Blame view
fs/eventpoll.c
44.5 KB
1da177e4c
|
1 |
/* |
5071f97ec
|
2 3 |
* fs/eventpoll.c (Efficient event retrieval implementation) * Copyright (C) 2001,...,2009 Davide Libenzi |
1da177e4c
|
4 5 6 7 8 9 10 11 12 |
* * 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. * * Davide Libenzi <davidel@xmailserver.org> * */ |
1da177e4c
|
13 14 15 16 17 18 19 20 21 22 |
#include <linux/init.h> #include <linux/kernel.h> #include <linux/sched.h> #include <linux/fs.h> #include <linux/file.h> #include <linux/signal.h> #include <linux/errno.h> #include <linux/mm.h> #include <linux/slab.h> #include <linux/poll.h> |
1da177e4c
|
23 24 25 26 27 |
#include <linux/string.h> #include <linux/list.h> #include <linux/hash.h> #include <linux/spinlock.h> #include <linux/syscalls.h> |
1da177e4c
|
28 29 30 31 32 |
#include <linux/rbtree.h> #include <linux/wait.h> #include <linux/eventpoll.h> #include <linux/mount.h> #include <linux/bitops.h> |
144efe3e3
|
33 |
#include <linux/mutex.h> |
da66f7cb0
|
34 |
#include <linux/anon_inodes.h> |
1da177e4c
|
35 36 37 38 |
#include <asm/uaccess.h> #include <asm/system.h> #include <asm/io.h> #include <asm/mman.h> |
60063497a
|
39 |
#include <linux/atomic.h> |
1da177e4c
|
40 |
|
1da177e4c
|
41 42 43 44 |
/* * LOCKING: * There are three level of locking required by epoll : * |
144efe3e3
|
45 |
* 1) epmutex (mutex) |
c7ea76302
|
46 47 |
* 2) ep->mtx (mutex) * 3) ep->lock (spinlock) |
1da177e4c
|
48 49 50 51 52 53 54 55 56 |
* * The acquire order is the one listed above, from 1 to 3. * We need a spinlock (ep->lock) because we manipulate objects * from inside the poll callback, that might be triggered from * a wake_up() that in turn might be called from IRQ context. * So we can't sleep inside the poll callback and hence we need * a spinlock. During the event transfer loop (from kernel to * user space) we could end up sleeping due a copy_to_user(), so * we need a lock that will allow us to sleep. This lock is a |
d47de16c7
|
57 58 59 60 61 |
* mutex (ep->mtx). It is acquired during the event transfer loop, * during epoll_ctl(EPOLL_CTL_DEL) and during eventpoll_release_file(). * Then we also need a global mutex to serialize eventpoll_release_file() * and ep_free(). * This mutex is acquired by ep_free() during the epoll file |
1da177e4c
|
62 63 |
* cleanup path and it is also acquired by eventpoll_release_file() * if a file has been pushed inside an epoll set and it is then |
bf6a41db7
|
64 |
* close()d without a previous call to epoll_ctl(EPOLL_CTL_DEL). |
22bacca48
|
65 66 67 68 69 70 71 |
* It is also acquired when inserting an epoll fd onto another epoll * fd. We do this so that we walk the epoll tree and ensure that this * insertion does not create a cycle of epoll file descriptors, which * could lead to deadlock. We need a global mutex to prevent two * simultaneous inserts (A into B and B into A) from racing and * constructing a cycle without either insert observing that it is * going to. |
d8805e633
|
72 73 74 75 76 77 78 79 80 |
* It is necessary to acquire multiple "ep->mtx"es at once in the * case when one epoll fd is added to another. In this case, we * always acquire the locks in the order of nesting (i.e. after * epoll_ctl(e1, EPOLL_CTL_ADD, e2), e1->mtx will always be acquired * before e2->mtx). Since we disallow cycles of epoll file * descriptors, this ensures that the mutexes are well-ordered. In * order to communicate this nesting to lockdep, when walking a tree * of epoll file descriptors, we use the current recursion depth as * the lockdep subkey. |
d47de16c7
|
81 82 83 |
* It is possible to drop the "ep->mtx" and to use the global * mutex "epmutex" (together with "ep->lock") to have it working, * but having "ep->mtx" will make the interface more scalable. |
144efe3e3
|
84 |
* Events that require holding "epmutex" are very rare, while for |
d47de16c7
|
85 86 |
* normal operations the epoll private "ep->mtx" will guarantee * a better scalability. |
1da177e4c
|
87 |
*/ |
1da177e4c
|
88 89 |
/* Epoll private bits inside the event mask */ #define EP_PRIVATE_BITS (EPOLLONESHOT | EPOLLET) |
5071f97ec
|
90 91 |
/* Maximum number of nesting allowed inside epoll sets */ #define EP_MAX_NESTS 4 |
1da177e4c
|
92 |
|
b611967de
|
93 |
#define EP_MAX_EVENTS (INT_MAX / sizeof(struct epoll_event)) |
d47de16c7
|
94 |
#define EP_UNACTIVE_PTR ((void *) -1L) |
7ef9964e6
|
95 |
#define EP_ITEM_COST (sizeof(struct epitem) + sizeof(struct eppoll_entry)) |
1da177e4c
|
96 97 98 99 100 101 |
struct epoll_filefd { struct file *file; int fd; }; /* |
5071f97ec
|
102 103 |
* Structure used to track possible nested calls, for too deep recursions * and loop cycles. |
1da177e4c
|
104 |
*/ |
5071f97ec
|
105 |
struct nested_call_node { |
1da177e4c
|
106 |
struct list_head llink; |
5071f97ec
|
107 |
void *cookie; |
3fe4a975d
|
108 |
void *ctx; |
1da177e4c
|
109 110 111 |
}; /* |
5071f97ec
|
112 113 |
* This structure is used as collector for nested calls, to check for * maximum recursion dept and loop cycles. |
1da177e4c
|
114 |
*/ |
5071f97ec
|
115 116 |
struct nested_calls { struct list_head tasks_call_list; |
1da177e4c
|
117 118 119 120 |
spinlock_t lock; }; /* |
d47de16c7
|
121 122 123 124 |
* Each file descriptor added to the eventpoll interface will * have an entry of this type linked to the "rbr" RB tree. */ struct epitem { |
67647d0fb
|
125 |
/* RB tree node used to link this structure to the eventpoll RB tree */ |
d47de16c7
|
126 127 128 129 |
struct rb_node rbn; /* List header used to link this structure to the eventpoll ready list */ struct list_head rdllink; |
c7ea76302
|
130 131 132 133 134 |
/* * Works together "struct eventpoll"->ovflist in keeping the * single linked chain of items. */ struct epitem *next; |
d47de16c7
|
135 136 137 138 139 140 141 142 143 144 145 |
/* The file descriptor information this item refers to */ struct epoll_filefd ffd; /* Number of active wait queue attached to poll operations */ int nwait; /* List containing poll wait queues */ struct list_head pwqlist; /* The "container" of this item */ struct eventpoll *ep; |
d47de16c7
|
146 147 |
/* List header used to link this item to the "struct file" items list */ struct list_head fllink; |
c7ea76302
|
148 149 |
/* The structure that describe the interested events and the source fd */ struct epoll_event event; |
d47de16c7
|
150 151 152 |
}; /* |
1da177e4c
|
153 |
* This structure is stored inside the "private_data" member of the file |
bf6a41db7
|
154 |
* structure and represents the main data structure for the eventpoll |
1da177e4c
|
155 156 157 |
* interface. */ struct eventpoll { |
bf6a41db7
|
158 |
/* Protect the access to this structure */ |
c7ea76302
|
159 |
spinlock_t lock; |
1da177e4c
|
160 161 |
/* |
d47de16c7
|
162 163 164 165 |
* This mutex is used to ensure that files are not removed * while epoll is using them. This is held during the event * collection loop, the file cleanup path, the epoll file exit * code and the ctl operations. |
1da177e4c
|
166 |
*/ |
d47de16c7
|
167 |
struct mutex mtx; |
1da177e4c
|
168 169 170 171 172 173 174 175 176 |
/* Wait queue used by sys_epoll_wait() */ wait_queue_head_t wq; /* Wait queue used by file->poll() */ wait_queue_head_t poll_wait; /* List of ready file descriptors */ struct list_head rdllist; |
67647d0fb
|
177 |
/* RB tree root used to store monitored fd structs */ |
