?  search:
summary | log | graphiclog1 | graphiclog2 | commit | commitdiff | tree | refs | edit | fork
blame | history | raw | HEAD
Merge branch 'for-3.11' of git://linux-nfs.org/~bfields/linux
[linux-2.6.git] / drivers / block / xen-blkfront.c
bloba4660bbee8a6279bd4cc1a6def776ddc85bcf169
1 /*
2 * blkfront.c
3 *
4 * XenLinux virtual block device driver.
5 *
6 * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
7 * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
8 * Copyright (c) 2004, Christian Limpach
9 * Copyright (c) 2004, Andrew Warfield
10 * Copyright (c) 2005, Christopher Clark
11 * Copyright (c) 2005, XenSource Ltd
12 *
13 * This program is free software; you can redistribute it and/or
14 * modify it under the terms of the GNU General Public License version 2
15 * as published by the Free Software Foundation; or, when distributed
16 * separately from the Linux kernel or incorporated into other
17 * software packages, subject to the following license:
18 *
19 * Permission is hereby granted, free of charge, to any person obtaining a copy
20 * of this source file (the "Software"), to deal in the Software without
21 * restriction, including without limitation the rights to use, copy, modify,
22 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
23 * and to permit persons to whom the Software is furnished to do so, subject to
24 * the following conditions:
25 *
26 * The above copyright notice and this permission notice shall be included in
27 * all copies or substantial portions of the Software.
28 *
29 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
30 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
31 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
32 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
33 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
34 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
35 * IN THE SOFTWARE.
36 */
37
38 #include <linux/interrupt.h>
39 #include <linux/blkdev.h>
40 #include <linux/hdreg.h>
41 #include <linux/cdrom.h>
42 #include <linux/module.h>
43 #include <linux/slab.h>
44 #include <linux/mutex.h>
45 #include <linux/scatterlist.h>
46 #include <linux/bitmap.h>
47 #include <linux/list.h>
48
49 #include <xen/xen.h>
50 #include <xen/xenbus.h>
51 #include <xen/grant_table.h>
52 #include <xen/events.h>
53 #include <xen/page.h>
54 #include <xen/platform_pci.h>
55
56 #include <xen/interface/grant_table.h>
57 #include <xen/interface/io/blkif.h>
58 #include <xen/interface/io/protocols.h>
59
60 #include <asm/xen/hypervisor.h>
61
62 enum blkif_state {
63 BLKIF_STATE_DISCONNECTED,
64 BLKIF_STATE_CONNECTED,
65 BLKIF_STATE_SUSPENDED,
66 };
67
68 struct grant {
69 grant_ref_t gref;
70 unsigned long pfn;
71 struct list_head node;
72 };
73
74 struct blk_shadow {
75 struct blkif_request req;
76 struct request *request;
77 struct grant **grants_used;
78 struct grant **indirect_grants;
79 struct scatterlist *sg;
80 };
81
82 struct split_bio {
83 struct bio *bio;
84 atomic_t pending;
85 int err;
86 };
87
88 static DEFINE_MUTEX(blkfront_mutex);
89 static const struct block_device_operations xlvbd_block_fops;
90
91 /*
92 * Maximum number of segments in indirect requests, the actual value used by
93 * the frontend driver is the minimum of this value and the value provided
94 * by the backend driver.
