socket.c 55.0 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>
66
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
67 68 69
#include <linux/netdevice.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
A
Arjan van de Ven 已提交
70
#include <linux/mutex.h>
L
Linus Torvalds 已提交
71 72
#include <linux/wanrouter.h>
#include <linux/if_bridge.h>
73 74
#include <linux/if_frad.h>
#include <linux/if_vlan.h>
L
Linus Torvalds 已提交
75 76 77 78 79 80 81 82 83 84
#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>
85
#include <linux/audit.h>
86
#include <linux/wireless.h>
87
#include <linux/nsproxy.h>
L
Linus Torvalds 已提交
88 89 90 91 92 93 94 95 96 97

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

#include <net/compat.h>

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

static int sock_no_open(struct inode *irrelevant, struct file *dontcare);
98 99 100 101
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);
102
static int sock_mmap(struct file *file, struct vm_area_struct *vma);
L
Linus Torvalds 已提交
103 104 105 106

static int sock_close(struct inode *inode, struct file *file);
static unsigned int sock_poll(struct file *file,
			      struct poll_table_struct *wait);
107
static long sock_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
108 109
#ifdef CONFIG_COMPAT
static long compat_sock_ioctl(struct file *file,
110
			      unsigned int cmd, unsigned long arg);
111
#endif
L
Linus Torvalds 已提交
112 113 114
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
115 116 117
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
			        struct pipe_inode_info *pipe, size_t len,
				unsigned int flags);
L
Linus Torvalds 已提交
118 119 120 121 122 123

/*
 *	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.
 */

124
static const struct file_operations socket_file_ops = {
L
Linus Torvalds 已提交
125 126 127 128 129 130
	.owner =	THIS_MODULE,
	.llseek =	no_llseek,
	.aio_read =	sock_aio_read,
	.aio_write =	sock_aio_write,
	.poll =		sock_poll,
	.unlocked_ioctl = sock_ioctl,
131 132 133
#ifdef CONFIG_COMPAT
	.compat_ioctl = compat_sock_ioctl,
#endif
L
Linus Torvalds 已提交
134 135 136 137
	.mmap =		sock_mmap,
	.open =		sock_no_open,	/* special open code to disallow open via /proc */
	.release =	sock_close,
	.fasync =	sock_fasync,
138 139
	.sendpage =	sock_sendpage,
	.splice_write = generic_splice_sendpage,
J
Jens Axboe 已提交
140
	.splice_read =	sock_splice_read,
L
Linus Torvalds 已提交
141 142 143 144 145 146 147
};

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

static DEFINE_SPINLOCK(net_family_lock);
148
static const struct net_proto_family *net_families[NPROTO] __read_mostly;
L
Linus Torvalds 已提交
149 150 151 152 153 154 155 156

/*
 *	Statistics counters of the socket lists
 */

static DEFINE_PER_CPU(int, sockets_in_use) = 0;

/*
157 158 159
 * 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
160 161
 */

162
#define MAX_SOCK_ADDR	128		/* 108 for Unix domain -
L
Linus Torvalds 已提交
163 164
					   16 for IP, 16 for IPX,
					   24 for IPv6,
165
					   about 80 for AX.25
L
Linus Torvalds 已提交
166 167
					   must be at least one bigger than
					   the AF_UNIX size (see net/unix/af_unix.c
168
					   :unix_mkname()).
L
Linus Torvalds 已提交
169
					 */
170

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.
 */

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

/**
 *	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.
 */
209 210 211

int move_addr_to_user(void *kaddr, int klen, void __user *uaddr,
		      int __user *ulen)
L
Linus Torvalds 已提交
212 213 214 215
{
	int err;
	int len;

216 217
	err = get_user(len, ulen);
	if (err)
L
Linus Torvalds 已提交
218
		return err;
219 220 221
	if (len > klen)
		len = klen;
	if (len < 0 || len > MAX_SOCK_ADDR)
L
Linus Torvalds 已提交
222
		return -EINVAL;
223
	if (len) {
S
Steve Grubb 已提交
224 225
		if (audit_sockaddr(klen, kaddr))
			return -ENOMEM;
226
		if (copy_to_user(uaddr, kaddr, len))
L
Linus Torvalds 已提交
227 228 229
			return -EFAULT;
	}
	/*
230 231
	 *      "fromlen shall refer to the value before truncation.."
	 *                      1003.1g
L
Linus Torvalds 已提交
232 233 234 235 236 237
	 */
	return __put_user(klen, ulen);
}

#define SOCKFS_MAGIC 0x534F434B

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

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

244
	ei = kmem_cache_alloc(sock_inode_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
245 246 247
	if (!ei)
		return NULL;
	init_waitqueue_head(&ei->socket.wait);
248

L
Linus Torvalds 已提交
249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264
	ei->socket.fasync_list = NULL;
	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)
{
	kmem_cache_free(sock_inode_cachep,
			container_of(inode, struct socket_alloc, vfs_inode));
}

265
static void init_once(struct kmem_cache *cachep, void *foo)
L
Linus Torvalds 已提交
266
{
267
	struct socket_alloc *ei = (struct socket_alloc *)foo;
L
Linus Torvalds 已提交
268

C
Christoph Lameter 已提交
269
	inode_init_once(&ei->vfs_inode);
L
Linus Torvalds 已提交
270
}
271

L
Linus Torvalds 已提交
272 273 274
static int init_inodecache(void)
{
	sock_inode_cachep = kmem_cache_create("sock_inode_cache",
275 276 277 278 279
					      sizeof(struct socket_alloc),
					      0,
					      (SLAB_HWCACHE_ALIGN |
					       SLAB_RECLAIM_ACCOUNT |
					       SLAB_MEM_SPREAD),
280
					      init_once);
L
Linus Torvalds 已提交
281 282 283 284 285 286 287 288 289 290 291
	if (sock_inode_cachep == NULL)
		return -ENOMEM;
	return 0;
}

static struct super_operations sockfs_ops = {
	.alloc_inode =	sock_alloc_inode,
	.destroy_inode =sock_destroy_inode,
	.statfs =	simple_statfs,
};

292
static int sockfs_get_sb(struct file_system_type *fs_type,
293 294
			 int flags, const char *dev_name, void *data,
			 struct vfsmount *mnt)
L
Linus Torvalds 已提交
295
{
296 297
	return get_sb_pseudo(fs_type, "socket:", &sockfs_ops, SOCKFS_MAGIC,
			     mnt);
L
Linus Torvalds 已提交
298 299
}

300
static struct vfsmount *sock_mnt __read_mostly;
L
Linus Torvalds 已提交
301 302 303 304 305 306

static struct file_system_type sock_fs_type = {
	.name =		"sockfs",
	.get_sb =	sockfs_get_sb,
	.kill_sb =	kill_anon_super,
};
307

L
Linus Torvalds 已提交
308 309
static int sockfs_delete_dentry(struct dentry *dentry)
{
310 311 312 313 314 315 316 317
	/*
	 * At creation time, we pretended this dentry was hashed
	 * (by clearing DCACHE_UNHASHED bit in d_flags)
	 * At delete time, we restore the truth : not hashed.
	 * (so that dput() can proceed correctly)
	 */
	dentry->d_flags |= DCACHE_UNHASHED;
	return 0;
L
Linus Torvalds 已提交
318
}
319 320 321 322 323 324 325 326 327 328

/*
 * 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);
}

L
Linus Torvalds 已提交
329
static struct dentry_operations sockfs_dentry_operations = {
330
	.d_delete = sockfs_delete_dentry,
331
	.d_dname  = sockfs_dname,
L
Linus Torvalds 已提交
332 333 334 335 336
};

/*
 *	Obtains the first available file descriptor and sets it up for use.
 *
337 338
 *	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
339 340 341 342 343 344 345 346 347 348 349 350
 *	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.
 */

351
static int sock_alloc_fd(struct file **filep)
L
Linus Torvalds 已提交
352 353 354 355
{
	int fd;

	fd = get_unused_fd();
356
	if (likely(fd >= 0)) {
L
Linus Torvalds 已提交
357 358
		struct file *file = get_empty_filp();

359 360
		*filep = file;
		if (unlikely(!file)) {
L
Linus Torvalds 已提交
361
			put_unused_fd(fd);
362
			return -ENFILE;
L
Linus Torvalds 已提交
363
		}
364 365 366 367
	} else
		*filep = NULL;
	return fd;
}
L
Linus Torvalds 已提交
368

