socket.c 83.3 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>
91
#include <linux/xattr.h>
L
Linus Torvalds 已提交
92 93 94 95 96

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

#include <net/compat.h>
97
#include <net/wext.h>
98
#include <net/cls_cgroup.h>
L
Linus Torvalds 已提交
99 100 101 102

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

103 104 105 106 107 108
#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
109
static int sock_no_open(struct inode *irrelevant, struct file *dontcare);
110 111 112 113
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);
114
static int sock_mmap(struct file *file, struct vm_area_struct *vma);
L
Linus Torvalds 已提交
115 116 117 118

static int sock_close(struct inode *inode, struct file *file);
static unsigned int sock_poll(struct file *file,
			      struct poll_table_struct *wait);
119
static long sock_ioctl(struct file *file, unsigned int cmd, unsigned long arg);
120 121
#ifdef CONFIG_COMPAT
static long compat_sock_ioctl(struct file *file,
122
			      unsigned int cmd, unsigned long arg);
123
#endif
L
Linus Torvalds 已提交
124 125 126
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
127
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
128
				struct pipe_inode_info *pipe, size_t len,
J
Jens Axboe 已提交
129
				unsigned int flags);
L
Linus Torvalds 已提交
130 131 132 133 134 135

/*
 *	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.
 */

136
static const struct file_operations socket_file_ops = {
L
Linus Torvalds 已提交
137 138 139 140 141 142
	.owner =	THIS_MODULE,
	.llseek =	no_llseek,
	.aio_read =	sock_aio_read,
	.aio_write =	sock_aio_write,
	.poll =		sock_poll,
	.unlocked_ioctl = sock_ioctl,
143 144 145
#ifdef CONFIG_COMPAT
	.compat_ioctl = compat_sock_ioctl,
#endif
L
Linus Torvalds 已提交
146 147 148 149
	.mmap =		sock_mmap,
	.open =		sock_no_open,	/* special open code to disallow open via /proc */
	.release =	sock_close,
	.fasync =	sock_fasync,
150 151
	.sendpage =	sock_sendpage,
	.splice_write = generic_splice_sendpage,
J
Jens Axboe 已提交
152
	.splice_read =	sock_splice_read,
L
Linus Torvalds 已提交
153 154 155 156 157 158 159
};

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

static DEFINE_SPINLOCK(net_family_lock);
E
Eric Dumazet 已提交
160
static const struct net_proto_family __rcu *net_families[NPROTO] __read_mostly;
L
Linus Torvalds 已提交
161 162 163 164 165

/*
 *	Statistics counters of the socket lists
 */

166
static DEFINE_PER_CPU(int, sockets_in_use);
L
Linus Torvalds 已提交
167 168

/*
169 170 171
 * 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
172 173 174 175 176 177 178 179 180 181 182 183 184
 */

/**
 *	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.
 */

185
int move_addr_to_kernel(void __user *uaddr, int ulen, struct sockaddr_storage *kaddr)
L
Linus Torvalds 已提交
186
{
187
	if (ulen < 0 || ulen > sizeof(struct sockaddr_storage))
L
Linus Torvalds 已提交
188
		return -EINVAL;
189
	if (ulen == 0)
L
Linus Torvalds 已提交
190
		return 0;
191
	if (copy_from_user(kaddr, uaddr, ulen))
L
Linus Torvalds 已提交
192
		return -EFAULT;
193
	return audit_sockaddr(ulen, kaddr);
L
Linus Torvalds 已提交
194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211
}

/**
 *	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.
 */
212

213
static int move_addr_to_user(struct sockaddr_storage *kaddr, int klen,
S
stephen hemminger 已提交
214
			     void __user *uaddr, int __user *ulen)
L
Linus Torvalds 已提交
215 216 217 218
{
	int err;
	int len;

219 220
	err = get_user(len, ulen);
	if (err)
L
Linus Torvalds 已提交
221
		return err;
222 223
	if (len > klen)
		len = klen;
224
	if (len < 0 || len > sizeof(struct sockaddr_storage))
L
Linus Torvalds 已提交
225
		return -EINVAL;
226
	if (len) {
S
Steve Grubb 已提交
227 228
		if (audit_sockaddr(klen, kaddr))
			return -ENOMEM;
229
		if (copy_to_user(uaddr, kaddr, len))
L
Linus Torvalds 已提交
230 231 232
			return -EFAULT;
	}
	/*
233 234
	 *      "fromlen shall refer to the value before truncation.."
	 *                      1003.1g
L
Linus Torvalds 已提交
235 236 237 238
	 */
	return __put_user(klen, ulen);
}

239
static struct kmem_cache *sock_inode_cachep __read_mostly;
L
Linus Torvalds 已提交
240 241 242 243

static struct inode *sock_alloc_inode(struct super_block *sb)
{
	struct socket_alloc *ei;
244
	struct socket_wq *wq;
245

246
	ei = kmem_cache_alloc(sock_inode_cachep, GFP_KERNEL);
L
Linus Torvalds 已提交
247 248
	if (!ei)
		return NULL;
249 250
	wq = kmalloc(sizeof(*wq), GFP_KERNEL);
	if (!wq) {
251 252 253
		kmem_cache_free(sock_inode_cachep, ei);
		return NULL;
	}
254 255 256
	init_waitqueue_head(&wq->wait);
	wq->fasync_list = NULL;
	RCU_INIT_POINTER(ei->socket.wq, wq);
257

L
Linus Torvalds 已提交
258 259 260 261 262 263 264 265 266 267 268
	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)
{
269
	struct socket_alloc *ei;
270
	struct socket_wq *wq;
271 272

	ei = container_of(inode, struct socket_alloc, vfs_inode);
273
	wq = rcu_dereference_protected(ei->socket.wq, 1);
274
	kfree_rcu(wq, rcu);
275
	kmem_cache_free(sock_inode_cachep, ei);
L
Linus Torvalds 已提交
276 277
}

278
static void init_once(void *foo)
L
Linus Torvalds 已提交
279
{
280
	struct socket_alloc *ei = (struct socket_alloc *)foo;
L
Linus Torvalds 已提交
281

C
Christoph Lameter 已提交
282
	inode_init_once(&ei->vfs_inode);
L
Linus Torvalds 已提交
283
}
284

L
Linus Torvalds 已提交
285 286 287
static int init_inodecache(void)
{
	sock_inode_cachep = kmem_cache_create("sock_inode_cache",
288 289 290 291 292
					      sizeof(struct socket_alloc),
					      0,
					      (SLAB_HWCACHE_ALIGN |
					       SLAB_RECLAIM_ACCOUNT |
					       SLAB_MEM_SPREAD),
293
					      init_once);
L
Linus Torvalds 已提交
294 295 296 297 298
	if (sock_inode_cachep == NULL)
		return -ENOMEM;
	return 0;
}

299
static const struct super_operations sockfs_ops = {
300 301 302
	.alloc_inode	= sock_alloc_inode,
	.destroy_inode	= sock_destroy_inode,
	.statfs		= simple_statfs,
L
Linus Torvalds 已提交
303 304
};

305 306 307 308 309 310 311 312 313
/*
 * 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
314
static const struct dentry_operations sockfs_dentry_operations = {
315
	.d_dname  = sockfs_dname,
L
Linus Torvalds 已提交
316 317
};

318 319 320 321 322 323 324 325 326 327 328 329 330 331 332
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
333 334 335
/*
 *	Obtains the first available file descriptor and sets it up for use.
 *
336 337
 *	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
338 339 340 341 342 343 344 345 346 347 348 349
 *	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.
 */

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

356 357 358 359 360 361 362
	if (dname) {
		name.name = dname;
		name.len = strlen(name.name);
	} else if (sock->sk) {
		name.name = sock->sk->sk_prot_creator->name;
		name.len = strlen(name.name);
	}
363
	path.dentry = d_alloc_pseudo(sock_mnt->mnt_sb, &name);
364 365
	if (unlikely(!path.dentry))
		return ERR_PTR(-ENOMEM);
366
	path.mnt = mntget(sock_mnt);
367

368
	d_instantiate(path.dentry, SOCK_INODE(sock));
A
Al Viro 已提交
369
	SOCK_INODE(sock)->i_fop = &socket_file_ops;
370

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

	sock->file = file;
381
	file->f_flags = O_RDWR | (flags & O_NONBLOCK);
382 383
	file->f_pos = 0;
	file->private_data = sock;
384
	return file;
385
}
386
EXPORT_SYMBOL(sock_alloc_file);
387

388
static int sock_map_fd(struct socket *sock, int flags)
389 390
{
	struct file *newfile;
391 392 393
	int fd = get_unused_fd_flags(flags);
	if (unlikely(fd < 0))
		return fd;
394

395
	newfile = sock_alloc_file(sock, flags, NULL);
396
	if (likely(!IS_ERR(newfile))) {
397
		fd_install(fd, newfile);
398 399
		return fd;
	}
400

401 402
	put_unused_fd(fd);
	return PTR_ERR(newfile);
L
Linus Torvalds 已提交
403 404
}

405
struct socket *sock_from_file(struct file *file, int *err)
406 407 408 409
{
	if (file->f_op == &socket_file_ops)
		return file->private_data;	/* set in sock_map_fd */

E
Eric Dumazet 已提交
410 411
	*err = -ENOTSOCK;
	return NULL;
412
}
413
EXPORT_SYMBOL(sock_from_file);
414

L
Linus Torvalds 已提交
415
/**
416
 *	sockfd_lookup - Go from a file number to its socket slot
L
Linus Torvalds 已提交
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432
 *	@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;

433 434
	file = fget(fd);
	if (!file) {
L
Linus Torvalds 已提交
435 436 437
		*err = -EBADF;
		return NULL;
	}
438

439 440
	sock = sock_from_file(file, err);
	if (!sock)
L
Linus Torvalds 已提交
441
		fput(file);
442 443
	return sock;
}
444
EXPORT_SYMBOL(sockfd_lookup);
L
Linus Torvalds 已提交
445

446 447 448 449 450
static struct socket *sockfd_lookup_light(int fd, int *err, int *fput_needed)
{
	struct file *file;
	struct socket *sock;

451
	*err = -EBADF;
452 453 454 455 456 457
	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
458
	}
459
	return NULL;
L
Linus Torvalds 已提交
460 461
}

462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523
#define XATTR_SOCKPROTONAME_SUFFIX "sockprotoname"
#define XATTR_NAME_SOCKPROTONAME (XATTR_SYSTEM_PREFIX XATTR_SOCKPROTONAME_SUFFIX)
#define XATTR_NAME_SOCKPROTONAME_LEN (sizeof(XATTR_NAME_SOCKPROTONAME)-1)
static ssize_t sockfs_getxattr(struct dentry *dentry,
			       const char *name, void *value, size_t size)
{
	const char *proto_name;
	size_t proto_size;
	int error;

	error = -ENODATA;
	if (!strncmp(name, XATTR_NAME_SOCKPROTONAME, XATTR_NAME_SOCKPROTONAME_LEN)) {
		proto_name = dentry->d_name.name;
		proto_size = strlen(proto_name);

		if (value) {
			error = -ERANGE;
			if (proto_size + 1 > size)
				goto out;

			strncpy(value, proto_name, proto_size + 1);
		}
		error = proto_size + 1;
	}

out:
	return error;
}

static ssize_t sockfs_listxattr(struct dentry *dentry, char *buffer,
				size_t size)
{
	ssize_t len;
	ssize_t used = 0;

	len = security_inode_listsecurity(dentry->d_inode, buffer, size);
	if (len < 0)
		return len;
	used += len;
	if (buffer) {
		if (size < used)
			return -ERANGE;
		buffer += len;
	}

	len = (XATTR_NAME_SOCKPROTONAME_LEN + 1);
	used += len;
	if (buffer) {
		if (size < used)
			return -ERANGE;
		memcpy(buffer, XATTR_NAME_SOCKPROTONAME, len);
		buffer += len;
	}

	return used;
}

static const struct inode_operations sockfs_inode_ops = {
	.getxattr = sockfs_getxattr,
	.listxattr = sockfs_listxattr,
};

