Blame view
fs/select.c
25.4 KB
1da177e4c
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 |
/* * This file contains the procedures for the handling of select and poll * * Created for Linux based loosely upon Mathius Lattner's minix * patches by Peter MacDonald. Heavily edited by Linus. * * 4 February 1994 * COFF/ELF binary emulation. If the process has the STICKY_TIMEOUTS * flag set in its personality we do *not* modify the given timeout * parameter to reflect time remaining. * * 24 January 2000 * Changed sys_poll()/do_poll() to use PAGE_SIZE chunk-based allocation * of fds to overcome nfds < 16390 descriptors limit (Tigran Aivazian). */ |
022a16924
|
16 |
#include <linux/kernel.h> |
a99bbaf5e
|
17 |
#include <linux/sched.h> |
1da177e4c
|
18 |
#include <linux/syscalls.h> |
630d9c472
|
19 |
#include <linux/export.h> |
1da177e4c
|
20 |
#include <linux/slab.h> |
1da177e4c
|
21 22 23 |
#include <linux/poll.h> #include <linux/personality.h> /* for STICKY_TIMEOUTS */ #include <linux/file.h> |
9f3acc314
|
24 |
#include <linux/fdtable.h> |
1da177e4c
|
25 |
#include <linux/fs.h> |
b835996f6
|
26 |
#include <linux/rcupdate.h> |
8ff3e8e85
|
27 |
#include <linux/hrtimer.h> |
8bd75c77b
|
28 |
#include <linux/sched/rt.h> |
9745cdb36
|
29 |
#include <linux/freezer.h> |
076bb0c82
|
30 |
#include <net/busy_poll.h> |
1da177e4c
|
31 32 |
#include <asm/uaccess.h> |
90d6e24a3
|
33 34 35 36 37 38 39 40 41 42 43 44 |
/* * Estimate expected accuracy in ns from a timeval. * * After quite a bit of churning around, we've settled on * a simple thing of taking 0.1% of the timeout as the * slack, with a cap of 100 msec. * "nice" tasks get a 0.5% slack instead. * * Consider this comment an open invitation to come up with even * better solutions.. */ |
5ae87e79e
|
45 |
#define MAX_SLACK (100 * NSEC_PER_MSEC) |
96d2ab484
|
46 |
static long __estimate_accuracy(struct timespec *tv) |
90d6e24a3
|
47 |
{ |
96d2ab484
|
48 |
long slack; |
90d6e24a3
|
49 |
int divfactor = 1000; |
5ae87e79e
|
50 51 |
if (tv->tv_sec < 0) return 0; |
4ce105d30
|
52 |
if (task_nice(current) > 0) |
90d6e24a3
|
53 |
divfactor = divfactor / 5; |
5ae87e79e
|
54 55 |
if (tv->tv_sec > MAX_SLACK / (NSEC_PER_SEC/divfactor)) return MAX_SLACK; |
90d6e24a3
|
56 57 |
slack = tv->tv_nsec / divfactor; slack += tv->tv_sec * (NSEC_PER_SEC/divfactor); |
5ae87e79e
|
58 59 |
if (slack > MAX_SLACK) return MAX_SLACK; |
96d2ab484
|
60 |
|
90d6e24a3
|
61 62 |
return slack; } |
95aac7b1c
|
63 |
long select_estimate_accuracy(struct timespec *tv) |
90d6e24a3
|
64 65 66 67 68 69 70 |
{ unsigned long ret; struct timespec now; /* * Realtime tasks get a slack of 0 for obvious reasons. */ |
4ce105d30
|
71 |
if (rt_task(current)) |
90d6e24a3
|
72 73 74 75 76 77 78 79 80 |
return 0; ktime_get_ts(&now); now = timespec_sub(*tv, now); ret = __estimate_accuracy(&now); if (ret < current->timer_slack_ns) return current->timer_slack_ns; return ret; } |
1da177e4c
|
81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 |
struct poll_table_page { struct poll_table_page * next; struct poll_table_entry * entry; struct poll_table_entry entries[0]; }; #define POLL_TABLE_FULL(table) \ ((unsigned long)((table)->entry+1) > PAGE_SIZE + (unsigned long)(table)) /* * Ok, Peter made a complicated, but straightforward multiple_wait() function. * I have rewritten this, taking some shortcuts: This code may not be easy to * follow, but it should be free of race-conditions, and it's practical. If you * understand what I'm doing here, then you understand how the linux * sleep/wakeup mechanism works. * * Two very simple procedures, poll_wait() and poll_freewait() make all the * work. poll_wait() is an inline-function defined in <linux/poll.h>, * as all select/poll functions have to call it to add an entry to the * poll table. */ |
75c96f858
|
102 103 |
static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p); |
1da177e4c
|
104 105 106 107 |
void poll_initwait(struct poll_wqueues *pwq) { init_poll_funcptr(&pwq->pt, __pollwait); |
5f820f648
|
108 |
pwq->polling_task = current; |
b2add73db
|
109 |
pwq->triggered = 0; |
1da177e4c
|
