blob: fdb7413d7719ff9ff4b46af8e611c072c6918435 [file] [log] [blame]
Jens Axboe2e1df072012-02-09 11:15:02 +01001/*
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>
Bruce Crane43606c2012-02-20 09:34:24 +010033#include <inttypes.h>
Jens Axboe2e1df072012-02-09 11:15:02 +010034#include <sys/stat.h>
35#include <sys/wait.h>
36#include <sys/ipc.h>
Jens Axboe2e1df072012-02-09 11:15:02 +010037#include <sys/mman.h>
38
39#include "fio.h"
Olega5e0ee12013-03-12 00:06:53 -070040#ifndef FIO_NO_HAVE_SHM_H
41#include <sys/shm.h>
42#endif
Jens Axboe2e1df072012-02-09 11:15:02 +010043#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"
Jens Axboe44404c52013-01-24 14:20:09 -070053#include "lib/getrusage.h"
Huadong Liuf2a2ce02013-01-30 13:22:24 +010054#include "idletime.h"
Jens Axboe002fe732014-02-11 08:31:13 -070055#include "err.h"
Jens Axboe65f93ca2014-07-03 14:16:25 -060056#include "lib/tp.h"
Jens Axboe2e1df072012-02-09 11:15:02 +010057
Jens Axboe7fe36312014-10-23 23:16:50 -060058static pthread_t helper_thread;
59static pthread_mutex_t helper_lock;
60pthread_cond_t helper_cond;
61int helper_do_stat = 0;
Jens Axboe62244c92014-10-23 08:53:33 -060062
Jens Axboe2e1df072012-02-09 11:15:02 +010063static struct fio_mutex *startup_mutex;
Jens Axboe2e1df072012-02-09 11:15:02 +010064static struct flist_head *cgroup_list;
65static char *cgroup_mnt;
66static int exit_value;
67static volatile int fio_abort;
Jens Axboe3a5f6bd2013-04-11 12:04:38 +020068static unsigned int nr_process = 0;
69static unsigned int nr_thread = 0;
Jens Axboe2e1df072012-02-09 11:15:02 +010070
Shaohua Li6eaf09d2012-09-14 08:49:43 +020071struct io_log *agg_io_log[DDIR_RWDIR_CNT];
Jens Axboe2e1df072012-02-09 11:15:02 +010072
Jens Axboea3efc912012-02-09 11:25:24 +010073int groupid = 0;
74unsigned int thread_number = 0;
Jens Axboe108fea72012-11-14 13:09:45 -070075unsigned int stat_number = 0;
Jens Axboea3efc912012-02-09 11:25:24 +010076int shm_id = 0;
77int temp_stall_ts;
78unsigned long done_secs = 0;
Jens Axboe7fe36312014-10-23 23:16:50 -060079volatile int helper_exit = 0;
Jens Axboea3efc912012-02-09 11:25:24 +010080
Jens Axboe2e1df072012-02-09 11:15:02 +010081#define PAGE_ALIGN(buf) \
Bruce Crane43606c2012-02-20 09:34:24 +010082 (char *) (((uintptr_t) (buf) + page_mask) & ~page_mask)
Jens Axboe2e1df072012-02-09 11:15:02 +010083
84#define JOB_START_TIMEOUT (5 * 1000)
85
86static 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);
Jens Axboebc2e69c2014-10-13 11:57:15 -060093 log_info_flush();
Jens Axboe2e1df072012-02-09 11:15:02 +010094 exit_value = 128;
95 }
96
97 fio_terminate_threads(TERMINATE_ALL);
98 }
99}
100
Jens Axboe4c6d91e2012-03-30 10:30:35 +0200101static void sig_show_status(int sig)
102{
103 show_running_run_stats();
104}
105
Jens Axboe2e1df072012-02-09 11:15:02 +0100106static 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
Bruce Cran2f694502012-10-10 16:34:13 +0100120/* 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
Jens Axboe4c6d91e2012-03-30 10:30:35 +0200128 memset(&act, 0, sizeof(act));
129 act.sa_handler = sig_show_status;
130 act.sa_flags = SA_RESTART;
131 sigaction(SIGUSR1, &act, NULL);
132
Jens Axboe2e1df072012-02-09 11:15:02 +0100133 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 */
144static 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 {
Jens Axboe49cba9b2014-04-14 09:51:10 -0600189 if (spent)
190 rate = ((bytes - td->rate_bytes[ddir]) * 1000) / spent;
191 else
192 rate = 0;
193
Jens Axboe2e1df072012-02-09 11:15:02 +0100194 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 {
Jens Axboe4c707a32014-04-15 08:26:57 -0600211 if (spent)
212 rate = ((iops - td->rate_blocks[ddir]) * 1000) / spent;
213 else
214 rate = 0;
215
Jens Axboe2e1df072012-02-09 11:15:02 +0100216 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
232static int check_min_rate(struct thread_data *td, struct timeval *now,
Jens Axboe100f49f2013-01-23 10:15:57 -0700233 uint64_t *bytes_done)
Jens Axboe2e1df072012-02-09 11:15:02 +0100234{
235 int ret = 0;
236
Shaohua Li6eaf09d2012-09-14 08:49:43 +0200237 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);
Jens Axboe2e1df072012-02-09 11:15:02 +0100243
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 */
251static void cleanup_pending_aio(struct thread_data *td)
252{
Jens Axboe2e1df072012-02-09 11:15:02 +0100253 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) {
Jens Axboe2ae0b202013-05-28 14:16:55 +0200266 struct io_u *io_u;
267 int i;
Jens Axboe2e1df072012-02-09 11:15:02 +0100268
Jens Axboe2ae0b202013-05-28 14:16:55 +0200269 io_u_qiter(&td->io_u_all, io_u, i) {
270 if (io_u->flags & IO_U_F_FLIGHT) {
Jens Axboe2e1df072012-02-09 11:15:02 +0100271 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 */
286static 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
302requeue:
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}
Jens Axboea3efc912012-02-09 11:25:24 +0100327
Jens Axboe61ee0f82013-02-11 11:26:55 +0100328static 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
Jens Axboe2e1df072012-02-09 11:15:02 +0100343static inline void __update_tv_cache(struct thread_data *td)
344{
345 fio_gettime(&td->tv_cache, NULL);
346}
347
348static 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
354static 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;
Jens Axboe0de5b262014-02-21 15:26:01 -0800360 if (utime_since(&td->epoch, t) >= td->o.timeout)
Jens Axboe2e1df072012-02-09 11:15:02 +0100361 return 1;
362
363 return 0;
364}
365
366static 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) {
Dmitry Monakhov8b28bd42012-09-23 15:46:09 +0400372 int err = td->error;
373 enum error_type_bit eb;
Jens Axboe2e1df072012-02-09 11:15:02 +0100374
375 if (ret < 0)
376 err = -ret;
Jens Axboe2e1df072012-02-09 11:15:02 +0100377
Dmitry Monakhov8b28bd42012-09-23 15:46:09 +0400378 eb = td_error_type(ddir, err);
379 if (!(td->o.continue_on_error & (1 << eb)))
Jens Axboe2e1df072012-02-09 11:15:02 +0100380 return 1;
381
Dmitry Monakhov8b28bd42012-09-23 15:46:09 +0400382 if (td_non_fatal_error(td, eb, err)) {
Jens Axboe2e1df072012-02-09 11:15:02 +0100383 /*
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);
Jens Axboe3b44e8b2014-07-09 11:24:12 +0200397 fio_mark_td_terminate(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100398 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
Jens Axboec97f1ad2013-03-28 15:08:49 -0600412static 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
Jens Axboe24528e72014-12-12 08:49:12 -0700421static 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
Jens Axboeb2c52da2014-12-12 11:34:28 -0700435 if (time && (__should_check_rate(td, DDIR_READ) ||
Jens Axboe24528e72014-12-12 08:49:12 -0700436 __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
Jens Axboe2e1df072012-02-09 11:15:02 +0100449/*
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 */
Jens Axboe100f49f2013-01-23 10:15:57 -0700453static void do_verify(struct thread_data *td, uint64_t verify_bytes)
Jens Axboe2e1df072012-02-09 11:15:02 +0100454{
Jens Axboe100f49f2013-01-23 10:15:57 -0700455 uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 };
Jens Axboe2e1df072012-02-09 11:15:02 +0100456 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
Jens Axboec97f1ad2013-03-28 15:08:49 -0600476 check_update_rusage(td);
477
Jens Axboe2e1df072012-02-09 11:15:02 +0100478 if (td->error)
479 return;
480
481 td_set_runstate(td, TD_VERIFYING);
482
483 io_u = NULL;
484 while (!td->terminate) {
Jens Axboefbccf462013-01-08 21:02:14 +0100485 enum fio_ddir ddir;
Jens Axboe2e1df072012-02-09 11:15:02 +0100486 int ret2, full;
487
488 update_tv_cache(td);
Jens Axboec97f1ad2013-03-28 15:08:49 -0600489 check_update_rusage(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100490
491 if (runtime_exceeded(td, &td->tv_cache)) {
492 __update_tv_cache(td);
493 if (runtime_exceeded(td, &td->tv_cache)) {
Jens Axboe3b44e8b2014-07-09 11:24:12 +0200494 fio_mark_td_terminate(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100495 break;
496 }
497 }
498
Dan Ehrenberg9e684a42012-02-20 11:05:14 +0100499 if (flow_threshold_exceeded(td))
500 continue;
501
Jens Axboe44cbc6d2013-01-21 09:47:03 -0700502 if (!td->o.experimental_verify) {
503 io_u = __get_io_u(td);
504 if (!io_u)
505 break;
Jens Axboe2e1df072012-02-09 11:15:02 +0100506
Jens Axboe44cbc6d2013-01-21 09:47:03 -0700507 if (get_next_verify(td, io_u)) {
508 put_io_u(td, io_u);
509 break;
510 }
Jens Axboe2e1df072012-02-09 11:15:02 +0100511
Jens Axboe44cbc6d2013-01-21 09:47:03 -0700512 if (td_io_prep(td, io_u)) {
513 put_io_u(td, io_u);
514 break;
515 }
516 } else {
Jens Axboe100f49f2013-01-23 10:15:57 -0700517 if (ddir_rw_sum(bytes_done) + td->o.rw_min_bs > verify_bytes)