L
Linus Torvalds 已提交
524 525
/**
 *	sock_alloc	-	allocate a socket
526
 *
L
Linus Torvalds 已提交
527 528 529 530 531 532 533
 *	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)
{
534 535
	struct inode *inode;
	struct socket *sock;
L
Linus Torvalds 已提交
536

537
	inode = new_inode_pseudo(sock_mnt->mnt_sb);
L
Linus Torvalds 已提交
538 539 540 541 542
	if (!inode)
		return NULL;

	sock = SOCKET_I(inode);

543
	kmemcheck_annotate_bitfield(sock, type);
544
	inode->i_ino = get_next_ino();
545
	inode->i_mode = S_IFSOCK | S_IRWXUGO;
546 547
	inode->i_uid = current_fsuid();
	inode->i_gid = current_fsgid();
548
	inode->i_op = &sockfs_inode_ops;
L
Linus Torvalds 已提交
549

550
	this_cpu_add(sockets_in_use, 1);
L
Linus Torvalds 已提交
551 552 553 554 555 556 557 558
	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.
 */
559

L
Linus Torvalds 已提交
560 561 562 563 564
static int sock_no_open(struct inode *irrelevant, struct file *dontcare)
{
	return -ENXIO;
}

565
const struct file_operations bad_sock_fops = {
L
Linus Torvalds 已提交
566 567
	.owner = THIS_MODULE,
	.open = sock_no_open,
568
	.llseek = noop_llseek,
L
Linus Torvalds 已提交
569 570 571 572 573 574 575 576
};

/**
 *	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
577
 *	an inode not a file.
L
Linus Torvalds 已提交
578
 */
579

L
Linus Torvalds 已提交
580 581 582 583 584 585 586 587 588 589
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);
	}

590
	if (rcu_dereference_protected(sock->wq, 1)->fasync_list)
L
Linus Torvalds 已提交
591 592
		printk(KERN_ERR "sock_release: fasync list not empty!\n");

593 594 595
	if (test_bit(SOCK_EXTERNALLY_ALLOCATED, &sock->flags))
		return;

596
	this_cpu_sub(sockets_in_use, 1);
L
Linus Torvalds 已提交
597 598 599 600
	if (!sock->file) {
		iput(SOCK_INODE(sock));
		return;
	}
601
	sock->file = NULL;
L
Linus Torvalds 已提交
602
}
603
EXPORT_SYMBOL(sock_release);
L
Linus Torvalds 已提交
604

605
int sock_tx_timestamp(struct sock *sk, __u8 *tx_flags)
606
{
607
	*tx_flags = 0;
608
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_HARDWARE))
609
		*tx_flags |= SKBTX_HW_TSTAMP;
610
	if (sock_flag(sk, SOCK_TIMESTAMPING_TX_SOFTWARE))
611
		*tx_flags |= SKBTX_SW_TSTAMP;
612 613
	if (sock_flag(sk, SOCK_WIFI_STATUS))
		*tx_flags |= SKBTX_WIFI_STATUS;
614 615 616 617
	return 0;
}
EXPORT_SYMBOL(sock_tx_timestamp);

618 619
static inline int __sock_sendmsg_nosec(struct kiocb *iocb, struct socket *sock,
				       struct msghdr *msg, size_t size)
L
Linus Torvalds 已提交
620 621 622 623 624 625 626 627 628 629 630
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);

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

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

631 632 633 634 635 636 637 638
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
639 640 641 642 643 644 645 646 647 648 649 650 651
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;
}
652
EXPORT_SYMBOL(sock_sendmsg);
L
Linus Torvalds 已提交
653

654
static int sock_sendmsg_nosec(struct socket *sock, struct msghdr *msg, size_t size)
655 656 657 658 659 660 661 662 663 664 665 666 667
{
	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
668 669 670 671 672 673 674 675 676 677 678
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
	 */
679
	msg->msg_iov = (struct iovec *)vec;
L
Linus Torvalds 已提交
680 681 682 683 684
	msg->msg_iovlen = num;
	result = sock_sendmsg(sock, msg, size);
	set_fs(oldfs);
	return result;
}
685
EXPORT_SYMBOL(kernel_sendmsg);
L
Linus Torvalds 已提交
686

687 688 689 690 691 692 693 694 695 696
static int ktime2ts(ktime_t kt, struct timespec *ts)
{
	if (kt.tv64) {
		*ts = ktime_to_timespec(kt);
		return 1;
	} else {
		return 0;
	}
}

697 698 699 700 701 702
/*
 * 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)
{
703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
	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 {
721
			skb_get_timestampns(skb, &ts[0]);
722
			put_cmsg(msg, SOL_SOCKET, SCM_TIMESTAMPNS,
723
				 sizeof(ts[0]), &ts[0]);
724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740
		}
	}


	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;
741
	}
742 743 744
	if (!empty)
		put_cmsg(msg, SOL_SOCKET,
			 SCM_TIMESTAMPING, sizeof(ts), &ts);
745
}
746 747
EXPORT_SYMBOL_GPL(__sock_recv_timestamp);

748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763
void __sock_recv_wifi_status(struct msghdr *msg, struct sock *sk,
	struct sk_buff *skb)
{
	int ack;

	if (!sock_flag(sk, SOCK_WIFI_STATUS))
		return;
	if (!skb->wifi_acked_valid)
		return;

	ack = skb->wifi_acked;

	put_cmsg(msg, SOL_SOCKET, SCM_WIFI_STATUS, sizeof(ack), &ack);
}
EXPORT_SYMBOL_GPL(__sock_recv_wifi_status);

S
stephen hemminger 已提交
764 765
static inline void sock_recv_drops(struct msghdr *msg, struct sock *sk,
				   struct sk_buff *skb)
766 767 768 769 770 771
{
	if (sock_flag(sk, SOCK_RXQ_OVFL) && skb && skb->dropcount)
		put_cmsg(msg, SOL_SOCKET, SO_RXQ_OVFL,
			sizeof(__u32), &skb->dropcount);
}

772
void __sock_recv_ts_and_drops(struct msghdr *msg, struct sock *sk,
773 774 775 776 777
	struct sk_buff *skb)
{
	sock_recv_timestamp(msg, sk, skb);
	sock_recv_drops(msg, sk, skb);
}
778
EXPORT_SYMBOL_GPL(__sock_recv_ts_and_drops);
779

780 781
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
782 783 784 785 786 787 788 789 790 791 792 793
{
	struct sock_iocb *si = kiocb_to_siocb(iocb);

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

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

794 795 796 797 798 799 800 801
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);
}

802
int sock_recvmsg(struct socket *sock, struct msghdr *msg,
L
Linus Torvalds 已提交
803 804 805 806 807 808
		 size_t size, int flags)
{
	struct kiocb iocb;
	struct sock_iocb siocb;
	int ret;

809
	init_sync_kiocb(&iocb, NULL);
L
Linus Torvalds 已提交
810 811 812 813 814 815
	iocb.private = &siocb;
	ret = __sock_recvmsg(&iocb, sock, msg, size, flags);
	if (-EIOCBQUEUED == ret)
		ret = wait_on_sync_kiocb(&iocb);
	return ret;
}
816
EXPORT_SYMBOL(sock_recvmsg);
L
Linus Torvalds 已提交
817

818 819 820 821 822 823 824 825 826 827 828 829 830 831 832
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;
}

833 834 835 836 837 838 839 840 841 842 843 844 845 846 847
/**
 * 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.
 */
848 849
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
850 851 852 853 854 855 856 857 858
{
	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
	 */
859
	msg->msg_iov = (struct iovec *)vec, msg->msg_iovlen = num;
L
Linus Torvalds 已提交
860 861 862 863
	result = sock_recvmsg(sock, msg, size, flags);
	set_fs(oldfs);
	return result;
}
864
EXPORT_SYMBOL(kernel_recvmsg);
L
Linus Torvalds 已提交
865 866 867 868 869 870

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

871 872
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
873 874 875 876
{
	struct socket *sock;
	int flags;

877 878
	sock = file->private_data;

879 880 881
	flags = (file->f_flags & O_NONBLOCK) ? MSG_DONTWAIT : 0;
	/* more is a combination of MSG_MORE and MSG_SENDPAGE_NOTLAST */
	flags |= more;
882

883
	return kernel_sendpage(sock, page, offset, size, flags);
884
}
L
Linus Torvalds 已提交
885

J
Jens Axboe 已提交
886
static ssize_t sock_splice_read(struct file *file, loff_t *ppos,
887
				struct pipe_inode_info *pipe, size_t len,
J
Jens Axboe 已提交
888 889 890 891
				unsigned int flags)
{
	struct socket *sock = file->private_data;

892 893 894
	if (unlikely(!sock->ops->splice_read))
		return -EINVAL;

J
Jens Axboe 已提交
895 896 897
	return sock->ops->splice_read(sock, ppos, pipe, len, flags);
}

898
static struct sock_iocb *alloc_sock_iocb(struct kiocb *iocb,
899
					 struct sock_iocb *siocb)
900 901 902 903 904
{
	if (!is_sync_kiocb(iocb)) {
		siocb = kmalloc(sizeof(*siocb), GFP_KERNEL);
		if (!siocb)
			return NULL;
L
Linus Torvalds 已提交
905 906 907
		iocb->ki_dtor = sock_aio_dtor;
	}

908 909 910
	siocb->kiocb = iocb;
	iocb->private = siocb;
	return siocb;
L
Linus Torvalds 已提交
911 912
}

913
static ssize_t do_sock_read(struct msghdr *msg, struct kiocb *iocb,
914 915
		struct file *file, const struct iovec *iov,
		unsigned long nr_segs)
916 917 918 919
{
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
920

921 922
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
923

924 925 926 927
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
928
	msg->msg_iov = (struct iovec *)iov;
929 930 931 932 933 934
	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);
}

935 936
static ssize_t sock_aio_read(struct kiocb *iocb, const struct iovec *iov,
				unsigned long nr_segs, loff_t pos)
937 938 939
{
	struct sock_iocb siocb, *x;

L
Linus Torvalds 已提交
940 941
	if (pos != 0)
		return -ESPIPE;
942 943

	if (iocb->ki_left == 0)	/* Match SYS5 behaviour */
L
Linus Torvalds 已提交
944 945
		return 0;

946 947

	x = alloc_sock_iocb(iocb, &siocb);
948 949
	if (!x)
		return -ENOMEM;
950
	return do_sock_read(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
951 952
}

953
static ssize_t do_sock_write(struct msghdr *msg, struct kiocb *iocb,
954 955
			struct file *file, const struct iovec *iov,
			unsigned long nr_segs)
L
Linus Torvalds 已提交
956
{
957 958 959
	struct socket *sock = file->private_data;
	size_t size = 0;
	int i;
L
Linus Torvalds 已提交
960

961 962
	for (i = 0; i < nr_segs; i++)
		size += iov[i].iov_len;
L
Linus Torvalds 已提交
963

