Blame view
fs/select.c
24.2 KB
1da177e4c Linux-2.6.12-rc2 |
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 ROUND_UP macro cl... |
16 |
#include <linux/kernel.h> |
a99bbaf5e headers: remove s... |
17 |
#include <linux/sched.h> |
1da177e4c Linux-2.6.12-rc2 |
18 19 20 |
#include <linux/syscalls.h> #include <linux/module.h> #include <linux/slab.h> |
1da177e4c Linux-2.6.12-rc2 |
21 22 23 |
#include <linux/poll.h> #include <linux/personality.h> /* for STICKY_TIMEOUTS */ #include <linux/file.h> |
9f3acc314 [PATCH] split lin... |
24 |
#include <linux/fdtable.h> |
1da177e4c Linux-2.6.12-rc2 |
25 |
#include <linux/fs.h> |
b835996f6 [PATCH] files: lo... |
26 |
#include <linux/rcupdate.h> |
8ff3e8e85 select: switch se... |
27 |
#include <linux/hrtimer.h> |
1da177e4c Linux-2.6.12-rc2 |
28 29 |
#include <asm/uaccess.h> |
90d6e24a3 hrtimer: make sel... |
30 31 32 33 34 35 36 37 38 39 40 41 |
/* * 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 poll/select: avoi... |
42 |
#define MAX_SLACK (100 * NSEC_PER_MSEC) |
96d2ab484 hrtimer: fix sign... |
43 |
static long __estimate_accuracy(struct timespec *tv) |
90d6e24a3 hrtimer: make sel... |
44 |
{ |
96d2ab484 hrtimer: fix sign... |
45 |
long slack; |
90d6e24a3 hrtimer: make sel... |
46 |
int divfactor = 1000; |
5ae87e79e poll/select: avoi... |
47 48 |
if (tv->tv_sec < 0) return 0; |
4ce105d30 hrtimer: incorpor... |
49 |
if (task_nice(current) > 0) |
90d6e24a3 hrtimer: make sel... |
50 |
divfactor = divfactor / 5; |
5ae87e79e poll/select: avoi... |
51 52 |
if (tv->tv_sec > MAX_SLACK / (NSEC_PER_SEC/divfactor)) return MAX_SLACK; |
90d6e24a3 hrtimer: make sel... |
53 54 |
slack = tv->tv_nsec / divfactor; slack += tv->tv_sec * (NSEC_PER_SEC/divfactor); |
5ae87e79e poll/select: avoi... |
55 56 |
if (slack > MAX_SLACK) return MAX_SLACK; |
96d2ab484 hrtimer: fix sign... |
57 |
|
90d6e24a3 hrtimer: make sel... |
58 59 |
return slack; } |
96d2ab484 hrtimer: fix sign... |
60 |
static long estimate_accuracy(struct timespec *tv) |
90d6e24a3 hrtimer: make sel... |
61 62 63 64 65 66 67 |
{ unsigned long ret; struct timespec now; /* * Realtime tasks get a slack of 0 for obvious reasons. */ |
4ce105d30 hrtimer: incorpor... |
68 |
if (rt_task(current)) |
90d6e24a3 hrtimer: make sel... |
69 70 71 72 73 74 75 76 77 |
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 Linux-2.6.12-rc2 |
78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 |
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 [PATCH] make some... |
99 100 |
static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p); |
1da177e4c Linux-2.6.12-rc2 |
101 102 103 104 |
void poll_initwait(struct poll_wqueues *pwq) { init_poll_funcptr(&pwq->pt, __pollwait); |
5f820f648 poll: allow f_op-... |
105 |
pwq->polling_task = current; |
b2add73db poll/select: init... |
106 |
pwq->triggered = 0; |
1da177e4c Linux-2.6.12-rc2 |
107 108 |
pwq->error = 0; pwq->table = NULL; |
70674f95c [PATCH] Optimize ... |
109 |
pwq->inline_index = 0; |
1da177e4c Linux-2.6.12-rc2 |
110 |
} |
1da177e4c Linux-2.6.12-rc2 |
111 |
EXPORT_SYMBOL(poll_initwait); |
70674f95c [PATCH] Optimize ... |
112 113 |
static void free_poll_entry(struct poll_table_entry *entry) { |
ccf6780dc Style fix in fs/s... |
114 |
remove_wait_queue(entry->wait_address, &entry->wait); |
70674f95c [PATCH] Optimize ... |
115 116 |