110 111 |
pwq->error = 0; pwq->table = NULL; |
70674f95c
|
112 |
pwq->inline_index = 0; |
1da177e4c
|
113 |
} |
1da177e4c
|
114 |
EXPORT_SYMBOL(poll_initwait); |
70674f95c
|
115 116 |
static void free_poll_entry(struct poll_table_entry *entry) { |
ccf6780dc
|
117 |
remove_wait_queue(entry->wait_address, &entry->wait); |
70674f95c
|
118 119 |
fput(entry->filp); } |
1da177e4c
|
120 121 122 |
void poll_freewait(struct poll_wqueues *pwq) { struct poll_table_page * p = pwq->table; |
70674f95c
|
123 124 125 |
int i; for (i = 0; i < pwq->inline_index; i++) free_poll_entry(pwq->inline_entries + i); |
1da177e4c
|
126 127 128 129 130 131 132 |
while (p) { struct poll_table_entry * entry; struct poll_table_page *old; entry = p->entry; do { entry--; |
70674f95c
|
133 |
free_poll_entry(entry); |
1da177e4c
|
134 135 136 137 138 139 |
} while (entry > p->entries); old = p; p = p->next; free_page((unsigned long) old); } } |
1da177e4c
|
140 |
EXPORT_SYMBOL(poll_freewait); |
5f820f648
|
141 |
static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p) |
1da177e4c
|
142 |
{ |
1da177e4c
|
143 |
struct poll_table_page *table = p->table; |
70674f95c
|
144 145 |
if (p->inline_index < N_INLINE_POLL_ENTRIES) return p->inline_entries + p->inline_index++; |
1da177e4c
|
146 147 148 149 150 151 |
if (!table || POLL_TABLE_FULL(table)) { struct poll_table_page *new_table; new_table = (struct poll_table_page *) __get_free_page(GFP_KERNEL); if (!new_table) { p->error = -ENOMEM; |
70674f95c
|
152 |
return NULL; |
1da177e4c
|
153 154 155 156 157 158 |
} new_table->entry = new_table->entries; new_table->next = table; p->table = new_table; table = new_table; } |
70674f95c
|
159 160 |
return table->entry++; } |
4938d7e02
|
161 |
static int __pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) |
5f820f648
|
162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 |
{ struct poll_wqueues *pwq = wait->private; DECLARE_WAITQUEUE(dummy_wait, pwq->polling_task); /* * Although this function is called under waitqueue lock, LOCK * doesn't imply write barrier and the users expect write * barrier semantics on wakeup functions. The following * smp_wmb() is equivalent to smp_wmb() in try_to_wake_up() * and is paired with set_mb() in poll_schedule_timeout. */ smp_wmb(); pwq->triggered = 1; /* * Perform the default wake up operation using a dummy * waitqueue. * * TODO: This is hacky but there currently is no interface to * pass in @sync. @sync is scheduled to be removed and once * that happens, wake_up_process() can be used directly. */ return default_wake_function(&dummy_wait, mode, sync, key); } |
4938d7e02
|
186 187 188 189 190 191 192 193 194 |
static int pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) { struct poll_table_entry *entry; entry = container_of(wait, struct poll_table_entry, wait); if (key && !((unsigned long)key & entry->key)) return 0; return __pollwake(wait, mode, sync, key); } |
70674f95c
|
195 196 197 198 |
/* Add a new entry */ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) { |
5f820f648
|
199 200 |
struct poll_wqueues *pwq = container_of(p, struct poll_wqueues, pt); struct poll_table_entry *entry = poll_get_entry(pwq); |
70674f95c
|
201 202 |
if (!entry) return; |
cb0942b81
|
203 |
entry->filp = get_file(filp); |
70674f95c
|
204 |
entry->wait_address = wait_address; |
626cf2366
|
205 |
entry->key = p->_key; |
5f820f648
|
206 207 |
init_waitqueue_func_entry(&entry->wait, pollwake); entry->wait.private = pwq; |
ccf6780dc
|
208 |
add_wait_queue(wait_address, &entry->wait); |
1da177e4c
|
209 |
} |
5f820f648
|
210 211 212 213 214 215 216 |
int poll_schedule_timeout(struct poll_wqueues *pwq, int state, ktime_t *expires, unsigned long slack) { int rc = -EINTR; set_current_state(state); if (!pwq->triggered) |
59612d187
|
217 |
rc = schedule_hrtimeout_range(expires, slack, HRTIMER_MODE_ABS); |
5f820f648
|
218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 |
