Blame view
fs/select.c
25.5 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 |
#include <linux/syscalls.h> |
630d9c472 fs: reduce the us... |
19 |
#include <linux/export.h> |
1da177e4c Linux-2.6.12-rc2 |
20 |
#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> |
8bd75c77b sched/rt: Move rt... |
28 |
#include <linux/sched/rt.h> |
9745cdb36 select: use freez... |
29 |
#include <linux/freezer.h> |
076bb0c82 net: rename inclu... |
30 |
#include <net/busy_poll.h> |
1da177e4c Linux-2.6.12-rc2 |
31 32 |
#include <asm/uaccess.h> |
90d6e24a3 hrtimer: make sel... |
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 poll/select: avoi... |
45 |
#define MAX_SLACK (100 * NSEC_PER_MSEC) |
766b9f928 fs: poll/select/r... |
46 |
static long __estimate_accuracy(struct timespec64 *tv) |
90d6e24a3 hrtimer: make sel... |
47 |
{ |
96d2ab484 hrtimer: fix sign... |
48 |
long slack; |
90d6e24a3 hrtimer: make sel... |
49 |
int divfactor = 1000; |
5ae87e79e poll/select: avoi... |
50 51 |
if (tv->tv_sec < 0) return 0; |
4ce105d30 hrtimer: incorpor... |
52 |
if (task_nice(current) > 0) |
90d6e24a3 hrtimer: make sel... |
53 |
divfactor = divfactor / 5; |
5ae87e79e poll/select: avoi... |
54 55 |
if (tv->tv_sec > MAX_SLACK / (NSEC_PER_SEC/divfactor)) return MAX_SLACK; |
90d6e24a3 hrtimer: make sel... |
56 57 |
slack = tv->tv_nsec / divfactor; slack += tv->tv_sec * (NSEC_PER_SEC/divfactor); |
5ae87e79e poll/select: avoi... |
58 59 |
if (slack > MAX_SLACK) return MAX_SLACK; |
96d2ab484 hrtimer: fix sign... |
60 |
|
90d6e24a3 hrtimer: make sel... |
61 62 |
return slack; } |
766b9f928 fs: poll/select/r... |
63 |
u64 select_estimate_accuracy(struct timespec64 *tv) |
90d6e24a3 hrtimer: make sel... |
64 |
{ |
da8b44d5a timer: convert ti... |
65 |
u64 ret; |
766b9f928 fs: poll/select/r... |
66 |
struct timespec64 now; |
90d6e24a3 hrtimer: make sel... |
67 68 69 70 |
/* * Realtime tasks get a slack of 0 for obvious reasons. */ |
4ce105d30 hrtimer: incorpor... |
71 |
if (rt_task(current)) |
90d6e24a3 hrtimer: make sel... |
72 |
return 0; |
766b9f928 fs: poll/select/r... |
73 74 |
ktime_get_ts64(&now); now = timespec64_sub(*tv, now); |
90d6e24a3 hrtimer: make sel... |
75 76 77 78 79 |
ret = __estimate_accuracy(&now); if (ret < current->timer_slack_ns) return current->timer_slack_ns; return ret; } |
1da177e4c Linux-2.6.12-rc2 |
80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 |
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... |
101 102 |
static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p); |
1da177e4c Linux-2.6.12-rc2 |
103 104 105 106 |
void poll_initwait(struct poll_wqueues *pwq) { init_poll_funcptr(&pwq->pt, __pollwait); |
5f820f648 poll: allow f_op-... |
107 |
pwq->polling_task = current; |
b2add73db poll/select: init... |
108 |
pwq->triggered = 0; |
1da177e4c Linux-2.6.12-rc2 |
109 110 |
pwq->error = 0; pwq->table = NULL; |
70674f95c [PATCH] Optimize ... |
111 |
pwq->inline_index = 0; |
1da177e4c Linux-2.6.12-rc2 |
112 |
} |
1da177e4c Linux-2.6.12-rc2 |
113 |
EXPORT_SYMBOL(poll_initwait); |
70674f95c [PATCH] Optimize ... |
114 115 |
static void free_poll_entry(struct poll_table_entry *entry) { |
ccf6780dc Style fix in fs/s... |
116 |
remove_wait_queue(entry->wait_address, &entry->wait); |
70674f95c [PATCH] Optimize ... |
117 118 |
fput(entry->filp); } |
1da177e4c Linux-2.6.12-rc2 |
119 120 121 |
void poll_freewait(struct poll_wqueues *pwq) { struct poll_table_page * p = pwq->table; |
70674f95c [PATCH] Optimize ... |
122 123 124 |
int i; for (i = 0; i < pwq->inline_index; i++) free_poll_entry(pwq->inline_entries + i); |
1da177e4c Linux-2.6.12-rc2 |
125 126 127 128 129 130 131 |
while (p) { struct poll_table_entry * entry; struct poll_table_page *old; entry = p->entry; do { entry--; |
70674f95c [PATCH] Optimize ... |
132 |
free_poll_entry(entry); |
1da177e4c Linux-2.6.12-rc2 |
133 134 135 136 137 138 |
} while (entry > p->entries); old = p; p = p->next; free_page((unsigned long) old); } } |
