xref: /openbmc/qemu/block/rbd.c (revision 2993683b)
1 /*
2  * QEMU Block driver for RADOS (Ceph)
3  *
4  * Copyright (C) 2010-2011 Christian Brunner <chb@muc.de>,
5  *                         Josh Durgin <josh.durgin@dreamhost.com>
6  *
7  * This work is licensed under the terms of the GNU GPL, version 2.  See
8  * the COPYING file in the top-level directory.
9  *
10  * Contributions after 2012-01-13 are licensed under the terms of the
11  * GNU GPL, version 2 or (at your option) any later version.
12  */
13 
14 #include <inttypes.h>
15 
16 #include "qemu-common.h"
17 #include "qemu/error-report.h"
18 #include "block/block_int.h"
19 
20 #include <rbd/librbd.h>
21 
22 /*
23  * When specifying the image filename use:
24  *
25  * rbd:poolname/devicename[@snapshotname][:option1=value1[:option2=value2...]]
26  *
27  * poolname must be the name of an existing rados pool.
28  *
29  * devicename is the name of the rbd image.
30  *
31  * Each option given is used to configure rados, and may be any valid
32  * Ceph option, "id", or "conf".
33  *
34  * The "id" option indicates what user we should authenticate as to
35  * the Ceph cluster.  If it is excluded we will use the Ceph default
36  * (normally 'admin').
37  *
38  * The "conf" option specifies a Ceph configuration file to read.  If
39  * it is not specified, we will read from the default Ceph locations
40  * (e.g., /etc/ceph/ceph.conf).  To avoid reading _any_ configuration
41  * file, specify conf=/dev/null.
42  *
43  * Configuration values containing :, @, or = can be escaped with a
44  * leading "\".
45  */
46 
47 /* rbd_aio_discard added in 0.1.2 */
48 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 2)
49 #define LIBRBD_SUPPORTS_DISCARD
50 #else
51 #undef LIBRBD_SUPPORTS_DISCARD
52 #endif
53 
54 #define OBJ_MAX_SIZE (1UL << OBJ_DEFAULT_OBJ_ORDER)
55 
56 #define RBD_MAX_CONF_NAME_SIZE 128
57 #define RBD_MAX_CONF_VAL_SIZE 512
58 #define RBD_MAX_CONF_SIZE 1024
59 #define RBD_MAX_POOL_NAME_SIZE 128
60 #define RBD_MAX_SNAP_NAME_SIZE 128
61 #define RBD_MAX_SNAPS 100
62 
63 typedef enum {
64     RBD_AIO_READ,
65     RBD_AIO_WRITE,
66     RBD_AIO_DISCARD,
67     RBD_AIO_FLUSH
68 } RBDAIOCmd;
69 
70 typedef struct RBDAIOCB {
71     BlockDriverAIOCB common;
72     QEMUBH *bh;
73     int64_t ret;
74     QEMUIOVector *qiov;
75     char *bounce;
76     RBDAIOCmd cmd;
77     int64_t sector_num;
78     int error;
79     struct BDRVRBDState *s;
80     int cancelled;
81     int status;
82 } RBDAIOCB;
83 
84 typedef struct RADOSCB {
85     int rcbid;
86     RBDAIOCB *acb;
87     struct BDRVRBDState *s;
88     int done;
89     int64_t size;
90     char *buf;
91     int64_t ret;
92 } RADOSCB;
93 
94 #define RBD_FD_READ 0
95 #define RBD_FD_WRITE 1
96 
97 typedef struct BDRVRBDState {
98     int fds[2];
99     rados_t cluster;
100     rados_ioctx_t io_ctx;
101     rbd_image_t image;
102     char name[RBD_MAX_IMAGE_NAME_SIZE];
103     int qemu_aio_count;
104     char *snap;
105     int event_reader_pos;
106     RADOSCB *event_rcb;
107 } BDRVRBDState;
108 
109 static void rbd_aio_bh_cb(void *opaque);
110 
111 static int qemu_rbd_next_tok(char *dst, int dst_len,
112                              char *src, char delim,
113                              const char *name,
114                              char **p)
115 {
116     int l;
117     char *end;
118 
119     *p = NULL;
120 
121     if (delim != '\0') {
122         for (end = src; *end; ++end) {
123             if (*end == delim) {
124                 break;
125             }
126             if (*end == '\\' && end[1] != '\0') {
127                 end++;
128             }
129         }
130         if (*end == delim) {
131             *p = end + 1;
132             *end = '\0';
133         }
134     }
135     l = strlen(src);
136     if (l >= dst_len) {
137         error_report("%s too long", name);
138         return -EINVAL;
139     } else if (l == 0) {
140         error_report("%s too short", name);
141         return -EINVAL;
142     }
143 
144     pstrcpy(dst, dst_len, src);
145 