1da177e4c
|
178 |
struct rb_root rbr; |
d47de16c7
|
179 180 181 |
/* * This is a single linked list that chains all the "struct epitem" that |
25985edce
|
182 |
* happened while transferring ready events to userspace w/out |
d47de16c7
|
183 184 185 |
* holding ->lock. */ struct epitem *ovflist; |
7ef9964e6
|
186 187 188 |
/* The user that created the eventpoll descriptor */ struct user_struct *user; |
1da177e4c
|
189 190 191 192 193 194 195 196 |
}; /* Wait structure used by the poll hooks */ struct eppoll_entry { /* List header used to link this structure to the "struct epitem" */ struct list_head llink; /* The "base" pointer is set to the container "struct epitem" */ |
4f0989dbf
|
197 |
struct epitem *base; |
1da177e4c
|
198 199 200 201 202 203 204 205 206 207 |
/* * Wait queue item that will be linked to the target file wait * queue head. */ wait_queue_t wait; /* The wait queue head that linked the "wait" wait queue item */ wait_queue_head_t *whead; }; |
1da177e4c
|
208 209 210 211 212 |
/* Wrapper struct used by poll queueing */ struct ep_pqueue { poll_table pt; struct epitem *epi; }; |
5071f97ec
|
213 214 215 216 217 |
/* Used by the ep_send_events() function as callback private data */ struct ep_send_events_data { int maxevents; struct epoll_event __user *events; }; |
1da177e4c
|
218 |
/* |
7ef9964e6
|
219 220 |
* Configuration options available inside /proc/sys/fs/epoll/ */ |
7ef9964e6
|
221 |
/* Maximum number of epoll watched descriptors, per user */ |
52bd19f76
|
222 |
static long max_user_watches __read_mostly; |
7ef9964e6
|
223 224 |
/* |
d47de16c7
|
225 |
* This mutex is used to serialize ep_free() and eventpoll_release_file(). |
1da177e4c
|
226 |
*/ |
7ef9964e6
|
227 |
static DEFINE_MUTEX(epmutex); |
1da177e4c
|
228 |
|
22bacca48
|
229 230 |
/* Used to check for epoll file descriptor inclusion loops */ static struct nested_calls poll_loop_ncalls; |
5071f97ec
|
231 232 233 234 235 |
/* Used for safe wake up implementation */ static struct nested_calls poll_safewake_ncalls; /* Used to call file's f_op->poll() under the nested calls boundaries */ static struct nested_calls poll_readywalk_ncalls; |
1da177e4c
|
236 237 |
/* Slab cache used to allocate "struct epitem" */ |
e18b890bb
|
238 |
static struct kmem_cache *epi_cache __read_mostly; |
1da177e4c
|
239 240 |
/* Slab cache used to allocate "struct eppoll_entry" */ |
e18b890bb
|
241 |
static struct kmem_cache *pwq_cache __read_mostly; |
1da177e4c
|
242 |
|
7ef9964e6
|
243 244 245 |
#ifdef CONFIG_SYSCTL #include <linux/sysctl.h> |
52bd19f76
|
246 247 |
static long zero; static long long_max = LONG_MAX; |
7ef9964e6
|
248 249 250 |
ctl_table epoll_table[] = { { |
7ef9964e6
|
251 252 |
.procname = "max_user_watches", .data = &max_user_watches, |
52bd19f76
|
253 |
.maxlen = sizeof(max_user_watches), |
7ef9964e6
|
254 |
.mode = 0644, |
52bd19f76
|
255 |
.proc_handler = proc_doulongvec_minmax, |
7ef9964e6
|
256 |
.extra1 = &zero, |
52bd19f76
|
257 |
.extra2 = &long_max, |
7ef9964e6
|
258 |
}, |
ab09203e3
|
259 |
{ } |
7ef9964e6
|
260 261 |
}; #endif /* CONFIG_SYSCTL */ |
b030a4dd6
|
262 |
|
67647d0fb
|
263 |
/* Setup the structure that is used as key for the RB tree */ |
b030a4dd6
|
264 265 266 267 268 269 |
static inline void ep_set_ffd(struct epoll_filefd *ffd, struct file *file, int fd) { ffd->file = file; ffd->fd = fd; } |
67647d0fb
|
270 |
/* Compare RB tree keys */ |
b030a4dd6
|
271 272 273 274 275 276 |
static inline int ep_cmp_ffd(struct epoll_filefd *p1, struct epoll_filefd *p2) { return (p1->file > p2->file ? +1: (p1->file < p2->file ? -1 : p1->fd - p2->fd)); } |
b030a4dd6
|
277 278 279 280 281 282 283 |
/* Tells us if the item is currently linked */ static inline int ep_is_linked(struct list_head *p) { return !list_empty(p); } /* Get the "struct epitem" from a wait queue pointer */ |
cdac75e6f
|
284 |
static inline struct epitem *ep_item_from_wait(wait_queue_t *p) |
b030a4dd6
|
285 286 287 288 289 |
{ return container_of(p, struct eppoll_entry, wait)->base; } /* Get the "struct epitem" from an epoll queue wrapper */ |
cdac75e6f
|
290 |
static inline struct epitem *ep_item_from_epqueue(poll_table *p) |
b030a4dd6
|
291 292 293 294 295 |
{ return container_of(p, struct ep_pqueue, pt)->epi; } /* Tells if the epoll_ctl(2) operation needs an event copy from userspace */ |
6192bd536
|
296 |
static inline int ep_op_has_event(int op) |
b030a4dd6
|
297 298 299 |
{ return op != EPOLL_CTL_DEL; } |
1da177e4c
|
300 |
/* Initialize the poll safe wake up structure */ |
5071f97ec
|
301 |
static void ep_nested_calls_init(struct nested_calls *ncalls) |
1da177e4c
|
302 |
{ |
5071f97ec
|
303 304 |
INIT_LIST_HEAD(&ncalls->tasks_call_list); spin_lock_init(&ncalls->lock); |
1da177e4c
|
305 |
} |
5071f97ec
|
306 |
/** |
3fb0e584a
|
307 308 309 310 311 312 313 314 315 316 317 318 319 |
* ep_events_available - Checks if ready events might be available. * * @ep: Pointer to the eventpoll context. * * Returns: Returns a value different than zero if ready events are available, * or zero otherwise. */ static inline int ep_events_available(struct eventpoll *ep) { return !list_empty(&ep->rdllist) || ep->ovflist != EP_UNACTIVE_PTR; } /** |
5071f97ec
|
320 321 322 323 324 325 326 327 328 329 |
* ep_call_nested - Perform a bound (possibly) nested call, by checking * that the recursion limit is not exceeded, and that * the same nested call (by the meaning of same cookie) is * no re-entered. * * @ncalls: Pointer to the nested_calls structure to be used for this call. * @max_nests: Maximum number of allowed nesting calls. * @nproc: Nested call core function pointer. * @priv: Opaque data to be passed to the @nproc callback. * @cookie: Cookie to be used to identify this nested call. |
3fe4a975d
|
330 |
* @ctx: This instance context. |
5071f97ec
|
331 332 333 |
* * Returns: Returns the code returned by the @nproc callback, or -1 if * the maximum recursion limit has been exceeded. |
1da177e4c
|
334 |
*/ |
5071f97ec
|
335 336 |
static int ep_call_nested(struct nested_calls *ncalls, int max_nests, int (*nproc)(void *, void *, int), void *priv, |
3fe4a975d
|
337 |
void *cookie, void *ctx) |
1da177e4c
|
338 |
{ |
5071f97ec
|
339 |
int error, call_nests = 0; |
1da177e4c
|
340 |
unsigned long flags; |
5071f97ec
|
341 342 343 |
struct list_head *lsthead = &ncalls->tasks_call_list; struct nested_call_node *tncur; struct nested_call_node tnode; |
1da177e4c
|
344 |
|
5071f97ec
|
345 |
spin_lock_irqsave(&ncalls->lock, flags); |
1da177e4c
|
346 |
|
5071f97ec
|
347 348 349 350 351 |
/* * Try to see if the current task is already inside this wakeup call. * We use a list here, since the population inside this set is always * very much limited. */ |
b70c39409
|
352 |
list_for_each_entry(tncur, lsthead, llink) { |
3fe4a975d
|
353 |
if (tncur->ctx == ctx && |
5071f97ec
|
354 |
(tncur->cookie == cookie || ++call_nests > max_nests)) { |
1da177e4c
|
355 356 357 358 |
/* * Ops ... loop detected or maximum nest level reached. * We abort this wake by breaking the cycle itself. */ |
abff55cee
|
359 360 |
error = -1; goto out_unlock; |
1da177e4c
|
361 362 |
} } |
5071f97ec
|
363 |
/* Add the current task and cookie to the list */ |
3fe4a975d
|
364 |
tnode.ctx = ctx; |
5071f97ec
|
365 |
tnode.cookie = cookie; |
1da177e4c
|
366 |
list_add(&tnode.llink, lsthead); |
