Blame view
fs/fcntl.c
19.4 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 |
/* * linux/fs/fcntl.c * * Copyright (C) 1991, 1992 Linus Torvalds */ #include <linux/syscalls.h> #include <linux/init.h> #include <linux/mm.h> #include <linux/fs.h> #include <linux/file.h> |
9f3acc314
|
12 |
#include <linux/fdtable.h> |
16f7e0fe2
|
13 |
#include <linux/capability.h> |
1da177e4c
|
14 |
#include <linux/dnotify.h> |
1da177e4c
|
15 16 |
#include <linux/slab.h> #include <linux/module.h> |
35f3d14db
|
17 |
#include <linux/pipe_fs_i.h> |
1da177e4c
|
18 19 |
#include <linux/security.h> #include <linux/ptrace.h> |
7ed20e1ad
|
20 |
#include <linux/signal.h> |
ab2af1f50
|
21 |
#include <linux/rcupdate.h> |
b488893a3
|
22 |
#include <linux/pid_namespace.h> |
1d151c337
|
23 |
#include <linux/user_namespace.h> |
1da177e4c
|
24 25 26 27 |
#include <asm/poll.h> #include <asm/siginfo.h> #include <asm/uaccess.h> |
fc9b52cd8
|
28 |
void set_close_on_exec(unsigned int fd, int flag) |
1da177e4c
|
29 30 |
{ struct files_struct *files = current->files; |
badf16621
|
31 |
struct fdtable *fdt; |
1da177e4c
|
32 |
spin_lock(&files->file_lock); |
badf16621
|
33 |
fdt = files_fdtable(files); |
1da177e4c
|
34 |
if (flag) |
1dce27c5a
|
35 |
__set_close_on_exec(fd, fdt); |
1da177e4c
|
36 |
else |
1dce27c5a
|
37 |
__clear_close_on_exec(fd, fdt); |
1da177e4c
|
38 39 |
spin_unlock(&files->file_lock); } |
1dce27c5a
|
40 |
static bool get_close_on_exec(unsigned int fd) |
1da177e4c
|
41 42 |
{ struct files_struct *files = current->files; |
badf16621
|
43 |
struct fdtable *fdt; |
1dce27c5a
|
44 |
bool res; |
b835996f6
|
45 |
rcu_read_lock(); |
badf16621
|
46 |
fdt = files_fdtable(files); |
1dce27c5a
|
47 |
res = close_on_exec(fd, fdt); |
b835996f6
|
48 |
rcu_read_unlock(); |
1da177e4c
|
49 50 |
return res; } |
a26eab240
|
51 |
SYSCALL_DEFINE3(dup3, unsigned int, oldfd, unsigned int, newfd, int, flags) |
1da177e4c
|
52 53 54 55 |
{ int err = -EBADF; struct file * file, *tofree; struct files_struct * files = current->files; |
badf16621
|
56 |
struct fdtable *fdt; |
1da177e4c
|
57 |
|
336dd1f70
|
58 59 |
if ((flags & ~O_CLOEXEC) != 0) return -EINVAL; |
6c5d0512a
|
60 61 |
if (unlikely(oldfd == newfd)) return -EINVAL; |
1da177e4c
|
62 |
spin_lock(&files->file_lock); |
1da177e4c
|
63 |
err = expand_files(files, newfd); |
1b7e190b4
|
64 65 66 |
file = fcheck(oldfd); if (unlikely(!file)) goto Ebadf; |
4e1e018ec
|
67 68 |
if (unlikely(err < 0)) { if (err == -EMFILE) |
1b7e190b4
|
69 70 |
goto Ebadf; goto out_unlock; |
4e1e018ec
|
71 |
} |
1b7e190b4
|
72 73 74 75 76 77 78 79 80 81 82 83 84 85 |
/* * We need to detect attempts to do dup2() over allocated but still * not finished descriptor. NB: OpenBSD avoids that at the price of * extra work in their equivalent of fget() - they insert struct * file immediately after grabbing descriptor, mark it larval if * more work (e.g. actual opening) is needed and make sure that * fget() treats larval files as absent. Potentially interesting, * but while extra work in fget() is trivial, locking implications * and amount of surgery on open()-related paths in VFS are not. * FreeBSD fails with -EBADF in the same situation, NetBSD "solution" * deadlocks in rather amusing ways, AFAICS. All of that is out of * scope of POSIX or SUS, since neither considers shared descriptor * tables and this condition does not arise without those. */ |