964 965 966 967
	msg->msg_name = NULL;
	msg->msg_namelen = 0;
	msg->msg_control = NULL;
	msg->msg_controllen = 0;
968
	msg->msg_iov = (struct iovec *)iov;
969 970 971 972
	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
973

974
	return __sock_sendmsg(iocb, sock, msg, size);
L
Linus Torvalds 已提交
975 976
}

977 978
static ssize_t sock_aio_write(struct kiocb *iocb, const struct iovec *iov,
			  unsigned long nr_segs, loff_t pos)
979 980
{
	struct sock_iocb siocb, *x;
L
Linus Torvalds 已提交
981

982 983
	if (pos != 0)
		return -ESPIPE;
984 985

	x = alloc_sock_iocb(iocb, &siocb);
986 987
	if (!x)
		return -ENOMEM;
L
Linus Torvalds 已提交
988

989
	return do_sock_write(&x->async_msg, iocb, iocb->ki_filp, iov, nr_segs);
L
Linus Torvalds 已提交
990 991 992 993 994 995 996
}

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

A
Arjan van de Ven 已提交
997
static DEFINE_MUTEX(br_ioctl_mutex);
998
static int (*br_ioctl_hook) (struct net *, unsigned int cmd, void __user *arg);
L
Linus Torvalds 已提交
999

1000
void brioctl_set(int (*hook) (struct net *, unsigned int, void __user *))
L
Linus Torvalds 已提交
1001
{
A
Arjan van de Ven 已提交
1002
	mutex_lock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
1003
	br_ioctl_hook = hook;
A
Arjan van de Ven 已提交
1004
	mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
1005 1006 1007
}
EXPORT_SYMBOL(brioctl_set);

A
Arjan van de Ven 已提交
1008
static DEFINE_MUTEX(vlan_ioctl_mutex);
1009
static int (*vlan_ioctl_hook) (struct net *, void __user *arg);
L
Linus Torvalds 已提交
1010

1011
void vlan_ioctl_set(int (*hook) (struct net *, void __user *))
L
Linus Torvalds 已提交
1012
{
A
Arjan van de Ven 已提交
1013
	mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1014
	vlan_ioctl_hook = hook;
A
Arjan van de Ven 已提交
1015
	mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1016 1017 1018
}
EXPORT_SYMBOL(vlan_ioctl_set);

A
Arjan van de Ven 已提交
1019
static DEFINE_MUTEX(dlci_ioctl_mutex);
1020
static int (*dlci_ioctl_hook) (unsigned int, void __user *);
L
Linus Torvalds 已提交
1021

1022
void dlci_ioctl_set(int (*hook) (unsigned int, void __user *))
L
Linus Torvalds 已提交
1023
{
A
Arjan van de Ven 已提交
1024
	mutex_lock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
1025
	dlci_ioctl_hook = hook;
A
Arjan van de Ven 已提交
1026
	mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
1027 1028 1029
}
EXPORT_SYMBOL(dlci_ioctl_set);

1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047
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
1048 1049 1050 1051 1052 1053 1054 1055
/*
 *	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;
1056
	struct sock *sk;
L
Linus Torvalds 已提交
1057 1058
	void __user *argp = (void __user *)arg;
	int pid, err;
1059
	struct net *net;
L
Linus Torvalds 已提交
1060

1061
	sock = file->private_data;
1062
	sk = sock->sk;
1063
	net = sock_net(sk);
L
Linus Torvalds 已提交
1064
	if (cmd >= SIOCDEVPRIVATE && cmd <= (SIOCDEVPRIVATE + 15)) {
1065
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
1066
	} else
J
Johannes Berg 已提交
1067
#ifdef CONFIG_WEXT_CORE
L
Linus Torvalds 已提交
1068
	if (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST) {
1069
		err = dev_ioctl(net, cmd, argp);
L
Linus Torvalds 已提交
1070
	} else
J
Johannes Berg 已提交
1071
#endif
1072
		switch (cmd) {
L
Linus Torvalds 已提交
1073 1074 1075 1076 1077 1078 1079 1080 1081
		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:
1082
			err = put_user(f_getown(sock->file),
1083
				       (int __user *)argp);
L
Linus Torvalds 已提交
1084 1085 1086 1087 1088 1089 1090 1091 1092
			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
1093
			mutex_lock(&br_ioctl_mutex);
1094
			if (br_ioctl_hook)
1095
				err = br_ioctl_hook(net, cmd, argp);
A
Arjan van de Ven 已提交
1096
			mutex_unlock(&br_ioctl_mutex);
L
Linus Torvalds 已提交
1097 1098 1099 1100 1101 1102 1103
			break;
		case SIOCGIFVLAN:
		case SIOCSIFVLAN:
			err = -ENOPKG;
			if (!vlan_ioctl_hook)
				request_module("8021q");

A
Arjan van de Ven 已提交
1104
			mutex_lock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1105
			if (vlan_ioctl_hook)
1106
				err = vlan_ioctl_hook(net, argp);
A
Arjan van de Ven 已提交
1107
			mutex_unlock(&vlan_ioctl_mutex);
L
Linus Torvalds 已提交
1108 1109 1110 1111 1112 1113 1114
			break;
		case SIOCADDDLCI:
		case SIOCDELDLCI:
			err = -ENOPKG;
			if (!dlci_ioctl_hook)
				request_module("dlci");

1115 1116
			mutex_lock(&dlci_ioctl_mutex);
			if (dlci_ioctl_hook)
L
Linus Torvalds 已提交
1117
				err = dlci_ioctl_hook(cmd, argp);
1118
			mutex_unlock(&dlci_ioctl_mutex);
L
Linus Torvalds 已提交
1119 1120
			break;
		default:
1121
			err = sock_do_ioctl(net, sock, cmd, arg);
L
Linus Torvalds 已提交
1122
			break;
1123
		}
L
Linus Torvalds 已提交
1124 1125 1126 1127 1128 1129 1130
	return err;
}

int sock_create_lite(int family, int type, int protocol, struct socket **res)
{
	int err;
	struct socket *sock = NULL;
1131

L
Linus Torvalds 已提交
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142
	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
1143 1144 1145 1146
	err = security_socket_post_create(sock, family, type, protocol, 1);
	if (err)
		goto out_release;

L
Linus Torvalds 已提交
1147 1148 1149
out:
	*res = sock;
	return err;
V
Venkat Yekkirala 已提交
1150 1151 1152 1153
out_release:
	sock_release(sock);
	sock = NULL;
	goto out;
L
Linus Torvalds 已提交
1154
}
1155
EXPORT_SYMBOL(sock_create_lite);
L
Linus Torvalds 已提交
1156 1157

/* No kernel lock held - perfect */
1158
static unsigned int sock_poll(struct file *file, poll_table *wait)
L
Linus Torvalds 已提交
1159 1160 1161 1162
{
	struct socket *sock;

	/*
1163
	 *      We can't return errors to poll, so it's either yes or no.
L
Linus Torvalds 已提交
1164
	 */
1165
	sock = file->private_data;
L
Linus Torvalds 已提交
1166 1167 1168
	return sock->ops->poll(file, sock, wait);
}

1169
static int sock_mmap(struct file *file, struct vm_area_struct *vma)
L
Linus Torvalds 已提交
1170
{
1171
	struct socket *sock = file->private_data;
L
Linus Torvalds 已提交
1172 1173 1174 1175

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

1176
static int sock_close(struct inode *inode, struct file *filp)
L
Linus Torvalds 已提交
1177 1178
{
	/*
1179 1180
	 *      It was possible the inode is NULL we were
	 *      closing an unfinished socket.
L
Linus Torvalds 已提交
1181 1182
	 */

1183
	if (!inode) {
L
Linus Torvalds 已提交
1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198
		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)
1199
 *	   or under socket lock
L
Linus Torvalds 已提交
1200 1201 1202 1203
 */

static int sock_fasync(int fd, struct file *filp, int on)
{
1204 1205
	struct socket *sock = filp->private_data;
	struct sock *sk = sock->sk;
1206
	struct socket_wq *wq;
L
Linus Torvalds 已提交
1207

1208
	if (sk == NULL)
L
Linus Torvalds 已提交
1209 1210 1211
		return -EINVAL;

	lock_sock(sk);
1212 1213
	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
1214

1215
	if (!wq->fasync_list)
1216 1217
		sock_reset_flag(sk, SOCK_FASYNC);
	else
E
Eric Dumazet 已提交
1218
		sock_set_flag(sk, SOCK_FASYNC);
L
Linus Torvalds 已提交
1219

1220
	release_sock(sk);
L
Linus Torvalds 已提交
1221 1222 1223
	return 0;
}

1224
/* This function may be called only under socket lock or callback_lock or rcu_lock */
L
Linus Torvalds 已提交
1225 1226 1227

int sock_wake_async(struct socket *sock, int how, int band)
{
1228 1229 1230 1231 1232 1233 1234 1235
	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
1236
		return -1;
1237
	}
1238
	switch (how) {
1239
	case SOCK_WAKE_WAITD:
L
Linus Torvalds 已提交
1240 1241 1242
		if (test_bit(SOCK_ASYNC_WAITDATA, &sock->flags))
			break;
		goto call_kill;
1243
	case SOCK_WAKE_SPACE:
L
Linus Torvalds 已提交
1244 1245 1246
		if (!test_and_clear_bit(SOCK_ASYNC_NOSPACE, &sock->flags))
			break;
		/* fall through */
1247
	case SOCK_WAKE_IO:
1248
call_kill:
1249
		kill_fasync(&wq->fasync_list, SIGIO, band);
L
Linus Torvalds 已提交
1250
		break;
1251
	case SOCK_WAKE_URG:
1252
		kill_fasync(&wq->fasync_list, SIGURG, band);
L
Linus Torvalds 已提交
1253
	}
1254
	rcu_read_unlock();
L
Linus Torvalds 已提交
1255 1256
	return 0;
}
1257
EXPORT_SYMBOL(sock_wake_async);
L
Linus Torvalds 已提交
1258

P
Pavel Emelyanov 已提交
1259
int __sock_create(struct net *net, int family, int type, int protocol,
1260
			 struct socket **res, int kern)
L
Linus Torvalds 已提交
1261 1262 1263
{
	int err;
	struct socket *sock;
1264
	const struct net_proto_family *pf;
L
Linus Torvalds 已提交
1265 1266

	/*
1267
	 *      Check protocol is in range
L
Linus Torvalds 已提交
1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
	 */
	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) {
1280
		static int warned;
L
Linus Torvalds 已提交
1281 1282
		if (!warned) {
			warned = 1;
1283 1284
			printk(KERN_INFO "%s uses obsolete (PF_INET,SOCK_PACKET)\n",
			       current->comm);
L
Linus Torvalds 已提交
1285 1286 1287 1288 1289 1290 1291
		}
		family = PF_PACKET;
	}

	err = security_socket_create(family, type, protocol, kern);
	if (err)
		return err;
1292

1293 1294 1295 1296 1297 1298 1299
	/*
	 *	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) {
1300
		net_warn_ratelimited("socket: no more sockets\n");
1301 1302 1303 1304 1305 1306
		return -ENFILE;	/* Not exactly a match, but its the
				   closest posix thing */
	}

	sock->type = type;

1307
#ifdef CONFIG_MODULES
1308 1309 1310
	/* 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
1311 1312 1313
	 * requested real, full-featured networking support upon configuration.
	 * Otherwise module support will break!
	 */
E
Eric Dumazet 已提交
1314
	if (rcu_access_pointer(net_families[family]) == NULL)
1315
		request_module("net-pf-%d", family);
L
Linus Torvalds 已提交
1316 1317
#endif