fput(entry->filp); } |
1da177e4c Linux-2.6.12-rc2 |
117 118 119 |
void poll_freewait(struct poll_wqueues *pwq) { struct poll_table_page * p = pwq->table; |
70674f95c [PATCH] Optimize ... |
120 121 122 |
int i; for (i = 0; i < pwq->inline_index; i++) free_poll_entry(pwq->inline_entries + i); |
1da177e4c Linux-2.6.12-rc2 |
123 124 125 126 127 128 129 |
while (p) { struct poll_table_entry * entry; struct poll_table_page *old; entry = p->entry; do { entry--; |
70674f95c [PATCH] Optimize ... |
130 |
free_poll_entry(entry); |
1da177e4c Linux-2.6.12-rc2 |
131 132 133 134 135 136 |
} while (entry > p->entries); old = p; p = p->next; free_page((unsigned long) old); } } |
1da177e4c Linux-2.6.12-rc2 |
137 |
EXPORT_SYMBOL(poll_freewait); |
5f820f648 poll: allow f_op-... |
138 |
static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p) |
1da177e4c Linux-2.6.12-rc2 |
139 |
{ |
1da177e4c Linux-2.6.12-rc2 |
140 |
struct poll_table_page *table = p->table; |
70674f95c [PATCH] Optimize ... |
141 142 |
if (p->inline_index < N_INLINE_POLL_ENTRIES) return p->inline_entries + p->inline_index++; |
1da177e4c Linux-2.6.12-rc2 |
143 144 145 146 147 148 |
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 [PATCH] Optimize ... |
149 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
150 151 152 153 154 155 |
} new_table->entry = new_table->entries; new_table->next = table; p->table = new_table; table = new_table; } |
70674f95c [PATCH] Optimize ... |
156 157 |
return table->entry++; } |
4938d7e02 poll: avoid extra... |
158 |
static int __pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) |
5f820f648 poll: allow f_op-... |
159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 |
{ 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 poll: avoid extra... |
183 184 185 186 187 188 189 190 191 |
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 [PATCH] Optimize ... |
192 193 194 195 |
/* Add a new entry */ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) { |
5f820f648 poll: allow f_op-... |
196 197 |
struct poll_wqueues *pwq = container_of(p, struct poll_wqueues, pt); struct poll_table_entry *entry = poll_get_entry(pwq); |
70674f95c [PATCH] Optimize ... |
198 199 200 201 202 |
if (!entry) return; get_file(filp); entry->filp = filp; entry->wait_address = wait_address; |
4938d7e02 poll: avoid extra... |
203 |
entry->key = p->key; |
5f820f648 poll: allow f_op-... |
204 205 |
init_waitqueue_func_entry(&entry->wait, pollwake); entry->wait.private = pwq; |
ccf6780dc Style fix in fs/s... |
206 |
add_wait_queue(wait_address, &entry->wait); |
1da177e4c Linux-2.6.12-rc2 |
207 |
} |
5f820f648 poll: allow f_op-... |
208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 |
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) rc = schedule_hrtimeout_range(expires, slack, HRTIMER_MODE_ABS); __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 select: add poll_... |
234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 |
/** * 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) { 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 Linux-2.6.12-rc2 |
306 307 308 309 310 311 312 313 314 315 316 |
#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 [PATCH] files: br... |
317 |
struct fdtable *fdt; |
1da177e4c Linux-2.6.12-rc2 |
318 319 320 321 |