__set_current_state(TASK_RUNNING); /* * Prepare for the next iteration. * * The following set_mb() serves two purposes. First, it's * the counterpart rmb of the wmb in pollwake() such that data * written before wake up is always visible after wake up. * Second, the full barrier guarantees that triggered clearing * doesn't pass event check of the next iteration. Note that * this problem doesn't exist for the first iteration as * add_wait_queue() has full barrier semantics. */ set_mb(pwq->triggered, 0); return rc; } EXPORT_SYMBOL(poll_schedule_timeout); |
b773ad40a
|
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 |
/** * poll_select_set_timeout - helper function to setup the timeout value * @to: pointer to timespec variable for the final timeout * @sec: seconds (from user space) * @nsec: nanoseconds (from user space) * * Note, we do not use a timespec for the user space value here, That * way we can use the function for timeval and compat interfaces as well. * * Returns -EINVAL if sec/nsec are not normalized. Otherwise 0. */ int poll_select_set_timeout(struct timespec *to, long sec, long nsec) { struct timespec ts = {.tv_sec = sec, .tv_nsec = nsec}; if (!timespec_valid(&ts)) return -EINVAL; /* Optimize for the zero timeout value here */ if (!sec && !nsec) { to->tv_sec = to->tv_nsec = 0; } else { ktime_get_ts(to); *to = timespec_add_safe(*to, ts); } return 0; } static int poll_select_copy_remaining(struct timespec *end_time, void __user *p, int timeval, int ret) { struct timespec rts; struct timeval rtv; if (!p) return ret; if (current->personality & STICKY_TIMEOUTS) goto sticky; /* No update for zero timeout */ if (!end_time->tv_sec && !end_time->tv_nsec) return ret; ktime_get_ts(&rts); rts = timespec_sub(*end_time, rts); if (rts.tv_sec < 0) rts.tv_sec = rts.tv_nsec = 0; if (timeval) { |
65329bf46
|
286 287 |
if (sizeof(rtv) > sizeof(rtv.tv_sec) + sizeof(rtv.tv_usec)) memset(&rtv, 0, sizeof(rtv)); |
b773ad40a
|
288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 |
rtv.tv_sec = rts.tv_sec; rtv.tv_usec = rts.tv_nsec / NSEC_PER_USEC; if (!copy_to_user(p, &rtv, sizeof(rtv))) return ret; } else if (!copy_to_user(p, &rts, sizeof(rts))) return ret; /* * If an application puts its timeval in read-only memory, we * don't want the Linux-specific update to the timeval to * cause a fault after the select has completed * successfully. However, because we're not updating the * timeval, we can't restart the system call. */ sticky: if (ret == -ERESTARTNOHAND) ret = -EINTR; return ret; } |
1da177e4c
|
310 311 312 313 314 315 316 317 318 319 320 |
#define FDS_IN(fds, n) (fds->in + n) #define FDS_OUT(fds, n) (fds->out + n) #define FDS_EX(fds, n) (fds->ex + n) #define BITS(fds, n) (*FDS_IN(fds, n)|*FDS_OUT(fds, n)|*FDS_EX(fds, n)) static int max_select_fd(unsigned long n, fd_set_bits *fds) { unsigned long *open_fds; unsigned long set; int max; |
badf16621
|
321 |
struct fdtable *fdt; |
1da177e4c
|
322 323 |
/* handle last in-complete long-word first */ |
8ded2bbc1
|
324 325 |
set = ~(~0UL << (n & (BITS_PER_LONG-1))); n /= BITS_PER_LONG; |
badf16621
|
326 |
fdt = files_fdtable(current->files); |
1fd36adcd
|
327 |
open_fds = fdt->open_fds + n; |
1da177e4c
|
328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 |
max = 0; if (set) { set &= BITS(fds, n); if (set) { if (!(set & ~*open_fds)) goto get_max; return -EBADF; } } while (n) { open_fds--; n--; set = BITS(fds, n); if (!set) continue; if (set & ~*open_fds) return -EBADF; if (max) continue; get_max: do { max++; set >>= 1; } while (set); |
8ded2bbc1
|
352 |
max += n * BITS_PER_LONG; |
1da177e4c
|
353 354 355 356 |
} return max; } |
1da177e4c
|
357 358 359 |
#define POLLIN_SET (POLLRDNORM | POLLRDBAND | POLLIN | POLLHUP | POLLERR) #define POLLOUT_SET (POLLWRBAND | POLLWRNORM | POLLOUT | POLLERR) #define POLLEX_SET (POLLPRI) |
4938d7e02
|
360 |
static inline void wait_key_set(poll_table *wait, unsigned long in, |
2d48d67fa
|
361 362 |
unsigned long out, unsigned long bit, unsigned int ll_flag) |
4938d7e02
|
363 |
{ |
2d48d67fa
|
364 |
wait->_key = POLLEX_SET | ll_flag; |
626cf2366
|
365 366 367 368 |
if (in & bit) wait->_key |= POLLIN_SET; if (out & bit) wait->_key |= POLLOUT_SET; |
4938d7e02
|
369 |
} |
8ff3e8e85
|
370 |
int do_select(int n, fd_set_bits *fds, struct timespec *end_time) |