1da177e4c Linux-2.6.12-rc2 |
139 |
EXPORT_SYMBOL(poll_freewait); |
5f820f648 poll: allow f_op-... |
140 |
static struct poll_table_entry *poll_get_entry(struct poll_wqueues *p) |
1da177e4c Linux-2.6.12-rc2 |
141 |
{ |
1da177e4c Linux-2.6.12-rc2 |
142 |
struct poll_table_page *table = p->table; |
70674f95c [PATCH] Optimize ... |
143 144 |
if (p->inline_index < N_INLINE_POLL_ENTRIES) return p->inline_entries + p->inline_index++; |
1da177e4c Linux-2.6.12-rc2 |
145 146 147 148 149 150 |
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 ... |
151 |
return NULL; |
1da177e4c Linux-2.6.12-rc2 |
152 153 154 155 156 157 |
} new_table->entry = new_table->entries; new_table->next = table; p->table = new_table; table = new_table; } |
70674f95c [PATCH] Optimize ... |
158 159 |
return table->entry++; } |
4938d7e02 poll: avoid extra... |
160 |
static int __pollwake(wait_queue_t *wait, unsigned mode, int sync, void *key) |
5f820f648 poll: allow f_op-... |
161 162 163 164 165 166 167 168 169 |
{ 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() |
b92b8b35a locking/arch: Ren... |
170 |
* and is paired with smp_store_mb() in poll_schedule_timeout. |
5f820f648 poll: allow f_op-... |
171 172 173 174 175 176 177 178 179 180 181 182 183 184 |
*/ 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... |
185 186 187 188 189 190 191 192 193 |
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 ... |
194 195 196 197 |
/* Add a new entry */ static void __pollwait(struct file *filp, wait_queue_head_t *wait_address, poll_table *p) { |
5f820f648 poll: allow f_op-... |
198 199 |
struct poll_wqueues *pwq = container_of(p, struct poll_wqueues, pt); struct poll_table_entry *entry = poll_get_entry(pwq); |
70674f95c [PATCH] Optimize ... |
200 201 |
if (!entry) return; |
cb0942b81 make get_file() r... |
202 |
entry->filp = get_file(filp); |
70674f95c [PATCH] Optimize ... |
203 |
entry->wait_address = wait_address; |
626cf2366 poll: add poll_re... |
204 |
entry->key = p->_key; |
5f820f648 poll: allow f_op-... |
205 206 |
init_waitqueue_func_entry(&entry->wait, pollwake); entry->wait.private = pwq; |
ccf6780dc Style fix in fs/s... |
207 |
add_wait_queue(wait_address, &entry->wait); |
1da177e4c Linux-2.6.12-rc2 |
208 |
} |
5f820f648 poll: allow f_op-... |
209 210 211 212 213 214 215 |
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 Revert "select: u... |
216 |
rc = schedule_hrtimeout_range(expires, slack, HRTIMER_MODE_ABS); |
5f820f648 poll: allow f_op-... |
217 218 219 220 221 |
__set_current_state(TASK_RUNNING); /* * Prepare for the next iteration. * |
b92b8b35a locking/arch: Ren... |
222 |
* The following smp_store_mb() serves two purposes. First, it's |
5f820f648 poll: allow f_op-... |
223 224 225 226 227 228 229 |
* 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. */ |
b92b8b35a locking/arch: Ren... |
230 |
smp_store_mb(pwq->triggered, 0); |
5f820f648 poll: allow f_op-... |
231 232 233 234 |
return rc; } EXPORT_SYMBOL(poll_schedule_timeout); |
b773ad40a select: add poll_... |
235 236 |
/** * poll_select_set_timeout - helper function to setup the timeout value |
766b9f928 fs: poll/select/r... |
237 |
* @to: pointer to timespec64 variable for the final timeout |
b773ad40a select: add poll_... |
238 239 240 241 242 243 244 245 |
* @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. */ |
766b9f928 fs: poll/select/r... |
246 |
int poll_select_set_timeout(struct timespec64 *to, time64_t sec, long nsec) |
b773ad40a select: add poll_... |
247 |
{ |
766b9f928 fs: poll/select/r... |
248 |
struct timespec64 ts = {.tv_sec = sec, .tv_nsec = nsec}; |
b773ad40a select: add poll_... |
249 |
|
766b9f928 fs: poll/select/r... |
250 |
if (!timespec64_valid(&ts)) |
b773ad40a select: add poll_... |
251 252 253 254 255 256 |
return -EINVAL; /* Optimize for the zero timeout value here */ if (!sec && !nsec) { to->tv_sec = to->tv_nsec = 0; } else { |
766b9f928 fs: poll/select/r... |
257 258 |
