2233 2234
					    &msg_sys, flags & ~MSG_WAITFORONE,
					    datagrams);
2235 2236 2237 2238 2239 2240
			if (err < 0)
				break;
			err = __put_user(err, &compat_entry->msg_len);
			++compat_entry;
		} else {
			err = __sys_recvmsg(sock, (struct msghdr __user *)entry,
2241 2242
					    &msg_sys, flags & ~MSG_WAITFORONE,
					    datagrams);
2243 2244 2245 2246 2247 2248
			if (err < 0)
				break;
			err = put_user(err, &entry->msg_len);
			++entry;
		}

2249 2250 2251 2252
		if (err)
			break;
		++datagrams;

2253 2254 2255 2256
		/* MSG_WAITFORONE turns on MSG_DONTWAIT after one packet */
		if (flags & MSG_WAITFORONE)
			flags |= MSG_DONTWAIT;

2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276
		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
2277

2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324
	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
2325 2326
/* Argument list sizes for sys_socketcall */
#define AL(x) ((x) * sizeof(unsigned long))
2327
static const unsigned char nargs[21] = {
2328 2329 2330
	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),
2331
	AL(4), AL(5), AL(4)
2332 2333
};

L
Linus Torvalds 已提交
2334 2335 2336
#undef AL

/*
2337
 *	System call vectors.
L
Linus Torvalds 已提交
2338 2339 2340
 *
 *	Argument checking cleaned up. Saved 20% in size.
 *  This function doesn't need to set the kernel lock because
2341
 *  it is set by the callees.
L
Linus Torvalds 已提交
2342 2343
 */

2344
SYSCALL_DEFINE2(socketcall, int, call, unsigned long __user *, args)
L
Linus Torvalds 已提交
2345 2346
{
	unsigned long a[6];
2347
	unsigned long a0, a1;
L
Linus Torvalds 已提交
2348
	int err;
2349
	unsigned int len;
L
Linus Torvalds 已提交
2350

2351
	if (call < 1 || call > SYS_SENDMMSG)
L
Linus Torvalds 已提交
2352 2353
		return -EINVAL;

2354 2355 2356 2357
	len = nargs[call];
	if (len > sizeof(a))
		return -EINVAL;

L
Linus Torvalds 已提交
2358
	/* copy_from_user should be SMP safe. */
2359
	if (copy_from_user(a, args, len))
L
Linus Torvalds 已提交
2360
		return -EFAULT;
2361

A
Al Viro 已提交
2362
	audit_socketcall(nargs[call] / sizeof(unsigned long), a);
2363

2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380
	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
2381 2382
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], 0);
2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425
		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;
2426 2427 2428
	case SYS_SENDMMSG:
		err = sys_sendmmsg(a0, (struct mmsghdr __user *)a1, a[2], a[3]);
		break;
2429 2430 2431
	case SYS_RECVMSG:
		err = sys_recvmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
2432 2433 2434 2435
	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
2436 2437 2438
	case SYS_ACCEPT4:
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], a[3]);
U
Ulrich Drepper 已提交
2439
		break;
2440 2441 2442
	default:
		err = -EINVAL;
		break;
L
Linus Torvalds 已提交
2443 2444 2445 2446
	}
	return err;
}

2447
#endif				/* __ARCH_WANT_SYS_SOCKETCALL */
L
Linus Torvalds 已提交
2448

2449 2450 2451 2452
/**
 *	sock_register - add a socket protocol handler
 *	@ops: description of protocol
 *
L
Linus Torvalds 已提交
2453 2454
 *	This function is called by a protocol handler that wants to
 *	advertise its address family, and have it linked into the
2455 2456
 *	socket interface. The value ops->family coresponds to the
 *	socket system call protocol family.
L
Linus Torvalds 已提交
2457
 */
2458
int sock_register(const struct net_proto_family *ops)
L
Linus Torvalds 已提交
2459 2460 2461 2462
{
	int err;

