Commit | Line | Data |
---|---|---|
2e1df07d JA |
1 | /* |
2 | * fio - the flexible io tester | |
3 | * | |
4 | * Copyright (C) 2005 Jens Axboe <axboe@suse.de> | |
5 | * Copyright (C) 2006-2012 Jens Axboe <axboe@kernel.dk> | |
6 | * | |
7 | * The license below covers all files distributed with fio unless otherwise | |
8 | * noted in the file itself. | |
9 | * | |
10 | * This program is free software; you can redistribute it and/or modify | |
11 | * it under the terms of the GNU General Public License version 2 as | |
12 | * published by the Free Software Foundation. | |
13 | * | |
14 | * This program is distributed in the hope that it will be useful, | |
15 | * but WITHOUT ANY WARRANTY; without even the implied warranty of | |
16 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
17 | * GNU General Public License for more details. | |
18 | * | |
19 | * You should have received a copy of the GNU General Public License | |
20 | * along with this program; if not, write to the Free Software | |
21 | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | |
22 | * | |
23 | */ | |
24 | #include <unistd.h> | |
25 | #include <fcntl.h> | |
26 | #include <string.h> | |
27 | #include <limits.h> | |
28 | #include <signal.h> | |
29 | #include <time.h> | |
30 | #include <locale.h> | |
31 | #include <assert.h> | |
32 | #include <time.h> | |
e43606c2 | 33 | #include <inttypes.h> |
2e1df07d JA |
34 | #include <sys/stat.h> |
35 | #include <sys/wait.h> | |
36 | #include <sys/ipc.h> | |
2e1df07d JA |
37 | #include <sys/mman.h> |
38 | ||
39 | #include "fio.h" | |
a5e0ee11 O |
40 | #ifndef FIO_NO_HAVE_SHM_H |
41 | #include <sys/shm.h> | |
42 | #endif | |
2e1df07d JA |
43 | #include "hash.h" |
44 | #include "smalloc.h" | |
45 | #include "verify.h" | |
46 | #include "trim.h" | |
47 | #include "diskutil.h" | |
48 | #include "cgroup.h" | |
49 | #include "profile.h" | |
50 | #include "lib/rand.h" | |
51 | #include "memalign.h" | |
52 | #include "server.h" | |
44404c5a | 53 | #include "lib/getrusage.h" |
f2a2ce0e | 54 | #include "idletime.h" |
002fe734 | 55 | #include "err.h" |
78d55e72 | 56 | #include "lib/tp.h" |
2e1df07d | 57 | |
5ddc6707 JA |
58 | static pthread_t helper_thread; |
59 | static pthread_mutex_t helper_lock; | |
60 | pthread_cond_t helper_cond; | |
61 | int helper_do_stat = 0; | |
8aab824f | 62 | |
2e1df07d | 63 | static struct fio_mutex *startup_mutex; |
2e1df07d JA |
64 | static struct flist_head *cgroup_list; |
65 | static char *cgroup_mnt; | |
66 | static int exit_value; | |
67 | static volatile int fio_abort; | |
3a5f6bde JA |
68 | static unsigned int nr_process = 0; |
69 | static unsigned int nr_thread = 0; | |
2e1df07d | 70 | |
6eaf09d6 | 71 | struct io_log *agg_io_log[DDIR_RWDIR_CNT]; |
2e1df07d | 72 | |
a3efc919 JA |
73 | int groupid = 0; |
74 | unsigned int thread_number = 0; | |
108fea77 | 75 | unsigned int stat_number = 0; |
a3efc919 JA |
76 | int shm_id = 0; |
77 | int temp_stall_ts; | |
78 | unsigned long done_secs = 0; | |
5ddc6707 | 79 | volatile int helper_exit = 0; |
a3efc919 | 80 | |
2e1df07d | 81 | #define PAGE_ALIGN(buf) \ |
e43606c2 | 82 | (char *) (((uintptr_t) (buf) + page_mask) & ~page_mask) |
2e1df07d JA |
83 | |
84 | #define JOB_START_TIMEOUT (5 * 1000) | |
85 | ||
86 | static void sig_int(int sig) | |
87 | { | |
88 | if (threads) { | |
89 | if (is_backend) | |
90 | fio_server_got_signal(sig); | |
91 | else { | |
92 | log_info("\nfio: terminating on signal %d\n", sig); | |
e411c301 | 93 | log_info_flush(); |
2e1df07d JA |
94 | exit_value = 128; |
95 | } | |
96 | ||
97 | fio_terminate_threads(TERMINATE_ALL); | |
98 | } | |
99 | } | |
100 | ||
4c6d91e8 JA |
101 | static void sig_show_status(int sig) |
102 | { | |
103 | show_running_run_stats(); | |
104 | } | |
105 | ||
2e1df07d JA |
106 | static void set_sig_handlers(void) |
107 | { | |
108 | struct sigaction act; | |
109 | ||
110 | memset(&act, 0, sizeof(act)); | |
111 | act.sa_handler = sig_int; | |
112 | act.sa_flags = SA_RESTART; | |
113 | sigaction(SIGINT, &act, NULL); | |
114 | ||
115 | memset(&act, 0, sizeof(act)); | |
116 | act.sa_handler = sig_int; | |
117 | act.sa_flags = SA_RESTART; | |
118 | sigaction(SIGTERM, &act, NULL); | |
119 | ||
2f694507 BC |
120 | /* Windows uses SIGBREAK as a quit signal from other applications */ |
121 | #ifdef WIN32 | |
122 | memset(&act, 0, sizeof(act)); | |
123 | act.sa_handler = sig_int; | |
124 | act.sa_flags = SA_RESTART; | |
125 | sigaction(SIGBREAK, &act, NULL); | |
126 | #endif | |
127 | ||
4c6d91e8 JA |
128 | memset(&act, 0, sizeof(act)); |
129 | act.sa_handler = sig_show_status; | |
130 | act.sa_flags = SA_RESTART; | |
131 | sigaction(SIGUSR1, &act, NULL); | |
132 | ||
2e1df07d JA |
133 | if (is_backend) { |
134 | memset(&act, 0, sizeof(act)); | |
135 | act.sa_handler = sig_int; | |
136 | act.sa_flags = SA_RESTART; | |
137 | sigaction(SIGPIPE, &act, NULL); | |
138 | } | |
139 | } | |
140 | ||
141 | /* | |
142 | * Check if we are above the minimum rate given. | |
143 | */ | |
144 | static int __check_min_rate(struct thread_data *td, struct timeval *now, | |
145 | enum fio_ddir ddir) | |
146 | { | |
147 | unsigned long long bytes = 0; | |
148 | unsigned long iops = 0; | |
149 | unsigned long spent; | |
150 | unsigned long rate; | |
151 | unsigned int ratemin = 0; | |
152 | unsigned int rate_iops = 0; | |
153 | unsigned int rate_iops_min = 0; | |
154 | ||
155 | assert(ddir_rw(ddir)); | |
156 | ||
157 | if (!td->o.ratemin[ddir] && !td->o.rate_iops_min[ddir]) | |
158 | return 0; | |
159 | ||
160 | /* | |
161 | * allow a 2 second settle period in the beginning | |
162 | */ | |
163 | if (mtime_since(&td->start, now) < 2000) | |
164 | return 0; | |
165 | ||
166 | iops += td->this_io_blocks[ddir]; | |
167 | bytes += td->this_io_bytes[ddir]; | |
168 | ratemin += td->o.ratemin[ddir]; | |
169 | rate_iops += td->o.rate_iops[ddir]; | |
170 | rate_iops_min += td->o.rate_iops_min[ddir]; | |
171 | ||
172 | /* | |
173 | * if rate blocks is set, sample is running | |
174 | */ | |
175 | if (td->rate_bytes[ddir] || td->rate_blocks[ddir]) { | |
176 | spent = mtime_since(&td->lastrate[ddir], now); | |
177 | if (spent < td->o.ratecycle) | |
178 | return 0; | |
179 | ||
180 | if (td->o.rate[ddir]) { | |
181 | /* | |
182 | * check bandwidth specified rate | |
183 | */ | |
184 | if (bytes < td->rate_bytes[ddir]) { | |
185 | log_err("%s: min rate %u not met\n", td->o.name, | |
186 | ratemin); | |
187 | return 1; | |
188 | } else { | |
49cba9b3 JA |
189 | if (spent) |
190 | rate = ((bytes - td->rate_bytes[ddir]) * 1000) / spent; | |
191 | else | |
192 | rate = 0; | |
193 | ||
2e1df07d JA |
194 | if (rate < ratemin || |
195 | bytes < td->rate_bytes[ddir]) { | |
196 | log_err("%s: min rate %u not met, got" | |
197 | " %luKB/sec\n", td->o.name, | |
198 | ratemin, rate); | |
199 | return 1; | |
200 | } | |
201 | } | |
202 | } else { | |
203 | /* | |
204 | * checks iops specified rate | |
205 | */ | |
206 | if (iops < rate_iops) { | |
207 | log_err("%s: min iops rate %u not met\n", | |
208 | td->o.name, rate_iops); | |
209 | return 1; | |
210 | } else { | |
4c707a3b JA |
211 | if (spent) |
212 | rate = ((iops - td->rate_blocks[ddir]) * 1000) / spent; | |
213 | else | |
214 | rate = 0; | |
215 | ||
2e1df07d JA |
216 | if (rate < rate_iops_min || |
217 | iops < td->rate_blocks[ddir]) { | |
218 | log_err("%s: min iops rate %u not met," | |
219 | " got %lu\n", td->o.name, | |
220 | rate_iops_min, rate); | |
221 | } | |
222 | } | |
223 | } | |
224 | } | |
225 | ||
226 | td->rate_bytes[ddir] = bytes; | |
227 | td->rate_blocks[ddir] = iops; | |
228 | memcpy(&td->lastrate[ddir], now, sizeof(*now)); | |
229 | return 0; | |
230 | } | |
231 | ||
232 | static int check_min_rate(struct thread_data *td, struct timeval *now, | |
100f49f1 | 233 | uint64_t *bytes_done) |
2e1df07d JA |
234 | { |
235 | int ret = 0; | |
236 | ||
6eaf09d6 SL |
237 | if (bytes_done[DDIR_READ]) |
238 | ret |= __check_min_rate(td, now, DDIR_READ); | |
239 | if (bytes_done[DDIR_WRITE]) | |
240 | ret |= __check_min_rate(td, now, DDIR_WRITE); | |
241 | if (bytes_done[DDIR_TRIM]) | |
242 | ret |= __check_min_rate(td, now, DDIR_TRIM); | |
2e1df07d JA |
243 | |
244 | return ret; | |
245 | } | |
246 | ||
247 | /* | |
248 | * When job exits, we can cancel the in-flight IO if we are using async | |
249 | * io. Attempt to do so. | |
250 | */ | |
251 | static void cleanup_pending_aio(struct thread_data *td) | |
252 | { | |
2e1df07d JA |
253 | int r; |
254 | ||
255 | /* | |
256 | * get immediately available events, if any | |
257 | */ | |
258 | r = io_u_queued_complete(td, 0, NULL); | |
259 | if (r < 0) | |
260 | return; | |
261 | ||
262 | /* | |
263 | * now cancel remaining active events | |
264 | */ | |
265 | if (td->io_ops->cancel) { | |
2ae0b204 JA |
266 | struct io_u *io_u; |
267 | int i; | |
2e1df07d | 268 | |
2ae0b204 JA |
269 | io_u_qiter(&td->io_u_all, io_u, i) { |
270 | if (io_u->flags & IO_U_F_FLIGHT) { | |
2e1df07d JA |
271 | r = td->io_ops->cancel(td, io_u); |
272 | if (!r) | |
273 | put_io_u(td, io_u); | |
274 | } | |
275 | } | |
276 | } | |
277 | ||
278 | if (td->cur_depth) | |
279 | r = io_u_queued_complete(td, td->cur_depth, NULL); | |
280 | } | |
281 | ||
282 | /* | |
283 | * Helper to handle the final sync of a file. Works just like the normal | |