5071f97ec
|
367 |
spin_unlock_irqrestore(&ncalls->lock, flags); |
1da177e4c
|
368 |
|
5071f97ec
|
369 370 |
/* Call the nested function */ error = (*nproc)(priv, cookie, call_nests); |
1da177e4c
|
371 372 |
/* Remove the current task from the list */ |
5071f97ec
|
373 |
spin_lock_irqsave(&ncalls->lock, flags); |
1da177e4c
|
374 |
list_del(&tnode.llink); |
3fe4a975d
|
375 |
out_unlock: |
5071f97ec
|
376 377 378 379 |
spin_unlock_irqrestore(&ncalls->lock, flags); return error; } |
2dfa4eeab
|
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 |
#ifdef CONFIG_DEBUG_LOCK_ALLOC static inline void ep_wake_up_nested(wait_queue_head_t *wqueue, unsigned long events, int subclass) { unsigned long flags; spin_lock_irqsave_nested(&wqueue->lock, flags, subclass); wake_up_locked_poll(wqueue, events); spin_unlock_irqrestore(&wqueue->lock, flags); } #else static inline void ep_wake_up_nested(wait_queue_head_t *wqueue, unsigned long events, int subclass) { wake_up_poll(wqueue, events); } #endif |
5071f97ec
|
397 398 |
static int ep_poll_wakeup_proc(void *priv, void *cookie, int call_nests) { |
2dfa4eeab
|
399 400 |
ep_wake_up_nested((wait_queue_head_t *) cookie, POLLIN, 1 + call_nests); |
5071f97ec
|
401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 |
return 0; } /* * Perform a safe wake up of the poll wait list. The problem is that * with the new callback'd wake up system, it is possible that the * poll callback is reentered from inside the call to wake_up() done * on the poll wait queue head. The rule is that we cannot reenter the * wake up code from the same task more than EP_MAX_NESTS times, * and we cannot reenter the same wait queue head at all. This will * enable to have a hierarchy of epoll file descriptor of no more than * EP_MAX_NESTS deep. */ static void ep_poll_safewake(wait_queue_head_t *wq) { |
3fe4a975d
|
416 |
int this_cpu = get_cpu(); |
5071f97ec
|
417 |
ep_call_nested(&poll_safewake_ncalls, EP_MAX_NESTS, |
3fe4a975d
|
418 419 420 |
ep_poll_wakeup_proc, NULL, wq, (void *) (long) this_cpu); put_cpu(); |
1da177e4c
|
421 |
} |
1da177e4c
|
422 |
/* |
d1bc90dd5
|
423 424 425 |
* This function unregisters poll callbacks from the associated file * descriptor. Must be called with "mtx" held (or "epmutex" if called from * ep_free). |
1da177e4c
|
426 |
*/ |
7699acd13
|
427 |
static void ep_unregister_pollwait(struct eventpoll *ep, struct epitem *epi) |
1da177e4c
|
428 |
{ |
7699acd13
|
429 430 |
struct list_head *lsthead = &epi->pwqlist; struct eppoll_entry *pwq; |
1da177e4c
|
431 |
|
d1bc90dd5
|
432 433 |
while (!list_empty(lsthead)) { pwq = list_first_entry(lsthead, struct eppoll_entry, llink); |
1da177e4c
|
434 |
|
d1bc90dd5
|
435 436 437 |
list_del(&pwq->llink); remove_wait_queue(pwq->whead, &pwq->wait); kmem_cache_free(pwq_cache, pwq); |
1da177e4c
|
438 |
} |
1da177e4c
|
439 |
} |
5071f97ec
|
440 441 442 443 444 445 446 447 |
/** * ep_scan_ready_list - Scans the ready list in a way that makes possible for * the scan code, to call f_op->poll(). Also allows for * O(NumReady) performance. * * @ep: Pointer to the epoll private data structure. * @sproc: Pointer to the scan callback. * @priv: Private opaque data passed to the @sproc callback. |
d8805e633
|
448 |
* @depth: The current depth of recursive f_op->poll calls. |
5071f97ec
|
449 450 451 452 453 454 |
* * Returns: The same integer error code returned by the @sproc callback. */ static int ep_scan_ready_list(struct eventpoll *ep, int (*sproc)(struct eventpoll *, struct list_head *, void *), |
d8805e633
|
455 456 |
void *priv, int depth) |
5071f97ec
|
457 458 459 460 |
{ int error, pwake = 0; unsigned long flags; struct epitem *epi, *nepi; |
296e236e9
|
461 |
LIST_HEAD(txlist); |
5071f97ec
|
462 463 464 |
/* * We need to lock this because we could be hit by |
e057e15ff
|
465 |
* eventpoll_release_file() and epoll_ctl(). |
5071f97ec
|
466 |
*/ |
d8805e633
|
467 |
mutex_lock_nested(&ep->mtx, depth); |
5071f97ec
|
468 469 470 471 472 473 474 475 476 477 |
/* * Steal the ready list, and re-init the original one to the * empty list. Also, set ep->ovflist to NULL so that events * happening while looping w/out locks, are not lost. We cannot * have the poll callback to queue directly on ep->rdllist, * because we want the "sproc" callback to be able to do it * in a lockless way. */ spin_lock_irqsave(&ep->lock, flags); |
296e236e9
|
478 |
list_splice_init(&ep->rdllist, &txlist); |
5071f97ec
|
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 |
ep->ovflist = NULL; spin_unlock_irqrestore(&ep->lock, flags); /* * Now call the callback function. */ error = (*sproc)(ep, &txlist, priv); spin_lock_irqsave(&ep->lock, flags); /* * During the time we spent inside the "sproc" callback, some * other events might have been queued by the poll callback. * We re-insert them inside the main ready-list here. */ for (nepi = ep->ovflist; (epi = nepi) != NULL; nepi = epi->next, epi->next = EP_UNACTIVE_PTR) { /* * We need to check if the item is already in the list. * During the "sproc" callback execution time, items are * queued into ->ovflist but the "txlist" might already * contain them, and the list_splice() below takes care of them. */ if (!ep_is_linked(&epi->rdllink)) list_add_tail(&epi->rdllink, &ep->rdllist); } /* * We need to set back ep->ovflist to EP_UNACTIVE_PTR, so that after * releasing the lock, events will be queued in the normal way inside * ep->rdllist. */ ep->ovflist = EP_UNACTIVE_PTR; /* * Quickly re-inject items left on "txlist". */ list_splice(&txlist, &ep->rdllist); if (!list_empty(&ep->rdllist)) { /* |
296e236e9
|
518 519 |
* Wake up (if active) both the eventpoll wait list and * the ->poll() wait list (delayed after we release the lock). |
5071f97ec
|
520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 |
*/ if (waitqueue_active(&ep->wq)) wake_up_locked(&ep->wq); if (waitqueue_active(&ep->poll_wait)) pwake++; } spin_unlock_irqrestore(&ep->lock, flags); mutex_unlock(&ep->mtx); /* We have to call this outside the lock */ if (pwake) ep_poll_safewake(&ep->poll_wait); return error; } |
1da177e4c
|
536 |
/* |
7699acd13
|
537 |
* Removes a "struct epitem" from the eventpoll RB tree and deallocates |
c7ea76302
|
538 |
* all the associated resources. Must be called with "mtx" held. |
7699acd13
|
539 540 541 |
*/ static int ep_remove(struct eventpoll *ep, struct epitem *epi) { |
7699acd13
|
542 543 |
unsigned long flags; struct file *file = epi->ffd.file; |
1da177e4c
|
544 545 |
/* |
7699acd13
|
546 547 548 549 550 551 |
* Removes poll wait queue hooks. We _have_ to do this without holding * the "ep->lock" otherwise a deadlock might occur. This because of the * sequence of the lock acquisition. Here we do "ep->lock" then the wait * queue head lock when unregistering the wait queue. The wakeup callback * will run by holding the wait queue head lock and will call our callback * that will try to get "ep->lock". |
1da177e4c
|
552 |
*/ |
7699acd13
|
553 |
ep_unregister_pollwait(ep, epi); |
1da177e4c
|
554 |
|
7699acd13
|
555 |
/* Remove the current item from the list of epoll hooks */ |
684999149
|
556 |
spin_lock(&file->f_lock); |
7699acd13
|
557 558 |
if (ep_is_linked(&epi->fllink)) list_del_init(&epi->fllink); |
684999149
|
559 |
spin_unlock(&file->f_lock); |
1da177e4c
|
560 |
|
cdac75e6f
|
561 |
rb_erase(&epi->rbn, &ep->rbr); |