95 */
96
97 static unsigned int xen_blkif_max_segments = 32;
98 module_param_named(max, xen_blkif_max_segments, int, S_IRUGO);
99 MODULE_PARM_DESC(max, "Maximum amount of segments in indirect requests (default is 32)");
100
101 #define BLK_RING_SIZE __CONST_RING_SIZE(blkif, PAGE_SIZE)
102
103 /*
104 * We have one of these per vbd, whether ide, scsi or 'other'. They
105 * hang in private_data off the gendisk structure. We may end up
106 * putting all kinds of interesting stuff here :-)
107 */
108 struct blkfront_info
109 {
110 spinlock_t io_lock;
111 struct mutex mutex;
112 struct xenbus_device *xbdev;
113 struct gendisk *gd;
114 int vdevice;
115 blkif_vdev_t handle;
116 enum blkif_state connected;
117 int ring_ref;
118 struct blkif_front_ring ring;
119 unsigned int evtchn, irq;
120 struct request_queue *rq;
121 struct work_struct work;
122 struct gnttab_free_callback callback;
123 struct blk_shadow shadow[BLK_RING_SIZE];
124 struct list_head persistent_gnts;
125 unsigned int persistent_gnts_c;
126 unsigned long shadow_free;
127 unsigned int feature_flush;
128 unsigned int flush_op;
129 unsigned int feature_discard:1;
130 unsigned int feature_secdiscard:1;
131 unsigned int discard_granularity;
132 unsigned int discard_alignment;
133 unsigned int feature_persistent:1;
134 unsigned int max_indirect_segments;
135 int is_ready;
136 };
137
138 static unsigned int nr_minors;
139 static unsigned long *minors;
140 static DEFINE_SPINLOCK(minor_lock);
141
142 #define MAXIMUM_OUTSTANDING_BLOCK_REQS \
143 (BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE)
144 #define GRANT_INVALID_REF 0
145
146 #define PARTS_PER_DISK 16
147 #define PARTS_PER_EXT_DISK 256
148
149 #define BLKIF_MAJOR(dev) ((dev)>>8)
150 #define BLKIF_MINOR(dev) ((dev) & 0xff)
151
152 #define EXT_SHIFT 28
153 #define EXTENDED (1<<EXT_SHIFT)
154 #define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
155 #define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
156 #define EMULATED_HD_DISK_MINOR_OFFSET (0)
157 #define EMULATED_HD_DISK_NAME_OFFSET (EMULATED_HD_DISK_MINOR_OFFSET / 256)
158 #define EMULATED_SD_DISK_MINOR_OFFSET (0)
159 #define EMULATED_SD_DISK_NAME_OFFSET (EMULATED_SD_DISK_MINOR_OFFSET / 256)
160
161 #define DEV_NAME "xvd" /* name in /dev */
162
163 #define SEGS_PER_INDIRECT_FRAME \
164 (PAGE_SIZE/sizeof(struct blkif_request_segment_aligned))
165 #define INDIRECT_GREFS(_segs) \
166 ((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)
167
168 static int blkfront_setup_indirect(struct blkfront_info *info);
169
170 static int get_id_from_freelist(struct blkfront_info *info)
171 {
172 unsigned long free = info->shadow_free;
173 BUG_ON(free >= BLK_RING_SIZE);
174 info->shadow_free = info->shadow[free].req.u.rw.id;
175 info->shadow[free].req.u.rw.id = 0x0fffffee; /* debug */
176 return free;
177 }
178
179 static int add_id_to_freelist(struct blkfront_info *info,
180 unsigned long id)
181 {
182 if (info->shadow[id].req.u.rw.id != id)
183 return -EINVAL;
184 if (info->shadow[id].request == NULL)
185 return -EINVAL;
186 info->shadow[id].req.u.rw.id = info->shadow_free;
187 info->shadow[id].request = NULL;
188 info->shadow_free = id;
189 return 0;
190 }
191
192 static int fill_grant_buffer(struct blkfront_info *info, int num)
193 {
194 struct page *granted_page;
195 struct grant *gnt_list_entry, *n;
196 int i = 0;
197
198 while(i < num) {
199 gnt_list_entry = kzalloc(sizeof(struct grant), GFP_NOIO);
200 if (!gnt_list_entry)
201 goto out_of_memory;
202
203 granted_page = alloc_page(GFP_NOIO);
204 if (!granted_page) {
205 kfree(gnt_list_entry);
206 goto out_of_memory;
207 }
208
209 gnt_list_entry->pfn = page_to_pfn(granted_page);
210 gnt_list_entry->gref = GRANT_INVALID_REF;
211 list_add(&gnt_list_entry->node, &info->persistent_gnts);
212 i++;
213 }
214
215 return 0;
216
217 out_of_memory:
218 list_for_each_entry_safe(gnt_list_entry, n,
219 &info->persistent_gnts, node) {
220 list_del(&gnt_list_entry->node);
221 __free_page(pfn_to_page(gnt_list_entry->pfn));
222 kfree(gnt_list_entry);
223 i--;
224 }
225 BUG_ON(i != 0);
226 return -ENOMEM;
227 }
228
229 static struct grant *get_grant(grant_ref_t *gref_head,
230 struct blkfront_info *info)
231 {
232 struct grant *gnt_list_entry;
233 unsigned long buffer_mfn;
234
235 BUG_ON(list_empty(&info->persistent_gnts));