369 370
static int sock_attach_fd(struct socket *sock, struct file *file)
{
371
	struct dentry *dentry;
372
	struct qstr name = { .name = "" };
373

374 375
	dentry = d_alloc(sock_mnt->mnt_sb->s_root, &name);
	if (unlikely(!dentry))
376 377
		return -ENOMEM;

378
	dentry->d_op = &sockfs_dentry_operations;
379 380 381 382 383
	/*
	 * We dont want to push this dentry into global dentry hash table.
	 * We pretend dentry is already hashed, by unsetting DCACHE_UNHASHED
	 * This permits a working /proc/$pid/fd/XXX on sockets
	 */
384 385
	dentry->d_flags &= ~DCACHE_UNHASHED;
	d_instantiate(dentry, SOCK_INODE(sock));
386 387

	sock->file = file;
388 389 390
	init_file(file, sock_mnt, dentry, FMODE_READ | FMODE_WRITE,
		  &socket_file_ops);
	SOCK_INODE(sock)->i_fop = &socket_file_ops;
391 392 393
	file->f_flags = O_RDWR;
	file->f_pos = 0;
	file->private_data = sock;
L
Linus Torvalds 已提交
394

395 396 397 398 399 400 401 402 403 404 405 406 407
	return 0;
}

int sock_map_fd(struct socket *sock)
{
	struct file *newfile;
	int fd = sock_alloc_fd(&newfile);

	if (likely(fd >= 0)) {
		int err = sock_attach_fd(sock, newfile);

		if (unlikely(err < 0)) {
			put_filp(newfile);
L
Linus Torvalds 已提交
408
			put_unused_fd(fd);
409
			return err;
L
Linus Torvalds 已提交
410
		}
411
		fd_install(fd, newfile);
L
Linus Torvalds 已提交
412 413 414 415
	}
	return fd;
}

416 417 418 419 420
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
421 422
	*err = -ENOTSOCK;
	return NULL;
423 424
}

L
Linus Torvalds 已提交
425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442
/**
 *	sockfd_lookup	- 	Go from a file number to its socket slot
 *	@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;

443 444
	file = fget(fd);
	if (!file) {
L
Linus Torvalds 已提交
445 446 447
		*err = -EBADF;
		return NULL;
	}
448

449 450
	sock = sock_from_file(file, err);
	if (!sock)
L
Linus Torvalds 已提交
451
		fput(file);
452 453
	return sock;
}
L
Linus Torvalds 已提交
454

455 456 457 458 459
static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed)
{
	struct file *file;
	struct socket *sock;

460
	*err = -EBADF;
461 462 463 464 465 466
	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
467
	}
468
	return NULL;
L
Linus Torvalds 已提交
469 470 471 472
}

/**
 *	sock_alloc	-	allocate a socket
473
 *
L
Linus Torvalds 已提交
474 475 476 477 478 479 480
 *	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)
{
481 482
	struct inode *inode;
	struct socket *sock;
L
Linus Torvalds 已提交
483 484 485 486 487 488 489

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

	sock = SOCKET_I(inode);

490
	inode->i_mode = S_IFSOCK | S_IRWXUGO;
L
Linus Torvalds 已提交
491 492 493 494 495 496 497 498 499 500 501 502 503
	inode->i_uid = current->fsuid;
	inode->i_gid = current->fsgid;

	get_cpu_var(sockets_in_use)++;
	put_cpu_var(sockets_in_use);
	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.
 */
504

L
Linus Torvalds 已提交
505 506 507 508 509
static int sock_no_open(struct inode *irrelevant, struct file *dontcare)
{
	return -ENXIO;
}

510
const struct file_operations bad_sock_fops = {
L
Linus Torvalds 已提交
511 512 513 514 515 516 517 518 519 520
	.owner = THIS_MODULE,
	.open = sock_no_open,
};

/**
 *	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
521
 *	an inode not a file.
L
Linus Torvalds 已提交
522
 */
523

L
Linus Torvalds 已提交
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542
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);
	}

	if (sock->fasync_list)
		printk(KERN_ERR "sock_release: fasync list not empty!\n");

	get_cpu_var(sockets_in_use)--;
	put_cpu_var(sockets_in_use);
	if (!sock->file) {
		iput(SOCK_INODE(sock));
		return;
	}
543
	sock->file = NULL;
L
Linus Torvalds 已提交
544 545
}

546
static inline int __sock_sendmsg(struct kiocb *iocb, struct socket *sock,
L
Linus Torvalds 已提交
547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588
				 struct msghdr *msg, size_t size)
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);
	int err;

	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;

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

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

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

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

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
	 */
589
	msg->msg_iov = (struct iovec *)vec;
L
Linus Torvalds 已提交
590 591 592 593 594 595
	msg->msg_iovlen = num;
	result = sock_sendmsg(sock, msg, size);
	set_fs(oldfs);
	return result;
}

596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624
/*
 * 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)
{
	ktime_t kt = skb->tstamp;

	if (!sock_flag(sk, SOCK_RCVTSTAMPNS)) {
		struct timeval tv;
		/* Race occurred between timestamp enabling and packet
		   receiving.  Fill in the current time for now. */
		if (kt.tv64 == 0)
			kt = ktime_get_real();
		skb->tstamp = kt;
		tv = ktime_to_timeval(kt);
		put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMP, sizeof(tv), &tv);
	} else {
		struct timespec ts;
		/* Race occurred between timestamp enabling and packet
		   receiving.  Fill in the current time for now. */
		if (kt.tv64 == 0)
			kt = ktime_get_real();
		skb->tstamp = kt;
		ts = ktime_to_timespec(kt);
		put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPNS, sizeof(ts), &ts);
	}
}

625 626
EXPORT_SYMBOL_GPL(__sock_recv_timestamp);

627
static inline int __sock_recvmsg(struct kiocb *iocb, struct socket *sock,
L
Linus Torvalds 已提交
628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645
				 struct msghdr *msg, size_t size, int flags)
{
	int err;
	struct sock_iocb *si = kiocb_to_siocb(iocb);

	si->sock = sock;
	si->scm = NULL;
	si->msg = msg;
	si->size = size;
	si->flags = flags;

	err = security_socket_recvmsg(sock, msg, size, flags);
	if (err)
		return err;

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

646
int sock_recvmsg(struct socket *sock, struct msghdr *msg,
L
Linus Torvalds 已提交
647 648 649 650 651 652
		 size_t size, int flags)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

653
	init_sync_kiocb(&iocb, NULL);
L
Linus Torvalds 已提交
654 655 656 657 658 659 660
	iocb.private = &siocb;
	ret = __sock_recvmsg(&iocb, sock, msg, size, flags);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}

661 662
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
663 664 665 666 667 668 669 670 671
{
	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
	 */
672
	msg->msg_iov = (struct iovec *)vec, msg->msg_iovlen = num;
L
Linus Torvalds 已提交
673 674 675 676 677 678 679 680 681 682
	result = sock_recvmsg(sock, msg, size, flags);
	set_fs(oldfs);
	return result;
}

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

683 684
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
685 686 687 688
{
	struct socket *sock;
	int flags;

689 690 691 692 693 694 695 696
	sock = file->private_data;

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

	return sock->ops->sendpage(sock, page, offset, size, flags);
}
L
Linus Torvalds 已提交
697

J
Jens Axboe 已提交
698 699 700 701 702 703
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
			        struct pipe_inode_info *pipe, size_t len,
				unsigned int flags)
{
	struct socket *sock = file->private_data;

704 705 706
	if (unlikely(!sock->ops->splice_read))
		return -EINVAL;

J
Jens Axboe 已提交
707 708 709
	return sock->ops->splice_read(sock, ppos, pipe, len, flags);
}

710
static struct sock_iocb *alloc_sock_iocb(struct kiocb *iocb,
711
					 struct sock_iocb *siocb)
712 713 714 715 716
{
	if (!is_sync_kiocb(iocb)) {
		siocb = kmalloc(sizeof(*siocb), GFP_KERNEL);
		if (!siocb)
			return NULL;
L
Linus Torvalds 已提交
717 718 719
		iocb->ki_dtor = sock_aio_dtor;
	}

