Commit | Line | Data |
---|---|---|
10ba535a JA |
1 | #include <unistd.h> |
2 | #include <fcntl.h> | |
3 | #include <string.h> | |
4 | #include <signal.h> | |
5 | #include <time.h> | |
0c6e7517 | 6 | #include <assert.h> |
10ba535a JA |
7 | |
8 | #include "fio.h" | |
5973cafb | 9 | #include "hash.h" |
10ba535a | 10 | |
5945b9b4 JA |
11 | /* |
12 | * Change this define to play with the timeout handling | |
13 | */ | |
14 | #undef FIO_USE_TIMEOUT | |
15 | ||
97601024 JA |
16 | struct io_completion_data { |
17 | int nr; /* input */ | |
97601024 JA |
18 | |
19 | int error; /* output */ | |
20 | unsigned long bytes_done[2]; /* output */ | |
21 | struct timeval time; /* output */ | |
22 | }; | |
23 | ||
10ba535a JA |
24 | /* |
25 | * The ->file_map[] contains a map of blocks we have or have not done io | |
26 | * to yet. Used to make sure we cover the entire range in a fair fashion. | |
27 | */ | |
aec2de20 | 28 | static int random_map_free(struct fio_file *f, const unsigned long long block) |
10ba535a | 29 | { |
aec2de20 JA |
30 | unsigned int idx = RAND_MAP_IDX(f, block); |
31 | unsigned int bit = RAND_MAP_BIT(f, block); | |
10ba535a | 32 | |
84422acd JA |
33 | dprint(FD_RANDOM, "free: b=%llu, idx=%u, bit=%u\n", block, idx, bit); |
34 | ||
10ba535a JA |
35 | return (f->file_map[idx] & (1UL << bit)) == 0; |
36 | } | |
37 | ||
df415585 JA |
38 | /* |
39 | * Mark a given offset as used in the map. | |
40 | */ | |
9bf2061e | 41 | static void mark_random_map(struct thread_data *td, struct io_u *io_u) |
df415585 | 42 | { |
2dc1bbeb | 43 | unsigned int min_bs = td->o.rw_min_bs; |
9bf2061e | 44 | struct fio_file *f = io_u->file; |
a00735e6 JA |
45 | unsigned long long block; |
46 | unsigned int blocks; | |
c685b5b2 | 47 | unsigned int nr_blocks; |
df415585 | 48 | |
b9c5b644 | 49 | block = (io_u->offset - f->file_offset) / (unsigned long long) min_bs; |
a00735e6 | 50 | blocks = 0; |
c685b5b2 JA |
51 | nr_blocks = (io_u->buflen + min_bs - 1) / min_bs; |
52 | ||
53 | while (blocks < nr_blocks) { | |
df415585 JA |
54 | unsigned int idx, bit; |
55 | ||
1e3d53ac JA |
56 | /* |
57 | * If we have a mixed random workload, we may | |
58 | * encounter blocks we already did IO to. | |
59 | */ | |
aec2de20 | 60 | if ((td->o.ddir_nr == 1) && !random_map_free(f, block)) |
df415585 JA |
61 | break; |
62 | ||
aec2de20 JA |
63 | idx = RAND_MAP_IDX(f, block); |
64 | bit = RAND_MAP_BIT(f, block); | |
df415585 | 65 | |
0032bf9f | 66 | fio_assert(td, idx < f->num_maps); |
df415585 JA |
67 | |
68 | f->file_map[idx] |= (1UL << bit); | |
69 | block++; | |
70 | blocks++; | |
71 | } | |
72 | ||
a00735e6 JA |
73 | if ((blocks * min_bs) < io_u->buflen) |
74 | io_u->buflen = blocks * min_bs; | |
df415585 JA |
75 | } |
76 | ||
2ba1c290 | 77 | static inline unsigned long long last_block(struct thread_data *td, |
4ba66134 JA |
78 | struct fio_file *f, |
79 | enum fio_ddir ddir) | |
2ba1c290 JA |
80 | { |
81 | unsigned long long max_blocks; | |
82 | ||
84422acd | 83 | max_blocks = f->io_size / (unsigned long long) td->o.min_bs[ddir]; |
2ba1c290 JA |
84 | if (!max_blocks) |
85 | return 0; | |
86 | ||
67778e88 | 87 | return max_blocks; |
2ba1c290 JA |
88 | } |
89 | ||
10ba535a JA |
90 | /* |
91 | * Return the next free block in the map. | |
92 | */ | |
93 | static int get_next_free_block(struct thread_data *td, struct fio_file *f, | |
4ba66134 | 94 | enum fio_ddir ddir, unsigned long long *b) |
10ba535a | 95 | { |
84422acd | 96 | unsigned long long min_bs = td->o.rw_min_bs; |
10ba535a JA |
97 | int i; |
98 | ||
c685b5b2 JA |
99 | i = f->last_free_lookup; |
100 | *b = (i * BLOCKS_PER_MAP); | |
84422acd | 101 | while ((*b) * min_bs < f->real_file_size) { |
10ba535a | 102 | if (f->file_map[i] != -1UL) { |
b12ebc65 | 103 | *b += fio_ffz(f->file_map[i]); |
4ba66134 | 104 | if (*b > last_block(td, f, ddir)) |
2ba1c290 | 105 | break; |
c685b5b2 | 106 | f->last_free_lookup = i; |
10ba535a JA |
107 | return 0; |
108 | } | |
109 | ||
110 | *b += BLOCKS_PER_MAP; | |
111 | i++; | |
112 | } | |
113 | ||
2ba1c290 | 114 | dprint(FD_IO, "failed finding a free block\n"); |
10ba535a JA |
115 | return 1; |
116 | } | |
117 | ||
ec4015da | 118 | static int get_next_rand_offset(struct thread_data *td, struct fio_file *f, |
4ba66134 | 119 | enum fio_ddir ddir, unsigned long long *b) |
ec4015da | 120 | { |
84422acd | 121 | unsigned long long r; |
ec4015da JA |
122 | int loops = 5; |
123 | ||
124 | do { | |
125 | r = os_random_long(&td->random_state); | |
84422acd | 126 | dprint(FD_RANDOM, "off rand %llu\n", r); |
5ec10eaa JA |
127 | *b = (last_block(td, f, ddir) - 1) |