ktime_get_ts64(to); *to = timespec64_add_safe(*to, ts); |
b773ad40a select: add poll_... |
259 260 261 |
} return 0; } |
766b9f928 fs: poll/select/r... |
262 263 |
static int poll_select_copy_remaining(struct timespec64 *end_time, void __user *p, |
b773ad40a select: add poll_... |
264 265 |
int timeval, int ret) { |
766b9f928 fs: poll/select/r... |
266 |
struct timespec64 rts64; |
b773ad40a select: add poll_... |
267 268 269 270 271 272 273 274 275 276 277 278 |
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; |
766b9f928 fs: poll/select/r... |
279 280 281 282 283 284 |
ktime_get_ts64(&rts64); rts64 = timespec64_sub(*end_time, rts64); if (rts64.tv_sec < 0) rts64.tv_sec = rts64.tv_nsec = 0; rts = timespec64_to_timespec(rts64); |
b773ad40a select: add poll_... |
285 286 |
if (timeval) { |
65329bf46 fs/select.c: fix ... |
287 288 |
if (sizeof(rtv) > sizeof(rtv.tv_sec) + sizeof(rtv.tv_usec)) memset(&rtv, 0, sizeof(rtv)); |
766b9f928 fs: poll/select/r... |
289 290 |
rtv.tv_sec = rts64.tv_sec; rtv.tv_usec = rts64.tv_nsec / NSEC_PER_USEC; |
b773ad40a select: add poll_... |
291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 |
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 |
311 312 313 314 315 316 317 318 319 320 321 |
#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... |
322 |
struct fdtable *fdt; |
1da177e4c Linux-2.6.12-rc2 |
323 324 |
/* handle last in-complete long-word first */ |
8ded2bbc1 posix_types.h: Cl... |
325 326 |
set = ~(~0UL << (n & (BITS_PER_LONG-1))); n /= BITS_PER_LONG; |
badf16621 [PATCH] files: br... |
327 |
fdt = files_fdtable(current->files); |
1fd36adcd Replace the fd_se... |
328 |
open_fds = fdt->open_fds + n; |
1da177e4c Linux-2.6.12-rc2 |
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); |
8ded2bbc1 posix_types.h: Cl... |
353 |
max += n * BITS_PER_LONG; |
1da177e4c Linux-2.6.12-rc2 |
354 355 356 357 |
} return max; } |
1da177e4c Linux-2.6.12-rc2 |
358 359 360 |
#define POLLIN_SET (POLLRDNORM | POLLRDBAND | POLLIN | POLLHUP | POLLERR) #define POLLOUT_SET (POLLWRBAND | POLLWRNORM | POLLOUT | POLLERR) #define POLLEX_SET (POLLPRI) |
4938d7e02 poll: avoid extra... |
361 |
static inline void wait_key_set(poll_table *wait, unsigned long in, |
2d48d67fa net: poll/select ... |
362 363 |
unsigned long out, unsigned long bit, unsigned int ll_flag) |
4938d7e02 poll: avoid extra... |
364 |
{ |
2d48d67fa net: poll/select ... |
365 |
wait->_key = POLLEX_SET | ll_flag; |
626cf2366 poll: add poll_re... |
366 367 368 369 |
if (in & bit) wait->_key |= POLLIN_SET; if (out & bit) wait->_key |= POLLOUT_SET; |
4938d7e02 poll: avoid extra... |
370 |
} |
766b9f928 fs: poll/select/r... |
371 |
int do_select(int n, fd_set_bits *fds, struct timespec64 *end_time) |
1da177e4c Linux-2.6.12-rc2 |
372 |
{ |
8ff3e8e85 select: switch se... |
373 |
ktime_t expire, *to = NULL; |
1da177e4c Linux-2.6.12-rc2 |
374 375 |
struct poll_wqueues table; poll_table *wait; |
8ff3e8e85 select: switch se... |
376 |
int retval, i, timed_out = 0; |
da8b44d5a timer: convert ti... |
377 |
u64 slack = 0; |
cbf55001b net: rename low l... |
378 |
unsigned int busy_flag = net_busy_loop_on() ? POLL_BUSY_LOOP : 0; |
76b1e9b98 net/fs: change bu... |
379 |
unsigned long busy_end = 0; |
1da177e4c Linux-2.6.12-rc2 |
380 |
|
b835996f6 [PATCH] files: lo... |
381 |
rcu_read_lock(); |
1da177e4c Linux-2.6.12-rc2 |
382 |
retval = max_select_fd(n, fds); |
b835996f6 [PATCH] files: lo... |
383 |
rcu_read_unlock(); |
1da177e4c Linux-2.6.12-rc2 |
384 385 386 387 388 389 390 |
if (retval < 0) return retval; n = retval; poll_initwait(&table); wait = &table.pt; |
8ff3e8e85 select: switch se... |
391 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
626cf2366 poll: add poll_re... |
392 |
wait->_qproc = NULL; |
8ff3e8e85 select: switch se... |
393 394 |
timed_out = 1; } |
96d2ab484 hrtimer: fix sign... |
395 |
if (end_time && !timed_out) |
231f3d393 select: rename es... |
396 |
slack = select_estimate_accuracy(end_time); |