1da177e4c
|
86 |
err = -EBUSY; |
badf16621
|
87 88 |
fdt = files_fdtable(files); tofree = fdt->fd[newfd]; |
1dce27c5a
|
89 |
if (!tofree && fd_is_open(newfd, fdt)) |
1b7e190b4
|
90 91 |
goto out_unlock; get_file(file); |
ab2af1f50
|
92 |
rcu_assign_pointer(fdt->fd[newfd], file); |
1dce27c5a
|
93 |
__set_open_fd(newfd, fdt); |
336dd1f70
|
94 |
if (flags & O_CLOEXEC) |
1dce27c5a
|
95 |
__set_close_on_exec(newfd, fdt); |
336dd1f70
|
96 |
else |
1dce27c5a
|
97 |
__clear_close_on_exec(newfd, fdt); |
1da177e4c
|
98 99 100 101 |
spin_unlock(&files->file_lock); if (tofree) filp_close(tofree, files); |
1da177e4c
|
102 |
|
1b7e190b4
|
103 104 105 106 107 |
return newfd; Ebadf: err = -EBADF; out_unlock: |
1da177e4c
|
108 |
spin_unlock(&files->file_lock); |
1b7e190b4
|
109 |
return err; |
1da177e4c
|
110 |
} |
336dd1f70
|
111 |
|
a26eab240
|
112 |
SYSCALL_DEFINE2(dup2, unsigned int, oldfd, unsigned int, newfd) |
336dd1f70
|
113 |
{ |
6c5d0512a
|
114 115 |
if (unlikely(newfd == oldfd)) { /* corner case */ struct files_struct *files = current->files; |
2b79bc4f7
|
116 |
int retval = oldfd; |
6c5d0512a
|
117 118 |
rcu_read_lock(); if (!fcheck_files(files, oldfd)) |
2b79bc4f7
|
119 |
retval = -EBADF; |
6c5d0512a
|
120 |
rcu_read_unlock(); |
2b79bc4f7
|
121 |
return retval; |
6c5d0512a
|
122 |
} |
336dd1f70
|
123 124 |
return sys_dup3(oldfd, newfd, 0); } |
1da177e4c
|
125 |
|
a26eab240
|
126 |
SYSCALL_DEFINE1(dup, unsigned int, fildes) |
1da177e4c
|
127 128 |
{ int ret = -EBADF; |
1abf0c718
|
129 |
struct file *file = fget_raw(fildes); |
1027abe88
|
130 131 132 133 134 135 136 137 |
if (file) { ret = get_unused_fd(); if (ret >= 0) fd_install(ret, file); else fput(file); } |
1da177e4c
|
138 139 |
return ret; } |
76398425b
|
140 |
#define SETFL_MASK (O_APPEND | O_NONBLOCK | O_NDELAY | O_DIRECT | O_NOATIME) |
1da177e4c
|
141 142 143 |
static int setfl(int fd, struct file * filp, unsigned long arg) { |
0f7fc9e4d
|
144 |
struct inode * inode = filp->f_path.dentry->d_inode; |
1da177e4c
|
145 |
int error = 0; |
7d95c8f27
|
146 147 148 149 150 |
/* * O_APPEND cannot be cleared if the file is marked as append-only * and the file is open for write. */ if (((arg ^ filp->f_flags) & O_APPEND) && IS_APPEND(inode)) |
1da177e4c
|
151 152 153 154 |
return -EPERM; /* O_NOATIME can only be set by the owner or superuser */ if ((arg & O_NOATIME) && !(filp->f_flags & O_NOATIME)) |
2e1496707
|
155 |
if (!inode_owner_or_capable(inode)) |
1da177e4c
|
156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 |
return -EPERM; /* required for strict SunOS emulation */ if (O_NONBLOCK != O_NDELAY) if (arg & O_NDELAY) arg |= O_NONBLOCK; if (arg & O_DIRECT) { if (!filp->f_mapping || !filp->f_mapping->a_ops || !filp->f_mapping->a_ops->direct_IO) return -EINVAL; } if (filp->f_op && filp->f_op->check_flags) error = filp->f_op->check_flags(arg); if (error) return error; |
218d11a8b
|
173 |
/* |
76398425b
|
174 |
* ->fasync() is responsible for setting the FASYNC bit. |
218d11a8b
|
175 |
*/ |
76398425b
|
176 177 178 179 180 |
if (((arg ^ filp->f_flags) & FASYNC) && filp->f_op && filp->f_op->fasync) { error = filp->f_op->fasync(fd, filp, (arg & FASYNC) != 0); if (error < 0) goto out; |
60aa49243
|
181 182 |
if (error > 0) error = 0; |
1da177e4c
|
183 |
} |
db1dd4d37
|
184 |
spin_lock(&filp->f_lock); |