236 gnt_list_entry = list_first_entry(&info->persistent_gnts, struct grant,
237 node);
238 list_del(&gnt_list_entry->node);
239
240 if (gnt_list_entry->gref != GRANT_INVALID_REF) {
241 info->persistent_gnts_c--;
242 return gnt_list_entry;
243 }
244
245 /* Assign a gref to this page */
246 gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head);
247 BUG_ON(gnt_list_entry->gref == -ENOSPC);
248 buffer_mfn = pfn_to_mfn(gnt_list_entry->pfn);
249 gnttab_grant_foreign_access_ref(gnt_list_entry->gref,
250 info->xbdev->otherend_id,
251 buffer_mfn, 0);
252 return gnt_list_entry;
253 }
254
255 static const char *op_name(int op)
256 {
257 static const char *const names[] = {
258 [BLKIF_OP_READ] = "read",
259 [BLKIF_OP_WRITE] = "write",
260 [BLKIF_OP_WRITE_BARRIER] = "barrier",
261 [BLKIF_OP_FLUSH_DISKCACHE] = "flush",
262 [BLKIF_OP_DISCARD] = "discard" };
263
264 if (op < 0 || op >= ARRAY_SIZE(names))
265 return "unknown";
266
267 if (!names[op])
268 return "reserved";
269
270 return names[op];
271 }
272 static int xlbd_reserve_minors(unsigned int minor, unsigned int nr)
273 {
274 unsigned int end = minor + nr;
275 int rc;
276
277 if (end > nr_minors) {
278 unsigned long *bitmap, *old;
279
280 bitmap = kcalloc(BITS_TO_LONGS(end), sizeof(*bitmap),
281 GFP_KERNEL);
282 if (bitmap == NULL)
283 return -ENOMEM;
284
285 spin_lock(&minor_lock);
286 if (end > nr_minors) {
287 old = minors;
288 memcpy(bitmap, minors,
289 BITS_TO_LONGS(nr_minors) * sizeof(*bitmap));
290 minors = bitmap;
291 nr_minors = BITS_TO_LONGS(end) * BITS_PER_LONG;
292 } else
293 old = bitmap;
294 spin_unlock(&minor_lock);
295 kfree(old);
296 }
297
298 spin_lock(&minor_lock);
299 if (find_next_bit(minors, end, minor) >= end) {
300 bitmap_set(minors, minor, nr);
301 rc = 0;
302 } else
303 rc = -EBUSY;
304 spin_unlock(&minor_lock);
305
306 return rc;
307 }
308
309 static void xlbd_release_minors(unsigned int minor, unsigned int nr)
310 {
311 unsigned int end = minor + nr;
312
313 BUG_ON(end > nr_minors);
314 spin_lock(&minor_lock);
315 bitmap_clear(minors, minor, nr);
316 spin_unlock(&minor_lock);
317 }
318
319 static void blkif_restart_queue_callback(void *arg)
320 {
321 struct blkfront_info *info = (struct blkfront_info *)arg;
322 schedule_work(&info->work);
323 }
324
325 static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
326 {
327 /* We don't have real geometry info, but let's at least return
328 values consistent with the size of the device */
329 sector_t nsect = get_capacity(bd->bd_disk);
330 sector_t cylinders = nsect;
331
332 hg->heads = 0xff;
333 hg->sectors = 0x3f;
334 sector_div(cylinders, hg->heads * hg->sectors);
335 hg->cylinders = cylinders;
336 if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect)
337 hg->cylinders = 0xffff;
338 return 0;
339 }
340
341 static int blkif_ioctl(struct block_device *bdev, fmode_t mode,
342 unsigned command, unsigned long argument)
343 {
344 struct blkfront_info *info = bdev->bd_disk->private_data;
345 int i;
346
347 dev_dbg(&info->xbdev->dev, "command: 0x%x, argument: 0x%lx\n",
348 command, (long)argument);
349
350 switch (command) {
351 case CDROMMULTISESSION:
352 dev_dbg(&info->xbdev->dev, "FIXME: support multisession CDs later\n");
353 for (i = 0; i < sizeof(struct cdrom_multisession); i++)
354 if (put_user(0, (char __user *)(argument + i)))
355 return -EFAULT;
356 return 0;
357
358 case CDROM_GET_CAPABILITY: {
359 struct gendisk *gd = info->gd;
360 if (gd->flags & GENHD_FL_CD)
361 return 0;
362 return -EINVAL;
363 }
364
365 default:
366 /*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
367 command);*/
368 return -EINVAL; /* same return as native Linux */
369 }
370
371 return 0;
372 }
373
374 /*
375 * Generate a Xen blkfront IO request from a blk layer request. Reads
376 * and writes are handled as expected.
377 *
378 * @req: a request struct
379 */
380 static int blkif_queue_request(struct request *req)
381 {
382 struct blkfront_info *info = req->rq_disk->private_data;
383 struct blkif_request *ring_req;
384 unsigned long id;
385 unsigned int fsect, lsect;
386 int i, ref, n;
387 struct blkif_request_segment_aligned *segments = NULL;
388
389 /*
390 * Used to store if we are able to queue the request by just using
391 * existing persistent grants, or if we have to get new grants,
392 * as there are not sufficiently many free.