720 721 722
	siocb->kiocb = iocb;
	iocb->private = siocb;
	return siocb;
L
Linus Torvalds 已提交
723 724
}

725
static ssize_t do_sock_read(struct msghdr *msg, struct kiocb *iocb,
726 727
		struct file *file, const struct iovec *iov,
		unsigned long nr_segs)
728 729 730 731
{
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
732

733 734
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
735

736 737 738 739
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
740
	msg->msg_iov = (struct iovec *)iov;
741 742 743 744 745 746
	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);
}

747 748
static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
				unsigned long nr_segs, loff_t pos)
749 750 751
{
	struct sock_iocb siocb, *x;

L
Linus Torvalds 已提交
752 753
	if (pos != 0)
		return -ESPIPE;
754 755

	if (iocb->ki_left == 0)	/* Match SYS5 behaviour */
L
Linus Torvalds 已提交
756 757
		return 0;

758 759

	x = alloc_sock_iocb(iocb, &siocb);
760 761
	if (!x)
		return -ENOMEM;
762
	return do_sock_read(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
763 764
}

765
static ssize_t do_sock_write(struct msghdr *msg, struct kiocb *iocb,
766 767
			struct file *file, const struct iovec *iov,
			unsigned long nr_segs)
L
Linus Torvalds 已提交
768
{
769 770 771
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
772

773 774
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
775

776 777 778 779
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
780
	msg->msg_iov = (struct iovec *)iov;
781 782 783 784
	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
785

786
	return __sock_sendmsg(iocb, sock, msg, size);
L
Linus Torvalds 已提交
787 788
}

789 790
static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov,
			  unsigned long nr_segs, loff_t pos)
791 792
{
	struct sock_iocb siocb, *x;
L
Linus Torvalds 已提交
793

794 795
	if (pos != 0)
		return -ESPIPE;
796 797

	x = alloc_sock_iocb(iocb, &siocb);
798 799
	if (!x)
		return -ENOMEM;
L
Linus Torvalds 已提交
800

801
	return do_sock_write(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
802 803 804 805 806 807 808
}

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

A
Arjan van de Ven 已提交
809
static DEFINE_MUTEX(br_ioctl_mutex);
810
static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg) = NULL;
L
Linus Torvalds 已提交
811

812
void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *))
L
Linus Torvalds 已提交
813
{
A
Arjan van de Ven 已提交
814
	mutex_lock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
815
	br_ioctl_hook = hook;
A
Arjan van de Ven 已提交
816
	mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
817
}
818

L
Linus Torvalds 已提交
819 820
EXPORT_SYMBOL(brioctl_set);

A
Arjan van de Ven 已提交
821
static DEFINE_MUTEX(vlan_ioctl_mutex);
822
static int (*vlan_ioctl_hook) (struct net *, void __user *arg);
L
Linus Torvalds 已提交
823

824
void vlan_ioctl_set(int (*hook) (struct net *, void __user *))
L
Linus Torvalds 已提交
825
{
A
Arjan van de Ven 已提交
826
	mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
827
	vlan_ioctl_hook = hook;
A
Arjan van de Ven 已提交
828
	mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
829
}
830

L
Linus Torvalds 已提交
831 832
EXPORT_SYMBOL(vlan_ioctl_set);

A
Arjan van de Ven 已提交
833
static DEFINE_MUTEX(dlci_ioctl_mutex);
834
static int (*dlci_ioctl_hook) (unsigned int, void __user *);
L
Linus Torvalds 已提交
835

836
void dlci_ioctl_set(int (*hook) (unsigned int, void __user *))
L
Linus Torvalds 已提交
837
{
A
Arjan van de Ven 已提交
838
	mutex_lock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
839
	dlci_ioctl_hook = hook;
A
Arjan van de Ven 已提交
840
	mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
841
}
842

L
Linus Torvalds 已提交
843 844 845 846 847 848 849 850 851 852
EXPORT_SYMBOL(dlci_ioctl_set);

/*
 *	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;
853
	struct sock *sk;
L
Linus Torvalds 已提交
854 855
	void __user *argp = (void __user *)arg;
	int pid, err;
856
	struct net *net;
L
Linus Torvalds 已提交
857

858
	sock = file->private_data;
859
	sk = sock->sk;
860
	net = sock_net(sk);
L
Linus Torvalds 已提交
861
	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
862
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
863
	} else
864
#ifdef CONFIG_WIRELESS_EXT
L
Linus Torvalds 已提交
865
	if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST) {
866
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
867
	} else
868 869
#endif				/* CONFIG_WIRELESS_EXT */
		switch (cmd) {
L
Linus Torvalds 已提交
870 871 872 873 874 875 876 877 878
		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:
879
			err = put_user(f_getown(sock->file),
880
				       (int __user *)argp);
L
Linus Torvalds 已提交
881 882 883 884 885 886 887 888 889
			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
890
			mutex_lock(&br_ioctl_mutex);
891
			if (br_ioctl_hook)
892
				err = br_ioctl_hook(net, cmd, argp);
A
Arjan van de Ven 已提交
893
			mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
894 895 896 897 898 899 900
			break;
		case SIOCGIFVLAN:
		case SIOCSIFVLAN:
			err = -ENOPKG;
			if (!vlan_ioctl_hook)
				request_module("8021q");

A
Arjan van de Ven 已提交
901
			mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
902
			if (vlan_ioctl_hook)
903
				err = vlan_ioctl_hook(net, argp);
A
Arjan van de Ven 已提交
904
			mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
905 906 907 908 909 910 911
			break;
		case SIOCADDDLCI:
		case SIOCDELDLCI:
			err = -ENOPKG;
			if (!dlci_ioctl_hook)
				request_module("dlci");

912 913
			mutex_lock(&dlci_ioctl_mutex);
			if (dlci_ioctl_hook)
L
Linus Torvalds 已提交
914
				err = dlci_ioctl_hook(cmd, argp);
915
			mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
916 917 918
			break;
		default:
			err = sock->ops->ioctl(sock, cmd, arg);
919 920 921 922 923 924

			/*
			 * If this ioctl is unknown try to hand it down
			 * to the NIC driver.
			 */
			if (err == -ENOIOCTLCMD)
925
				err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
926
			break;
927
		}
L
Linus Torvalds 已提交
928 929 930 931 932 933 934
	return err;
}

int sock_create_lite(int family, int type, int protocol, struct socket **res)
{
	int err;
	struct socket *sock = NULL;
935

L
Linus Torvalds 已提交
936 937 938 939 940 941 942 943 944 945 946
	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
947 948 949 950
	err = security_socket_post_create(sock, family, type, protocol, 1);
	if (err)
		goto out_release;

L
Linus Torvalds 已提交
951 952 953
out:
	*res = sock;
	return err;
V
Venkat Yekkirala 已提交
954 955 956 957
out_release:
	sock_release(sock);
	sock = NULL;
	goto out;
L
Linus Torvalds 已提交
958 959 960
}

/* No kernel lock held - perfect */
961
static unsigned int sock_poll(struct file *file, poll_table *wait)
L
Linus Torvalds 已提交
962 963 964 965
{
	struct socket *sock;

	/*
966
	 *      We can't return errors to poll, so it's either yes or no.
L
Linus Torvalds 已提交
967
	 */
968
	sock = file->private_data;
L
Linus Torvalds 已提交
969 970 971
	return sock->ops->poll(file, sock, wait);
}

972
static int sock_mmap(struct file *file, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
973
{
974
	struct socket *sock = file->private_data;
L
Linus Torvalds 已提交
975 976 977 978

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

979
static int sock_close(struct inode *inode, struct file *filp)
L
Linus Torvalds 已提交
980 981
{
	/*
982 983
	 *      It was possible the inode is NULL we were
	 *      closing an unfinished socket.
L
Linus Torvalds 已提交
984 985
	 */

986
	if (!inode) {
L
Linus Torvalds 已提交
987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011
		printk(KERN_DEBUG "sock_close: NULL inode\n");
		return 0;
	}
	sock_fasync(-1, filp, 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)
 *	   or under socket lock.
 *	3. fasync_list can be used from softirq context, so that
 *	   modification under socket lock have to be enhanced with
 *	   write_lock_bh(&sk->sk_callback_lock).
 *							--ANK (990710)
 */

static int sock_fasync(int fd, struct file *filp, int on)
{
1012
	struct fasync_struct *fa, *fna = NULL, **prev;
L
Linus Torvalds 已提交
1013 1014 1015
	struct socket *sock;
	struct sock *sk;