1da177e4c
|
371 |
{ |
8ff3e8e85
|
372 |
ktime_t expire, *to = NULL; |
1da177e4c
|
373 374 |
struct poll_wqueues table; poll_table *wait; |
8ff3e8e85
|
375 |
int retval, i, timed_out = 0; |
90d6e24a3
|
376 |
unsigned long slack = 0; |
cbf55001b
|
377 |
unsigned int busy_flag = net_busy_loop_on() ? POLL_BUSY_LOOP : 0; |
76b1e9b98
|
378 |
unsigned long busy_end = 0; |
1da177e4c
|
379 |
|
b835996f6
|
380 |
rcu_read_lock(); |
1da177e4c
|
381 |
retval = max_select_fd(n, fds); |
b835996f6
|
382 |
rcu_read_unlock(); |
1da177e4c
|
383 384 385 386 387 388 389 |
if (retval < 0) return retval; n = retval; poll_initwait(&table); wait = &table.pt; |
8ff3e8e85
|
390 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
626cf2366
|
391 |
wait->_qproc = NULL; |
8ff3e8e85
|
392 393 |
timed_out = 1; } |
96d2ab484
|
394 |
if (end_time && !timed_out) |
231f3d393
|
395 |
slack = select_estimate_accuracy(end_time); |
90d6e24a3
|
396 |
|
1da177e4c
|
397 398 399 |
retval = 0; for (;;) { unsigned long *rinp, *routp, *rexp, *inp, *outp, *exp; |
cbf55001b
|
400 |
bool can_busy_loop = false; |
1da177e4c
|
401 |
|
1da177e4c
|
402 403 404 405 406 407 |
inp = fds->in; outp = fds->out; exp = fds->ex; rinp = fds->res_in; routp = fds->res_out; rexp = fds->res_ex; for (i = 0; i < n; ++rinp, ++routp, ++rexp) { unsigned long in, out, ex, all_bits, bit = 1, mask, j; unsigned long res_in = 0, res_out = 0, res_ex = 0; |
1da177e4c
|
408 409 410 411 |
in = *inp++; out = *outp++; ex = *exp++; all_bits = in | out | ex; if (all_bits == 0) { |
8ded2bbc1
|
412 |
i += BITS_PER_LONG; |
1da177e4c
|
413 414 |
continue; } |
8ded2bbc1
|
415 |
for (j = 0; j < BITS_PER_LONG; ++j, ++i, bit <<= 1) { |
2903ff019
|
416 |
struct fd f; |
1da177e4c
|
417 418 419 420 |
if (i >= n) break; if (!(bit & all_bits)) continue; |
2903ff019
|
421 422 423 424 |
f = fdget(i); if (f.file) { const struct file_operations *f_op; f_op = f.file->f_op; |
1da177e4c
|
425 |
mask = DEFAULT_POLLMASK; |
72c2d5319
|
426 |
if (f_op->poll) { |
2d48d67fa
|
427 |
wait_key_set(wait, in, out, |
cbf55001b
|
428 |
bit, busy_flag); |
2903ff019
|
429 |
mask = (*f_op->poll)(f.file, wait); |
4938d7e02
|
430 |
} |
2903ff019
|
431 |
fdput(f); |
1da177e4c
|
432 433 434 |
if ((mask & POLLIN_SET) && (in & bit)) { res_in |= bit; retval++; |
626cf2366
|
435 |
wait->_qproc = NULL; |
1da177e4c
|
436 437 438 439 |
} if ((mask & POLLOUT_SET) && (out & bit)) { res_out |= bit; retval++; |
626cf2366
|
440 |
wait->_qproc = NULL; |
1da177e4c
|
441 442 443 444 |
} if ((mask & POLLEX_SET) && (ex & bit)) { res_ex |= bit; retval++; |
626cf2366
|
445 |
wait->_qproc = NULL; |
1da177e4c
|
446 |
} |
2d48d67fa
|
447 |
/* got something, stop busy polling */ |
cbf55001b
|
448 449 450 451 452 453 454 455 456 457 |
if (retval) { can_busy_loop = false; busy_flag = 0; /* * only remember a returned * POLL_BUSY_LOOP if we asked for it */ } else if (busy_flag & mask) can_busy_loop = true; |
1da177e4c
|
458 |
} |
1da177e4c
|
459 460 461 462 463 464 465 |
} if (res_in) *rinp = res_in; if (res_out) *routp = res_out; if (res_ex) *rexp = res_ex; |
55d853849
|
466 |
cond_resched(); |
1da177e4c
|
467 |
} |
626cf2366
|
468 |
wait->_qproc = NULL; |
8ff3e8e85
|
469 |
if (retval || timed_out || signal_pending(current)) |
1da177e4c
|
470 |
break; |
f5264481c
|
471 |
if (table.error) { |
1da177e4c
|
472 473 474 |
retval = table.error; break; } |
9f72949f6
|
475 |
|
cbf55001b
|
476 |
/* only if found POLL_BUSY_LOOP sockets && not out of time */ |
76b1e9b98
|
477 478 479 480 481 482 483 484 485 |
if (can_busy_loop && !need_resched()) { if (!busy_end) { busy_end = busy_loop_end_time(); continue; } if (!busy_loop_timeout(busy_end)) continue; } busy_flag = 0; |
2d48d67fa
|
486 |
|
8ff3e8e85
|
487 488 489 490 491 492 493 494 |
/* * If this is the first loop and we have a timeout * given, then we convert to ktime_t and set the to * pointer to the expiry value. */ if (end_time && !to) { expire = timespec_to_ktime(*end_time); to = &expire; |
9f72949f6
|
495 |
} |
8ff3e8e85
|
496 |
|
5f820f648
|
497 498 |
