blob: a67ffe3113c3a81b385b52f769fda3fab57e4c87 [file] [log] [blame]
David 'Digit' Turnercb42a1b2010-12-23 02:54:08 +01001/*
2 * QEMU posix-aio emulation
3 *
4 * Copyright IBM, Corp. 2008
5 *
6 * Authors:
7 * Anthony Liguori <aliguori@us.ibm.com>
8 *
9 * This work is licensed under the terms of the GNU GPL, version 2. See
10 * the COPYING file in the top-level directory.
11 *
12 */
13
14#include <sys/ioctl.h>
15#include <sys/types.h>
16#include <pthread.h>
17#include <unistd.h>
18#include <errno.h>
19#include <time.h>
20#include <signal.h>
21#include <string.h>
22#include <stdlib.h>
23#include <stdio.h>
24
25#include "qemu-queue.h"
26#include "osdep.h"
27#include "qemu-common.h"
28#include "block_int.h"
29
30#include "block/raw-posix-aio.h"
31
32
33struct qemu_paiocb {
34 BlockDriverAIOCB common;
35 int aio_fildes;
36 union {
37 struct iovec *aio_iov;
38 void *aio_ioctl_buf;
39 };
40 int aio_niov;
41 size_t aio_nbytes;
42#define aio_ioctl_cmd aio_nbytes /* for QEMU_AIO_IOCTL */
43 int ev_signo;
44 off_t aio_offset;
45
46 QTAILQ_ENTRY(qemu_paiocb) node;
47 int aio_type;
48 ssize_t ret;
49 int active;
50 struct qemu_paiocb *next;
51
52 int async_context_id;
53};
54
55typedef struct PosixAioState {
56 int rfd, wfd;
57 struct qemu_paiocb *first_aio;
58} PosixAioState;
59
60
61static pthread_mutex_t lock = PTHREAD_MUTEX_INITIALIZER;
62static pthread_cond_t cond = PTHREAD_COND_INITIALIZER;
63static pthread_t thread_id;
64static pthread_attr_t attr;
65static int max_threads = 64;
66static int cur_threads = 0;
67static int idle_threads = 0;
68static QTAILQ_HEAD(, qemu_paiocb) request_list;
69
70#ifdef CONFIG_PREADV
71static int preadv_present = 1;
72#else
73static int preadv_present = 0;
74#endif
75
76static void die2(int err, const char *what)
77{
78 fprintf(stderr, "%s failed: %s\n", what, strerror(err));
79 abort();
80}
81
82static void die(const char *what)
83{
84 die2(errno, what);
85}
86
87static void mutex_lock(pthread_mutex_t *mutex)
88{
89 int ret = pthread_mutex_lock(mutex);
90 if (ret) die2(ret, "pthread_mutex_lock");
91}
92
93static void mutex_unlock(pthread_mutex_t *mutex)
94{
95 int ret = pthread_mutex_unlock(mutex);
96 if (ret) die2(ret, "pthread_mutex_unlock");
97}
98
99static int cond_timedwait(pthread_cond_t *cond, pthread_mutex_t *mutex,
100 struct timespec *ts)
101{
102 int ret = pthread_cond_timedwait(cond, mutex, ts);
103 if (ret && ret != ETIMEDOUT) die2(ret, "pthread_cond_timedwait");
104 return ret;
105}
106
107static void cond_signal(pthread_cond_t *cond)
108{
109 int ret = pthread_cond_signal(cond);
110 if (ret) die2(ret, "pthread_cond_signal");
111}
112
113static void thread_create(pthread_t *thread, pthread_attr_t *attr,
114 void *(*start_routine)(void*), void *arg)
115{
116 int ret = pthread_create(thread, attr, start_routine, arg);
117 if (ret) die2(ret, "pthread_create");
118}
119
120static ssize_t handle_aiocb_ioctl(struct qemu_paiocb *aiocb)
121{
122 int ret;
123
124 ret = ioctl(aiocb->aio_fildes, aiocb->aio_ioctl_cmd, aiocb->aio_ioctl_buf);
125 if (ret == -1)
126 return -errno;
127
128 /*
129 * This looks weird, but the aio code only consideres a request
130 * successfull if it has written the number full number of bytes.