/* handle last in-complete long-word first */ set = ~(~0UL << (n & (__NFDBITS-1))); n /= __NFDBITS; |
badf16621 [PATCH] files: br... |
322 323 |
fdt = files_fdtable(current->files); open_fds = fdt->open_fds->fds_bits+n; |
1da177e4c Linux-2.6.12-rc2 |
324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 |
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); max += n * __NFDBITS; } return max; } |
1da177e4c Linux-2.6.12-rc2 |
353 354 355 |
#define POLLIN_SET (POLLRDNORM | POLLRDBAND | POLLIN | POLLHUP | POLLERR) #define POLLOUT_SET (POLLWRBAND | POLLWRNORM | POLLOUT | POLLERR) #define POLLEX_SET (POLLPRI) |
4938d7e02 poll: avoid extra... |
356 357 358 359 360 361 362 363 364 365 366 |
static inline void wait_key_set(poll_table *wait, unsigned long in, unsigned long out, unsigned long bit) { if (wait) { wait->key = POLLEX_SET; if (in & bit) wait->key |= POLLIN_SET; if (out & bit) wait->key |= POLLOUT_SET; } } |
8ff3e8e85 select: switch se... |
367 |
int do_select(int n, fd_set_bits *fds, struct timespec *end_time) |
1da177e4c Linux-2.6.12-rc2 |
368 |
{ |
8ff3e8e85 select: switch se... |
369 |
ktime_t expire, *to = NULL; |
1da177e4c Linux-2.6.12-rc2 |
370 371 |
struct poll_wqueues table; poll_table *wait; |
8ff3e8e85 select: switch se... |
372 |
int retval, i, timed_out = 0; |
90d6e24a3 hrtimer: make sel... |
373 |
unsigned long slack = 0; |
1da177e4c Linux-2.6.12-rc2 |
374 |
|
b835996f6 [PATCH] files: lo... |
375 |
rcu_read_lock(); |
1da177e4c Linux-2.6.12-rc2 |
376 |
retval = max_select_fd(n, fds); |
b835996f6 [PATCH] files: lo... |
377 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
378 379 380 381 382 383 384 |
if (retval < 0) return retval; n = retval; poll_initwait(&table); wait = &table.pt; |
8ff3e8e85 select: switch se... |
385 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
1da177e4c Linux-2.6.12-rc2 |
386 |
wait = NULL; |
8ff3e8e85 select: switch se... |
387 388 |
timed_out = 1; } |
96d2ab484 hrtimer: fix sign... |
389 |
if (end_time && !timed_out) |
90d6e24a3 hrtimer: make sel... |
390 |
slack = estimate_accuracy(end_time); |
1da177e4c Linux-2.6.12-rc2 |
391 392 393 |
retval = 0; for (;;) { unsigned long *rinp, *routp, *rexp, *inp, *outp, *exp; |
1da177e4c Linux-2.6.12-rc2 |
394 395 396 397 398 399 |
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; |
99ac48f54 [PATCH] mark f_op... |
400 |
const struct file_operations *f_op = NULL; |
1da177e4c Linux-2.6.12-rc2 |
401 402 403 404 405 406 407 408 409 410 |
struct file *file = NULL; in = *inp++; out = *outp++; ex = *exp++; all_bits = in | out | ex; if (all_bits == 0) { i += __NFDBITS; continue; } for (j = 0; j < __NFDBITS; ++j, ++i, bit <<= 1) { |
e4a1f129f [PATCH] use fget_... |
411 |
int fput_needed; |
1da177e4c Linux-2.6.12-rc2 |
412 413 414 415 |
if (i >= n) break; if (!(bit & all_bits)) continue; |
e4a1f129f [PATCH] use fget_... |
416 |
file = fget_light(i, &fput_needed); |
1da177e4c Linux-2.6.12-rc2 |
417 418 419 |
if (file) { f_op = file->f_op; mask = DEFAULT_POLLMASK; |
4938d7e02 poll: avoid extra... |
420 421 422 423 |
if (f_op && f_op->poll) { wait_key_set(wait, in, out, bit); mask = (*f_op->poll)(file, wait); } |
e4a1f129f [PATCH] use fget_... |
424 |
fput_light(file, fput_needed); |
1da177e4c Linux-2.6.12-rc2 |
425 426 427 |
if ((mask & POLLIN_SET) && (in & bit)) { res_in |= bit; retval++; |
4938d7e02 poll: avoid extra... |