284 | * io path, just does everything sync. | |
285 | */ | |
286 | static int fio_io_sync(struct thread_data *td, struct fio_file *f) | |
287 | { | |
288 | struct io_u *io_u = __get_io_u(td); | |
289 | int ret; | |
290 | ||
291 | if (!io_u) | |
292 | return 1; | |
293 | ||
294 | io_u->ddir = DDIR_SYNC; | |
295 | io_u->file = f; | |
296 | ||
297 | if (td_io_prep(td, io_u)) { | |
298 | put_io_u(td, io_u); | |
299 | return 1; | |
300 | } | |
301 | ||
302 | requeue: | |
303 | ret = td_io_queue(td, io_u); | |
304 | if (ret < 0) { | |
305 | td_verror(td, io_u->error, "td_io_queue"); | |
306 | put_io_u(td, io_u); | |
307 | return 1; | |
308 | } else if (ret == FIO_Q_QUEUED) { | |
309 | if (io_u_queued_complete(td, 1, NULL) < 0) | |
310 | return 1; | |
311 | } else if (ret == FIO_Q_COMPLETED) { | |
312 | if (io_u->error) { | |
313 | td_verror(td, io_u->error, "td_io_queue"); | |
314 | return 1; | |
315 | } | |
316 | ||
317 | if (io_u_sync_complete(td, io_u, NULL) < 0) | |
318 | return 1; | |
319 | } else if (ret == FIO_Q_BUSY) { | |
320 | if (td_io_commit(td)) | |
321 | return 1; | |
322 | goto requeue; | |
323 | } | |
324 | ||
325 | return 0; | |
326 | } | |
a3efc919 | 327 | |
61ee0f86 JA |
328 | static int fio_file_fsync(struct thread_data *td, struct fio_file *f) |
329 | { | |
330 | int ret; | |
331 | ||
332 | if (fio_file_open(f)) | |
333 | return fio_io_sync(td, f); | |
334 | ||
335 | if (td_io_open_file(td, f)) | |
336 | return 1; | |
337 | ||
338 | ret = fio_io_sync(td, f); | |
339 | td_io_close_file(td, f); | |
340 | return ret; | |
341 | } | |
342 | ||
2e1df07d JA |
343 | static inline void __update_tv_cache(struct thread_data *td) |
344 | { | |
345 | fio_gettime(&td->tv_cache, NULL); | |
346 | } | |
347 | ||
348 | static inline void update_tv_cache(struct thread_data *td) | |
349 | { | |
350 | if ((++td->tv_cache_nr & td->tv_cache_mask) == td->tv_cache_mask) | |
351 | __update_tv_cache(td); | |
352 | } | |
353 | ||
354 | static inline int runtime_exceeded(struct thread_data *td, struct timeval *t) | |
355 | { | |
356 | if (in_ramp_time(td)) | |
357 | return 0; | |
358 | if (!td->o.timeout) | |
359 | return 0; | |
0de5b26f | 360 | if (utime_since(&td->epoch, t) >= td->o.timeout) |
2e1df07d JA |
361 | return 1; |
362 | ||
363 | return 0; | |
364 | } | |
365 | ||
366 | static int break_on_this_error(struct thread_data *td, enum fio_ddir ddir, | |
367 | int *retptr) | |
368 | { | |
369 | int ret = *retptr; | |
370 | ||
371 | if (ret < 0 || td->error) { | |
8b28bd41 DM |
372 | int err = td->error; |
373 | enum error_type_bit eb; | |
2e1df07d JA |
374 | |
375 | if (ret < 0) | |
376 | err = -ret; | |
2e1df07d | 377 | |
8b28bd41 DM |
378 | eb = td_error_type(ddir, err); |
379 | if (!(td->o.continue_on_error & (1 << eb))) | |
2e1df07d JA |
380 | return 1; |
381 | ||
8b28bd41 | 382 | if (td_non_fatal_error(td, eb, err)) { |
2e1df07d JA |
383 | /* |
384 | * Continue with the I/Os in case of | |
385 | * a non fatal error. | |
386 | */ | |
387 | update_error_count(td, err); | |
388 | td_clear_error(td); | |
389 | *retptr = 0; | |
390 | return 0; | |
391 | } else if (td->o.fill_device && err == ENOSPC) { | |
392 | /* | |
393 | * We expect to hit this error if | |
394 | * fill_device option is set. | |
395 | */ | |
396 | td_clear_error(td); | |
ebea2133 | 397 | fio_mark_td_terminate(td); |
2e1df07d JA |
398 | return 1; |
399 | } else { | |
400 | /* | |
401 | * Stop the I/O in case of a fatal | |
402 | * error. | |
403 | */ | |
404 | update_error_count(td, err); | |
405 | return 1; | |
406 | } | |
407 | } | |
408 | ||
409 | return 0; | |
410 | } | |
411 | ||
c97f1ad6 JA |
412 | static void check_update_rusage(struct thread_data *td) |
413 | { | |
414 | if (td->update_rusage) { | |
415 | td->update_rusage = 0; | |
416 | update_rusage_stat(td); | |
417 | fio_mutex_up(td->rusage_sem); | |
418 | } | |
419 | } | |
420 | ||
69fea81e JA |
421 | static int wait_for_completions(struct thread_data *td, struct timeval *time, |
422 | uint64_t *bytes_done) | |
423 | { | |
424 | const int full = queue_full(td); | |
425 | int min_evts = 0; | |
426 | int ret; | |
427 | ||
428 | /* | |
429 | * if the queue is full, we MUST reap at least 1 event | |
430 | */ | |
431 | min_evts = min(td->o.iodepth_batch_complete, td->cur_depth); | |
432 | if (full && !min_evts) | |
433 | min_evts = 1; | |
434 | ||
66b6c5ef | 435 | if (time && (__should_check_rate(td, DDIR_READ) || |
69fea81e JA |
436 | __should_check_rate(td, DDIR_WRITE) || |
437 | __should_check_rate(td, DDIR_TRIM))) | |
438 | fio_gettime(time, NULL); | |
439 | ||
440 | do { | |
441 | ret = io_u_queued_complete(td, min_evts, bytes_done); | |
442 | if (ret < 0) | |
443 | break; | |
444 | } while (full && (td->cur_depth > td->o.iodepth_low)); | |
445 | ||
446 | return ret; | |
447 | } | |
448 | ||
2e1df07d JA |
449 | /* |
450 | * The main verify engine. Runs over the writes we previously submitted, | |
451 | * reads the blocks back in, and checks the crc/md5 of the data. | |
452 | */ | |
100f49f1 | 453 | static void do_verify(struct thread_data *td, uint64_t verify_bytes) |
2e1df07d | 454 | { |
100f49f1 | 455 | uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 }; |
2e1df07d JA |
456 | struct fio_file *f; |
457 | struct io_u *io_u; | |
458 | int ret, min_events; | |
459 | unsigned int i; | |
460 | ||
461 | dprint(FD_VERIFY, "starting loop\n"); | |
462 | ||
463 | /* | |
464 | * sync io first and invalidate cache, to make sure we really | |
465 | * read from disk. | |
466 | */ | |
467 | for_each_file(td, f, i) { | |
468 | if (!fio_file_open(f)) | |
469 | continue; | |
470 | if (fio_io_sync(td, f)) | |
471 | break; | |
472 | if (file_invalidate_cache(td, f)) | |
473 | break; | |
474 | } | |
475 | ||
c97f1ad6 JA |
476 | check_update_rusage(td); |
477 | ||
2e1df07d JA |
478 | if (td->error) |
479 | return; | |
480 | ||
481 | td_set_runstate(td, TD_VERIFYING); | |
482 | ||
483 | io_u = NULL; | |
484 | while (!td->terminate) { | |
fbccf46c | 485 | enum fio_ddir ddir; |
2e1df07d JA |
486 | int ret2, full; |
487 | ||
488 | update_tv_cache(td); | |
c97f1ad6 | 489 | check_update_rusage(td); |
2e1df07d JA |
490 | |
491 | if (runtime_exceeded(td, &td->tv_cache)) { | |
492 | __update_tv_cache(td); | |
493 | if (runtime_exceeded(td, &td->tv_cache)) { | |
ebea2133 | 494 | fio_mark_td_terminate(td); |
2e1df07d JA |
495 | break; |
496 | } | |
497 | } | |
498 | ||
9e684a49 DE |
499 | if (flow_threshold_exceeded(td)) |
500 | continue; | |
501 | ||
44cbc6da JA |
502 | if (!td->o.experimental_verify) { |
503 | io_u = __get_io_u(td); | |
504 | if (!io_u) | |
505 | break; | |
2e1df07d | 506 | |
44cbc6da JA |
507 | if (get_next_verify(td, io_u)) { |
508 | put_io_u(td, io_u); | |
509 | break; | |
510 | } | |
2e1df07d | 511 | |
44cbc6da JA |
512 | if (td_io_prep(td, io_u)) { |
513 | put_io_u(td, io_u); | |
514 | break; | |
515 | } | |
516 | } else { | |
100f49f1 JA |
517 | if (ddir_rw_sum(bytes_done) + td->o.rw_min_bs > verify_bytes) |
518 | break; | |
519 | ||
bcd5abfa | 520 | while ((io_u = get_io_u(td)) != NULL) { |
002fe734 JA |
521 | if (IS_ERR(io_u)) { |
522 | io_u = NULL; | |
523 | ret = FIO_Q_BUSY; | |
524 | goto reap; | |
525 | } | |
526 | ||
bcd5abfa JA |
527 | /* |
528 | * We are only interested in the places where | |
529 | * we wrote or trimmed IOs. Turn those into | |
530 | * reads for verification purposes. | |
531 | */ | |
532 | if (io_u->ddir == DDIR_READ) { | |
533 | /* | |
534 | * Pretend we issued it for rwmix | |
535 | * accounting | |
536 | */ | |
537 | td->io_issues[DDIR_READ]++; | |
538 | put_io_u(td, io_u); | |
539 | continue; | |
540 | } else if (io_u->ddir == DDIR_TRIM) { | |
541 | io_u->ddir = DDIR_READ; | |
542 | io_u->flags |= IO_U_F_TRIMMED; | |
543 | break; | |
544 | } else if (io_u->ddir == DDIR_WRITE) { | |
545 | io_u->ddir = DDIR_READ; | |
546 | break; | |
547 | } else { | |
548 | put_io_u(td, io_u); | |
549 | continue; | |
550 | } | |
551 | } | |
44cbc6da | 552 | |
bcd5abfa | 553 | if (!io_u) |
44cbc6da | 554 | break; |
2e1df07d JA |
555 | } |
556 | ||
ca09be4b JA |
557 | if (verify_state_should_stop(td, io_u)) { |
558 | put_io_u(td, io_u); | |
559 | break; | |
560 | } | |
561 | ||
2e1df07d JA |
562 | if (td->o.verify_async) |
563 | io_u->end_io = verify_io_u_async; | |
564 | else | |
565 | io_u->end_io = verify_io_u; | |
566 | ||
fbccf46c | 567 | ddir = io_u->ddir; |
10adc4a7 GG |
568 | if (!td->o.disable_slat) |
569 | fio_gettime(&io_u->start_time, NULL); | |
fbccf46c | 570 | |
2e1df07d JA |
571 | ret = td_io_queue(td, io_u); |
572 | switch (ret) { | |
573 | case FIO_Q_COMPLETED: | |
574 | if (io_u->error) { | |
575 | ret = -io_u->error; | |
576 | clear_io_u(td, io_u); | |
577 | } else if (io_u->resid) { | |
578 | int bytes = io_u->xfer_buflen - io_u->resid; | |
579 | ||
580 | /* | |
581 | * zero read, fail | |
582 | */ | |
583 | if (!bytes) { | |
584 | td_verror(td, EIO, "full resid"); | |
585 | put_io_u(td, io_u); | |
586 | break; | |
587 | } | |
588 | ||
589 | io_u->xfer_buflen = io_u->resid; | |
590 | io_u->xfer_buf += bytes; | |
591 | io_u->offset += bytes; | |
592 | ||
593 | if (ddir_rw(io_u->ddir)) | |
594 | td->ts.short_io_u[io_u->ddir]++; | |
595 | ||
596 | f = io_u->file; | |
597 | if (io_u->offset == f->real_file_size) | |