1da177e4c
|
562 |
|
c7ea76302
|
563 564 565 566 |
spin_lock_irqsave(&ep->lock, flags); if (ep_is_linked(&epi->rdllink)) list_del_init(&epi->rdllink); spin_unlock_irqrestore(&ep->lock, flags); |
1da177e4c
|
567 |
|
7699acd13
|
568 |
/* At this point it is safe to free the eventpoll item */ |
c7ea76302
|
569 |
kmem_cache_free(epi_cache, epi); |
1da177e4c
|
570 |
|
52bd19f76
|
571 |
atomic_long_dec(&ep->user->epoll_watches); |
7ef9964e6
|
572 |
|
c7ea76302
|
573 |
return 0; |
1da177e4c
|
574 |
} |
7699acd13
|
575 |
static void ep_free(struct eventpoll *ep) |
1da177e4c
|
576 |
{ |
7699acd13
|
577 578 |
struct rb_node *rbp; struct epitem *epi; |
1da177e4c
|
579 |
|
7699acd13
|
580 581 |
/* We need to release all tasks waiting for these file */ if (waitqueue_active(&ep->poll_wait)) |
5071f97ec
|
582 |
ep_poll_safewake(&ep->poll_wait); |
1da177e4c
|
583 |
|
7699acd13
|
584 585 586 |
/* * We need to lock this because we could be hit by * eventpoll_release_file() while we're freeing the "struct eventpoll". |
d47de16c7
|
587 |
* We do not need to hold "ep->mtx" here because the epoll file |
7699acd13
|
588 589 |
* is on the way to be removed and no one has references to it * anymore. The only hit might come from eventpoll_release_file() but |
25985edce
|
590 |
* holding "epmutex" is sufficient here. |
7699acd13
|
591 592 |
*/ mutex_lock(&epmutex); |
1da177e4c
|
593 594 |
/* |
7699acd13
|
595 |
* Walks through the whole tree by unregistering poll callbacks. |
1da177e4c
|
596 |
*/ |
7699acd13
|
597 598 599 600 601 |
for (rbp = rb_first(&ep->rbr); rbp; rbp = rb_next(rbp)) { epi = rb_entry(rbp, struct epitem, rbn); ep_unregister_pollwait(ep, epi); } |
1da177e4c
|
602 603 |
/* |
7699acd13
|
604 605 |
* Walks through the whole tree by freeing each "struct epitem". At this * point we are sure no poll callbacks will be lingering around, and also by |
d47de16c7
|
606 |
* holding "epmutex" we can be sure that no file cleanup code will hit |
7699acd13
|
607 |
* us during this operation. So we can avoid the lock on "ep->lock". |
1da177e4c
|
608 |
*/ |
c80544dc0
|
609 |
while ((rbp = rb_first(&ep->rbr)) != NULL) { |
7699acd13
|
610 611 612 |
epi = rb_entry(rbp, struct epitem, rbn); ep_remove(ep, epi); } |
1da177e4c
|
613 |
|
7699acd13
|
614 |
mutex_unlock(&epmutex); |
d47de16c7
|
615 |
mutex_destroy(&ep->mtx); |
7ef9964e6
|
616 |
free_uid(ep->user); |
f0ee9aabb
|
617 |
kfree(ep); |
7699acd13
|
618 |
} |
1da177e4c
|
619 |
|
7699acd13
|
620 621 622 |
static int ep_eventpoll_release(struct inode *inode, struct file *file) { struct eventpoll *ep = file->private_data; |
1da177e4c
|
623 |
|
f0ee9aabb
|
624 |
if (ep) |
7699acd13
|
625 |
ep_free(ep); |
7699acd13
|
626 |
|
7699acd13
|
627 |
return 0; |
1da177e4c
|
628 |
} |
296e236e9
|
629 630 |
static int ep_read_events_proc(struct eventpoll *ep, struct list_head *head, void *priv) |
5071f97ec
|
631 632 633 634 635 636 637 |
{ struct epitem *epi, *tmp; list_for_each_entry_safe(epi, tmp, head, rdllink) { if (epi->ffd.file->f_op->poll(epi->ffd.file, NULL) & epi->event.events) return POLLIN | POLLRDNORM; |
296e236e9
|
638 |
else { |
5071f97ec
|
639 640 641 642 643 644 |
/* * Item has been dropped into the ready list by the poll * callback, but it's not actually ready, as far as * caller requested events goes. We can remove it here. */ list_del_init(&epi->rdllink); |
296e236e9
|
645 |
} |
5071f97ec
|
646 647 648 649 650 651 652 |
} return 0; } static int ep_poll_readyevents_proc(void *priv, void *cookie, int call_nests) { |
d8805e633
|
653 |
return ep_scan_ready_list(priv, ep_read_events_proc, NULL, call_nests + 1); |
5071f97ec
|
654 |
} |
7699acd13
|
655 656 |
static unsigned int ep_eventpoll_poll(struct file *file, poll_table *wait) { |
5071f97ec
|
657 |
int pollflags; |
7699acd13
|
658 |
struct eventpoll *ep = file->private_data; |
1da177e4c
|
659 |
|
7699acd13
|
660 661 |
/* Insert inside our poll wait queue */ poll_wait(file, &ep->poll_wait, wait); |
5071f97ec
|
662 663 664 665 666 667 668 |
/* * Proceed to find out if wanted events are really available inside * the ready list. This need to be done under ep_call_nested() * supervision, since the call to f_op->poll() done on listed files * could re-enter here. */ pollflags = ep_call_nested(&poll_readywalk_ncalls, EP_MAX_NESTS, |
3fe4a975d
|
669 |
ep_poll_readyevents_proc, ep, ep, current); |
7699acd13
|
670 |
|
296e236e9
|
671 |
return pollflags != -1 ? pollflags : 0; |
7699acd13
|
672 673 674 675 676 |
} /* File callbacks that implement the eventpoll file behaviour */ static const struct file_operations eventpoll_fops = { .release = ep_eventpoll_release, |
6038f373a
|
677 678 |
.poll = ep_eventpoll_poll, .llseek = noop_llseek, |
7699acd13
|
679 |
}; |
10bd2473b
|
680 |
/* Fast test to see if the file is an eventpoll file */ |
7699acd13
|
681 682 683 684 |
static inline int is_file_epoll(struct file *f) { return f->f_op == &eventpoll_fops; } |
b611967de
|
685 686 |
/* |
7699acd13
|
687 688 689 |
* This is called from eventpoll_release() to unlink files from the eventpoll * interface. We need to have this facility to cleanup correctly files that are * closed without being removed from the eventpoll interface. |
b611967de
|
690 |
*/ |
7699acd13
|
691 |
void eventpoll_release_file(struct file *file) |
b611967de
|
692 |
{ |
7699acd13
|
693 694 695 |
struct list_head *lsthead = &file->f_ep_links; struct eventpoll *ep; struct epitem *epi; |
b611967de
|
696 697 |
/* |
684999149
|
698 |
* We don't want to get "file->f_lock" because it is not |
7699acd13
|
699 |
* necessary. It is not necessary because we're in the "struct file" |
25985edce
|
700 |
* cleanup path, and this means that no one is using this file anymore. |
5071f97ec
|
701 |
* So, for example, epoll_ctl() cannot hit here since if we reach this |
67647d0fb
|
702 |
* point, the file counter already went to zero and fget() would fail. |
d47de16c7
|
703 |
* The only hit might come from ep_free() but by holding the mutex |
7699acd13
|
704 |
* will correctly serialize the operation. We do need to acquire |
d47de16c7
|
705 |
* "ep->mtx" after "epmutex" because ep_remove() requires it when called |
7699acd13
|
706 |
* from anywhere but ep_free(). |
684999149
|
707 708 |
* * Besides, ep_remove() acquires the lock, so we can't hold it here. |
b611967de
|
709 |
*/ |
7699acd13
|
710 |
mutex_lock(&epmutex); |
b611967de
|
711 |
|
7699acd13
|
712 713 |
while (!list_empty(lsthead)) { epi = list_first_entry(lsthead, struct epitem, fllink); |
b611967de
|
714 |
|
7699acd13
|
715 716 |
ep = epi->ep; list_del_init(&epi->fllink); |
d8805e633
|
717 |
mutex_lock_nested(&ep->mtx, 0); |
7699acd13
|
718 |
ep_remove(ep, epi); |
d47de16c7
|
719 |
mutex_unlock(&ep->mtx); |
b611967de
|
720 |
} |
7699acd13
|
721 |
mutex_unlock(&epmutex); |
b611967de
|
722 |
} |
53d2be79d
|
723 |
static int ep_alloc(struct eventpoll **pep) |
1da177e4c
|
724 |
{ |
7ef9964e6
|
725 726 727 |
int error; struct user_struct *user; struct eventpoll *ep; |
1da177e4c
|
728 |
|
7ef9964e6
|
729 |
user = get_current_user(); |
7ef9964e6
|
730 731 732 733 |
error = -ENOMEM; ep = kzalloc(sizeof(*ep), GFP_KERNEL); if (unlikely(!ep)) goto free_uid; |
1da177e4c
|
734 |
|
c7ea76302
|
735 |
spin_lock_init(&ep->lock); |
d47de16c7
|
736 |
mutex_init(&ep->mtx); |
1da177e4c
|
737 738 739 740 |