1318 1319 1320 1321 1322
	rcu_read_lock();
	pf = rcu_dereference(net_families[family]);
	err = -EAFNOSUPPORT;
	if (!pf)
		goto out_release;
L
Linus Torvalds 已提交
1323 1324 1325 1326 1327

	/*
	 * We will call the ->create function, that possibly is in a loadable
	 * module, so we have to bump that loadable module refcnt first.
	 */
1328
	if (!try_module_get(pf->owner))
L
Linus Torvalds 已提交
1329 1330
		goto out_release;

1331 1332 1333
	/* Now protected by module ref count */
	rcu_read_unlock();

1334
	err = pf->create(net, sock, protocol, kern);
1335
	if (err < 0)
L
Linus Torvalds 已提交
1336
		goto out_module_put;
1337

L
Linus Torvalds 已提交
1338 1339 1340 1341
	/*
	 * Now to bump the refcnt of the [loadable] module that owns this
	 * socket at sock_release time we decrement its refcnt.
	 */
1342 1343 1344
	if (!try_module_get(sock->ops->owner))
		goto out_module_busy;

L
Linus Torvalds 已提交
1345 1346 1347 1348
	/*
	 * Now that we're done with the ->create function, the [loadable]
	 * module can have its refcnt decremented
	 */
1349
	module_put(pf->owner);
V
Venkat Yekkirala 已提交
1350 1351
	err = security_socket_post_create(sock, family, type, protocol, kern);
	if (err)
1352
		goto out_sock_release;
1353
	*res = sock;
L
Linus Torvalds 已提交
1354

1355 1356 1357 1358
	return 0;

out_module_busy:
	err = -EAFNOSUPPORT;
L
Linus Torvalds 已提交
1359
out_module_put:
1360 1361 1362
	sock->ops = NULL;
	module_put(pf->owner);
out_sock_release:
L
Linus Torvalds 已提交
1363
	sock_release(sock);
1364 1365 1366 1367 1368
	return err;

out_release:
	rcu_read_unlock();
	goto out_sock_release;
L
Linus Torvalds 已提交
1369
}
P
Pavel Emelyanov 已提交
1370
EXPORT_SYMBOL(__sock_create);
L
Linus Torvalds 已提交
1371 1372 1373

int sock_create(int family, int type, int protocol, struct socket **res)
{
1374
	return __sock_create(current->nsproxy->net_ns, family, type, protocol, res, 0);
L
Linus Torvalds 已提交
1375
}
1376
EXPORT_SYMBOL(sock_create);
L
Linus Torvalds 已提交
1377 1378 1379

int sock_create_kern(int family, int type, int protocol, struct socket **res)
{
1380
	return __sock_create(&init_net, family, type, protocol, res, 1);
L
Linus Torvalds 已提交
1381
}
1382
EXPORT_SYMBOL(sock_create_kern);
L
Linus Torvalds 已提交
1383

1384
SYSCALL_DEFINE3(socket, int, family, int, type, int, protocol)
L
Linus Torvalds 已提交
1385 1386 1387
{
	int retval;
	struct socket *sock;
1388 1389
	int flags;

1390 1391 1392 1393 1394 1395
	/* 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);

1396
	flags = type & ~SOCK_TYPE_MASK;
1397
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1398 1399
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1400

U
Ulrich Drepper 已提交
1401 1402 1403
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1404 1405 1406 1407
	retval = sock_create(family, type, protocol, &sock);
	if (retval < 0)
		goto out;

1408
	retval = sock_map_fd(sock, flags & (O_CLOEXEC | O_NONBLOCK));
L
Linus Torvalds 已提交
1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424
	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.
 */

1425 1426
SYSCALL_DEFINE4(socketpair, int, family, int, type, int, protocol,
		int __user *, usockvec)
L
Linus Torvalds 已提交
1427 1428 1429
{
	struct socket *sock1, *sock2;
	int fd1, fd2, err;
A
Al Viro 已提交
1430
	struct file *newfile1, *newfile2;
1431 1432 1433
	int flags;

	flags = type & ~SOCK_TYPE_MASK;
1434
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
1435 1436
		return -EINVAL;
	type &= SOCK_TYPE_MASK;
L
Linus Torvalds 已提交
1437

U
Ulrich Drepper 已提交
1438 1439 1440
	if (SOCK_NONBLOCK != O_NONBLOCK && (flags & SOCK_NONBLOCK))
		flags = (flags & ~SOCK_NONBLOCK) | O_NONBLOCK;

L
Linus Torvalds 已提交
1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454
	/*
	 * 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);
1455
	if (err < 0)
L
Linus Torvalds 已提交
1456 1457
		goto out_release_both;

1458
	fd1 = get_unused_fd_flags(flags);
1459 1460
	if (unlikely(fd1 < 0)) {
		err = fd1;
A
Al Viro 已提交
1461
		goto out_release_both;
1462
	}
1463
	fd2 = get_unused_fd_flags(flags);
1464 1465
	if (unlikely(fd2 < 0)) {
		err = fd2;
1466 1467 1468 1469
		put_unused_fd(fd1);
		goto out_release_both;
	}

1470
	newfile1 = sock_alloc_file(sock1, flags, NULL);
1471 1472 1473 1474 1475 1476 1477
	if (unlikely(IS_ERR(newfile1))) {
		err = PTR_ERR(newfile1);
		put_unused_fd(fd1);
		put_unused_fd(fd2);
		goto out_release_both;
	}

1478
	newfile2 = sock_alloc_file(sock2, flags, NULL);
1479 1480
	if (IS_ERR(newfile2)) {
		err = PTR_ERR(newfile2);
1481 1482
		fput(newfile1);
		put_unused_fd(fd1);
1483
		put_unused_fd(fd2);
1484 1485
		sock_release(sock2);
		goto out;
A
Al Viro 已提交
1486 1487
	}

A
Al Viro 已提交
1488
	audit_fd_pair(fd1, fd2);
A
Al Viro 已提交
1489 1490
	fd_install(fd1, newfile1);
	fd_install(fd2, newfile2);
L
Linus Torvalds 已提交
1491 1492 1493 1494
	/* fd1 and fd2 may be already another descriptors.
	 * Not kernel problem.
	 */

1495
	err = put_user(fd1, &usockvec[0]);
L
Linus Torvalds 已提交
1496 1497 1498 1499 1500 1501 1502 1503 1504 1505
	if (!err)
		err = put_user(fd2, &usockvec[1]);
	if (!err)
		return 0;

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

out_release_both:
1506
	sock_release(sock2);
L
Linus Torvalds 已提交
1507
out_release_1:
1508
	sock_release(sock1);
L
Linus Torvalds 已提交
1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520
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).
 */

1521
SYSCALL_DEFINE3(bind, int, fd, struct sockaddr __user *, umyaddr, int, addrlen)
L
Linus Torvalds 已提交
1522 1523
{
	struct socket *sock;
1524
	struct sockaddr_storage address;
1525
	int err, fput_needed;
L
Linus Torvalds 已提交
1526

1527
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
1528
	if (sock) {
1529
		err = move_addr_to_kernel(umyaddr, addrlen, &address);
1530 1531
		if (err >= 0) {
			err = security_socket_bind(sock,
1532
						   (struct sockaddr *)&address,
1533
						   addrlen);
1534 1535
			if (!err)
				err = sock->ops->bind(sock,
1536
						      (struct sockaddr *)
1537
						      &address, addrlen);
L
Linus Torvalds 已提交
1538
		}
1539
		fput_light(sock->file, fput_needed);
1540
	}
L
Linus Torvalds 已提交
1541 1542 1543 1544 1545 1546 1547 1548 1549
	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.
 */

1550
SYSCALL_DEFINE2(listen, int, fd, int, backlog)
L
Linus Torvalds 已提交
1551 1552
{
	struct socket *sock;
1553
	int err, fput_needed;
1554
	int somaxconn;
1555 1556 1557

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock) {
1558
		somaxconn = sock_net(sock->sk)->core.sysctl_somaxconn;
1559
		if ((unsigned int)backlog > somaxconn)
1560
			backlog = somaxconn;
L
Linus Torvalds 已提交
1561 1562

		err = security_socket_listen(sock, backlog);
1563 1564
		if (!err)
			err = sock->ops->listen(sock, backlog);
L
Linus Torvalds 已提交
1565

1566
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582
	}
	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.
 */

1583 1584
SYSCALL_DEFINE4(accept4, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen, int, flags)
L
Linus Torvalds 已提交
1585 1586
{
	struct socket *sock, *newsock;
1587
	struct file *newfile;
1588
	int err, len, newfd, fput_needed;
1589
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1590

1591
	if (flags & ~(SOCK_CLOEXEC | SOCK_NONBLOCK))
U
Ulrich Drepper 已提交
1592 1593 1594 1595 1596
		return -EINVAL;

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

1597
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1598 1599 1600 1601
	if (!sock)
		goto out;

	err = -ENFILE;
1602 1603
	newsock = sock_alloc();
	if (!newsock)
L
Linus Torvalds 已提交
1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614
		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);

1615
	newfd = get_unused_fd_flags(flags);
1616 1617
	if (unlikely(newfd < 0)) {
		err = newfd;
1618 1619
		sock_release(newsock);
		goto out_put;
1620
	}
1621
	newfile = sock_alloc_file(newsock, flags, sock->sk->sk_prot_creator->name);
1622 1623 1624 1625 1626 1627
	if (unlikely(IS_ERR(newfile))) {
		err = PTR_ERR(newfile);
		put_unused_fd(newfd);
		sock_release(newsock);
		goto out_put;
	}
1628

1629 1630
	err = security_socket_accept(sock, newsock);
	if (err)
1631
		goto out_fd;
1632

L
Linus Torvalds 已提交
1633 1634
	err = sock->ops->accept(sock, newsock, sock->file->f_flags);
	if (err < 0)
1635
		goto out_fd;
L
Linus Torvalds 已提交
1636 1637

	if (upeer_sockaddr) {
1638
		if (newsock->ops->getname(newsock, (struct sockaddr *)&address,
1639
					  &len, 2) < 0) {
L
Linus Torvalds 已提交
1640
			err = -ECONNABORTED;
1641
			goto out_fd;
L
Linus Torvalds 已提交
1642
		}
1643
		err = move_addr_to_user(&address,
1644
					len, upeer_sockaddr, upeer_addrlen);
L
Linus Torvalds 已提交
1645
		if (err < 0)
1646
			goto out_fd;
L
Linus Torvalds 已提交
1647 1648 1649 1650
	}

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

1651 1652
	fd_install(newfd, newfile);
	err = newfd;
L
Linus Torvalds 已提交
1653 1654

out_put:
1655
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1656 1657
out:
	return err;
1658
out_fd:
1659
	fput(newfile);
1660
	put_unused_fd(newfd);
L
Linus Torvalds 已提交
1661 1662 1663
	goto out_put;
}

1664 1665
SYSCALL_DEFINE3(accept, int, fd, struct sockaddr __user *, upeer_sockaddr,
		int __user *, upeer_addrlen)
U
Ulrich Drepper 已提交
1666
{
U
Ulrich Drepper 已提交
1667
	return sys_accept4(fd, upeer_sockaddr, upeer_addrlen, 0);
U
Ulrich Drepper 已提交
1668 1669
}

L
Linus Torvalds 已提交
1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681
/*
 *	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.
 */