1016
	if (on) {
1017
		fna = kmalloc(sizeof(struct fasync_struct), GFP_KERNEL);
1018
		if (fna == NULL)
L
Linus Torvalds 已提交
1019 1020 1021
			return -ENOMEM;
	}

1022
	sock = filp->private_data;
L
Linus Torvalds 已提交
1023

1024 1025
	sk = sock->sk;
	if (sk == NULL) {
L
Linus Torvalds 已提交
1026 1027 1028 1029 1030 1031
		kfree(fna);
		return -EINVAL;
	}

	lock_sock(sk);

1032
	prev = &(sock->fasync_list);
L
Linus Torvalds 已提交
1033

1034 1035
	for (fa = *prev; fa != NULL; prev = &fa->fa_next, fa = *prev)
		if (fa->fa_file == filp)
L
Linus Torvalds 已提交
1036 1037
			break;

1038 1039
	if (on) {
		if (fa != NULL) {
L
Linus Torvalds 已提交
1040
			write_lock_bh(&sk->sk_callback_lock);
1041
			fa->fa_fd = fd;
L
Linus Torvalds 已提交
1042 1043 1044 1045 1046
			write_unlock_bh(&sk->sk_callback_lock);

			kfree(fna);
			goto out;
		}
1047 1048 1049 1050
		fna->fa_file = filp;
		fna->fa_fd = fd;
		fna->magic = FASYNC_MAGIC;
		fna->fa_next = sock->fasync_list;
L
Linus Torvalds 已提交
1051
		write_lock_bh(&sk->sk_callback_lock);
1052
		sock->fasync_list = fna;
L
Linus Torvalds 已提交
1053
		write_unlock_bh(&sk->sk_callback_lock);
1054 1055
	} else {
		if (fa != NULL) {
L
Linus Torvalds 已提交
1056
			write_lock_bh(&sk->sk_callback_lock);
1057
			*prev = fa->fa_next;
L
Linus Torvalds 已提交
1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073
			write_unlock_bh(&sk->sk_callback_lock);
			kfree(fa);
		}
	}

out:
	release_sock(sock->sk);
	return 0;
}

/* This function may be called only under socket lock or callback_lock */

int sock_wake_async(struct socket *sock, int how, int band)
{
	if (!sock || !sock->fasync_list)
		return -1;
1074
	switch (how) {
1075
	case SOCK_WAKE_WAITD:
L
Linus Torvalds 已提交
1076 1077 1078
		if (test_bit(SOCK_ASYNC_WAITDATA, &sock->flags))
			break;
		goto call_kill;
1079
	case SOCK_WAKE_SPACE:
L
Linus Torvalds 已提交
1080 1081 1082
		if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags))
			break;
		/* fall through */
1083
	case SOCK_WAKE_IO:
1084
call_kill:
L
Linus Torvalds 已提交
1085 1086
		__kill_fasync(sock->fasync_list, SIGIO, band);
		break;
1087
	case SOCK_WAKE_URG:
L
Linus Torvalds 已提交
1088 1089 1090 1091 1092
		__kill_fasync(sock->fasync_list, SIGURG, band);
	}
	return 0;
}

1093
static int __sock_create(struct net *net, int family, int type, int protocol,
1094
			 struct socket **res, int kern)
L
Linus Torvalds 已提交
1095 1096 1097
{
	int err;
	struct socket *sock;
1098
	const struct net_proto_family *pf;
L
Linus Torvalds 已提交
1099 1100

	/*
1101
	 *      Check protocol is in range
L
Linus Torvalds 已提交
1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
	 */
	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) {
1114
		static int warned;
L
Linus Torvalds 已提交
1115 1116
		if (!warned) {
			warned = 1;
1117 1118
			printk(KERN_INFO "%s uses obsolete (PF_INET,SOCK_PACKET)\n",
			       current->comm);
L
Linus Torvalds 已提交
1119 1120 1121 1122 1123 1124 1125
		}
		family = PF_PACKET;
	}

	err = security_socket_create(family, type, protocol, kern);
	if (err)
		return err;
1126

1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141
	/*
	 *	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;

L
Linus Torvalds 已提交
1142
#if defined(CONFIG_KMOD)
1143 1144 1145
	/* 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
1146 1147 1148
	 * requested real, full-featured networking support upon configuration.
	 * Otherwise module support will break!
	 */
1149
	if (net_families[family] == NULL)
1150
		request_module("net-pf-%d", family);
L
Linus Torvalds 已提交
1151 1152
#endif

1153 1154 1155 1156 1157
	rcu_read_lock();
	pf = rcu_dereference(net_families[family]);
	err = -EAFNOSUPPORT;
	if (!pf)
		goto out_release;
L
Linus Torvalds 已提交
1158 1159 1160 1161 1162

	/*
	 * We will call the ->create function, that possibly is in a loadable
	 * module, so we have to bump that loadable module refcnt first.
	 */
1163
	if (!try_module_get(pf->owner))
L
Linus Torvalds 已提交
1164 1165
		goto out_release;

1166 1167 1168
	/* Now protected by module ref count */
	rcu_read_unlock();

1169
	err = pf->create(net, sock, protocol);
1170
	if (err < 0)
L
Linus Torvalds 已提交
1171
		goto out_module_put;
1172

L
Linus Torvalds 已提交
1173 1174 1175 1176
	/*
	 * Now to bump the refcnt of the [loadable] module that owns this
	 * socket at sock_release time we decrement its refcnt.
	 */
1177 1178 1179
	if (!try_module_get(sock->ops->owner))
		goto out_module_busy;

L
Linus Torvalds 已提交
1180 1181 1182 1183
	/*
	 * Now that we're done with the ->create function, the [loadable]
	 * module can have its refcnt decremented
	 */
1184
	module_put(pf->owner);
V
Venkat Yekkirala 已提交
1185 1186
	err = security_socket_post_create(sock, family, type, protocol, kern);
	if (err)
1187
		goto out_sock_release;
1188
	*res = sock;
L
Linus Torvalds 已提交
1189

1190 1191 1192 1193
	return 0;

out_module_busy:
	err = -EAFNOSUPPORT;
L
Linus Torvalds 已提交
1194
out_module_put:
1195 1196 1197
	sock->ops = NULL;
	module_put(pf->owner);
out_sock_release:
L
Linus Torvalds 已提交
1198
	sock_release(sock);
1199 1200 1201 1202 1203
	return err;

out_release:
	rcu_read_unlock();
	goto out_sock_release;
L
Linus Torvalds 已提交
1204 1205 1206 1207
}

int sock_create(int family, int type, int protocol, struct socket **res)
{
1208
	return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0);
L
Linus Torvalds 已提交
1209 1210 1211 1212
}

int sock_create_kern(int family, int type, int protocol, struct socket **res)
{
1213
	return __sock_create(&init_net, family, type, protocol, res, 1);
L
Linus Torvalds 已提交
1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241
}

asmlinkage long sys_socket(int family, int type, int protocol)
{
	int retval;
	struct socket *sock;

	retval = sock_create(family, type, protocol, &sock);
	if (retval < 0)
		goto out;

	retval = sock_map_fd(sock);
	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.
 */

1242 1243
asmlinkage long sys_socketpair(int family, int type, int protocol,
			       int __user *usockvec)
L
Linus Torvalds 已提交
1244 1245 1246
{
	struct socket *sock1, *sock2;
	int fd1, fd2, err;
A
Al Viro 已提交
1247
	struct file *newfile1, *newfile2;
L
Linus Torvalds 已提交
1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262

	/*
	 * 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);
1263
	if (err < 0)
L
Linus Torvalds 已提交
1264 1265
		goto out_release_both;

A
Al Viro 已提交
1266
	fd1 = sock_alloc_fd(&newfile1);
1267 1268
	if (unlikely(fd1 < 0)) {
		err = fd1;
A
Al Viro 已提交
1269
		goto out_release_both;
1270
	}
L
Linus Torvalds 已提交
1271

A
Al Viro 已提交
1272 1273
	fd2 = sock_alloc_fd(&newfile2);
	if (unlikely(fd2 < 0)) {
1274
		err = fd2;
A
Al Viro 已提交
1275 1276
		put_filp(newfile1);
		put_unused_fd(fd1);
L
Linus Torvalds 已提交
1277
		goto out_release_both;
A
Al Viro 已提交
1278
	}
L
Linus Torvalds 已提交
1279