131 *
132 * Now we overload aio_nbytes as aio_ioctl_cmd for the ioctl command,
133 * so in fact we return the ioctl command here to make posix_aio_read()
134 * happy..
135 */
136 return aiocb->aio_nbytes;
137}
138
139static ssize_t handle_aiocb_flush(struct qemu_paiocb *aiocb)
140{
141 int ret;
142
143 ret = qemu_fdatasync(aiocb->aio_fildes);
144 if (ret == -1)
145 return -errno;
146 return 0;
147}
148
149#ifdef CONFIG_PREADV
150
151static ssize_t
152qemu_preadv(int fd, const struct iovec *iov, int nr_iov, off_t offset)
153{
154 return preadv(fd, iov, nr_iov, offset);
155}
156
157static ssize_t
158qemu_pwritev(int fd, const struct iovec *iov, int nr_iov, off_t offset)
159{
160 return pwritev(fd, iov, nr_iov, offset);
161}
162
163#else
164
165static ssize_t
166qemu_preadv(int fd, const struct iovec *iov, int nr_iov, off_t offset)
167{
168 return -ENOSYS;
169}
170
171static ssize_t
172qemu_pwritev(int fd, const struct iovec *iov, int nr_iov, off_t offset)
173{
174 return -ENOSYS;
175}
176
177#endif
178
179static ssize_t handle_aiocb_rw_vector(struct qemu_paiocb *aiocb)
180{
181 size_t offset = 0;
182 ssize_t len;
183
184 do {
185 if (aiocb->aio_type & QEMU_AIO_WRITE)
186 len = qemu_pwritev(aiocb->aio_fildes,
187 aiocb->aio_iov,
188 aiocb->aio_niov,
189 aiocb->aio_offset + offset);
190 else
191 len = qemu_preadv(aiocb->aio_fildes,
192 aiocb->aio_iov,
193 aiocb->aio_niov,
194 aiocb->aio_offset + offset);
195 } while (len == -1 && errno == EINTR);
196
197 if (len == -1)
198 return -errno;
199 return len;
200}
201
202static ssize_t handle_aiocb_rw_linear(struct qemu_paiocb *aiocb, char *buf)
203{
204 ssize_t offset = 0;
205 ssize_t len;
206
207 while (offset < aiocb->aio_nbytes) {
208 if (aiocb->aio_type & QEMU_AIO_WRITE)
209 len = pwrite(aiocb->aio_fildes,
210 (const char *)buf + offset,
211 aiocb->aio_nbytes - offset,
212 aiocb->aio_offset + offset);
213 else
214 len = pread(aiocb->aio_fildes,
215 buf + offset,
216 aiocb->aio_nbytes - offset,
217 aiocb->aio_offset + offset);
218
219 if (len == -1 && errno == EINTR)
220 continue;
221 else if (len == -1) {
222 offset = -errno;
223 break;
224 } else if (len == 0)
225 break;
226
227 offset += len;
228 }
229
230 return offset;
231}
232
233static ssize_t handle_aiocb_rw(struct qemu_paiocb *aiocb)
234{
235 ssize_t nbytes;
236 char *buf;
237
238 if (!(aiocb->aio_type & QEMU_AIO_MISALIGNED)) {
239 /*
240 * If there is just a single buffer, and it is properly aligned
241 * we can just use plain pread/pwrite without any problems.
242 */
243 if (aiocb->aio_niov == 1)
244 return handle_aiocb_rw_linear(aiocb, aiocb->aio_iov->iov_base);
245
246 /*
247 * We have more than one iovec, and all are properly aligned.
248 *
249 * Try preadv/pwritev first and fall back to linearizing the
250 * buffer if it's not supported.
251 */
252 if (preadv_present) {
253 nbytes = handle_aiocb_rw_vector(aiocb);
254 if (nbytes == aiocb->aio_nbytes)
255 return nbytes;
256 if (nbytes < 0 && nbytes != -ENOSYS)
257 return nbytes;
258 preadv_present = 0;
259 }
260
261 /*
262 * XXX(hch): short read/write. no easy way to handle the reminder
263 * using these interfaces. For now retry using plain
264 * pread/pwrite?