518 break;
519
Jens Axboebcd5abf2013-01-23 09:27:25 -0700520 while ((io_u = get_io_u(td)) != NULL) {
Jens Axboe002fe732014-02-11 08:31:13 -0700521 if (IS_ERR(io_u)) {
522 io_u = NULL;
523 ret = FIO_Q_BUSY;
524 goto reap;
525 }
526
Jens Axboebcd5abf2013-01-23 09:27:25 -0700527 /*
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 }
Jens Axboe44cbc6d2013-01-21 09:47:03 -0700552
Jens Axboebcd5abf2013-01-23 09:27:25 -0700553 if (!io_u)
Jens Axboe44cbc6d2013-01-21 09:47:03 -0700554 break;
Jens Axboe2e1df072012-02-09 11:15:02 +0100555 }
556
Jens Axboede54cfd2014-11-10 20:34:00 -0700557 if (verify_state_should_stop(td, io_u)) {
558 put_io_u(td, io_u);
559 break;
560 }
561
Jens Axboe2e1df072012-02-09 11:15:02 +0100562 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
Jens Axboefbccf462013-01-08 21:02:14 +0100567 ddir = io_u->ddir;
Gwendal Grignou73187622015-02-08 19:53:20 -0800568 if (!td->o.disable_slat)
569 fio_gettime(&io_u->start_time, NULL);
Jens Axboefbccf462013-01-08 21:02:14 +0100570
Jens Axboe2e1df072012-02-09 11:15:02 +0100571 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 {
602sync_done:
Jens Axboe100f49f2013-01-23 10:15:57 -0700603 ret = io_u_sync_complete(td, io_u, bytes_done);
Jens Axboe2e1df072012-02-09 11:15:02 +0100604 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
Jens Axboefbccf462013-01-08 21:02:14 +0100622 if (break_on_this_error(td, ddir, &ret))
Jens Axboe2e1df072012-02-09 11:15:02 +0100623 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 */
Jens Axboe002fe732014-02-11 08:31:13 -0700630reap:
Jens Axboe2e1df072012-02-09 11:15:02 +0100631 full = queue_full(td) || (ret == FIO_Q_BUSY && td->cur_depth);
Jens Axboe24528e72014-12-12 08:49:12 -0700632 if (full || !td->o.iodepth_batch_complete)
633 ret = wait_for_completions(td, NULL, bytes_done);
Jens Axboe2e1df072012-02-09 11:15:02 +0100634
Jens Axboe2e1df072012-02-09 11:15:02 +0100635 if (ret < 0)
636 break;
637 }
638
Jens Axboec97f1ad2013-03-28 15:08:49 -0600639 check_update_rusage(td);
640
Jens Axboe2e1df072012-02-09 11:15:02 +0100641 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
Jens Axboe3939fe82014-03-11 20:16:38 -0600654static 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
Jens Axboe842d2e22015-02-24 14:50:57 -0800661 number_ios = ddir_rw_sum(td->io_blocks);
Jens Axboe3939fe82014-03-11 20:16:38 -0600662 number_ios += td->io_u_queued + td->io_u_in_flight;
663
Jens Axboe842d2e22015-02-24 14:50:57 -0800664 return number_ios >= (td->o.number_ios * td->loops);
Jens Axboe3939fe82014-03-11 20:16:38 -0600665}
666
Justin Eno1b7ab802015-01-28 14:13:28 -0800667static int io_issue_bytes_exceeded(struct thread_data *td)
Jens Axboef7078f72012-03-07 09:24:05 +0100668{
Jens Axboe77731b22014-04-28 12:08:47 -0600669 unsigned long long bytes, limit;
Jens Axboef7078f72012-03-07 09:24:05 +0100670
671 if (td_rw(td))
Jens Axboe5d61d112015-01-16 10:03:11 -0700672 bytes = td->io_issue_bytes[DDIR_READ] + td->io_issue_bytes[DDIR_WRITE];
Jens Axboef7078f72012-03-07 09:24:05 +0100673 else if (td_write(td))
Jens Axboe5d61d112015-01-16 10:03:11 -0700674 bytes = td->io_issue_bytes[DDIR_WRITE];
Shaohua Li6eaf09d2012-09-14 08:49:43 +0200675 else if (td_read(td))
Jens Axboe5d61d112015-01-16 10:03:11 -0700676 bytes = td->io_issue_bytes[DDIR_READ];
Jens Axboef7078f72012-03-07 09:24:05 +0100677 else
Jens Axboe5d61d112015-01-16 10:03:11 -0700678 bytes = td->io_issue_bytes[DDIR_TRIM];
Jens Axboef7078f72012-03-07 09:24:05 +0100679
Jens Axboe77731b22014-04-28 12:08:47 -0600680 if (td->o.io_limit)
681 limit = td->o.io_limit;
682 else
683 limit = td->o.size;
684
Jens Axboe842d2e22015-02-24 14:50:57 -0800685 limit *= td->loops;
Jens Axboe77731b22014-04-28 12:08:47 -0600686 return bytes >= limit || exceeds_number_ios(td);
Jens Axboef7078f72012-03-07 09:24:05 +0100687}
688
Justin Eno1b7ab802015-01-28 14:13:28 -0800689static 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
Jens Axboe842d2e22015-02-24 14:50:57 -0800707 limit *= td->loops;
Justin Eno1b7ab802015-01-28 14:13:28 -0800708 return bytes >= limit || exceeds_number_ios(td);
709}
710
Jens Axboe2e1df072012-02-09 11:15:02 +0100711/*
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.
Jens Axboe100f49f2013-01-23 10:15:57 -0700714 *
715 * Returns number of bytes written and trimmed.
Jens Axboe2e1df072012-02-09 11:15:02 +0100716 */
Jens Axboe100f49f2013-01-23 10:15:57 -0700717static uint64_t do_io(struct thread_data *td)
Jens Axboe2e1df072012-02-09 11:15:02 +0100718{
Jens Axboe100f49f2013-01-23 10:15:57 -0700719 uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 };
Jens Axboe2e1df072012-02-09 11:15:02 +0100720 unsigned int i;
721 int ret = 0;
Jens Axboec2703bf2014-02-05 12:36:02 -0700722 uint64_t total_bytes, bytes_issued = 0;
Jens Axboe2e1df072012-02-09 11:15:02 +0100723
724 if (in_ramp_time(td))
725 td_set_runstate(td, TD_RAMP);
726 else
727 td_set_runstate(td, TD_RUNNING);
728
Jens Axboe3e260a42013-12-09 12:38:53 -0700729 lat_target_init(td);
730
Justin Eno1b357402015-01-29 12:27:58 -0800731 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);
Jens Axboe78a64692014-02-05 13:15:22 -0700738 /*
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 */
Jens Axboe78a64692014-02-05 13:15:22 -0700743 if (td->o.verify != VERIFY_NONE &&
744 (td_write(td) && td->o.verify_backlog))
Jens Axboec2703bf2014-02-05 12:36:02 -0700745 total_bytes += td->o.size;
746
Jens Axboef7078f72012-03-07 09:24:05 +0100747 while ((td->o.read_iolog_file && !flist_empty(&td->io_log_list)) ||
Justin Eno1b7ab802015-01-28 14:13:28 -0800748 (!flist_empty(&td->trim_list)) || !io_issue_bytes_exceeded(td) ||
Daniel Ehrenbergc04e4662012-03-16 18:54:15 +0100749 td->o.time_based) {
Jens Axboe2e1df072012-02-09 11:15:02 +0100750 struct timeval comp_time;
Jens Axboe2e1df072012-02-09 11:15:02 +0100751 struct io_u *io_u;
752 int ret2, full;
753 enum fio_ddir ddir;
754
Jens Axboec97f1ad2013-03-28 15:08:49 -0600755 check_update_rusage(td);
756
Yufei Ren7d7803f2012-10-22 21:29:38 -0400757 if (td->terminate || td->done)
Jens Axboe2e1df072012-02-09 11:15:02 +0100758 break;
759
760 update_tv_cache(td);
761
762 if (runtime_exceeded(td, &td->tv_cache)) {
763 __update_tv_cache(td);
764 if (runtime_exceeded(td, &td->tv_cache)) {
Jens Axboe3b44e8b2014-07-09 11:24:12 +0200765 fio_mark_td_terminate(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100766 break;
767 }
768 }
769
Dan Ehrenberg9e684a42012-02-20 11:05:14 +0100770 if (flow_threshold_exceeded(td))
771 continue;
772
Jens Axboec2703bf2014-02-05 12:36:02 -0700773 if (bytes_issued >= total_bytes)
Juan Casse20876c52013-09-16 09:22:10 -0700774 break;
775
Jens Axboe2e1df072012-02-09 11:15:02 +0100776 io_u = get_io_u(td);
Jens Axboe002fe732014-02-11 08:31:13 -0700777 if (IS_ERR_OR_NULL(io_u)) {
778 int err = PTR_ERR(io_u);
779
780 io_u = NULL;
781 if (err == -EBUSY) {
782 ret = FIO_Q_BUSY;
783 goto reap;
784 }
Jens Axboe3e260a42013-12-09 12:38:53 -0700785 if (td->o.latency_target)
786 goto reap;
Jens Axboe2e1df072012-02-09 11:15:02 +0100787 break;
Jens Axboe3e260a42013-12-09 12:38:53 -0700788 }
Jens Axboe2e1df072012-02-09 11:15:02 +0100789
790 ddir = io_u->ddir;
791
792 /*
Jens Axboe82af2a72012-03-13 13:45:58 +0100793 * Add verification end_io handler if:
794 * - Asked to verify (!td_rw(td))
795 * - Or the io_u is from our verify list (mixed write/ver)
Jens Axboe2e1df072012-02-09 11:15:02 +0100796 */
797 if (td->o.verify != VERIFY_NONE && io_u->ddir == DDIR_READ &&
Jens Axboe82af2a72012-03-13 13:45:58 +0100798 ((io_u->flags & IO_U_F_VER_LIST) || !td_rw(td))) {
Puthikorn Voravootivatc4b61172014-02-05 10:17:11 -0700799
800 if (!td->o.verify_pattern_bytes) {
Jens Axboef6787012014-11-05 18:39:23 -0700801 io_u->rand_seed = __rand(&td->verify_state);
Puthikorn Voravootivatc4b61172014-02-05 10:17:11 -0700802 if (sizeof(int) != sizeof(long *))
Jens Axboef6787012014-11-05 18:39:23 -0700803 io_u->rand_seed *= __rand(&td->verify_state);
Puthikorn Voravootivatc4b61172014-02-05 10:17:11 -0700804 }
805
Jens Axboede54cfd2014-11-10 20:34:00 -0700806 if (verify_state_should_stop(td, io_u)) {
807 put_io_u(td, io_u);
808 break;
809 }
810
Jens Axboe2e1df072012-02-09 11:15:02 +0100811 if (td->o.verify_async)
812 io_u->end_io = verify_io_u_async;
813 else
814 io_u->end_io = verify_io_u;
815 td_set_runstate(td, TD_VERIFYING);
816 } else if (in_ramp_time(td))
817 td_set_runstate(td, TD_RAMP);
818 else
819 td_set_runstate(td, TD_RUNNING);
820
Puthikorn Voravootivat9a50c5c2014-02-05 10:28:15 -0800821 /*
Jens Axboef9401282014-02-06 12:17:37 -0700822 * Always log IO before it's issued, so we know the specific
823 * order of it. The logged unit will track when the IO has
824 * completed.