A
Al Viro 已提交
1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296
	err = sock_attach_fd(sock1, newfile1);
	if (unlikely(err < 0)) {
		goto out_fd2;
	}

	err = sock_attach_fd(sock2, newfile2);
	if (unlikely(err < 0)) {
		fput(newfile1);
		goto out_fd1;
	}

	err = audit_fd_pair(fd1, fd2);
	if (err < 0) {
		fput(newfile1);
		fput(newfile2);
		goto out_fd;
	}
L
Linus Torvalds 已提交
1297

A
Al Viro 已提交
1298 1299
	fd_install(fd1, newfile1);
	fd_install(fd2, newfile2);
L
Linus Torvalds 已提交
1300 1301 1302 1303
	/* fd1 and fd2 may be already another descriptors.
	 * Not kernel problem.
	 */

1304
	err = put_user(fd1, &usockvec[0]);
L
Linus Torvalds 已提交
1305 1306 1307 1308 1309 1310 1311 1312 1313 1314
	if (!err)
		err = put_user(fd2, &usockvec[1]);
	if (!err)
		return 0;

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

out_release_both:
1315
	sock_release(sock2);
L
Linus Torvalds 已提交
1316
out_release_1:
1317
	sock_release(sock1);
L
Linus Torvalds 已提交
1318 1319
out:
	return err;
A
Al Viro 已提交
1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330

out_fd2:
	put_filp(newfile1);
	sock_release(sock1);
out_fd1:
	put_filp(newfile2);
	sock_release(sock2);
out_fd:
	put_unused_fd(fd1);
	put_unused_fd(fd2);
	goto out;
L
Linus Torvalds 已提交
1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344
}

/*
 *	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).
 */

asmlinkage long sys_bind(int fd, struct sockaddr __user *umyaddr, int addrlen)
{
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
1345
	int err, fput_needed;
L
Linus Torvalds 已提交
1346

1347
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1348
	if (sock) {
1349 1350 1351 1352 1353
		err = move_addr_to_kernel(umyaddr, addrlen, address);
		if (err >= 0) {
			err = security_socket_bind(sock,
						   (struct sockaddr *)address,
						   addrlen);
1354 1355
			if (!err)
				err = sock->ops->bind(sock,
1356 1357
						      (struct sockaddr *)
						      address, addrlen);
L
Linus Torvalds 已提交
1358
		}
1359
		fput_light(sock->file, fput_needed);
1360
	}
L
Linus Torvalds 已提交
1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372
	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.
 */

asmlinkage long sys_listen(int fd, int backlog)
{
	struct socket *sock;
1373
	int err, fput_needed;
1374
	int somaxconn;
1375 1376 1377

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock) {
1378
		somaxconn = sock_net(sock->sk)->core.sysctl_somaxconn;
1379 1380
		if ((unsigned)backlog > somaxconn)
			backlog = somaxconn;
L
Linus Torvalds 已提交
1381 1382

		err = security_socket_listen(sock, backlog);
1383 1384
		if (!err)
			err = sock->ops->listen(sock, backlog);
L
Linus Torvalds 已提交
1385

1386
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402
	}
	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.
 */

1403 1404
asmlinkage long sys_accept(int fd, struct sockaddr __user *upeer_sockaddr,
			   int __user *upeer_addrlen)
L
Linus Torvalds 已提交
1405 1406
{
	struct socket *sock, *newsock;
1407
	struct file *newfile;
1408
	int err, len, newfd, fput_needed;
L
Linus Torvalds 已提交
1409 1410
	char address[MAX_SOCK_ADDR];

1411
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1412 1413 1414 1415
	if (!sock)
		goto out;

	err = -ENFILE;
1416
	if (!(newsock = sock_alloc()))
L
Linus Torvalds 已提交
1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427
		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);

1428 1429 1430
	newfd = sock_alloc_fd(&newfile);
	if (unlikely(newfd < 0)) {
		err = newfd;
1431 1432
		sock_release(newsock);
		goto out_put;
1433 1434 1435 1436
	}

	err = sock_attach_fd(newsock, newfile);
	if (err < 0)
1437
		goto out_fd_simple;
1438

1439 1440
	err = security_socket_accept(sock, newsock);
	if (err)
1441
		goto out_fd;
1442

L
Linus Torvalds 已提交
1443 1444
	err = sock->ops->accept(sock, newsock, sock->file->f_flags);
	if (err < 0)
1445
		goto out_fd;
L
Linus Torvalds 已提交
1446 1447

	if (upeer_sockaddr) {
1448 1449
		if (newsock->ops->getname(newsock, (struct sockaddr *)address,
					  &len, 2) < 0) {
L
Linus Torvalds 已提交
1450
			err = -ECONNABORTED;
1451
			goto out_fd;
L
Linus Torvalds 已提交
1452
		}
1453 1454
		err = move_addr_to_user(address, len, upeer_sockaddr,
					upeer_addrlen);
L
Linus Torvalds 已提交
1455
		if (err < 0)
1456
			goto out_fd;
L
Linus Torvalds 已提交
1457 1458 1459 1460
	}

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

1461 1462
	fd_install(newfd, newfile);
	err = newfd;
L
Linus Torvalds 已提交
1463 1464 1465 1466

	security_socket_post_accept(sock, newsock);

out_put:
1467
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1468 1469
out:
	return err;
1470 1471 1472 1473 1474
out_fd_simple:
	sock_release(newsock);
	put_filp(newfile);
	put_unused_fd(newfd);
	goto out_put;
1475
out_fd:
1476
	fput(newfile);
1477
	put_unused_fd(newfd);
L
Linus Torvalds 已提交
1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492
	goto out_put;
}

/*
 *	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.
 */

1493 1494
asmlinkage long sys_connect(int fd, struct sockaddr __user *uservaddr,
			    int addrlen)
L
Linus Torvalds 已提交
1495 1496 1497
{
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
1498
	int err, fput_needed;
L
Linus Torvalds 已提交
1499

1500
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1501 1502 1503 1504 1505 1506
	if (!sock)
		goto out;
	err = move_addr_to_kernel(uservaddr, addrlen, address);
	if (err < 0)
		goto out_put;

1507 1508
	err =
	    security_socket_connect(sock, (struct sockaddr *)address, addrlen);
L
Linus Torvalds 已提交
1509 1510 1511
	if (err)
		goto out_put;

1512
	err = sock->ops->connect(sock, (struct sockaddr *)address, addrlen,
L
Linus Torvalds 已提交
1513 1514
				 sock->file->f_flags);
out_put:
1515
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1516 1517 1518 1519 1520 1521 1522 1523 1524
out:
	return err;
}

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

1525 1526
asmlinkage long sys_getsockname(int fd, struct sockaddr __user *usockaddr,
				int __user *usockaddr_len)
L
Linus Torvalds 已提交
1527 1528 1529
{
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
1530
	int len, err, fput_needed;
1531

1532
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
	if (!sock)
		goto out;

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

	err = sock->ops->getname(sock, (struct sockaddr *)address, &len, 0);
	if (err)
		goto out_put;
	err = move_addr_to_user(address, len, usockaddr, usockaddr_len);

out_put:
1546
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1547 1548 1549 1550 1551 1552 1553 1554 1555
out:
	return err;
}

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

1556 1557
asmlinkage long sys_getpeername(int fd, struct sockaddr __user *usockaddr,
				int __user *usockaddr_len)
L
Linus Torvalds 已提交
1558 1559 1560
{
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
1561
	int len, err, fput_needed;
L
Linus Torvalds 已提交
1562

1563 1564
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1565 1566
		err = security_socket_getpeername(sock);
		if (err) {
1567
			fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1568 1569 1570
			return err;
		}

1571 1572 1573
		err =
		    sock->ops->getname(sock, (struct sockaddr *)address, &len,
				       1);
L
Linus Torvalds 已提交
1574
		if (!err)
1575 1576
			err = move_addr_to_user(address, len, usockaddr,
						usockaddr_len);
1577
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1578 1579 1580 1581 1582 1583 1584 1585 1586 1587
	}
	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.
 */