128 | * (r / ((unsigned long long) RAND_MAX + 1.0)); | |
2ba1c290 | 129 | |
43c63a78 JA |
130 | /* |
131 | * if we are not maintaining a random map, we are done. | |
132 | */ | |
303032ae | 133 | if (!file_randommap(td, f)) |
43c63a78 JA |
134 | return 0; |
135 | ||
136 | /* | |
84422acd | 137 | * calculate map offset and check if it's free |
43c63a78 | 138 | */ |
aec2de20 | 139 | if (random_map_free(f, *b)) |
43c63a78 JA |
140 | return 0; |
141 | ||
84422acd JA |
142 | dprint(FD_RANDOM, "get_next_rand_offset: offset %llu busy\n", |
143 | *b); | |
43c63a78 | 144 | } while (--loops); |
ec4015da JA |
145 | |
146 | /* | |
43c63a78 JA |
147 | * we get here, if we didn't suceed in looking up a block. generate |
148 | * a random start offset into the filemap, and find the first free | |
149 | * block from there. | |
ec4015da | 150 | */ |
43c63a78 JA |
151 | loops = 10; |
152 | do { | |
153 | f->last_free_lookup = (f->num_maps - 1) * (r / (RAND_MAX+1.0)); | |
4ba66134 | 154 | if (!get_next_free_block(td, f, ddir, b)) |
43c63a78 | 155 | return 0; |
ec4015da | 156 | |
43c63a78 JA |
157 | r = os_random_long(&td->random_state); |
158 | } while (--loops); | |
159 | ||
160 | /* | |
161 | * that didn't work either, try exhaustive search from the start | |
162 | */ | |
163 | f->last_free_lookup = 0; | |
4ba66134 | 164 | return get_next_free_block(td, f, ddir, b); |
ec4015da JA |
165 | } |
166 | ||
10ba535a JA |
167 | /* |
168 | * For random io, generate a random new block and see if it's used. Repeat | |
169 | * until we find a free one. For sequential io, just return the end of | |
170 | * the last io issued. | |
171 | */ | |
9bf2061e | 172 | static int get_next_offset(struct thread_data *td, struct io_u *io_u) |
10ba535a | 173 | { |
9bf2061e | 174 | struct fio_file *f = io_u->file; |
ec4015da | 175 | unsigned long long b; |
4ba66134 | 176 | enum fio_ddir ddir = io_u->ddir; |
10ba535a | 177 | |
ec4015da JA |
178 | if (td_random(td) && (td->o.ddir_nr && !--td->ddir_nr)) { |
179 | td->ddir_nr = td->o.ddir_nr; | |
211097b2 | 180 | |
4ba66134 | 181 | if (get_next_rand_offset(td, f, ddir, &b)) |
bca4ed4d | 182 | return 1; |
43063a1c | 183 | } else { |
0c3d768a | 184 | if (f->last_pos >= f->real_file_size) { |
4ba66134 JA |
185 | if (!td_random(td) || |
186 | get_next_rand_offset(td, f, ddir, &b)) | |
0c3d768a | 187 | return 1; |
bcdedd0a | 188 | } else |
4ba66134 | 189 | b = (f->last_pos - f->file_offset) / td->o.min_bs[ddir]; |
43063a1c | 190 | } |
10ba535a | 191 | |
009bd847 JA |
192 | io_u->offset = b * td->o.min_bs[ddir]; |
193 | if (io_u->offset >= f->io_size) { | |
194 | dprint(FD_IO, "get_next_offset: offset %llu >= io_size %llu\n", | |
195 | io_u->offset, f->io_size); | |
196 | return 1; | |
197 | } | |
198 | ||
199 | io_u->offset += f->file_offset; | |
2ba1c290 JA |
200 | if (io_u->offset >= f->real_file_size) { |
201 | dprint(FD_IO, "get_next_offset: offset %llu >= size %llu\n", | |
202 | io_u->offset, f->real_file_size); | |
10ba535a | 203 | return 1; |
2ba1c290 | 204 | } |
10ba535a JA |
205 | |
206 | return 0; | |
207 | } | |
208 | ||
9bf2061e | 209 | static unsigned int get_next_buflen(struct thread_data *td, struct io_u *io_u) |
10ba535a | 210 | { |
bca4ed4d | 211 | const int ddir = io_u->ddir; |
10ba535a JA |
212 | unsigned int buflen; |
213 | long r; | |
214 | ||
2dc1bbeb JA |
215 | if (td->o.min_bs[ddir] == td->o.max_bs[ddir]) |
216 | buflen = td->o.min_bs[ddir]; | |
10ba535a JA |
217 | else { |
218 | r = os_random_long(&td->bsrange_state); | |
5ec10eaa JA |
219 | if (!td->o.bssplit_nr) { |
220 | buflen = (unsigned int) | |
221 | (1 + (double) (td->o.max_bs[ddir] - 1) | |
222 | * r / (RAND_MAX + 1.0)); | |
223 | } else { | |
564ca972 JA |
224 | long perc = 0; |
225 | unsigned int i; | |
226 | ||
227 | for (i = 0; i < td->o.bssplit_nr; i++) { | |
228 | struct bssplit *bsp = &td->o.bssplit[i]; | |
229 | ||
230 | buflen = bsp->bs; | |
231 | perc += bsp->perc; | |
232 | if (r <= ((LONG_MAX / 100L) * perc)) | |
233 | break; | |
234 | } | |
235 | } | |
5ec10eaa JA |
236 | if (!td->o.bs_unaligned) { |
237 | buflen = (buflen + td->o.min_bs[ddir] - 1) | |
238 | & ~(td->o.min_bs[ddir] - 1); | |
239 | } | |
10ba535a JA |
240 | } |
241 | ||
4ba66134 JA |
242 | if (io_u->offset + buflen > io_u->file->real_file_size) { |
243 | dprint(FD_IO, "lower buflen %u -> %u (ddir=%d)\n", buflen, | |
244 | td->o.min_bs[ddir], ddir); | |
6a5e6884 | 245 | buflen = td->o.min_bs[ddir]; |
4ba66134 | 246 | } |
6a5e6884 | 247 | |
10ba535a JA |
248 | return buflen; |
249 | } | |
250 | ||
afe24a5a JA |