146     return 0;
147 }
148 
149 static void qemu_rbd_unescape(char *src)
150 {
151     char *p;
152 
153     for (p = src; *src; ++src, ++p) {
154         if (*src == '\\' && src[1] != '\0') {
155             src++;
156         }
157         *p = *src;
158     }
159     *p = '\0';
160 }
161 
162 static int qemu_rbd_parsename(const char *filename,
163                               char *pool, int pool_len,
164                               char *snap, int snap_len,
165                               char *name, int name_len,
166                               char *conf, int conf_len)
167 {
168     const char *start;
169     char *p, *buf;
170     int ret;
171 
172     if (!strstart(filename, "rbd:", &start)) {
173         return -EINVAL;
174     }
175 
176     buf = g_strdup(start);
177     p = buf;
178     *snap = '\0';
179     *conf = '\0';
180 
181     ret = qemu_rbd_next_tok(pool, pool_len, p, '/', "pool name", &p);
182     if (ret < 0 || !p) {
183         ret = -EINVAL;
184         goto done;
185     }
186     qemu_rbd_unescape(pool);
187 
188     if (strchr(p, '@')) {
189         ret = qemu_rbd_next_tok(name, name_len, p, '@', "object name", &p);
190         if (ret < 0) {
191             goto done;
192         }
193         ret = qemu_rbd_next_tok(snap, snap_len, p, ':', "snap name", &p);
194         qemu_rbd_unescape(snap);
195     } else {
196         ret = qemu_rbd_next_tok(name, name_len, p, ':', "object name", &p);
197     }
198     qemu_rbd_unescape(name);
199     if (ret < 0 || !p) {
200         goto done;
201     }
202 
203     ret = qemu_rbd_next_tok(conf, conf_len, p, '\0', "configuration", &p);
204 
205 done:
206     g_free(buf);
207     return ret;
208 }
209 
210 static char *qemu_rbd_parse_clientname(const char *conf, char *clientname)
211 {
212     const char *p = conf;
213 
214     while (*p) {
215         int len;
216         const char *end = strchr(p, ':');
217 
218         if (end) {
219             len = end - p;
220         } else {
221             len = strlen(p);
222         }
223 
224         if (strncmp(p, "id=", 3) == 0) {
225             len -= 3;
226             strncpy(clientname, p + 3, len);
227             clientname[len] = '\0';
228             return clientname;
229         }
230         if (end == NULL) {
231             break;
232         }
233         p = end + 1;
234     }
235     return NULL;
236 }
237 
238 static int qemu_rbd_set_conf(rados_t cluster, const char *conf)
239 {
240     char *p, *buf;
241     char name[RBD_MAX_CONF_NAME_SIZE];
242     char value[RBD_MAX_CONF_VAL_SIZE];
243     int ret = 0;
244 
245     buf = g_strdup(conf);
246     p = buf;
247 
248     while (p) {
249         ret = qemu_rbd_next_tok(name, sizeof(name), p,
250                                 '=', "conf option name", &p);
251         if (ret < 0) {
252             break;
253         }
254         qemu_rbd_unescape(name);
255 
256         if (!p) {
257             error_report("conf option %s has no value", name);
258             ret = -EINVAL;
259             break;
260         }
261 
262         ret = qemu_rbd_next_tok(value, sizeof(value), p,
263                                 ':', "conf option value", &p);
264         if (ret < 0) {
265             break;
266         }
267         qemu_rbd_unescape(value);
268 
269         if (strcmp(name, "conf") == 0) {
270             ret = rados_conf_read_file(cluster, value);
271             if (ret < 0) {
272                 error_report("error reading conf file %s", value);
273                 break;
274             }
275         } else if (strcmp(name, "id") == 0) {
276             /* ignore, this is parsed by qemu_rbd_parse_clientname() */
277         } else {
278             ret = rados_conf_set(cluster, name, value);
279             if (ret < 0) {
280                 error_report("invalid conf option %s", name);
281                 ret = -EINVAL;
282                 break;
283             }
284         }
285     }
286 
287     g_free(buf);
288     return ret;
289 }
290 