1588 1589 1590
asmlinkage long sys_sendto(int fd, void __user *buff, size_t len,
			   unsigned flags, struct sockaddr __user *addr,
			   int addr_len)
L
Linus Torvalds 已提交
1591 1592 1593 1594 1595 1596
{
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
	int err;
	struct msghdr msg;
	struct iovec iov;
1597 1598
	int fput_needed;

1599 1600
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (!sock)
1601
		goto out;
1602

1603 1604 1605 1606 1607 1608 1609 1610
	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;
1611
	if (addr) {
L
Linus Torvalds 已提交
1612 1613 1614
		err = move_addr_to_kernel(addr, addr_len, address);
		if (err < 0)
			goto out_put;
1615 1616
		msg.msg_name = address;
		msg.msg_namelen = addr_len;
L
Linus Torvalds 已提交
1617 1618 1619 1620 1621 1622
	}
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
	msg.msg_flags = flags;
	err = sock_sendmsg(sock, &msg, len);

1623
out_put:
1624
	fput_light(sock->file, fput_needed);
1625
out:
L
Linus Torvalds 已提交
1626 1627 1628 1629
	return err;
}

/*
1630
 *	Send a datagram down a socket.
L
Linus Torvalds 已提交
1631 1632
 */

1633
asmlinkage long sys_send(int fd, void __user *buff, size_t len, unsigned flags)
L
Linus Torvalds 已提交
1634 1635 1636 1637 1638
{
	return sys_sendto(fd, buff, len, flags, NULL, 0);
}

/*
1639
 *	Receive a frame from the socket and optionally record the address of the
L
Linus Torvalds 已提交
1640 1641 1642 1643
 *	sender. We verify the buffers are writable and if needed move the
 *	sender address from kernel to user space.
 */

1644 1645 1646
asmlinkage long sys_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
1647 1648 1649 1650 1651
{
	struct socket *sock;
	struct iovec iov;
	struct msghdr msg;
	char address[MAX_SOCK_ADDR];
1652
	int err, err2;
1653 1654
	int fput_needed;

1655
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1656
	if (!sock)
1657
		goto out;
L
Linus Torvalds 已提交
1658

1659 1660 1661 1662 1663 1664 1665 1666
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_iovlen = 1;
	msg.msg_iov = &iov;
	iov.iov_len = size;
	iov.iov_base = ubuf;
	msg.msg_name = address;
	msg.msg_namelen = MAX_SOCK_ADDR;
L
Linus Torvalds 已提交
1667 1668
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
1669
	err = sock_recvmsg(sock, &msg, size, flags);
L
Linus Torvalds 已提交
1670

1671 1672 1673 1674
	if (err >= 0 && addr != NULL) {
		err2 = move_addr_to_user(address, msg.msg_namelen, addr, addr_len);
		if (err2 < 0)
			err = err2;
L
Linus Torvalds 已提交
1675
	}
1676 1677

	fput_light(sock->file, fput_needed);
1678
out:
L
Linus Torvalds 已提交
1679 1680 1681 1682
	return err;
}

/*
1683
 *	Receive a datagram from a socket.
L
Linus Torvalds 已提交
1684 1685
 */

1686 1687
asmlinkage long sys_recv(int fd, void __user *ubuf, size_t size,
			 unsigned flags)
L
Linus Torvalds 已提交
1688 1689 1690 1691 1692 1693 1694 1695 1696
{
	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.
 */

1697 1698
asmlinkage long sys_setsockopt(int fd, int level, int optname,
			       char __user *optval, int optlen)
L
Linus Torvalds 已提交
1699
{
1700
	int err, fput_needed;
L
Linus Torvalds 已提交
1701 1702 1703 1704
	struct socket *sock;

	if (optlen < 0)
		return -EINVAL;
1705 1706 1707 1708

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
		err = security_socket_setsockopt(sock, level, optname);
1709 1710
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1711 1712

		if (level == SOL_SOCKET)
1713 1714 1715
			err =
			    sock_setsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1716
		else
1717 1718 1719
			err =
			    sock->ops->setsockopt(sock, level, optname, optval,
						  optlen);
1720 1721
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726 1727 1728 1729 1730
	}
	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.
 */

1731 1732
asmlinkage long sys_getsockopt(int fd, int level, int optname,
			       char __user *optval, int __user *optlen)
L
Linus Torvalds 已提交
1733
{
1734
	int err, fput_needed;
L
Linus Torvalds 已提交
1735 1736
	struct socket *sock;

1737 1738
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
1739 1740 1741
		err = security_socket_getsockopt(sock, level, optname);
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1742 1743

		if (level == SOL_SOCKET)
1744 1745 1746
			err =
			    sock_getsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1747
		else
1748 1749 1750
			err =
			    sock->ops->getsockopt(sock, level, optname, optval,
						  optlen);
1751 1752
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1753 1754 1755 1756 1757 1758 1759 1760 1761 1762
	}
	return err;
}

/*
 *	Shutdown a socket.
 */

asmlinkage long sys_shutdown(int fd, int how)
{
1763
	int err, fput_needed;
L
Linus Torvalds 已提交
1764 1765
	struct socket *sock;

1766 1767
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1768
		err = security_socket_shutdown(sock, how);
1769 1770 1771
		if (!err)
			err = sock->ops->shutdown(sock, how);
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1772 1773 1774 1775
	}
	return err;
}

1776
/* A couple of helpful macros for getting the address of the 32/64 bit
L
Linus Torvalds 已提交
1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788
 * fields which are the same type (int / unsigned) on our platforms.
 */
#define COMPAT_MSG(msg, member)	((MSG_CMSG_COMPAT & flags) ? &msg##_compat->member : &msg->member)
#define COMPAT_NAMELEN(msg)	COMPAT_MSG(msg, msg_namelen)
#define COMPAT_FLAGS(msg)	COMPAT_MSG(msg, msg_flags)

/*
 *	BSD sendmsg interface
 */

asmlinkage long sys_sendmsg(int fd, struct msghdr __user *msg, unsigned flags)
{
1789 1790
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
1791 1792 1793
	struct socket *sock;
	char address[MAX_SOCK_ADDR];
	struct iovec iovstack[UIO_FASTIOV], *iov = iovstack;
1794
	unsigned char ctl[sizeof(struct cmsghdr) + 20]
1795 1796
	    __attribute__ ((aligned(sizeof(__kernel_size_t))));
	/* 20 is size of ipv6_pktinfo */
L
Linus Torvalds 已提交
1797 1798 1799
	unsigned char *ctl_buf = ctl;
	struct msghdr msg_sys;
	int err, ctl_len, iov_size, total_len;
1800
	int fput_needed;
1801

L
Linus Torvalds 已提交
1802 1803 1804 1805
	err = -EFAULT;
	if (MSG_CMSG_COMPAT & flags) {
		if (get_compat_msghdr(&msg_sys, msg_compat))
			return -EFAULT;
1806 1807
	}
	else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr)))
L
Linus Torvalds 已提交
1808 1809
		return -EFAULT;

1810
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1811
	if (!sock)
L
Linus Torvalds 已提交
1812 1813 1814 1815 1816 1817 1818
		goto out;

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

1819
	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832
	err = -ENOMEM;
	iov_size = msg_sys.msg_iovlen * sizeof(struct iovec);
	if (msg_sys.msg_iovlen > UIO_FASTIOV) {
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
			goto out_put;
	}