	if (ops->family >= NPROTO) {
2463 2464
		printk(KERN_CRIT "protocol %d >= NPROTO(%d)\n", ops->family,
		       NPROTO);
L
Linus Torvalds 已提交
2465 2466
		return -ENOBUFS;
	}
2467 2468

	spin_lock(&net_family_lock);
E
Eric Dumazet 已提交
2469 2470
	if (rcu_dereference_protected(net_families[ops->family],
				      lockdep_is_held(&net_family_lock)))
2471 2472
		err = -EEXIST;
	else {
E
Eric Dumazet 已提交
2473
		rcu_assign_pointer(net_families[ops->family], ops);
L
Linus Torvalds 已提交
2474 2475
		err = 0;
	}
2476 2477
	spin_unlock(&net_family_lock);

2478
	printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family);
L
Linus Torvalds 已提交
2479 2480
	return err;
}
2481
EXPORT_SYMBOL(sock_register);
L
Linus Torvalds 已提交
2482

2483 2484 2485 2486
/**
 *	sock_unregister - remove a protocol handler
 *	@family: protocol family to remove
 *
L
Linus Torvalds 已提交
2487 2488
 *	This function is called by a protocol handler that wants to
 *	remove its address family, and have it unlinked from the
2489 2490 2491 2492 2493 2494
 *	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
2495
 */
2496
void sock_unregister(int family)
L
Linus Torvalds 已提交
2497
{
2498
	BUG_ON(family < 0 || family >= NPROTO);
L
Linus Torvalds 已提交
2499

2500
	spin_lock(&net_family_lock);
E
Eric Dumazet 已提交
2501
	rcu_assign_pointer(net_families[family], NULL);
2502 2503 2504 2505
	spin_unlock(&net_family_lock);

	synchronize_rcu();

2506
	printk(KERN_INFO "NET: Unregistered protocol family %d\n", family);
L
Linus Torvalds 已提交
2507
}
2508
EXPORT_SYMBOL(sock_unregister);
L
Linus Torvalds 已提交
2509

2510
static int __init sock_init(void)
L
Linus Torvalds 已提交
2511
{
N
Nick Piggin 已提交
2512 2513
	int err;

L
Linus Torvalds 已提交
2514
	/*
2515
	 *      Initialize sock SLAB cache.
L
Linus Torvalds 已提交
2516
	 */
2517

L
Linus Torvalds 已提交
2518 2519 2520
	sk_init();

	/*
2521
	 *      Initialize skbuff SLAB cache
L
Linus Torvalds 已提交
2522 2523 2524 2525
	 */
	skb_init();

	/*
2526
	 *      Initialize the protocols module.
L
Linus Torvalds 已提交
2527 2528 2529
	 */

	init_inodecache();
N
Nick Piggin 已提交
2530 2531 2532 2533

	err = register_filesystem(&sock_fs_type);
	if (err)
		goto out_fs;
L
Linus Torvalds 已提交
2534
	sock_mnt = kern_mount(&sock_fs_type);
N
Nick Piggin 已提交
2535 2536 2537 2538
	if (IS_ERR(sock_mnt)) {
		err = PTR_ERR(sock_mnt);
		goto out_mount;
	}
2539 2540

	/* The real protocol initialization is performed in later initcalls.
L
Linus Torvalds 已提交
2541 2542 2543 2544 2545
	 */

#ifdef CONFIG_NETFILTER
	netfilter_init();
#endif
2546

2547 2548 2549 2550
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING
	skb_timestamping_init();
#endif

N
Nick Piggin 已提交
2551 2552 2553 2554 2555 2556 2557
out:
	return err;

out_mount:
	unregister_filesystem(&sock_fs_type);
out_fs:
	goto out;
L
Linus Torvalds 已提交
2558 2559
}

2560 2561
core_initcall(sock_init);	/* early initcall */

L
Linus Torvalds 已提交
2562 2563 2564 2565 2566 2567
#ifdef CONFIG_PROC_FS
void socket_seq_show(struct seq_file *seq)
{
	int cpu;
	int counter = 0;

2568
	for_each_possible_cpu(cpu)
2569
	    counter += per_cpu(sockets_in_use, cpu);
L
Linus Torvalds 已提交
2570 2571 2572 2573 2574 2575 2576