1682 1683
SYSCALL_DEFINE3(connect, int, fd, struct sockaddr __user *, uservaddr,
		int, addrlen)
L
Linus Torvalds 已提交
1684 1685
{
	struct socket *sock;
1686
	struct sockaddr_storage address;
1687
	int err, fput_needed;
L
Linus Torvalds 已提交
1688

1689
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1690 1691
	if (!sock)
		goto out;
1692
	err = move_addr_to_kernel(uservaddr, addrlen, &address);
L
Linus Torvalds 已提交
1693 1694 1695
	if (err < 0)
		goto out_put;

1696
	err =
1697
	    security_socket_connect(sock, (struct sockaddr *)&address, addrlen);
L
Linus Torvalds 已提交
1698 1699 1700
	if (err)
		goto out_put;

1701
	err = sock->ops->connect(sock, (struct sockaddr *)&address, addrlen,
L
Linus Torvalds 已提交
1702 1703
				 sock->file->f_flags);
out_put:
1704
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1705 1706 1707 1708 1709 1710 1711 1712 1713
out:
	return err;
}

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

1714 1715
SYSCALL_DEFINE3(getsockname, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1716 1717
{
	struct socket *sock;
1718
	struct sockaddr_storage address;
1719
	int len, err, fput_needed;
1720

1721
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1722 1723 1724 1725 1726 1727 1728
	if (!sock)
		goto out;

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

1729
	err = sock->ops->getname(sock, (struct sockaddr *)&address, &len, 0);
L
Linus Torvalds 已提交
1730 1731
	if (err)
		goto out_put;
1732
	err = move_addr_to_user(&address, len, usockaddr, usockaddr_len);
L
Linus Torvalds 已提交
1733 1734

out_put:
1735
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1736 1737 1738 1739 1740 1741 1742 1743 1744
out:
	return err;
}

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

1745 1746
SYSCALL_DEFINE3(getpeername, int, fd, struct sockaddr __user *, usockaddr,
		int __user *, usockaddr_len)
L
Linus Torvalds 已提交
1747 1748
{
	struct socket *sock;
1749
	struct sockaddr_storage address;
1750
	int len, err, fput_needed;
L
Linus Torvalds 已提交
1751

1752 1753
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1754 1755
		err = security_socket_getpeername(sock);
		if (err) {
1756
			fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1757 1758 1759
			return err;
		}

1760
		err =
1761
		    sock->ops->getname(sock, (struct sockaddr *)&address, &len,
1762
				       1);
L
Linus Torvalds 已提交
1763
		if (!err)
1764
			err = move_addr_to_user(&address, len, usockaddr,
1765
						usockaddr_len);
1766
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1767 1768 1769 1770 1771 1772 1773 1774 1775 1776
	}
	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.
 */

1777
SYSCALL_DEFINE6(sendto, int, fd, void __user *, buff, size_t, len,
1778
		unsigned int, flags, struct sockaddr __user *, addr,
1779
		int, addr_len)
L
Linus Torvalds 已提交
1780 1781
{
	struct socket *sock;
1782
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1783 1784 1785
	int err;
	struct msghdr msg;
	struct iovec iov;
1786 1787
	int fput_needed;

1788 1789
	if (len > INT_MAX)
		len = INT_MAX;
1790 1791
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (!sock)
1792
		goto out;
1793

1794 1795 1796 1797 1798 1799 1800 1801
	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;
1802
	if (addr) {
1803
		err = move_addr_to_kernel(addr, addr_len, &address);
L
Linus Torvalds 已提交
1804 1805
		if (err < 0)
			goto out_put;
1806
		msg.msg_name = (struct sockaddr *)&address;
1807
		msg.msg_namelen = addr_len;
L
Linus Torvalds 已提交
1808 1809 1810 1811 1812 1813
	}
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
	msg.msg_flags = flags;
	err = sock_sendmsg(sock, &msg, len);

1814
out_put:
1815
	fput_light(sock->file, fput_needed);
1816
out:
L
Linus Torvalds 已提交
1817 1818 1819 1820
	return err;
}

/*
1821
 *	Send a datagram down a socket.
L
Linus Torvalds 已提交
1822 1823
 */

1824
SYSCALL_DEFINE4(send, int, fd, void __user *, buff, size_t, len,
1825
		unsigned int, flags)
L
Linus Torvalds 已提交
1826 1827 1828 1829 1830
{
	return sys_sendto(fd, buff, len, flags, NULL, 0);
}

/*
1831
 *	Receive a frame from the socket and optionally record the address of the
L
Linus Torvalds 已提交
1832 1833 1834 1835
 *	sender. We verify the buffers are writable and if needed move the
 *	sender address from kernel to user space.
 */

1836
SYSCALL_DEFINE6(recvfrom, int, fd, void __user *, ubuf, size_t, size,
1837
		unsigned int, flags, struct sockaddr __user *, addr,
1838
		int __user *, addr_len)
L
Linus Torvalds 已提交
1839 1840 1841 1842
{
	struct socket *sock;
	struct iovec iov;
	struct msghdr msg;
1843
	struct sockaddr_storage address;
1844
	int err, err2;
1845 1846
	int fput_needed;

1847 1848
	if (size > INT_MAX)
		size = INT_MAX;
1849
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
L
Linus Torvalds 已提交
1850
	if (!sock)
1851
		goto out;
L
Linus Torvalds 已提交
1852

1853 1854 1855 1856 1857 1858
	msg.msg_control = NULL;
	msg.msg_controllen = 0;
	msg.msg_iovlen = 1;
	msg.msg_iov = &iov;
	iov.iov_len = size;
	iov.iov_base = ubuf;
1859 1860
	msg.msg_name = (struct sockaddr *)&address;
	msg.msg_namelen = sizeof(address);
L
Linus Torvalds 已提交
1861 1862
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
1863
	err = sock_recvmsg(sock, &msg, size, flags);
L
Linus Torvalds 已提交
1864

1865
	if (err >= 0 && addr != NULL) {
1866
		err2 = move_addr_to_user(&address,
1867
					 msg.msg_namelen, addr, addr_len);
1868 1869
		if (err2 < 0)
			err = err2;
L
Linus Torvalds 已提交
1870
	}
1871 1872

	fput_light(sock->file, fput_needed);
1873
out:
L
Linus Torvalds 已提交
1874 1875 1876 1877
	return err;
}

/*
1878
 *	Receive a datagram from a socket.
L
Linus Torvalds 已提交
1879 1880
 */

1881
asmlinkage long sys_recv(int fd, void __user *ubuf, size_t size,
1882
			 unsigned int flags)
L
Linus Torvalds 已提交
1883 1884 1885 1886 1887 1888 1889 1890 1891
{
	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.
 */

1892 1893
SYSCALL_DEFINE5(setsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int, optlen)
L
Linus Torvalds 已提交
1894
{
1895
	int err, fput_needed;
L
Linus Torvalds 已提交
1896 1897 1898 1899
	struct socket *sock;

	if (optlen < 0)
		return -EINVAL;
1900 1901 1902 1903

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
		err = security_socket_setsockopt(sock, level, optname);
1904 1905
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1906 1907

		if (level == SOL_SOCKET)
1908 1909 1910
			err =
			    sock_setsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1911
		else
1912 1913 1914
			err =
			    sock->ops->setsockopt(sock, level, optname, optval,
						  optlen);
1915 1916
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1917 1918 1919 1920 1921 1922 1923 1924 1925
	}
	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.
 */

1926 1927
SYSCALL_DEFINE5(getsockopt, int, fd, int, level, int, optname,
		char __user *, optval, int __user *, optlen)
L
Linus Torvalds 已提交
1928
{
1929
	int err, fput_needed;
L
Linus Torvalds 已提交
1930 1931
	struct socket *sock;

1932 1933
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
1934 1935 1936
		err = security_socket_getsockopt(sock, level, optname);
		if (err)
			goto out_put;
L
Linus Torvalds 已提交
1937 1938

		if (level == SOL_SOCKET)
1939 1940 1941
			err =
			    sock_getsockopt(sock, level, optname, optval,
					    optlen);
L
Linus Torvalds 已提交
1942
		else
1943 1944 1945
			err =
			    sock->ops->getsockopt(sock, level, optname, optval,
						  optlen);
1946 1947
out_put:
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1948 1949 1950 1951 1952 1953 1954 1955
	}
	return err;
}

/*
 *	Shutdown a socket.
 */

1956
SYSCALL_DEFINE2(shutdown, int, fd, int, how)
L
Linus Torvalds 已提交
1957
{
1958
	int err, fput_needed;
L
Linus Torvalds 已提交
1959 1960
	struct socket *sock;

1961 1962
	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock != NULL) {
L
Linus Torvalds 已提交
1963
		err = security_socket_shutdown(sock, how);
1964 1965 1966
		if (!err)
			err = sock->ops->shutdown(sock, how);
		fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
1967 1968 1969 1970
	}
	return err;
}

1971
/* A couple of helpful macros for getting the address of the 32/64 bit
L
Linus Torvalds 已提交
1972 1973 1974 1975 1976 1977
 * 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)

1978 1979 1980 1981 1982
struct used_address {
	struct sockaddr_storage name;
	unsigned int name_len;
};

1983
static int __sys_sendmsg(struct socket *sock, struct msghdr __user *msg,
1984
			 struct msghdr *msg_sys, unsigned int flags,
1985
			 struct used_address *used_address)
L
Linus Torvalds 已提交
1986
{
1987 1988
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
1989
	struct sockaddr_storage address;
L
Linus Torvalds 已提交
1990
	struct iovec iovstack[UIO_FASTIOV], *iov = iovstack;
1991
	unsigned char ctl[sizeof(struct cmsghdr) + 20]
1992 1993
	    __attribute__ ((aligned(sizeof(__kernel_size_t))));
	/* 20 is size of ipv6_pktinfo */
L
Linus Torvalds 已提交
1994
	unsigned char *ctl_buf = ctl;
E
Eric Dumazet 已提交
1995
	int err, ctl_len, total_len;
1996

L
Linus Torvalds 已提交
1997 1998
	err = -EFAULT;
	if (MSG_CMSG_COMPAT & flags) {
1999
		if (get_compat_msghdr(msg_sys, msg_compat))
L
Linus Torvalds 已提交
2000
			return -EFAULT;
2001
	} else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr)))
L
Linus Torvalds 已提交
2002 2003
		return -EFAULT;

2004
	if (msg_sys->msg_iovlen > UIO_FASTIOV) {
E
Eric Dumazet 已提交
2005 2006 2007 2008 2009 2010
		err = -EMSGSIZE;
		if (msg_sys->msg_iovlen > UIO_MAXIOV)
			goto out;
		err = -ENOMEM;
		iov = kmalloc(msg_sys->msg_iovlen * sizeof(struct iovec),
			      GFP_KERNEL);
L
Linus Torvalds 已提交
2011
		if (!iov)
2012
			goto out;
L
Linus Torvalds 已提交
2013 2014 2015 2016
	}

	/* This will also move the address data into kernel space */
	if (MSG_CMSG_COMPAT & flags) {
2017
		err = verify_compat_iovec(msg_sys, iov, &address, VERIFY_READ);
L
Linus Torvalds 已提交
2018
	} else
2019
		err = verify_iovec(msg_sys, iov, &address, VERIFY_READ);
2020
	if (err < 0)
L
Linus Torvalds 已提交
2021 2022 2023 2024 2025
		goto out_freeiov;
	total_len = err;

	err = -ENOBUFS;