291 static int qemu_rbd_create(const char *filename, QEMUOptionParameter *options)
292 {
293     int64_t bytes = 0;
294     int64_t objsize;
295     int obj_order = 0;
296     char pool[RBD_MAX_POOL_NAME_SIZE];
297     char name[RBD_MAX_IMAGE_NAME_SIZE];
298     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
299     char conf[RBD_MAX_CONF_SIZE];
300     char clientname_buf[RBD_MAX_CONF_SIZE];
301     char *clientname;
302     rados_t cluster;
303     rados_ioctx_t io_ctx;
304     int ret;
305 
306     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
307                            snap_buf, sizeof(snap_buf),
308                            name, sizeof(name),
309                            conf, sizeof(conf)) < 0) {
310         return -EINVAL;
311     }
312 
313     /* Read out options */
314     while (options && options->name) {
315         if (!strcmp(options->name, BLOCK_OPT_SIZE)) {
316             bytes = options->value.n;
317         } else if (!strcmp(options->name, BLOCK_OPT_CLUSTER_SIZE)) {
318             if (options->value.n) {
319                 objsize = options->value.n;
320                 if ((objsize - 1) & objsize) {    /* not a power of 2? */
321                     error_report("obj size needs to be power of 2");
322                     return -EINVAL;
323                 }
324                 if (objsize < 4096) {
325                     error_report("obj size too small");
326                     return -EINVAL;
327                 }
328                 obj_order = ffs(objsize) - 1;
329             }
330         }
331         options++;
332     }
333 
334     clientname = qemu_rbd_parse_clientname(conf, clientname_buf);
335     if (rados_create(&cluster, clientname) < 0) {
336         error_report("error initializing");
337         return -EIO;
338     }
339 
340     if (strstr(conf, "conf=") == NULL) {
341         /* try default location, but ignore failure */
342         rados_conf_read_file(cluster, NULL);
343     }
344 
345     if (conf[0] != '\0' &&
346         qemu_rbd_set_conf(cluster, conf) < 0) {
347         error_report("error setting config options");
348         rados_shutdown(cluster);
349         return -EIO;
350     }
351 
352     if (rados_connect(cluster) < 0) {
353         error_report("error connecting");
354         rados_shutdown(cluster);
355         return -EIO;
356     }
357 
358     if (rados_ioctx_create(cluster, pool, &io_ctx) < 0) {
359         error_report("error opening pool %s", pool);
360         rados_shutdown(cluster);
361         return -EIO;
362     }
363 
364     ret = rbd_create(io_ctx, name, bytes, &obj_order);
365     rados_ioctx_destroy(io_ctx);
366     rados_shutdown(cluster);
367 
368     return ret;
369 }
370 
371 /*
372  * This aio completion is being called from qemu_rbd_aio_event_reader()
373  * and runs in qemu context. It schedules a bh, but just in case the aio
374  * was not cancelled before.
375  */
376 static void qemu_rbd_complete_aio(RADOSCB *rcb)
377 {
378     RBDAIOCB *acb = rcb->acb;
379     int64_t r;
380 
381     r = rcb->ret;
382 
383     if (acb->cmd != RBD_AIO_READ) {
384         if (r < 0) {
385             acb->ret = r;
386             acb->error = 1;
387         } else if (!acb->error) {
388             acb->ret = rcb->size;
389         }
390     } else {
391         if (r < 0) {
392             memset(rcb->buf, 0, rcb->size);
393             acb->ret = r;
394             acb->error = 1;
395         } else if (r < rcb->size) {
396             memset(rcb->buf + r, 0, rcb->size - r);
397             if (!acb->error) {
398                 acb->ret = rcb->size;
399             }
400         } else if (!acb->error) {
401             acb->ret = r;
402         }
403     }
404     /* Note that acb->bh can be NULL in case where the aio was cancelled */
405     acb->bh = qemu_bh_new(rbd_aio_bh_cb, acb);
406     qemu_bh_schedule(acb->bh);
407     g_free(rcb);
408 }
409 
410 /*
411  * aio fd read handler. It runs in the qemu context and calls the
412  * completion handling of completed rados aio operations.