90d6e24a3 hrtimer: make sel... |
397 |
|
1da177e4c Linux-2.6.12-rc2 |
398 399 400 |
retval = 0; for (;;) { unsigned long *rinp, *routp, *rexp, *inp, *outp, *exp; |
cbf55001b net: rename low l... |
401 |
bool can_busy_loop = false; |
1da177e4c Linux-2.6.12-rc2 |
402 |
|
1da177e4c Linux-2.6.12-rc2 |
403 404 405 406 407 408 |
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 Linux-2.6.12-rc2 |
409 410 411 412 |
in = *inp++; out = *outp++; ex = *exp++; all_bits = in | out | ex; if (all_bits == 0) { |
8ded2bbc1 posix_types.h: Cl... |
413 |
i += BITS_PER_LONG; |
1da177e4c Linux-2.6.12-rc2 |
414 415 |
continue; } |
8ded2bbc1 posix_types.h: Cl... |
416 |
for (j = 0; j < BITS_PER_LONG; ++j, ++i, bit <<= 1) { |
2903ff019 switch simple cas... |
417 |
struct fd f; |
1da177e4c Linux-2.6.12-rc2 |
418 419 420 421 |
if (i >= n) break; if (!(bit & all_bits)) continue; |
2903ff019 switch simple cas... |
422 423 424 425 |
f = fdget(i); if (f.file) { const struct file_operations *f_op; f_op = f.file->f_op; |
1da177e4c Linux-2.6.12-rc2 |
426 |
mask = DEFAULT_POLLMASK; |
72c2d5319 file->f_op is nev... |
427 |
if (f_op->poll) { |
2d48d67fa net: poll/select ... |
428 |
wait_key_set(wait, in, out, |
cbf55001b net: rename low l... |
429 |
bit, busy_flag); |
2903ff019 switch simple cas... |
430 |
mask = (*f_op->poll)(f.file, wait); |
4938d7e02 poll: avoid extra... |
431 |
} |
2903ff019 switch simple cas... |
432 |
fdput(f); |
1da177e4c Linux-2.6.12-rc2 |
433 434 435 |
if ((mask & POLLIN_SET) && (in & bit)) { res_in |= bit; retval++; |
626cf2366 poll: add poll_re... |
436 |
wait->_qproc = NULL; |
1da177e4c Linux-2.6.12-rc2 |
437 438 439 440 |
} if ((mask & POLLOUT_SET) && (out & bit)) { res_out |= bit; retval++; |
626cf2366 poll: add poll_re... |
441 |
wait->_qproc = NULL; |
1da177e4c Linux-2.6.12-rc2 |
442 443 444 445 |
} if ((mask & POLLEX_SET) && (ex & bit)) { res_ex |= bit; retval++; |
626cf2366 poll: add poll_re... |
446 |
wait->_qproc = NULL; |
1da177e4c Linux-2.6.12-rc2 |
447 |
} |
2d48d67fa net: poll/select ... |
448 |
/* got something, stop busy polling */ |
cbf55001b net: rename low l... |
449 450 451 452 453 454 455 456 457 458 |
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 Linux-2.6.12-rc2 |
459 |
} |
1da177e4c Linux-2.6.12-rc2 |
460 461 462 463 464 465 466 |
} if (res_in) *rinp = res_in; if (res_out) *routp = res_out; if (res_ex) *rexp = res_ex; |
55d853849 Fix performance r... |
467 |
cond_resched(); |
1da177e4c Linux-2.6.12-rc2 |
468 |
} |
626cf2366 poll: add poll_re... |
469 |
wait->_qproc = NULL; |
8ff3e8e85 select: switch se... |
470 |
if (retval || timed_out || signal_pending(current)) |
1da177e4c Linux-2.6.12-rc2 |
471 |
break; |
f5264481c trivial: small cl... |
472 |
if (table.error) { |
1da177e4c Linux-2.6.12-rc2 |
473 474 475 |
retval = table.error; break; } |
9f72949f6 [PATCH] Add psele... |
476 |
|
cbf55001b net: rename low l... |
477 |
/* only if found POLL_BUSY_LOOP sockets && not out of time */ |
76b1e9b98 net/fs: change bu... |
478 479 480 481 482 483 484 485 486 |
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 net: poll/select ... |
487 |
|
8ff3e8e85 select: switch se... |
488 489 490 491 492 493 |
/* * 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) { |
766b9f928 fs: poll/select/r... |
494 |
expire = timespec64_to_ktime(*end_time); |
8ff3e8e85 select: switch se... |
495 |
to = &expire; |
9f72949f6 [PATCH] Add psele... |
496 |
} |
8ff3e8e85 select: switch se... |
497 |
|
5f820f648 poll: allow f_op-... |
498 499 |
if (!poll_schedule_timeout(&table, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85 select: switch se... |
500 |
timed_out = 1; |
1da177e4c Linux-2.6.12-rc2 |
501 |
} |
1da177e4c Linux-2.6.12-rc2 |
502 503 |
poll_freewait(&table); |
1da177e4c Linux-2.6.12-rc2 |
504 505 |
return retval; } |
1da177e4c Linux-2.6.12-rc2 |
506 507 508 509 510 511 512 513 |
/* * 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 [PATCH] make osf_... |