265 */
266 }
267
268 /*
269 * Ok, we have to do it the hard way, copy all segments into
270 * a single aligned buffer.
271 */
272 buf = qemu_memalign(512, aiocb->aio_nbytes);
273 if (aiocb->aio_type & QEMU_AIO_WRITE) {
274 char *p = buf;
275 int i;
276
277 for (i = 0; i < aiocb->aio_niov; ++i) {
278 memcpy(p, aiocb->aio_iov[i].iov_base, aiocb->aio_iov[i].iov_len);
279 p += aiocb->aio_iov[i].iov_len;
280 }
281 }
282
283 nbytes = handle_aiocb_rw_linear(aiocb, buf);
284 if (!(aiocb->aio_type & QEMU_AIO_WRITE)) {
285 char *p = buf;
286 size_t count = aiocb->aio_nbytes, copy;
287 int i;
288
289 for (i = 0; i < aiocb->aio_niov && count; ++i) {
290 copy = count;
291 if (copy > aiocb->aio_iov[i].iov_len)
292 copy = aiocb->aio_iov[i].iov_len;
293 memcpy(aiocb->aio_iov[i].iov_base, p, copy);
294 p += copy;
295 count -= copy;
296 }
297 }
298 qemu_vfree(buf);
299
300 return nbytes;
301}
302
303static void *aio_thread(void *unused)
304{
305 pid_t pid;
306
307 pid = getpid();
308
309 while (1) {
310 struct qemu_paiocb *aiocb;
311 ssize_t ret = 0;
312 qemu_timeval tv;
313 struct timespec ts;
314
315 qemu_gettimeofday(&tv);
316 ts.tv_sec = tv.tv_sec + 10;
317 ts.tv_nsec = 0;
318
319 mutex_lock(&lock);
320
321 while (QTAILQ_EMPTY(&request_list) &&
322 !(ret == ETIMEDOUT)) {
323 ret = cond_timedwait(&cond, &lock, &ts);
324 }
325
326 if (QTAILQ_EMPTY(&request_list))
327 break;
328
329 aiocb = QTAILQ_FIRST(&request_list);
330 QTAILQ_REMOVE(&request_list, aiocb, node);
331 aiocb->active = 1;
332 idle_threads--;
333 mutex_unlock(&lock);
334
335 switch (aiocb->aio_type & QEMU_AIO_TYPE_MASK) {
336 case QEMU_AIO_READ:
337 case QEMU_AIO_WRITE:
338 ret = handle_aiocb_rw(aiocb);
339 break;
340 case QEMU_AIO_FLUSH:
341 ret = handle_aiocb_flush(aiocb);
342 break;
343 case QEMU_AIO_IOCTL:
344 ret = handle_aiocb_ioctl(aiocb);
345 break;
346 default:
347 fprintf(stderr, "invalid aio request (0x%x)\n", aiocb->aio_type);
348 ret = -EINVAL;
349 break;
350 }
351
352 mutex_lock(&lock);
353 aiocb->ret = ret;
354 idle_threads++;
355 mutex_unlock(&lock);
356
357 if (kill(pid, aiocb->ev_signo)) die("kill failed");
358 }
359
360 idle_threads--;
361 cur_threads--;
362 mutex_unlock(&lock);
363
364 return NULL;
365}
366
367static void spawn_thread(void)
368{
369 sigset_t set, oldset;
370
371 cur_threads++;
372 idle_threads++;
373
374 /* block all signals */
375 if (sigfillset(&set)) die("sigfillset");
376 if (sigprocmask(SIG_SETMASK, &set, &oldset)) die("sigprocmask");
377
378 thread_create(&thread_id, &attr, aio_thread, NULL);
379
380 if (sigprocmask(SIG_SETMASK, &oldset, NULL)) die("sigprocmask restore");
381}
382
383static void qemu_paio_submit(struct qemu_paiocb *aiocb)
384{
385 aiocb->ret = -EINPROGRESS;
386 aiocb->active = 0;
387 mutex_lock(&lock);
388 if (idle_threads == 0 && cur_threads < max_threads)
389 spawn_thread();
390 QTAILQ_INSERT_TAIL(&request_list, aiocb, node);
391 mutex_unlock(&lock);
392 cond_signal(&cond);
393}