393 */
394 bool new_persistent_gnts;
395 grant_ref_t gref_head;
396 struct grant *gnt_list_entry = NULL;
397 struct scatterlist *sg;
398 int nseg, max_grefs;
399
400 if (unlikely(info->connected != BLKIF_STATE_CONNECTED))
401 return 1;
402
403 max_grefs = info->max_indirect_segments ?
404 info->max_indirect_segments +
405 INDIRECT_GREFS(info->max_indirect_segments) :
406 BLKIF_MAX_SEGMENTS_PER_REQUEST;
407
408 /* Check if we have enough grants to allocate a requests */
409 if (info->persistent_gnts_c < max_grefs) {
410 new_persistent_gnts = 1;
411 if (gnttab_alloc_grant_references(
412 max_grefs - info->persistent_gnts_c,
413 &gref_head) < 0) {
414 gnttab_request_free_callback(
415 &info->callback,
416 blkif_restart_queue_callback,
417 info,
418 max_grefs);
419 return 1;
420 }
421 } else
422 new_persistent_gnts = 0;
423
424 /* Fill out a communications ring structure. */
425 ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
426 id = get_id_from_freelist(info);
427 info->shadow[id].request = req;
428
429 if (unlikely(req->cmd_flags & (REQ_DISCARD | REQ_SECURE))) {
430 ring_req->operation = BLKIF_OP_DISCARD;
431 ring_req->u.discard.nr_sectors = blk_rq_sectors(req);
432 ring_req->u.discard.id = id;
433 ring_req->u.discard.sector_number = (blkif_sector_t)blk_rq_pos(req);
434 if ((req->cmd_flags & REQ_SECURE) && info->feature_secdiscard)
435 ring_req->u.discard.flag = BLKIF_DISCARD_SECURE;
436 else
437 ring_req->u.discard.flag = 0;
438 } else {
439 BUG_ON(info->max_indirect_segments == 0 &&
440 req->nr_phys_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST);
441 BUG_ON(info->max_indirect_segments &&
442 req->nr_phys_segments > info->max_indirect_segments);
443 nseg = blk_rq_map_sg(req->q, req, info->shadow[id].sg);
444 ring_req->u.rw.id = id;
445 if (nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST) {
446 /*
447 * The indirect operation can only be a BLKIF_OP_READ or
448 * BLKIF_OP_WRITE
449 */
450 BUG_ON(req->cmd_flags & (REQ_FLUSH | REQ_FUA));
451 ring_req->operation = BLKIF_OP_INDIRECT;
452 ring_req->u.indirect.indirect_op = rq_data_dir(req) ?
453 BLKIF_OP_WRITE : BLKIF_OP_READ;
454 ring_req->u.indirect.sector_number = (blkif_sector_t)blk_rq_pos(req);
455 ring_req->u.indirect.handle = info->handle;
456 ring_req->u.indirect.nr_segments = nseg;
457 } else {
458 ring_req->u.rw.sector_number = (blkif_sector_t)blk_rq_pos(req);
459 ring_req->u.rw.handle = info->handle;
460 ring_req->operation = rq_data_dir(req) ?
461 BLKIF_OP_WRITE : BLKIF_OP_READ;
462 if (req->cmd_flags & (REQ_FLUSH | REQ_FUA)) {
463 /*
464 * Ideally we can do an unordered flush-to-disk. In case the
465 * backend onlysupports barriers, use that. A barrier request
466 * a superset of FUA, so we can implement it the same
467 * way. (It's also a FLUSH+FUA, since it is
468 * guaranteed ordered WRT previous writes.)
469 */
470 ring_req->operation = info->flush_op;
471 }
472 ring_req->u.rw.nr_segments = nseg;
473 }
474 for_each_sg(info->shadow[id].sg, sg, nseg, i) {
475 fsect = sg->offset >> 9;
476 lsect = fsect + (sg->length >> 9) - 1;
477
478 if ((ring_req->operation == BLKIF_OP_INDIRECT) &&
479 (i % SEGS_PER_INDIRECT_FRAME == 0)) {
480 if (segments)
481 kunmap_atomic(segments);
482
483 n = i / SEGS_PER_INDIRECT_FRAME;
484 gnt_list_entry = get_grant(&gref_head, info);
485 info->shadow[id].indirect_grants[n] = gnt_list_entry;
486 segments = kmap_atomic(pfn_to_page(gnt_list_entry->pfn));
487 ring_req->u.indirect.indirect_grefs[n] = gnt_list_entry->gref;
488 }
489
490 gnt_list_entry = get_grant(&gref_head, info);
491 ref = gnt_list_entry->gref;
492
493 info->shadow[id].grants_used[i] = gnt_list_entry;
494
495 if (rq_data_dir(req)) {
496 char *bvec_data;
497 void *shared_data;
498
499 BUG_ON(sg->offset + sg->length > PAGE_SIZE);
500
501 shared_data = kmap_atomic(pfn_to_page(gnt_list_entry->pfn));
502 bvec_data = kmap_atomic(sg_page(sg));
503
504 /*
505 * this does not wipe data stored outside the
506 * range sg->offset..sg->offset+sg->length.