413  */
414 static void qemu_rbd_aio_event_reader(void *opaque)
415 {
416     BDRVRBDState *s = opaque;
417 
418     ssize_t ret;
419 
420     do {
421         char *p = (char *)&s->event_rcb;
422 
423         /* now read the rcb pointer that was sent from a non qemu thread */
424         ret = read(s->fds[RBD_FD_READ], p + s->event_reader_pos,
425                    sizeof(s->event_rcb) - s->event_reader_pos);
426         if (ret > 0) {
427             s->event_reader_pos += ret;
428             if (s->event_reader_pos == sizeof(s->event_rcb)) {
429                 s->event_reader_pos = 0;
430                 qemu_rbd_complete_aio(s->event_rcb);
431                 s->qemu_aio_count--;
432             }
433         }
434     } while (ret < 0 && errno == EINTR);
435 }
436 
437 static int qemu_rbd_aio_flush_cb(void *opaque)
438 {
439     BDRVRBDState *s = opaque;
440 
441     return (s->qemu_aio_count > 0);
442 }
443 
444 /* TODO Convert to fine grained options */
445 static QemuOptsList runtime_opts = {
446     .name = "rbd",
447     .head = QTAILQ_HEAD_INITIALIZER(runtime_opts.head),
448     .desc = {
449         {
450             .name = "filename",
451             .type = QEMU_OPT_STRING,
452             .help = "Specification of the rbd image",
453         },
454         { /* end of list */ }
455     },
456 };
457 
458 static int qemu_rbd_open(BlockDriverState *bs, QDict *options, int flags)
459 {
460     BDRVRBDState *s = bs->opaque;
461     char pool[RBD_MAX_POOL_NAME_SIZE];
462     char snap_buf[RBD_MAX_SNAP_NAME_SIZE];
463     char conf[RBD_MAX_CONF_SIZE];
464     char clientname_buf[RBD_MAX_CONF_SIZE];
465     char *clientname;
466     QemuOpts *opts;
467     Error *local_err = NULL;
468     const char *filename;
469     int r;
470 
471     opts = qemu_opts_create_nofail(&runtime_opts);
472     qemu_opts_absorb_qdict(opts, options, &local_err);
473     if (error_is_set(&local_err)) {
474         qerror_report_err(local_err);
475         error_free(local_err);
476         qemu_opts_del(opts);
477         return -EINVAL;
478     }
479 
480     filename = qemu_opt_get(opts, "filename");
481 
482     if (qemu_rbd_parsename(filename, pool, sizeof(pool),
483                            snap_buf, sizeof(snap_buf),
484                            s->name, sizeof(s->name),
485                            conf, sizeof(conf)) < 0) {
486         r = -EINVAL;
487         goto failed_opts;
488     }
489 
490     clientname = qemu_rbd_parse_clientname(conf, clientname_buf);
491     r = rados_create(&s->cluster, clientname);
492     if (r < 0) {
493         error_report("error initializing");
494         goto failed_opts;
495     }
496 
497     s->snap = NULL;
498     if (snap_buf[0] != '\0') {
499         s->snap = g_strdup(snap_buf);
500     }
501 
502     /*
503      * Fallback to more conservative semantics if setting cache
504      * options fails. Ignore errors from setting rbd_cache because the
505      * only possible error is that the option does not exist, and
506      * librbd defaults to no caching. If write through caching cannot
507      * be set up, fall back to no caching.