1da177e4c
|
185 |
filp->f_flags = (arg & SETFL_MASK) | (filp->f_flags & ~SETFL_MASK); |
db1dd4d37
|
186 |
spin_unlock(&filp->f_lock); |
76398425b
|
187 |
|
1da177e4c
|
188 |
out: |
1da177e4c
|
189 190 |
return error; } |
609d7fa95
|
191 |
static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, |
2f38d70fb
|
192 |
int force) |
1da177e4c
|
193 |
{ |
80e1e8239
|
194 |
write_lock_irq(&filp->f_owner.lock); |
1da177e4c
|
195 |
if (force || !filp->f_owner.pid) { |
609d7fa95
|
196 197 198 |
put_pid(filp->f_owner.pid); filp->f_owner.pid = get_pid(pid); filp->f_owner.pid_type = type; |
2f38d70fb
|
199 200 201 202 203 204 |
if (pid) { const struct cred *cred = current_cred(); filp->f_owner.uid = cred->uid; filp->f_owner.euid = cred->euid; } |
1da177e4c
|
205 |
} |
80e1e8239
|
206 |
write_unlock_irq(&filp->f_owner.lock); |
1da177e4c
|
207 |
} |
609d7fa95
|
208 209 |
int __f_setown(struct file *filp, struct pid *pid, enum pid_type type, int force) |
1da177e4c
|
210 211 |
{ int err; |
2f38d70fb
|
212 |
|
1da177e4c
|
213 214 215 |
err = security_file_set_fowner(filp); if (err) return err; |
2f38d70fb
|
216 |
f_modown(filp, pid, type, force); |
1da177e4c
|
217 218 |
return 0; } |
609d7fa95
|
219 |
EXPORT_SYMBOL(__f_setown); |
1da177e4c
|
220 |
|
609d7fa95
|
221 222 223 224 225 226 227 228 229 230 231 232 |
int f_setown(struct file *filp, unsigned long arg, int force) { enum pid_type type; struct pid *pid; int who = arg; int result; type = PIDTYPE_PID; if (who < 0) { type = PIDTYPE_PGID; who = -who; } rcu_read_lock(); |
b488893a3
|
233 |
pid = find_vpid(who); |
609d7fa95
|
234 235 236 237 |
result = __f_setown(filp, pid, type, force); rcu_read_unlock(); return result; } |
1da177e4c
|
238 239 240 241 |
EXPORT_SYMBOL(f_setown); void f_delown(struct file *filp) { |
2f38d70fb
|
242 |
f_modown(filp, NULL, PIDTYPE_PID, 1); |
609d7fa95
|
243 244 245 246 247 |
} pid_t f_getown(struct file *filp) { pid_t pid; |
43fa1adb9
|
248 |
read_lock(&filp->f_owner.lock); |
6c5f3e7b4
|
249 |
pid = pid_vnr(filp->f_owner.pid); |
609d7fa95
|
250 251 |
if (filp->f_owner.pid_type == PIDTYPE_PGID) pid = -pid; |
43fa1adb9
|
252 |
read_unlock(&filp->f_owner.lock); |
609d7fa95
|
253 |
return pid; |
1da177e4c
|
254 |
} |
ba0a6c9f6
|
255 256 257 258 259 260 261 262 263 264 |
static int f_setown_ex(struct file *filp, unsigned long arg) { struct f_owner_ex * __user owner_p = (void * __user)arg; struct f_owner_ex owner; struct pid *pid; int type; int ret; ret = copy_from_user(&owner, owner_p, sizeof(owner)); if (ret) |
5b54470da
|
265 |
return -EFAULT; |
ba0a6c9f6
|
266 267 268 269 270 271 272 273 274 |
switch (owner.type) { case F_OWNER_TID: type = PIDTYPE_MAX; break; case F_OWNER_PID: type = PIDTYPE_PID; break; |
978b4053a
|
275 |
case F_OWNER_PGRP: |
ba0a6c9f6
|
276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 |
type = PIDTYPE_PGID; break; default: return -EINVAL; } rcu_read_lock(); pid = find_vpid(owner.pid); if (owner.pid && !pid) ret = -ESRCH; else ret = __f_setown(filp, pid, type, 1); rcu_read_unlock(); return ret; } static int f_getown_ex(struct file *filp, unsigned long arg) { struct f_owner_ex * __user owner_p = (void * __user)arg; struct f_owner_ex owner; int ret = 0; read_lock(&filp->f_owner.lock); owner.pid = pid_vnr(filp->f_owner.pid); switch (filp->f_owner.pid_type) { case PIDTYPE_MAX: owner.type = F_OWNER_TID; break; case PIDTYPE_PID: owner.type = F_OWNER_PID; break; case PIDTYPE_PGID: |