Puthikorn Voravootivat9a50c5c2014-02-05 10:28:15 -0800825 */
Puthikorn Voravootivatc4b61172014-02-05 10:17:11 -0700826 if (td_write(td) && io_u->ddir == DDIR_WRITE &&
827 td->o.do_verify &&
828 td->o.verify != VERIFY_NONE &&
Jens Axboef9401282014-02-06 12:17:37 -0700829 !td->o.experimental_verify)
Puthikorn Voravootivatc4b61172014-02-05 10:17:11 -0700830 log_io_piece(td, io_u);
831
Jens Axboe2e1df072012-02-09 11:15:02 +0100832 ret = td_io_queue(td, io_u);
833 switch (ret) {
834 case FIO_Q_COMPLETED:
835 if (io_u->error) {
836 ret = -io_u->error;
Jens Axboe890b6652014-05-06 19:06:51 -0600837 unlog_io_piece(td, io_u);
Jens Axboe2e1df072012-02-09 11:15:02 +0100838 clear_io_u(td, io_u);
839 } else if (io_u->resid) {
840 int bytes = io_u->xfer_buflen - io_u->resid;
841 struct fio_file *f = io_u->file;
842
Juan Casse20876c52013-09-16 09:22:10 -0700843 bytes_issued += bytes;
Jens Axboe890b6652014-05-06 19:06:51 -0600844
845 trim_io_piece(td, io_u);
846
Jens Axboe2e1df072012-02-09 11:15:02 +0100847 /*
848 * zero read, fail
849 */
850 if (!bytes) {
Jens Axboe890b6652014-05-06 19:06:51 -0600851 unlog_io_piece(td, io_u);
Jens Axboe2e1df072012-02-09 11:15:02 +0100852 td_verror(td, EIO, "full resid");
853 put_io_u(td, io_u);
854 break;
855 }
856
857 io_u->xfer_buflen = io_u->resid;
858 io_u->xfer_buf += bytes;
859 io_u->offset += bytes;
860
861 if (ddir_rw(io_u->ddir))
862 td->ts.short_io_u[io_u->ddir]++;
863
864 if (io_u->offset == f->real_file_size)
865 goto sync_done;
866
867 requeue_io_u(td, &io_u);
868 } else {
869sync_done:
Shaohua Li6eaf09d2012-09-14 08:49:43 +0200870 if (__should_check_rate(td, DDIR_READ) ||
871 __should_check_rate(td, DDIR_WRITE) ||
872 __should_check_rate(td, DDIR_TRIM))
Jens Axboe2e1df072012-02-09 11:15:02 +0100873 fio_gettime(&comp_time, NULL);
874
875 ret = io_u_sync_complete(td, io_u, bytes_done);
876 if (ret < 0)
877 break;
Juan Casse20876c52013-09-16 09:22:10 -0700878 bytes_issued += io_u->xfer_buflen;
Jens Axboe2e1df072012-02-09 11:15:02 +0100879 }
880 break;
881 case FIO_Q_QUEUED:
882 /*
883 * if the engine doesn't have a commit hook,
884 * the io_u is really queued. if it does have such
885 * a hook, it has to call io_u_queued() itself.
886 */
887 if (td->io_ops->commit == NULL)
888 io_u_queued(td, io_u);
Juan Casse20876c52013-09-16 09:22:10 -0700889 bytes_issued += io_u->xfer_buflen;
Jens Axboe2e1df072012-02-09 11:15:02 +0100890 break;
891 case FIO_Q_BUSY:
Jens Axboe890b6652014-05-06 19:06:51 -0600892 unlog_io_piece(td, io_u);
Jens Axboe2e1df072012-02-09 11:15:02 +0100893 requeue_io_u(td, &io_u);
894 ret2 = td_io_commit(td);
895 if (ret2 < 0)
896 ret = ret2;
897 break;
898 default:
899 assert(ret < 0);
900 put_io_u(td, io_u);
901 break;
902 }
903
904 if (break_on_this_error(td, ddir, &ret))
905 break;
906
907 /*
908 * See if we need to complete some commands. Note that we
909 * can get BUSY even without IO queued, if the system is
910 * resource starved.
911 */
Jens Axboe3e260a42013-12-09 12:38:53 -0700912reap:
Jens Axboe2e1df072012-02-09 11:15:02 +0100913 full = queue_full(td) || (ret == FIO_Q_BUSY && td->cur_depth);
Jens Axboe24528e72014-12-12 08:49:12 -0700914 if (full || !td->o.iodepth_batch_complete)
915 ret = wait_for_completions(td, &comp_time, bytes_done);
Jens Axboe2e1df072012-02-09 11:15:02 +0100916 if (ret < 0)
917 break;
Jens Axboed5abee02012-10-22 20:36:26 +0200918 if (!ddir_rw_sum(bytes_done) && !(td->io_ops->flags & FIO_NOIO))
Jens Axboe2e1df072012-02-09 11:15:02 +0100919 continue;
920
921 if (!in_ramp_time(td) && should_check_rate(td, bytes_done)) {
922 if (check_min_rate(td, &comp_time, bytes_done)) {
923 if (exitall_on_terminate)
924 fio_terminate_threads(td->groupid);
925 td_verror(td, EIO, "check_min_rate");
926 break;
927 }
928 }
Jens Axboe3e260a42013-12-09 12:38:53 -0700929 if (!in_ramp_time(td) && td->o.latency_target)
930 lat_target_check(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100931
932 if (td->o.thinktime) {
933 unsigned long long b;
934
Jens Axboe342f4be2012-09-14 08:59:20 +0200935 b = ddir_rw_sum(td->io_blocks);
Jens Axboe2e1df072012-02-09 11:15:02 +0100936 if (!(b % td->o.thinktime_blocks)) {
937 int left;
938
Jens Axboe002e7182013-05-17 12:39:53 +0200939 io_u_quiesce(td);
940
Jens Axboe2e1df072012-02-09 11:15:02 +0100941 if (td->o.thinktime_spin)
942 usec_spin(td->o.thinktime_spin);
943
944 left = td->o.thinktime - td->o.thinktime_spin;
945 if (left)
946 usec_sleep(td, left);
947 }
948 }
949 }
950
Jens Axboec97f1ad2013-03-28 15:08:49 -0600951 check_update_rusage(td);
952
Jens Axboe2e1df072012-02-09 11:15:02 +0100953 if (td->trim_entries)
Jens Axboe4e0a8fa2013-04-15 11:40:57 +0200954 log_err("fio: %lu trim entries leaked?\n", td->trim_entries);
Jens Axboe2e1df072012-02-09 11:15:02 +0100955
956 if (td->o.fill_device && td->error == ENOSPC) {
957 td->error = 0;
Jens Axboe3b44e8b2014-07-09 11:24:12 +0200958 fio_mark_td_terminate(td);
Jens Axboe2e1df072012-02-09 11:15:02 +0100959 }
960 if (!td->error) {
961 struct fio_file *f;
962
963 i = td->cur_depth;
964 if (i) {
Jens Axboe5bd5f712013-02-03 14:20:44 +0100965 ret = io_u_queued_complete(td, i, bytes_done);
Jens Axboe2e1df072012-02-09 11:15:02 +0100966 if (td->o.fill_device && td->error == ENOSPC)
967 td->error = 0;
968 }
969
970 if (should_fsync(td) && td->o.end_fsync) {
971 td_set_runstate(td, TD_FSYNCING);
972
973 for_each_file(td, f, i) {
Jens Axboe61ee0f82013-02-11 11:26:55 +0100974 if (!fio_file_fsync(td, f))
Jens Axboe2e1df072012-02-09 11:15:02 +0100975 continue;
Jens Axboe61ee0f82013-02-11 11:26:55 +0100976
977 log_err("fio: end_fsync failed for file %s\n",
978 f->file_name);
Jens Axboe2e1df072012-02-09 11:15:02 +0100979 }
980 }
981 } else
982 cleanup_pending_aio(td);
983
984 /*
985 * stop job if we failed doing any IO
986 */
Jens Axboe342f4be2012-09-14 08:59:20 +0200987 if (!ddir_rw_sum(td->this_io_bytes))
Jens Axboe2e1df072012-02-09 11:15:02 +0100988 td->done = 1;
Jens Axboe100f49f2013-01-23 10:15:57 -0700989
990 return bytes_done[DDIR_WRITE] + bytes_done[DDIR_TRIM];
Jens Axboe2e1df072012-02-09 11:15:02 +0100991}
992
993static void cleanup_io_u(struct thread_data *td)
994{
Jens Axboe2e1df072012-02-09 11:15:02 +0100995 struct io_u *io_u;
996
Jens Axboe2ae0b202013-05-28 14:16:55 +0200997 while ((io_u = io_u_qpop(&td->io_u_freelist)) != NULL) {
Jens Axboec73ed242012-12-06 20:30:38 +0100998
999 if (td->io_ops->io_u_free)
1000 td->io_ops->io_u_free(td, io_u);
1001
Jens Axboe2e1df072012-02-09 11:15:02 +01001002 fio_memfree(io_u, sizeof(*io_u));
1003 }
1004
1005 free_io_mem(td);
Jens Axboe2ae0b202013-05-28 14:16:55 +02001006
1007 io_u_rexit(&td->io_u_requeues);
1008 io_u_qexit(&td->io_u_freelist);
1009 io_u_qexit(&td->io_u_all);
Jens Axboede54cfd2014-11-10 20:34:00 -07001010
1011 if (td->last_write_comp)
1012 sfree(td->last_write_comp);
Jens Axboe2e1df072012-02-09 11:15:02 +01001013}
1014
1015static int init_io_u(struct thread_data *td)
1016{
1017 struct io_u *io_u;
Jens Axboe9c426842012-03-02 21:02:12 +01001018 unsigned int max_bs, min_write;
Jens Axboe2e1df072012-02-09 11:15:02 +01001019 int cl_align, i, max_units;
Jens Axboe2ae0b202013-05-28 14:16:55 +02001020 int data_xfer = 1, err;
Jens Axboe2e1df072012-02-09 11:15:02 +01001021 char *p;
1022
1023 max_units = td->o.iodepth;
Jens Axboe74f4b022013-03-22 22:56:55 -06001024 max_bs = td_max_bs(td);
Jens Axboe9c426842012-03-02 21:02:12 +01001025 min_write = td->o.min_bs[DDIR_WRITE];
Jens Axboe2e1df072012-02-09 11:15:02 +01001026 td->orig_buffer_size = (unsigned long long) max_bs
1027 * (unsigned long long) max_units;
1028
Dmitry Monakhov88045e02012-09-25 12:12:36 +02001029 if ((td->io_ops->flags & FIO_NOIO) || !(td_read(td) || td_write(td)))
Jens Axboe59d8d0f2012-09-24 14:25:33 +02001030 data_xfer = 0;
1031
Jens Axboe2ae0b202013-05-28 14:16:55 +02001032 err = 0;
1033 err += io_u_rinit(&td->io_u_requeues, td->o.iodepth);
1034 err += io_u_qinit(&td->io_u_freelist, td->o.iodepth);
1035 err += io_u_qinit(&td->io_u_all, td->o.iodepth);
1036
1037 if (err) {
1038 log_err("fio: failed setting up IO queues\n");
1039 return 1;
1040 }
1041
peter changfd8a09b2013-04-24 16:32:25 -06001042 /*
1043 * if we may later need to do address alignment, then add any
1044 * possible adjustment here so that we don't cause a buffer
1045 * overflow later. this adjustment may be too much if we get
1046 * lucky and the allocator gives us an aligned address.