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

	seq_printf(seq, "sockets: used %d\n", counter);
}
2577
#endif				/* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
2578

2579
#ifdef CONFIG_COMPAT
2580 2581
static int do_siocgstamp(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_timeval __user *up)
2582 2583 2584 2585 2586 2587
{
	mm_segment_t old_fs = get_fs();
	struct timeval ktv;
	int err;

	set_fs(KERNEL_DS);
2588
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&ktv);
2589 2590 2591 2592 2593 2594 2595 2596
	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;
}

2597 2598
static int do_siocgstampns(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_timespec __user *up)
2599 2600 2601 2602 2603 2604
{
	mm_segment_t old_fs = get_fs();
	struct timespec kts;
	int err;

	set_fs(KERNEL_DS);
2605
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&kts);
2606 2607 2608 2609 2610 2611 2612 2613
	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;
}

2614
static int dev_ifname32(struct net *net, struct compat_ifreq __user *uifr32)
2615 2616 2617 2618 2619
{
	struct ifreq __user *uifr;
	int err;

	uifr = compat_alloc_user_space(sizeof(struct ifreq));
2620
	if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq)))
2621 2622
		return -EFAULT;

2623
	err = dev_ioctl(net, SIOCGIFNAME, uifr);
2624 2625 2626
	if (err)
		return err;

2627
	if (copy_in_user(uifr32, uifr, sizeof(struct compat_ifreq)))
2628 2629 2630 2631 2632
		return -EFAULT;

	return 0;
}

2633
static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32)
2634
{
2635
	struct compat_ifconf ifc32;
2636 2637
	struct ifconf ifc;
	struct ifconf __user *uifc;
2638
	struct compat_ifreq __user *ifr32;
2639 2640 2641 2642
	struct ifreq __user *ifr;
	unsigned int i, j;
	int err;

2643
	if (copy_from_user(&ifc32, uifc32, sizeof(struct compat_ifconf)))
2644 2645 2646 2647 2648 2649 2650 2651
		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 {
2652 2653
		size_t len = ((ifc32.ifc_len / sizeof(struct compat_ifreq)) + 1) *
			sizeof(struct ifreq);
2654 2655 2656 2657
		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);
2658
		for (i = 0; i < ifc32.ifc_len; i += sizeof(struct compat_ifreq)) {
2659
			if (copy_in_user(ifr, ifr32, sizeof(struct compat_ifreq)))
2660 2661 2662 2663 2664 2665 2666 2667
				return -EFAULT;
			ifr++;
			ifr32++;
		}
	}
	if (copy_to_user(uifc, &ifc, sizeof(struct ifconf)))
		return -EFAULT;

2668
	err = dev_ioctl(net, SIOCGIFCONF, uifc);
2669 2670 2671 2672 2673 2674 2675 2676 2677
	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;
2678 2679 2680
	     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)))
2681 2682 2683 2684 2685 2686 2687 2688 2689 2690
			return -EFAULT;
		ifr32++;
		ifr++;
	}

	if (ifc32.ifcbuf == 0) {
		/* Translate from 64-bit structure multiple to
		 * a 32-bit one.
		 */
		i = ifc.ifc_len;
2691
		i = ((i / sizeof(struct ifreq)) * sizeof(struct compat_ifreq));
2692 2693 2694 2695
		ifc32.ifc_len = i;
	} else {
		ifc32.ifc_len = i;
	}
2696
	if (copy_to_user(uifc32, &ifc32, sizeof(struct compat_ifconf)))
2697 2698 2699 2700 2701
		return -EFAULT;

	return 0;
}

2702
static int ethtool_ioctl(struct net *net, struct compat_ifreq __user *ifr32)
2703
{
2704 2705 2706 2707
	struct compat_ethtool_rxnfc __user *compat_rxnfc;
	bool convert_in = false, convert_out = false;
	size_t buf_size = ALIGN(sizeof(struct ifreq), 8);
	struct ethtool_rxnfc __user *rxnfc;
2708
	struct ifreq __user *ifr;
2709 2710
	u32 rule_cnt = 0, actual_rule_cnt;
	u32 ethcmd;
2711
	u32 data;
2712
	int ret;
2713