978b4053a
|
312 |
owner.type = F_OWNER_PGRP; |
ba0a6c9f6
|
313 314 315 316 317 318 319 320 |
break; default: WARN_ON(1); ret = -EINVAL; break; } read_unlock(&filp->f_owner.lock); |
5b54470da
|
321 |
if (!ret) { |
ba0a6c9f6
|
322 |
ret = copy_to_user(owner_p, &owner, sizeof(owner)); |
5b54470da
|
323 324 325 |
if (ret) ret = -EFAULT; } |
ba0a6c9f6
|
326 327 |
return ret; } |
1d151c337
|
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 |
#ifdef CONFIG_CHECKPOINT_RESTORE static int f_getowner_uids(struct file *filp, unsigned long arg) { struct user_namespace *user_ns = current_user_ns(); uid_t * __user dst = (void * __user)arg; uid_t src[2]; int err; read_lock(&filp->f_owner.lock); src[0] = from_kuid(user_ns, filp->f_owner.uid); src[1] = from_kuid(user_ns, filp->f_owner.euid); read_unlock(&filp->f_owner.lock); err = put_user(src[0], &dst[0]); err |= put_user(src[1], &dst[1]); return err; } #else static int f_getowner_uids(struct file *filp, unsigned long arg) { return -EINVAL; } #endif |
1da177e4c
|
352 353 354 355 356 357 358 |
static long do_fcntl(int fd, unsigned int cmd, unsigned long arg, struct file *filp) { long err = -EINVAL; switch (cmd) { case F_DUPFD: |
22d2b35b2
|
359 |
case F_DUPFD_CLOEXEC: |
d554ed895
|
360 |
if (arg >= rlimit(RLIMIT_NOFILE)) |
4e1e018ec
|
361 |
break; |
1027abe88
|
362 363 364 365 366 |
err = alloc_fd(arg, cmd == F_DUPFD_CLOEXEC ? O_CLOEXEC : 0); if (err >= 0) { get_file(filp); fd_install(err, filp); } |
1da177e4c
|
367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 |
break; case F_GETFD: err = get_close_on_exec(fd) ? FD_CLOEXEC : 0; break; case F_SETFD: err = 0; set_close_on_exec(fd, arg & FD_CLOEXEC); break; case F_GETFL: err = filp->f_flags; break; case F_SETFL: err = setfl(fd, filp, arg); break; case F_GETLK: err = fcntl_getlk(filp, (struct flock __user *) arg); break; case F_SETLK: case F_SETLKW: |
c293621bb
|
386 |
err = fcntl_setlk(fd, filp, cmd, (struct flock __user *) arg); |
1da177e4c
|
387 388 389 390 391 392 393 394 395 |
break; case F_GETOWN: /* * XXX If f_owner is a process group, the * negative return value will get converted * into an error. Oops. If we keep the * current syscall conventions, the only way * to fix this will be in libc. */ |
609d7fa95
|
396 |
err = f_getown(filp); |
1da177e4c
|
397 398 399 400 401 |
force_successful_syscall_return(); break; case F_SETOWN: err = f_setown(filp, arg, 1); break; |
ba0a6c9f6
|
402 403 404 405 406 407 |
case F_GETOWN_EX: err = f_getown_ex(filp, arg); break; case F_SETOWN_EX: err = f_setown_ex(filp, arg); break; |
1d151c337
|
408 409 410 |
case F_GETOWNER_UIDS: err = f_getowner_uids(filp, arg); break; |
1da177e4c
|
411 412 413 414 415 |
case F_GETSIG: err = filp->f_owner.signum; break; case F_SETSIG: /* arg == 0 restores default behaviour. */ |
7ed20e1ad
|
416 |
if (!valid_signal(arg)) { |
1da177e4c
|
417 418 419 420 421 422 423 424 425 426 427 428 429 430 |
break; } err = 0; filp->f_owner.signum = arg; break; case F_GETLEASE: err = fcntl_getlease(filp); break; case F_SETLEASE: err = fcntl_setlease(fd, filp, arg); break; case F_NOTIFY: err = fcntl_dirnotify(fd, filp, arg); break; |
35f3d14db
|
431 432 433 434 |
case F_SETPIPE_SZ: case F_GETPIPE_SZ: err = pipe_fcntl(filp, cmd, arg); break; |