251 | static void set_rwmix_bytes(struct thread_data *td) |
252 | { | |
e4928662 | 253 | unsigned long issues; |
afe24a5a JA |
254 | unsigned int diff; |
255 | ||
256 | /* | |
257 | * we do time or byte based switch. this is needed because | |
258 | * buffered writes may issue a lot quicker than they complete, | |
259 | * whereas reads do not. | |
260 | */ | |
e4928662 | 261 | issues = td->io_issues[td->rwmix_ddir] - td->rwmix_issues; |
e47f799f | 262 | diff = td->o.rwmix[td->rwmix_ddir ^ 1]; |
afe24a5a | 263 | |
e4928662 JA |
264 | td->rwmix_issues = td->io_issues[td->rwmix_ddir] |
265 | + (issues * ((100 - diff)) / diff); | |
e47f799f JA |
266 | } |
267 | ||
268 | static inline enum fio_ddir get_rand_ddir(struct thread_data *td) | |
269 | { | |
270 | unsigned int v; | |
271 | long r; | |
272 | ||
273 | r = os_random_long(&td->rwmix_state); | |
274 | v = 1 + (int) (100.0 * (r / (RAND_MAX + 1.0))); | |
275 | if (v < td->o.rwmix[DDIR_READ]) | |
276 | return DDIR_READ; | |
277 | ||
278 | return DDIR_WRITE; | |
afe24a5a JA |
279 | } |
280 | ||
10ba535a JA |
281 | /* |
282 | * Return the data direction for the next io_u. If the job is a | |
283 | * mixed read/write workload, check the rwmix cycle and switch if | |
284 | * necessary. | |
285 | */ | |
1e97cce9 | 286 | static enum fio_ddir get_rw_ddir(struct thread_data *td) |
10ba535a JA |
287 | { |
288 | if (td_rw(td)) { | |
10ba535a JA |
289 | /* |
290 | * Check if it's time to seed a new data direction. | |
291 | */ | |
e4928662 | 292 | if (td->io_issues[td->rwmix_ddir] >= td->rwmix_issues) { |
e47f799f | 293 | unsigned long long max_bytes; |
5ec10eaa | 294 | enum fio_ddir ddir; |
e47f799f JA |
295 | |
296 | /* | |
297 | * Put a top limit on how many bytes we do for | |
298 | * one data direction, to avoid overflowing the | |
299 | * ranges too much | |
300 | */ | |
301 | ddir = get_rand_ddir(td); | |
302 | max_bytes = td->this_io_bytes[ddir]; | |
5ec10eaa JA |
303 | if (max_bytes >= |
304 | (td->o.size * td->o.rwmix[ddir] / 100)) { | |
e4928662 | 305 | if (!td->rw_end_set[ddir]) |
38d77cae | 306 | td->rw_end_set[ddir] = 1; |
e4928662 | 307 | |
e47f799f | 308 | ddir ^= 1; |
38d77cae | 309 | } |
e47f799f JA |
310 | |
311 | if (ddir != td->rwmix_ddir) | |
312 | set_rwmix_bytes(td); | |
313 | ||
314 | td->rwmix_ddir = ddir; | |
10ba535a JA |
315 | } |
316 | return td->rwmix_ddir; | |
317 | } else if (td_read(td)) | |
318 | return DDIR_READ; | |
319 | else | |
320 | return DDIR_WRITE; | |
321 | } | |
322 | ||
60f2c658 JA |
323 | static void put_file_log(struct thread_data *td, struct fio_file *f) |
324 | { | |
325 | int ret = put_file(td, f); | |
326 | ||
327 | if (ret) | |
328 | td_verror(td, ret, "file close"); | |
329 | } | |
330 | ||
10ba535a JA |
331 | void put_io_u(struct thread_data *td, struct io_u *io_u) |
332 | { | |
0c6e7517 JA |
333 | assert((io_u->flags & IO_U_F_FREE) == 0); |
334 | io_u->flags |= IO_U_F_FREE; | |
335 | ||
60f2c658 JA |
336 | if (io_u->file) |
337 | put_file_log(td, io_u->file); | |
2dbdab7e | 338 | |
10ba535a JA |
339 | io_u->file = NULL; |
340 | list_del(&io_u->list); | |
341 | list_add(&io_u->list, &td->io_u_freelist); | |
342 | td->cur_depth--; | |
343 | } | |
344 | ||
755200a3 JA |
345 | void requeue_io_u(struct thread_data *td, struct io_u **io_u) |
346 | { | |
347 | struct io_u *__io_u = *io_u; | |
348 | ||
4d2e0f49 | 349 | __io_u->flags |= IO_U_F_FREE; |
e4f54adb JA |
350 | if ((__io_u->flags & IO_U_F_FLIGHT) && (__io_u->ddir != DDIR_SYNC)) |
351 | td->io_issues[__io_u->ddir]--; | |
5ec10eaa | 352 | |
4d2e0f49 JA |
353 | __io_u->flags &= ~IO_U_F_FLIGHT; |
354 | ||
755200a3 JA |
355 | list_del(&__io_u->list); |
356 | list_add_tail(&__io_u->list, &td->io_u_requeues); | |
357 | td->cur_depth--; | |
358 | *io_u = NULL; | |
359 | } | |
360 | ||
9bf2061e | 361 | static int fill_io_u(struct thread_data *td, struct io_u *io_u) |
10ba535a | 362 | { |
b4c5e1ac JA |
363 | if (td->io_ops->flags & FIO_NOIO) |
364 | goto out; | |
365 | ||
87dc1ab1 JA |
366 | /* |
367 | * see if it's time to sync | |
368 | */ | |
2dc1bbeb JA |
369 | if (td->o.fsync_blocks && |
370 | !(td->io_issues[DDIR_WRITE] % td->o.fsync_blocks) && | |
371 | td->io_issues[DDIR_WRITE] && should_fsync(td)) { | |
87dc1ab1 | 372 | io_u->ddir = DDIR_SYNC; |
c38e9468 | 373 | goto out; |
87dc1ab1 JA |
374 | } |
375 | ||
a00735e6 JA |
376 | io_u->ddir = get_rw_ddir(td); |
377 | ||
48f5abd3 JA |
378 | /* |
379 | * See if it's time to switch to a new zone | |
380 | */ | |
381 | if (td->zone_bytes >= td->o.zone_size) { | |
382 | td->zone_bytes = 0; | |
383 | io_u->file->last_pos += td->o.zone_skip; | |