init_waitqueue_head(&ep->wq); init_waitqueue_head(&ep->poll_wait); INIT_LIST_HEAD(&ep->rdllist); ep->rbr = RB_ROOT; |
d47de16c7
|
741 |
ep->ovflist = EP_UNACTIVE_PTR; |
7ef9964e6
|
742 |
ep->user = user; |
1da177e4c
|
743 |
|
53d2be79d
|
744 |
*pep = ep; |
1da177e4c
|
745 |
|
1da177e4c
|
746 |
return 0; |
7ef9964e6
|
747 748 749 750 |
free_uid: free_uid(user); return error; |
1da177e4c
|
751 |
} |
1da177e4c
|
752 |
/* |
c7ea76302
|
753 754 755 |
* Search the file inside the eventpoll tree. The RB tree operations * are protected by the "mtx" mutex, and ep_find() must be called with * "mtx" held. |
1da177e4c
|
756 757 758 759 |
*/ static struct epitem *ep_find(struct eventpoll *ep, struct file *file, int fd) { int kcmp; |
1da177e4c
|
760 761 762 |
struct rb_node *rbp; struct epitem *epi, *epir = NULL; struct epoll_filefd ffd; |
b030a4dd6
|
763 |
ep_set_ffd(&ffd, file, fd); |
1da177e4c
|
764 765 |
for (rbp = ep->rbr.rb_node; rbp; ) { epi = rb_entry(rbp, struct epitem, rbn); |
b030a4dd6
|
766 |
kcmp = ep_cmp_ffd(&ffd, &epi->ffd); |
1da177e4c
|
767 768 769 770 771 |
if (kcmp > 0) rbp = rbp->rb_right; else if (kcmp < 0) rbp = rbp->rb_left; else { |
1da177e4c
|
772 773 774 775 |
epir = epi; break; } } |
1da177e4c
|
776 |
|
1da177e4c
|
777 778 |
return epir; } |
1da177e4c
|
779 |
/* |
7699acd13
|
780 |
* This is the callback that is passed to the wait queue wakeup |
bf6a41db7
|
781 |
* mechanism. It is called by the stored file descriptors when they |
7699acd13
|
782 |
* have events to report. |
1da177e4c
|
783 |
*/ |
7699acd13
|
784 |
static int ep_poll_callback(wait_queue_t *wait, unsigned mode, int sync, void *key) |
1da177e4c
|
785 |
{ |
7699acd13
|
786 787 788 789 |
int pwake = 0; unsigned long flags; struct epitem *epi = ep_item_from_wait(wait); struct eventpoll *ep = epi->ep; |
1da177e4c
|
790 |
|
c7ea76302
|
791 |
spin_lock_irqsave(&ep->lock, flags); |
1da177e4c
|
792 |
|
7699acd13
|
793 794 795 796 797 798 799 |
/* * If the event mask does not contain any poll(2) event, we consider the * descriptor to be disabled. This condition is likely the effect of the * EPOLLONESHOT bit that disables the descriptor when an event is received, * until the next EPOLL_CTL_MOD will be issued. */ if (!(epi->event.events & ~EP_PRIVATE_BITS)) |
d47de16c7
|
800 801 802 |
goto out_unlock; /* |
2dfa4eeab
|
803 804 805 806 807 808 809 810 811 |
* Check the events coming with the callback. At this stage, not * every device reports the events in the "key" parameter of the * callback. We need to be able to handle both cases here, hence the * test for "key" != NULL before the event match test. */ if (key && !((unsigned long) key & epi->event.events)) goto out_unlock; /* |
bf6a41db7
|
812 |
* If we are transferring events to userspace, we can hold no locks |
d47de16c7
|
813 |
* (because we're accessing user memory, and because of linux f_op->poll() |
bf6a41db7
|
814 |
* semantics). All the events that happen during that period of time are |
d47de16c7
|
815 816 817 818 819 820 821 822 823 |
* chained in ep->ovflist and requeued later on. */ if (unlikely(ep->ovflist != EP_UNACTIVE_PTR)) { if (epi->next == EP_UNACTIVE_PTR) { epi->next = ep->ovflist; ep->ovflist = epi; } goto out_unlock; } |
1da177e4c
|
824 |
|
7699acd13
|
825 |
/* If this file is already in the ready list we exit soon */ |
5071f97ec
|
826 827 |
if (!ep_is_linked(&epi->rdllink)) list_add_tail(&epi->rdllink, &ep->rdllist); |
7699acd13
|
828 |
|
7699acd13
|
829 830 831 832 833 |
/* * Wake up ( if active ) both the eventpoll wait list and the ->poll() * wait list. */ if (waitqueue_active(&ep->wq)) |
4a6e9e2ce
|
834 |
wake_up_locked(&ep->wq); |
7699acd13
|
835 836 |
if (waitqueue_active(&ep->poll_wait)) pwake++; |
d47de16c7
|
837 |
out_unlock: |
c7ea76302
|
838 |
spin_unlock_irqrestore(&ep->lock, flags); |
1da177e4c
|
839 |
|
7699acd13
|
840 841 |
/* We have to call this outside the lock */ if (pwake) |
5071f97ec
|
842 |
ep_poll_safewake(&ep->poll_wait); |
7699acd13
|
843 844 845 |
return 1; } |
1da177e4c
|
846 847 848 849 850 851 852 853 |
/* * This is the callback that is used to add our wait queue to the * target file wakeup lists. */ static void ep_ptable_queue_proc(struct file *file, wait_queue_head_t *whead, poll_table *pt) { |
b030a4dd6
|
854 |
struct epitem *epi = ep_item_from_epqueue(pt); |
1da177e4c
|
855 |
struct eppoll_entry *pwq; |
e94b17660
|
856 |
if (epi->nwait >= 0 && (pwq = kmem_cache_alloc(pwq_cache, GFP_KERNEL))) { |
1da177e4c
|
857 858 859 860 861 862 |
init_waitqueue_func_entry(&pwq->wait, ep_poll_callback); pwq->whead = whead; pwq->base = epi; add_wait_queue(whead, &pwq->wait); list_add_tail(&pwq->llink, &epi->pwqlist); epi->nwait++; |
296e236e9
|
863 |
} else { |
1da177e4c
|
864 865 |
/* We have to signal that an error occurred */ epi->nwait = -1; |
296e236e9
|
866 |
} |
1da177e4c
|
867 |
} |
1da177e4c
|
868 869 870 871 872 873 874 875 876 |
static void ep_rbtree_insert(struct eventpoll *ep, struct epitem *epi) { int kcmp; struct rb_node **p = &ep->rbr.rb_node, *parent = NULL; struct epitem *epic; while (*p) { parent = *p; epic = rb_entry(parent, struct epitem, rbn); |
b030a4dd6
|
877 |
kcmp = ep_cmp_ffd(&epi->ffd, &epic->ffd); |
1da177e4c
|
878 879 880 881 882 883 884 885 |
if (kcmp > 0) p = &parent->rb_right; else p = &parent->rb_left; } rb_link_node(&epi->rbn, parent, p); rb_insert_color(&epi->rbn, &ep->rbr); } |
c7ea76302
|
886 887 888 |
/* * Must be called with "mtx" held. */ |
1da177e4c
|
889 890 891 892 893 |
static int ep_insert(struct eventpoll *ep, struct epoll_event *event, struct file *tfile, int fd) { int error, revents, pwake = 0; unsigned long flags; |
52bd19f76
|
894 |
long user_watches; |
1da177e4c
|
895 896 |
struct epitem *epi; struct ep_pqueue epq; |
52bd19f76
|
897 898 |
user_watches = atomic_long_read(&ep->user->epoll_watches); if (unlikely(user_watches >= max_user_watches)) |
7ef9964e6
|
899 |
return -ENOSPC; |
e94b17660
|
900 |
if (!(epi = kmem_cache_alloc(epi_cache, GFP_KERNEL))) |
7ef9964e6
|
901 |
return -ENOMEM; |
1da177e4c
|
902 903 |
/* Item initialization follow here ... */ |
1da177e4c
|
904 905 |
INIT_LIST_HEAD(&epi->rdllink); INIT_LIST_HEAD(&epi->fllink); |
1da177e4c
|
906 907 |
INIT_LIST_HEAD(&epi->pwqlist); epi->ep = ep; |
b030a4dd6
|
908 |
ep_set_ffd(&epi->ffd, tfile, fd); |
1da177e4c
|
909 |
epi->event = *event; |
1da177e4c
|
910 |
epi->nwait = 0; |
d47de16c7
|
911 |
epi->next = EP_UNACTIVE_PTR; |
1da177e4c
|
912 913 914 915 916 917 918 919 |
/* Initialize the poll table using the queue callback */ epq.epi = epi; init_poll_funcptr(&epq.pt, ep_ptable_queue_proc); /* * Attach the item to the poll hooks and get current event bits. * We can safely use the file* here because its usage count has |
c7ea76302
|
920 921 922 |
* been increased by the caller of this function. Note that after * this operation completes, the poll callback can start hitting * the new item. |
1da177e4c
|
923 924 925 926 927 928 929 930 |
*/ revents = tfile->f_op->poll(tfile, &epq.pt); /* * We have to check if something went wrong during the poll wait queue * install process. Namely an allocation for a wait queue failed due * high memory pressure. */ |
7ef9964e6
|
931 |
error = -ENOMEM; |
1da177e4c
|
932 |
if (epi->nwait < 0) |
7699acd13
|
933 |
goto error_unregister; |
1da177e4c
|
934 935 |
/* Add the current item to the list of active epoll hook for this file */ |