1da177e4c
|
435 436 437 438 439 |
default: break; } return err; } |
1abf0c718
|
440 441 442 443 444 445 446 447 448 449 450 451 |
static int check_fcntl_cmd(unsigned cmd) { switch (cmd) { case F_DUPFD: case F_DUPFD_CLOEXEC: case F_GETFD: case F_SETFD: case F_GETFL: return 1; } return 0; } |
a26eab240
|
452 |
SYSCALL_DEFINE3(fcntl, unsigned int, fd, unsigned int, cmd, unsigned long, arg) |
1da177e4c
|
453 454 |
{ struct file *filp; |
545ec2c79
|
455 |
int fput_needed; |
1da177e4c
|
456 |
long err = -EBADF; |
545ec2c79
|
457 |
filp = fget_raw_light(fd, &fput_needed); |
1da177e4c
|
458 459 |
if (!filp) goto out; |
1abf0c718
|
460 |
if (unlikely(filp->f_mode & FMODE_PATH)) { |
545ec2c79
|
461 462 |
if (!check_fcntl_cmd(cmd)) goto out1; |
1abf0c718
|
463 |
} |
1da177e4c
|
464 |
err = security_file_fcntl(filp, cmd, arg); |
545ec2c79
|
465 466 |
if (!err) err = do_fcntl(fd, cmd, arg, filp); |
1da177e4c
|
467 |
|
545ec2c79
|
468 469 |
out1: fput_light(filp, fput_needed); |
1da177e4c
|
470 471 472 473 474 |
out: return err; } #if BITS_PER_LONG == 32 |
a26eab240
|
475 476 |
SYSCALL_DEFINE3(fcntl64, unsigned int, fd, unsigned int, cmd, unsigned long, arg) |
1da177e4c
|
477 478 |
{ struct file * filp; |
545ec2c79
|
479 480 |
long err = -EBADF; int fput_needed; |
1da177e4c
|
481 |
|
545ec2c79
|
482 |
filp = fget_raw_light(fd, &fput_needed); |
1da177e4c
|
483 484 |
if (!filp) goto out; |
1abf0c718
|
485 |
if (unlikely(filp->f_mode & FMODE_PATH)) { |
545ec2c79
|
486 487 |
if (!check_fcntl_cmd(cmd)) goto out1; |
1abf0c718
|
488 |
} |
1da177e4c
|
489 |
err = security_file_fcntl(filp, cmd, arg); |
545ec2c79
|
490 491 |
if (err) goto out1; |
1da177e4c
|
492 493 494 495 496 497 498 |
switch (cmd) { case F_GETLK64: err = fcntl_getlk64(filp, (struct flock64 __user *) arg); break; case F_SETLK64: case F_SETLKW64: |
c293621bb
|
499 500 |
err = fcntl_setlk64(fd, filp, cmd, (struct flock64 __user *) arg); |
1da177e4c
|
501 502 503 504 505 |
break; default: err = do_fcntl(fd, cmd, arg, filp); break; } |
545ec2c79
|
506 507 |
out1: fput_light(filp, fput_needed); |
1da177e4c
|
508 509 510 511 512 513 |
out: return err; } #endif /* Table to convert sigio signal codes into poll band bitmaps */ |
fa3536cc1
|
514 |
static const long band_table[NSIGPOLL] = { |
1da177e4c
|
515 516 517 518 519 520 521 522 523 524 525 |
POLLIN | POLLRDNORM, /* POLL_IN */ POLLOUT | POLLWRNORM | POLLWRBAND, /* POLL_OUT */ POLLIN | POLLRDNORM | POLLMSG, /* POLL_MSG */ POLLERR, /* POLL_ERR */ POLLPRI | POLLRDBAND, /* POLL_PRI */ POLLHUP | POLLERR /* POLL_HUP */ }; static inline int sigio_perm(struct task_struct *p, struct fown_struct *fown, int sig) { |
c69e8d9c0
|
526 527 528 529 530 |
const struct cred *cred; int ret; rcu_read_lock(); cred = __task_cred(p); |
8e96e3b7b
|
531 532 533 |
ret = ((uid_eq(fown->euid, GLOBAL_ROOT_UID) || uid_eq(fown->euid, cred->suid) || uid_eq(fown->euid, cred->uid) || uid_eq(fown->uid, cred->suid) || uid_eq(fown->uid, cred->uid)) && |
c69e8d9c0
|
534 535 536 |
!security_file_send_sigiotask(p, fown, sig)); rcu_read_unlock(); return ret; |
1da177e4c
|
537 538 539 |
} static void send_sigio_to_task(struct task_struct *p, |
8eeee4e2f
|
540 |
struct fown_struct *fown, |
ba0a6c9f6
|
541 |
int fd, int reason, int group) |
1da177e4c
|
542 |
{ |
8eeee4e2f
|
543 544 545 546 547 548 549 |