1047 */
Chris Masond01612f2013-11-15 15:52:58 -07001048 if (td->o.odirect || td->o.mem_align || td->o.oatomic ||
1049 (td->io_ops->flags & FIO_RAWIO))
peter changfd8a09b2013-04-24 16:32:25 -06001050 td->orig_buffer_size += page_mask + td->o.mem_align;
1051
Jens Axboe2e1df072012-02-09 11:15:02 +01001052 if (td->o.mem_type == MEM_SHMHUGE || td->o.mem_type == MEM_MMAPHUGE) {
1053 unsigned long bs;
1054
1055 bs = td->orig_buffer_size + td->o.hugepage_size - 1;
1056 td->orig_buffer_size = bs & ~(td->o.hugepage_size - 1);
1057 }
1058
1059 if (td->orig_buffer_size != (size_t) td->orig_buffer_size) {
1060 log_err("fio: IO memory too large. Reduce max_bs or iodepth\n");
1061 return 1;
1062 }
1063
Jens Axboe59d8d0f2012-09-24 14:25:33 +02001064 if (data_xfer && allocate_io_mem(td))
Jens Axboe2e1df072012-02-09 11:15:02 +01001065 return 1;
1066
Chris Masond01612f2013-11-15 15:52:58 -07001067 if (td->o.odirect || td->o.mem_align || td->o.oatomic ||
Jens Axboe2e1df072012-02-09 11:15:02 +01001068 (td->io_ops->flags & FIO_RAWIO))
1069 p = PAGE_ALIGN(td->orig_buffer) + td->o.mem_align;
1070 else
1071 p = td->orig_buffer;
1072
1073 cl_align = os_cache_line_size();
1074
1075 for (i = 0; i < max_units; i++) {
1076 void *ptr;
1077
1078 if (td->terminate)
1079 return 1;
1080
1081 ptr = fio_memalign(cl_align, sizeof(*io_u));
1082 if (!ptr) {
1083 log_err("fio: unable to allocate aligned memory\n");
1084 break;
1085 }
1086
1087 io_u = ptr;
1088 memset(io_u, 0, sizeof(*io_u));
Jens Axboe2ae0b202013-05-28 14:16:55 +02001089 INIT_FLIST_HEAD(&io_u->verify_list);
Jens Axboe2e1df072012-02-09 11:15:02 +01001090 dprint(FD_MEM, "io_u alloc %p, index %u\n", io_u, i);
1091
Jens Axboe59d8d0f2012-09-24 14:25:33 +02001092 if (data_xfer) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001093 io_u->buf = p;
1094 dprint(FD_MEM, "io_u %p, mem %p\n", io_u, io_u->buf);
1095
1096 if (td_write(td))
Jens Axboe9c426842012-03-02 21:02:12 +01001097 io_u_fill_buffer(td, io_u, min_write, max_bs);
Jens Axboe2e1df072012-02-09 11:15:02 +01001098 if (td_write(td) && td->o.verify_pattern_bytes) {
1099 /*
1100 * Fill the buffer with the pattern if we are
1101 * going to be doing writes.
1102 */
Jens Axboece35b1e2014-01-14 15:35:58 -07001103 fill_verify_pattern(td, io_u->buf, max_bs, io_u, 0, 0);
Jens Axboe2e1df072012-02-09 11:15:02 +01001104 }
1105 }
1106
1107 io_u->index = i;
1108 io_u->flags = IO_U_F_FREE;
Jens Axboe2ae0b202013-05-28 14:16:55 +02001109 io_u_qpush(&td->io_u_freelist, io_u);
1110
1111 /*
1112 * io_u never leaves this stack, used for iteration of all
1113 * io_u buffers.
1114 */
1115 io_u_qpush(&td->io_u_all, io_u);
Jens Axboec73ed242012-12-06 20:30:38 +01001116
1117 if (td->io_ops->io_u_init) {
1118 int ret = td->io_ops->io_u_init(td, io_u);
1119
1120 if (ret) {
1121 log_err("fio: failed to init engine data: %d\n", ret);
1122 return 1;
1123 }
1124 }
1125
Jens Axboe2e1df072012-02-09 11:15:02 +01001126 p += max_bs;
1127 }
1128
Jens Axboede54cfd2014-11-10 20:34:00 -07001129 if (td->o.verify != VERIFY_NONE) {
1130 td->last_write_comp = scalloc(max_units, sizeof(uint64_t));
1131 if (!td->last_write_comp) {
1132 log_err("fio: failed to alloc write comp data\n");
1133 return 1;
1134 }
1135 }
1136
Jens Axboe2e1df072012-02-09 11:15:02 +01001137 return 0;
1138}
1139
1140static int switch_ioscheduler(struct thread_data *td)
1141{
1142 char tmp[256], tmp2[128];
1143 FILE *f;
1144 int ret;
1145
1146 if (td->io_ops->flags & FIO_DISKLESSIO)
1147 return 0;
1148
1149 sprintf(tmp, "%s/queue/scheduler", td->sysfs_root);
1150
1151 f = fopen(tmp, "r+");
1152 if (!f) {
1153 if (errno == ENOENT) {
1154 log_err("fio: os or kernel doesn't support IO scheduler"
1155 " switching\n");
1156 return 0;
1157 }
1158 td_verror(td, errno, "fopen iosched");
1159 return 1;
1160 }
1161
1162 /*
1163 * Set io scheduler.
1164 */
1165 ret = fwrite(td->o.ioscheduler, strlen(td->o.ioscheduler), 1, f);
1166 if (ferror(f) || ret != 1) {
1167 td_verror(td, errno, "fwrite");
1168 fclose(f);
1169 return 1;
1170 }
1171
1172 rewind(f);
1173
1174 /*
1175 * Read back and check that the selected scheduler is now the default.
1176 */
Jens Axboe49c6f332014-04-14 11:36:01 -06001177 ret = fread(tmp, sizeof(tmp), 1, f);
Jens Axboe2e1df072012-02-09 11:15:02 +01001178 if (ferror(f) || ret < 0) {
1179 td_verror(td, errno, "fread");
1180 fclose(f);
1181 return 1;
1182 }
Jens Axboe49c6f332014-04-14 11:36:01 -06001183 tmp[sizeof(tmp) - 1] = '\0';
1184
Jens Axboe2e1df072012-02-09 11:15:02 +01001185
1186 sprintf(tmp2, "[%s]", td->o.ioscheduler);
1187 if (!strstr(tmp, tmp2)) {
1188 log_err("fio: io scheduler %s not found\n", td->o.ioscheduler);
1189 td_verror(td, EINVAL, "iosched_switch");
1190 fclose(f);
1191 return 1;
1192 }
1193
1194 fclose(f);
1195 return 0;
1196}
1197
1198static int keep_running(struct thread_data *td)
1199{
Jens Axboe77731b22014-04-28 12:08:47 -06001200 unsigned long long limit;
1201
Jens Axboe2e1df072012-02-09 11:15:02 +01001202 if (td->done)
1203 return 0;
1204 if (td->o.time_based)
1205 return 1;
1206 if (td->o.loops) {
1207 td->o.loops--;
1208 return 1;
1209 }
Jens Axboe3939fe82014-03-11 20:16:38 -06001210 if (exceeds_number_ios(td))
1211 return 0;
Jens Axboe26251d82014-03-11 10:21:18 -06001212
Jens Axboe77731b22014-04-28 12:08:47 -06001213 if (td->o.io_limit)
1214 limit = td->o.io_limit;
1215 else
1216 limit = td->o.size;
1217
1218 if (limit != -1ULL && ddir_rw_sum(td->io_bytes) < limit) {
Jens Axboe5bd5f712013-02-03 14:20:44 +01001219 uint64_t diff;
1220
1221 /*
1222 * If the difference is less than the minimum IO size, we
1223 * are done.
1224 */
Jens Axboe77731b22014-04-28 12:08:47 -06001225 diff = limit - ddir_rw_sum(td->io_bytes);
Jens Axboe74f4b022013-03-22 22:56:55 -06001226 if (diff < td_max_bs(td))
Jens Axboe5bd5f712013-02-03 14:20:44 +01001227 return 0;
1228
Jens Axboe002fe732014-02-11 08:31:13 -07001229 if (fio_files_done(td))
1230 return 0;
1231
Jens Axboe2e1df072012-02-09 11:15:02 +01001232 return 1;
Jens Axboe5bd5f712013-02-03 14:20:44 +01001233 }
Jens Axboe2e1df072012-02-09 11:15:02 +01001234
1235 return 0;
1236}
1237
Erwan Veluce486492013-07-17 23:04:46 +02001238static int exec_string(struct thread_options *o, const char *string, const char *mode)
Jens Axboe2e1df072012-02-09 11:15:02 +01001239{
Erwan Veluce486492013-07-17 23:04:46 +02001240 int ret, newlen = strlen(string) + strlen(o->name) + strlen(mode) + 9 + 1;
Jens Axboe2e1df072012-02-09 11:15:02 +01001241 char *str;
1242
1243 str = malloc(newlen);
Erwan Veluce486492013-07-17 23:04:46 +02001244 sprintf(str, "%s &> %s.%s.txt", string, o->name, mode);
Jens Axboe2e1df072012-02-09 11:15:02 +01001245
Erwan Veluce486492013-07-17 23:04:46 +02001246 log_info("%s : Saving output of %s in %s.%s.txt\n",o->name, mode, o->name, mode);
Jens Axboe2e1df072012-02-09 11:15:02 +01001247 ret = system(str);
1248 if (ret == -1)
1249 log_err("fio: exec of cmd <%s> failed\n", str);
1250
1251 free(str);
1252 return ret;
1253}
1254
1255/*
Juan Casse62167762013-09-17 14:06:13 -07001256 * Dry run to compute correct state of numberio for verification.
1257 */
1258static uint64_t do_dry_run(struct thread_data *td)
1259{
1260 uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 };
1261
1262 td_set_runstate(td, TD_RUNNING);
1263
1264 while ((td->o.read_iolog_file && !flist_empty(&td->io_log_list)) ||
Justin Eno1b7ab802015-01-28 14:13:28 -08001265 (!flist_empty(&td->trim_list)) || !io_complete_bytes_exceeded(td)) {
Juan Casse62167762013-09-17 14:06:13 -07001266 struct io_u *io_u;
1267 int ret;
1268
1269 if (td->terminate || td->done)
1270 break;
1271
1272 io_u = get_io_u(td);
1273 if (!io_u)
1274 break;
1275
1276 io_u->flags |= IO_U_F_FLIGHT;
1277 io_u->error = 0;
1278 io_u->resid = 0;
1279 if (ddir_rw(acct_ddir(io_u)))
1280 td->io_issues[acct_ddir(io_u)]++;
1281 if (ddir_rw(io_u->ddir)) {
1282 io_u_mark_depth(td, 1);
1283 td->ts.total_io_u[io_u->ddir]++;
1284 }
1285
Puthikorn Voravootivat2e63e962014-03-03 15:33:32 -08001286 if (td_write(td) && io_u->ddir == DDIR_WRITE &&
1287 td->o.do_verify &&
1288 td->o.verify != VERIFY_NONE &&
1289 !td->o.experimental_verify)
1290 log_io_piece(td, io_u);
1291
Juan Casse62167762013-09-17 14:06:13 -07001292 ret = io_u_sync_complete(td, io_u, bytes_done);
1293 (void) ret;
1294 }
1295
1296 return bytes_done[DDIR_WRITE] + bytes_done[DDIR_TRIM];
1297}
1298
1299/*
Jens Axboe2e1df072012-02-09 11:15:02 +01001300 * Entry point for the thread based jobs. The process based jobs end up
1301 * here as well, after a little setup.