598 | goto sync_done; | |
599 | ||
600 | requeue_io_u(td, &io_u); | |
601 | } else { | |
602 | sync_done: | |
100f49f1 | 603 | ret = io_u_sync_complete(td, io_u, bytes_done); |
2e1df07d JA |
604 | if (ret < 0) |
605 | break; | |
606 | } | |
607 | continue; | |
608 | case FIO_Q_QUEUED: | |
609 | break; | |
610 | case FIO_Q_BUSY: | |
611 | requeue_io_u(td, &io_u); | |
612 | ret2 = td_io_commit(td); | |
613 | if (ret2 < 0) | |
614 | ret = ret2; | |
615 | break; | |
616 | default: | |
617 | assert(ret < 0); | |
618 | td_verror(td, -ret, "td_io_queue"); | |
619 | break; | |
620 | } | |
621 | ||
fbccf46c | 622 | if (break_on_this_error(td, ddir, &ret)) |
2e1df07d JA |
623 | break; |
624 | ||
625 | /* | |
626 | * if we can queue more, do so. but check if there are | |
627 | * completed io_u's first. Note that we can get BUSY even | |
628 | * without IO queued, if the system is resource starved. | |
629 | */ | |
002fe734 | 630 | reap: |
2e1df07d | 631 | full = queue_full(td) || (ret == FIO_Q_BUSY && td->cur_depth); |
69fea81e JA |
632 | if (full || !td->o.iodepth_batch_complete) |
633 | ret = wait_for_completions(td, NULL, bytes_done); | |
2e1df07d | 634 | |
2e1df07d JA |
635 | if (ret < 0) |
636 | break; | |
637 | } | |
638 | ||
c97f1ad6 JA |
639 | check_update_rusage(td); |
640 | ||
2e1df07d JA |
641 | if (!td->error) { |
642 | min_events = td->cur_depth; | |
643 | ||
644 | if (min_events) | |
645 | ret = io_u_queued_complete(td, min_events, NULL); | |
646 | } else | |
647 | cleanup_pending_aio(td); | |
648 | ||
649 | td_set_runstate(td, TD_RUNNING); | |
650 | ||
651 | dprint(FD_VERIFY, "exiting loop\n"); | |
652 | } | |
653 | ||
3939fe85 JA |
654 | static unsigned int exceeds_number_ios(struct thread_data *td) |
655 | { | |
656 | unsigned long long number_ios; | |
657 | ||
658 | if (!td->o.number_ios) | |
659 | return 0; | |
660 | ||
cf8a46a7 | 661 | number_ios = ddir_rw_sum(td->io_blocks); |
3939fe85 JA |
662 | number_ios += td->io_u_queued + td->io_u_in_flight; |
663 | ||
cf8a46a7 | 664 | return number_ios >= (td->o.number_ios * td->loops); |
3939fe85 JA |
665 | } |
666 | ||
e28dd2cf | 667 | static int io_issue_bytes_exceeded(struct thread_data *td) |
f7078f7b | 668 | { |
77731b29 | 669 | unsigned long long bytes, limit; |
f7078f7b JA |
670 | |
671 | if (td_rw(td)) | |
74d6277f | 672 | bytes = td->io_issue_bytes[DDIR_READ] + td->io_issue_bytes[DDIR_WRITE]; |
f7078f7b | 673 | else if (td_write(td)) |
74d6277f | 674 | bytes = td->io_issue_bytes[DDIR_WRITE]; |
6eaf09d6 | 675 | else if (td_read(td)) |
74d6277f | 676 | bytes = td->io_issue_bytes[DDIR_READ]; |
f7078f7b | 677 | else |
74d6277f | 678 | bytes = td->io_issue_bytes[DDIR_TRIM]; |
f7078f7b | 679 | |
77731b29 JA |
680 | if (td->o.io_limit) |
681 | limit = td->o.io_limit; | |
682 | else | |
683 | limit = td->o.size; | |
684 | ||
cf8a46a7 | 685 | limit *= td->loops; |
77731b29 | 686 | return bytes >= limit || exceeds_number_ios(td); |
f7078f7b JA |
687 | } |
688 | ||
e28dd2cf JE |
689 | static int io_complete_bytes_exceeded(struct thread_data *td) |
690 | { | |
691 | unsigned long long bytes, limit; | |
692 | ||
693 | if (td_rw(td)) | |
694 | bytes = td->this_io_bytes[DDIR_READ] + td->this_io_bytes[DDIR_WRITE]; | |
695 | else if (td_write(td)) | |
696 | bytes = td->this_io_bytes[DDIR_WRITE]; | |
697 | else if (td_read(td)) | |
698 | bytes = td->this_io_bytes[DDIR_READ]; | |
699 | else | |
700 | bytes = td->this_io_bytes[DDIR_TRIM]; | |
701 | ||
702 | if (td->o.io_limit) | |
703 | limit = td->o.io_limit; | |
704 | else | |
705 | limit = td->o.size; | |
706 | ||
cf8a46a7 | 707 | limit *= td->loops; |
e28dd2cf JE |
708 | return bytes >= limit || exceeds_number_ios(td); |
709 | } | |
710 | ||
2e1df07d JA |
711 | /* |
712 | * Main IO worker function. It retrieves io_u's to process and queues | |
713 | * and reaps them, checking for rate and errors along the way. | |
100f49f1 JA |
714 | * |
715 | * Returns number of bytes written and trimmed. | |
2e1df07d | 716 | */ |
100f49f1 | 717 | static uint64_t do_io(struct thread_data *td) |
2e1df07d | 718 | { |
100f49f1 | 719 | uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 }; |
2e1df07d JA |
720 | unsigned int i; |
721 | int ret = 0; | |
c2703bf3 | 722 | uint64_t total_bytes, bytes_issued = 0; |
2e1df07d JA |
723 | |
724 | if (in_ramp_time(td)) | |
725 | td_set_runstate(td, TD_RAMP); | |
726 | else | |
727 | td_set_runstate(td, TD_RUNNING); | |
728 | ||
3e260a46 JA |
729 | lat_target_init(td); |
730 | ||
1e564979 JE |
731 | total_bytes = td->o.size; |
732 | /* | |
733 | * Allow random overwrite workloads to write up to io_limit | |
734 | * before starting verification phase as 'size' doesn't apply. | |
735 | */ | |
736 | if (td_write(td) && td_random(td) && td->o.norandommap) | |
737 | total_bytes = max(total_bytes, (uint64_t) td->o.io_limit); | |
78a6469c JA |
738 | /* |
739 | * If verify_backlog is enabled, we'll run the verify in this | |
740 | * handler as well. For that case, we may need up to twice the | |
741 | * amount of bytes. | |
742 | */ | |
78a6469c JA |
743 | if (td->o.verify != VERIFY_NONE && |
744 | (td_write(td) && td->o.verify_backlog)) | |
c2703bf3 JA |
745 | total_bytes += td->o.size; |
746 | ||
82a90686 | 747 | /* In trimwrite mode, each byte is trimmed and then written, so |
0e4dd95c | 748 | * allow total_bytes to be twice as big */ |
82a90686 | 749 | if (td_trimwrite(td)) |
0e4dd95c DE |
750 | total_bytes += td->total_io_size; |
751 | ||
f7078f7b | 752 | while ((td->o.read_iolog_file && !flist_empty(&td->io_log_list)) || |
e28dd2cf | 753 | (!flist_empty(&td->trim_list)) || !io_issue_bytes_exceeded(td) || |
c04e4661 | 754 | td->o.time_based) { |
2e1df07d | 755 | struct timeval comp_time; |
2e1df07d JA |
756 | struct io_u *io_u; |
757 | int ret2, full; | |
758 | enum fio_ddir ddir; | |
759 | ||
c97f1ad6 JA |
760 | check_update_rusage(td); |
761 | ||
7d7803fa | 762 | if (td->terminate || td->done) |
2e1df07d JA |
763 | break; |
764 | ||
765 | update_tv_cache(td); | |
766 | ||
767 | if (runtime_exceeded(td, &td->tv_cache)) { | |
768 | __update_tv_cache(td); | |
769 | if (runtime_exceeded(td, &td->tv_cache)) { | |
ebea2133 | 770 | fio_mark_td_terminate(td); |
2e1df07d JA |
771 | break; |
772 | } | |
773 | } | |
774 | ||
9e684a49 DE |
775 | if (flow_threshold_exceeded(td)) |
776 | continue; | |
777 | ||
c2703bf3 | 778 | if (bytes_issued >= total_bytes) |
20876c53 JC |
779 | break; |
780 | ||
2e1df07d | 781 | io_u = get_io_u(td); |
002fe734 JA |
782 | if (IS_ERR_OR_NULL(io_u)) { |
783 | int err = PTR_ERR(io_u); | |
784 | ||
785 | io_u = NULL; | |
786 | if (err == -EBUSY) { | |
787 | ret = FIO_Q_BUSY; | |
788 | goto reap; | |
789 | } | |
3e260a46 JA |
790 | if (td->o.latency_target) |
791 | goto reap; | |
2e1df07d | 792 | break; |
3e260a46 | 793 | } |
2e1df07d JA |
794 | |
795 | ddir = io_u->ddir; | |
796 | ||
797 | /* | |
82af2a7c JA |
798 | * Add verification end_io handler if: |
799 | * - Asked to verify (!td_rw(td)) | |
800 | * - Or the io_u is from our verify list (mixed write/ver) | |
2e1df07d JA |
801 | */ |
802 | if (td->o.verify != VERIFY_NONE && io_u->ddir == DDIR_READ && | |
82af2a7c | 803 | ((io_u->flags & IO_U_F_VER_LIST) || !td_rw(td))) { |
c4b6117b PV |
804 | |
805 | if (!td->o.verify_pattern_bytes) { | |
d6b72507 | 806 | io_u->rand_seed = __rand(&td->verify_state); |
c4b6117b | 807 | if (sizeof(int) != sizeof(long *)) |
d6b72507 | 808 | io_u->rand_seed *= __rand(&td->verify_state); |
c4b6117b PV |
809 | } |
810 | ||
ca09be4b JA |
811 | if (verify_state_should_stop(td, io_u)) { |
812 | put_io_u(td, io_u); | |
813 | break; | |
814 | } | |
815 | ||
2e1df07d JA |
816 | if (td->o.verify_async) |
817 | io_u->end_io = verify_io_u_async; | |
818 | else | |
819 | io_u->end_io = verify_io_u; | |
820 | td_set_runstate(td, TD_VERIFYING); | |
821 | } else if (in_ramp_time(td)) | |
822 | td_set_runstate(td, TD_RAMP); | |
823 | else | |
824 | td_set_runstate(td, TD_RUNNING); | |
825 | ||
9a50c5c5 | 826 | /* |
f9401285 JA |
827 | * Always log IO before it's issued, so we know the specific |
828 | * order of it. The logged unit will track when the IO has | |
829 | * completed. | |
9a50c5c5 | 830 | */ |
c4b6117b PV |
831 | if (td_write(td) && io_u->ddir == DDIR_WRITE && |
832 | td->o.do_verify && | |
833 | td->o.verify != VERIFY_NONE && | |
f9401285 | 834 | !td->o.experimental_verify) |
c4b6117b PV |
835 | log_io_piece(td, io_u); |
836 | ||
2e1df07d JA |
837 | ret = td_io_queue(td, io_u); |
838 | switch (ret) { | |
839 | case FIO_Q_COMPLETED: | |
840 | if (io_u->error) { | |
841 | ret = -io_u->error; | |
890b6656 | 842 | unlog_io_piece(td, io_u); |
2e1df07d JA |
843 | clear_io_u(td, io_u); |
844 | } else if (io_u->resid) { | |
845 | int bytes = io_u->xfer_buflen - io_u->resid; | |
846 | struct fio_file *f = io_u->file; | |
847 | ||
20876c53 | 848 | bytes_issued += bytes; |
890b6656 JA |
849 | |
850 | trim_io_piece(td, io_u); | |
851 | ||
2e1df07d JA |
852 | /* |
853 | * zero read, fail | |
854 | */ | |
855 | if (!bytes) { | |
890b6656 | 856 | unlog_io_piece(td, io_u); |
2e1df07d JA |
857 | td_verror(td, EIO, "full resid"); |
858 | put_io_u(td, io_u); | |
859 | break; | |
860 | } | |
861 | ||
862 | io_u->xfer_buflen = io_u->resid; | |
863 | io_u->xfer_buf += bytes; | |
864 | io_u->offset += bytes; | |