2714 2715
	if (get_user(data, &ifr32->ifr_ifru.ifru_data))
		return -EFAULT;
2716

2717 2718 2719
	compat_rxnfc = compat_ptr(data);

	if (get_user(ethcmd, &compat_rxnfc->cmd))
2720 2721
		return -EFAULT;

2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751
	/* Most ethtool structures are defined without padding.
	 * Unfortunately struct ethtool_rxnfc is an exception.
	 */
	switch (ethcmd) {
	default:
		break;
	case ETHTOOL_GRXCLSRLALL:
		/* Buffer size is variable */
		if (get_user(rule_cnt, &compat_rxnfc->rule_cnt))
			return -EFAULT;
		if (rule_cnt > KMALLOC_MAX_SIZE / sizeof(u32))
			return -ENOMEM;
		buf_size += rule_cnt * sizeof(u32);
		/* fall through */
	case ETHTOOL_GRXRINGS:
	case ETHTOOL_GRXCLSRLCNT:
	case ETHTOOL_GRXCLSRULE:
		convert_out = true;
		/* fall through */
	case ETHTOOL_SRXCLSRLDEL:
	case ETHTOOL_SRXCLSRLINS:
		buf_size += sizeof(struct ethtool_rxnfc);
		convert_in = true;
		break;
	}

	ifr = compat_alloc_user_space(buf_size);
	rxnfc = (void *)ifr + ALIGN(sizeof(struct ifreq), 8);

	if (copy_in_user(&ifr->ifr_name, &ifr32->ifr_name, IFNAMSIZ))
2752 2753
		return -EFAULT;

2754 2755
	if (put_user(convert_in ? rxnfc : compat_ptr(data),
		     &ifr->ifr_ifru.ifru_data))
2756 2757
		return -EFAULT;

2758
	if (convert_in) {
2759
		/* We expect there to be holes between fs.m_ext and
2760 2761
		 * fs.ring_cookie and at the end of fs, but nowhere else.
		 */
2762 2763 2764 2765
		BUILD_BUG_ON(offsetof(struct compat_ethtool_rxnfc, fs.m_ext) +
			     sizeof(compat_rxnfc->fs.m_ext) !=
			     offsetof(struct ethtool_rxnfc, fs.m_ext) +
			     sizeof(rxnfc->fs.m_ext));
2766 2767 2768 2769 2770 2771 2772
		BUILD_BUG_ON(
			offsetof(struct compat_ethtool_rxnfc, fs.location) -
			offsetof(struct compat_ethtool_rxnfc, fs.ring_cookie) !=
			offsetof(struct ethtool_rxnfc, fs.location) -
			offsetof(struct ethtool_rxnfc, fs.ring_cookie));

		if (copy_in_user(rxnfc, compat_rxnfc,
2773
				 (void *)(&rxnfc->fs.m_ext + 1) -
2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789
				 (void *)rxnfc) ||
		    copy_in_user(&rxnfc->fs.ring_cookie,
				 &compat_rxnfc->fs.ring_cookie,
				 (void *)(&rxnfc->fs.location + 1) -
				 (void *)&rxnfc->fs.ring_cookie) ||
		    copy_in_user(&rxnfc->rule_cnt, &compat_rxnfc->rule_cnt,
				 sizeof(rxnfc->rule_cnt)))
			return -EFAULT;
	}

	ret = dev_ioctl(net, SIOCETHTOOL, ifr);
	if (ret)
		return ret;

	if (convert_out) {
		if (copy_in_user(compat_rxnfc, rxnfc,
2790
				 (const void *)(&rxnfc->fs.m_ext + 1) -
2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820
				 (const void *)rxnfc) ||
		    copy_in_user(&compat_rxnfc->fs.ring_cookie,
				 &rxnfc->fs.ring_cookie,
				 (const void *)(&rxnfc->fs.location + 1) -
				 (const void *)&rxnfc->fs.ring_cookie) ||
		    copy_in_user(&compat_rxnfc->rule_cnt, &rxnfc->rule_cnt,
				 sizeof(rxnfc->rule_cnt)))
			return -EFAULT;