514 |
int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, |
766b9f928 fs: poll/select/r... |
515 |
fd_set __user *exp, struct timespec64 *end_time) |
1da177e4c Linux-2.6.12-rc2 |
516 517 |
{ fd_set_bits fds; |
29ff2db55 [PATCH] select() ... |
518 |
void *bits; |
bbea9f696 [PATCH] fdtable: ... |
519 |
int ret, max_fds; |
b04eb6aa0 [PATCH] select: d... |
520 |
unsigned int size; |
badf16621 [PATCH] files: br... |
521 |
struct fdtable *fdt; |
70674f95c [PATCH] Optimize ... |
522 |
/* Allocate small arguments on the stack to save memory and be faster */ |
30c14e40e [PATCH] avoid una... |
523 |
long stack_fds[SELECT_STACK_ALLOC/sizeof(long)]; |
1da177e4c Linux-2.6.12-rc2 |
524 |
|
1da177e4c Linux-2.6.12-rc2 |
525 526 527 |
ret = -EINVAL; if (n < 0) goto out_nofds; |
bbea9f696 [PATCH] fdtable: ... |
528 |
/* max_fds can increase, so grab it once to avoid race */ |
b835996f6 [PATCH] files: lo... |
529 |
rcu_read_lock(); |
badf16621 [PATCH] files: br... |
530 |
fdt = files_fdtable(current->files); |
bbea9f696 [PATCH] fdtable: ... |
531 |
max_fds = fdt->max_fds; |
b835996f6 [PATCH] files: lo... |
532 |
rcu_read_unlock(); |
bbea9f696 [PATCH] fdtable: ... |
533 534 |
if (n > max_fds) n = max_fds; |
1da177e4c Linux-2.6.12-rc2 |
535 536 537 538 539 540 |
/* * 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 |
541 |
size = FDS_BYTES(n); |
b04eb6aa0 [PATCH] select: d... |
542 543 544 545 |
bits = stack_fds; if (size > sizeof(stack_fds) / 6) { /* Not enough space in on-stack array; must use kmalloc */ ret = -ENOMEM; |
70674f95c [PATCH] Optimize ... |
546 |
bits = kmalloc(6 * size, GFP_KERNEL); |
b04eb6aa0 [PATCH] select: d... |
547 548 549 |
if (!bits) goto out_nofds; } |
29ff2db55 [PATCH] select() ... |
550 551 552 553 554 555 |
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 |
556 557 558 559 560 561 562 563 |
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... |
564 |
ret = do_select(n, &fds, end_time); |
1da177e4c Linux-2.6.12-rc2 |
565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 |
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 ... |
581 582 |
if (bits != stack_fds) kfree(bits); |
1da177e4c Linux-2.6.12-rc2 |
583 584 585 |
out_nofds: return ret; } |
5a8a82b1d [CVE-2009-0029] S... |
586 587 |
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... |
588 |
{ |
766b9f928 fs: poll/select/r... |
589 |
struct timespec64 end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
590 591 592 593 594 595 |
struct timeval tv; int ret; if (tvp) { if (copy_from_user(&tv, tvp, sizeof(tv))) return -EFAULT; |
8ff3e8e85 select: switch se... |
596 |
to = &end_time; |
4d36a9e65 select: deal with... |
597 598 599 |
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... |
600 |
return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
601 |
} |
8ff3e8e85 select: switch se... |
602 603 |
ret = core_sys_select(n, inp, outp, exp, to); ret = poll_select_copy_remaining(&end_time, tvp, 1, ret); |
9f72949f6 [PATCH] Add psele... |
604 605 606 |
return ret; } |
c9da9f212 [CVE-2009-0029] M... |
607 608 609 |
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... |
610 |
{ |
9f72949f6 [PATCH] Add psele... |
611 |
sigset_t ksigmask, sigsaved; |
766b9f928 fs: poll/select/r... |
612 613 |
struct timespec ts; struct timespec64 ts64, end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
614 615 616 617 618 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
766b9f928 fs: poll/select/r... |
619 |
ts64 = timespec_to_timespec64(ts); |
9f72949f6 [PATCH] Add psele... |
620 |
|
8ff3e8e85 select: switch se... |
621 |
to = &end_time; |
766b9f928 fs: poll/select/r... |
622 |
if (poll_select_set_timeout(to, ts64.tv_sec, ts64.tv_nsec)) |
9f72949f6 [PATCH] Add psele... |
623 |
return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
624 625 626 627 628 629 630 631 632 633 634 635 |
} 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... |
636 |