865 | ||
866 | if (ddir_rw(io_u->ddir)) | |
867 | td->ts.short_io_u[io_u->ddir]++; | |
868 | ||
869 | if (io_u->offset == f->real_file_size) | |
870 | goto sync_done; | |
871 | ||
872 | requeue_io_u(td, &io_u); | |
873 | } else { | |
874 | sync_done: | |
6eaf09d6 SL |
875 | if (__should_check_rate(td, DDIR_READ) || |
876 | __should_check_rate(td, DDIR_WRITE) || | |
877 | __should_check_rate(td, DDIR_TRIM)) | |
2e1df07d JA |
878 | fio_gettime(&comp_time, NULL); |
879 | ||
880 | ret = io_u_sync_complete(td, io_u, bytes_done); | |
881 | if (ret < 0) | |
882 | break; | |
20876c53 | 883 | bytes_issued += io_u->xfer_buflen; |
2e1df07d JA |
884 | } |
885 | break; | |
886 | case FIO_Q_QUEUED: | |
887 | /* | |
888 | * if the engine doesn't have a commit hook, | |
889 | * the io_u is really queued. if it does have such | |
890 | * a hook, it has to call io_u_queued() itself. | |
891 | */ | |
892 | if (td->io_ops->commit == NULL) | |
893 | io_u_queued(td, io_u); | |
20876c53 | 894 | bytes_issued += io_u->xfer_buflen; |
2e1df07d JA |
895 | break; |
896 | case FIO_Q_BUSY: | |
890b6656 | 897 | unlog_io_piece(td, io_u); |
2e1df07d JA |
898 | requeue_io_u(td, &io_u); |
899 | ret2 = td_io_commit(td); | |
900 | if (ret2 < 0) | |
901 | ret = ret2; | |
902 | break; | |
903 | default: | |
904 | assert(ret < 0); | |
905 | put_io_u(td, io_u); | |
906 | break; | |
907 | } | |
908 | ||
909 | if (break_on_this_error(td, ddir, &ret)) | |
910 | break; | |
911 | ||
912 | /* | |
913 | * See if we need to complete some commands. Note that we | |
914 | * can get BUSY even without IO queued, if the system is | |
915 | * resource starved. | |
916 | */ | |
3e260a46 | 917 | reap: |
2e1df07d | 918 | full = queue_full(td) || (ret == FIO_Q_BUSY && td->cur_depth); |
69fea81e JA |
919 | if (full || !td->o.iodepth_batch_complete) |
920 | ret = wait_for_completions(td, &comp_time, bytes_done); | |
2e1df07d JA |
921 | if (ret < 0) |
922 | break; | |
d5abee06 | 923 | if (!ddir_rw_sum(bytes_done) && !(td->io_ops->flags & FIO_NOIO)) |
2e1df07d JA |
924 | continue; |
925 | ||
926 | if (!in_ramp_time(td) && should_check_rate(td, bytes_done)) { | |
927 | if (check_min_rate(td, &comp_time, bytes_done)) { | |
928 | if (exitall_on_terminate) | |
929 | fio_terminate_threads(td->groupid); | |
930 | td_verror(td, EIO, "check_min_rate"); | |
931 | break; | |
932 | } | |
933 | } | |
3e260a46 JA |
934 | if (!in_ramp_time(td) && td->o.latency_target) |
935 | lat_target_check(td); | |
2e1df07d JA |
936 | |
937 | if (td->o.thinktime) { | |
938 | unsigned long long b; | |
939 | ||
342f4be4 | 940 | b = ddir_rw_sum(td->io_blocks); |
2e1df07d JA |
941 | if (!(b % td->o.thinktime_blocks)) { |
942 | int left; | |
943 | ||
002e7183 JA |
944 | io_u_quiesce(td); |
945 | ||
2e1df07d JA |
946 | if (td->o.thinktime_spin) |
947 | usec_spin(td->o.thinktime_spin); | |
948 | ||
949 | left = td->o.thinktime - td->o.thinktime_spin; | |
950 | if (left) | |
951 | usec_sleep(td, left); | |
952 | } | |
953 | } | |
954 | } | |
955 | ||
c97f1ad6 JA |
956 | check_update_rusage(td); |
957 | ||
2e1df07d | 958 | if (td->trim_entries) |
4e0a8fa2 | 959 | log_err("fio: %lu trim entries leaked?\n", td->trim_entries); |
2e1df07d JA |
960 | |
961 | if (td->o.fill_device && td->error == ENOSPC) { | |
962 | td->error = 0; | |
ebea2133 | 963 | fio_mark_td_terminate(td); |
2e1df07d JA |
964 | } |
965 | if (!td->error) { | |
966 | struct fio_file *f; | |
967 | ||
968 | i = td->cur_depth; | |
969 | if (i) { | |
5bd5f71a | 970 | ret = io_u_queued_complete(td, i, bytes_done); |
2e1df07d JA |
971 | if (td->o.fill_device && td->error == ENOSPC) |
972 | td->error = 0; | |
973 | } | |
974 | ||
975 | if (should_fsync(td) && td->o.end_fsync) { | |
976 | td_set_runstate(td, TD_FSYNCING); | |
977 | ||
978 | for_each_file(td, f, i) { | |
61ee0f86 | 979 | if (!fio_file_fsync(td, f)) |
2e1df07d | 980 | continue; |
61ee0f86 JA |
981 | |
982 | log_err("fio: end_fsync failed for file %s\n", | |
983 | f->file_name); | |
2e1df07d JA |
984 | } |
985 | } | |
986 | } else | |
987 | cleanup_pending_aio(td); | |
988 | ||
989 | /* | |
990 | * stop job if we failed doing any IO | |
991 | */ | |
342f4be4 | 992 | if (!ddir_rw_sum(td->this_io_bytes)) |
2e1df07d | 993 | td->done = 1; |
100f49f1 JA |
994 | |
995 | return bytes_done[DDIR_WRITE] + bytes_done[DDIR_TRIM]; | |
2e1df07d JA |
996 | } |
997 | ||
998 | static void cleanup_io_u(struct thread_data *td) | |
999 | { | |
2e1df07d JA |
1000 | struct io_u *io_u; |
1001 | ||
2ae0b204 | 1002 | while ((io_u = io_u_qpop(&td->io_u_freelist)) != NULL) { |
c73ed246 JA |
1003 | |
1004 | if (td->io_ops->io_u_free) | |
1005 | td->io_ops->io_u_free(td, io_u); | |
1006 | ||
2e1df07d JA |
1007 | fio_memfree(io_u, sizeof(*io_u)); |
1008 | } | |
1009 | ||
1010 | free_io_mem(td); | |
2ae0b204 JA |
1011 | |
1012 | io_u_rexit(&td->io_u_requeues); | |
1013 | io_u_qexit(&td->io_u_freelist); | |
1014 | io_u_qexit(&td->io_u_all); | |
ca09be4b JA |
1015 | |
1016 | if (td->last_write_comp) | |
1017 | sfree(td->last_write_comp); | |
2e1df07d JA |
1018 | } |
1019 | ||
1020 | static int init_io_u(struct thread_data *td) | |
1021 | { | |
1022 | struct io_u *io_u; | |
9c42684e | 1023 | unsigned int max_bs, min_write; |
2e1df07d | 1024 | int cl_align, i, max_units; |
2ae0b204 | 1025 | int data_xfer = 1, err; |
2e1df07d JA |
1026 | char *p; |
1027 | ||
1028 | max_units = td->o.iodepth; | |
74f4b020 | 1029 | max_bs = td_max_bs(td); |
9c42684e | 1030 | min_write = td->o.min_bs[DDIR_WRITE]; |
2e1df07d JA |
1031 | td->orig_buffer_size = (unsigned long long) max_bs |
1032 | * (unsigned long long) max_units; | |
1033 | ||
88045e04 | 1034 | if ((td->io_ops->flags & FIO_NOIO) || !(td_read(td) || td_write(td))) |
59d8d0f5 JA |
1035 | data_xfer = 0; |
1036 | ||
2ae0b204 JA |
1037 | err = 0; |
1038 | err += io_u_rinit(&td->io_u_requeues, td->o.iodepth); | |
1039 | err += io_u_qinit(&td->io_u_freelist, td->o.iodepth); | |
1040 | err += io_u_qinit(&td->io_u_all, td->o.iodepth); | |
1041 | ||
1042 | if (err) { | |
1043 | log_err("fio: failed setting up IO queues\n"); | |
1044 | return 1; | |
1045 | } | |
1046 | ||
fd8a09b8 | 1047 | /* |
1048 | * if we may later need to do address alignment, then add any | |
1049 | * possible adjustment here so that we don't cause a buffer | |
1050 | * overflow later. this adjustment may be too much if we get | |
1051 | * lucky and the allocator gives us an aligned address. | |
1052 | */ | |
d01612f3 CM |
1053 | if (td->o.odirect || td->o.mem_align || td->o.oatomic || |
1054 | (td->io_ops->flags & FIO_RAWIO)) | |
fd8a09b8 | 1055 | td->orig_buffer_size += page_mask + td->o.mem_align; |
1056 | ||
2e1df07d JA |
1057 | if (td->o.mem_type == MEM_SHMHUGE || td->o.mem_type == MEM_MMAPHUGE) { |
1058 | unsigned long bs; | |
1059 | ||
1060 | bs = td->orig_buffer_size + td->o.hugepage_size - 1; | |
1061 | td->orig_buffer_size = bs & ~(td->o.hugepage_size - 1); | |
1062 | } | |
1063 | ||
1064 | if (td->orig_buffer_size != (size_t) td->orig_buffer_size) { | |
1065 | log_err("fio: IO memory too large. Reduce max_bs or iodepth\n"); | |
1066 | return 1; | |
1067 | } | |
1068 | ||
59d8d0f5 | 1069 | if (data_xfer && allocate_io_mem(td)) |
2e1df07d JA |
1070 | return 1; |
1071 | ||
d01612f3 | 1072 | if (td->o.odirect || td->o.mem_align || td->o.oatomic || |
2e1df07d JA |
1073 | (td->io_ops->flags & FIO_RAWIO)) |
1074 | p = PAGE_ALIGN(td->orig_buffer) + td->o.mem_align; | |
1075 | else | |
1076 | p = td->orig_buffer; | |
1077 | ||
1078 | cl_align = os_cache_line_size(); | |
1079 | ||
1080 | for (i = 0; i < max_units; i++) { | |
1081 | void *ptr; | |
1082 | ||
1083 | if (td->terminate) | |
1084 | return 1; | |
1085 | ||
1086 | ptr = fio_memalign(cl_align, sizeof(*io_u)); | |
1087 | if (!ptr) { | |
1088 | log_err("fio: unable to allocate aligned memory\n"); | |
1089 | break; | |
1090 | } | |
1091 | ||
1092 | io_u = ptr; | |
1093 | memset(io_u, 0, sizeof(*io_u)); | |
2ae0b204 | 1094 | INIT_FLIST_HEAD(&io_u->verify_list); |
2e1df07d JA |
1095 | dprint(FD_MEM, "io_u alloc %p, index %u\n", io_u, i); |
1096 | ||
59d8d0f5 | 1097 | if (data_xfer) { |
2e1df07d JA |
1098 | io_u->buf = p; |
1099 | dprint(FD_MEM, "io_u %p, mem %p\n", io_u, io_u->buf); | |
1100 | ||
1101 | if (td_write(td)) | |
9c42684e | 1102 | io_u_fill_buffer(td, io_u, min_write, max_bs); |
2e1df07d JA |
1103 | if (td_write(td) && td->o.verify_pattern_bytes) { |
1104 | /* | |
1105 | * Fill the buffer with the pattern if we are | |
1106 | * going to be doing writes. | |
1107 | */ | |
ce35b1ec | 1108 | fill_verify_pattern(td, io_u->buf, max_bs, io_u, 0, 0); |
2e1df07d JA |
1109 | } |
1110 | } | |
1111 | ||
1112 | io_u->index = i; | |
1113 | io_u->flags = IO_U_F_FREE; | |
2ae0b204 JA |
1114 | io_u_qpush(&td->io_u_freelist, io_u); |
1115 | ||
1116 | /* | |
1117 | * io_u never leaves this stack, used for iteration of all | |
1118 | * io_u buffers. | |
1119 | */ | |
1120 | io_u_qpush(&td->io_u_all, io_u); | |
c73ed246 JA |
1121 | |
1122 | if (td->io_ops->io_u_init) { | |
1123 | int ret = td->io_ops->io_u_init(td, io_u); | |
1124 | ||
1125 | if (ret) { | |
1126 | log_err("fio: failed to init engine data: %d\n", ret); | |
1127 | return 1; | |
1128 | } | |
1129 | } | |
1130 | ||
2e1df07d JA |
1131 | p += max_bs; |
1132 | } | |
1133 | ||