428 |
wait = NULL; |
1da177e4c Linux-2.6.12-rc2 |
429 430 431 432 |
} if ((mask & POLLOUT_SET) && (out & bit)) { res_out |= bit; retval++; |
4938d7e02 poll: avoid extra... |
433 |
wait = NULL; |
1da177e4c Linux-2.6.12-rc2 |
434 435 436 437 |
} if ((mask & POLLEX_SET) && (ex & bit)) { res_ex |= bit; retval++; |
4938d7e02 poll: avoid extra... |
438 |
wait = NULL; |
1da177e4c Linux-2.6.12-rc2 |
439 440 |
} } |
1da177e4c Linux-2.6.12-rc2 |
441 442 443 444 445 446 447 |
} if (res_in) *rinp = res_in; if (res_out) *routp = res_out; if (res_ex) *rexp = res_ex; |
55d853849 Fix performance r... |
448 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
449 450 |
} wait = NULL; |
8ff3e8e85 select: switch se... |
451 |
if (retval || timed_out || signal_pending(current)) |
1da177e4c Linux-2.6.12-rc2 |
452 |
break; |
f5264481c trivial: small cl... |
453 |
if (table.error) { |
1da177e4c Linux-2.6.12-rc2 |
454 455 456 |
retval = table.error; break; } |
9f72949f6 [PATCH] Add psele... |
457 |
|
8ff3e8e85 select: switch se... |
458 459 460 461 462 463 464 465 |
/* * 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 [PATCH] Add psele... |
466 |
} |
8ff3e8e85 select: switch se... |
467 |
|
5f820f648 poll: allow f_op-... |
468 469 |
if (!poll_schedule_timeout(&table, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85 select: switch se... |
470 |
timed_out = 1; |
1da177e4c Linux-2.6.12-rc2 |
471 |
} |
1da177e4c Linux-2.6.12-rc2 |
472 473 |
poll_freewait(&table); |
1da177e4c Linux-2.6.12-rc2 |
474 475 |
return retval; } |
1da177e4c Linux-2.6.12-rc2 |
476 477 478 479 480 481 482 483 484 485 |
/* * 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. */ #define MAX_SELECT_SECONDS \ ((unsigned long) (MAX_SCHEDULE_TIMEOUT / HZ)-1) |
a2dcb44c3 [PATCH] make osf_... |
486 |
int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, |
8ff3e8e85 select: switch se... |
487 |
fd_set __user *exp, struct timespec *end_time) |
1da177e4c Linux-2.6.12-rc2 |
488 489 |
{ fd_set_bits fds; |
29ff2db55 [PATCH] select() ... |
490 |
void *bits; |
bbea9f696 [PATCH] fdtable: ... |
491 |
int ret, max_fds; |
b04eb6aa0 [PATCH] select: d... |
492 |
unsigned int size; |
badf16621 [PATCH] files: br... |
493 |
struct fdtable *fdt; |
70674f95c [PATCH] Optimize ... |
494 |
/* Allocate small arguments on the stack to save memory and be faster */ |
30c14e40e [PATCH] avoid una... |
495 |
long stack_fds[SELECT_STACK_ALLOC/sizeof(long)]; |
1da177e4c Linux-2.6.12-rc2 |
496 |
|
1da177e4c Linux-2.6.12-rc2 |
497 498 499 |
ret = -EINVAL; if (n < 0) goto out_nofds; |
bbea9f696 [PATCH] fdtable: ... |
500 |
/* max_fds can increase, so grab it once to avoid race */ |
b835996f6 [PATCH] files: lo... |
501 |
rcu_read_lock(); |
badf16621 [PATCH] files: br... |
502 |
fdt = files_fdtable(current->files); |
bbea9f696 [PATCH] fdtable: ... |
503 |
max_fds = fdt->max_fds; |
b835996f6 [PATCH] files: lo... |
504 |
rcu_read_unlock(); |
bbea9f696 [PATCH] fdtable: ... |
505 506 |
if (n > max_fds) n = max_fds; |
1da177e4c Linux-2.6.12-rc2 |
507 508 509 510 511 512 |
/* * 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 Linux-2.6.12-rc2 |
513 |
size = FDS_BYTES(n); |
b04eb6aa0 [PATCH] select: d... |
514 515 516 517 |
bits = stack_fds; if (size > sizeof(stack_fds) / 6) { /* Not enough space in on-stack array; must use kmalloc */ ret = -ENOMEM; |