507 * Therefore, blkback *could* see data from
508 * previous requests. This is OK as long as
509 * persistent grants are shared with just one
510 * domain. It may need refactoring if this
511 * changes
512 */
513 memcpy(shared_data + sg->offset,
514 bvec_data + sg->offset,
515 sg->length);
516
517 kunmap_atomic(bvec_data);
518 kunmap_atomic(shared_data);
519 }
520 if (ring_req->operation != BLKIF_OP_INDIRECT) {
521 ring_req->u.rw.seg[i] =
522 (struct blkif_request_segment) {
523 .gref = ref,
524 .first_sect = fsect,
525 .last_sect = lsect };
526 } else {
527 n = i % SEGS_PER_INDIRECT_FRAME;
528 segments[n] =
529 (struct blkif_request_segment_aligned) {
530 .gref = ref,
531 .first_sect = fsect,
532 .last_sect = lsect };
533 }
534 }
535 if (segments)
536 kunmap_atomic(segments);
537 }
538
539 info->ring.req_prod_pvt++;
540
541 /* Keep a private copy so we can reissue requests when recovering. */
542 info->shadow[id].req = *ring_req;
543
544 if (new_persistent_gnts)
545 gnttab_free_grant_references(gref_head);
546
547 return 0;
548 }
549
550
551 static inline void flush_requests(struct blkfront_info *info)
552 {
553 int notify;
554
555 RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify);
556
557 if (notify)
558 notify_remote_via_irq(info->irq);
559 }
560
561 /*
562 * do_blkif_request
563 * read a block; request is in a request queue
564 */
565 static void do_blkif_request(struct request_queue *rq)
566 {
567 struct blkfront_info *info = NULL;
568 struct request *req;
569 int queued;
570
571 pr_debug("Entered do_blkif_request\n");
572
573 queued = 0;
574
575 while ((req = blk_peek_request(rq)) != NULL) {
576 info = req->rq_disk->private_data;
577
578 if (RING_FULL(&info->ring))
579 goto wait;
580
581 blk_start_request(req);
582
583 if ((req->cmd_type != REQ_TYPE_FS) ||
584 ((req->cmd_flags & (REQ_FLUSH | REQ_FUA)) &&
585 !info->flush_op)) {
586 __blk_end_request_all(req, -EIO);
587 continue;
588 }
589
590 pr_debug("do_blk_req %p: cmd %p, sec %lx, "
591 "(%u/%u) buffer:%p [%s]\n",
592 req, req->cmd, (unsigned long)blk_rq_pos(req),
593 blk_rq_cur_sectors(req), blk_rq_sectors(req),
594 req->buffer, rq_data_dir(req) ? "write" : "read");
595
596 if (blkif_queue_request(req)) {
597 blk_requeue_request(rq, req);
598 wait:
599 /* Avoid pointless unplugs. */
600 blk_stop_queue(rq);
601 break;
602 }
603
604 queued++;
605 }
606
607 if (queued != 0)
608 flush_requests(info);
609 }
610
611 static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size,
612 unsigned int physical_sector_size,
613 unsigned int segments)
614 {
615 struct request_queue *rq;
616 struct blkfront_info *info = gd->private_data;
617
618 rq = blk_init_queue(do_blkif_request, &info->io_lock);
619 if (rq == NULL)
620 return -1;
621
622 queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq);
623
624 if (info->feature_discard) {
625 queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, rq);
626 blk_queue_max_discard_sectors(rq, get_capacity(gd));
627 rq->limits.discard_granularity = info->discard_granularity;
628 rq->limits.discard_alignment = info->discard_alignment;
629 if (info->feature_secdiscard)
630 queue_flag_set_unlocked(QUEUE_FLAG_SECDISCARD, rq);
631 }
632
633 /* Hard sector size and max sectors impersonate the equiv. hardware. */
634 blk_queue_logical_block_size(rq, sector_size);
635 blk_queue_physical_block_size(rq, physical_sector_size);
636 blk_queue_max_hw_sectors(rq, (segments * PAGE_SIZE) / 512);
637
638 /* Each segment in a request is up to an aligned page in size. */
639 blk_queue_segment_boundary(rq, PAGE_SIZE - 1);
640 blk_queue_max_segment_size(rq, PAGE_SIZE);
641
642 /* Ensure a merged request will fit in a single I/O ring slot. */
643 blk_queue_max_segments(rq, segments);
644
645 /* Make sure buffer addresses are sector-aligned. */
646 blk_queue_dma_alignment(rq, 511);
647
648 /* Make sure we don't use bounce buffers. */
649 blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);
650
651 gd->queue = rq;
652
653 return 0;
654 }
655
656
657 static void xlvbd_flush(struct blkfront_info *info)
658 {
659 blk_queue_flush(info->rq, info->feature_flush);
660 printk(KERN_INFO "blkfront: %s: %s: %s %s %s %s %s\n",
661 info->gd->disk_name,
662 info->flush_op == BLKIF_OP_WRITE_BARRIER ?