ca09be4b JA |
1134 | if (td->o.verify != VERIFY_NONE) { |
1135 | td->last_write_comp = scalloc(max_units, sizeof(uint64_t)); | |
1136 | if (!td->last_write_comp) { | |
1137 | log_err("fio: failed to alloc write comp data\n"); | |
1138 | return 1; | |
1139 | } | |
1140 | } | |
1141 | ||
2e1df07d JA |
1142 | return 0; |
1143 | } | |
1144 | ||
1145 | static int switch_ioscheduler(struct thread_data *td) | |
1146 | { | |
1147 | char tmp[256], tmp2[128]; | |
1148 | FILE *f; | |
1149 | int ret; | |
1150 | ||
1151 | if (td->io_ops->flags & FIO_DISKLESSIO) | |
1152 | return 0; | |
1153 | ||
1154 | sprintf(tmp, "%s/queue/scheduler", td->sysfs_root); | |
1155 | ||
1156 | f = fopen(tmp, "r+"); | |
1157 | if (!f) { | |
1158 | if (errno == ENOENT) { | |
1159 | log_err("fio: os or kernel doesn't support IO scheduler" | |
1160 | " switching\n"); | |
1161 | return 0; | |
1162 | } | |
1163 | td_verror(td, errno, "fopen iosched"); | |
1164 | return 1; | |
1165 | } | |
1166 | ||
1167 | /* | |
1168 | * Set io scheduler. | |
1169 | */ | |
1170 | ret = fwrite(td->o.ioscheduler, strlen(td->o.ioscheduler), 1, f); | |
1171 | if (ferror(f) || ret != 1) { | |
1172 | td_verror(td, errno, "fwrite"); | |
1173 | fclose(f); | |
1174 | return 1; | |
1175 | } | |
1176 | ||
1177 | rewind(f); | |
1178 | ||
1179 | /* | |
1180 | * Read back and check that the selected scheduler is now the default. | |
1181 | */ | |
49c6f33d | 1182 | ret = fread(tmp, sizeof(tmp), 1, f); |
2e1df07d JA |
1183 | if (ferror(f) || ret < 0) { |
1184 | td_verror(td, errno, "fread"); | |
1185 | fclose(f); | |
1186 | return 1; | |
1187 | } | |
49c6f33d JA |
1188 | tmp[sizeof(tmp) - 1] = '\0'; |
1189 | ||
2e1df07d JA |
1190 | |
1191 | sprintf(tmp2, "[%s]", td->o.ioscheduler); | |
1192 | if (!strstr(tmp, tmp2)) { | |
1193 | log_err("fio: io scheduler %s not found\n", td->o.ioscheduler); | |
1194 | td_verror(td, EINVAL, "iosched_switch"); | |
1195 | fclose(f); | |
1196 | return 1; | |
1197 | } | |
1198 | ||
1199 | fclose(f); | |
1200 | return 0; | |
1201 | } | |
1202 | ||
1203 | static int keep_running(struct thread_data *td) | |
1204 | { | |
77731b29 JA |
1205 | unsigned long long limit; |
1206 | ||
2e1df07d JA |
1207 | if (td->done) |
1208 | return 0; | |
1209 | if (td->o.time_based) | |
1210 | return 1; | |
1211 | if (td->o.loops) { | |
1212 | td->o.loops--; | |
1213 | return 1; | |
1214 | } | |
3939fe85 JA |
1215 | if (exceeds_number_ios(td)) |
1216 | return 0; | |
26251d8d | 1217 | |
77731b29 JA |
1218 | if (td->o.io_limit) |
1219 | limit = td->o.io_limit; | |
1220 | else | |
1221 | limit = td->o.size; | |
1222 | ||
1223 | if (limit != -1ULL && ddir_rw_sum(td->io_bytes) < limit) { | |
5bd5f71a JA |
1224 | uint64_t diff; |
1225 | ||
1226 | /* | |
1227 | * If the difference is less than the minimum IO size, we | |
1228 | * are done. | |
1229 | */ | |
77731b29 | 1230 | diff = limit - ddir_rw_sum(td->io_bytes); |
74f4b020 | 1231 | if (diff < td_max_bs(td)) |
5bd5f71a JA |
1232 | return 0; |
1233 | ||
002fe734 JA |
1234 | if (fio_files_done(td)) |
1235 | return 0; | |
1236 | ||
2e1df07d | 1237 | return 1; |
5bd5f71a | 1238 | } |
2e1df07d JA |
1239 | |
1240 | return 0; | |
1241 | } | |
1242 | ||
ce486495 | 1243 | static int exec_string(struct thread_options *o, const char *string, const char *mode) |
2e1df07d | 1244 | { |
ce486495 | 1245 | int ret, newlen = strlen(string) + strlen(o->name) + strlen(mode) + 9 + 1; |
2e1df07d JA |
1246 | char *str; |
1247 | ||
1248 | str = malloc(newlen); | |
ce486495 | 1249 | sprintf(str, "%s &> %s.%s.txt", string, o->name, mode); |
2e1df07d | 1250 | |
ce486495 | 1251 | log_info("%s : Saving output of %s in %s.%s.txt\n",o->name, mode, o->name, mode); |
2e1df07d JA |
1252 | ret = system(str); |
1253 | if (ret == -1) | |
1254 | log_err("fio: exec of cmd <%s> failed\n", str); | |
1255 | ||
1256 | free(str); | |
1257 | return ret; | |
1258 | } | |
1259 | ||
62167762 JC |
1260 | /* |
1261 | * Dry run to compute correct state of numberio for verification. | |
1262 | */ | |
1263 | static uint64_t do_dry_run(struct thread_data *td) | |
1264 | { | |
1265 | uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 }; | |
1266 | ||
1267 | td_set_runstate(td, TD_RUNNING); | |
1268 | ||
1269 | while ((td->o.read_iolog_file && !flist_empty(&td->io_log_list)) || | |
e28dd2cf | 1270 | (!flist_empty(&td->trim_list)) || !io_complete_bytes_exceeded(td)) { |
62167762 JC |
1271 | struct io_u *io_u; |
1272 | int ret; | |
1273 | ||
1274 | if (td->terminate || td->done) | |
1275 | break; | |
1276 | ||
1277 | io_u = get_io_u(td); | |
1278 | if (!io_u) | |
1279 | break; | |
1280 | ||
1281 | io_u->flags |= IO_U_F_FLIGHT; | |
1282 | io_u->error = 0; | |
1283 | io_u->resid = 0; | |
1284 | if (ddir_rw(acct_ddir(io_u))) | |
1285 | td->io_issues[acct_ddir(io_u)]++; | |
1286 | if (ddir_rw(io_u->ddir)) { | |
1287 | io_u_mark_depth(td, 1); | |
1288 | td->ts.total_io_u[io_u->ddir]++; | |
1289 | } | |
1290 | ||
2e63e96b PV |
1291 | if (td_write(td) && io_u->ddir == DDIR_WRITE && |
1292 | td->o.do_verify && | |
1293 | td->o.verify != VERIFY_NONE && | |
1294 | !td->o.experimental_verify) | |
1295 | log_io_piece(td, io_u); | |
1296 | ||
62167762 JC |
1297 | ret = io_u_sync_complete(td, io_u, bytes_done); |
1298 | (void) ret; | |
1299 | } | |
1300 | ||
1301 | return bytes_done[DDIR_WRITE] + bytes_done[DDIR_TRIM]; | |
1302 | } | |
1303 | ||
2e1df07d JA |
1304 | /* |
1305 | * Entry point for the thread based jobs. The process based jobs end up | |
1306 | * here as well, after a little setup. | |
1307 | */ | |
1308 | static void *thread_main(void *data) | |
1309 | { | |
1310 | unsigned long long elapsed; | |
1311 | struct thread_data *td = data; | |
4896473e | 1312 | struct thread_options *o = &td->o; |
2e1df07d JA |
1313 | pthread_condattr_t attr; |
1314 | int clear_state; | |
28727df7 | 1315 | int ret; |
2e1df07d | 1316 | |
4896473e | 1317 | if (!o->use_thread) { |
2e1df07d JA |
1318 | setsid(); |
1319 | td->pid = getpid(); | |
1320 | } else | |
1321 | td->pid = gettid(); | |
1322 | ||
4896473e | 1323 | fio_local_clock_init(o->use_thread); |
5d879392 | 1324 | |
2e1df07d JA |
1325 | dprint(FD_PROCESS, "jobs pid=%d started\n", (int) td->pid); |
1326 | ||
122c7725 JA |
1327 | if (is_backend) |
1328 | fio_server_send_start(td); | |
1329 | ||
2e1df07d JA |
1330 | INIT_FLIST_HEAD(&td->io_log_list); |
1331 | INIT_FLIST_HEAD(&td->io_hist_list); | |
1332 | INIT_FLIST_HEAD(&td->verify_list); | |
1333 | INIT_FLIST_HEAD(&td->trim_list); | |
1ae83d45 | 1334 | INIT_FLIST_HEAD(&td->next_rand_list); |
2e1df07d JA |
1335 | pthread_mutex_init(&td->io_u_lock, NULL); |
1336 | td->io_hist_tree = RB_ROOT; | |
1337 | ||
1338 | pthread_condattr_init(&attr); | |
1339 | pthread_cond_init(&td->verify_cond, &attr); | |
1340 | pthread_cond_init(&td->free_cond, &attr); | |
1341 | ||
1342 | td_set_runstate(td, TD_INITIALIZED); | |
1343 | dprint(FD_MUTEX, "up startup_mutex\n"); | |
1344 | fio_mutex_up(startup_mutex); | |
1345 | dprint(FD_MUTEX, "wait on td->mutex\n"); | |
1346 | fio_mutex_down(td->mutex); | |
1347 | dprint(FD_MUTEX, "done waiting on td->mutex\n"); | |
1348 | ||
2e1df07d JA |
1349 | /* |
1350 | * A new gid requires privilege, so we need to do this before setting | |
1351 | * the uid. | |
1352 | */ | |
4896473e | 1353 | if (o->gid != -1U && setgid(o->gid)) { |
2e1df07d JA |
1354 | td_verror(td, errno, "setgid"); |
1355 | goto err; | |
1356 | } | |
4896473e | 1357 | if (o->uid != -1U && setuid(o->uid)) { |
2e1df07d JA |
1358 | td_verror(td, errno, "setuid"); |
1359 | goto err; | |
1360 | } | |
1361 | ||
1362 | /* | |
1363 | * If we have a gettimeofday() thread, make sure we exclude that | |
1364 | * thread from this job | |
1365 | */ | |
4896473e JA |
1366 | if (o->gtod_cpu) |
1367 | fio_cpu_clear(&o->cpumask, o->gtod_cpu); | |
2e1df07d JA |
1368 | |
1369 | /* | |
1370 | * Set affinity first, in case it has an impact on the memory | |
1371 | * allocations. | |
1372 | */ | |
b2a9e649 | 1373 | if (fio_option_is_set(o, cpumask)) { |
c2acfbac | 1374 | if (o->cpus_allowed_policy == FIO_CPUS_SPLIT) { |
30cb4c65 | 1375 | ret = fio_cpus_split(&o->cpumask, td->thread_number - 1); |
c2acfbac JA |
1376 | if (!ret) { |
1377 | log_err("fio: no CPUs set\n"); | |
1378 | log_err("fio: Try increasing number of available CPUs\n"); | |
1379 | td_verror(td, EINVAL, "cpus_split"); | |
1380 | goto err; | |
1381 | } | |
1382 | } | |
28727df7 JA |
1383 | ret = fio_setaffinity(td->pid, o->cpumask); |
1384 | if (ret == -1) { | |
4896473e JA |
1385 | td_verror(td, errno, "cpu_set_affinity"); |
1386 | goto err; | |
1387 | } | |
2e1df07d JA |
1388 | } |
1389 | ||
67bf9823 | 1390 | #ifdef CONFIG_LIBNUMA |
d0b937ed | 1391 | /* numa node setup */ |
b2a9e649 JA |
1392 | if (fio_option_is_set(o, numa_cpunodes) || |
1393 | fio_option_is_set(o, numa_memnodes)) { | |
43522848 | 1394 | struct bitmask *mask; |
d0b937ed YR |
1395 | |
1396 | if (numa_available() < 0) { | |
1397 | td_verror(td, errno, "Does not support NUMA API\n"); | |
1398 | goto err; | |
1399 | } | |
1400 | ||
b2a9e649 | 1401 | if (fio_option_is_set(o, numa_cpunodes)) { |
43522848 DG |
1402 | mask = numa_parse_nodestring(o->numa_cpunodes); |
1403 | ret = numa_run_on_node_mask(mask); | |
1404 | numa_free_nodemask(mask); | |
d0b937ed YR |
1405 | if (ret == -1) { |