		if (ethcmd == ETHTOOL_GRXCLSRLALL) {
			/* As an optimisation, we only copy the actual
			 * number of rules that the underlying
			 * function returned.  Since Mallory might
			 * change the rule count in user memory, we
			 * check that it is less than the rule count
			 * originally given (as the user buffer size),
			 * which has been range-checked.
			 */
			if (get_user(actual_rule_cnt, &rxnfc->rule_cnt))
				return -EFAULT;
			if (actual_rule_cnt < rule_cnt)
				rule_cnt = actual_rule_cnt;
			if (copy_in_user(&compat_rxnfc->rule_locs[0],
					 &rxnfc->rule_locs[0],
					 rule_cnt * sizeof(u32)))
				return -EFAULT;
		}
	}

	return 0;
2821 2822
}

2823 2824 2825 2826 2827 2828
static int compat_siocwandev(struct net *net, struct compat_ifreq __user *uifr32)
{
	void __user *uptr;
	compat_uptr_t uptr32;
	struct ifreq __user *uifr;

2829
	uifr = compat_alloc_user_space(sizeof(*uifr));
2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843
	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);
}

2844 2845
static int bond_ioctl(struct net *net, unsigned int cmd,
			 struct compat_ifreq __user *ifr32)
2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858
{
	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:
2859
		if (copy_from_user(&kifr, ifr32, sizeof(struct compat_ifreq)))
2860 2861 2862
			return -EFAULT;

		old_fs = get_fs();
2863
		set_fs(KERNEL_DS);
2864 2865
		err = dev_ioctl(net, cmd,
				(struct ifreq __user __force *) &kifr);
2866
		set_fs(old_fs);
2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881

		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;

2882
		return dev_ioctl(net, cmd, uifr);
2883 2884
	default:
		return -EINVAL;
2885
	}
2886 2887
}

2888 2889
static int siocdevprivate_ioctl(struct net *net, unsigned int cmd,
				 struct compat_ifreq __user *u_ifreq32)
2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913
{
	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;

2914
	return dev_ioctl(net, cmd, u_ifreq64);
2915 2916
}

2917 2918
static int dev_ifsioc(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_ifreq __user *uifr32)
2919
{
2920
	struct ifreq __user *uifr;
2921 2922
	int err;

2923 2924 2925 2926 2927 2928
	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);

2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940
	if (!err) {
		switch (cmd) {
		case SIOCGIFFLAGS:
		case SIOCGIFMETRIC:
		case SIOCGIFMTU:
		case SIOCGIFMEM:
		case SIOCGIFHWADDR:
		case SIOCGIFINDEX:
		case SIOCGIFADDR:
		case SIOCGIFBRDADDR:
		case SIOCGIFDSTADDR:
		case SIOCGIFNETMASK:
2941
		case SIOCGIFPFLAGS:
2942
		case SIOCGIFTXQLEN:
2943 2944
		case SIOCGMIIPHY:
		case SIOCGMIIREG:
2945
			if (copy_in_user(uifr32, uifr, sizeof(*uifr32)))
2946 2947 2948 2949 2950 2951 2952
				err = -EFAULT;
			break;
		}
	}
	return err;
}

2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972
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();
2973
	set_fs(KERNEL_DS);
2974
	err = dev_ioctl(net, cmd, (void  __user __force *)&ifr);
2975
	set_fs(old_fs);
2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996

	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;

2997
	uifr = compat_alloc_user_space(sizeof(*uifr));
2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011
	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);
}

3012
struct rtentry32 {
3013
	u32		rt_pad1;
3014 3015 3016
	struct sockaddr rt_dst;         /* target address               */
	struct sockaddr rt_gateway;     /* gateway addr (RTF_GATEWAY)   */
	struct sockaddr rt_genmask;     /* target network mask (IP)     */
3017 3018 3019 3020 3021 3022 3023
	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! */
3024
	/* char * */ u32 rt_dev;        /* forcing the device at add    */
3025 3026
	u32		rt_mtu;         /* per route MTU/Window         */
	u32		rt_window;      /* Window clamping              */
3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042
	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;
};