	/* This will also move the address data into kernel space */
	if (MSG_CMSG_COMPAT & flags) {
		err = verify_compat_iovec(&msg_sys, iov, address, VERIFY_READ);
	} else
		err = verify_iovec(&msg_sys, iov, address, VERIFY_READ);
1833
	if (err < 0)
L
Linus Torvalds 已提交
1834 1835 1836 1837 1838 1839 1840
		goto out_freeiov;
	total_len = err;

	err = -ENOBUFS;

	if (msg_sys.msg_controllen > INT_MAX)
		goto out_freeiov;
1841
	ctl_len = msg_sys.msg_controllen;
L
Linus Torvalds 已提交
1842
	if ((MSG_CMSG_COMPAT & flags) && ctl_len) {
1843 1844 1845
		err =
		    cmsghdr_from_user_compat_to_kern(&msg_sys, sock->sk, ctl,
						     sizeof(ctl));
L
Linus Torvalds 已提交
1846 1847 1848
		if (err)
			goto out_freeiov;
		ctl_buf = msg_sys.msg_control;
A
Al Viro 已提交
1849
		ctl_len = msg_sys.msg_controllen;
L
Linus Torvalds 已提交
1850
	} else if (ctl_len) {
1851
		if (ctl_len > sizeof(ctl)) {
L
Linus Torvalds 已提交
1852
			ctl_buf = sock_kmalloc(sock->sk, ctl_len, GFP_KERNEL);
1853
			if (ctl_buf == NULL)
L
Linus Torvalds 已提交
1854 1855 1856 1857 1858 1859 1860 1861
				goto out_freeiov;
		}
		err = -EFAULT;
		/*
		 * Careful! Before this, msg_sys.msg_control contains a user pointer.
		 * Afterwards, it will be a kernel pointer. Thus the compiler-assisted
		 * checking falls down on this.
		 */
1862 1863
		if (copy_from_user(ctl_buf, (void __user *)msg_sys.msg_control,
				   ctl_len))
L
Linus Torvalds 已提交
1864 1865 1866 1867 1868 1869 1870 1871 1872 1873
			goto out_freectl;
		msg_sys.msg_control = ctl_buf;
	}
	msg_sys.msg_flags = flags;

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

out_freectl:
1874
	if (ctl_buf != ctl)
L
Linus Torvalds 已提交
1875 1876 1877 1878 1879
		sock_kfree_s(sock->sk, ctl_buf, ctl_len);
out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
out_put:
1880
	fput_light(sock->file, fput_needed);
1881
out:
L
Linus Torvalds 已提交
1882 1883 1884 1885 1886 1887 1888
	return err;
}

/*
 *	BSD recvmsg interface
 */

1889 1890
asmlinkage long sys_recvmsg(int fd, struct msghdr __user *msg,
			    unsigned int flags)
L
Linus Torvalds 已提交
1891
{
1892 1893
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
1894 1895
	struct socket *sock;
	struct iovec iovstack[UIO_FASTIOV];
1896
	struct iovec *iov = iovstack;
L
Linus Torvalds 已提交
1897 1898 1899
	struct msghdr msg_sys;
	unsigned long cmsg_ptr;
	int err, iov_size, total_len, len;
1900
	int fput_needed;
L
Linus Torvalds 已提交
1901 1902 1903 1904 1905 1906 1907

	/* kernel mode address */
	char addr[MAX_SOCK_ADDR];

	/* user mode address pointers */
	struct sockaddr __user *uaddr;
	int __user *uaddr_len;
1908

L
Linus Torvalds 已提交
1909 1910 1911
	if (MSG_CMSG_COMPAT & flags) {
		if (get_compat_msghdr(&msg_sys, msg_compat))
			return -EFAULT;
1912 1913 1914
	}
	else if (copy_from_user(&msg_sys, msg, sizeof(struct msghdr)))
		return -EFAULT;
L
Linus Torvalds 已提交
1915

1916
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1917 1918 1919 1920 1921 1922
	if (!sock)
		goto out;

	err = -EMSGSIZE;
	if (msg_sys.msg_iovlen > UIO_MAXIOV)
		goto out_put;
1923 1924

	/* Check whether to allocate the iovec area */
L
Linus Torvalds 已提交
1925 1926 1927 1928 1929 1930 1931 1932 1933
	err = -ENOMEM;
	iov_size = msg_sys.msg_iovlen * sizeof(struct iovec);
	if (msg_sys.msg_iovlen > UIO_FASTIOV) {
		iov = sock_kmalloc(sock->sk, iov_size, GFP_KERNEL);
		if (!iov)
			goto out_put;
	}

	/*
1934 1935
	 *      Save the user-mode address (verify_iovec will change the
	 *      kernel msghdr to use the kernel address space)
L
Linus Torvalds 已提交
1936
	 */
1937

S
Stephen Hemminger 已提交
1938
	uaddr = (__force void __user *)msg_sys.msg_name;
L
Linus Torvalds 已提交
1939 1940 1941 1942 1943 1944 1945
	uaddr_len = COMPAT_NAMELEN(msg);
	if (MSG_CMSG_COMPAT & flags) {
		err = verify_compat_iovec(&msg_sys, iov, addr, VERIFY_WRITE);
	} else
		err = verify_iovec(&msg_sys, iov, addr, VERIFY_WRITE);
	if (err < 0)
		goto out_freeiov;
1946
	total_len = err;
L
Linus Torvalds 已提交
1947 1948

	cmsg_ptr = (unsigned long)msg_sys.msg_control;
U
Ulrich Drepper 已提交
1949
	msg_sys.msg_flags = flags & (MSG_CMSG_CLOEXEC|MSG_CMSG_COMPAT);
1950

L
Linus Torvalds 已提交
1951 1952 1953 1954 1955 1956 1957 1958
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
	err = sock_recvmsg(sock, &msg_sys, total_len, flags);
	if (err < 0)
		goto out_freeiov;
	len = err;

	if (uaddr != NULL) {
1959 1960
		err = move_addr_to_user(addr, msg_sys.msg_namelen, uaddr,
					uaddr_len);
L
Linus Torvalds 已提交
1961 1962 1963
		if (err < 0)
			goto out_freeiov;
	}
1964 1965
	err = __put_user((msg_sys.msg_flags & ~MSG_CMSG_COMPAT),
			 COMPAT_FLAGS(msg));
L
Linus Torvalds 已提交
1966 1967 1968
	if (err)
		goto out_freeiov;
	if (MSG_CMSG_COMPAT & flags)
1969
		err = __put_user((unsigned long)msg_sys.msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
1970 1971
				 &msg_compat->msg_controllen);
	else
1972
		err = __put_user((unsigned long)msg_sys.msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
1973 1974 1975 1976 1977 1978 1979 1980 1981
				 &msg->msg_controllen);
	if (err)
		goto out_freeiov;
	err = len;

out_freeiov:
	if (iov != iovstack)
		sock_kfree_s(sock->sk, iov, iov_size);
out_put:
1982
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1983 1984 1985 1986 1987 1988 1989 1990
out:
	return err;
}

#ifdef __ARCH_WANT_SYS_SOCKETCALL

/* Argument list sizes for sys_socketcall */
#define AL(x) ((x) * sizeof(unsigned long))
1991 1992 1993 1994 1995 1996
static const unsigned char nargs[18]={
	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)
};

L
Linus Torvalds 已提交
1997 1998 1999
#undef AL

/*
2000
 *	System call vectors.
L
Linus Torvalds 已提交
2001 2002 2003
 *
 *	Argument checking cleaned up. Saved 20% in size.
 *  This function doesn't need to set the kernel lock because
2004
 *  it is set by the callees.
L
Linus Torvalds 已提交
2005 2006 2007 2008 2009
 */

asmlinkage long sys_socketcall(int call, unsigned long __user *args)
{
	unsigned long a[6];
2010
	unsigned long a0, a1;
L
Linus Torvalds 已提交
2011 2012
	int err;

2013
	if (call < 1 || call > SYS_RECVMSG)
L
Linus Torvalds 已提交
2014 2015 2016 2017 2018
		return -EINVAL;

	/* copy_from_user should be SMP safe. */
	if (copy_from_user(a, args, nargs[call]))
		return -EFAULT;
2019

2020
	err = audit_socketcall(nargs[call] / sizeof(unsigned long), a);
2021 2022 2023
	if (err)
		return err;