/* * F_SETSIG can change ->signum lockless in parallel, make * sure we read it once and use the same value throughout. */ int signum = ACCESS_ONCE(fown->signum); if (!sigio_perm(p, fown, signum)) |
1da177e4c
|
550 |
return; |
8eeee4e2f
|
551 |
switch (signum) { |
1da177e4c
|
552 553 554 555 556 557 558 559 |
siginfo_t si; default: /* Queue a rt signal with the appropriate fd as its value. We use SI_SIGIO as the source, not SI_KERNEL, since kernel signals always get delivered even if we can't queue. Failure to queue in this case _should_ be reported; we fall back to SIGIO in that case. --sct */ |
8eeee4e2f
|
560 |
si.si_signo = signum; |
1da177e4c
|
561 562 563 564 565 |
si.si_errno = 0; si.si_code = reason; /* Make sure we are called with one of the POLL_* reasons, otherwise we could leak kernel stack into userspace. */ |
f6298aab2
|
566 |
BUG_ON((reason & __SI_MASK) != __SI_POLL); |
1da177e4c
|
567 568 569 570 571 |
if (reason - POLL_IN >= NSIGPOLL) si.si_band = ~0L; else si.si_band = band_table[reason - POLL_IN]; si.si_fd = fd; |
ba0a6c9f6
|
572 |
if (!do_send_sig_info(signum, &si, p, group)) |
1da177e4c
|
573 574 575 |
break; /* fall-through: fall back on the old plain SIGIO signal */ case 0: |
ba0a6c9f6
|
576 |
do_send_sig_info(SIGIO, SEND_SIG_PRIV, p, group); |
1da177e4c
|
577 578 579 580 581 582 |
} } void send_sigio(struct fown_struct *fown, int fd, int band) { struct task_struct *p; |
609d7fa95
|
583 584 |
enum pid_type type; struct pid *pid; |
ba0a6c9f6
|
585 |
int group = 1; |
1da177e4c
|
586 587 |
read_lock(&fown->lock); |
ba0a6c9f6
|
588 |
|
609d7fa95
|
589 |
type = fown->pid_type; |
ba0a6c9f6
|
590 591 592 593 |
if (type == PIDTYPE_MAX) { group = 0; type = PIDTYPE_PID; } |
1da177e4c
|
594 595 596 597 598 |
pid = fown->pid; if (!pid) goto out_unlock_fown; read_lock(&tasklist_lock); |
609d7fa95
|
599 |
do_each_pid_task(pid, type, p) { |
ba0a6c9f6
|
600 |
send_sigio_to_task(p, fown, fd, band, group); |
609d7fa95
|
601 |
} while_each_pid_task(pid, type, p); |
1da177e4c
|
602 603 604 605 606 607 |
read_unlock(&tasklist_lock); out_unlock_fown: read_unlock(&fown->lock); } static void send_sigurg_to_task(struct task_struct *p, |
ba0a6c9f6
|
608 |
struct fown_struct *fown, int group) |
1da177e4c
|
609 610 |
{ if (sigio_perm(p, fown, SIGURG)) |
ba0a6c9f6
|
611 |
do_send_sig_info(SIGURG, SEND_SIG_PRIV, p, group); |
1da177e4c
|
612 613 614 615 616 |
} int send_sigurg(struct fown_struct *fown) { struct task_struct *p; |
609d7fa95
|
617 618 |
enum pid_type type; struct pid *pid; |
ba0a6c9f6
|
619 |
int group = 1; |
609d7fa95
|
620 |
int ret = 0; |
1da177e4c
|
621 622 |
read_lock(&fown->lock); |
ba0a6c9f6
|
623 |
|
609d7fa95
|
624 |
type = fown->pid_type; |
ba0a6c9f6
|
625 626 627 628 |
if (type == PIDTYPE_MAX) { group = 0; type = PIDTYPE_PID; } |
1da177e4c
|
629 630 631 632 633 634 635 |
pid = fown->pid; if (!pid) goto out_unlock_fown; ret = 1; read_lock(&tasklist_lock); |
609d7fa95
|
636 |
do_each_pid_task(pid, type, p) { |
ba0a6c9f6
|
637 |
send_sigurg_to_task(p, fown, group); |
609d7fa95
|
638 |
} while_each_pid_task(pid, type, p); |
1da177e4c
|
639 640 641 642 643 |
read_unlock(&tasklist_lock); out_unlock_fown: read_unlock(&fown->lock); return ret; } |
989a29792
|
644 |
static DEFINE_SPINLOCK(fasync_lock); |
e18b890bb
|
645 |
static struct kmem_cache *fasync_cache __read_mostly; |
1da177e4c
|
646 |