1302 */
1303static void *thread_main(void *data)
1304{
1305 unsigned long long elapsed;
1306 struct thread_data *td = data;
Jens Axboe48964732013-04-11 12:01:48 +02001307 struct thread_options *o = &td->o;
Jens Axboe2e1df072012-02-09 11:15:02 +01001308 pthread_condattr_t attr;
1309 int clear_state;
Jens Axboe28727df2012-03-29 08:33:15 +02001310 int ret;
Jens Axboe2e1df072012-02-09 11:15:02 +01001311
Jens Axboe48964732013-04-11 12:01:48 +02001312 if (!o->use_thread) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001313 setsid();
1314 td->pid = getpid();
1315 } else
1316 td->pid = gettid();
1317
Jens Axboe48964732013-04-11 12:01:48 +02001318 fio_local_clock_init(o->use_thread);
Jens Axboe5d879392012-12-18 09:12:27 +01001319
Jens Axboe2e1df072012-02-09 11:15:02 +01001320 dprint(FD_PROCESS, "jobs pid=%d started\n", (int) td->pid);
1321
Jens Axboe122c7722012-03-19 09:13:15 +01001322 if (is_backend)
1323 fio_server_send_start(td);
1324
Jens Axboe2e1df072012-02-09 11:15:02 +01001325 INIT_FLIST_HEAD(&td->io_log_list);
1326 INIT_FLIST_HEAD(&td->io_hist_list);
1327 INIT_FLIST_HEAD(&td->verify_list);
1328 INIT_FLIST_HEAD(&td->trim_list);
Jens Axboe1ae83d42013-01-12 01:44:15 -07001329 INIT_FLIST_HEAD(&td->next_rand_list);
Jens Axboe2e1df072012-02-09 11:15:02 +01001330 pthread_mutex_init(&td->io_u_lock, NULL);
1331 td->io_hist_tree = RB_ROOT;
1332
1333 pthread_condattr_init(&attr);
1334 pthread_cond_init(&td->verify_cond, &attr);
1335 pthread_cond_init(&td->free_cond, &attr);
1336
1337 td_set_runstate(td, TD_INITIALIZED);
1338 dprint(FD_MUTEX, "up startup_mutex\n");
1339 fio_mutex_up(startup_mutex);
1340 dprint(FD_MUTEX, "wait on td->mutex\n");
1341 fio_mutex_down(td->mutex);
1342 dprint(FD_MUTEX, "done waiting on td->mutex\n");
1343
1344 /*
Jens Axboe2e1df072012-02-09 11:15:02 +01001345 * A new gid requires privilege, so we need to do this before setting
1346 * the uid.
1347 */
Jens Axboe48964732013-04-11 12:01:48 +02001348 if (o->gid != -1U && setgid(o->gid)) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001349 td_verror(td, errno, "setgid");
1350 goto err;
1351 }
Jens Axboe48964732013-04-11 12:01:48 +02001352 if (o->uid != -1U && setuid(o->uid)) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001353 td_verror(td, errno, "setuid");
1354 goto err;
1355 }
1356
1357 /*
1358 * If we have a gettimeofday() thread, make sure we exclude that
1359 * thread from this job
1360 */
Jens Axboe48964732013-04-11 12:01:48 +02001361 if (o->gtod_cpu)
1362 fio_cpu_clear(&o->cpumask, o->gtod_cpu);
Jens Axboe2e1df072012-02-09 11:15:02 +01001363
1364 /*
1365 * Set affinity first, in case it has an impact on the memory
1366 * allocations.
1367 */
Jens Axboe94b360f2014-12-09 13:17:33 -07001368 if (fio_option_is_set(o, cpumask)) {
Jens Axboec2acfba2014-02-27 15:52:02 -08001369 if (o->cpus_allowed_policy == FIO_CPUS_SPLIT) {
Jens Axboe30cb4c62014-02-28 08:35:40 -08001370 ret = fio_cpus_split(&o->cpumask, td->thread_number - 1);
Jens Axboec2acfba2014-02-27 15:52:02 -08001371 if (!ret) {
1372 log_err("fio: no CPUs set\n");
1373 log_err("fio: Try increasing number of available CPUs\n");
1374 td_verror(td, EINVAL, "cpus_split");
1375 goto err;
1376 }
1377 }
Jens Axboe28727df2012-03-29 08:33:15 +02001378 ret = fio_setaffinity(td->pid, o->cpumask);
1379 if (ret == -1) {
Jens Axboe48964732013-04-11 12:01:48 +02001380 td_verror(td, errno, "cpu_set_affinity");
1381 goto err;
1382 }
Jens Axboe2e1df072012-02-09 11:15:02 +01001383 }
1384
Jens Axboe67bf9822013-01-10 11:23:19 +01001385#ifdef CONFIG_LIBNUMA
Yufei Rend0b937e2012-10-19 23:11:52 -04001386 /* numa node setup */
Jens Axboe94b360f2014-12-09 13:17:33 -07001387 if (fio_option_is_set(o, numa_cpunodes) ||
1388 fio_option_is_set(o, numa_memnodes)) {
Daniel Gollub43522842014-05-01 17:07:49 +02001389 struct bitmask *mask;
Yufei Rend0b937e2012-10-19 23:11:52 -04001390
1391 if (numa_available() < 0) {
1392 td_verror(td, errno, "Does not support NUMA API\n");
1393 goto err;
1394 }
1395
Jens Axboe94b360f2014-12-09 13:17:33 -07001396 if (fio_option_is_set(o, numa_cpunodes)) {
Daniel Gollub43522842014-05-01 17:07:49 +02001397 mask = numa_parse_nodestring(o->numa_cpunodes);
1398 ret = numa_run_on_node_mask(mask);
1399 numa_free_nodemask(mask);
Yufei Rend0b937e2012-10-19 23:11:52 -04001400 if (ret == -1) {
1401 td_verror(td, errno, \
1402 "numa_run_on_node_mask failed\n");
1403 goto err;
1404 }
1405 }
1406
Jens Axboe94b360f2014-12-09 13:17:33 -07001407 if (fio_option_is_set(o, numa_memnodes)) {
Daniel Gollub43522842014-05-01 17:07:49 +02001408 mask = NULL;
1409 if (o->numa_memnodes)
1410 mask = numa_parse_nodestring(o->numa_memnodes);
1411
Jens Axboe48964732013-04-11 12:01:48 +02001412 switch (o->numa_mem_mode) {
Yufei Rend0b937e2012-10-19 23:11:52 -04001413 case MPOL_INTERLEAVE:
Daniel Gollub43522842014-05-01 17:07:49 +02001414 numa_set_interleave_mask(mask);
Yufei Rend0b937e2012-10-19 23:11:52 -04001415 break;
1416 case MPOL_BIND:
Daniel Gollub43522842014-05-01 17:07:49 +02001417 numa_set_membind(mask);
Yufei Rend0b937e2012-10-19 23:11:52 -04001418 break;
1419 case MPOL_LOCAL:
1420 numa_set_localalloc();
1421 break;
1422 case MPOL_PREFERRED:
Jens Axboe48964732013-04-11 12:01:48 +02001423 numa_set_preferred(o->numa_mem_prefer_node);
Yufei Rend0b937e2012-10-19 23:11:52 -04001424 break;
1425 case MPOL_DEFAULT:
1426 default:
1427 break;
1428 }
1429
Daniel Gollub43522842014-05-01 17:07:49 +02001430 if (mask)
1431 numa_free_nodemask(mask);
1432
Yufei Rend0b937e2012-10-19 23:11:52 -04001433 }
1434 }
1435#endif
1436
Jens Axboe9a3f1102012-03-28 20:50:15 +02001437 if (fio_pin_memory(td))
1438 goto err;
1439
Jens Axboe2e1df072012-02-09 11:15:02 +01001440 /*
1441 * May alter parameters that init_io_u() will use, so we need to
1442 * do this first.
1443 */
1444 if (init_iolog(td))
1445 goto err;
1446
1447 if (init_io_u(td))
1448 goto err;
1449
Jens Axboe48964732013-04-11 12:01:48 +02001450 if (o->verify_async && verify_async_init(td))
Jens Axboe2e1df072012-02-09 11:15:02 +01001451 goto err;
1452
Jens Axboe1006b2d2014-12-09 14:10:05 -07001453 if (fio_option_is_set(o, ioprio) ||
1454 fio_option_is_set(o, ioprio_class)) {
Jens Axboe28727df2012-03-29 08:33:15 +02001455 ret = ioprio_set(IOPRIO_WHO_PROCESS, 0, o->ioprio_class, o->ioprio);
1456 if (ret == -1) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001457 td_verror(td, errno, "ioprio_set");
1458 goto err;
1459 }
1460 }
1461
Jens Axboe48964732013-04-11 12:01:48 +02001462 if (o->cgroup && cgroup_setup(td, cgroup_list, &cgroup_mnt))
Jens Axboe2e1df072012-02-09 11:15:02 +01001463 goto err;
1464
Bruce Cran649c10c2012-02-20 07:59:06 +01001465 errno = 0;
Jens Axboe48964732013-04-11 12:01:48 +02001466 if (nice(o->nice) == -1 && errno != 0) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001467 td_verror(td, errno, "nice");
1468 goto err;
1469 }
1470
Jens Axboe48964732013-04-11 12:01:48 +02001471 if (o->ioscheduler && switch_ioscheduler(td))
Jens Axboe2e1df072012-02-09 11:15:02 +01001472 goto err;
1473
Jens Axboe48964732013-04-11 12:01:48 +02001474 if (!o->create_serialize && setup_files(td))
Jens Axboe2e1df072012-02-09 11:15:02 +01001475 goto err;
1476
1477 if (td_io_init(td))
1478 goto err;
1479
1480 if (init_random_map(td))
1481 goto err;
1482
Erwan Veluce486492013-07-17 23:04:46 +02001483 if (o->exec_prerun && exec_string(o, o->exec_prerun, (const char *)"prerun"))
Jens Axboe48964732013-04-11 12:01:48 +02001484 goto err;
Jens Axboe2e1df072012-02-09 11:15:02 +01001485
Jens Axboe48964732013-04-11 12:01:48 +02001486 if (o->pre_read) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001487 if (pre_read_files(td) < 0)
1488 goto err;
1489 }
1490
Jens Axboe38a812d2014-07-03 09:10:39 -06001491 if (td->flags & TD_F_COMPRESS_LOG)
1492 tp_init(&td->tp_data);
1493
Jens Axboedc5bfbb2013-04-10 22:23:40 +02001494 fio_verify_init(td);
1495
Jens Axboe2e1df072012-02-09 11:15:02 +01001496 fio_gettime(&td->epoch, NULL);
Jens Axboe44404c52013-01-24 14:20:09 -07001497 fio_getrusage(&td->ru_start);
Jens Axboe2e1df072012-02-09 11:15:02 +01001498 clear_state = 0;
1499 while (keep_running(td)) {
Jens Axboe100f49f2013-01-23 10:15:57 -07001500 uint64_t verify_bytes;
1501
Jens Axboe2e1df072012-02-09 11:15:02 +01001502 fio_gettime(&td->start, NULL);
1503 memcpy(&td->bw_sample_time, &td->start, sizeof(td->start));
1504 memcpy(&td->iops_sample_time, &td->start, sizeof(td->start));
1505 memcpy(&td->tv_cache, &td->start, sizeof(td->start));
1506
Jens Axboe48964732013-04-11 12:01:48 +02001507 if (o->ratemin[DDIR_READ] || o->ratemin[DDIR_WRITE] ||
1508 o->ratemin[DDIR_TRIM]) {
Shaohua Li6eaf09d2012-09-14 08:49:43 +02001509 memcpy(&td->lastrate[DDIR_READ], &td->bw_sample_time,
Jens Axboe2e1df072012-02-09 11:15:02 +01001510 sizeof(td->bw_sample_time));
Shaohua Li6eaf09d2012-09-14 08:49:43 +02001511 memcpy(&td->lastrate[DDIR_WRITE], &td->bw_sample_time,
1512 sizeof(td->bw_sample_time));