1406 | td_verror(td, errno, \ | |
1407 | "numa_run_on_node_mask failed\n"); | |
1408 | goto err; | |
1409 | } | |
1410 | } | |
1411 | ||
b2a9e649 | 1412 | if (fio_option_is_set(o, numa_memnodes)) { |
43522848 DG |
1413 | mask = NULL; |
1414 | if (o->numa_memnodes) | |
1415 | mask = numa_parse_nodestring(o->numa_memnodes); | |
1416 | ||
4896473e | 1417 | switch (o->numa_mem_mode) { |
d0b937ed | 1418 | case MPOL_INTERLEAVE: |
43522848 | 1419 | numa_set_interleave_mask(mask); |
d0b937ed YR |
1420 | break; |
1421 | case MPOL_BIND: | |
43522848 | 1422 | numa_set_membind(mask); |
d0b937ed YR |
1423 | break; |
1424 | case MPOL_LOCAL: | |
1425 | numa_set_localalloc(); | |
1426 | break; | |
1427 | case MPOL_PREFERRED: | |
4896473e | 1428 | numa_set_preferred(o->numa_mem_prefer_node); |
d0b937ed YR |
1429 | break; |
1430 | case MPOL_DEFAULT: | |
1431 | default: | |
1432 | break; | |
1433 | } | |
1434 | ||
43522848 DG |
1435 | if (mask) |
1436 | numa_free_nodemask(mask); | |
1437 | ||
d0b937ed YR |
1438 | } |
1439 | } | |
1440 | #endif | |
1441 | ||
9a3f1100 JA |
1442 | if (fio_pin_memory(td)) |
1443 | goto err; | |
1444 | ||
2e1df07d JA |
1445 | /* |
1446 | * May alter parameters that init_io_u() will use, so we need to | |
1447 | * do this first. | |
1448 | */ | |
1449 | if (init_iolog(td)) | |
1450 | goto err; | |
1451 | ||
1452 | if (init_io_u(td)) | |
1453 | goto err; | |
1454 | ||
4896473e | 1455 | if (o->verify_async && verify_async_init(td)) |
2e1df07d JA |
1456 | goto err; |
1457 | ||
27d74836 JA |
1458 | if (fio_option_is_set(o, ioprio) || |
1459 | fio_option_is_set(o, ioprio_class)) { | |
28727df7 JA |
1460 | ret = ioprio_set(IOPRIO_WHO_PROCESS, 0, o->ioprio_class, o->ioprio); |
1461 | if (ret == -1) { | |
2e1df07d JA |
1462 | td_verror(td, errno, "ioprio_set"); |
1463 | goto err; | |
1464 | } | |
1465 | } | |
1466 | ||
4896473e | 1467 | if (o->cgroup && cgroup_setup(td, cgroup_list, &cgroup_mnt)) |
2e1df07d JA |
1468 | goto err; |
1469 | ||
649c10c5 | 1470 | errno = 0; |
4896473e | 1471 | if (nice(o->nice) == -1 && errno != 0) { |
2e1df07d JA |
1472 | td_verror(td, errno, "nice"); |
1473 | goto err; | |
1474 | } | |
1475 | ||
4896473e | 1476 | if (o->ioscheduler && switch_ioscheduler(td)) |
2e1df07d JA |
1477 | goto err; |
1478 | ||
4896473e | 1479 | if (!o->create_serialize && setup_files(td)) |
2e1df07d JA |
1480 | goto err; |
1481 | ||
1482 | if (td_io_init(td)) | |
1483 | goto err; | |
1484 | ||
1485 | if (init_random_map(td)) | |
1486 | goto err; | |
1487 | ||
ce486495 | 1488 | if (o->exec_prerun && exec_string(o, o->exec_prerun, (const char *)"prerun")) |
4896473e | 1489 | goto err; |
2e1df07d | 1490 | |
4896473e | 1491 | if (o->pre_read) { |
2e1df07d JA |
1492 | if (pre_read_files(td) < 0) |
1493 | goto err; | |
1494 | } | |
1495 | ||
aee2ab67 JA |
1496 | if (td->flags & TD_F_COMPRESS_LOG) |
1497 | tp_init(&td->tp_data); | |
1498 | ||
dc5bfbb2 JA |
1499 | fio_verify_init(td); |
1500 | ||
2e1df07d | 1501 | fio_gettime(&td->epoch, NULL); |
44404c5a | 1502 | fio_getrusage(&td->ru_start); |
2e1df07d JA |
1503 | clear_state = 0; |
1504 | while (keep_running(td)) { | |
100f49f1 JA |
1505 | uint64_t verify_bytes; |
1506 | ||
2e1df07d JA |
1507 | fio_gettime(&td->start, NULL); |
1508 | memcpy(&td->bw_sample_time, &td->start, sizeof(td->start)); | |
1509 | memcpy(&td->iops_sample_time, &td->start, sizeof(td->start)); | |
1510 | memcpy(&td->tv_cache, &td->start, sizeof(td->start)); | |
1511 | ||
4896473e JA |
1512 | if (o->ratemin[DDIR_READ] || o->ratemin[DDIR_WRITE] || |
1513 | o->ratemin[DDIR_TRIM]) { | |
6eaf09d6 | 1514 | memcpy(&td->lastrate[DDIR_READ], &td->bw_sample_time, |
2e1df07d | 1515 | sizeof(td->bw_sample_time)); |
6eaf09d6 SL |
1516 | memcpy(&td->lastrate[DDIR_WRITE], &td->bw_sample_time, |
1517 | sizeof(td->bw_sample_time)); | |
1518 | memcpy(&td->lastrate[DDIR_TRIM], &td->bw_sample_time, | |
2e1df07d JA |
1519 | sizeof(td->bw_sample_time)); |
1520 | } | |
1521 | ||
1522 | if (clear_state) | |
1523 | clear_io_state(td); | |
1524 | ||
1525 | prune_io_piece_log(td); | |
1526 | ||
62167762 JC |
1527 | if (td->o.verify_only && (td_write(td) || td_rw(td))) |
1528 | verify_bytes = do_dry_run(td); | |
1529 | else | |
1530 | verify_bytes = do_io(td); | |
2e1df07d JA |
1531 | |
1532 | clear_state = 1; | |
1533 | ||
40c666c8 JA |
1534 | /* |
1535 | * Make sure we've successfully updated the rusage stats | |
1536 | * before waiting on the stat mutex. Otherwise we could have | |
1537 | * the stat thread holding stat mutex and waiting for | |
1538 | * the rusage_sem, which would never get upped because | |
1539 | * this thread is waiting for the stat mutex. | |
1540 | */ | |
1541 | check_update_rusage(td); | |
1542 | ||
e5437a07 | 1543 | fio_mutex_down(stat_mutex); |
2e1df07d | 1544 | if (td_read(td) && td->io_bytes[DDIR_READ]) { |
e5437a07 | 1545 | elapsed = mtime_since_now(&td->start); |
2e1df07d JA |
1546 | td->ts.runtime[DDIR_READ] += elapsed; |
1547 | } | |
1548 | if (td_write(td) && td->io_bytes[DDIR_WRITE]) { | |
e5437a07 | 1549 | elapsed = mtime_since_now(&td->start); |
2e1df07d JA |
1550 | td->ts.runtime[DDIR_WRITE] += elapsed; |
1551 | } | |
6eaf09d6 | 1552 | if (td_trim(td) && td->io_bytes[DDIR_TRIM]) { |
e5437a07 | 1553 | elapsed = mtime_since_now(&td->start); |
6eaf09d6 SL |
1554 | td->ts.runtime[DDIR_TRIM] += elapsed; |
1555 | } | |
e5437a07 VT |
1556 | fio_gettime(&td->start, NULL); |
1557 | fio_mutex_up(stat_mutex); | |
2e1df07d JA |
1558 | |
1559 | if (td->error || td->terminate) | |
1560 | break; | |
1561 | ||
4896473e JA |
1562 | if (!o->do_verify || |
1563 | o->verify == VERIFY_NONE || | |
2e1df07d JA |
1564 | (td->io_ops->flags & FIO_UNIDIR)) |
1565 | continue; | |
1566 | ||
1567 | clear_io_state(td); | |
1568 | ||
1569 | fio_gettime(&td->start, NULL); | |
1570 | ||
100f49f1 | 1571 | do_verify(td, verify_bytes); |
2e1df07d | 1572 | |
40c666c8 JA |
1573 | /* |
1574 | * See comment further up for why this is done here. | |
1575 | */ | |
1576 | check_update_rusage(td); | |
1577 | ||
e5437a07 VT |
1578 | fio_mutex_down(stat_mutex); |
1579 | td->ts.runtime[DDIR_READ] += mtime_since_now(&td->start); | |
1580 | fio_gettime(&td->start, NULL); | |
1581 | fio_mutex_up(stat_mutex); | |
2e1df07d JA |
1582 | |
1583 | if (td->error || td->terminate) | |
1584 | break; | |
1585 | } | |
1586 | ||
1587 | update_rusage_stat(td); | |
2e1df07d | 1588 | td->ts.total_run_time = mtime_since_now(&td->epoch); |
6eaf09d6 SL |
1589 | td->ts.io_bytes[DDIR_READ] = td->io_bytes[DDIR_READ]; |
1590 | td->ts.io_bytes[DDIR_WRITE] = td->io_bytes[DDIR_WRITE]; | |
1591 | td->ts.io_bytes[DDIR_TRIM] = td->io_bytes[DDIR_TRIM]; | |
2e1df07d | 1592 | |
ca09be4b JA |
1593 | if (td->o.verify_state_save && !(td->flags & TD_F_VSTATE_SAVED) && |
1594 | (td->o.verify != VERIFY_NONE && td_write(td))) { | |
1595 | struct all_io_list *state; | |
1596 | size_t sz; | |
1597 | ||
1598 | state = get_all_io_list(td->thread_number, &sz); | |
1599 | if (state) { | |
1600 | __verify_save_state(state, "local"); | |
1601 | free(state); | |
1602 | } | |
1603 | } | |
1604 | ||
9a3f1100 JA |
1605 | fio_unpin_memory(td); |
1606 | ||
905e3d4f | 1607 | fio_writeout_logs(td); |
2e1df07d | 1608 | |
aee2ab67 JA |
1609 | if (td->flags & TD_F_COMPRESS_LOG) |
1610 | tp_exit(&td->tp_data); | |
1611 | ||
4896473e | 1612 | if (o->exec_postrun) |
ce486495 | 1613 | exec_string(o, o->exec_postrun, (const char *)"postrun"); |
2e1df07d JA |
1614 | |
1615 | if (exitall_on_terminate) | |
1616 | fio_terminate_threads(td->groupid); | |
1617 | ||
1618 | err: | |
1619 | if (td->error) | |
1620 | log_info("fio: pid=%d, err=%d/%s\n", (int) td->pid, td->error, | |
1621 | td->verror); | |
1622 | ||
4896473e | 1623 | if (o->verify_async) |
2e1df07d JA |
1624 | verify_async_exit(td); |
1625 | ||
1626 | close_and_free_files(td); | |
2e1df07d | 1627 | cleanup_io_u(td); |
32dbca2c | 1628 | close_ioengine(td); |
2e1df07d | 1629 | cgroup_shutdown(td, &cgroup_mnt); |
ca09be4b | 1630 | verify_free_state(td); |
2e1df07d | 1631 | |
b2a9e649 | 1632 | if (fio_option_is_set(o, cpumask)) { |
8a1db9a1 JA |
1633 | ret = fio_cpuset_exit(&o->cpumask); |
1634 | if (ret) | |
1635 | td_verror(td, ret, "fio_cpuset_exit"); | |
2e1df07d JA |
1636 | } |
1637 | ||
1638 | /* | |
1639 | * do this very late, it will log file closing as well | |
1640 | */ | |
4896473e | 1641 | if (o->write_iolog_file) |
2e1df07d JA |
1642 | write_iolog_close(td); |
1643 | ||
ea66e04f JA |
1644 | fio_mutex_remove(td->mutex); |
1645 | td->mutex = NULL; | |
1646 | ||
2e1df07d | 1647 | td_set_runstate(td, TD_EXITED); |
fda2cfac JA |
1648 | |
1649 | /* | |
1650 | * Do this last after setting our runstate to exited, so we | |
1651 | * know that the stat thread is signaled. | |
1652 | */ | |
1653 | check_update_rusage(td); | |
1654 | ||
e43606c2 | 1655 | return (void *) (uintptr_t) td->error; |
2e1df07d JA |
1656 | } |
1657 | ||
1658 | ||
1659 | /* | |
1660 | * We cannot pass the td data into a forked process, so attach the td and | |
1661 | * pass it to the thread worker. | |
1662 | */ | |
1663 | static int fork_main(int shmid, int offset) | |
1664 | { | |
1665 | struct thread_data *td; | |
1666 | void *data, *ret; | |
1667 | ||
c8931876 | 1668 | #if !defined(__hpux) && !defined(CONFIG_NO_SHM) |