|
989a29792
|
647 648 649 650 651 |
static void fasync_free_rcu(struct rcu_head *head) { kmem_cache_free(fasync_cache, container_of(head, struct fasync_struct, fa_rcu)); } |
1da177e4c
|
652 |
/* |
53281b6d3
|
653 654 655 656 657 658 659 |
* Remove a fasync entry. If successfully removed, return * positive and clear the FASYNC flag. If no entry exists, * do nothing and return 0. * * NOTE! It is very important that the FASYNC flag always * match the state "is the filp on a fasync list". * |
1da177e4c
|
660 |
*/ |
f7347ce4e
|
661 |
int fasync_remove_entry(struct file *filp, struct fasync_struct **fapp) |
1da177e4c
|
662 663 |
{ struct fasync_struct *fa, **fp; |
1da177e4c
|
664 |
int result = 0; |
53281b6d3
|
665 |
spin_lock(&filp->f_lock); |
989a29792
|
666 |
spin_lock(&fasync_lock); |
53281b6d3
|
667 668 669 |
for (fp = fapp; (fa = *fp) != NULL; fp = &fa->fa_next) { if (fa->fa_file != filp) continue; |
989a29792
|
670 671 672 673 |
spin_lock_irq(&fa->fa_lock); fa->fa_file = NULL; spin_unlock_irq(&fa->fa_lock); |
53281b6d3
|
674 |
*fp = fa->fa_next; |
989a29792
|
675 |
call_rcu(&fa->fa_rcu, fasync_free_rcu); |
53281b6d3
|
676 677 678 |
filp->f_flags &= ~FASYNC; result = 1; break; |
1da177e4c
|
679 |
} |
989a29792
|
680 |
spin_unlock(&fasync_lock); |
53281b6d3
|
681 682 683 |
spin_unlock(&filp->f_lock); return result; } |
f7347ce4e
|
684 685 686 687 |
struct fasync_struct *fasync_alloc(void) { return kmem_cache_alloc(fasync_cache, GFP_KERNEL); } |
53281b6d3
|
688 |
/* |
f7347ce4e
|
689 690 691 |
* NOTE! This can be used only for unused fasync entries: * entries that actually got inserted on the fasync list * need to be released by rcu - see fasync_remove_entry. |
53281b6d3
|
692 |
*/ |
f7347ce4e
|
693 |
void fasync_free(struct fasync_struct *new) |
53281b6d3
|
694 |
{ |
f7347ce4e
|
695 696 |
kmem_cache_free(fasync_cache, new); } |
53281b6d3
|
697 |
|
f7347ce4e
|
698 699 700 |
/* * Insert a new entry into the fasync list. Return the pointer to the * old one if we didn't use the new one. |
55f335a88
|
701 702 703 |
* * NOTE! It is very important that the FASYNC flag always * match the state "is the filp on a fasync list". |
f7347ce4e
|
704 705 706 707 |
*/ struct fasync_struct *fasync_insert_entry(int fd, struct file *filp, struct fasync_struct **fapp, struct fasync_struct *new) { struct fasync_struct *fa, **fp; |
4a6a44996
|
708 |
|
4a6a44996
|
709 |
spin_lock(&filp->f_lock); |
989a29792
|
710 |
spin_lock(&fasync_lock); |
1da177e4c
|
711 |
for (fp = fapp; (fa = *fp) != NULL; fp = &fa->fa_next) { |
53281b6d3
|
712 713 |
if (fa->fa_file != filp) continue; |
989a29792
|
714 715 |
spin_lock_irq(&fa->fa_lock); |
53281b6d3
|
716 |
fa->fa_fd = fd; |
989a29792
|
717 |
spin_unlock_irq(&fa->fa_lock); |
53281b6d3
|
718 |
goto out; |
1da177e4c
|
719 |
} |
989a29792
|
720 |
spin_lock_init(&new->fa_lock); |
53281b6d3
|
721 722 723 724 |
new->magic = FASYNC_MAGIC; new->fa_file = filp; new->fa_fd = fd; new->fa_next = *fapp; |
989a29792
|
725 |
rcu_assign_pointer(*fapp, new); |
53281b6d3
|
726 |
filp->f_flags |= FASYNC; |
1da177e4c
|
727 |
out: |
989a29792
|
728 |
spin_unlock(&fasync_lock); |
4a6a44996
|
729 |
spin_unlock(&filp->f_lock); |
f7347ce4e
|
730 731 732 733 734 735 |
return fa; } /* * Add a fasync entry. Return negative on error, positive if * added, and zero if did nothing but change an existing one. |