2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092
	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:
		err =
		    sys_accept(a0, (struct sockaddr __user *)a1,
			       (int __user *)a[2]);
		break;
	case SYS_GETSOCKNAME:
		err =
		    sys_getsockname(a0, (struct sockaddr __user *)a1,
				    (int __user *)a[2]);
		break;
	case SYS_GETPEERNAME:
		err =
		    sys_getpeername(a0, (struct sockaddr __user *)a1,
				    (int __user *)a[2]);
		break;
	case SYS_SOCKETPAIR:
		err = sys_socketpair(a0, a1, a[2], (int __user *)a[3]);
		break;
	case SYS_SEND:
		err = sys_send(a0, (void __user *)a1, a[2], a[3]);
		break;
	case SYS_SENDTO:
		err = sys_sendto(a0, (void __user *)a1, a[2], a[3],
				 (struct sockaddr __user *)a[4], a[5]);
		break;
	case SYS_RECV:
		err = sys_recv(a0, (void __user *)a1, a[2], a[3]);
		break;
	case SYS_RECVFROM:
		err = sys_recvfrom(a0, (void __user *)a1, a[2], a[3],
				   (struct sockaddr __user *)a[4],
				   (int __user *)a[5]);
		break;
	case SYS_SHUTDOWN:
		err = sys_shutdown(a0, a1);
		break;
	case SYS_SETSOCKOPT:
		err = sys_setsockopt(a0, a1, a[2], (char __user *)a[3], a[4]);
		break;
	case SYS_GETSOCKOPT:
		err =
		    sys_getsockopt(a0, a1, a[2], (char __user *)a[3],
				   (int __user *)a[4]);
		break;
	case SYS_SENDMSG:
		err = sys_sendmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
	case SYS_RECVMSG:
		err = sys_recvmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
	default:
		err = -EINVAL;
		break;
L
Linus Torvalds 已提交
2093 2094 2095 2096
	}
	return err;
}

2097
#endif				/* __ARCH_WANT_SYS_SOCKETCALL */
L
Linus Torvalds 已提交
2098

2099 2100 2101 2102
/**
 *	sock_register - add a socket protocol handler
 *	@ops: description of protocol
 *
L
Linus Torvalds 已提交
2103 2104
 *	This function is called by a protocol handler that wants to
 *	advertise its address family, and have it linked into the
2105 2106
 *	socket interface. The value ops->family coresponds to the
 *	socket system call protocol family.
L
Linus Torvalds 已提交
2107
 */
2108
int sock_register(const struct net_proto_family *ops)
L
Linus Torvalds 已提交
2109 2110 2111 2112
{
	int err;

	if (ops->family >= NPROTO) {
2113 2114
		printk(KERN_CRIT "protocol %d >= NPROTO(%d)\n", ops->family,
		       NPROTO);
L
Linus Torvalds 已提交
2115 2116
		return -ENOBUFS;
	}
2117 2118 2119 2120 2121

	spin_lock(&net_family_lock);
	if (net_families[ops->family])
		err = -EEXIST;
	else {
2122
		net_families[ops->family] = ops;
L
Linus Torvalds 已提交
2123 2124
		err = 0;
	}
2125 2126
	spin_unlock(&net_family_lock);

2127
	printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family);
L
Linus Torvalds 已提交
2128 2129 2130
	return err;
}

2131 2132 2133 2134
/**
 *	sock_unregister - remove a protocol handler
 *	@family: protocol family to remove
 *
L
Linus Torvalds 已提交
2135 2136
 *	This function is called by a protocol handler that wants to
 *	remove its address family, and have it unlinked from the
2137 2138 2139 2140 2141 2142
 *	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
2143
 */
2144
void sock_unregister(int family)
L
Linus Torvalds 已提交
2145
{
2146
	BUG_ON(family < 0 || family >= NPROTO);
L
Linus Torvalds 已提交
2147

2148
	spin_lock(&net_family_lock);
2149
	net_families[family] = NULL;
2150 2151 2152 2153
	spin_unlock(&net_family_lock);

	synchronize_rcu();

2154
	printk(KERN_INFO "NET: Unregistered protocol family %d\n", family);
L
Linus Torvalds 已提交
2155 2156
}

2157
static int __init sock_init(void)
L
Linus Torvalds 已提交
2158 2159
{
	/*
2160
	 *      Initialize sock SLAB cache.
L
Linus Torvalds 已提交
2161
	 */
2162

L
Linus Torvalds 已提交
2163 2164 2165
	sk_init();

	/*
2166
	 *      Initialize skbuff SLAB cache
L
Linus Torvalds 已提交
2167 2168 2169 2170
	 */
	skb_init();

	/*
2171
	 *      Initialize the protocols module.
L
Linus Torvalds 已提交
2172 2173 2174 2175 2176
	 */

	init_inodecache();
	register_filesystem(&sock_fs_type);
	sock_mnt = kern_mount(&sock_fs_type);
2177 2178

	/* The real protocol initialization is performed in later initcalls.
L
Linus Torvalds 已提交
2179 2180 2181 2182 2183
	 */

#ifdef CONFIG_NETFILTER
	netfilter_init();
#endif
2184 2185

	return 0;
L
Linus Torvalds 已提交
2186 2187
}

2188 2189
core_initcall(sock_init);	/* early initcall */

L
Linus Torvalds 已提交
2190 2191 2192 2193 2194 2195
#ifdef CONFIG_PROC_FS
void socket_seq_show(struct seq_file *seq)
{
	int cpu;
	int counter = 0;

2196
	for_each_possible_cpu(cpu)
2197
	    counter += per_cpu(sockets_in_use, cpu);
L
Linus Torvalds 已提交
2198 2199 2200 2201 2202 2203 2204

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

	seq_printf(seq, "sockets: used %d\n", counter);
}
2205
#endif				/* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
2206

2207 2208
#ifdef CONFIG_COMPAT
static long compat_sock_ioctl(struct file *file, unsigned cmd,
2209
			      unsigned long arg)
2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220
{
	struct socket *sock = file->private_data;
	int ret = -ENOIOCTLCMD;

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

	return ret;
}
#endif

2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243
int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen)
{
	return sock->ops->bind(sock, addr, addrlen);
}

int kernel_listen(struct socket *sock, int backlog)
{
	return sock->ops->listen(sock, backlog);
}

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);
2244
		*newsock = NULL;
2245 2246 2247 2248 2249 2250 2251 2252 2253 2254
		goto done;
	}

	(*newsock)->ops = sock->ops;

done:
	return err;
}

int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen,
2255
		   int flags)
2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 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
{
	return sock->ops->connect(sock, addr, addrlen, flags);
}

int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 0);
}

int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 1);
}

int kernel_getsockopt(struct socket *sock, int level, int optname,
			char *optval, int *optlen)
{
	mm_segment_t oldfs = get_fs();
	int err;

	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
		err = sock_getsockopt(sock, level, optname, optval, optlen);
	else
		err = sock->ops->getsockopt(sock, level, optname, optval,
					    optlen);
	set_fs(oldfs);
	return err;
}

int kernel_setsockopt(struct socket *sock, int level, int optname,
			char *optval, int optlen)
{
	mm_segment_t oldfs = get_fs();
	int err;

	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
		err = sock_setsockopt(sock, level, optname, optval, optlen);
	else
		err = sock->ops->setsockopt(sock, level, optname, optval,
					    optlen);
	set_fs(oldfs);
	return err;
}

int kernel_sendpage(struct socket *sock, struct page *page, int offset,
		    size_t size, int flags)
{
	if (sock->ops->sendpage)
		return sock->ops->sendpage(sock, page, offset, size, flags);

	return sock_no_sendpage(sock, page, offset, size, flags);
}

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;
}

2325 2326 2327 2328 2329
int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how)
{
	return sock->ops->shutdown(sock, how);
}

L
Linus Torvalds 已提交
2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342
EXPORT_SYMBOL(sock_create);
EXPORT_SYMBOL(sock_create_kern);
EXPORT_SYMBOL(sock_create_lite);
EXPORT_SYMBOL(sock_map_fd);
EXPORT_SYMBOL(sock_recvmsg);
EXPORT_SYMBOL(sock_register);
EXPORT_SYMBOL(sock_release);
EXPORT_SYMBOL(sock_sendmsg);
EXPORT_SYMBOL(sock_unregister);
EXPORT_SYMBOL(sock_wake_async);
EXPORT_SYMBOL(sockfd_lookup);
EXPORT_SYMBOL(kernel_sendmsg);
EXPORT_SYMBOL(kernel_recvmsg);
2343 2344 2345 2346 2347 2348 2349 2350 2351 2352
EXPORT_SYMBOL(kernel_bind);
EXPORT_SYMBOL(kernel_listen);
EXPORT_SYMBOL(kernel_accept);
EXPORT_SYMBOL(kernel_connect);
EXPORT_SYMBOL(kernel_getsockname);
EXPORT_SYMBOL(kernel_getpeername);
EXPORT_SYMBOL(kernel_getsockopt);
EXPORT_SYMBOL(kernel_setsockopt);
EXPORT_SYMBOL(kernel_sendpage);
EXPORT_SYMBOL(kernel_sock_ioctl);
2353
EXPORT_SYMBOL(kernel_sock_shutdown);