684999149
|
936 |
spin_lock(&tfile->f_lock); |
1da177e4c
|
937 |
list_add_tail(&epi->fllink, &tfile->f_ep_links); |
684999149
|
938 |
spin_unlock(&tfile->f_lock); |
1da177e4c
|
939 |
|
c7ea76302
|
940 941 942 943 |
/* * Add the current item to the RB tree. All RB tree operations are * protected by "mtx", and ep_insert() is called with "mtx" held. */ |
1da177e4c
|
944 |
ep_rbtree_insert(ep, epi); |
c7ea76302
|
945 946 |
/* We have to drop the new item inside our item list to keep track of it */ spin_lock_irqsave(&ep->lock, flags); |
1da177e4c
|
947 |
/* If the file is already "ready" we drop it inside the ready list */ |
b030a4dd6
|
948 |
if ((revents & event->events) && !ep_is_linked(&epi->rdllink)) { |
1da177e4c
|
949 950 951 952 |
list_add_tail(&epi->rdllink, &ep->rdllist); /* Notify waiting tasks that events are available */ if (waitqueue_active(&ep->wq)) |
4a6e9e2ce
|
953 |
wake_up_locked(&ep->wq); |
1da177e4c
|
954 955 956 |
if (waitqueue_active(&ep->poll_wait)) pwake++; } |
c7ea76302
|
957 |
spin_unlock_irqrestore(&ep->lock, flags); |
1da177e4c
|
958 |
|
52bd19f76
|
959 |
atomic_long_inc(&ep->user->epoll_watches); |
7ef9964e6
|
960 |
|
1da177e4c
|
961 962 |
/* We have to call this outside the lock */ if (pwake) |
5071f97ec
|
963 |
ep_poll_safewake(&ep->poll_wait); |
1da177e4c
|
964 |
|
1da177e4c
|
965 |
return 0; |
7699acd13
|
966 |
error_unregister: |
1da177e4c
|
967 968 969 970 |
ep_unregister_pollwait(ep, epi); /* * We need to do this because an event could have been arrived on some |
67647d0fb
|
971 972 973 |
* allocated wait queue. Note that we don't care about the ep->ovflist * list, since that is used/cleaned only inside a section bound by "mtx". * And ep_insert() is called with "mtx" held. |
1da177e4c
|
974 |
*/ |
c7ea76302
|
975 |
spin_lock_irqsave(&ep->lock, flags); |
b030a4dd6
|
976 |
if (ep_is_linked(&epi->rdllink)) |
6192bd536
|
977 |
list_del_init(&epi->rdllink); |
c7ea76302
|
978 |
spin_unlock_irqrestore(&ep->lock, flags); |
1da177e4c
|
979 |
|
b030a4dd6
|
980 |
kmem_cache_free(epi_cache, epi); |
7ef9964e6
|
981 |
|
1da177e4c
|
982 983 |
return error; } |
1da177e4c
|
984 985 |
/* * Modify the interest event mask by dropping an event if the new mask |
c7ea76302
|
986 |
* has a match in the current file status. Must be called with "mtx" held. |
1da177e4c
|
987 988 989 990 991 |
*/ static int ep_modify(struct eventpoll *ep, struct epitem *epi, struct epoll_event *event) { int pwake = 0; unsigned int revents; |
1da177e4c
|
992 993 |
/* |
e057e15ff
|
994 995 996 |
* Set the new event interest mask before calling f_op->poll(); * otherwise we might miss an event that happens between the * f_op->poll() call and the new event set registering. |
1da177e4c
|
997 998 |
*/ epi->event.events = event->events; |
e057e15ff
|
999 |
epi->event.data = event->data; /* protected by mtx */ |
1da177e4c
|
1000 1001 1002 1003 1004 1005 |
/* * Get current event bits. We can safely use the file* here because * its usage count has been increased by the caller of this function. */ revents = epi->ffd.file->f_op->poll(epi->ffd.file, NULL); |
1da177e4c
|
1006 |
/* |
c7ea76302
|
1007 |
* If the item is "hot" and it is not registered inside the ready |
67647d0fb
|
1008 |
* list, push it inside. |
1da177e4c
|
1009 |
*/ |
c7ea76302
|
1010 |
if (revents & event->events) { |
e057e15ff
|
1011 |
spin_lock_irq(&ep->lock); |
c7ea76302
|
1012 1013 1014 1015 1016 |
if (!ep_is_linked(&epi->rdllink)) { list_add_tail(&epi->rdllink, &ep->rdllist); /* Notify waiting tasks that events are available */ if (waitqueue_active(&ep->wq)) |
4a6e9e2ce
|
1017 |
wake_up_locked(&ep->wq); |
c7ea76302
|
1018 1019 |
if (waitqueue_active(&ep->poll_wait)) pwake++; |
7699acd13
|
1020 |
} |
e057e15ff
|
1021 |
spin_unlock_irq(&ep->lock); |
7699acd13
|
1022 |
} |
1da177e4c
|
1023 |
|
7699acd13
|
1024 1025 |
/* We have to call this outside the lock */ if (pwake) |
5071f97ec
|
1026 |
ep_poll_safewake(&ep->poll_wait); |
1da177e4c
|
1027 |
|
7699acd13
|
1028 |
return 0; |
1da177e4c
|
1029 |
} |
296e236e9
|
1030 1031 |
static int ep_send_events_proc(struct eventpoll *ep, struct list_head *head, void *priv) |
1da177e4c
|
1032 |
{ |
5071f97ec
|
1033 1034 |
struct ep_send_events_data *esed = priv; int eventcnt; |
296e236e9
|
1035 |
unsigned int revents; |
5071f97ec
|
1036 1037 |
struct epitem *epi; struct epoll_event __user *uevent; |
1da177e4c
|
1038 |
|
296e236e9
|
1039 |
/* |
5071f97ec
|
1040 1041 1042 |
* We can loop without lock because we are passed a task private list. * Items cannot vanish during the loop because ep_scan_ready_list() is * holding "mtx" during this call. |
296e236e9
|
1043 |
*/ |
5071f97ec
|
1044 1045 1046 |
for (eventcnt = 0, uevent = esed->events; !list_empty(head) && eventcnt < esed->maxevents;) { epi = list_first_entry(head, struct epitem, rdllink); |
d47de16c7
|
1047 1048 |
list_del_init(&epi->rdllink); |
1da177e4c
|
1049 |
|
296e236e9
|
1050 1051 |
revents = epi->ffd.file->f_op->poll(epi->ffd.file, NULL) & epi->event.events; |
5071f97ec
|
1052 |
|
296e236e9
|
1053 |
/* |
5071f97ec
|
1054 1055 1056 1057 |
* If the event mask intersect the caller-requested one, * deliver the event to userspace. Again, ep_scan_ready_list() * is holding "mtx", so no operations coming from userspace * can change the item. |
296e236e9
|
1058 1059 |
*/ if (revents) { |
5071f97ec
|
1060 |
if (__put_user(revents, &uevent->events) || |
d03058828
|
1061 1062 |
__put_user(epi->event.data, &uevent->data)) { list_add(&epi->rdllink, head); |
296e236e9
|
1063 |
return eventcnt ? eventcnt : -EFAULT; |
d03058828
|
1064 |
} |
296e236e9
|
1065 |
eventcnt++; |
5071f97ec
|
1066 |
uevent++; |
296e236e9
|
1067 1068 1069 1070 1071 1072 1073 1074 |
if (epi->event.events & EPOLLONESHOT) epi->event.events &= EP_PRIVATE_BITS; else if (!(epi->event.events & EPOLLET)) { /* * If this file has been added with Level * Trigger mode, we need to insert back inside * the ready list, so that the next call to * epoll_wait() will check again the events |
25985edce
|
1075 |
* availability. At this point, no one can insert |
296e236e9
|
1076 1077 1078 1079 1080 1081 1082 1083 1084 |
* into ep->rdllist besides us. The epoll_ctl() * callers are locked out by * ep_scan_ready_list() holding "mtx" and the * poll callback will queue them in ep->ovflist. */ list_add_tail(&epi->rdllink, &ep->rdllist); } } } |
5071f97ec
|
1085 1086 1087 |
return eventcnt; } |
d47de16c7
|
1088 |
|
296e236e9
|
1089 1090 |
static int ep_send_events(struct eventpoll *ep, struct epoll_event __user *events, int maxevents) |
5071f97ec
|
1091 1092 |
{ struct ep_send_events_data esed; |
1da177e4c
|
1093 |
|
5071f97ec
|
1094 1095 |
esed.maxevents = maxevents; esed.events = events; |
6192bd536
|
1096 |
|
d8805e633
|
1097 |
return ep_scan_ready_list(ep, ep_send_events_proc, &esed, 0); |
1da177e4c
|
1098 |
} |
0781b909b
|
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 |
static inline struct timespec ep_set_mstimeout(long ms) { struct timespec now, ts = { .tv_sec = ms / MSEC_PER_SEC, .tv_nsec = NSEC_PER_MSEC * (ms % MSEC_PER_SEC), }; ktime_get_ts(&now); return timespec_add_safe(now, ts); } |
f4d93ad74
|
1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 |