663 "barrier" : (info->flush_op == BLKIF_OP_FLUSH_DISKCACHE ?
664 "flush diskcache" : "barrier or flush"),
665 info->feature_flush ? "enabled;" : "disabled;",
666 "persistent grants:",
667 info->feature_persistent ? "enabled;" : "disabled;",
668 "indirect descriptors:",
669 info->max_indirect_segments ? "enabled;" : "disabled;");
670 }
671
672 static int xen_translate_vdev(int vdevice, int *minor, unsigned int *offset)
673 {
674 int major;
675 major = BLKIF_MAJOR(vdevice);
676 *minor = BLKIF_MINOR(vdevice);
677 switch (major) {
678 case XEN_IDE0_MAJOR:
679 *offset = (*minor / 64) + EMULATED_HD_DISK_NAME_OFFSET;
680 *minor = ((*minor / 64) * PARTS_PER_DISK) +
681 EMULATED_HD_DISK_MINOR_OFFSET;
682 break;
683 case XEN_IDE1_MAJOR:
684 *offset = (*minor / 64) + 2 + EMULATED_HD_DISK_NAME_OFFSET;
685 *minor = (((*minor / 64) + 2) * PARTS_PER_DISK) +
686 EMULATED_HD_DISK_MINOR_OFFSET;
687 break;
688 case XEN_SCSI_DISK0_MAJOR:
689 *offset = (*minor / PARTS_PER_DISK) + EMULATED_SD_DISK_NAME_OFFSET;
690 *minor = *minor + EMULATED_SD_DISK_MINOR_OFFSET;
691 break;
692 case XEN_SCSI_DISK1_MAJOR:
693 case XEN_SCSI_DISK2_MAJOR:
694 case XEN_SCSI_DISK3_MAJOR:
695 case XEN_SCSI_DISK4_MAJOR:
696 case XEN_SCSI_DISK5_MAJOR:
697 case XEN_SCSI_DISK6_MAJOR:
698 case XEN_SCSI_DISK7_MAJOR:
699 *offset = (*minor / PARTS_PER_DISK) +
700 ((major - XEN_SCSI_DISK1_MAJOR + 1) * 16) +
701 EMULATED_SD_DISK_NAME_OFFSET;
702 *minor = *minor +
703 ((major - XEN_SCSI_DISK1_MAJOR + 1) * 16 * PARTS_PER_DISK) +
704 EMULATED_SD_DISK_MINOR_OFFSET;
705 break;
706 case XEN_SCSI_DISK8_MAJOR:
707 case XEN_SCSI_DISK9_MAJOR:
708 case XEN_SCSI_DISK10_MAJOR:
709 case XEN_SCSI_DISK11_MAJOR:
710 case XEN_SCSI_DISK12_MAJOR:
711 case XEN_SCSI_DISK13_MAJOR:
712 case XEN_SCSI_DISK14_MAJOR:
713 case XEN_SCSI_DISK15_MAJOR:
714 *offset = (*minor / PARTS_PER_DISK) +
715 ((major - XEN_SCSI_DISK8_MAJOR + 8) * 16) +
716 EMULATED_SD_DISK_NAME_OFFSET;
717 *minor = *minor +
718 ((major - XEN_SCSI_DISK8_MAJOR + 8) * 16 * PARTS_PER_DISK) +
719 EMULATED_SD_DISK_MINOR_OFFSET;
720 break;
721 case XENVBD_MAJOR:
722 *offset = *minor / PARTS_PER_DISK;
723 break;
724 default:
725 printk(KERN_WARNING "blkfront: your disk configuration is "
726 "incorrect, please use an xvd device instead\n");
727 return -ENODEV;
728 }
729 return 0;
730 }
731
732 static char *encode_disk_name(char *ptr, unsigned int n)
733 {
734 if (n >= 26)
735 ptr = encode_disk_name(ptr, n / 26 - 1);
736 *ptr = 'a' + n % 26;
737 return ptr + 1;
738 }
739
740 static int xlvbd_alloc_gendisk(blkif_sector_t capacity,
741 struct blkfront_info *info,
742 u16 vdisk_info, u16 sector_size,
743 unsigned int physical_sector_size)
744 {
745 struct gendisk *gd;
746 int nr_minors = 1;
747 int err;
748 unsigned int offset;
749 int minor;
750 int nr_parts;
751 char *ptr;
752
753 BUG_ON(info->gd != NULL);
754 BUG_ON(info->rq != NULL);
755
756 if ((info->vdevice>>EXT_SHIFT) > 1) {
757 /* this is above the extended range; something is wrong */
758 printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", info->vdevice);
759 return -ENODEV;
760 }
761
762 if (!VDEV_IS_EXTENDED(info->vdevice)) {