1513 memcpy(&td->lastrate[DDIR_TRIM], &td->bw_sample_time,
Jens Axboe2e1df072012-02-09 11:15:02 +01001514 sizeof(td->bw_sample_time));
1515 }
1516
1517 if (clear_state)
1518 clear_io_state(td);
1519
1520 prune_io_piece_log(td);
1521
Juan Casse62167762013-09-17 14:06:13 -07001522 if (td->o.verify_only && (td_write(td) || td_rw(td)))
1523 verify_bytes = do_dry_run(td);
1524 else
1525 verify_bytes = do_io(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001526
1527 clear_state = 1;
1528
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001529 fio_mutex_down(stat_mutex);
Jens Axboe2e1df072012-02-09 11:15:02 +01001530 if (td_read(td) && td->io_bytes[DDIR_READ]) {
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001531 elapsed = mtime_since_now(&td->start);
Jens Axboe2e1df072012-02-09 11:15:02 +01001532 td->ts.runtime[DDIR_READ] += elapsed;
1533 }
1534 if (td_write(td) && td->io_bytes[DDIR_WRITE]) {
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001535 elapsed = mtime_since_now(&td->start);
Jens Axboe2e1df072012-02-09 11:15:02 +01001536 td->ts.runtime[DDIR_WRITE] += elapsed;
1537 }
Shaohua Li6eaf09d2012-09-14 08:49:43 +02001538 if (td_trim(td) && td->io_bytes[DDIR_TRIM]) {
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001539 elapsed = mtime_since_now(&td->start);
Shaohua Li6eaf09d2012-09-14 08:49:43 +02001540 td->ts.runtime[DDIR_TRIM] += elapsed;
1541 }
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001542 fio_gettime(&td->start, NULL);
1543 fio_mutex_up(stat_mutex);
Jens Axboe2e1df072012-02-09 11:15:02 +01001544
1545 if (td->error || td->terminate)
1546 break;
1547
Jens Axboe48964732013-04-11 12:01:48 +02001548 if (!o->do_verify ||
1549 o->verify == VERIFY_NONE ||
Jens Axboe2e1df072012-02-09 11:15:02 +01001550 (td->io_ops->flags & FIO_UNIDIR))
1551 continue;
1552
1553 clear_io_state(td);
1554
1555 fio_gettime(&td->start, NULL);
1556
Jens Axboe100f49f2013-01-23 10:15:57 -07001557 do_verify(td, verify_bytes);
Jens Axboe2e1df072012-02-09 11:15:02 +01001558
Vasily Tarasovc38e38f2014-11-09 20:22:24 -07001559 fio_mutex_down(stat_mutex);
1560 td->ts.runtime[DDIR_READ] += mtime_since_now(&td->start);
1561 fio_gettime(&td->start, NULL);
1562 fio_mutex_up(stat_mutex);
Jens Axboe2e1df072012-02-09 11:15:02 +01001563
1564 if (td->error || td->terminate)
1565 break;
1566 }
1567
1568 update_rusage_stat(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001569 td->ts.total_run_time = mtime_since_now(&td->epoch);
Shaohua Li6eaf09d2012-09-14 08:49:43 +02001570 td->ts.io_bytes[DDIR_READ] = td->io_bytes[DDIR_READ];
1571 td->ts.io_bytes[DDIR_WRITE] = td->io_bytes[DDIR_WRITE];
1572 td->ts.io_bytes[DDIR_TRIM] = td->io_bytes[DDIR_TRIM];
Jens Axboe2e1df072012-02-09 11:15:02 +01001573
Jens Axboede54cfd2014-11-10 20:34:00 -07001574 if (td->o.verify_state_save && !(td->flags & TD_F_VSTATE_SAVED) &&
1575 (td->o.verify != VERIFY_NONE && td_write(td))) {
1576 struct all_io_list *state;
1577 size_t sz;
1578
1579 state = get_all_io_list(td->thread_number, &sz);
1580 if (state) {
1581 __verify_save_state(state, "local");
1582 free(state);
1583 }
1584 }
1585
Jens Axboe9a3f1102012-03-28 20:50:15 +02001586 fio_unpin_memory(td);
1587
Jens Axboe905e3d42014-04-02 20:01:27 -06001588 fio_writeout_logs(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001589
Jens Axboe38a812d2014-07-03 09:10:39 -06001590 if (td->flags & TD_F_COMPRESS_LOG)
1591 tp_exit(&td->tp_data);
1592
Jens Axboe48964732013-04-11 12:01:48 +02001593 if (o->exec_postrun)
Erwan Veluce486492013-07-17 23:04:46 +02001594 exec_string(o, o->exec_postrun, (const char *)"postrun");
Jens Axboe2e1df072012-02-09 11:15:02 +01001595
1596 if (exitall_on_terminate)
1597 fio_terminate_threads(td->groupid);
1598
1599err:
1600 if (td->error)
1601 log_info("fio: pid=%d, err=%d/%s\n", (int) td->pid, td->error,
1602 td->verror);
1603
Jens Axboe48964732013-04-11 12:01:48 +02001604 if (o->verify_async)
Jens Axboe2e1df072012-02-09 11:15:02 +01001605 verify_async_exit(td);
1606
1607 close_and_free_files(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001608 cleanup_io_u(td);
Bruce Cran32dbca22012-12-06 20:32:40 +01001609 close_ioengine(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001610 cgroup_shutdown(td, &cgroup_mnt);
Jens Axboede54cfd2014-11-10 20:34:00 -07001611 verify_free_state(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001612
Jens Axboe94b360f2014-12-09 13:17:33 -07001613 if (fio_option_is_set(o, cpumask)) {
Jens Axboe0f7f9a92014-11-06 09:21:10 -07001614 ret = fio_cpuset_exit(&o->cpumask);
1615 if (ret)
1616 td_verror(td, ret, "fio_cpuset_exit");
Jens Axboe2e1df072012-02-09 11:15:02 +01001617 }
1618
1619 /*
1620 * do this very late, it will log file closing as well
1621 */
Jens Axboe48964732013-04-11 12:01:48 +02001622 if (o->write_iolog_file)
Jens Axboe2e1df072012-02-09 11:15:02 +01001623 write_iolog_close(td);
1624
Jens Axboeea66e042014-02-10 13:57:09 -07001625 fio_mutex_remove(td->mutex);
1626 td->mutex = NULL;
1627
Jens Axboe2e1df072012-02-09 11:15:02 +01001628 td_set_runstate(td, TD_EXITED);
Jens Axboe04247862014-10-23 23:04:37 -06001629
1630 /*
1631 * Do this last after setting our runstate to exited, so we
1632 * know that the stat thread is signaled.
1633 */
1634 check_update_rusage(td);
1635
Bruce Crane43606c2012-02-20 09:34:24 +01001636 return (void *) (uintptr_t) td->error;
Jens Axboe2e1df072012-02-09 11:15:02 +01001637}
1638
1639
1640/*
1641 * We cannot pass the td data into a forked process, so attach the td and
1642 * pass it to the thread worker.
1643 */
1644static int fork_main(int shmid, int offset)
1645{
1646 struct thread_data *td;
1647 void *data, *ret;
1648
Jens Axboeef7035a2014-06-18 15:30:09 -07001649#if !defined(__hpux) && !defined(CONFIG_NO_SHM)
Jens Axboe2e1df072012-02-09 11:15:02 +01001650 data = shmat(shmid, NULL, 0);
1651 if (data == (void *) -1) {
1652 int __err = errno;
1653
1654 perror("shmat");
1655 return __err;
1656 }
1657#else
1658 /*
1659 * HP-UX inherits shm mappings?
1660 */
1661 data = threads;
1662#endif
1663
1664 td = data + offset * sizeof(struct thread_data);
1665 ret = thread_main(td);
1666 shmdt(data);
Bruce Crane43606c2012-02-20 09:34:24 +01001667 return (int) (uintptr_t) ret;
Jens Axboe2e1df072012-02-09 11:15:02 +01001668}
1669
Jens Axboe958b8912014-07-09 08:46:12 +02001670static void dump_td_info(struct thread_data *td)
1671{
1672 log_err("fio: job '%s' hasn't exited in %lu seconds, it appears to "
1673 "be stuck. Doing forceful exit of this job.\n", td->o.name,
1674 (unsigned long) time_since_now(&td->terminate_time));
1675}
1676
Jens Axboe2e1df072012-02-09 11:15:02 +01001677/*
1678 * Run over the job map and reap the threads that have exited, if any.
1679 */
1680static void reap_threads(unsigned int *nr_running, unsigned int *t_rate,
1681 unsigned int *m_rate)
1682{
1683 struct thread_data *td;
1684 unsigned int cputhreads, realthreads, pending;
1685 int i, status, ret;
1686
1687 /*
1688 * reap exited threads (TD_EXITED -> TD_REAPED)
1689 */
1690 realthreads = pending = cputhreads = 0;
1691 for_each_td(td, i) {
1692 int flags = 0;
1693
1694 /*
1695 * ->io_ops is NULL for a thread that has closed its
1696 * io engine
1697 */
1698 if (td->io_ops && !strcmp(td->io_ops->name, "cpuio"))
1699 cputhreads++;
1700 else
1701 realthreads++;
1702
1703 if (!td->pid) {
1704 pending++;
1705 continue;
1706 }
1707 if (td->runstate == TD_REAPED)
1708 continue;
1709 if (td->o.use_thread) {
1710 if (td->runstate == TD_EXITED) {
1711 td_set_runstate(td, TD_REAPED);
1712 goto reaped;
1713 }
1714 continue;
1715 }
1716
1717 flags = WNOHANG;
1718 if (td->runstate == TD_EXITED)
1719 flags = 0;
1720
1721 /*
1722 * check if someone quit or got killed in an unusual way
1723 */
1724 ret = waitpid(td->pid, &status, flags);
1725 if (ret < 0) {
1726 if (errno == ECHILD) {
1727 log_err("fio: pid=%d disappeared %d\n",
1728 (int) td->pid, td->runstate);
Jens Axboea5e371a2012-04-02 09:47:09 -07001729 td->sig = ECHILD;
Jens Axboe2e1df072012-02-09 11:15:02 +01001730 td_set_runstate(td, TD_REAPED);
1731 goto reaped;
1732 }
1733 perror("waitpid");
1734 } else if (ret == td->pid) {
1735 if (WIFSIGNALED(status)) {
1736 int sig = WTERMSIG(status);
1737
Jens Axboe36d80bc2012-11-30 21:46:06 +01001738 if (sig != SIGTERM && sig != SIGUSR2)
Jens Axboe2e1df072012-02-09 11:15:02 +01001739 log_err("fio: pid=%d, got signal=%d\n",
1740 (int) td->pid, sig);
Jens Axboea5e371a2012-04-02 09:47:09 -07001741 td->sig = sig;
Jens Axboe2e1df072012-02-09 11:15:02 +01001742 td_set_runstate(td, TD_REAPED);
1743 goto reaped;
1744 }
1745 if (WIFEXITED(status)) {
1746 if (WEXITSTATUS(status) && !td->error)
1747 td->error = WEXITSTATUS(status);
1748
1749 td_set_runstate(td, TD_REAPED);
1750 goto reaped;
1751 }
1752 }
1753
1754 /*
Jens Axboe958b8912014-07-09 08:46:12 +02001755 * If the job is stuck, do a forceful timeout of it and
1756 * move on.