2026
	if (msg_sys->msg_controllen > INT_MAX)
L
Linus Torvalds 已提交
2027
		goto out_freeiov;
2028
	ctl_len = msg_sys->msg_controllen;
L
Linus Torvalds 已提交
2029
	if ((MSG_CMSG_COMPAT & flags) && ctl_len) {
2030
		err =
2031
		    cmsghdr_from_user_compat_to_kern(msg_sys, sock->sk, ctl,
2032
						     sizeof(ctl));
L
Linus Torvalds 已提交
2033 2034
		if (err)
			goto out_freeiov;
2035 2036
		ctl_buf = msg_sys->msg_control;
		ctl_len = msg_sys->msg_controllen;
L
Linus Torvalds 已提交
2037
	} else if (ctl_len) {
2038
		if (ctl_len > sizeof(ctl)) {
L
Linus Torvalds 已提交
2039
			ctl_buf = sock_kmalloc(sock->sk, ctl_len, GFP_KERNEL);
2040
			if (ctl_buf == NULL)
L
Linus Torvalds 已提交
2041 2042 2043 2044
				goto out_freeiov;
		}
		err = -EFAULT;
		/*
2045
		 * Careful! Before this, msg_sys->msg_control contains a user pointer.
L
Linus Torvalds 已提交
2046 2047 2048
		 * Afterwards, it will be a kernel pointer. Thus the compiler-assisted
		 * checking falls down on this.
		 */
2049
		if (copy_from_user(ctl_buf,
2050
				   (void __user __force *)msg_sys->msg_control,
2051
				   ctl_len))
L
Linus Torvalds 已提交
2052
			goto out_freectl;
2053
		msg_sys->msg_control = ctl_buf;
L
Linus Torvalds 已提交
2054
	}
2055
	msg_sys->msg_flags = flags;
L
Linus Torvalds 已提交
2056 2057

	if (sock->file->f_flags & O_NONBLOCK)
2058
		msg_sys->msg_flags |= MSG_DONTWAIT;
2059 2060 2061 2062 2063 2064
	/*
	 * 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.
	 */
2065 2066 2067
	if (used_address && msg_sys->msg_name &&
	    used_address->name_len == msg_sys->msg_namelen &&
	    !memcmp(&used_address->name, msg_sys->msg_name,
2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078
		    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;
2079 2080 2081
		if (msg_sys->msg_name)
			memcpy(&used_address->name, msg_sys->msg_name,
			       used_address->name_len);
2082
	}
L
Linus Torvalds 已提交
2083 2084

out_freectl:
2085
	if (ctl_buf != ctl)
L
Linus Torvalds 已提交
2086 2087 2088
		sock_kfree_s(sock->sk, ctl_buf, ctl_len);
out_freeiov:
	if (iov != iovstack)
E
Eric Dumazet 已提交
2089
		kfree(iov);
2090 2091 2092 2093 2094 2095 2096 2097
out:
	return err;
}

/*
 *	BSD sendmsg interface
 */

2098
SYSCALL_DEFINE3(sendmsg, int, fd, struct msghdr __user *, msg, unsigned int, flags)
2099 2100 2101 2102 2103 2104 2105 2106
{
	int fput_needed, err;
	struct msghdr msg_sys;
	struct socket *sock = sockfd_lookup_light(fd, &err, &fput_needed);

	if (!sock)
		goto out;

2107
	err = __sys_sendmsg(sock, msg, &msg_sys, flags, NULL);
2108

2109
	fput_light(sock->file, fput_needed);
2110
out:
L
Linus Torvalds 已提交
2111 2112 2113
	return err;
}

2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125
/*
 *	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;
2126
	struct used_address used_address;
2127

2128 2129
	if (vlen > UIO_MAXIOV)
		vlen = UIO_MAXIOV;
2130 2131 2132 2133 2134 2135 2136

	datagrams = 0;

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

2137
	used_address.name_len = UINT_MAX;
2138 2139
	entry = mmsg;
	compat_entry = (struct compat_mmsghdr __user *)mmsg;
2140
	err = 0;
2141 2142 2143 2144

	while (datagrams < vlen) {
		if (MSG_CMSG_COMPAT & flags) {
			err = __sys_sendmsg(sock, (struct msghdr __user *)compat_entry,
2145
					    &msg_sys, flags, &used_address);
2146 2147 2148 2149 2150 2151
			if (err < 0)
				break;
			err = __put_user(err, &compat_entry->msg_len);
			++compat_entry;
		} else {
			err = __sys_sendmsg(sock, (struct msghdr __user *)entry,
2152
					    &msg_sys, flags, &used_address);
2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165
			if (err < 0)
				break;
			err = put_user(err, &entry->msg_len);
			++entry;
		}

		if (err)
			break;
		++datagrams;
	}

	fput_light(sock->file, fput_needed);

2166 2167
	/* We only return an error if no datagrams were able to be sent */
	if (datagrams != 0)
2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178
		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);
}

2179
static int __sys_recvmsg(struct socket *sock, struct msghdr __user *msg,
2180
			 struct msghdr *msg_sys, unsigned int flags, int nosec)
L
Linus Torvalds 已提交
2181
{
2182 2183
	struct compat_msghdr __user *msg_compat =
	    (struct compat_msghdr __user *)msg;
L
Linus Torvalds 已提交
2184
	struct iovec iovstack[UIO_FASTIOV];
2185
	struct iovec *iov = iovstack;
L
Linus Torvalds 已提交
2186
	unsigned long cmsg_ptr;
E
Eric Dumazet 已提交
2187
	int err, total_len, len;
L
Linus Torvalds 已提交
2188 2189

	/* kernel mode address */
2190
	struct sockaddr_storage addr;
L
Linus Torvalds 已提交
2191 2192 2193 2194

	/* user mode address pointers */
	struct sockaddr __user *uaddr;
	int __user *uaddr_len;
2195

L
Linus Torvalds 已提交
2196
	if (MSG_CMSG_COMPAT & flags) {
2197
		if (get_compat_msghdr(msg_sys, msg_compat))
L
Linus Torvalds 已提交
2198
			return -EFAULT;
2199
	} else if (copy_from_user(msg_sys, msg, sizeof(struct msghdr)))
2200
		return -EFAULT;
L
Linus Torvalds 已提交
2201

2202
	if (msg_sys->msg_iovlen > UIO_FASTIOV) {
E
Eric Dumazet 已提交
2203 2204 2205 2206 2207 2208
		err = -EMSGSIZE;
		if (msg_sys->msg_iovlen > UIO_MAXIOV)
			goto out;
		err = -ENOMEM;
		iov = kmalloc(msg_sys->msg_iovlen * sizeof(struct iovec),
			      GFP_KERNEL);
L
Linus Torvalds 已提交
2209
		if (!iov)
2210
			goto out;
L
Linus Torvalds 已提交
2211 2212 2213
	}

	/*
2214 2215
	 *      Save the user-mode address (verify_iovec will change the
	 *      kernel msghdr to use the kernel address space)
L
Linus Torvalds 已提交
2216
	 */
2217

2218
	uaddr = (__force void __user *)msg_sys->msg_name;
L
Linus Torvalds 已提交
2219 2220
	uaddr_len = COMPAT_NAMELEN(msg);
	if (MSG_CMSG_COMPAT & flags) {
2221
		err = verify_compat_iovec(msg_sys, iov, &addr, VERIFY_WRITE);
L
Linus Torvalds 已提交
2222
	} else
2223
		err = verify_iovec(msg_sys, iov, &addr, VERIFY_WRITE);
L
Linus Torvalds 已提交
2224 2225
	if (err < 0)
		goto out_freeiov;
2226
	total_len = err;
L
Linus Torvalds 已提交
2227

2228 2229
	cmsg_ptr = (unsigned long)msg_sys->msg_control;
	msg_sys->msg_flags = flags & (MSG_CMSG_CLOEXEC|MSG_CMSG_COMPAT);
2230

L
Linus Torvalds 已提交
2231 2232
	if (sock->file->f_flags & O_NONBLOCK)
		flags |= MSG_DONTWAIT;
2233 2234
	err = (nosec ? sock_recvmsg_nosec : sock_recvmsg)(sock, msg_sys,
							  total_len, flags);
L
Linus Torvalds 已提交
2235 2236 2237 2238 2239
	if (err < 0)
		goto out_freeiov;
	len = err;

	if (uaddr != NULL) {
2240
		err = move_addr_to_user(&addr,
2241
					msg_sys->msg_namelen, uaddr,
2242
					uaddr_len);
L
Linus Torvalds 已提交
2243 2244 2245
		if (err < 0)
			goto out_freeiov;
	}
2246
	err = __put_user((msg_sys->msg_flags & ~MSG_CMSG_COMPAT),
2247
			 COMPAT_FLAGS(msg));
L
Linus Torvalds 已提交
2248 2249 2250
	if (err)
		goto out_freeiov;
	if (MSG_CMSG_COMPAT & flags)
2251
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2252 2253
				 &msg_compat->msg_controllen);
	else
2254
		err = __put_user((unsigned long)msg_sys->msg_control - cmsg_ptr,
L
Linus Torvalds 已提交
2255 2256 2257 2258 2259 2260 2261
				 &msg->msg_controllen);
	if (err)
		goto out_freeiov;
	err = len;

out_freeiov:
	if (iov != iovstack)
E
Eric Dumazet 已提交
2262
		kfree(iov);
2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282
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);

2283
	fput_light(sock->file, fput_needed);
L
Linus Torvalds 已提交
2284 2285 2286 2287
out:
	return err;
}

2288 2289 2290 2291 2292 2293 2294 2295 2296 2297
/*
 *     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;
2298
	struct compat_mmsghdr __user *compat_entry;
2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317
	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;
2318
	compat_entry = (struct compat_mmsghdr __user *)mmsg;
2319 2320 2321 2322 2323

	while (datagrams < vlen) {
		/*
		 * No need to ask LSM for more than the first datagram.
		 */
2324 2325
		if (MSG_CMSG_COMPAT & flags) {
			err = __sys_recvmsg(sock, (struct msghdr __user *)compat_entry,
2326 2327
					    &msg_sys, flags & ~MSG_WAITFORONE,
					    datagrams);
2328 2329 2330 2331 2332 2333
			if (err < 0)
				break;
			err = __put_user(err, &compat_entry->msg_len);
			++compat_entry;
		} else {
			err = __sys_recvmsg(sock, (struct msghdr __user *)entry,
2334 2335
					    &msg_sys, flags & ~MSG_WAITFORONE,
					    datagrams);
2336 2337 2338 2339 2340 2341
			if (err < 0)
				break;
			err = put_user(err, &entry->msg_len);
			++entry;
		}

2342 2343 2344 2345
		if (err)
			break;
		++datagrams;

2346 2347 2348 2349
		/* MSG_WAITFORONE turns on MSG_DONTWAIT after one packet */
		if (flags & MSG_WAITFORONE)
			flags |= MSG_DONTWAIT;

2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369
		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
2370

2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 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
	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
2418 2419
/* Argument list sizes for sys_socketcall */
#define AL(x) ((x) * sizeof(unsigned long))
2420
static const unsigned char nargs[21] = {
2421 2422 2423
	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),
2424
	AL(4), AL(5), AL(4)
2425 2426
};

L
Linus Torvalds 已提交
2427 2428 2429
#undef AL

/*
2430
 *	System call vectors.
L
Linus Torvalds 已提交
2431 2432 2433
 *
 *	Argument checking cleaned up. Saved 20% in size.
 *  This function doesn't need to set the kernel lock because
2434
 *  it is set by the callees.
L
Linus Torvalds 已提交
2435 2436
 */