384 | td->io_skip_bytes += td->o.zone_skip; | |
385 | } | |
386 | ||
10ba535a | 387 | /* |
c685b5b2 JA |
388 | * No log, let the seq/rand engine retrieve the next buflen and |
389 | * position. | |
10ba535a | 390 | */ |
2ba1c290 JA |
391 | if (get_next_offset(td, io_u)) { |
392 | dprint(FD_IO, "io_u %p, failed getting offset\n", io_u); | |
bca4ed4d | 393 | return 1; |
2ba1c290 | 394 | } |
10ba535a | 395 | |
9bf2061e | 396 | io_u->buflen = get_next_buflen(td, io_u); |
2ba1c290 JA |
397 | if (!io_u->buflen) { |
398 | dprint(FD_IO, "io_u %p, failed getting buflen\n", io_u); | |
bca4ed4d | 399 | return 1; |
2ba1c290 | 400 | } |
bca4ed4d | 401 | |
2ba1c290 JA |
402 | if (io_u->offset + io_u->buflen > io_u->file->real_file_size) { |
403 | dprint(FD_IO, "io_u %p, offset too large\n", io_u); | |
4ba66134 JA |
404 | dprint(FD_IO, " off=%llu/%lu > %llu\n", io_u->offset, |
405 | io_u->buflen, io_u->file->real_file_size); | |
6a5e6884 | 406 | return 1; |
2ba1c290 | 407 | } |
6a5e6884 | 408 | |
bca4ed4d JA |
409 | /* |
410 | * mark entry before potentially trimming io_u | |
411 | */ | |
303032ae | 412 | if (td_random(td) && file_randommap(td, io_u->file)) |
9bf2061e | 413 | mark_random_map(td, io_u); |
bca4ed4d JA |
414 | |
415 | /* | |
416 | * If using a write iolog, store this entry. | |
417 | */ | |
c38e9468 | 418 | out: |
2ba1c290 | 419 | dprint_io_u(io_u, "fill_io_u"); |
d9d91e39 | 420 | td->zone_bytes += io_u->buflen; |
f29b25a3 | 421 | log_io_u(td, io_u); |
bca4ed4d | 422 | return 0; |
10ba535a JA |
423 | } |
424 | ||
838bc709 JA |
425 | static void __io_u_mark_map(unsigned int *map, unsigned int nr) |
426 | { | |
427 | int index = 0; | |
428 | ||
429 | switch (nr) { | |
430 | default: | |
431 | index = 6; | |
432 | break; | |
433 | case 33 ... 64: | |
434 | index = 5; | |
435 | break; | |
436 | case 17 ... 32: | |
437 | index = 4; | |
438 | break; | |
439 | case 9 ... 16: | |
440 | index = 3; | |
441 | break; | |
442 | case 5 ... 8: | |
443 | index = 2; | |
444 | break; | |
445 | case 1 ... 4: | |
446 | index = 1; | |
447 | case 0: | |
448 | break; | |
449 | } | |
450 | ||
451 | map[index]++; | |
452 | } | |
453 | ||
454 | void io_u_mark_submit(struct thread_data *td, unsigned int nr) | |
455 | { | |
456 | __io_u_mark_map(td->ts.io_u_submit, nr); | |
457 | td->ts.total_submit++; | |
458 | } | |
459 | ||
460 | void io_u_mark_complete(struct thread_data *td, unsigned int nr) | |
461 | { | |
462 | __io_u_mark_map(td->ts.io_u_complete, nr); | |
463 | td->ts.total_complete++; | |
464 | } | |
465 | ||
d8005759 | 466 | void io_u_mark_depth(struct thread_data *td, unsigned int nr) |
71619dc2 JA |
467 | { |
468 | int index = 0; | |
469 | ||
470 | switch (td->cur_depth) { | |
471 | default: | |
a783e61a JA |
472 | index = 6; |
473 | break; | |
71619dc2 | 474 | case 32 ... 63: |
a783e61a JA |
475 | index = 5; |
476 | break; | |
71619dc2 | 477 | case 16 ... 31: |
a783e61a JA |
478 | index = 4; |
479 | break; | |
71619dc2 | 480 | case 8 ... 15: |
a783e61a JA |
481 | index = 3; |
482 | break; | |
71619dc2 | 483 | case 4 ... 7: |
a783e61a JA |
484 | index = 2; |
485 | break; | |
71619dc2 | 486 | case 2 ... 3: |
a783e61a | 487 | index = 1; |
71619dc2 JA |
488 | case 1: |
489 | break; | |
490 | } | |
491 | ||
3bec7ae0 | 492 | td->ts.io_u_map[index] += nr; |
71619dc2 JA |
493 | } |
494 | ||
04a0feae JA |
495 | static void io_u_mark_lat_usec(struct thread_data *td, unsigned long usec) |
496 | { | |
497 | int index = 0; | |
498 | ||
499 | assert(usec < 1000); | |
500 | ||
501 | switch (usec) { | |
502 | case 750 ... 999: | |
503 | index = 9; | |
504 | break; | |
505 | case 500 ... 749: | |
506 | index = 8; | |
507 | break; | |
508 | case 250 ... 499: | |
509 | index = 7; | |
510 | break; | |
511 | case 100 ... 249: | |
512 | index = 6; | |
513 | break; | |
514 | case 50 ... 99: | |
515 | index = 5; | |
516 | break; | |
517 | case 20 ... 49: | |
518 | index = 4; | |
519 | break; | |
520 | case 10 ... 19: | |
521 | index = 3; | |
522 | break; | |
523 | case 4 ... 9: | |
524 | index = 2; | |
525 | break; | |
526 | case 2 ... 3: | |
527 | index = 1; | |
528 | case 0 ... 1: | |
529 | break; | |
530 | } | |
531 | ||
532 | assert(index < FIO_IO_U_LAT_U_NR); | |
533 | td->ts.io_u_lat_u[index]++; | |
534 | } | |
535 | ||
536 | static void io_u_mark_lat_msec(struct thread_data *td, unsigned long msec) | |
ec118304 JA |
537 | { |
538 | int index = 0; | |
539 | ||
540 | switch (msec) { | |
541 | default: | |
04a0feae JA |
542 | index = 11; |
543 | break; | |
8abdce66 | 544 | case 1000 ... 1999: |