if (!poll_schedule_timeout(&table, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85
|
499 |
timed_out = 1; |
1da177e4c
|
500 |
} |
1da177e4c
|
501 502 |
poll_freewait(&table); |
1da177e4c
|
503 504 |
return retval; } |
1da177e4c
|
505 506 507 508 509 510 511 512 |
/* * We can actually return ERESTARTSYS instead of EINTR, but I'd * like to be certain this leads to no problems. So I return * EINTR just for safety. * * Update: ERESTARTSYS breaks at least the xview clock binary, so * I'm trying ERESTARTNOHAND which restart only when you want to. */ |
a2dcb44c3
|
513 |
int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, |
8ff3e8e85
|
514 |
fd_set __user *exp, struct timespec *end_time) |
1da177e4c
|
515 516 |
{ fd_set_bits fds; |
29ff2db55
|
517 |
void *bits; |
bbea9f696
|
518 |
int ret, max_fds; |
b04eb6aa0
|
519 |
unsigned int size; |
badf16621
|
520 |
struct fdtable *fdt; |
70674f95c
|
521 |
/* Allocate small arguments on the stack to save memory and be faster */ |
30c14e40e
|
522 |
long stack_fds[SELECT_STACK_ALLOC/sizeof(long)]; |
1da177e4c
|
523 |
|
1da177e4c
|
524 525 526 |
ret = -EINVAL; if (n < 0) goto out_nofds; |
bbea9f696
|
527 |
/* max_fds can increase, so grab it once to avoid race */ |
b835996f6
|
528 |
rcu_read_lock(); |
badf16621
|
529 |
fdt = files_fdtable(current->files); |
bbea9f696
|
530 |
max_fds = fdt->max_fds; |
b835996f6
|
531 |
rcu_read_unlock(); |
bbea9f696
|
532 533 |
if (n > max_fds) n = max_fds; |
1da177e4c
|
534 535 536 537 538 539 |
/* * We need 6 bitmaps (in/out/ex for both incoming and outgoing), * since we used fdset we need to allocate memory in units of * long-words. */ |
1da177e4c
|
540 |
size = FDS_BYTES(n); |
b04eb6aa0
|
541 542 543 544 |
bits = stack_fds; if (size > sizeof(stack_fds) / 6) { /* Not enough space in on-stack array; must use kmalloc */ ret = -ENOMEM; |
70674f95c
|
545 |
bits = kmalloc(6 * size, GFP_KERNEL); |
b04eb6aa0
|
546 547 548 |
if (!bits) goto out_nofds; } |
29ff2db55
|
549 550 551 552 553 554 |
fds.in = bits; fds.out = bits + size; fds.ex = bits + 2*size; fds.res_in = bits + 3*size; fds.res_out = bits + 4*size; fds.res_ex = bits + 5*size; |
1da177e4c
|
555 556 557 558 559 560 561 562 |
if ((ret = get_fd_set(n, inp, fds.in)) || (ret = get_fd_set(n, outp, fds.out)) || (ret = get_fd_set(n, exp, fds.ex))) goto out; zero_fd_set(n, fds.res_in); zero_fd_set(n, fds.res_out); zero_fd_set(n, fds.res_ex); |
8ff3e8e85
|
563 |
ret = do_select(n, &fds, end_time); |
1da177e4c
|
564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 |
if (ret < 0) goto out; if (!ret) { ret = -ERESTARTNOHAND; if (signal_pending(current)) goto out; ret = 0; } if (set_fd_set(n, inp, fds.res_in) || set_fd_set(n, outp, fds.res_out) || set_fd_set(n, exp, fds.res_ex)) ret = -EFAULT; out: |
70674f95c
|
580 581 |
if (bits != stack_fds) kfree(bits); |
1da177e4c
|
582 583 584 |
out_nofds: return ret; } |
5a8a82b1d
|
585 586 |
SYSCALL_DEFINE5(select, int, n, fd_set __user *, inp, fd_set __user *, outp, fd_set __user *, exp, struct timeval __user *, tvp) |
9f72949f6
|
587 |
{ |
8ff3e8e85
|
588 |
struct timespec end_time, *to = NULL; |
9f72949f6
|
589 590 591 592 593 594 |
struct timeval tv; int ret; if (tvp) { if (copy_from_user(&tv, tvp, sizeof(tv))) return -EFAULT; |
8ff3e8e85
|
595 |
to = &end_time; |
4d36a9e65
|
596 597 598 |
if (poll_select_set_timeout(to, tv.tv_sec + (tv.tv_usec / USEC_PER_SEC), (tv.tv_usec % USEC_PER_SEC) * NSEC_PER_USEC)) |
9f72949f6
|
599 |
return -EINVAL; |
9f72949f6
|
600 |
} |
8ff3e8e85
|
601 602 |
ret = core_sys_select(n, inp, outp, exp, to); ret = poll_select_copy_remaining(&end_time, tvp, 1, ret); |
9f72949f6
|
603 604 605 |
return ret; } |
c9da9f212
|
606 607 608 |
static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, struct timespec __user *tsp, const sigset_t __user *sigmask, size_t sigsetsize) |
9f72949f6
|
609 |
{ |
9f72949f6
|
610 |
sigset_t ksigmask, sigsaved; |
8ff3e8e85
|
611 |
struct timespec ts, end_time, *to = NULL; |
9f72949f6
|
612 613 614 615 616 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
8ff3e8e85
|
617 618 |
to = &end_time; if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) |