f7347ce4e
|
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 |
*/ static int fasync_add_entry(int fd, struct file *filp, struct fasync_struct **fapp) { struct fasync_struct *new; new = fasync_alloc(); if (!new) return -ENOMEM; /* * fasync_insert_entry() returns the old (update) entry if * it existed. * * So free the (unused) new entry and return 0 to let the * caller know that we didn't add any new fasync entries. */ if (fasync_insert_entry(fd, filp, fapp, new)) { fasync_free(new); return 0; } return 1; |
1da177e4c
|
758 |
} |
53281b6d3
|
759 760 761 762 763 764 765 766 767 768 769 770 |
/* * fasync_helper() is used by almost all character device drivers * to set up the fasync queue, and for regular files by the file * lease code. It returns negative on error, 0 if it did no changes * and positive if it added/deleted the entry. */ int fasync_helper(int fd, struct file * filp, int on, struct fasync_struct **fapp) { if (!on) return fasync_remove_entry(filp, fapp); return fasync_add_entry(fd, filp, fapp); } |
1da177e4c
|
771 |
EXPORT_SYMBOL(fasync_helper); |
989a29792
|
772 773 774 775 |
/* * rcu_read_lock() is held */ static void kill_fasync_rcu(struct fasync_struct *fa, int sig, int band) |
1da177e4c
|
776 777 |
{ while (fa) { |
989a29792
|
778 |
struct fown_struct *fown; |
f4985dc71
|
779 |
unsigned long flags; |
1da177e4c
|
780 781 782 783 784 785 |
if (fa->magic != FASYNC_MAGIC) { printk(KERN_ERR "kill_fasync: bad magic number in " "fasync_struct! "); return; } |
f4985dc71
|
786 |
spin_lock_irqsave(&fa->fa_lock, flags); |
989a29792
|
787 788 789 790 791 792 793 794 |
if (fa->fa_file) { fown = &fa->fa_file->f_owner; /* Don't send SIGURG to processes which have not set a queued signum: SIGURG has its own default signalling mechanism. */ if (!(sig == SIGURG && fown->signum == 0)) send_sigio(fown, fa->fa_fd, band); } |
f4985dc71
|
795 |
spin_unlock_irqrestore(&fa->fa_lock, flags); |
989a29792
|
796 |
fa = rcu_dereference(fa->fa_next); |
1da177e4c
|
797 798 |
} } |
1da177e4c
|
799 800 801 802 803 804 |
void kill_fasync(struct fasync_struct **fp, int sig, int band) { /* First a quick test without locking: usually * the list is empty. */ if (*fp) { |
989a29792
|
805 806 807 |
rcu_read_lock(); kill_fasync_rcu(rcu_dereference(*fp), sig, band); rcu_read_unlock(); |
1da177e4c
|
808 809 810 |
} } EXPORT_SYMBOL(kill_fasync); |
454eedb89
|
811 |
static int __init fcntl_init(void) |
1da177e4c
|
812 |
{ |
3ab04d5cf
|
813 814 815 816 817 |
/* * Please add new bits here to ensure allocation uniqueness. * Exceptions: O_NONBLOCK is a two bit define on parisc; O_NDELAY * is defined as O_NONBLOCK on some platforms and not on others. */ |
1abf0c718
|
818 |
BUILD_BUG_ON(19 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( |
454eedb89
|
819 820 |
O_RDONLY | O_WRONLY | O_RDWR | O_CREAT | O_EXCL | O_NOCTTY | |
3ab04d5cf
|
821 |
O_TRUNC | O_APPEND | /* O_NONBLOCK | */ |
454eedb89
|
822 823 824 |
__O_SYNC | O_DSYNC | FASYNC | O_DIRECT | O_LARGEFILE | O_DIRECTORY | O_NOFOLLOW | O_NOATIME | O_CLOEXEC | |
1abf0c718
|
825 |
__FMODE_EXEC | O_PATH |
454eedb89
|
826 |
)); |
1da177e4c
|
827 |
fasync_cache = kmem_cache_create("fasync_cache", |
20c2df83d
|
828 |
sizeof(struct fasync_struct), 0, SLAB_PANIC, NULL); |
1da177e4c
|
829 830 |
return 0; } |
454eedb89
|
831 |
module_init(fcntl_init) |