04a0feae JA |
545 | index = 10; |
546 | break; | |
8abdce66 | 547 | case 750 ... 999: |
04a0feae JA |
548 | index = 9; |
549 | break; | |
8abdce66 | 550 | case 500 ... 749: |
04a0feae JA |
551 | index = 8; |
552 | break; | |
8abdce66 | 553 | case 250 ... 499: |
04a0feae JA |
554 | index = 7; |
555 | break; | |
8abdce66 | 556 | case 100 ... 249: |
04a0feae JA |
557 | index = 6; |
558 | break; | |
8abdce66 | 559 | case 50 ... 99: |
04a0feae JA |
560 | index = 5; |
561 | break; | |
8abdce66 | 562 | case 20 ... 49: |
04a0feae JA |
563 | index = 4; |
564 | break; | |
8abdce66 | 565 | case 10 ... 19: |
04a0feae JA |
566 | index = 3; |
567 | break; | |
8abdce66 | 568 | case 4 ... 9: |
04a0feae JA |
569 | index = 2; |
570 | break; | |
ec118304 | 571 | case 2 ... 3: |
04a0feae | 572 | index = 1; |
ec118304 JA |
573 | case 0 ... 1: |
574 | break; | |
575 | } | |
576 | ||
04a0feae JA |
577 | assert(index < FIO_IO_U_LAT_M_NR); |
578 | td->ts.io_u_lat_m[index]++; | |
579 | } | |
580 | ||
581 | static void io_u_mark_latency(struct thread_data *td, unsigned long usec) | |
582 | { | |
583 | if (usec < 1000) | |
584 | io_u_mark_lat_usec(td, usec); | |
585 | else | |
586 | io_u_mark_lat_msec(td, usec / 1000); | |
ec118304 JA |
587 | } |
588 | ||
0aabe160 JA |
589 | /* |
590 | * Get next file to service by choosing one at random | |
591 | */ | |
1c178180 JA |
592 | static struct fio_file *get_next_file_rand(struct thread_data *td, int goodf, |
593 | int badf) | |
0aabe160 | 594 | { |
0aabe160 | 595 | struct fio_file *f; |
1c178180 | 596 | int fno; |
0aabe160 JA |
597 | |
598 | do { | |
7c83c089 JA |
599 | long r = os_random_long(&td->next_file_state); |
600 | ||
5ec10eaa JA |
601 | fno = (unsigned int) ((double) td->o.nr_files |
602 | * (r / (RAND_MAX + 1.0))); | |
126d65c6 | 603 | f = td->files[fno]; |
059e63c0 JA |
604 | if (f->flags & FIO_FILE_DONE) |
605 | continue; | |
1c178180 | 606 | |
2ba1c290 JA |
607 | if ((!goodf || (f->flags & goodf)) && !(f->flags & badf)) { |
608 | dprint(FD_FILE, "get_next_file_rand: %p\n", f); | |
0aabe160 | 609 | return f; |
2ba1c290 | 610 | } |
0aabe160 JA |
611 | } while (1); |
612 | } | |
613 | ||
614 | /* | |
615 | * Get next file to service by doing round robin between all available ones | |
616 | */ | |
1c178180 JA |
617 | static struct fio_file *get_next_file_rr(struct thread_data *td, int goodf, |
618 | int badf) | |
3d7c391d JA |
619 | { |
620 | unsigned int old_next_file = td->next_file; | |
621 | struct fio_file *f; | |
622 | ||
623 | do { | |
126d65c6 | 624 | f = td->files[td->next_file]; |
3d7c391d JA |
625 | |
626 | td->next_file++; | |
2dc1bbeb | 627 | if (td->next_file >= td->o.nr_files) |
3d7c391d JA |
628 | td->next_file = 0; |
629 | ||
d5ed68ea JA |
630 | if (f->flags & FIO_FILE_DONE) { |
631 | f = NULL; | |
059e63c0 | 632 | continue; |
d5ed68ea | 633 | } |
059e63c0 | 634 | |
1c178180 | 635 | if ((!goodf || (f->flags & goodf)) && !(f->flags & badf)) |
3d7c391d JA |
636 | break; |
637 | ||
638 | f = NULL; | |
639 | } while (td->next_file != old_next_file); | |
640 | ||
2ba1c290 | 641 | dprint(FD_FILE, "get_next_file_rr: %p\n", f); |
3d7c391d JA |
642 | return f; |
643 | } | |
644 | ||
bdb4e2e9 JA |
645 | static struct fio_file *get_next_file(struct thread_data *td) |
646 | { | |
1907dbc6 JA |
647 | struct fio_file *f; |
648 | ||
2dc1bbeb | 649 | assert(td->o.nr_files <= td->files_index); |
1c178180 | 650 | |
2ba1c290 | 651 | if (!td->nr_open_files || td->nr_done_files >= td->o.nr_files) { |
5ec10eaa JA |
652 | dprint(FD_FILE, "get_next_file: nr_open=%d, nr_done=%d," |
653 | " nr_files=%d\n", td->nr_open_files, | |
654 | td->nr_done_files, | |
655 | td->o.nr_files); | |
bdb4e2e9 | 656 | return NULL; |
2ba1c290 | 657 | } |
bdb4e2e9 | 658 | |
1907dbc6 | 659 | f = td->file_service_file; |
f11bd94d | 660 | if (f && (f->flags & FIO_FILE_OPEN) && td->file_service_left--) |
2ba1c290 | 661 | goto out; |
1907dbc6 | 662 | |
2dc1bbeb | 663 | if (td->o.file_service_type == FIO_FSERVICE_RR) |
1c178180 | 664 | f = get_next_file_rr(td, FIO_FILE_OPEN, FIO_FILE_CLOSING); |
bdb4e2e9 | 665 | else |
1c178180 | 666 | f = get_next_file_rand(td, FIO_FILE_OPEN, FIO_FILE_CLOSING); |
1907dbc6 JA |
667 | |
668 | td->file_service_file = f; | |
669 | td->file_service_left = td->file_service_nr - 1; | |
2ba1c290 JA |
670 | out: |
671 | dprint(FD_FILE, "get_next_file: %p\n", f); | |
1907dbc6 | 672 | return f; |
bdb4e2e9 JA |
673 | } |
674 | ||
1c178180 JA |
675 | static struct fio_file *find_next_new_file(struct thread_data *td) |
676 | { | |
677 | struct fio_file *f; | |
678 | ||
1020a139 JA |