508      */
509     if (flags & BDRV_O_NOCACHE) {
510         rados_conf_set(s->cluster, "rbd_cache", "false");
511     } else {
512         rados_conf_set(s->cluster, "rbd_cache", "true");
513     }
514 
515     if (strstr(conf, "conf=") == NULL) {
516         /* try default location, but ignore failure */
517         rados_conf_read_file(s->cluster, NULL);
518     }
519 
520     if (conf[0] != '\0') {
521         r = qemu_rbd_set_conf(s->cluster, conf);
522         if (r < 0) {
523             error_report("error setting config options");
524             goto failed_shutdown;
525         }
526     }
527 
528     r = rados_connect(s->cluster);
529     if (r < 0) {
530         error_report("error connecting");
531         goto failed_shutdown;
532     }
533 
534     r = rados_ioctx_create(s->cluster, pool, &s->io_ctx);
535     if (r < 0) {
536         error_report("error opening pool %s", pool);
537         goto failed_shutdown;
538     }
539 
540     r = rbd_open(s->io_ctx, s->name, &s->image, s->snap);
541     if (r < 0) {
542         error_report("error reading header from %s", s->name);
543         goto failed_open;
544     }
545 
546     bs->read_only = (s->snap != NULL);
547 
548     s->event_reader_pos = 0;
549     r = qemu_pipe(s->fds);
550     if (r < 0) {
551         error_report("error opening eventfd");
552         goto failed;
553     }
554     fcntl(s->fds[0], F_SETFL, O_NONBLOCK);
555     fcntl(s->fds[1], F_SETFL, O_NONBLOCK);
556     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], qemu_rbd_aio_event_reader,
557                             NULL, qemu_rbd_aio_flush_cb, s);
558 
559 
560     qemu_opts_del(opts);
561     return 0;
562 
563 failed:
564     rbd_close(s->image);
565 failed_open:
566     rados_ioctx_destroy(s->io_ctx);
567 failed_shutdown:
568     rados_shutdown(s->cluster);
569     g_free(s->snap);
570 failed_opts:
571     qemu_opts_del(opts);
572     return r;
573 }
574 
575 static void qemu_rbd_close(BlockDriverState *bs)
576 {
577     BDRVRBDState *s = bs->opaque;
578 
579     close(s->fds[0]);
580     close(s->fds[1]);
581     qemu_aio_set_fd_handler(s->fds[RBD_FD_READ], NULL, NULL, NULL, NULL);
582 
583     rbd_close(s->image);
584     rados_ioctx_destroy(s->io_ctx);
585     g_free(s->snap);
586     rados_shutdown(s->cluster);
587 }
588 
589 /*
590  * Cancel aio. Since we don't reference acb in a non qemu threads,
591  * it is safe to access it here.
592  */
593 static void qemu_rbd_aio_cancel(BlockDriverAIOCB *blockacb)
594 {
595     RBDAIOCB *acb = (RBDAIOCB *) blockacb;
596     acb->cancelled = 1;
597 
598     while (acb->status == -EINPROGRESS) {
599         qemu_aio_wait();
600     }
601 
602     qemu_aio_release(acb);
603 }
604 
605 static const AIOCBInfo rbd_aiocb_info = {
606     .aiocb_size = sizeof(RBDAIOCB),
607     .cancel = qemu_rbd_aio_cancel,
608 };
609 
610 static int qemu_rbd_send_pipe(BDRVRBDState *s, RADOSCB *rcb)
611 {
612     int ret = 0;
613     while (1) {
614         fd_set wfd;
615         int fd = s->fds[RBD_FD_WRITE];
616 
617         /* send the op pointer to the qemu thread that is responsible
618            for the aio/op completion. Must do it in a qemu thread context */
619         ret = write(fd, (void *)&rcb, sizeof(rcb));
620         if (ret >= 0) {
621             break;
622         }
623         if (errno == EINTR) {
624             continue;
625         }
626         if (errno != EAGAIN) {
627             break;
628         }
629 
630         FD_ZERO(&wfd);
631         FD_SET(fd, &wfd);
632         do {
633             ret = select(fd + 1, NULL, &wfd, NULL, NULL);
634         } while (ret < 0 && errno == EINTR);
635     }
636 
637     return ret;
638 }
639 
640 /*
641  * This is the callback function for rbd_aio_read and _write
642  *
643  * Note: this function is being called from a non qemu thread so
644  * we need to be careful about what we do here. Generally we only
645  * write to the block notification pipe, and do the rest of the
646  * io completion handling from qemu_rbd_aio_event_reader() which
647  * runs in a qemu context.