394
395static ssize_t qemu_paio_return(struct qemu_paiocb *aiocb)
396{
397 ssize_t ret;
398
399 mutex_lock(&lock);
400 ret = aiocb->ret;
401 mutex_unlock(&lock);
402
403 return ret;
404}
405
406static int qemu_paio_error(struct qemu_paiocb *aiocb)
407{
408 ssize_t ret = qemu_paio_return(aiocb);
409
410 if (ret < 0)
411 ret = -ret;
412 else
413 ret = 0;
414
415 return ret;
416}
417
418static int posix_aio_process_queue(void *opaque)
419{
420 PosixAioState *s = opaque;
421 struct qemu_paiocb *acb, **pacb;
422 int ret;
423 int result = 0;
424 int async_context_id = get_async_context_id();
425
426 for(;;) {
427 pacb = &s->first_aio;
428 for(;;) {
429 acb = *pacb;
430 if (!acb)
431 return result;
432
433 /* we're only interested in requests in the right context */
434 if (acb->async_context_id != async_context_id) {
435 pacb = &acb->next;
436 continue;
437 }
438
439 ret = qemu_paio_error(acb);
440 if (ret == ECANCELED) {
441 /* remove the request */
442 *pacb = acb->next;
443 qemu_aio_release(acb);
444 result = 1;
445 } else if (ret != EINPROGRESS) {
446 /* end of aio */
447 if (ret == 0) {
448 ret = qemu_paio_return(acb);
449 if (ret == acb->aio_nbytes)
450 ret = 0;
451 else
452 ret = -EINVAL;
453 } else {
454 ret = -ret;
455 }
456 /* remove the request */
457 *pacb = acb->next;
458 /* call the callback */
459 acb->common.cb(acb->common.opaque, ret);
460 qemu_aio_release(acb);
461 result = 1;
462 break;
463 } else {
464 pacb = &acb->next;
465 }
466 }
467 }
468
469 return result;
470}
471
472static void posix_aio_read(void *opaque)
473{
474 PosixAioState *s = opaque;
475 ssize_t len;
476
477 /* read all bytes from signal pipe */
478 for (;;) {
479 char bytes[16];
480
481 len = read(s->rfd, bytes, sizeof(bytes));
482 if (len == -1 && errno == EINTR)
483 continue; /* try again */
484 if (len == sizeof(bytes))
485 continue; /* more to read */
486 break;
487 }
488
489 posix_aio_process_queue(s);
490}
491
492static int posix_aio_flush(void *opaque)
493{
494 PosixAioState *s = opaque;
495 return !!s->first_aio;
496}
497
498static PosixAioState *posix_aio_state;
499
500static void aio_signal_handler(int signum)
501{
502 if (posix_aio_state) {
503 char byte = 0;
504 ssize_t ret;
505
506 ret = write(posix_aio_state->wfd, &byte, sizeof(byte));
507 if (ret < 0 && errno != EAGAIN)
508 die("write()");
509 }
510
511 qemu_service_io();
512}
513
514static void paio_remove(struct qemu_paiocb *acb)
515{
516 struct qemu_paiocb **pacb;
517
518 /* remove the callback from the queue */
519 pacb = &posix_aio_state->first_aio;
520 for(;;) {
521 if (*pacb == NULL) {
522 fprintf(stderr, "paio_remove: aio request not found!\n");
523 break;
524 } else if (*pacb == acb) {
525 *pacb = acb->next;
526 qemu_aio_release(acb);
527 break;
528 }
529 pacb = &(*pacb)->next;
530 }
531}
532
533static void paio_cancel(BlockDriverAIOCB *blockacb)
534{
535 struct qemu_paiocb *acb = (struct qemu_paiocb *)blockacb;
536 int active = 0;
537