70674f95c [PATCH] Optimize ... |
518 |
bits = kmalloc(6 * size, GFP_KERNEL); |
b04eb6aa0 [PATCH] select: d... |
519 520 521 |
if (!bits) goto out_nofds; } |
29ff2db55 [PATCH] select() ... |
522 523 524 525 526 527 |
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 Linux-2.6.12-rc2 |
528 529 530 531 532 533 534 535 |
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 select: switch se... |
536 |
ret = do_select(n, &fds, end_time); |
1da177e4c Linux-2.6.12-rc2 |
537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 |
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 [PATCH] Optimize ... |
553 554 |
if (bits != stack_fds) kfree(bits); |
1da177e4c Linux-2.6.12-rc2 |
555 556 557 |
out_nofds: return ret; } |
5a8a82b1d [CVE-2009-0029] S... |
558 559 |
SYSCALL_DEFINE5(select, int, n, fd_set __user *, inp, fd_set __user *, outp, fd_set __user *, exp, struct timeval __user *, tvp) |
9f72949f6 [PATCH] Add psele... |
560 |
{ |
8ff3e8e85 select: switch se... |
561 |
struct timespec end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
562 563 564 565 566 567 |
struct timeval tv; int ret; if (tvp) { if (copy_from_user(&tv, tvp, sizeof(tv))) return -EFAULT; |
8ff3e8e85 select: switch se... |
568 |
to = &end_time; |
4d36a9e65 select: deal with... |
569 570 571 |
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 [PATCH] Add psele... |
572 |
return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
573 |
} |
8ff3e8e85 select: switch se... |
574 575 |
ret = core_sys_select(n, inp, outp, exp, to); ret = poll_select_copy_remaining(&end_time, tvp, 1, ret); |
9f72949f6 [PATCH] Add psele... |
576 577 578 |
return ret; } |
f3de272b8 signals: use HAVE... |
579 |
#ifdef HAVE_SET_RESTORE_SIGMASK |
c9da9f212 [CVE-2009-0029] M... |
580 581 582 |
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 [PATCH] Add psele... |
583 |
{ |
9f72949f6 [PATCH] Add psele... |
584 |
sigset_t ksigmask, sigsaved; |
8ff3e8e85 select: switch se... |
585 |
struct timespec ts, end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
586 587 588 589 590 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
8ff3e8e85 select: switch se... |
591 592 |
to = &end_time; if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) |
9f72949f6 [PATCH] Add psele... |
593 |
return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
594 595 596 597 598 599 600 601 602 603 604 605 |
} 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 Fix timeouts in s... |
606 |
ret = core_sys_select(n, inp, outp, exp, to); |
8ff3e8e85 select: switch se... |
607 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6 [PATCH] Add psele... |
608 609 610 611 612 613 614 615 616 617 |
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 signals: add set_... |
618 |
set_restore_sigmask(); |
9f72949f6 [PATCH] Add psele... |
619 620 621 622 623 624 625 626 627 628 629 630 631 |
} } 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 [CVE-2009-0029] S... |
632 633 634 |
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 [PATCH] Add psele... |
635 636 637 638 639 640 |
{ size_t sigsetsize = 0; sigset_t __user *up = NULL; if (sig) { if (!access_ok(VERIFY_READ, sig, sizeof(void *)+sizeof(size_t)) |
e110ab94e [PATCH] fix __use... |
641 |
|| __get_user(up, (sigset_t __user * __user *)sig) |
9f72949f6 [PATCH] Add psele... |