1757 */
1758 if (td->terminate &&
1759 time_since_now(&td->terminate_time) >= FIO_REAP_TIMEOUT) {
1760 dump_td_info(td);
1761 td_set_runstate(td, TD_REAPED);
1762 goto reaped;
1763 }
1764
1765 /*
Jens Axboe2e1df072012-02-09 11:15:02 +01001766 * thread is not dead, continue
1767 */
1768 pending++;
1769 continue;
1770reaped:
1771 (*nr_running)--;
Jens Axboe342f4be2012-09-14 08:59:20 +02001772 (*m_rate) -= ddir_rw_sum(td->o.ratemin);
1773 (*t_rate) -= ddir_rw_sum(td->o.rate);
Jens Axboe2e1df072012-02-09 11:15:02 +01001774 if (!td->pid)
1775 pending--;
1776
1777 if (td->error)
1778 exit_value++;
1779
1780 done_secs += mtime_since_now(&td->epoch) / 1000;
Jens Axboe4a887522013-05-23 12:33:08 +02001781 profile_td_exit(td);
Jens Axboe2e1df072012-02-09 11:15:02 +01001782 }
1783
1784 if (*nr_running == cputhreads && !pending && realthreads)
1785 fio_terminate_threads(TERMINATE_ALL);
1786}
1787
Jens Axboede54cfd2014-11-10 20:34:00 -07001788static int __check_trigger_file(void)
1789{
1790 struct stat sb;
1791
1792 if (!trigger_file)
1793 return 0;
1794
1795 if (stat(trigger_file, &sb))
1796 return 0;
1797
1798 if (unlink(trigger_file) < 0)
1799 log_err("fio: failed to unlink %s: %s\n", trigger_file,
1800 strerror(errno));
1801
1802 return 1;
1803}
1804
1805static int trigger_timedout(void)
1806{
1807 if (trigger_timeout)
1808 return time_since_genesis() >= trigger_timeout;
1809
1810 return 0;
1811}
1812
1813void exec_trigger(const char *cmd)
1814{
1815 int ret;
1816
1817 if (!cmd)
1818 return;
1819
1820 ret = system(cmd);
1821 if (ret == -1)
1822 log_err("fio: failed executing %s trigger\n", cmd);
1823}
1824
1825void check_trigger_file(void)
1826{
1827 if (__check_trigger_file() || trigger_timedout()) {
Jens Axboe07f512c2014-11-19 09:21:16 -07001828 if (nr_clients)
1829 fio_clients_send_trigger(trigger_remote_cmd);
1830 else {
Jens Axboede54cfd2014-11-10 20:34:00 -07001831 verify_save_state();
1832 fio_terminate_threads(TERMINATE_ALL);
1833 exec_trigger(trigger_cmd);
1834 }
1835 }
1836}
1837
1838static int fio_verify_load_state(struct thread_data *td)
1839{
1840 int ret;
1841
1842 if (!td->o.verify_state)
1843 return 0;
1844
1845 if (is_backend) {
1846 void *data;
1847
1848 ret = fio_server_get_verify_state(td->o.name,
1849 td->thread_number - 1, &data);
1850 if (!ret)
1851 verify_convert_assign_state(td, data);
1852 } else
1853 ret = verify_load_state(td, "local");
1854
1855 return ret;
1856}
1857
Jens Axboe06464902013-04-24 20:38:54 -06001858static void do_usleep(unsigned int usecs)
1859{
1860 check_for_running_stats();
Jens Axboede54cfd2014-11-10 20:34:00 -07001861 check_trigger_file();
Jens Axboe06464902013-04-24 20:38:54 -06001862 usleep(usecs);
1863}
1864
Jens Axboe2e1df072012-02-09 11:15:02 +01001865/*
1866 * Main function for kicking off and reaping jobs, as needed.
1867 */
1868static void run_threads(void)
1869{
1870 struct thread_data *td;
Jens Axboe2e1df072012-02-09 11:15:02 +01001871 unsigned int i, todo, nr_running, m_rate, t_rate, nr_started;
Jens Axboe0de5b262014-02-21 15:26:01 -08001872 uint64_t spent;
Jens Axboe2e1df072012-02-09 11:15:02 +01001873
Jens Axboe2e1df072012-02-09 11:15:02 +01001874 if (fio_gtod_offload && fio_start_gtod_thread())
1875 return;
Bruce Cran334185e2013-11-01 12:32:40 -06001876
Huadong Liuf2a2ce02013-01-30 13:22:24 +01001877 fio_idle_prof_init();
Jens Axboe2e1df072012-02-09 11:15:02 +01001878
1879 set_sig_handlers();
1880
Jens Axboe3a5f6bd2013-04-11 12:04:38 +02001881 nr_thread = nr_process = 0;
1882 for_each_td(td, i) {
1883 if (td->o.use_thread)
1884 nr_thread++;
1885 else
1886 nr_process++;
1887 }
1888
Jens Axboef3afa572012-09-17 13:34:16 +02001889 if (output_format == FIO_OUTPUT_NORMAL) {
Jens Axboe2e1df072012-02-09 11:15:02 +01001890 log_info("Starting ");
1891 if (nr_thread)
1892 log_info("%d thread%s", nr_thread,
1893 nr_thread > 1 ? "s" : "");
1894 if (nr_process) {
1895 if (nr_thread)
1896 log_info(" and ");
1897 log_info("%d process%s", nr_process,
1898 nr_process > 1 ? "es" : "");
1899 }
1900 log_info("\n");
Jens Axboebc2e69c2014-10-13 11:57:15 -06001901 log_info_flush();
Jens Axboe2e1df072012-02-09 11:15:02 +01001902 }
1903
1904 todo = thread_number;
1905 nr_running = 0;
1906 nr_started = 0;
1907 m_rate = t_rate = 0;
1908
1909 for_each_td(td, i) {
1910 print_status_init(td->thread_number - 1);
1911
1912 if (!td->o.create_serialize)
1913 continue;
1914
Jens Axboede54cfd2014-11-10 20:34:00 -07001915 if (fio_verify_load_state(td))
1916 goto reap;
1917
Jens Axboe2e1df072012-02-09 11:15:02 +01001918 /*
1919 * do file setup here so it happens sequentially,
1920 * we don't want X number of threads getting their
1921 * client data interspersed on disk
1922 */
1923 if (setup_files(td)) {
Jens Axboede54cfd2014-11-10 20:34:00 -07001924reap:
Jens Axboe2e1df072012-02-09 11:15:02 +01001925 exit_value++;
1926 if (td->error)
1927 log_err("fio: pid=%d, err=%d/%s\n",
1928 (int) td->pid, td->error, td->verror);
1929 td_set_runstate(td, TD_REAPED);
1930 todo--;
1931 } else {
1932 struct fio_file *f;
1933 unsigned int j;
1934
1935 /*
1936 * for sharing to work, each job must always open
1937 * its own files. so close them, if we opened them
1938 * for creation
1939 */
1940 for_each_file(td, f, j) {
1941 if (fio_file_open(f))
1942 td_io_close_file(td, f);
1943 }
1944 }
1945 }
1946
Huadong Liuf2a2ce02013-01-30 13:22:24 +01001947 /* start idle threads before io threads start to run */
1948 fio_idle_prof_start();
1949
Jens Axboe2e1df072012-02-09 11:15:02 +01001950 set_genesis_time();
1951
1952 while (todo) {
1953 struct thread_data *map[REAL_MAX_JOBS];
1954 struct timeval this_start;
1955 int this_jobs = 0, left;
1956
1957 /*
1958 * create threads (TD_NOT_CREATED -> TD_CREATED)
1959 */
1960 for_each_td(td, i) {
1961 if (td->runstate != TD_NOT_CREATED)
1962 continue;
1963
1964 /*
1965 * never got a chance to start, killed by other
1966 * thread for some reason
1967 */
1968 if (td->terminate) {
1969 todo--;
1970 continue;
1971 }
1972
1973 if (td->o.start_delay) {
Jens Axboe0de5b262014-02-21 15:26:01 -08001974 spent = utime_since_genesis();
Jens Axboe2e1df072012-02-09 11:15:02 +01001975
Christian Ehrhardt74454ce2014-02-20 14:20:01 +01001976 if (td->o.start_delay > spent)
Jens Axboe2e1df072012-02-09 11:15:02 +01001977 continue;
1978 }
1979
1980 if (td->o.stonewall && (nr_started || nr_running)) {
1981 dprint(FD_PROCESS, "%s: stonewall wait\n",
1982 td->o.name);
1983 break;
1984 }
1985
1986 init_disk_util(td);
1987
Jens Axboec97f1ad2013-03-28 15:08:49 -06001988 td->rusage_sem = fio_mutex_init(FIO_MUTEX_LOCKED);
1989 td->update_rusage = 0;
1990
Jens Axboe2e1df072012-02-09 11:15:02 +01001991 /*
1992 * Set state to created. Thread will transition
1993 * to TD_INITIALIZED when it's done setting up.
1994 */
1995 td_set_runstate(td, TD_CREATED);
1996 map[this_jobs++] = td;
1997 nr_started++;
1998
1999 if (td->o.use_thread) {
2000 int ret;
2001
2002 dprint(FD_PROCESS, "will pthread_create\n");
2003 ret = pthread_create(&td->thread, NULL,
2004 thread_main, td);
2005 if (ret) {
2006 log_err("pthread_create: %s\n",
2007 strerror(ret));
2008 nr_started--;
2009 break;
2010 }
2011 ret = pthread_detach(td->thread);
2012 if (ret)
2013 log_err("pthread_detach: %s",
2014 strerror(ret));
2015 } else {
2016 pid_t pid;
2017 dprint(FD_PROCESS, "will fork\n");
2018 pid = fork();
2019 if (!pid) {
2020 int ret = fork_main(shm_id, i);
2021
2022 _exit(ret);
2023 } else if (i == fio_debug_jobno)
2024 *fio_debug_jobp = pid;
2025 }
2026 dprint(FD_MUTEX, "wait on startup_mutex\n");
2027 if (fio_mutex_down_timeout(startup_mutex, 10)) {
2028 log_err("fio: job startup hung? exiting.\n");
2029 fio_terminate_threads(TERMINATE_ALL);
2030 fio_abort = 1;
2031 nr_started--;
2032 break;
2033 }
2034 dprint(FD_MUTEX, "done waiting on startup_mutex\n");
2035 }
2036
2037 /*
2038 * Wait for the started threads to transition to
2039 * TD_INITIALIZED.