679 | if (!td->nr_open_files || td->nr_done_files >= td->o.nr_files) |
680 | return NULL; | |
681 | ||
2dc1bbeb | 682 | if (td->o.file_service_type == FIO_FSERVICE_RR) |
1c178180 JA |
683 | f = get_next_file_rr(td, 0, FIO_FILE_OPEN); |
684 | else | |
685 | f = get_next_file_rand(td, 0, FIO_FILE_OPEN); | |
686 | ||
687 | return f; | |
688 | } | |
689 | ||
429f6675 JA |
690 | static int set_io_u_file(struct thread_data *td, struct io_u *io_u) |
691 | { | |
692 | struct fio_file *f; | |
693 | ||
694 | do { | |
695 | f = get_next_file(td); | |
696 | if (!f) | |
697 | return 1; | |
698 | ||
699 | set_file: | |
700 | io_u->file = f; | |
701 | get_file(f); | |
702 | ||
703 | if (!fill_io_u(td, io_u)) | |
704 | break; | |
705 | ||
009bd847 JA |
706 | /* |
707 | * optimization to prevent close/open of the same file. This | |
708 | * way we preserve queueing etc. | |
709 | */ | |
710 | if (td->o.nr_files == 1 && td->o.time_based) { | |
60f2c658 | 711 | put_file_log(td, f); |
009bd847 JA |
712 | fio_file_reset(f); |
713 | goto set_file; | |
714 | } | |
715 | ||
429f6675 JA |
716 | /* |
717 | * td_io_close() does a put_file() as well, so no need to | |
718 | * do that here. | |
719 | */ | |
720 | io_u->file = NULL; | |
721 | td_io_close_file(td, f); | |
722 | f->flags |= FIO_FILE_DONE; | |
723 | td->nr_done_files++; | |
724 | ||
725 | /* | |
726 | * probably not the right place to do this, but see | |
727 | * if we need to open a new file | |
728 | */ | |
729 | if (td->nr_open_files < td->o.open_files && | |
730 | td->o.open_files != td->o.nr_files) { | |
731 | f = find_next_new_file(td); | |
732 | ||
733 | if (!f || td_io_open_file(td, f)) | |
734 | return 1; | |
735 | ||
736 | goto set_file; | |
737 | } | |
738 | } while (1); | |
739 | ||
740 | return 0; | |
741 | } | |
742 | ||
743 | ||
10ba535a JA |
744 | struct io_u *__get_io_u(struct thread_data *td) |
745 | { | |
746 | struct io_u *io_u = NULL; | |
747 | ||
755200a3 JA |
748 | if (!list_empty(&td->io_u_requeues)) |
749 | io_u = list_entry(td->io_u_requeues.next, struct io_u, list); | |
750 | else if (!queue_full(td)) { | |
10ba535a JA |
751 | io_u = list_entry(td->io_u_freelist.next, struct io_u, list); |
752 | ||
6040dabc | 753 | io_u->buflen = 0; |
10ba535a | 754 | io_u->resid = 0; |
755200a3 | 755 | io_u->file = NULL; |
d7762cf8 | 756 | io_u->end_io = NULL; |
755200a3 JA |
757 | } |
758 | ||
759 | if (io_u) { | |
0c6e7517 JA |
760 | assert(io_u->flags & IO_U_F_FREE); |
761 | io_u->flags &= ~IO_U_F_FREE; | |
762 | ||
755200a3 | 763 | io_u->error = 0; |
10ba535a JA |
764 | list_del(&io_u->list); |
765 | list_add(&io_u->list, &td->io_u_busylist); | |
766 | td->cur_depth++; | |
767 | } | |
768 | ||
769 | return io_u; | |
770 | } | |
771 | ||
772 | /* | |
773 | * Return an io_u to be processed. Gets a buflen and offset, sets direction, | |
774 | * etc. The returned io_u is fully ready to be prepped and submitted. | |
775 | */ | |
3d7c391d | 776 | struct io_u *get_io_u(struct thread_data *td) |
10ba535a | 777 | { |
3d7c391d | 778 | struct fio_file *f; |
10ba535a JA |
779 | struct io_u *io_u; |
780 | ||
781 | io_u = __get_io_u(td); | |
2ba1c290 JA |
782 | if (!io_u) { |
783 | dprint(FD_IO, "__get_io_u failed\n"); | |
10ba535a | 784 | return NULL; |
2ba1c290 | 785 | } |
10ba535a | 786 | |
755200a3 JA |
787 | /* |
788 | * from a requeue, io_u already setup | |
789 | */ | |
790 | if (io_u->file) | |
77f392bf | 791 | goto out; |
755200a3 | 792 | |
429f6675 JA |
793 | /* |
794 | * If using an iolog, grab next piece if any available. | |
795 | */ | |
796 | if (td->o.read_iolog_file) { | |
797 | if (read_iolog_get(td, io_u)) | |
798 | goto err_put; | |
2ba1c290 JA |
799 | } else if (set_io_u_file(td, io_u)) { |
800 | dprint(FD_IO, "io_u %p, setting file failed\n", io_u); | |
429f6675 | 801 | goto err_put; |
2ba1c290 | 802 | } |
5ec10eaa | 803 | |
429f6675 JA |
804 | f = io_u->file; |
805 | assert(f->flags & FIO_FILE_OPEN); | |
97af62ce | 806 | |
87dc1ab1 | 807 | if (io_u->ddir != DDIR_SYNC) { |
d0656a93 | 808 | if (!io_u->buflen && !(td->io_ops->flags & FIO_NOIO)) { |
2ba1c290 | 809 | dprint(FD_IO, "get_io_u: zero buflen on %p\n", io_u); |
429f6675 | 810 | goto err_put; |
2ba1c290 | 811 | } |
10ba535a | 812 | |
36167d82 | 813 | f->last_pos = io_u->offset + io_u->buflen; |
10ba535a | 814 | |
2dc1bbeb | 815 | if (td->o.verify != VERIFY_NONE) |
87dc1ab1 JA |
816 | populate_verify_io_u(td, io_u); |
817 | } | |
10ba535a | 818 | |
165faf16 JA |
819 | /* |
820 | * Set io data pointers. | |
821 | */ | |
d460eb31 | 822 | io_u->endpos = io_u->offset + io_u->buflen; |