2e1df07d JA |
1669 | data = shmat(shmid, NULL, 0); |
1670 | if (data == (void *) -1) { | |
1671 | int __err = errno; | |
1672 | ||
1673 | perror("shmat"); | |
1674 | return __err; | |
1675 | } | |
1676 | #else | |
1677 | /* | |
1678 | * HP-UX inherits shm mappings? | |
1679 | */ | |
1680 | data = threads; | |
1681 | #endif | |
1682 | ||
1683 | td = data + offset * sizeof(struct thread_data); | |
1684 | ret = thread_main(td); | |
1685 | shmdt(data); | |
e43606c2 | 1686 | return (int) (uintptr_t) ret; |
2e1df07d JA |
1687 | } |
1688 | ||
cba5460c JA |
1689 | static void dump_td_info(struct thread_data *td) |
1690 | { | |
1691 | log_err("fio: job '%s' hasn't exited in %lu seconds, it appears to " | |
1692 | "be stuck. Doing forceful exit of this job.\n", td->o.name, | |
1693 | (unsigned long) time_since_now(&td->terminate_time)); | |
1694 | } | |
1695 | ||
2e1df07d JA |
1696 | /* |
1697 | * Run over the job map and reap the threads that have exited, if any. | |
1698 | */ | |
1699 | static void reap_threads(unsigned int *nr_running, unsigned int *t_rate, | |
1700 | unsigned int *m_rate) | |
1701 | { | |
1702 | struct thread_data *td; | |
1703 | unsigned int cputhreads, realthreads, pending; | |
1704 | int i, status, ret; | |
1705 | ||
1706 | /* | |
1707 | * reap exited threads (TD_EXITED -> TD_REAPED) | |
1708 | */ | |
1709 | realthreads = pending = cputhreads = 0; | |
1710 | for_each_td(td, i) { | |
1711 | int flags = 0; | |
1712 | ||
1713 | /* | |
1714 | * ->io_ops is NULL for a thread that has closed its | |
1715 | * io engine | |
1716 | */ | |
1717 | if (td->io_ops && !strcmp(td->io_ops->name, "cpuio")) | |
1718 | cputhreads++; | |
1719 | else | |
1720 | realthreads++; | |
1721 | ||
1722 | if (!td->pid) { | |
1723 | pending++; | |
1724 | continue; | |
1725 | } | |
1726 | if (td->runstate == TD_REAPED) | |
1727 | continue; | |
1728 | if (td->o.use_thread) { | |
1729 | if (td->runstate == TD_EXITED) { | |
1730 | td_set_runstate(td, TD_REAPED); | |
1731 | goto reaped; | |
1732 | } | |
1733 | continue; | |
1734 | } | |
1735 | ||
1736 | flags = WNOHANG; | |
1737 | if (td->runstate == TD_EXITED) | |
1738 | flags = 0; | |
1739 | ||
1740 | /* | |
1741 | * check if someone quit or got killed in an unusual way | |
1742 | */ | |
1743 | ret = waitpid(td->pid, &status, flags); | |
1744 | if (ret < 0) { | |
1745 | if (errno == ECHILD) { | |
1746 | log_err("fio: pid=%d disappeared %d\n", | |
1747 | (int) td->pid, td->runstate); | |
a5e371a6 | 1748 | td->sig = ECHILD; |
2e1df07d JA |
1749 | td_set_runstate(td, TD_REAPED); |
1750 | goto reaped; | |
1751 | } | |
1752 | perror("waitpid"); | |
1753 | } else if (ret == td->pid) { | |
1754 | if (WIFSIGNALED(status)) { | |
1755 | int sig = WTERMSIG(status); | |
1756 | ||
36d80bc7 | 1757 | if (sig != SIGTERM && sig != SIGUSR2) |
2e1df07d JA |
1758 | log_err("fio: pid=%d, got signal=%d\n", |
1759 | (int) td->pid, sig); | |
a5e371a6 | 1760 | td->sig = sig; |
2e1df07d JA |
1761 | td_set_runstate(td, TD_REAPED); |
1762 | goto reaped; | |
1763 | } | |
1764 | if (WIFEXITED(status)) { | |
1765 | if (WEXITSTATUS(status) && !td->error) | |
1766 | td->error = WEXITSTATUS(status); | |
1767 | ||
1768 | td_set_runstate(td, TD_REAPED); | |
1769 | goto reaped; | |
1770 | } | |
1771 | } | |
1772 | ||
cba5460c JA |
1773 | /* |
1774 | * If the job is stuck, do a forceful timeout of it and | |
1775 | * move on. | |
1776 | */ | |
1777 | if (td->terminate && | |
1778 | time_since_now(&td->terminate_time) >= FIO_REAP_TIMEOUT) { | |
1779 | dump_td_info(td); | |
1780 | td_set_runstate(td, TD_REAPED); | |
1781 | goto reaped; | |
1782 | } | |
1783 | ||
2e1df07d JA |
1784 | /* |
1785 | * thread is not dead, continue | |
1786 | */ | |
1787 | pending++; | |
1788 | continue; | |
1789 | reaped: | |
1790 | (*nr_running)--; | |
342f4be4 JA |
1791 | (*m_rate) -= ddir_rw_sum(td->o.ratemin); |
1792 | (*t_rate) -= ddir_rw_sum(td->o.rate); | |
2e1df07d JA |
1793 | if (!td->pid) |
1794 | pending--; | |
1795 | ||
1796 | if (td->error) | |
1797 | exit_value++; | |
1798 | ||
1799 | done_secs += mtime_since_now(&td->epoch) / 1000; | |
4a88752a | 1800 | profile_td_exit(td); |
2e1df07d JA |
1801 | } |
1802 | ||
1803 | if (*nr_running == cputhreads && !pending && realthreads) | |
1804 | fio_terminate_threads(TERMINATE_ALL); | |
1805 | } | |
1806 | ||
ca09be4b JA |
1807 | static int __check_trigger_file(void) |
1808 | { | |
1809 | struct stat sb; | |
1810 | ||
1811 | if (!trigger_file) | |
1812 | return 0; | |
1813 | ||
1814 | if (stat(trigger_file, &sb)) | |
1815 | return 0; | |
1816 | ||
1817 | if (unlink(trigger_file) < 0) | |
1818 | log_err("fio: failed to unlink %s: %s\n", trigger_file, | |
1819 | strerror(errno)); | |
1820 | ||
1821 | return 1; | |
1822 | } | |
1823 | ||
1824 | static int trigger_timedout(void) | |
1825 | { | |
1826 | if (trigger_timeout) | |
1827 | return time_since_genesis() >= trigger_timeout; | |
1828 | ||
1829 | return 0; | |
1830 | } | |
1831 | ||
1832 | void exec_trigger(const char *cmd) | |
1833 | { | |
1834 | int ret; | |
1835 | ||
1836 | if (!cmd) | |
1837 | return; | |
1838 | ||
1839 | ret = system(cmd); | |
1840 | if (ret == -1) | |
1841 | log_err("fio: failed executing %s trigger\n", cmd); | |
1842 | } | |
1843 | ||
1844 | void check_trigger_file(void) | |
1845 | { | |
1846 | if (__check_trigger_file() || trigger_timedout()) { | |
796fb3ce JA |
1847 | if (nr_clients) |
1848 | fio_clients_send_trigger(trigger_remote_cmd); | |
1849 | else { | |
ca09be4b JA |
1850 | verify_save_state(); |
1851 | fio_terminate_threads(TERMINATE_ALL); | |
1852 | exec_trigger(trigger_cmd); | |
1853 | } | |
1854 | } | |
1855 | } | |
1856 | ||
1857 | static int fio_verify_load_state(struct thread_data *td) | |
1858 | { | |
1859 | int ret; | |
1860 | ||
1861 | if (!td->o.verify_state) | |
1862 | return 0; | |
1863 | ||
1864 | if (is_backend) { | |
1865 | void *data; | |
1866 | ||
1867 | ret = fio_server_get_verify_state(td->o.name, | |
1868 | td->thread_number - 1, &data); | |
1869 | if (!ret) | |
1870 | verify_convert_assign_state(td, data); | |
1871 | } else | |
1872 | ret = verify_load_state(td, "local"); | |
1873 | ||
1874 | return ret; | |
1875 | } | |
1876 | ||
06464907 JA |
1877 | static void do_usleep(unsigned int usecs) |
1878 | { | |
1879 | check_for_running_stats(); | |
ca09be4b | 1880 | check_trigger_file(); |
06464907 JA |
1881 | usleep(usecs); |
1882 | } | |
1883 | ||
2e1df07d JA |
1884 | /* |
1885 | * Main function for kicking off and reaping jobs, as needed. | |
1886 | */ | |
1887 | static void run_threads(void) | |
1888 | { | |
1889 | struct thread_data *td; | |
2e1df07d | 1890 | unsigned int i, todo, nr_running, m_rate, t_rate, nr_started; |
0de5b26f | 1891 | uint64_t spent; |
2e1df07d | 1892 | |
2e1df07d JA |
1893 | if (fio_gtod_offload && fio_start_gtod_thread()) |
1894 | return; | |
334185e9 | 1895 | |
f2a2ce0e | 1896 | fio_idle_prof_init(); |
2e1df07d JA |
1897 | |
1898 | set_sig_handlers(); | |
1899 | ||
3a5f6bde JA |
1900 | nr_thread = nr_process = 0; |
1901 | for_each_td(td, i) { | |
1902 | if (td->o.use_thread) | |
1903 | nr_thread++; | |
1904 | else | |
1905 | nr_process++; | |
1906 | } | |
1907 | ||
f3afa57e | 1908 | if (output_format == FIO_OUTPUT_NORMAL) { |
2e1df07d JA |
1909 | log_info("Starting "); |
1910 | if (nr_thread) | |
1911 | log_info("%d thread%s", nr_thread, | |
1912 | nr_thread > 1 ? "s" : ""); | |
1913 | if (nr_process) { | |
1914 | if (nr_thread) | |
1915 | log_info(" and "); | |
1916 | log_info("%d process%s", nr_process, | |
1917 | nr_process > 1 ? "es" : ""); | |
1918 | } | |
1919 | log_info("\n"); | |
e411c301 | 1920 | log_info_flush(); |
2e1df07d JA |
1921 | } |
1922 | ||
1923 | todo = thread_number; | |
1924 | nr_running = 0; | |
1925 | nr_started = 0; | |
1926 | m_rate = t_rate = 0; | |
1927 | ||
1928 | for_each_td(td, i) { | |
1929 | print_status_init(td->thread_number - 1); | |
1930 | ||
1931 | if (!td->o.create_serialize) | |
1932 | continue; | |
1933 | ||
ca09be4b JA |
1934 | if (fio_verify_load_state(td)) |
1935 | goto reap; | |
1936 | ||
2e1df07d JA |
1937 | /* |
1938 | * do file setup here so it happens sequentially, | |
1939 | * we don't want X number of threads getting their | |
1940 | * client data interspersed on disk | |
1941 | */ | |
1942 | if (setup_files(td)) { | |
ca09be4b | 1943 | reap: |
2e1df07d JA |
1944 | exit_value++; |
1945 | if (td->error) | |
1946 | log_err("fio: pid=%d, err=%d/%s\n", | |
1947 | (int) td->pid, td->error, td->verror); | |
1948 | td_set_runstate(td, TD_REAPED); | |
1949 | todo--; | |
1950 | } else { | |
1951 | struct fio_file *f; | |
1952 | unsigned int j; | |
1953 | ||
1954 | /* | |
1955 | * for sharing to work, each job must always open | |
1956 | * its own files. so close them, if we opened them | |
1957 | * for creation | |
1958 | */ | |
1959 | for_each_file(td, f, j) { | |
1960 | if (fio_file_open(f)) | |
1961 | td_io_close_file(td, f); | |
1962 | } | |
1963 | } | |
1964 | } | |
1965 | ||
f2a2ce0e HL |
1966 | /* start idle threads before io threads start to run */ |
1967 | fio_idle_prof_start(); | |
1968 | ||
2e1df07d JA |
1969 | set_genesis_time(); |
1970 | ||
1971 | while (todo) { | |
1972 | struct thread_data *map[REAL_MAX_JOBS]; | |
1973 | struct timeval this_start; | |
1974 | int this_jobs = 0, left; | |