/** * ep_poll - Retrieves ready events, and delivers them to the caller supplied * event buffer. * * @ep: Pointer to the eventpoll context. * @events: Pointer to the userspace buffer where the ready events should be * stored. * @maxevents: Size (in terms of number of events) of the caller event buffer. * @timeout: Maximum timeout for the ready events fetch operation, in * milliseconds. If the @timeout is zero, the function will not block, * while if the @timeout is less than zero, the function will block * until at least one event has been retrieved (or an error * occurred). * * Returns: Returns the number of ready events which have been fetched, or an * error code, in case of error. */ |
1da177e4c
|
1126 1127 1128 |
static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events, int maxevents, long timeout) { |
f4d93ad74
|
1129 |
int res = 0, eavail, timed_out = 0; |
1da177e4c
|
1130 |
unsigned long flags; |
f4d93ad74
|
1131 |
long slack = 0; |
1da177e4c
|
1132 |
wait_queue_t wait; |
95aac7b1c
|
1133 1134 1135 |
ktime_t expires, *to = NULL; if (timeout > 0) { |
0781b909b
|
1136 |
struct timespec end_time = ep_set_mstimeout(timeout); |
95aac7b1c
|
1137 1138 1139 1140 |
slack = select_estimate_accuracy(&end_time); to = &expires; *to = timespec_to_ktime(end_time); } else if (timeout == 0) { |
f4d93ad74
|
1141 1142 1143 1144 |
/* * Avoid the unnecessary trip to the wait queue loop, if the * caller specified a non blocking operation. */ |
95aac7b1c
|
1145 |
timed_out = 1; |
f4d93ad74
|
1146 1147 |
spin_lock_irqsave(&ep->lock, flags); goto check_events; |
95aac7b1c
|
1148 |
} |
1da177e4c
|
1149 |
|
f4d93ad74
|
1150 |
fetch_events: |
c7ea76302
|
1151 |
spin_lock_irqsave(&ep->lock, flags); |
1da177e4c
|
1152 |
|
3fb0e584a
|
1153 |
if (!ep_events_available(ep)) { |
1da177e4c
|
1154 1155 1156 1157 1158 1159 |
/* * We don't have any available event to return to the caller. * We need to sleep here, and we will be wake up by * ep_poll_callback() when events will become available. */ init_waitqueue_entry(&wait, current); |
a93d2f174
|
1160 |
__add_wait_queue_exclusive(&ep->wq, &wait); |
1da177e4c
|
1161 1162 1163 1164 1165 1166 1167 1168 |
for (;;) { /* * We don't want to sleep if the ep_poll_callback() sends us * a wakeup in between. That's why we set the task state * to TASK_INTERRUPTIBLE before doing the checks. */ set_current_state(TASK_INTERRUPTIBLE); |
3fb0e584a
|
1169 |
if (ep_events_available(ep) || timed_out) |
1da177e4c
|
1170 1171 1172 1173 1174 |
break; if (signal_pending(current)) { res = -EINTR; break; } |
c7ea76302
|
1175 |
spin_unlock_irqrestore(&ep->lock, flags); |
95aac7b1c
|
1176 1177 |
if (!schedule_hrtimeout_range(to, slack, HRTIMER_MODE_ABS)) timed_out = 1; |
c7ea76302
|
1178 |
spin_lock_irqsave(&ep->lock, flags); |
1da177e4c
|
1179 |
} |
3419b23a9
|
1180 |
__remove_wait_queue(&ep->wq, &wait); |
1da177e4c
|
1181 1182 1183 |
set_current_state(TASK_RUNNING); } |
f4d93ad74
|
1184 |
check_events: |
1da177e4c
|
1185 |
/* Is it worth to try to dig for events ? */ |
3fb0e584a
|
1186 |
eavail = ep_events_available(ep); |
1da177e4c
|
1187 |
|
c7ea76302
|
1188 |
spin_unlock_irqrestore(&ep->lock, flags); |
1da177e4c
|
1189 1190 1191 1192 1193 1194 1195 |
/* * Try to transfer events to user space. In case we get 0 events and * there's still timeout left over, we go trying again in search of * more luck. */ if (!res && eavail && |
95aac7b1c
|
1196 |
!(res = ep_send_events(ep, events, maxevents)) && !timed_out) |
f4d93ad74
|
1197 |
goto fetch_events; |
1da177e4c
|
1198 1199 1200 |
return res; } |
22bacca48
|
1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 |
/** * ep_loop_check_proc - Callback function to be passed to the @ep_call_nested() * API, to verify that adding an epoll file inside another * epoll structure, does not violate the constraints, in * terms of closed loops, or too deep chains (which can * result in excessive stack usage). * * @priv: Pointer to the epoll file to be currently checked. * @cookie: Original cookie for this call. This is the top-of-the-chain epoll * data structure pointer. * @call_nests: Current dept of the @ep_call_nested() call stack. * * Returns: Returns zero if adding the epoll @file inside current epoll * structure @ep does not violate the constraints, or -1 otherwise. */ static int ep_loop_check_proc(void *priv, void *cookie, int call_nests) { int error = 0; struct file *file = priv; struct eventpoll *ep = file->private_data; struct rb_node *rbp; struct epitem *epi; |
d8805e633
|
1223 |
mutex_lock_nested(&ep->mtx, call_nests + 1); |
22bacca48
|
1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 |
for (rbp = rb_first(&ep->rbr); rbp; rbp = rb_next(rbp)) { epi = rb_entry(rbp, struct epitem, rbn); if (unlikely(is_file_epoll(epi->ffd.file))) { error = ep_call_nested(&poll_loop_ncalls, EP_MAX_NESTS, ep_loop_check_proc, epi->ffd.file, epi->ffd.file->private_data, current); if (error != 0) break; } } mutex_unlock(&ep->mtx); return error; } /** * ep_loop_check - Performs a check to verify that adding an epoll file (@file) * another epoll file (represented by @ep) does not create * closed loops or too deep chains. * * @ep: Pointer to the epoll private data structure. * @file: Pointer to the epoll file to be checked. * * Returns: Returns zero if adding the epoll @file inside current epoll * structure @ep does not violate the constraints, or -1 otherwise. */ static int ep_loop_check(struct eventpoll *ep, struct file *file) { return ep_call_nested(&poll_loop_ncalls, EP_MAX_NESTS, ep_loop_check_proc, file, ep, current); } |
7699acd13
|
1255 |
/* |
523723bb5
|
1256 |
* Open an eventpoll file descriptor. |
7699acd13
|
1257 |
*/ |
5a8a82b1d
|
1258 |
SYSCALL_DEFINE1(epoll_create1, int, flags) |
7699acd13
|
1259 |
{ |
bb57c3edc
|
1260 1261 |
int error; struct eventpoll *ep = NULL; |
7699acd13
|
1262 |
|
e38b36f32
|
1263 1264 |
/* Check the EPOLL_* constant for consistency. */ BUILD_BUG_ON(EPOLL_CLOEXEC != O_CLOEXEC); |
296e236e9
|
1265 1266 |
if (flags & ~EPOLL_CLOEXEC) return -EINVAL; |
7699acd13
|
1267 |
/* |
bb57c3edc
|
1268 |
* Create the internal data structure ("struct eventpoll"). |
7699acd13
|
1269 |
*/ |
9fe5ad9c8
|
1270 |
error = ep_alloc(&ep); |
bb57c3edc
|
1271 1272 |
if (error < 0) return error; |
7699acd13
|
1273 1274 |
/* * Creates all the items needed to setup an eventpoll file. That is, |
2030a42ce
|
1275 |
* a file structure and a free file descriptor. |
7699acd13
|
1276 |
*/ |
bb57c3edc
|
1277 |
error = anon_inode_getfd("[eventpoll]", &eventpoll_fops, ep, |
628ff7c1d
|
1278 |
O_RDWR | (flags & O_CLOEXEC)); |
bb57c3edc
|
1279 |
if (error < 0) |
2030a42ce
|
1280 |
ep_free(ep); |
7699acd13
|
1281 |
|
bb57c3edc
|
1282 |
return error; |
7699acd13
|
1283 |
} |
5a8a82b1d
|
1284 |
SYSCALL_DEFINE1(epoll_create, int, size) |
a0998b50c
|
1285 |
{ |
bfe3891a5
|
1286 |
if (size <= 0) |
9fe5ad9c8
|
1287 1288 1289 |
return -EINVAL; return sys_epoll_create1(0); |
a0998b50c
|
1290 |
} |
7699acd13
|
1291 1292 1293 |
/* * The following function implements the controller interface for * the eventpoll file that enables the insertion/removal/change of |
67647d0fb
|
1294 |
* file descriptors inside the interest set. |
7699acd13
|
1295 |
*/ |
5a8a82b1d
|
1296 1297 |
SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd, struct epoll_event __user *, event) |
7699acd13
|
1298 1299 |
{ int error; |
22bacca48
|
1300 |
int did_lock_epmutex = 0; |
7699acd13