538 mutex_lock(&lock);
539 if (!acb->active) {
540 QTAILQ_REMOVE(&request_list, acb, node);
541 acb->ret = -ECANCELED;
542 } else if (acb->ret == -EINPROGRESS) {
543 active = 1;
544 }
545 mutex_unlock(&lock);
546
547 if (active) {
548 /* fail safe: if the aio could not be canceled, we wait for
549 it */
550 while (qemu_paio_error(acb) == EINPROGRESS)
551 ;
552 }
553
554 paio_remove(acb);
555}
556
557static AIOPool raw_aio_pool = {
558 .aiocb_size = sizeof(struct qemu_paiocb),
559 .cancel = paio_cancel,
560};
561
562BlockDriverAIOCB *paio_submit(BlockDriverState *bs, int fd,
563 int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
564 BlockDriverCompletionFunc *cb, void *opaque, int type)
565{
566 struct qemu_paiocb *acb;
567
568 acb = qemu_aio_get(&raw_aio_pool, bs, cb, opaque);
569 if (!acb)
570 return NULL;
571 acb->aio_type = type;
572 acb->aio_fildes = fd;
573 acb->ev_signo = SIGUSR2;
574 acb->async_context_id = get_async_context_id();
575
576 if (qiov) {
577 acb->aio_iov = qiov->iov;
578 acb->aio_niov = qiov->niov;
579 }
580 acb->aio_nbytes = nb_sectors * 512;
581 acb->aio_offset = sector_num * 512;
582
583 acb->next = posix_aio_state->first_aio;
584 posix_aio_state->first_aio = acb;
585
586 qemu_paio_submit(acb);
587 return &acb->common;
588}
589
590BlockDriverAIOCB *paio_ioctl(BlockDriverState *bs, int fd,
591 unsigned long int req, void *buf,
592 BlockDriverCompletionFunc *cb, void *opaque)
593{
594 struct qemu_paiocb *acb;
595
596 acb = qemu_aio_get(&raw_aio_pool, bs, cb, opaque);
597 if (!acb)
598 return NULL;
599 acb->aio_type = QEMU_AIO_IOCTL;
600 acb->aio_fildes = fd;
601 acb->ev_signo = SIGUSR2;
602 acb->aio_offset = 0;
603 acb->aio_ioctl_buf = buf;
604 acb->aio_ioctl_cmd = req;
605
606 acb->next = posix_aio_state->first_aio;
607 posix_aio_state->first_aio = acb;
608
609 qemu_paio_submit(acb);
610 return &acb->common;
611}
612
613int paio_init(void)
614{
615 struct sigaction act;
616 PosixAioState *s;
617 int fds[2];
618 int ret;
619
620 if (posix_aio_state)
621 return 0;
622
623 s = qemu_malloc(sizeof(PosixAioState));
624
625 sigfillset(&act.sa_mask);
626 act.sa_flags = 0; /* do not restart syscalls to interrupt select() */
627 act.sa_handler = aio_signal_handler;
628 sigaction(SIGUSR2, &act, NULL);
629
630 s->first_aio = NULL;
631 if (qemu_pipe(fds) == -1) {
632 fprintf(stderr, "failed to create pipe\n");
633 return -1;
634 }
635
636 s->rfd = fds[0];
637 s->wfd = fds[1];
638
639 fcntl(s->rfd, F_SETFL, O_NONBLOCK);
640 fcntl(s->wfd, F_SETFL, O_NONBLOCK);
641
642 qemu_aio_set_fd_handler(s->rfd, posix_aio_read, NULL, posix_aio_flush,
643 posix_aio_process_queue, s);
644
645 ret = pthread_attr_init(&attr);
646 if (ret)
647 die2(ret, "pthread_attr_init");
648
649 ret = pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_DETACHED);
650 if (ret)
651 die2(ret, "pthread_attr_setdetachstate");
652
653 QTAILQ_INIT(&request_list);
654
655 posix_aio_state = s;
656 return 0;
657}