ret = core_sys_select(n, inp, outp, exp, to); |
8ff3e8e85 select: switch se... |
637 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6 [PATCH] Add psele... |
638 639 640 641 642 643 644 645 646 647 |
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_... |
648 |
set_restore_sigmask(); |
9f72949f6 [PATCH] Add psele... |
649 650 651 652 653 654 655 656 657 658 659 660 661 |
} } 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... |
662 663 664 |
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... |
665 666 667 668 669 670 |
{ 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... |
671 |
|| __get_user(up, (sigset_t __user * __user *)sig) |
9f72949f6 [PATCH] Add psele... |
672 |
|| __get_user(sigsetsize, |
e110ab94e [PATCH] fix __use... |
673 |
(size_t __user *)(sig+sizeof(void *)))) |
9f72949f6 [PATCH] Add psele... |
674 675 |
return -EFAULT; } |
c9da9f212 [CVE-2009-0029] M... |
676 |
return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize); |
9f72949f6 [PATCH] Add psele... |
677 |
} |
9f72949f6 [PATCH] Add psele... |
678 |
|
5d0e52830 Add generic sys_o... |
679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 |
#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 |
695 696 697 698 699 700 701 |
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... |
702 703 704 705 706 |
/* * 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 poll: add poll_re... |
707 |
* if pwait->_qproc is non-NULL. |
4a4b69f79 [PATCH] Poll clea... |
708 |
*/ |
2d48d67fa net: poll/select ... |
709 |
static inline unsigned int do_pollfd(struct pollfd *pollfd, poll_table *pwait, |
cbf55001b net: rename low l... |
710 711 |
bool *can_busy_poll, unsigned int busy_flag) |
1da177e4c Linux-2.6.12-rc2 |
712 |
{ |
4a4b69f79 [PATCH] Poll clea... |
713 714 715 716 717 718 |
unsigned int mask; int fd; mask = 0; fd = pollfd->fd; if (fd >= 0) { |
2903ff019 switch simple cas... |
719 |
struct fd f = fdget(fd); |
4a4b69f79 [PATCH] Poll clea... |
720 |
mask = POLLNVAL; |
2903ff019 switch simple cas... |
721 |
if (f.file) { |
4a4b69f79 [PATCH] Poll clea... |
722 |
mask = DEFAULT_POLLMASK; |
72c2d5319 file->f_op is nev... |
723 |
if (f.file->f_op->poll) { |
626cf2366 poll: add poll_re... |
724 |
pwait->_key = pollfd->events|POLLERR|POLLHUP; |
cbf55001b net: rename low l... |
725 |
pwait->_key |= busy_flag; |
2903ff019 switch simple cas... |
726 |
mask = f.file->f_op->poll(f.file, pwait); |
cbf55001b net: rename low l... |
727 728 |
if (mask & busy_flag) *can_busy_poll = true; |
4938d7e02 poll: avoid extra... |
729 |
} |
4a4b69f79 [PATCH] Poll clea... |
730 731 |
/* Mask out unneeded events. */ mask &= pollfd->events | POLLERR | POLLHUP; |
2903ff019 switch simple cas... |
732 |
fdput(f); |
1da177e4c Linux-2.6.12-rc2 |
733 |
} |
1da177e4c Linux-2.6.12-rc2 |
734 |
} |
4a4b69f79 [PATCH] Poll clea... |
735 736 737 |
pollfd->revents = mask; return mask; |
1da177e4c Linux-2.6.12-rc2 |
738 |
} |
ccec5ee30 poll: plug an unu... |
739 |
static int do_poll(struct poll_list *list, struct poll_wqueues *wait, |
766b9f928 fs: poll/select/r... |
740 |
struct timespec64 *end_time) |
1da177e4c Linux-2.6.12-rc2 |
741 |
{ |
1da177e4c Linux-2.6.12-rc2 |
742 |
poll_table* pt = &wait->pt; |
8ff3e8e85 select: switch se... |
743 744 |
ktime_t expire, *to = NULL; int timed_out = 0, count = 0; |
da8b44d5a timer: convert ti... |
745 |
u64 slack = 0; |
cbf55001b net: rename low l... |
746 |
unsigned int busy_flag = net_busy_loop_on() ? POLL_BUSY_LOOP : 0; |
76b1e9b98 net/fs: change bu... |
747 |
unsigned long busy_end = 0; |
1da177e4c Linux-2.6.12-rc2 |
748 |
|
9f72949f6 [PATCH] Add psele... |
749 |
/* Optimise the no-wait case */ |
8ff3e8e85 select: switch se... |
750 |
if (end_time && !end_time->tv_sec && !end_time->tv_nsec) { |
626cf2366 poll: add poll_re... |
751 |
pt->_qproc = NULL; |
8ff3e8e85 select: switch se... |
752 753 |
timed_out = 1; } |
9bf084f70 do_poll: return -... |
754 |
|
96d2ab484 hrtimer: fix sign... |
755 |