9f72949f6
|
619 |
return -EINVAL; |
9f72949f6
|
620 621 622 623 624 625 626 627 628 629 630 631 |
} if (sigmask) { /* XXX: Don't preclude handling different sized sigset_t's. */ 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); } |
62568510b
|
632 |
ret = core_sys_select(n, inp, outp, exp, to); |
8ff3e8e85
|
633 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6
|
634 635 636 637 638 639 640 641 642 643 |
if (ret == -ERESTARTNOHAND) { /* * Don't restore the signal mask yet. Let do_signal() deliver * the signal on the way back to userspace, before the signal * mask is restored. */ if (sigmask) { memcpy(¤t->saved_sigmask, &sigsaved, sizeof(sigsaved)); |
4e4c22c71
|
644 |
set_restore_sigmask(); |
9f72949f6
|
645 646 647 648 649 650 651 652 653 654 655 656 657 |
} } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); return ret; } /* * Most architectures can't handle 7-argument syscalls. So we provide a * 6-argument version where the sixth argument is a pointer to a structure * which has a pointer to the sigset_t itself followed by a size_t containing * the sigset size. */ |
d4e82042c
|
658 659 660 |
SYSCALL_DEFINE6(pselect6, int, n, fd_set __user *, inp, fd_set __user *, outp, fd_set __user *, exp, struct timespec __user *, tsp, void __user *, sig) |
9f72949f6
|
661 662 663 664 665 666 |
{ size_t sigsetsize = 0; sigset_t __user *up = NULL; if (sig) { if (!access_ok(VERIFY_READ, sig, sizeof(void *)+sizeof(size_t)) |
e110ab94e
|
667 |
|| __get_user(up, (sigset_t __user * __user *)sig) |
9f72949f6
|
668 |
|| __get_user(sigsetsize, |
e110ab94e
|
669 |
(size_t __user *)(sig+sizeof(void *)))) |
9f72949f6
|
670 671 |
return -EFAULT; } |
c9da9f212
|
672 |
return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize); |
9f72949f6
|
673 |
} |
9f72949f6
|
674 |
|
5d0e52830
|
675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 |
#ifdef __ARCH_WANT_SYS_OLD_SELECT struct sel_arg_struct { unsigned long n; fd_set __user *inp, *outp, *exp; struct timeval __user *tvp; }; SYSCALL_DEFINE1(old_select, struct sel_arg_struct __user *, arg) { struct sel_arg_struct a; if (copy_from_user(&a, arg, sizeof(a))) return -EFAULT; return sys_select(a.n, a.inp, a.outp, a.exp, a.tvp); } #endif |
1da177e4c
|
691 692 693 694 695 696 697 |
struct poll_list { struct poll_list *next; int len; struct pollfd entries[0]; }; #define POLLFD_PER_PAGE ((PAGE_SIZE-sizeof(struct poll_list)) / sizeof(struct pollfd)) |
4a4b69f79
|
698 699 700 701 702 |
/* * Fish for pollable events on the pollfd->fd file descriptor. We're only * interested in events matching the pollfd->events mask, and the result * matching that mask is both recorded in pollfd->revents and returned. The * pwait poll_table will be used by the fd-provided poll handler for waiting, |
626cf2366
|
703 |
* if pwait->_qproc is non-NULL. |
4a4b69f79
|
704 |
*/ |
2d48d67fa
|
705 |
static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait, |
cbf55001b
|
706 707 |
bool *can_busy_poll, unsigned int busy_flag) |
1da177e4c
|
708 |
{ |
4a4b69f79
|
709 710 711 712 713 714 |
unsigned int mask; int fd; mask = 0; fd = pollfd->fd; if (fd >= 0) { |
2903ff019
|
715 |
struct fd f = fdget(fd); |
4a4b69f79
|
716 |
mask = POLLNVAL; |
2903ff019
|
717 |
if (f.file) { |
4a4b69f79
|
718 |
mask = DEFAULT_POLLMASK; |
72c2d5319
|
719 |
if (f.file->f_op->poll) { |
626cf2366
|
720 |
pwait->_key = pollfd->events|POLLERR|POLLHUP; |
cbf55001b
|
721 |
pwait->_key |= busy_flag; |
2903ff019
|
722 |
mask = f.file->f_op->poll(f.file, pwait); |
cbf55001b
|
723 724 |
if (mask & busy_flag) *can_busy_poll = true; |
4938d7e02
|
725 |
} |
4a4b69f79
|
726 727 |
/* Mask out unneeded events. */ mask &= pollfd->events | POLLERR | POLLHUP; |
2903ff019
|
728 |
fdput(f); |
1da177e4c
|
729 |
} |
1da177e4c
|
730 |
} |
4a4b69f79
|
731 732 733 |
pollfd->revents = mask; return mask; |
1da177e4c
|
734 735 736 |
} static int do_poll(unsigned int nfds, struct poll_list *list, |
8ff3e8e85
|
737 |
struct poll_wqueues *wait, struct timespec *end_time) |
1da177e4c
|
738 |
{ |
1da177e4c
|
739 |
poll_table* pt = &wait->pt; |
8ff3e8e85