648  */
649 static void rbd_finish_aiocb(rbd_completion_t c, RADOSCB *rcb)
650 {
651     int ret;
652     rcb->ret = rbd_aio_get_return_value(c);
653     rbd_aio_release(c);
654     ret = qemu_rbd_send_pipe(rcb->s, rcb);
655     if (ret < 0) {
656         error_report("failed writing to acb->s->fds");
657         g_free(rcb);
658     }
659 }
660 
661 /* Callback when all queued rbd_aio requests are complete */
662 
663 static void rbd_aio_bh_cb(void *opaque)
664 {
665     RBDAIOCB *acb = opaque;
666 
667     if (acb->cmd == RBD_AIO_READ) {
668         qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
669     }
670     qemu_vfree(acb->bounce);
671     acb->common.cb(acb->common.opaque, (acb->ret > 0 ? 0 : acb->ret));
672     qemu_bh_delete(acb->bh);
673     acb->bh = NULL;
674     acb->status = 0;
675 
676     if (!acb->cancelled) {
677         qemu_aio_release(acb);
678     }
679 }
680 
681 static int rbd_aio_discard_wrapper(rbd_image_t image,
682                                    uint64_t off,
683                                    uint64_t len,
684                                    rbd_completion_t comp)
685 {
686 #ifdef LIBRBD_SUPPORTS_DISCARD
687     return rbd_aio_discard(image, off, len, comp);
688 #else
689     return -ENOTSUP;
690 #endif
691 }
692 
693 static int rbd_aio_flush_wrapper(rbd_image_t image,
694                                  rbd_completion_t comp)
695 {
696 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
697     return rbd_aio_flush(image, comp);
698 #else
699     return -ENOTSUP;
700 #endif
701 }
702 
703 static BlockDriverAIOCB *rbd_start_aio(BlockDriverState *bs,
704                                        int64_t sector_num,
705                                        QEMUIOVector *qiov,
706                                        int nb_sectors,
707                                        BlockDriverCompletionFunc *cb,
708                                        void *opaque,
709                                        RBDAIOCmd cmd)
710 {
711     RBDAIOCB *acb;
712     RADOSCB *rcb;
713     rbd_completion_t c;
714     int64_t off, size;
715     char *buf;
716     int r;
717 
718     BDRVRBDState *s = bs->opaque;
719 
720     acb = qemu_aio_get(&rbd_aiocb_info, bs, cb, opaque);
721     acb->cmd = cmd;
722     acb->qiov = qiov;
723     if (cmd == RBD_AIO_DISCARD || cmd == RBD_AIO_FLUSH) {
724         acb->bounce = NULL;
725     } else {
726         acb->bounce = qemu_blockalign(bs, qiov->size);
727     }
728     acb->ret = 0;
729     acb->error = 0;
730     acb->s = s;
731     acb->cancelled = 0;
732     acb->bh = NULL;
733     acb->status = -EINPROGRESS;
734 
735     if (cmd == RBD_AIO_WRITE) {
736         qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
737     }
738 
739     buf = acb->bounce;
740 
741     off = sector_num * BDRV_SECTOR_SIZE;
742     size = nb_sectors * BDRV_SECTOR_SIZE;
743 
744     s->qemu_aio_count++; /* All the RADOSCB */
745 
746     rcb = g_malloc(sizeof(RADOSCB));
747     rcb->done = 0;
748     rcb->acb = acb;
749     rcb->buf = buf;
750     rcb->s = acb->s;
751     rcb->size = size;
752     r = rbd_aio_create_completion(rcb, (rbd_callback_t) rbd_finish_aiocb, &c);
753     if (r < 0) {
754         goto failed;
755     }
756 
757     switch (cmd) {
758     case RBD_AIO_WRITE:
759         r = rbd_aio_write(s->image, off, size, buf, c);
760         break;
761     case RBD_AIO_READ:
762         r = rbd_aio_read(s->image, off, size, buf, c);
763         break;
764     case RBD_AIO_DISCARD:
765         r = rbd_aio_discard_wrapper(s->image, off, size, c);
766         break;
767     case RBD_AIO_FLUSH:
768         r = rbd_aio_flush_wrapper(s->image, c);
769         break;
770     default:
771         r = -EINVAL;
772     }
773 
774     if (r < 0) {
775         goto failed;
776     }
777 
778     return &acb->common;
779 
780 failed:
781     g_free(rcb);
782     s->qemu_aio_count--;
783     qemu_aio_release(acb);
784     return NULL;
785 }
786 
787 static BlockDriverAIOCB *qemu_rbd_aio_readv(BlockDriverState *bs,
788                                             int64_t sector_num,
789                                             QEMUIOVector *qiov,
790                                             int nb_sectors,
791                                             BlockDriverCompletionFunc *cb,
792                                             void *opaque)
793 {
794     return rbd_start_aio(bs, sector_num, qiov, nb_sectors, cb, opaque,
795                          RBD_AIO_READ);
796 }
797 
798 static BlockDriverAIOCB *qemu_rbd_aio_writev(BlockDriverState *bs,
799                                              int64_t sector_num,
800                                              QEMUIOVector *qiov,
801                                              int nb_sectors,
802                                              BlockDriverCompletionFunc *cb,
803                                              void *opaque)
804 {
805     return rbd_start_aio(bs, sector_num, qiov, nb_sectors, cb, opaque,
806                          RBD_AIO_WRITE);
807 }
808 
809 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
810 static BlockDriverAIOCB *qemu_rbd_aio_flush(BlockDriverState *bs,
811                                             BlockDriverCompletionFunc *cb,
812                                             void *opaque)
813 {
814     return rbd_start_aio(bs, 0, NULL, 0, cb, opaque, RBD_AIO_FLUSH);
815 }
816 
817 #else
818 
819 static int qemu_rbd_co_flush(BlockDriverState *bs)
820 {
821 #if LIBRBD_VERSION_CODE >= LIBRBD_VERSION(0, 1, 1)
822     /* rbd_flush added in 0.1.1 */
823     BDRVRBDState *s = bs->opaque;
824     return rbd_flush(s->image);
825 #else
826     return 0;
827 #endif
828 }
829 #endif
830 
831 static int qemu_rbd_getinfo(BlockDriverState *bs, BlockDriverInfo *bdi)
832 {
833     BDRVRBDState *s = bs->opaque;
834     rbd_image_info_t info;
835     int r;
836 
837     r = rbd_stat(s->image, &info, sizeof(info));
838     if (r < 0) {
839         return r;
840     }
841 
842     bdi->cluster_size = info.obj_size;
843     return 0;
844 }
845 
846 static int64_t qemu_rbd_getlength(BlockDriverState *bs)
847 {
848     BDRVRBDState *s = bs->opaque;
849     rbd_image_info_t info;
850     int r;
851 
852     r = rbd_stat(s->image, &info, sizeof(info));
853     if (r < 0) {
854         return r;
855     }
856 
857     return info.size;
858 }
859 
860 static int qemu_rbd_truncate(BlockDriverState *bs, int64_t offset)
861 {
862     BDRVRBDState *s = bs->opaque;
863     int r;
864 
865     r = rbd_resize(s->image, offset);
866     if (r < 0) {
867         return r;
868     }
869 
870     return 0;
871 }
872 
873 static int qemu_rbd_snap_create(BlockDriverState *bs,
874                                 QEMUSnapshotInfo *sn_info)
875 {
876     BDRVRBDState *s = bs->opaque;
877     int r;
878 
879     if (sn_info->name[0] == '\0') {
880         return -EINVAL; /* we need a name for rbd snapshots */
881     }
882 
883     /*
884      * rbd snapshots are using the name as the user controlled unique identifier
885      * we can't use the rbd snapid for that purpose, as it can't be set
886      */
887     if (sn_info->id_str[0] != '\0' &&
888         strcmp(sn_info->id_str, sn_info->name) != 0) {
889         return -EINVAL;
890     }
891 
892     if (strlen(sn_info->name) >= sizeof(sn_info->id_str)) {
893         return -ERANGE;
894     }
895 
896     r = rbd_snap_create(s->image, sn_info->name);
897     if (r < 0) {
898         error_report("failed to create snap: %s", strerror(-r));
899         return r;
900     }
901 
902     return 0;
903 }
904 
905 static int qemu_rbd_snap_remove(BlockDriverState *bs,
906                                 const char *snapshot_name)