if (end_time && !timed_out) |
231f3d393 select: rename es... |
756 |
slack = select_estimate_accuracy(end_time); |
90d6e24a3 hrtimer: make sel... |
757 |
|
1da177e4c Linux-2.6.12-rc2 |
758 759 |
for (;;) { struct poll_list *walk; |
cbf55001b net: rename low l... |
760 |
bool can_busy_loop = false; |
9f72949f6 [PATCH] Add psele... |
761 |
|
4a4b69f79 [PATCH] Poll clea... |
762 763 764 765 766 767 768 769 |
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 poll: add poll_re... |
770 |
* and kill poll_table->_qproc, so we don't |
4a4b69f79 [PATCH] Poll clea... |
771 772 773 774 |
* needlessly register any other waiters after * this. They'll get immediately deregistered * when we break out and return. */ |
cbf55001b net: rename low l... |
775 776 |
if (do_pollfd(pfd, pt, &can_busy_loop, busy_flag)) { |
4a4b69f79 [PATCH] Poll clea... |
777 |
count++; |
626cf2366 poll: add poll_re... |
778 |
pt->_qproc = NULL; |
cbf55001b net: rename low l... |
779 780 781 |
/* found something, stop busy polling */ busy_flag = 0; can_busy_loop = false; |
4a4b69f79 [PATCH] Poll clea... |
782 783 |
} } |
1da177e4c Linux-2.6.12-rc2 |
784 |
} |
4a4b69f79 [PATCH] Poll clea... |
785 786 |
/* * All waiters have already been registered, so don't provide |
626cf2366 poll: add poll_re... |
787 |
* a poll_table->_qproc to them on the next loop iteration. |
4a4b69f79 [PATCH] Poll clea... |
788 |
*/ |
626cf2366 poll: add poll_re... |
789 |
pt->_qproc = NULL; |
9bf084f70 do_poll: return -... |
790 791 792 793 794 |
if (!count) { count = wait->error; if (signal_pending(current)) count = -EINTR; } |
8ff3e8e85 select: switch se... |
795 |
if (count || timed_out) |
1da177e4c Linux-2.6.12-rc2 |
796 |
break; |
9f72949f6 [PATCH] Add psele... |
797 |
|
cbf55001b net: rename low l... |
798 |
/* only if found POLL_BUSY_LOOP sockets && not out of time */ |
76b1e9b98 net/fs: change bu... |
799 800 801 802 803 804 805 806 807 |
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 net: avoid callin... |
808 |
|
8ff3e8e85 select: switch se... |
809 810 811 812 813 814 |
/* * 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) { |
766b9f928 fs: poll/select/r... |
815 |
expire = timespec64_to_ktime(*end_time); |
8ff3e8e85 select: switch se... |
816 |
to = &expire; |
9f72949f6 [PATCH] Add psele... |
817 |
} |
5f820f648 poll: allow f_op-... |
818 |
if (!poll_schedule_timeout(wait, TASK_INTERRUPTIBLE, to, slack)) |
8ff3e8e85 select: switch se... |
819 |
timed_out = 1; |
1da177e4c Linux-2.6.12-rc2 |
820 |
} |
1da177e4c Linux-2.6.12-rc2 |
821 822 |
return count; } |
70674f95c [PATCH] Optimize ... |
823 824 |
#define N_STACK_PPS ((sizeof(stack_pps) - sizeof(struct poll_list)) / \ sizeof(struct pollfd)) |
8ff3e8e85 select: switch se... |
825 |
int do_sys_poll(struct pollfd __user *ufds, unsigned int nfds, |
766b9f928 fs: poll/select/r... |
826 |
struct timespec64 *end_time) |
1da177e4c Linux-2.6.12-rc2 |
827 828 |
{ struct poll_wqueues table; |
252e5725c do_sys_poll: simp... |
829 |
int err = -EFAULT, fdcount, len, size; |
30c14e40e [PATCH] avoid una... |
830 831 832 833 |
/* 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... |
834 835 836 |
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 |
837 |
|
d554ed895 fs: use rlimit he... |
838 |
if (nfds > rlimit(RLIMIT_NOFILE)) |
1da177e4c Linux-2.6.12-rc2 |
839 |
return -EINVAL; |
252e5725c do_sys_poll: simp... |
840 841 842 843 844 845 |
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 |
846 |
|
252e5725c do_sys_poll: simp... |
847 848 849 850 851 852 853 |
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 |
854 |
|
252e5725c do_sys_poll: simp... |
855 856 857 858 859 |
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 |
860 861 |
goto out_fds; } |
1da177e4c Linux-2.6.12-rc2 |
862 |
} |
9f72949f6 [PATCH] Add psele... |
863 |
|
252e5725c do_sys_poll: simp... |
864 |
poll_initwait(&table); |
ccec5ee30 poll: plug an unu... |
865 |
fdcount = do_poll(head, &table, end_time); |