763 err = xen_translate_vdev(info->vdevice, &minor, &offset);
764 if (err)
765 return err;
766 nr_parts = PARTS_PER_DISK;
767 } else {
768 minor = BLKIF_MINOR_EXT(info->vdevice);
769 nr_parts = PARTS_PER_EXT_DISK;
770 offset = minor / nr_parts;
771 if (xen_hvm_domain() && offset < EMULATED_HD_DISK_NAME_OFFSET + 4)
772 printk(KERN_WARNING "blkfront: vdevice 0x%x might conflict with "
773 "emulated IDE disks,\n\t choose an xvd device name"
774 "from xvde on\n", info->vdevice);
775 }
776 if (minor >> MINORBITS) {
777 pr_warn("blkfront: %#x's minor (%#x) out of range; ignoring\n",
778 info->vdevice, minor);
779 return -ENODEV;
780 }
781
782 if ((minor % nr_parts) == 0)
783 nr_minors = nr_parts;
784
785 err = xlbd_reserve_minors(minor, nr_minors);
786 if (err)
787 goto out;
788 err = -ENODEV;
789
790 gd = alloc_disk(nr_minors);
791 if (gd == NULL)
792 goto release;
793
794 strcpy(gd->disk_name, DEV_NAME);
795 ptr = encode_disk_name(gd->disk_name + sizeof(DEV_NAME) - 1, offset);
796 BUG_ON(ptr >= gd->disk_name + DISK_NAME_LEN);
797 if (nr_minors > 1)
798 *ptr = 0;
799 else
800 snprintf(ptr, gd->disk_name + DISK_NAME_LEN - ptr,
801 "%d", minor & (nr_parts - 1));
802
803 gd->major = XENVBD_MAJOR;
804 gd->first_minor = minor;
805 gd->fops = &xlvbd_block_fops;
806 gd->private_data = info;
807 gd->driverfs_dev = &(info->xbdev->dev);
808 set_capacity(gd, capacity);
809
810 if (xlvbd_init_blk_queue(gd, sector_size, physical_sector_size,
811 info->max_indirect_segments ? :
812 BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
813 del_gendisk(gd);
814 goto release;
815 }
816
817 info->rq = gd->queue;
818 info->gd = gd;
819
820 xlvbd_flush(info);
821
822 if (vdisk_info & VDISK_READONLY)
823 set_disk_ro(gd, 1);
824
825 if (vdisk_info & VDISK_REMOVABLE)
826 gd->flags |= GENHD_FL_REMOVABLE;
827
828 if (vdisk_info & VDISK_CDROM)
829 gd->flags |= GENHD_FL_CD;
830
831 return 0;
832
833 release:
834 xlbd_release_minors(minor, nr_minors);
835 out:
836 return err;
837 }
838
839 static void xlvbd_release_gendisk(struct blkfront_info *info)
840 {
841 unsigned int minor, nr_minors;
842 unsigned long flags;
843
844 if (info->rq == NULL)
845 return;
846
847 spin_lock_irqsave(&info->io_lock, flags);
848
849 /* No more blkif_request(). */
850 blk_stop_queue(info->rq);
851
852 /* No more gnttab callback work. */
853 gnttab_cancel_free_callback(&info->callback);
854 spin_unlock_irqrestore(&info->io_lock, flags);
855
856 /* Flush gnttab callback work. Must be done with no locks held. */
857 flush_work(&info->work);
858
859 del_gendisk(info->gd);
860
861 minor = info->gd->first_minor;
862 nr_minors = info->gd->minors;
863 xlbd_release_minors(minor, nr_minors);
864
865 blk_cleanup_queue(info->rq);
866 info->rq = NULL;
867
868 put_disk(info->gd);
869 info->gd = NULL;
870 }
871
872 static void kick_pending_request_queues(struct blkfront_info *info)
873 {
874 if (!RING_FULL(&info->ring)) {
875 /* Re-enable calldowns. */
876 blk_start_queue(info->rq);
877 /* Kick things off immediately. */
878 do_blkif_request(info->rq);
879 }
880 }
881
882 static void blkif_restart_queue(struct work_struct *work)
883 {
884 struct blkfront_info *info = container_of(work, struct blkfront_info, work);