2437
SYSCALL_DEFINE2(socketcall, int, call, unsigned long __user *, args)
L
Linus Torvalds 已提交
2438 2439
{
	unsigned long a[6];
2440
	unsigned long a0, a1;
L
Linus Torvalds 已提交
2441
	int err;
2442
	unsigned int len;
L
Linus Torvalds 已提交
2443

2444
	if (call < 1 || call > SYS_SENDMMSG)
L
Linus Torvalds 已提交
2445 2446
		return -EINVAL;

2447 2448 2449 2450
	len = nargs[call];
	if (len > sizeof(a))
		return -EINVAL;

L
Linus Torvalds 已提交
2451
	/* copy_from_user should be SMP safe. */
2452
	if (copy_from_user(a, args, len))
L
Linus Torvalds 已提交
2453
		return -EFAULT;
2454

A
Al Viro 已提交
2455
	audit_socketcall(nargs[call] / sizeof(unsigned long), a);
2456

2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473
	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
2474 2475
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], 0);
2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518
		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;
2519 2520 2521
	case SYS_SENDMMSG:
		err = sys_sendmmsg(a0, (struct mmsghdr __user *)a1, a[2], a[3]);
		break;
2522 2523 2524
	case SYS_RECVMSG:
		err = sys_recvmsg(a0, (struct msghdr __user *)a1, a[2]);
		break;
2525 2526 2527 2528
	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
2529 2530 2531
	case SYS_ACCEPT4:
		err = sys_accept4(a0, (struct sockaddr __user *)a1,
				  (int __user *)a[2], a[3]);
U
Ulrich Drepper 已提交
2532
		break;
2533 2534 2535
	default:
		err = -EINVAL;
		break;
L
Linus Torvalds 已提交
2536 2537 2538 2539
	}
	return err;
}

2540
#endif				/* __ARCH_WANT_SYS_SOCKETCALL */
L
Linus Torvalds 已提交
2541

2542 2543 2544 2545
/**
 *	sock_register - add a socket protocol handler
 *	@ops: description of protocol
 *
L
Linus Torvalds 已提交
2546 2547
 *	This function is called by a protocol handler that wants to
 *	advertise its address family, and have it linked into the
2548 2549
 *	socket interface. The value ops->family coresponds to the
 *	socket system call protocol family.
L
Linus Torvalds 已提交
2550
 */
2551
int sock_register(const struct net_proto_family *ops)
L
Linus Torvalds 已提交
2552 2553 2554 2555
{
	int err;

	if (ops->family >= NPROTO) {
2556 2557
		printk(KERN_CRIT "protocol %d >= NPROTO(%d)\n", ops->family,
		       NPROTO);
L
Linus Torvalds 已提交
2558 2559
		return -ENOBUFS;
	}
2560 2561

	spin_lock(&net_family_lock);
E
Eric Dumazet 已提交
2562 2563
	if (rcu_dereference_protected(net_families[ops->family],
				      lockdep_is_held(&net_family_lock)))
2564 2565
		err = -EEXIST;
	else {
2566
		rcu_assign_pointer(net_families[ops->family], ops);
L
Linus Torvalds 已提交
2567 2568
		err = 0;
	}
2569 2570
	spin_unlock(&net_family_lock);

2571
	printk(KERN_INFO "NET: Registered protocol family %d\n", ops->family);
L
Linus Torvalds 已提交
2572 2573
	return err;
}
2574
EXPORT_SYMBOL(sock_register);
L
Linus Torvalds 已提交
2575

2576 2577 2578 2579
/**
 *	sock_unregister - remove a protocol handler
 *	@family: protocol family to remove
 *
L
Linus Torvalds 已提交
2580 2581
 *	This function is called by a protocol handler that wants to
 *	remove its address family, and have it unlinked from the
2582 2583 2584 2585 2586 2587
 *	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
2588
 */
2589
void sock_unregister(int family)
L
Linus Torvalds 已提交
2590
{
2591
	BUG_ON(family < 0 || family >= NPROTO);
L
Linus Torvalds 已提交
2592

2593
	spin_lock(&net_family_lock);
2594
	RCU_INIT_POINTER(net_families[family], NULL);
2595 2596 2597 2598
	spin_unlock(&net_family_lock);

	synchronize_rcu();

2599
	printk(KERN_INFO "NET: Unregistered protocol family %d\n", family);
L
Linus Torvalds 已提交
2600
}
2601
EXPORT_SYMBOL(sock_unregister);
L
Linus Torvalds 已提交
2602

2603
static int __init sock_init(void)
L
Linus Torvalds 已提交
2604
{
N
Nick Piggin 已提交
2605
	int err;
2606 2607 2608 2609 2610 2611
	/*
	 *      Initialize the network sysctl infrastructure.
	 */
	err = net_sysctl_init();
	if (err)
		goto out;
N
Nick Piggin 已提交
2612

L
Linus Torvalds 已提交
2613
	/*
2614
	 *      Initialize skbuff SLAB cache
L
Linus Torvalds 已提交
2615 2616 2617 2618
	 */
	skb_init();

	/*
2619
	 *      Initialize the protocols module.
L
Linus Torvalds 已提交
2620 2621 2622
	 */

	init_inodecache();
N
Nick Piggin 已提交
2623 2624 2625 2626

	err = register_filesystem(&sock_fs_type);
	if (err)
		goto out_fs;
L
Linus Torvalds 已提交
2627
	sock_mnt = kern_mount(&sock_fs_type);
N
Nick Piggin 已提交
2628 2629 2630 2631
	if (IS_ERR(sock_mnt)) {
		err = PTR_ERR(sock_mnt);
		goto out_mount;
	}
2632 2633

	/* The real protocol initialization is performed in later initcalls.
L
Linus Torvalds 已提交
2634 2635 2636 2637 2638
	 */

#ifdef CONFIG_NETFILTER
	netfilter_init();
#endif
2639

2640 2641 2642 2643
#ifdef CONFIG_NETWORK_PHY_TIMESTAMPING
	skb_timestamping_init();
#endif

N
Nick Piggin 已提交
2644 2645 2646 2647 2648 2649 2650
out:
	return err;

out_mount:
	unregister_filesystem(&sock_fs_type);
out_fs:
	goto out;
L
Linus Torvalds 已提交
2651 2652
}

2653 2654
core_initcall(sock_init);	/* early initcall */

L
Linus Torvalds 已提交
2655 2656 2657 2658 2659 2660
#ifdef CONFIG_PROC_FS
void socket_seq_show(struct seq_file *seq)
{
	int cpu;
	int counter = 0;

2661
	for_each_possible_cpu(cpu)
2662
	    counter += per_cpu(sockets_in_use, cpu);
L
Linus Torvalds 已提交
2663 2664 2665 2666 2667 2668 2669

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

	seq_printf(seq, "sockets: used %d\n", counter);
}
2670
#endif				/* CONFIG_PROC_FS */
L
Linus Torvalds 已提交
2671

2672
#ifdef CONFIG_COMPAT
2673
static int do_siocgstamp(struct net *net, struct socket *sock,
2674
			 unsigned int cmd, void __user *up)
2675 2676 2677 2678 2679 2680
{
	mm_segment_t old_fs = get_fs();
	struct timeval ktv;
	int err;

	set_fs(KERNEL_DS);
2681
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&ktv);
2682
	set_fs(old_fs);
2683
	if (!err)
2684
		err = compat_put_timeval(&ktv, up);
2685

2686 2687 2688
	return err;
}

2689
static int do_siocgstampns(struct net *net, struct socket *sock,
2690
			   unsigned int cmd, void __user *up)
2691 2692 2693 2694 2695 2696
{
	mm_segment_t old_fs = get_fs();
	struct timespec kts;
	int err;

	set_fs(KERNEL_DS);
2697
	err = sock_do_ioctl(net, sock, cmd, (unsigned long)&kts);
2698
	set_fs(old_fs);
2699
	if (!err)
2700
		err = compat_put_timespec(&kts, up);
2701

2702 2703 2704
	return err;
}

2705
static int dev_ifname32(struct net *net, struct compat_ifreq __user *uifr32)
2706 2707 2708 2709 2710
{
	struct ifreq __user *uifr;
	int err;

	uifr = compat_alloc_user_space(sizeof(struct ifreq));
2711
	if (copy_in_user(uifr, uifr32, sizeof(struct compat_ifreq)))
2712 2713
		return -EFAULT;

2714
	err = dev_ioctl(net, SIOCGIFNAME, uifr);
2715 2716 2717
	if (err)
		return err;

2718
	if (copy_in_user(uifr32, uifr, sizeof(struct compat_ifreq)))
2719 2720 2721 2722 2723
		return -EFAULT;

	return 0;
}

2724
static int dev_ifconf(struct net *net, struct compat_ifconf __user *uifc32)
2725
{
2726
	struct compat_ifconf ifc32;
2727 2728
	struct ifconf ifc;
	struct ifconf __user *uifc;
2729
	struct compat_ifreq __user *ifr32;
2730 2731 2732 2733
	struct ifreq __user *ifr;
	unsigned int i, j;
	int err;

2734
	if (copy_from_user(&ifc32, uifc32, sizeof(struct compat_ifconf)))
2735 2736
		return -EFAULT;

2737
	memset(&ifc, 0, sizeof(ifc));
2738 2739 2740 2741 2742 2743
	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 {
2744 2745
		size_t len = ((ifc32.ifc_len / sizeof(struct compat_ifreq)) + 1) *
			sizeof(struct ifreq);
2746 2747 2748 2749
		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);
2750
		for (i = 0; i < ifc32.ifc_len; i += sizeof(struct compat_ifreq)) {
2751
			if (copy_in_user(ifr, ifr32, sizeof(struct compat_ifreq)))
2752 2753 2754 2755 2756 2757 2758 2759
				return -EFAULT;
			ifr++;
			ifr32++;
		}
	}
	if (copy_to_user(uifc, &ifc, sizeof(struct ifconf)))
		return -EFAULT;

2760
	err = dev_ioctl(net, SIOCGIFCONF, uifc);
2761 2762 2763 2764 2765 2766 2767 2768 2769
	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;
2770 2771 2772
	     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)))
2773 2774 2775 2776 2777 2778 2779 2780 2781 2782
			return -EFAULT;
		ifr32++;
		ifr++;
	}

	if (ifc32.ifcbuf == 0) {
		/* Translate from 64-bit structure multiple to
		 * a 32-bit one.
		 */
		i = ifc.ifc_len;
2783
		i = ((i / sizeof(struct ifreq)) * sizeof(struct compat_ifreq));
2784 2785 2786 2787
		ifc32.ifc_len = i;
	} else {
		ifc32.ifc_len = i;
	}
2788
	if (copy_to_user(uifc32, &ifc32, sizeof(struct compat_ifconf)))
2789 2790 2791 2792 2793
		return -EFAULT;

	return 0;
}

2794
static int ethtool_ioctl(struct net *net, struct compat_ifreq __user *ifr32)
2795
{
2796 2797 2798 2799
	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;
2800
	struct ifreq __user *ifr;
2801 2802
	u32 rule_cnt = 0, actual_rule_cnt;
	u32 ethcmd;
2803
	u32 data;
2804
	int ret;
2805

2806 2807
	if (get_user(data, &ifr32->ifr_ifru.ifru_data))
		return -EFAULT;
2808

2809 2810 2811
	compat_rxnfc = compat_ptr(data);

	if (get_user(ethcmd, &compat_rxnfc->cmd))
2812 2813
		return -EFAULT;