907 {
908     BDRVRBDState *s = bs->opaque;
909     int r;
910 
911     r = rbd_snap_remove(s->image, snapshot_name);
912     return r;
913 }
914 
915 static int qemu_rbd_snap_rollback(BlockDriverState *bs,
916                                   const char *snapshot_name)
917 {
918     BDRVRBDState *s = bs->opaque;
919     int r;
920 
921     r = rbd_snap_rollback(s->image, snapshot_name);
922     return r;
923 }
924 
925 static int qemu_rbd_snap_list(BlockDriverState *bs,
926                               QEMUSnapshotInfo **psn_tab)
927 {
928     BDRVRBDState *s = bs->opaque;
929     QEMUSnapshotInfo *sn_info, *sn_tab = NULL;
930     int i, snap_count;
931     rbd_snap_info_t *snaps;
932     int max_snaps = RBD_MAX_SNAPS;
933 
934     do {
935         snaps = g_malloc(sizeof(*snaps) * max_snaps);
936         snap_count = rbd_snap_list(s->image, snaps, &max_snaps);
937         if (snap_count < 0) {
938             g_free(snaps);
939         }
940     } while (snap_count == -ERANGE);
941 
942     if (snap_count <= 0) {
943         goto done;
944     }
945 
946     sn_tab = g_malloc0(snap_count * sizeof(QEMUSnapshotInfo));
947 
948     for (i = 0; i < snap_count; i++) {
949         const char *snap_name = snaps[i].name;
950 
951         sn_info = sn_tab + i;
952         pstrcpy(sn_info->id_str, sizeof(sn_info->id_str), snap_name);
953         pstrcpy(sn_info->name, sizeof(sn_info->name), snap_name);
954 
955         sn_info->vm_state_size = snaps[i].size;
956         sn_info->date_sec = 0;
957         sn_info->date_nsec = 0;
958         sn_info->vm_clock_nsec = 0;
959     }
960     rbd_snap_list_end(snaps);
961 
962  done:
963     *psn_tab = sn_tab;
964     return snap_count;
965 }
966 
967 #ifdef LIBRBD_SUPPORTS_DISCARD
968 static BlockDriverAIOCB* qemu_rbd_aio_discard(BlockDriverState *bs,
969                                               int64_t sector_num,
970                                               int nb_sectors,
971                                               BlockDriverCompletionFunc *cb,
972                                               void *opaque)
973 {
974     return rbd_start_aio(bs, sector_num, NULL, nb_sectors, cb, opaque,
975                          RBD_AIO_DISCARD);
976 }
977 #endif
978 
979 static QEMUOptionParameter qemu_rbd_create_options[] = {
980     {
981      .name = BLOCK_OPT_SIZE,
982      .type = OPT_SIZE,
983      .help = "Virtual disk size"
984     },
985     {
986      .name = BLOCK_OPT_CLUSTER_SIZE,
987      .type = OPT_SIZE,
988      .help = "RBD object size"
989     },
990     {NULL}
991 };
992 
993 static BlockDriver bdrv_rbd = {
994     .format_name        = "rbd",
995     .instance_size      = sizeof(BDRVRBDState),
996     .bdrv_file_open     = qemu_rbd_open,
997     .bdrv_close         = qemu_rbd_close,
998     .bdrv_create        = qemu_rbd_create,
999     .bdrv_get_info      = qemu_rbd_getinfo,
1000     .create_options     = qemu_rbd_create_options,
1001     .bdrv_getlength     = qemu_rbd_getlength,
1002     .bdrv_truncate      = qemu_rbd_truncate,
1003     .protocol_name      = "rbd",
1004 
1005     .bdrv_aio_readv         = qemu_rbd_aio_readv,
1006     .bdrv_aio_writev        = qemu_rbd_aio_writev,
1007 
1008 #ifdef LIBRBD_SUPPORTS_AIO_FLUSH
1009     .bdrv_aio_flush         = qemu_rbd_aio_flush,
1010 #else
1011     .bdrv_co_flush_to_disk  = qemu_rbd_co_flush,
1012 #endif
1013 
1014 #ifdef LIBRBD_SUPPORTS_DISCARD
1015     .bdrv_aio_discard       = qemu_rbd_aio_discard,
1016 #endif
1017 
1018     .bdrv_snapshot_create   = qemu_rbd_snap_create,
1019     .bdrv_snapshot_delete   = qemu_rbd_snap_remove,
1020     .bdrv_snapshot_list     = qemu_rbd_snap_list,
1021     .bdrv_snapshot_goto     = qemu_rbd_snap_rollback,
1022 };
1023 
1024 static void bdrv_rbd_init(void)
1025 {
1026     bdrv_register(&bdrv_rbd);
1027 }
1028 
1029 block_init(bdrv_rbd_init);
1030