642 |
|| __get_user(sigsetsize, |
e110ab94e [PATCH] fix __use... |
643 |
(size_t __user *)(sig+sizeof(void *)))) |
9f72949f6 [PATCH] Add psele... |
644 645 |
return -EFAULT; } |
c9da9f212 [CVE-2009-0029] M... |
646 |
return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize); |
9f72949f6 [PATCH] Add psele... |
647 |
} |
f3de272b8 signals: use HAVE... |
648 |
#endif /* HAVE_SET_RESTORE_SIGMASK */ |
9f72949f6 [PATCH] Add psele... |
649 |
|
5d0e52830 Add generic sys_o... |
650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 |
#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 Linux-2.6.12-rc2 |
666 667 668 669 670 671 672 |
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 [PATCH] Poll clea... |
673 674 675 676 677 678 679 680 |
/* * 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, * if non-NULL. */ static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait) |
1da177e4c Linux-2.6.12-rc2 |
681 |
{ |
4a4b69f79 [PATCH] Poll clea... |
682 683 684 685 686 687 688 689 690 691 692 693 694 |
unsigned int mask; int fd; mask = 0; fd = pollfd->fd; if (fd >= 0) { int fput_needed; struct file * file; file = fget_light(fd, &fput_needed); mask = POLLNVAL; if (file != NULL) { mask = DEFAULT_POLLMASK; |
4938d7e02 poll: avoid extra... |
695 696 697 698 |
if (file->f_op && file->f_op->poll) { if (pwait) pwait->key = pollfd->events | POLLERR | POLLHUP; |
4a4b69f79 [PATCH] Poll clea... |
699 |
mask = file->f_op->poll(file, pwait); |
4938d7e02 poll: avoid extra... |
700 |
} |
4a4b69f79 [PATCH] Poll clea... |
701 702 703 |
/* Mask out unneeded events. */ mask &= pollfd->events | POLLERR | POLLHUP; fput_light(file, fput_needed); |
1da177e4c Linux-2.6.12-rc2 |
704 |
} |
1da177e4c Linux-2.6.12-rc2 |
705 |
} |
4a4b69f79 [PATCH] Poll clea... |
706 707 708 |
pollfd->revents = mask; return mask; |
1da177e4c Linux-2.6.12-rc2 |
709 710 711 |
} static int do_poll(unsigned int nfds, struct poll_list *list, |
8ff3e8e85 select: switch se... |
712 |
struct poll_wqueues *wait, struct timespec *end_time) |
1da177e4c Linux-2.6.12-rc2 |
713 |
{ |
1da177e4c Linux-2.6.12-rc2 |
714 |
poll_table* pt = &wait->pt; |
8ff3e8e85 select: switch se... |
715 716 |
ktime_t expire, *to = NULL; int timed_out = 0, count = 0; |
90d6e24a3 hrtimer: make sel... |
717 |
unsigned long slack = 0; |
1da177e4c Linux-2.6.12-rc2 |
718 |
|
9f72949f6 [PATCH] Add psele... |
719 |
/* Optimise the no-wait case */ |
8ff3e8e85 select: switch se... |
720 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
1da177e4c Linux-2.6.12-rc2 |
721 |
pt = NULL; |
8ff3e8e85 select: switch se... |
722 723 |
timed_out = 1; } |
9bf084f70 do_poll: return -... |
724 |
|
96d2ab484 hrtimer: fix sign... |
725 |
if (end_time && !timed_out) |
90d6e24a3 hrtimer: make sel... |
726 |
slack = estimate_accuracy(end_time); |
1da177e4c Linux-2.6.12-rc2 |
727 728 |
for (;;) { struct poll_list *walk; |
9f72949f6 [PATCH] Add psele... |
729 |
|
4a4b69f79 [PATCH] Poll clea... |
730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 |
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 * and kill the poll_table, so we don't * needlessly register any other waiters after * this. They'll get immediately deregistered * when we break out and return. */ if (do_pollfd(pfd, pt)) { count++; pt = NULL; } } |
1da177e4c Linux-2.6.12-rc2 |
748 |
} |