|
1301 1302 1303 1304 |
struct file *file, *tfile; struct eventpoll *ep; struct epitem *epi; struct epoll_event epds; |
7699acd13
|
1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 |
error = -EFAULT; if (ep_op_has_event(op) && copy_from_user(&epds, event, sizeof(struct epoll_event))) goto error_return; /* Get the "struct file *" for the eventpoll file */ error = -EBADF; file = fget(epfd); if (!file) goto error_return; /* Get the "struct file *" for the target file */ tfile = fget(fd); if (!tfile) goto error_fput; /* The target file descriptor must support poll */ error = -EPERM; if (!tfile->f_op || !tfile->f_op->poll) goto error_tgt_fput; /* * We have to check that the file structure underneath the file descriptor * the user passed to us _is_ an eventpoll file. And also we do not permit * adding an epoll file descriptor inside itself. */ error = -EINVAL; if (file == tfile || !is_file_epoll(file)) goto error_tgt_fput; /* * At this point it is safe to assume that the "private_data" contains * our own data structure. */ ep = file->private_data; |
22bacca48
|
1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 |
/* * When we insert an epoll file descriptor, inside another epoll file * descriptor, there is the change of creating closed loops, which are * better be handled here, than in more critical paths. * * We hold epmutex across the loop check and the insert in this case, in * order to prevent two separate inserts from racing and each doing the * insert "at the same time" such that ep_loop_check passes on both * before either one does the insert, thereby creating a cycle. */ if (unlikely(is_file_epoll(tfile) && op == EPOLL_CTL_ADD)) { mutex_lock(&epmutex); did_lock_epmutex = 1; error = -ELOOP; if (ep_loop_check(ep, tfile) != 0) goto error_tgt_fput; } |
d8805e633
|
1357 |
mutex_lock_nested(&ep->mtx, 0); |
7699acd13
|
1358 |
|
67647d0fb
|
1359 1360 1361 1362 1363 |
/* * Try to lookup the file inside our RB tree, Since we grabbed "mtx" * above, we can be sure to be able to use the item looked up by * ep_find() till we release the mutex. */ |
7699acd13
|
1364 1365 1366 1367 1368 1369 1370 |
epi = ep_find(ep, tfile, fd); error = -EINVAL; switch (op) { case EPOLL_CTL_ADD: if (!epi) { epds.events |= POLLERR | POLLHUP; |
7699acd13
|
1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 |
error = ep_insert(ep, &epds, tfile, fd); } else error = -EEXIST; break; case EPOLL_CTL_DEL: if (epi) error = ep_remove(ep, epi); else error = -ENOENT; break; case EPOLL_CTL_MOD: if (epi) { epds.events |= POLLERR | POLLHUP; error = ep_modify(ep, epi, &epds); } else error = -ENOENT; break; } |
d47de16c7
|
1389 |
mutex_unlock(&ep->mtx); |
7699acd13
|
1390 1391 |
error_tgt_fput: |
22bacca48
|
1392 1393 |
if (unlikely(did_lock_epmutex)) mutex_unlock(&epmutex); |
7699acd13
|
1394 1395 1396 1397 |
fput(tfile); error_fput: fput(file); error_return: |
7699acd13
|
1398 1399 1400 1401 1402 1403 1404 1405 |
return error; } /* * Implement the event wait interface for the eventpoll file. It is the kernel * part of the user space epoll_wait(2). */ |
5a8a82b1d
|
1406 1407 |
SYSCALL_DEFINE4(epoll_wait, int, epfd, struct epoll_event __user *, events, int, maxevents, int, timeout) |
7699acd13
|
1408 1409 1410 1411 |
{ int error; struct file *file; struct eventpoll *ep; |
7699acd13
|
1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 |
/* The maximum number of event must be greater than zero */ if (maxevents <= 0 || maxevents > EP_MAX_EVENTS) return -EINVAL; /* Verify that the area passed by the user is writeable */ if (!access_ok(VERIFY_WRITE, events, maxevents * sizeof(struct epoll_event))) { error = -EFAULT; goto error_return; } /* Get the "struct file *" for the eventpoll file */ error = -EBADF; file = fget(epfd); if (!file) goto error_return; /* * We have to check that the file structure underneath the fd * the user passed to us _is_ an eventpoll file. */ error = -EINVAL; if (!is_file_epoll(file)) goto error_fput; /* * At this point it is safe to assume that the "private_data" contains * our own data structure. */ ep = file->private_data; /* Time to fish for events ... */ error = ep_poll(ep, events, maxevents, timeout); error_fput: fput(file); error_return: |
7699acd13
|
1448 1449 1450 |
return error; } |
f3de272b8
|
1451 |
#ifdef HAVE_SET_RESTORE_SIGMASK |
7699acd13
|
1452 1453 1454 1455 1456 |
/* * Implement the event wait interface for the eventpoll file. It is the kernel * part of the user space epoll_pwait(2). */ |
5a8a82b1d
|
1457 1458 1459 |
SYSCALL_DEFINE6(epoll_pwait, int, epfd, struct epoll_event __user *, events, int, maxevents, int, timeout, const sigset_t __user *, sigmask, size_t, sigsetsize) |
7699acd13
|
1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 |
{ int error; sigset_t ksigmask, sigsaved; /* * If the caller wants a certain signal mask to be set during the wait, * we apply it here. */ if (sigmask) { if (sigsetsize != sizeof(sigset_t)) return -EINVAL; if (copy_from_user(&ksigmask, sigmask, sizeof(ksigmask))) return -EFAULT; sigdelsetmask(&ksigmask, sigmask(SIGKILL) | sigmask(SIGSTOP)); sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved); } error = sys_epoll_wait(epfd, events, maxevents, timeout); /* * If we changed the signal mask, we need to restore the original one. * In case we've got a signal while waiting, we do not restore the * signal mask yet, and we allow do_signal() to deliver the signal on * the way back to userspace, before the signal mask is restored. */ if (sigmask) { if (error == -EINTR) { memcpy(¤t->saved_sigmask, &sigsaved, |
c7ea76302
|
1488 |
sizeof(sigsaved)); |
4e4c22c71
|
1489 |
set_restore_sigmask(); |
7699acd13
|
1490 1491 1492 1493 1494 1495 |
} else sigprocmask(SIG_SETMASK, &sigsaved, NULL); } return error; } |
f3de272b8
|
1496 |
#endif /* HAVE_SET_RESTORE_SIGMASK */ |
7699acd13
|
1497 |
|
1da177e4c
|
1498 1499 |
static int __init eventpoll_init(void) { |
7ef9964e6
|
1500 1501 1502 |
struct sysinfo si; si_meminfo(&si); |
9df04e1f2
|
1503 1504 1505 1506 |
/* * Allows top 4% of lomem to be allocated for epoll watches (per user). */ max_user_watches = (((si.totalram - si.totalhigh) / 25) << PAGE_SHIFT) / |
7ef9964e6
|
1507 |
EP_ITEM_COST; |
52bd19f76
|
1508 |
BUG_ON(max_user_watches < 0); |
1da177e4c
|
1509 |
|
22bacca48
|
1510 1511 1512 1513 1514 |
/* * Initialize the structure used to perform epoll file descriptor * inclusion loops checks. */ ep_nested_calls_init(&poll_loop_ncalls); |
1da177e4c
|
1515 |
/* Initialize the structure used to perform safe poll wait head wake ups */ |
5071f97ec
|
1516 1517 1518 1519 |
ep_nested_calls_init(&poll_safewake_ncalls); /* Initialize the structure used to perform file's f_op->poll() calls */ ep_nested_calls_init(&poll_readywalk_ncalls); |
1da177e4c
|
1520 1521 1522 |
/* Allocates slab cache used to allocate "struct epitem" items */ epi_cache = kmem_cache_create("eventpoll_epi", sizeof(struct epitem), |
bb57c3edc
|
1523 |
0, SLAB_HWCACHE_ALIGN | SLAB_PANIC, NULL); |
1da177e4c
|
1524 1525 1526 |
/* Allocates slab cache used to allocate "struct eppoll_entry" */ pwq_cache = kmem_cache_create("eventpoll_pwq", |
bb57c3edc
|
1527 |
sizeof(struct eppoll_entry), 0, SLAB_PANIC, NULL); |
1da177e4c
|
1528 |
|
1da177e4c
|
1529 |
return 0; |
1da177e4c
|
1530 |
} |
cea692418
|
1531 |
fs_initcall(eventpoll_init); |