2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830
	/* 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:
2831
	case ETHTOOL_SRXCLSRLINS:
2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843
		convert_out = true;
		/* fall through */
	case ETHTOOL_SRXCLSRLDEL:
		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))
2844 2845
		return -EFAULT;

2846 2847
	if (put_user(convert_in ? rxnfc : compat_ptr(data),
		     &ifr->ifr_ifru.ifru_data))
2848 2849
		return -EFAULT;

2850
	if (convert_in) {
2851
		/* We expect there to be holes between fs.m_ext and
2852 2853
		 * fs.ring_cookie and at the end of fs, but nowhere else.
		 */
2854 2855 2856 2857
		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));
2858 2859 2860 2861 2862 2863 2864
		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,
2865
				 (void *)(&rxnfc->fs.m_ext + 1) -
2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881
				 (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,
2882
				 (const void *)(&rxnfc->fs.m_ext + 1) -
2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912
				 (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;
2913 2914
}

2915 2916 2917 2918 2919 2920
static int compat_siocwandev(struct net *net, struct compat_ifreq __user *uifr32)
{
	void __user *uptr;
	compat_uptr_t uptr32;
	struct ifreq __user *uifr;

2921
	uifr = compat_alloc_user_space(sizeof(*uifr));
2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935
	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);
}

2936 2937
static int bond_ioctl(struct net *net, unsigned int cmd,
			 struct compat_ifreq __user *ifr32)
2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950
{
	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:
2951
		if (copy_from_user(&kifr, ifr32, sizeof(struct compat_ifreq)))
2952 2953 2954
			return -EFAULT;

		old_fs = get_fs();
2955
		set_fs(KERNEL_DS);
2956 2957
		err = dev_ioctl(net, cmd,
				(struct ifreq __user __force *) &kifr);
2958
		set_fs(old_fs);
2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973

		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;

2974
		return dev_ioctl(net, cmd, uifr);
2975
	default:
2976
		return -ENOIOCTLCMD;
2977
	}
2978 2979
}

2980 2981
static int siocdevprivate_ioctl(struct net *net, unsigned int cmd,
				 struct compat_ifreq __user *u_ifreq32)
2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005
{
	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;

3006
	return dev_ioctl(net, cmd, u_ifreq64);
3007 3008
}

3009 3010
static int dev_ifsioc(struct net *net, struct socket *sock,
			 unsigned int cmd, struct compat_ifreq __user *uifr32)
3011
{
3012
	struct ifreq __user *uifr;
3013 3014
	int err;

3015 3016 3017 3018 3019 3020
	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);

3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032
	if (!err) {
		switch (cmd) {
		case SIOCGIFFLAGS:
		case SIOCGIFMETRIC:
		case SIOCGIFMTU:
		case SIOCGIFMEM:
		case SIOCGIFHWADDR:
		case SIOCGIFINDEX:
		case SIOCGIFADDR:
		case SIOCGIFBRDADDR:
		case SIOCGIFDSTADDR:
		case SIOCGIFNETMASK:
3033
		case SIOCGIFPFLAGS:
3034
		case SIOCGIFTXQLEN:
3035 3036
		case SIOCGMIIPHY:
		case SIOCGMIIREG:
3037
			if (copy_in_user(uifr32, uifr, sizeof(*uifr32)))
3038 3039 3040 3041 3042 3043 3044
				err = -EFAULT;
			break;
		}
	}
	return err;
}

3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064
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();
3065
	set_fs(KERNEL_DS);
3066
	err = dev_ioctl(net, cmd, (void  __user __force *)&ifr);
3067
	set_fs(old_fs);
3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088

	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;

3089
	uifr = compat_alloc_user_space(sizeof(*uifr));
3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103
	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);
}

3104
struct rtentry32 {
3105
	u32		rt_pad1;
3106 3107 3108
	struct sockaddr rt_dst;         /* target address               */
	struct sockaddr rt_gateway;     /* gateway addr (RTF_GATEWAY)   */
	struct sockaddr rt_genmask;     /* target network mask (IP)     */
3109 3110 3111 3112 3113 3114 3115
	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! */
3116
	/* char * */ u32 rt_dev;        /* forcing the device at add    */
3117 3118
	u32		rt_mtu;         /* per route MTU/Window         */
	u32		rt_window;      /* Window clamping              */
3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134
	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;
};

3135 3136
static int routing_ioctl(struct net *net, struct socket *sock,
			 unsigned int cmd, void __user *argp)
3137 3138 3139 3140 3141 3142 3143 3144 3145
{
	int ret;
	void *r = NULL;
	struct in6_rtmsg r6;
	struct rtentry r4;
	char devname[16];
	u32 rtdev;
	mm_segment_t old_fs = get_fs();

3146 3147
	if (sock && sock->sk && sock->sk->sk_family == AF_INET6) { /* ipv6 */
		struct in6_rtmsg32 __user *ur6 = argp;
3148
		ret = copy_from_user(&r6.rtmsg_dst, &(ur6->rtmsg_dst),
3149
			3 * sizeof(struct in6_addr));
3150 3151 3152 3153 3154 3155 3156
		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));
3157 3158 3159

		r = (void *) &r6;
	} else { /* ipv4 */
3160
		struct rtentry32 __user *ur4 = argp;
3161
		ret = copy_from_user(&r4.rt_dst, &(ur4->rt_dst),
3162
					3 * sizeof(struct sockaddr));
3163 3164 3165 3166 3167 3168
		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));
3169
		if (rtdev) {
3170
			ret |= copy_from_user(devname, compat_ptr(rtdev), 15);
3171 3172
			r4.rt_dev = (char __user __force *)devname;
			devname[15] = 0;
3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183
		} else
			r4.rt_dev = NULL;

		r = (void *) &r4;
	}

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

3184
	set_fs(KERNEL_DS);
3185
	ret = sock_do_ioctl(net, sock, cmd, (unsigned long) r);
3186
	set_fs(old_fs);
3187 3188 3189 3190 3191 3192 3193

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
3194
 * use compatible ioctls
3195
 */
3196
static int old_bridge_ioctl(compat_ulong_t __user *argp)
3197
{
3198
	compat_ulong_t tmp;
3199

3200
	if (get_user(tmp, argp))
3201 3202 3203 3204 3205 3206
		return -EFAULT;
	if (tmp == BRCTL_GET_VERSION)
		return BRCTL_VERSION + 1;
	return -EINVAL;
}

3207 3208 3209 3210 3211 3212
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);
3213

3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226
	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);
3227 3228
	case SIOCWANDEV:
		return compat_siocwandev(net, argp);
3229 3230 3231
	case SIOCGIFMAP:
	case SIOCSIFMAP:
		return compat_sioc_ifmap(net, cmd, argp);
3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245
	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);
3246 3247
	case SIOCSHWTSTAMP:
		return compat_siocshwtstamp(net, argp);
3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289

	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:
3290 3291 3292 3293
	case SIOCSIFNAME:
	case SIOCGMIIPHY:
	case SIOCGMIIREG:
	case SIOCSMIIREG:
3294
		return dev_ifsioc(net, sock, cmd, argp);
3295

3296 3297 3298 3299
	case SIOCSARP:
	case SIOCGARP:
	case SIOCDARP:
	case SIOCATMARK:
3300 3301 3302
		return sock_do_ioctl(net, sock, cmd, arg);
	}

3303 3304
	return -ENOIOCTLCMD;
}
3305

3306
static long compat_sock_ioctl(struct file *file, unsigned int cmd,
3307
			      unsigned long arg)
3308 3309 3310
{
	struct socket *sock = file->private_data;
	int ret = -ENOIOCTLCMD;
3311 3312 3313 3314 3315
	struct sock *sk;
	struct net *net;

	sk = sock->sk;
	net = sock_net(sk);
3316 3317 3318 3319

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

3320 3321 3322 3323
	if (ret == -ENOIOCTLCMD &&
	    (cmd >= SIOCIWFIRST && cmd <= SIOCIWLAST))
		ret = compat_wext_handle_ioctl(net, cmd, arg);

3324 3325 3326
	if (ret == -ENOIOCTLCMD)
		ret = compat_sock_ioctl_trans(file, sock, cmd, arg);

3327 3328 3329 3330
	return ret;
}
#endif

3331 3332 3333 3334
int kernel_bind(struct socket *sock, struct sockaddr *addr, int addrlen)
{
	return sock->ops->bind(sock, addr, addrlen);
}
3335
EXPORT_SYMBOL(kernel_bind);
3336 3337 3338 3339 3340

int kernel_listen(struct socket *sock, int backlog)
{
	return sock->ops->listen(sock, backlog);
}
3341
EXPORT_SYMBOL(kernel_listen);
3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355

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);
3356
		*newsock = NULL;
3357 3358 3359 3360
		goto done;
	}

	(*newsock)->ops = sock->ops;
3361
	__module_get((*newsock)->ops->owner);
3362 3363 3364 3365

done:
	return err;
}
3366
EXPORT_SYMBOL(kernel_accept);
3367 3368

int kernel_connect(struct socket *sock, struct sockaddr *addr, int addrlen,
3369
		   int flags)
3370 3371 3372
{
	return sock->ops->connect(sock, addr, addrlen, flags);
}
3373
EXPORT_SYMBOL(kernel_connect);
3374 3375 3376 3377 3378 3379

int kernel_getsockname(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 0);
}
3380
EXPORT_SYMBOL(kernel_getsockname);
3381 3382 3383 3384 3385 3386

int kernel_getpeername(struct socket *sock, struct sockaddr *addr,
			 int *addrlen)
{
	return sock->ops->getname(sock, addr, addrlen, 1);
}
3387
EXPORT_SYMBOL(kernel_getpeername);
3388 3389 3390 3391 3392

int kernel_getsockopt(struct socket *sock, int level, int optname,
			char *optval, int *optlen)
{
	mm_segment_t oldfs = get_fs();
3393 3394
	char __user *uoptval;
	int __user *uoptlen;
3395 3396
	int err;

3397 3398 3399
	uoptval = (char __user __force *) optval;
	uoptlen = (int __user __force *) optlen;

3400 3401
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3402
		err = sock_getsockopt(sock, level, optname, uoptval, uoptlen);
3403
	else
3404 3405
		err = sock->ops->getsockopt(sock, level, optname, uoptval,
					    uoptlen);
3406 3407 3408
	set_fs(oldfs);
	return err;
}
3409
EXPORT_SYMBOL(kernel_getsockopt);
3410 3411

int kernel_setsockopt(struct socket *sock, int level, int optname,
3412
			char *optval, unsigned int optlen)
3413 3414
{
	mm_segment_t oldfs = get_fs();
3415
	char __user *uoptval;
3416 3417
	int err;

3418 3419
	uoptval = (char __user __force *) optval;

3420 3421
	set_fs(KERNEL_DS);
	if (level == SOL_SOCKET)
3422
		err = sock_setsockopt(sock, level, optname, uoptval, optlen);
3423
	else
3424
		err = sock->ops->setsockopt(sock, level, optname, uoptval,
3425 3426 3427 3428
					    optlen);
	set_fs(oldfs);
	return err;
}
3429
EXPORT_SYMBOL(kernel_setsockopt);
3430 3431 3432 3433 3434 3435 3436 3437 3438

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);
}
3439
EXPORT_SYMBOL(kernel_sendpage);
3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451

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;
}
3452
EXPORT_SYMBOL(kernel_sock_ioctl);
3453

3454 3455 3456 3457 3458
int kernel_sock_shutdown(struct socket *sock, enum sock_shutdown_cmd how)
{
	return sock->ops->shutdown(sock, how);
}
EXPORT_SYMBOL(kernel_sock_shutdown);