|
740 741 |
ktime_t expire, *to = NULL; int timed_out = 0, count = 0; |
90d6e24a3
|
742 |
unsigned long slack = 0; |
cbf55001b
|
743 |
unsigned int busy_flag = net_busy_loop_on() ? POLL_BUSY_LOOP : 0; |
76b1e9b98
|
744 |
unsigned long busy_end = 0; |
1da177e4c
|
745 |
|
9f72949f6
|
746 |
/* Optimise the no-wait case */ |
8ff3e8e85
|
747 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
626cf2366
|
748 |
pt->_qproc = NULL; |
8ff3e8e85
|
749 750 |
timed_out = 1; } |
9bf084f70
|
751 |
|
96d2ab484
|
752 |
if (end_time && !timed_out) |
231f3d393
|
753 |
slack = select_estimate_accuracy(end_time); |
90d6e24a3
|
754 |
|
1da177e4c
|
755 756 |
for (;;) { struct poll_list *walk; |
cbf55001b
|
757 |
bool can_busy_loop = false; |
9f72949f6
|
758 |
|
4a4b69f79
|
759 760 761 762 763 764 765 766 |
for (walk = list; walk != NULL; walk = walk->next) { struct pollfd * pfd, * pfd_end; pfd = walk->entries; pfd_end = pfd + walk->len; for (; pfd != pfd_end; pfd++) { /* * Fish for events. If we found one, record it |
626cf2366
|
767 |
* and kill poll_table->_qproc, so we don't |
4a4b69f79
|
768 769 770 771 |
* needlessly register any other waiters after * this. They'll get immediately deregistered * when we break out and return. */ |
cbf55001b
|
772 773 |
if (do_pollfd(pfd, pt, &can_busy_loop, busy_flag)) { |
4a4b69f79
|
774 |
count++; |
626cf2366
|
775 |
pt->_qproc = NULL; |
cbf55001b
|
776 777 778 |
/* found something, stop busy polling */ busy_flag = 0; can_busy_loop = false; |
4a4b69f79
|
779 780 |
} } |
1da177e4c
|
781 |
} |
4a4b69f79
|
782 783 |
/* * All waiters have already been registered, so don't provide |
626cf2366
|
784 |
* a poll_table->_qproc to them on the next loop iteration. |
4a4b69f79
|
785 |
*/ |
626cf2366
|
786 |
pt->_qproc = NULL; |
9bf084f70
|
787 788 789 790 791 |
if (!count) { count = wait->error; if (signal_pending(current)) count = -EINTR; } |
8ff3e8e85
|
792 |
if (count || timed_out) |
1da177e4c
|
793 |
break; |
9f72949f6
|
794 |
|
cbf55001b
|
795 |
/* only if found POLL_BUSY_LOOP sockets && not out of time */ |
76b1e9b98
|
796 797 798 799 800 801 802 803 804 |
if (can_busy_loop && !need_resched()) { if (!busy_end) { busy_end = busy_loop_end_time(); continue; } if (!busy_loop_timeout(busy_end)) continue; } busy_flag = 0; |
91e2fd337
|
805 |
|
8ff3e8e85
|
806 807 808 809 810 811 812 813 |
/* * If this is the first loop and we have a timeout * given, then we convert to ktime_t and set the to * pointer to the expiry value. */ if (end_time && !to) { expire = timespec_to_ktime(*end_time); to = &expire; |
9f72949f6
|
814 |
} |
5f820f648
|
815 |
if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85
|
816 |
timed_out = 1; |
1da177e4c
|
817 |
} |
1da177e4c
|
818 819 |
return count; } |
70674f95c
|
820 821 |
#define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \ sizeof(struct pollfd)) |
8ff3e8e85
|
822 823 |
int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, struct timespec *end_time) |
1da177e4c
|
824 825 |
{ struct poll_wqueues table; |
252e5725c
|
826 |
int err = -EFAULT, fdcount, len, size; |
30c14e40e
|
827 828 829 830 |
/* Allocate small arguments on the stack to save memory and be faster - use long to make sure the buffer is aligned properly on 64 bit archs to avoid unaligned access */ long stack_pps[POLL_STACK_ALLOC/sizeof(long)]; |
252e5725c
|
831 832 833 |
struct poll_list *const head = (struct poll_list *)stack_pps; struct poll_list *walk = head; unsigned long todo = nfds; |
1da177e4c
|
834 |
|
d554ed895
|
835 |
if (nfds > rlimit(RLIMIT_NOFILE)) |
1da177e4c
|
836 |
return -EINVAL; |
252e5725c
|
837 838 839 840 841 842 |
len = min_t(unsigned int, nfds, N_STACK_PPS); for (;;) { walk->next = NULL; walk->len = len; if (!len) break; |
1da177e4c
|
843 |
|
252e5725c
|
844 845 846 847 848 849 850 |
if (copy_from_user(walk->entries, ufds + nfds-todo, sizeof(struct pollfd) * walk->len)) goto out_fds; todo -= walk->len; if (!todo) break; |
1da177e4c
|
851 |
|
252e5725c
|
852 853 854 855 856 |
len = min(todo, POLLFD_PER_PAGE); size = sizeof(struct poll_list) + sizeof(struct pollfd) * len; walk = walk->next = kmalloc(size, GFP_KERNEL); if (!walk) { err = -ENOMEM; |