252e5725c do_sys_poll: simp... |
866 |
poll_freewait(&table); |
1da177e4c Linux-2.6.12-rc2 |
867 |
|
252e5725c do_sys_poll: simp... |
868 |
for (walk = head; walk; walk = walk->next) { |
1da177e4c Linux-2.6.12-rc2 |
869 870 |
struct pollfd *fds = walk->entries; int j; |
252e5725c do_sys_poll: simp... |
871 872 |
for (j = 0; j < walk->len; j++, ufds++) if (__put_user(fds[j].revents, &ufds->revents)) |
1da177e4c Linux-2.6.12-rc2 |
873 |
goto out_fds; |
1da177e4c Linux-2.6.12-rc2 |
874 |
} |
252e5725c do_sys_poll: simp... |
875 |
|
1da177e4c Linux-2.6.12-rc2 |
876 |
err = fdcount; |
1da177e4c Linux-2.6.12-rc2 |
877 |
out_fds: |
252e5725c do_sys_poll: simp... |
878 879 880 881 882 |
walk = head->next; while (walk) { struct poll_list *pos = walk; walk = walk->next; kfree(pos); |
1da177e4c Linux-2.6.12-rc2 |
883 |
} |
252e5725c do_sys_poll: simp... |
884 |
|
1da177e4c Linux-2.6.12-rc2 |
885 886 |
return err; } |
9f72949f6 [PATCH] Add psele... |
887 |
|
3075d9da0 Use ERESTART_REST... |
888 889 |
static long do_restart_poll(struct restart_block *restart_block) { |
8ff3e8e85 select: switch se... |
890 891 |
struct pollfd __user *ufds = restart_block->poll.ufds; int nfds = restart_block->poll.nfds; |
766b9f928 fs: poll/select/r... |
892 |
struct timespec64 *to = NULL, end_time; |
3075d9da0 Use ERESTART_REST... |
893 |
int ret; |
8ff3e8e85 select: switch se... |
894 895 896 897 898 899 900 |
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... |
901 902 |
if (ret == -EINTR) { restart_block->fn = do_restart_poll; |
3075d9da0 Use ERESTART_REST... |
903 904 905 906 |
ret = -ERESTART_RESTARTBLOCK; } return ret; } |
5a8a82b1d [CVE-2009-0029] S... |
907 |
SYSCALL_DEFINE3(poll, struct pollfd __user *, ufds, unsigned int, nfds, |
faf309009 sys_poll: fix inc... |
908 |
int, timeout_msecs) |
9f72949f6 [PATCH] Add psele... |
909 |
{ |
766b9f928 fs: poll/select/r... |
910 |
struct timespec64 end_time, *to = NULL; |
3075d9da0 Use ERESTART_REST... |
911 |
int ret; |
9f72949f6 [PATCH] Add psele... |
912 |
|
8ff3e8e85 select: switch se... |
913 914 915 916 |
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... |
917 |
} |
8ff3e8e85 select: switch se... |
918 |
ret = do_sys_poll(ufds, nfds, to); |
3075d9da0 Use ERESTART_REST... |
919 920 |
if (ret == -EINTR) { struct restart_block *restart_block; |
8ff3e8e85 select: switch se... |
921 |
|
f56141e3e all arches, signa... |
922 |
restart_block = ¤t->restart_block; |
3075d9da0 Use ERESTART_REST... |
923 |
restart_block->fn = do_restart_poll; |
8ff3e8e85 select: switch se... |
924 925 926 927 928 929 930 931 932 |
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... |
933 934 935 |
ret = -ERESTART_RESTARTBLOCK; } return ret; |
9f72949f6 [PATCH] Add psele... |
936 |
} |
d4e82042c [CVE-2009-0029] S... |
937 938 939 |
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... |
940 941 |
{ sigset_t ksigmask, sigsaved; |
766b9f928 fs: poll/select/r... |
942 943 |
struct timespec ts; struct timespec64 end_time, *to = NULL; |
9f72949f6 [PATCH] Add psele... |
944 945 946 947 948 |
int ret; if (tsp) { if (copy_from_user(&ts, tsp, sizeof(ts))) return -EFAULT; |
8ff3e8e85 select: switch se... |
949 950 951 |
to = &end_time; if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec)) return -EINVAL; |
9f72949f6 [PATCH] Add psele... |
952 953 954 955 956 957 958 959 960 961 962 963 |
} 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... |
964 |
ret = do_sys_poll(ufds, nfds, to); |
9f72949f6 [PATCH] Add psele... |
965 966 967 968 969 970 971 972 973 974 975 |
/* 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_... |
976 |
set_restore_sigmask(); |
9f72949f6 [PATCH] Add psele... |
977 978 979 980 |
} ret = -ERESTARTNOHAND; } else if (sigmask) sigprocmask(SIG_SETMASK, &sigsaved, NULL); |
8ff3e8e85 select: switch se... |
981 |
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret); |
9f72949f6 [PATCH] Add psele... |
982 983 984 |
return ret; } |