1975 | ||
1976 | /* | |
1977 | * create threads (TD_NOT_CREATED -> TD_CREATED) | |
1978 | */ | |
1979 | for_each_td(td, i) { | |
1980 | if (td->runstate != TD_NOT_CREATED) | |
1981 | continue; | |
1982 | ||
1983 | /* | |
1984 | * never got a chance to start, killed by other | |
1985 | * thread for some reason | |
1986 | */ | |
1987 | if (td->terminate) { | |
1988 | todo--; | |
1989 | continue; | |
1990 | } | |
1991 | ||
1992 | if (td->o.start_delay) { | |
0de5b26f | 1993 | spent = utime_since_genesis(); |
2e1df07d | 1994 | |
74454ce4 | 1995 | if (td->o.start_delay > spent) |
2e1df07d JA |
1996 | continue; |
1997 | } | |
1998 | ||
1999 | if (td->o.stonewall && (nr_started || nr_running)) { | |
2000 | dprint(FD_PROCESS, "%s: stonewall wait\n", | |
2001 | td->o.name); | |
2002 | break; | |
2003 | } | |
2004 | ||
2005 | init_disk_util(td); | |
2006 | ||
c97f1ad6 JA |
2007 | td->rusage_sem = fio_mutex_init(FIO_MUTEX_LOCKED); |
2008 | td->update_rusage = 0; | |
2009 | ||
2e1df07d JA |
2010 | /* |
2011 | * Set state to created. Thread will transition | |
2012 | * to TD_INITIALIZED when it's done setting up. | |
2013 | */ | |
2014 | td_set_runstate(td, TD_CREATED); | |
2015 | map[this_jobs++] = td; | |
2016 | nr_started++; | |
2017 | ||
2018 | if (td->o.use_thread) { | |
2019 | int ret; | |
2020 | ||
2021 | dprint(FD_PROCESS, "will pthread_create\n"); | |
2022 | ret = pthread_create(&td->thread, NULL, | |
2023 | thread_main, td); | |
2024 | if (ret) { | |
2025 | log_err("pthread_create: %s\n", | |
2026 | strerror(ret)); | |
2027 | nr_started--; | |
2028 | break; | |
2029 | } | |
2030 | ret = pthread_detach(td->thread); | |
2031 | if (ret) | |
2032 | log_err("pthread_detach: %s", | |
2033 | strerror(ret)); | |
2034 | } else { | |
2035 | pid_t pid; | |
2036 | dprint(FD_PROCESS, "will fork\n"); | |
2037 | pid = fork(); | |
2038 | if (!pid) { | |
2039 | int ret = fork_main(shm_id, i); | |
2040 | ||
2041 | _exit(ret); | |
2042 | } else if (i == fio_debug_jobno) | |
2043 | *fio_debug_jobp = pid; | |
2044 | } | |
2045 | dprint(FD_MUTEX, "wait on startup_mutex\n"); | |
2046 | if (fio_mutex_down_timeout(startup_mutex, 10)) { | |
2047 | log_err("fio: job startup hung? exiting.\n"); | |
2048 | fio_terminate_threads(TERMINATE_ALL); | |
2049 | fio_abort = 1; | |
2050 | nr_started--; | |
2051 | break; | |
2052 | } | |
2053 | dprint(FD_MUTEX, "done waiting on startup_mutex\n"); | |
2054 | } | |
2055 | ||
2056 | /* | |
2057 | * Wait for the started threads to transition to | |
2058 | * TD_INITIALIZED. | |
2059 | */ | |
2060 | fio_gettime(&this_start, NULL); | |
2061 | left = this_jobs; | |
2062 | while (left && !fio_abort) { | |
2063 | if (mtime_since_now(&this_start) > JOB_START_TIMEOUT) | |
2064 | break; | |
2065 | ||
06464907 | 2066 | do_usleep(100000); |
2e1df07d JA |
2067 | |
2068 | for (i = 0; i < this_jobs; i++) { | |
2069 | td = map[i]; | |
2070 | if (!td) | |
2071 | continue; | |
2072 | if (td->runstate == TD_INITIALIZED) { | |
2073 | map[i] = NULL; | |
2074 | left--; | |
2075 | } else if (td->runstate >= TD_EXITED) { | |
2076 | map[i] = NULL; | |
2077 | left--; | |
2078 | todo--; | |
2079 | nr_running++; /* work-around... */ | |
2080 | } | |
2081 | } | |
2082 | } | |
2083 | ||
2084 | if (left) { | |
4e87c37a JA |
2085 | log_err("fio: %d job%s failed to start\n", left, |
2086 | left > 1 ? "s" : ""); | |
2e1df07d JA |
2087 | for (i = 0; i < this_jobs; i++) { |
2088 | td = map[i]; | |
2089 | if (!td) | |
2090 | continue; | |
2091 | kill(td->pid, SIGTERM); | |
2092 | } | |
2093 | break; | |
2094 | } | |
2095 | ||
2096 | /* | |
2097 | * start created threads (TD_INITIALIZED -> TD_RUNNING). | |
2098 | */ | |
2099 | for_each_td(td, i) { | |
2100 | if (td->runstate != TD_INITIALIZED) | |
2101 | continue; | |
2102 | ||
2103 | if (in_ramp_time(td)) | |
2104 | td_set_runstate(td, TD_RAMP); | |
2105 | else | |
2106 | td_set_runstate(td, TD_RUNNING); | |
2107 | nr_running++; | |
2108 | nr_started--; | |
342f4be4 JA |
2109 | m_rate += ddir_rw_sum(td->o.ratemin); |
2110 | t_rate += ddir_rw_sum(td->o.rate); | |
2e1df07d JA |
2111 | todo--; |
2112 | fio_mutex_up(td->mutex); | |
2113 | } | |
2114 | ||
2115 | reap_threads(&nr_running, &t_rate, &m_rate); | |
2116 | ||
122c7725 | 2117 | if (todo) |
06464907 | 2118 | do_usleep(100000); |
2e1df07d JA |
2119 | } |
2120 | ||
2121 | while (nr_running) { | |
2122 | reap_threads(&nr_running, &t_rate, &m_rate); | |
06464907 | 2123 | do_usleep(10000); |
2e1df07d JA |
2124 | } |
2125 | ||
f2a2ce0e HL |
2126 | fio_idle_prof_stop(); |
2127 | ||
2e1df07d | 2128 | update_io_ticks(); |
2e1df07d JA |
2129 | } |
2130 | ||
5ddc6707 | 2131 | static void wait_for_helper_thread_exit(void) |
9ec7779f | 2132 | { |
8aab824f JA |
2133 | void *ret; |
2134 | ||
98b4b0a2 | 2135 | helper_exit = 1; |
5ddc6707 JA |
2136 | pthread_cond_signal(&helper_cond); |
2137 | pthread_join(helper_thread, &ret); | |
9ec7779f JA |
2138 | } |
2139 | ||
27357187 JA |
2140 | static void free_disk_util(void) |
2141 | { | |
27357187 | 2142 | disk_util_prune_entries(); |
8aab824f | 2143 | |
5ddc6707 | 2144 | pthread_cond_destroy(&helper_cond); |
27357187 JA |
2145 | } |
2146 | ||
5ddc6707 | 2147 | static void *helper_thread_main(void *data) |
2e1df07d | 2148 | { |
9ec7779f JA |
2149 | int ret = 0; |
2150 | ||
2e1df07d JA |
2151 | fio_mutex_up(startup_mutex); |
2152 | ||
8aab824f JA |
2153 | while (!ret) { |
2154 | uint64_t sec = DISK_UTIL_MSEC / 1000; | |
2155 | uint64_t nsec = (DISK_UTIL_MSEC % 1000) * 1000000; | |
2156 | struct timespec ts; | |
2157 | struct timeval tv; | |
2158 | ||
2159 | gettimeofday(&tv, NULL); | |
2160 | ts.tv_sec = tv.tv_sec + sec; | |
2161 | ts.tv_nsec = (tv.tv_usec * 1000) + nsec; | |
f893b76d | 2162 | |
823952af | 2163 | if (ts.tv_nsec >= 1000000000ULL) { |
8aab824f JA |
2164 | ts.tv_nsec -= 1000000000ULL; |
2165 | ts.tv_sec++; | |
2166 | } | |
2167 | ||
5ddc6707 | 2168 | pthread_cond_timedwait(&helper_cond, &helper_lock, &ts); |
8aab824f | 2169 | |
9ec7779f | 2170 | ret = update_io_ticks(); |
2e1df07d | 2171 | |
5ddc6707 JA |
2172 | if (helper_do_stat) { |
2173 | helper_do_stat = 0; | |
2174 | __show_running_run_stats(); | |
2175 | } | |
2176 | ||
2e1df07d JA |
2177 | if (!is_backend) |
2178 | print_thread_status(); | |
2179 | } | |
2180 | ||
2181 | return NULL; | |
2182 | } | |
2183 | ||
5ddc6707 | 2184 | static int create_helper_thread(void) |
2e1df07d JA |
2185 | { |
2186 | int ret; | |
2187 | ||
9ec7779f JA |
2188 | setup_disk_util(); |
2189 | ||
5ddc6707 JA |
2190 | pthread_cond_init(&helper_cond, NULL); |
2191 | pthread_mutex_init(&helper_lock, NULL); | |
8aab824f | 2192 | |
5ddc6707 | 2193 | ret = pthread_create(&helper_thread, NULL, helper_thread_main, NULL); |
2e1df07d | 2194 | if (ret) { |
5ddc6707 | 2195 | log_err("Can't create helper thread: %s\n", strerror(ret)); |
2e1df07d JA |
2196 | return 1; |
2197 | } | |
2198 | ||
2e1df07d JA |
2199 | dprint(FD_MUTEX, "wait on startup_mutex\n"); |
2200 | fio_mutex_down(startup_mutex); | |
2201 | dprint(FD_MUTEX, "done waiting on startup_mutex\n"); | |
2202 | return 0; | |
2203 | } | |
2204 | ||
2e1df07d JA |
2205 | int fio_backend(void) |
2206 | { | |
2207 | struct thread_data *td; | |
2208 | int i; | |
2209 | ||
2210 | if (exec_profile) { | |
2211 | if (load_profile(exec_profile)) | |
2212 | return 1; | |
2213 | free(exec_profile); | |
2214 | exec_profile = NULL; | |
2215 | } | |
2216 | if (!thread_number) | |
2217 | return 0; | |
2218 | ||
2219 | if (write_bw_log) { | |
aee2ab67 JA |
2220 | struct log_params p = { |
2221 | .log_type = IO_LOG_TYPE_BW, | |
2222 | }; | |
2223 | ||
2224 | setup_log(&agg_io_log[DDIR_READ], &p, "agg-read_bw.log"); | |
2225 | setup_log(&agg_io_log[DDIR_WRITE], &p, "agg-write_bw.log"); | |
2226 | setup_log(&agg_io_log[DDIR_TRIM], &p, "agg-trim_bw.log"); | |
2e1df07d JA |
2227 | } |
2228 | ||
521da527 | 2229 | startup_mutex = fio_mutex_init(FIO_MUTEX_LOCKED); |
2e1df07d JA |
2230 | if (startup_mutex == NULL) |
2231 | return 1; | |
2e1df07d JA |
2232 | |
2233 | set_genesis_time(); | |
cef9175e | 2234 | stat_init(); |
5ddc6707 | 2235 | create_helper_thread(); |
2e1df07d JA |
2236 | |
2237 | cgroup_list = smalloc(sizeof(*cgroup_list)); | |
2238 | INIT_FLIST_HEAD(cgroup_list); | |
2239 | ||
2240 | run_threads(); | |
2241 | ||
5ddc6707 | 2242 | wait_for_helper_thread_exit(); |
8aab824f | 2243 | |
2e1df07d | 2244 | if (!fio_abort) { |
83f7b64e | 2245 | __show_run_stats(); |
2e1df07d | 2246 | if (write_bw_log) { |
cb7e0ace JA |
2247 | for (i = 0; i < DDIR_RWDIR_CNT; i++) { |
2248 | struct io_log *log = agg_io_log[i]; | |
2249 | ||
aee2ab67 | 2250 | flush_log(log); |
518dac09 | 2251 | free_log(log); |
cb7e0ace | 2252 | } |
2e1df07d JA |
2253 | } |
2254 | } | |
2255 | ||
fda2cfac | 2256 | for_each_td(td, i) { |
2e1df07d | 2257 | fio_options_free(td); |
8049adc1 JA |
2258 | if (td->rusage_sem) { |
2259 | fio_mutex_remove(td->rusage_sem); | |
2260 | td->rusage_sem = NULL; | |
2261 | } | |
fda2cfac | 2262 | } |
2e1df07d | 2263 | |
a462baef | 2264 | free_disk_util(); |
2e1df07d JA |
2265 | cgroup_kill(cgroup_list); |
2266 | sfree(cgroup_list); | |
2267 | sfree(cgroup_mnt); | |
2268 | ||
2269 | fio_mutex_remove(startup_mutex); | |
cef9175e | 2270 | stat_exit(); |
2e1df07d JA |
2271 | return exit_value; |
2272 | } |