3043 3044
static int routing_ioctl(struct net *net, struct socket *sock,
			 unsigned int cmd, void __user *argp)
3045 3046 3047 3048 3049 3050 3051 3052 3053
{
	int ret;
	void *r = NULL;
	struct in6_rtmsg r6;
	struct rtentry r4;
	char devname[16];
	u32 rtdev;
	mm_segment_t old_fs = get_fs();

3054 3055
	if (sock && sock->sk && sock->sk->sk_family == AF_INET6) { /* ipv6 */
		struct in6_rtmsg32 __user *ur6 = argp;
3056
		ret = copy_from_user(&r6.rtmsg_dst, &(ur6->rtmsg_dst),
3057
			3 * sizeof(struct in6_addr));
3058 3059 3060 3061 3062 3063 3064
		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));
3065 3066 3067

		r = (void *) &r6;
	} else { /* ipv4 */
3068
		struct rtentry32 __user *ur4 = argp;
3069
		ret = copy_from_user(&r4.rt_dst, &(ur4->rt_dst),
3070
					3 * sizeof(struct sockaddr));
3071 3072 3073 3074 3075 3076
		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));
3077
		if (rtdev) {
3078
			ret |= copy_from_user(devname, compat_ptr(rtdev), 15);
3079 3080
			r4.rt_dev = (char __user __force *)devname;
			devname[15] = 0;
3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091
		} else
			r4.rt_dev = NULL;

		r = (void *) &r4;
	}

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

3092
	set_fs(KERNEL_DS);
3093
	ret = sock_do_ioctl(net, sock, cmd, (unsigned long) r);
3094
	set_fs(old_fs);
3095 3096 3097 3098 3099 3100 3101

out:
	return ret;
}

/* Since old style bridge ioctl's endup using SIOCDEVPRIVATE
 * for some operations; this forces use of the newer bridge-utils that
L
Lucas De Marchi 已提交
3102
 * use compatible ioctls
3103
 */
3104
static int old_bridge_ioctl(compat_ulong_t __user *argp)
3105
{
3106
	compat_ulong_t tmp;
3107

3108
	if (get_user(tmp, argp))
3109 3110 3111 3112 3113 3114
		return -EFAULT;
	if (tmp == BRCTL_GET_VERSION)
		return BRCTL_VERSION + 1;
	return -EINVAL;
}

3115 3116 3117 3118 3119 3120
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);
3121

3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134
	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);
3135 3136
	case SIOCWANDEV:
		return compat_siocwandev(net, argp);
3137 3138 3139
	case SIOCGIFMAP:
	case SIOCSIFMAP:
		return compat_sioc_ifmap(net, cmd, argp);
3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153
	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);
3154 3155
	case SIOCSHWTSTAMP:
		return compat_siocshwtstamp(net, argp);
3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197

	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:
3198 3199 3200 3201
	case SIOCSIFNAME:
	case SIOCGMIIPHY:
	case SIOCGMIIREG:
	case SIOCSMIIREG:
3202
		return dev_ifsioc(net, sock, cmd, argp);
3203

3204 3205 3206 3207
	case SIOCSARP:
	case SIOCGARP:
	case SIOCDARP:
	case SIOCATMARK:
3208 3209 3210 3211 3212 3213 3214 3215
		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:
3216 3217 3218
	case SIOCSRARP:
	case SIOCGRARP:
	case SIOCDRARP:
3219 3220 3221 3222
	case SIOCSIFLINK:
	case SIOCGIFSLAVE:
	case SIOCSIFSLAVE:
		return -EINVAL;
3223 3224 3225 3226
	}

	return -ENOIOCTLCMD;
}
3227

3228
static long compat_sock_ioctl(struct file *file, unsigned cmd,
3229
			      unsigned long arg)
3230 3231 3232
{
	struct socket *sock = file->private_data;
	int ret = -ENOIOCTLCMD;
3233 3234 3235 3236 3237
	struct sock *sk;
	struct net *net;

	sk = sock->sk;
	net = sock_net(sk);
3238 3239 3240 3241

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

3242 3243 3244 3245
	if (ret == -ENOIOCTLCMD &&
	    (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST))
		ret = compat_wext_handle_ioctl(net, cmd, arg);