4a4b69f79 [PATCH] Poll clea... |
749 750 751 752 |
/* * All waiters have already been registered, so don't provide * a poll_table to them on the next loop iteration. */ |
1da177e4c Linux-2.6.12-rc2 |
753 |
pt = NULL; |
9bf084f70 do_poll: return -... |
754 755 756 757 758 |
if (!count) { count = wait->error; if (signal_pending(current)) count = -EINTR; } |
8ff3e8e85 select: switch se... |
759 |
if (count || timed_out) |
1da177e4c Linux-2.6.12-rc2 |
760 |
break; |
9f72949f6 [PATCH] Add psele... |
761 |
|
8ff3e8e85 select: switch se... |
762 763 764 765 766 767 768 769 |
/* * 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 [PATCH] Add psele... |
770 |
} |
5f820f648 poll: allow f_op-... |
771 |
if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85 select: switch se... |
772 |
timed_out = 1; |
1da177e4c Linux-2.6.12-rc2 |
773 |
} |
1da177e4c Linux-2.6.12-rc2 |
774 775 |
return count; } |
70674f95c [PATCH] Optimize ... |
776 777 |
#define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \ sizeof(struct pollfd)) |
8ff3e8e85 select: switch se... |
778 779 |
int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, struct timespec *end_time) |
1da177e4c Linux-2.6.12-rc2 |
780 781 |
{ struct poll_wqueues table; |
252e5725c do_sys_poll: simp... |
782 |
int err = -EFAULT, fdcount, len, size; |
30c14e40e [PATCH] avoid una... |
783 784 785 786 |
/* 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 do_sys_poll: simp... |
787 788 789 |
struct poll_list *const head = (struct poll_list *)stack_pps; struct poll_list *walk = head; unsigned long todo = nfds; |
1da177e4c Linux-2.6.12-rc2 |
790 |
|
d554ed895 fs: use rlimit he... |
791 |
if (nfds > rlimit(RLIMIT_NOFILE)) |
1da177e4c Linux-2.6.12-rc2 |
792 |
return -EINVAL; |
252e5725c do_sys_poll: simp... |
793 794 795 796 797 798 |
len = min_t(unsigned int, nfds, N_STACK_PPS); for (;;) { walk->next = NULL; walk->len = len; if (!len) break; |
1da177e4c Linux-2.6.12-rc2 |
799 |
|
252e5725c do_sys_poll: simp... |
800 801 802 803 804 805 806 |
if (copy_from_user(walk->entries, ufds + nfds-todo, sizeof(struct pollfd) * walk->len)) goto out_fds; todo -= walk->len; if (!todo) break; |
1da177e4c Linux-2.6.12-rc2 |
807 |
|
252e5725c do_sys_poll: simp... |
808 809 810 811 812 |
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 Linux-2.6.12-rc2 |
813 814 |
goto out_fds; } |
1da177e4c Linux-2.6.12-rc2 |
815 |
} |
9f72949f6 [PATCH] Add psele... |
816 |
|
252e5725c do_sys_poll: simp... |
817 |
poll_initwait(&table); |
8ff3e8e85 select: switch se... |
818 |
fdcount = do_poll(nfds, head, &table, end_time); |
252e5725c do_sys_poll: simp... |
819 |
poll_freewait(&table); |
1da177e4c Linux-2.6.12-rc2 |
820 |
|
252e5725c do_sys_poll: simp... |
821 |
for (walk = head; walk; walk = walk->next) { |
1da177e4c Linux-2.6.12-rc2 |
822 823 |
struct pollfd *fds = walk->entries; int j; |
252e5725c do_sys_poll: simp... |
824 825 |
for (j = 0; j < walk->len; j++, ufds++) if (__put_user(fds[j].revents, &ufds->revents)) |
1da177e4c Linux-2.6.12-rc2 |
826 |
goto out_fds; |
1da177e4c Linux-2.6.12-rc2 |
827 |
} |
252e5725c do_sys_poll: simp... |
828 |
|
1da177e4c Linux-2.6.12-rc2 |
829 |
err = fdcount; |
1da177e4c Linux-2.6.12-rc2 |
830 |
out_fds: |
252e5725c do_sys_poll: simp... |
831 832 833 834 835 |
walk = head->next; while (walk) { struct poll_list *pos = walk; walk = walk->next; kfree(pos); |