cec6b55d JA |
823 | io_u->xfer_buf = io_u->buf; |
824 | io_u->xfer_buflen = io_u->buflen; | |
5973cafb JA |
825 | |
826 | if (td->o.refill_buffers && io_u->ddir == DDIR_WRITE) | |
827 | io_u_fill_buffer(td, io_u, io_u->xfer_buflen); | |
6ac7a331 | 828 | out: |
429f6675 JA |
829 | if (!td_io_prep(td, io_u)) { |
830 | fio_gettime(&io_u->start_time, NULL); | |
831 | return io_u; | |
36167d82 | 832 | } |
429f6675 | 833 | err_put: |
2ba1c290 | 834 | dprint(FD_IO, "get_io_u failed\n"); |
429f6675 JA |
835 | put_io_u(td, io_u); |
836 | return NULL; | |
10ba535a JA |
837 | } |
838 | ||
5451792e JA |
839 | void io_u_log_error(struct thread_data *td, struct io_u *io_u) |
840 | { | |
841 | const char *msg[] = { "read", "write", "sync" }; | |
842 | ||
843 | log_err("fio: io_u error"); | |
844 | ||
845 | if (io_u->file) | |
846 | log_err(" on file %s", io_u->file->file_name); | |
847 | ||
848 | log_err(": %s\n", strerror(io_u->error)); | |
849 | ||
5ec10eaa JA |
850 | log_err(" %s offset=%llu, buflen=%lu\n", msg[io_u->ddir], |
851 | io_u->offset, io_u->xfer_buflen); | |
5451792e JA |
852 | |
853 | if (!td->error) | |
854 | td_verror(td, io_u->error, "io_u error"); | |
855 | } | |
856 | ||
97601024 JA |
857 | static void io_completed(struct thread_data *td, struct io_u *io_u, |
858 | struct io_completion_data *icd) | |
10ba535a | 859 | { |
d85f5118 | 860 | unsigned long usec; |
10ba535a | 861 | |
2ba1c290 JA |
862 | dprint_io_u(io_u, "io complete"); |
863 | ||
0c6e7517 JA |
864 | assert(io_u->flags & IO_U_F_FLIGHT); |
865 | io_u->flags &= ~IO_U_F_FLIGHT; | |
866 | ||
87dc1ab1 JA |
867 | if (io_u->ddir == DDIR_SYNC) { |
868 | td->last_was_sync = 1; | |
869 | return; | |
870 | } | |
871 | ||
872 | td->last_was_sync = 0; | |
873 | ||
10ba535a JA |
874 | if (!io_u->error) { |
875 | unsigned int bytes = io_u->buflen - io_u->resid; | |
1e97cce9 | 876 | const enum fio_ddir idx = io_u->ddir; |
3af6ef39 | 877 | int ret; |
10ba535a JA |
878 | |
879 | td->io_blocks[idx]++; | |
880 | td->io_bytes[idx] += bytes; | |
10ba535a JA |
881 | td->this_io_bytes[idx] += bytes; |
882 | ||
d85f5118 | 883 | usec = utime_since(&io_u->issue_time, &icd->time); |
10ba535a | 884 | |
d85f5118 | 885 | add_clat_sample(td, idx, usec); |
02bcaa8c | 886 | add_bw_sample(td, idx, &icd->time); |
04a0feae | 887 | io_u_mark_latency(td, usec); |
10ba535a | 888 | |
660a1cb5 | 889 | if (td_write(td) && idx == DDIR_WRITE && |
8670579e | 890 | td->o.do_verify && |
41128405 | 891 | td->o.verify != VERIFY_NONE) |
10ba535a JA |
892 | log_io_piece(td, io_u); |
893 | ||
894 | icd->bytes_done[idx] += bytes; | |
3af6ef39 | 895 | |
d7762cf8 | 896 | if (io_u->end_io) { |
36690c9b | 897 | ret = io_u->end_io(td, io_u); |
3af6ef39 JA |
898 | if (ret && !icd->error) |
899 | icd->error = ret; | |
900 | } | |
5451792e | 901 | } else { |
10ba535a | 902 | icd->error = io_u->error; |
5451792e JA |
903 | io_u_log_error(td, io_u); |
904 | } | |
10ba535a JA |
905 | } |
906 | ||
d7762cf8 | 907 | static void init_icd(struct io_completion_data *icd, int nr) |
10ba535a | 908 | { |
02bcaa8c JA |
909 | fio_gettime(&icd->time, NULL); |
910 | ||
3af6ef39 JA |
911 | icd->nr = nr; |
912 | ||
10ba535a JA |
913 | icd->error = 0; |
914 | icd->bytes_done[0] = icd->bytes_done[1] = 0; | |
36167d82 JA |
915 | } |
916 | ||
97601024 JA |
917 | static void ios_completed(struct thread_data *td, |
918 | struct io_completion_data *icd) | |
36167d82 JA |
919 | { |
920 | struct io_u *io_u; | |
921 | int i; | |
922 | ||
10ba535a JA |
923 | for (i = 0; i < icd->nr; i++) { |
924 | io_u = td->io_ops->event(td, i); | |
925 | ||
926 | io_completed(td, io_u, icd); | |
927 | put_io_u(td, io_u); | |
928 | } | |
929 | } | |
97601024 | 930 | |
e7e6cfb4 JA |
931 | /* |
932 | * Complete a single io_u for the sync engines. | |
933 | */ | |
d7762cf8 | 934 | long io_u_sync_complete(struct thread_data *td, struct io_u *io_u) |
97601024 JA |
935 | { |
936 | struct io_completion_data icd; | |
937 | ||
d7762cf8 | 938 | init_icd(&icd, 1); |
97601024 JA |
939 | io_completed(td, io_u, &icd); |
940 | put_io_u(td, io_u); | |
941 | ||
942 | if (!icd.error) | |
943 | return icd.bytes_done[0] + icd.bytes_done[1]; | |
944 | ||
37e974a5 | 945 | td_verror(td, icd.error, "io_u_sync_complete"); |
97601024 JA |
946 | return -1; |
947 | } | |
948 | ||
e7e6cfb4 JA |
949 | /* |
950 | * Called to complete min_events number of io for the async engines. | |
951 | */ | |
d7762cf8 | 952 | long io_u_queued_complete(struct thread_data *td, int min_events) |
97601024 | 953 | { |
97601024 | 954 | struct io_completion_data icd; |