3246 3247 3248
	if (ret == -ENOIOCTLCMD)
		ret = compat_sock_ioctl_trans(file, sock, cmd, arg);

3249 3250 3251 3252
	return ret;
}
#endif

3253 3254 3255 3256
int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen)
{
	return sock->ops->bind(sock, addr, addrlen);
}
3257
EXPORT_SYMBOL(kernel_bind);
3258 3259 3260 3261 3262

int kernel_listen(struct socket *sock, int backlog)
{
	return sock->ops->listen(sock, backlog);
}
3263
EXPORT_SYMBOL(kernel_listen);
3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277

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);
3278
		*newsock = NULL;
3279 3280 3281 3282
		goto done;
	}

	(*newsock)->ops = sock->ops;
3283
	__module_get((*newsock)->ops->owner);
3284 3285 3286 3287

done:
	return err;
}
3288
EXPORT_SYMBOL(kernel_accept);
3289 3290

int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen,
3291
		   int flags)
3292 3293 3294
{
	return sock->ops->connect(sock, addr, addrlen, flags);
}
3295
EXPORT_SYMBOL(kernel_connect);
3296 3297 3298 3299 3300 3301

int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 0);
}
3302
EXPORT_SYMBOL(kernel_getsockname);
3303 3304 3305 3306 3307 3308

int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 1);
}
3309
EXPORT_SYMBOL(kernel_getpeername);
3310 3311 3312 3313 3314

int kernel_getsockopt(struct socket *sock, int level, int optname,
			char *optval, int *optlen)
{
	mm_segment_t oldfs = get_fs();
3315 3316
	char __user *uoptval;
	int __user *uoptlen;
3317 3318
	int err;

3319 3320 3321
	uoptval = (char __user __force *) optval;
	uoptlen = (int __user __force *) optlen;

3322 3323
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3324
		err = sock_getsockopt(sock, level, optname, uoptval, uoptlen);
3325
	else
3326 3327
		err = sock->ops->getsockopt(sock, level, optname, uoptval,
					    uoptlen);
3328 3329 3330
	set_fs(oldfs);
	return err;
}
3331
EXPORT_SYMBOL(kernel_getsockopt);
3332 3333

int kernel_setsockopt(struct socket *sock, int level, int optname,
3334
			char *optval, unsigned int optlen)
3335 3336
{
	mm_segment_t oldfs = get_fs();
3337
	char __user *uoptval;
3338 3339
	int err;

3340 3341
	uoptval = (char __user __force *) optval;

3342 3343
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3344
		err = sock_setsockopt(sock, level, optname, uoptval, optlen);
3345
	else
3346
		err = sock->ops->setsockopt(sock, level, optname, uoptval,
3347 3348 3349 3350
					    optlen);
	set_fs(oldfs);
	return err;
}
3351
EXPORT_SYMBOL(kernel_setsockopt);
3352 3353 3354 3355

int kernel_sendpage(struct socket *sock, struct page *page, int offset,
		    size_t size, int flags)
{
3356 3357
	sock_update_classid(sock->sk);

3358 3359 3360 3361 3362
	if (sock->ops->sendpage)
		return sock->ops->sendpage(sock, page, offset, size, flags);

	return sock_no_sendpage(sock, page, offset, size, flags);
}
3363
EXPORT_SYMBOL(kernel_sendpage);
3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375

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;
}
3376
EXPORT_SYMBOL(kernel_sock_ioctl);
3377

3378 3379 3380 3381 3382
int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how)
{
	return sock->ops->shutdown(sock, how);
}
EXPORT_SYMBOL(kernel_sock_shutdown);