1da177e4c Linux-2.6.12-rc2 |
836 |
} |
252e5725c do_sys_poll: simp... |
837 |
|
1da177e4c Linux-2.6.12-rc2 |
838 839 |
return err; } |
9f72949f6 [PATCH] Add psele... |
840 |
|
3075d9da0 Use ERESTART_REST... |
841 842 |
static long do_restart_poll(struct restart_block *restart_block) { |
8ff3e8e85 select: switch se... |
843 844 845 |
struct pollfd __user *ufds = restart_block->poll.ufds; int nfds = restart_block->poll.nfds; struct timespec *to = NULL, end_time; |
3075d9da0 Use ERESTART_REST... |
846 |
int ret; |
8ff3e8e85 select: switch se... |
847 848 849 850 851 852 853 |
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 Use ERESTART_REST... |
854 855 |
if (ret == -EINTR) { restart_block->fn = do_restart_poll; |
3075d9da0 Use ERESTART_REST... |
856 857 858 859 |
ret = -ERESTART_RESTARTBLOCK; } return ret; } |
5a8a82b1d [CVE-2009-0029] S... |
860 861 |
SYSCALL_DEFINE3(poll, struct pollfd __user *, ufds, unsigned int, nfds, long, timeout_msecs) |
9f72949f6 [PATCH] Add psele... |
862 |
{ |
8ff3e8e85 select: switch se... |
863 |
struct timespec end_time, *to = NULL; |
3075d9da0 Use ERESTART_REST... |
864 |
int ret; |
9f72949f6 [PATCH] Add psele... |
865 |
|
8ff3e8e85 select: switch se... |
866 867 868 869 |
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 [PATCH] Add psele... |
870 |
} |
8ff3e8e85 select: switch se... |
871 |
ret = do_sys_poll(ufds, nfds, to); |
3075d9da0 Use ERESTART_REST... |
872 873 |
if (ret == -EINTR) { struct restart_block *restart_block; |
8ff3e8e85 select: switch se... |
874 |
|
3075d9da0 Use ERESTART_REST... |
875 876 |
restart_block = ¤t_thread_info()->restart_block; restart_block->fn = do_restart_poll; |
8ff3e8e85 select: switch se... |
877 878 879 880 881 882 883 884 885 |
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 Use ERESTART_REST... |
886 887 888 |
ret = -ERESTART_RESTARTBLOCK; } return ret; |
9f72949f6 [PATCH] Add psele... |
889 |
} |
f3de272b8 signals: use HAVE... |
890 |
#ifdef HAVE_SET_RESTORE_SIGMASK |
d4e82042c [CVE-2009-0029] S... |
891 892 893 |
SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds, struct timespec __user *, tsp, const sigset_t __user *, sigmask, size_t, sigsetsize) |
9f72949f6 [PATCH] Add psele... |
894 895 |
{ sigset_t ksigmask, sigsaved; |
8ff3e8e85 select: switch se... |
896 |
struct timespec ts, end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
897 898 899 900 901 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
8ff3e8e85 select: switch se... |
902 903 904 |
to = &end_time; if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
905 906 907 908 909 910 911 912 913 914 915 916 |
} 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 select: switch se... |
917 |
ret = do_sys_poll(ufds, nfds, to); |
9f72949f6 [PATCH] Add psele... |
918 919 920 921 922 923 924 925 926 927 928 |
/* 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 signals: add set_... |
929 |
set_restore_sigmask(); |
9f72949f6 [PATCH] Add psele... |
930 931 932 933 |
} ret = -ERESTARTNOHAND; } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); |
8ff3e8e85 select: switch se... |
934 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6 [PATCH] Add psele... |
935 936 937 |
return ret; } |
f3de272b8 signals: use HAVE... |
938 |
#endif /* HAVE_SET_RESTORE_SIGMASK */ |