00de55ef | 955 | struct timespec *tvp = NULL; |
97601024 | 956 | int ret; |
4d06a338 | 957 | struct timespec ts = { .tv_sec = 0, .tv_nsec = 0, }; |
97601024 | 958 | |
b271fe62 JA |
959 | dprint(FD_IO, "io_u_queued_completed: min=%d\n", min_events); |
960 | ||
4d06a338 | 961 | if (!min_events) |
00de55ef | 962 | tvp = &ts; |
97601024 | 963 | |
00de55ef | 964 | ret = td_io_getevents(td, min_events, td->cur_depth, tvp); |
97601024 | 965 | if (ret < 0) { |
e1161c32 | 966 | td_verror(td, -ret, "td_io_getevents"); |
97601024 JA |
967 | return ret; |
968 | } else if (!ret) | |
969 | return ret; | |
970 | ||
d7762cf8 | 971 | init_icd(&icd, ret); |
97601024 JA |
972 | ios_completed(td, &icd); |
973 | if (!icd.error) | |
974 | return icd.bytes_done[0] + icd.bytes_done[1]; | |
975 | ||
37e974a5 | 976 | td_verror(td, icd.error, "io_u_queued_complete"); |
97601024 JA |
977 | return -1; |
978 | } | |
7e77dd02 JA |
979 | |
980 | /* | |
981 | * Call when io_u is really queued, to update the submission latency. | |
982 | */ | |
983 | void io_u_queued(struct thread_data *td, struct io_u *io_u) | |
984 | { | |
985 | unsigned long slat_time; | |
986 | ||
d85f5118 | 987 | slat_time = utime_since(&io_u->start_time, &io_u->issue_time); |
7e77dd02 JA |
988 | add_slat_sample(td, io_u->ddir, slat_time); |
989 | } | |
433afcb4 | 990 | |
5973cafb JA |
991 | /* |
992 | * "randomly" fill the buffer contents | |
993 | */ | |
994 | void io_u_fill_buffer(struct thread_data *td, struct io_u *io_u, | |
995 | unsigned int max_bs) | |
996 | { | |
997 | long *ptr = io_u->buf; | |
998 | ||
999 | if (!td->o.zero_buffers) { | |
1000 | while ((void *) ptr - io_u->buf < max_bs) { | |
1001 | *ptr = rand() * GOLDEN_RATIO_PRIME; | |
1002 | ptr++; | |
1003 | } | |
1004 | } else | |
1005 | memset(ptr, 0, max_bs); | |
1006 | } | |
1007 | ||
55bc9728 | 1008 | #ifdef FIO_USE_TIMEOUT |
433afcb4 JA |
1009 | void io_u_set_timeout(struct thread_data *td) |
1010 | { | |
1011 | assert(td->cur_depth); | |
1012 | ||
1013 | td->timer.it_interval.tv_sec = 0; | |
1014 | td->timer.it_interval.tv_usec = 0; | |
1015 | td->timer.it_value.tv_sec = IO_U_TIMEOUT + IO_U_TIMEOUT_INC; | |
1016 | td->timer.it_value.tv_usec = 0; | |
1017 | setitimer(ITIMER_REAL, &td->timer, NULL); | |
1018 | fio_gettime(&td->timeout_end, NULL); | |
1019 | } | |
5945b9b4 JA |
1020 | |
1021 | static void io_u_dump(struct io_u *io_u) | |
1022 | { | |
1023 | unsigned long t_start = mtime_since_now(&io_u->start_time); | |
1024 | unsigned long t_issue = mtime_since_now(&io_u->issue_time); | |
1025 | ||
1026 | log_err("io_u=%p, t_start=%lu, t_issue=%lu\n", io_u, t_start, t_issue); | |
5ec10eaa JA |
1027 | log_err(" buf=%p/%p, len=%lu/%lu, offset=%llu\n", io_u->buf, |
1028 | io_u->xfer_buf, io_u->buflen, | |
1029 | io_u->xfer_buflen, | |
1030 | io_u->offset); | |
5945b9b4 JA |
1031 | log_err(" ddir=%d, fname=%s\n", io_u->ddir, io_u->file->file_name); |
1032 | } | |
55bc9728 JA |
1033 | #else |
1034 | void io_u_set_timeout(struct thread_data fio_unused *td) | |
1035 | { | |
1036 | } | |
1037 | #endif | |
433afcb4 | 1038 | |
55bc9728 | 1039 | #ifdef FIO_USE_TIMEOUT |
433afcb4 JA |
1040 | static void io_u_timeout_handler(int fio_unused sig) |
1041 | { | |
1042 | struct thread_data *td, *__td; | |
1043 | pid_t pid = getpid(); | |
5945b9b4 JA |
1044 | struct list_head *entry; |
1045 | struct io_u *io_u; | |
433afcb4 JA |
1046 | int i; |
1047 | ||
1048 | log_err("fio: io_u timeout\n"); | |
1049 | ||
1050 | /* | |
1051 | * TLS would be nice... | |
1052 | */ | |
1053 | td = NULL; | |
1054 | for_each_td(__td, i) { | |
1055 | if (__td->pid == pid) { | |
1056 | td = __td; | |
1057 | break; | |
1058 | } | |
1059 | } | |
1060 | ||
1061 | if (!td) { | |
1062 | log_err("fio: io_u timeout, can't find job\n"); | |
1063 | exit(1); | |
1064 | } | |
1065 | ||
1066 | if (!td->cur_depth) { | |
1067 | log_err("fio: timeout without pending work?\n"); | |
1068 | return; | |
1069 | } | |
1070 | ||
15506d09 | 1071 | log_err("fio: io_u timeout: job=%s, pid=%d\n", td->o.name, td->pid); |
5945b9b4 JA |
1072 | |
1073 | list_for_each(entry, &td->io_u_busylist) { | |
1074 | io_u = list_entry(entry, struct io_u, list); | |
1075 | ||
1076 | io_u_dump(io_u); | |
1077 | } | |
1078 | ||
1079 | td_verror(td, ETIMEDOUT, "io_u timeout"); | |
433afcb4 JA |
1080 | exit(1); |
1081 | } | |
55bc9728 | 1082 | #endif |
433afcb4 JA |
1083 | |
1084 | void io_u_init_timeout(void) | |
1085 | { | |
55bc9728 | 1086 | #ifdef FIO_USE_TIMEOUT |
433afcb4 | 1087 | signal(SIGALRM, io_u_timeout_handler); |
55bc9728 | 1088 | #endif |
433afcb4 | 1089 | } |