1da177e4c
|
857 858 |
goto out_fds; } |
1da177e4c
|
859 |
} |
9f72949f6
|
860 |
|
252e5725c
|
861 |
poll_initwait(&table); |
8ff3e8e85
|
862 |
fdcount = do_poll(nfds, head, &table, end_time); |
252e5725c
|
863 |
poll_freewait(&table); |
1da177e4c
|
864 |
|
252e5725c
|
865 |
for (walk = head; walk; walk = walk->next) { |
1da177e4c
|
866 867 |
struct pollfd *fds = walk->entries; int j; |
252e5725c
|
868 869 |
for (j = 0; j < walk->len; j++, ufds++) if (__put_user(fds[j].revents, &ufds->revents)) |
1da177e4c
|
870 |
goto out_fds; |
1da177e4c
|
871 |
} |
252e5725c
|
872 |
|
1da177e4c
|
873 |
err = fdcount; |
1da177e4c
|
874 |
out_fds: |
252e5725c
|
875 876 877 878 879 |
walk = head->next; while (walk) { struct poll_list *pos = walk; walk = walk->next; kfree(pos); |
1da177e4c
|
880 |
} |
252e5725c
|
881 |
|
1da177e4c
|
882 883 |
return err; } |
9f72949f6
|
884 |
|
3075d9da0
|
885 886 |
static long do_restart_poll(struct restart_block *restart_block) { |
8ff3e8e85
|
887 888 889 |
struct pollfd __user *ufds = restart_block->poll.ufds; int nfds = restart_block->poll.nfds; struct timespec *to = NULL, end_time; |
3075d9da0
|
890 |
int ret; |
8ff3e8e85
|
891 892 893 894 895 896 897 |
if (restart_block->poll.has_timeout) { end_time.tv_sec = restart_block->poll.tv_sec; end_time.tv_nsec = restart_block->poll.tv_nsec; to = &end_time; } ret = do_sys_poll(ufds, nfds, to); |
3075d9da0
|
898 899 |
if (ret == -EINTR) { restart_block->fn = do_restart_poll; |
3075d9da0
|
900 901 902 903 |
ret = -ERESTART_RESTARTBLOCK; } return ret; } |
5a8a82b1d
|
904 |
SYSCALL_DEFINE3(poll, struct pollfd __user *, ufds, unsigned int, nfds, |
faf309009
|
905 |
int, timeout_msecs) |
9f72949f6
|
906 |
{ |
8ff3e8e85
|
907 |
struct timespec end_time, *to = NULL; |
3075d9da0
|
908 |
int ret; |
9f72949f6
|
909 |
|
8ff3e8e85
|
910 911 912 913 |
if (timeout_msecs >= 0) { to = &end_time; poll_select_set_timeout(to, timeout_msecs / MSEC_PER_SEC, NSEC_PER_MSEC * (timeout_msecs % MSEC_PER_SEC)); |
9f72949f6
|
914 |
} |
8ff3e8e85
|
915 |
ret = do_sys_poll(ufds, nfds, to); |
3075d9da0
|
916 917 |
if (ret == -EINTR) { struct restart_block *restart_block; |
8ff3e8e85
|
918 |
|
3075d9da0
|
919 920 |
restart_block = ¤t_thread_info()->restart_block; restart_block->fn = do_restart_poll; |
8ff3e8e85
|
921 922 923 924 925 926 927 928 929 |
restart_block->poll.ufds = ufds; restart_block->poll.nfds = nfds; if (timeout_msecs >= 0) { restart_block->poll.tv_sec = end_time.tv_sec; restart_block->poll.tv_nsec = end_time.tv_nsec; restart_block->poll.has_timeout = 1; } else restart_block->poll.has_timeout = 0; |
3075d9da0
|
930 931 932 |
ret = -ERESTART_RESTARTBLOCK; } return ret; |
9f72949f6
|
933 |
} |
d4e82042c
|
934 935 936 |
SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds, struct timespec __user *, tsp, const sigset_t __user *, sigmask, size_t, sigsetsize) |
9f72949f6
|
937 938 |
{ sigset_t ksigmask, sigsaved; |
8ff3e8e85
|
939 |
struct timespec ts, end_time, *to = NULL; |
9f72949f6
|
940 941 942 943 944 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
8ff3e8e85
|
945 946 947 |
to = &end_time; if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) return -EINVAL; |
9f72949f6
|
948 949 950 951 952 953 954 955 956 957 958 959 |
} if (sigmask) { /* XXX: Don't preclude handling different sized sigset_t's. */ 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); } |
8ff3e8e85
|
960 |
ret = do_sys_poll(ufds, nfds, to); |
9f72949f6
|
961 962 963 964 965 966 967 968 969 970 971 |
/* We can restart this syscall, usually */ if (ret == -EINTR) { /* * Don't restore the signal mask yet. Let do_signal() deliver * the signal on the way back to userspace, before the signal * mask is restored. */ if (sigmask) { memcpy(¤t->saved_sigmask, &sigsaved, sizeof(sigsaved)); |
4e4c22c71
|
972 |
set_restore_sigmask(); |
9f72949f6
|
973 974 975 976 |
} ret = -ERESTARTNOHAND; } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); |
8ff3e8e85
|
977 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6
|
978 979 980 |
return ret; } |