2040 */
2041 fio_gettime(&this_start, NULL);
2042 left = this_jobs;
2043 while (left && !fio_abort) {
2044 if (mtime_since_now(&this_start) > JOB_START_TIMEOUT)
2045 break;
2046
Jens Axboe06464902013-04-24 20:38:54 -06002047 do_usleep(100000);
Jens Axboe2e1df072012-02-09 11:15:02 +01002048
2049 for (i = 0; i < this_jobs; i++) {
2050 td = map[i];
2051 if (!td)
2052 continue;
2053 if (td->runstate == TD_INITIALIZED) {
2054 map[i] = NULL;
2055 left--;
2056 } else if (td->runstate >= TD_EXITED) {
2057 map[i] = NULL;
2058 left--;
2059 todo--;
2060 nr_running++; /* work-around... */
2061 }
2062 }
2063 }
2064
2065 if (left) {
Jens Axboe4e87c372012-02-15 14:27:08 +01002066 log_err("fio: %d job%s failed to start\n", left,
2067 left > 1 ? "s" : "");
Jens Axboe2e1df072012-02-09 11:15:02 +01002068 for (i = 0; i < this_jobs; i++) {
2069 td = map[i];
2070 if (!td)
2071 continue;
2072 kill(td->pid, SIGTERM);
2073 }
2074 break;
2075 }
2076
2077 /*
2078 * start created threads (TD_INITIALIZED -> TD_RUNNING).
2079 */
2080 for_each_td(td, i) {
2081 if (td->runstate != TD_INITIALIZED)
2082 continue;
2083
2084 if (in_ramp_time(td))
2085 td_set_runstate(td, TD_RAMP);
2086 else
2087 td_set_runstate(td, TD_RUNNING);
2088 nr_running++;
2089 nr_started--;
Jens Axboe342f4be2012-09-14 08:59:20 +02002090 m_rate += ddir_rw_sum(td->o.ratemin);
2091 t_rate += ddir_rw_sum(td->o.rate);
Jens Axboe2e1df072012-02-09 11:15:02 +01002092 todo--;
2093 fio_mutex_up(td->mutex);
2094 }
2095
2096 reap_threads(&nr_running, &t_rate, &m_rate);
2097
Jens Axboe122c7722012-03-19 09:13:15 +01002098 if (todo)
Jens Axboe06464902013-04-24 20:38:54 -06002099 do_usleep(100000);
Jens Axboe2e1df072012-02-09 11:15:02 +01002100 }
2101
2102 while (nr_running) {
2103 reap_threads(&nr_running, &t_rate, &m_rate);
Jens Axboe06464902013-04-24 20:38:54 -06002104 do_usleep(10000);
Jens Axboe2e1df072012-02-09 11:15:02 +01002105 }
2106
Huadong Liuf2a2ce02013-01-30 13:22:24 +01002107 fio_idle_prof_stop();
2108
Jens Axboe2e1df072012-02-09 11:15:02 +01002109 update_io_ticks();
Jens Axboe2e1df072012-02-09 11:15:02 +01002110}
2111
Jens Axboe7fe36312014-10-23 23:16:50 -06002112static void wait_for_helper_thread_exit(void)
Jens Axboe9ec77792012-08-02 08:27:41 +02002113{
Jens Axboe62244c92014-10-23 08:53:33 -06002114 void *ret;
2115
Jens Axboe1266d3f2014-10-23 23:47:01 -06002116 helper_exit = 1;
Jens Axboe7fe36312014-10-23 23:16:50 -06002117 pthread_cond_signal(&helper_cond);
2118 pthread_join(helper_thread, &ret);
Jens Axboe9ec77792012-08-02 08:27:41 +02002119}
2120
Jens Axboe27357182012-09-27 11:31:50 +02002121static void free_disk_util(void)
2122{
Jens Axboe27357182012-09-27 11:31:50 +02002123 disk_util_prune_entries();
Jens Axboe62244c92014-10-23 08:53:33 -06002124
Jens Axboe7fe36312014-10-23 23:16:50 -06002125 pthread_cond_destroy(&helper_cond);
Jens Axboe27357182012-09-27 11:31:50 +02002126}
2127
Jens Axboe7fe36312014-10-23 23:16:50 -06002128static void *helper_thread_main(void *data)
Jens Axboe2e1df072012-02-09 11:15:02 +01002129{
Jens Axboe9ec77792012-08-02 08:27:41 +02002130 int ret = 0;
2131
Jens Axboe2e1df072012-02-09 11:15:02 +01002132 fio_mutex_up(startup_mutex);
2133
Jens Axboe62244c92014-10-23 08:53:33 -06002134 while (!ret) {
2135 uint64_t sec = DISK_UTIL_MSEC / 1000;
2136 uint64_t nsec = (DISK_UTIL_MSEC % 1000) * 1000000;
2137 struct timespec ts;
2138 struct timeval tv;
2139
2140 gettimeofday(&tv, NULL);
2141 ts.tv_sec = tv.tv_sec + sec;
2142 ts.tv_nsec = (tv.tv_usec * 1000) + nsec;
Jens Axboe42ec5752014-12-15 19:54:57 -07002143
Jens Axboe581b5df2014-12-15 13:45:53 -07002144 if (ts.tv_nsec >= 1000000000ULL) {
Jens Axboe62244c92014-10-23 08:53:33 -06002145 ts.tv_nsec -= 1000000000ULL;
2146 ts.tv_sec++;
2147 }
2148
Jens Axboe7fe36312014-10-23 23:16:50 -06002149 pthread_cond_timedwait(&helper_cond, &helper_lock, &ts);
Jens Axboe62244c92014-10-23 08:53:33 -06002150
Jens Axboe9ec77792012-08-02 08:27:41 +02002151 ret = update_io_ticks();
Jens Axboe2e1df072012-02-09 11:15:02 +01002152
Jens Axboe7fe36312014-10-23 23:16:50 -06002153 if (helper_do_stat) {
2154 helper_do_stat = 0;
2155 __show_running_run_stats();
2156 }
2157
Jens Axboe2e1df072012-02-09 11:15:02 +01002158 if (!is_backend)
2159 print_thread_status();
2160 }
2161
2162 return NULL;
2163}
2164
Jens Axboe7fe36312014-10-23 23:16:50 -06002165static int create_helper_thread(void)
Jens Axboe2e1df072012-02-09 11:15:02 +01002166{
2167 int ret;
2168
Jens Axboe9ec77792012-08-02 08:27:41 +02002169 setup_disk_util();
2170
Jens Axboe7fe36312014-10-23 23:16:50 -06002171 pthread_cond_init(&helper_cond, NULL);
2172 pthread_mutex_init(&helper_lock, NULL);
Jens Axboe62244c92014-10-23 08:53:33 -06002173
Jens Axboe7fe36312014-10-23 23:16:50 -06002174 ret = pthread_create(&helper_thread, NULL, helper_thread_main, NULL);
Jens Axboe2e1df072012-02-09 11:15:02 +01002175 if (ret) {
Jens Axboe7fe36312014-10-23 23:16:50 -06002176 log_err("Can't create helper thread: %s\n", strerror(ret));
Jens Axboe2e1df072012-02-09 11:15:02 +01002177 return 1;
2178 }
2179
Jens Axboe2e1df072012-02-09 11:15:02 +01002180 dprint(FD_MUTEX, "wait on startup_mutex\n");
2181 fio_mutex_down(startup_mutex);
2182 dprint(FD_MUTEX, "done waiting on startup_mutex\n");
2183 return 0;
2184}
2185
Jens Axboe2e1df072012-02-09 11:15:02 +01002186int fio_backend(void)
2187{
2188 struct thread_data *td;
2189 int i;
2190
2191 if (exec_profile) {
2192 if (load_profile(exec_profile))
2193 return 1;
2194 free(exec_profile);
2195 exec_profile = NULL;
2196 }
2197 if (!thread_number)
2198 return 0;
2199
2200 if (write_bw_log) {
Jens Axboe38a812d2014-07-03 09:10:39 -06002201 struct log_params p = {
2202 .log_type = IO_LOG_TYPE_BW,
2203 };
2204
2205 setup_log(&agg_io_log[DDIR_READ], &p, "agg-read_bw.log");
2206 setup_log(&agg_io_log[DDIR_WRITE], &p, "agg-write_bw.log");
2207 setup_log(&agg_io_log[DDIR_TRIM], &p, "agg-trim_bw.log");
Jens Axboe2e1df072012-02-09 11:15:02 +01002208 }
2209
Jens Axboe521da522012-08-02 11:21:36 +02002210 startup_mutex = fio_mutex_init(FIO_MUTEX_LOCKED);
Jens Axboe2e1df072012-02-09 11:15:02 +01002211 if (startup_mutex == NULL)
2212 return 1;
Jens Axboe2e1df072012-02-09 11:15:02 +01002213
2214 set_genesis_time();
Jens Axboecef91752013-04-26 17:05:57 -06002215 stat_init();
Jens Axboe7fe36312014-10-23 23:16:50 -06002216 create_helper_thread();
Jens Axboe2e1df072012-02-09 11:15:02 +01002217
2218 cgroup_list = smalloc(sizeof(*cgroup_list));
2219 INIT_FLIST_HEAD(cgroup_list);
2220
2221 run_threads();
2222
Jens Axboe7fe36312014-10-23 23:16:50 -06002223 wait_for_helper_thread_exit();
Jens Axboe62244c92014-10-23 08:53:33 -06002224
Jens Axboe2e1df072012-02-09 11:15:02 +01002225 if (!fio_abort) {
Jens Axboe2e627242014-07-25 09:51:56 +02002226 __show_run_stats();
Jens Axboe2e1df072012-02-09 11:15:02 +01002227 if (write_bw_log) {
Jens Axboe987c4f42014-07-01 16:29:12 -06002228 for (i = 0; i < DDIR_RWDIR_CNT; i++) {
2229 struct io_log *log = agg_io_log[i];
2230
Jens Axboe38a812d2014-07-03 09:10:39 -06002231 flush_log(log);
Jens Axboe633d8252014-07-02 13:36:34 -06002232 free_log(log);
Jens Axboe987c4f42014-07-01 16:29:12 -06002233 }
Jens Axboe2e1df072012-02-09 11:15:02 +01002234 }
2235 }
2236
Jens Axboe04247862014-10-23 23:04:37 -06002237 for_each_td(td, i) {
Jens Axboe2e1df072012-02-09 11:15:02 +01002238 fio_options_free(td);
Jens Axboeb89f89e2014-11-10 14:08:50 -07002239 if (td->rusage_sem) {
2240 fio_mutex_remove(td->rusage_sem);
2241 td->rusage_sem = NULL;
2242 }
Jens Axboe04247862014-10-23 23:04:37 -06002243 }
Jens Axboe2e1df072012-02-09 11:15:02 +01002244
Jens Axboea462bae2012-03-30 08:33:27 +02002245 free_disk_util();
Jens Axboe2e1df072012-02-09 11:15:02 +01002246 cgroup_kill(cgroup_list);
2247 sfree(cgroup_list);
2248 sfree(cgroup_mnt);
2249
2250 fio_mutex_remove(startup_mutex);
Jens Axboecef91752013-04-26 17:05:57 -06002251 stat_exit();
Jens Axboe2e1df072012-02-09 11:15:02 +01002252 return exit_value;
2253}