885
886 spin_lock_irq(&info->io_lock);
887 if (info->connected == BLKIF_STATE_CONNECTED)
888 kick_pending_request_queues(info);
889 spin_unlock_irq(&info->io_lock);
890 }
891
892 static void blkif_free(struct blkfront_info *info, int suspend)
893 {
894 struct grant *persistent_gnt;
895 struct grant *n;
896 int i, j, segs;
897
898 /* Prevent new requests being issued until we fix things up. */
899 spin_lock_irq(&info->io_lock);
900 info->connected = suspend ?
901 BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
902 /* No more blkif_request(). */
903 if (info->rq)
904 blk_stop_queue(info->rq);
905
906 /* Remove all persistent grants */
907 if (!list_empty(&info->persistent_gnts)) {
908 list_for_each_entry_safe(persistent_gnt, n,
909 &info->persistent_gnts, node) {
910 list_del(&persistent_gnt->node);
911 if (persistent_gnt->gref != GRANT_INVALID_REF) {
912 gnttab_end_foreign_access(persistent_gnt->gref,
913 0, 0UL);
914 info->persistent_gnts_c--;
915 }
916 __free_page(pfn_to_page(persistent_gnt->pfn));
917 kfree(persistent_gnt);
918 }
919 }
920 BUG_ON(info->persistent_gnts_c != 0);
921
922 for (i = 0; i < BLK_RING_SIZE; i++) {
923 /*
924 * Clear persistent grants present in requests already
925 * on the shared ring
926 */
927 if (!info->shadow[i].request)
928 goto free_shadow;
929
930 segs = info->shadow[i].req.operation == BLKIF_OP_INDIRECT ?
931 info->shadow[i].req.u.indirect.nr_segments :
932 info->shadow[i].req.u.rw.nr_segments;
933 for (j = 0; j < segs; j++) {
934 persistent_gnt = info->shadow[i].grants_used[j];
935 gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL);
936 __free_page(pfn_to_page(persistent_gnt->pfn));
937 kfree(persistent_gnt);
938 }
939
940 if (info->shadow[i].req.operation != BLKIF_OP_INDIRECT)
941 /*
942 * If this is not an indirect operation don't try to
943 * free indirect segments
944 */
945 goto free_shadow;
946
947 for (j = 0; j < INDIRECT_GREFS(segs); j++) {
948 persistent_gnt = info->shadow[i].indirect_grants[j];
949 gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL);
950 __free_page(pfn_to_page(persistent_gnt->pfn));
951 kfree(persistent_gnt);
952 }
953
954 free_shadow:
955 kfree(info->shadow[i].grants_used);
956 info->shadow[i].grants_used = NULL;
957 kfree(info->shadow[i].indirect_grants);
958 info->shadow[i].indirect_grants = NULL;
959 kfree(info->shadow[i].sg);
960 info->shadow[i].sg = NULL;
961 }
962
963 /* No more gnttab callback work. */
964 gnttab_cancel_free_callback(&info->callback);
965 spin_unlock_irq(&info->io_lock);
966
967 /* Flush gnttab callback work. Must be done with no locks held. */
968 flush_work(&info->work);
969
970 /* Free resources associated with old device channel. */
971 if (info->ring_ref != GRANT_INVALID_REF) {
972 gnttab_end_foreign_access(info->ring_ref, 0,
973 (unsigned long)info->ring.sring);
974 info->ring_ref = GRANT_INVALID_REF;
975 info->ring.sring = NULL;
976 }
977 if (info->irq)
978 unbind_from_irqhandler(info->irq, info);
979 info->evtchn = info->irq = 0;
980
981 }
982
983 static void blkif_completion(struct blk_shadow *s, struct blkfront_info *info,
984 struct blkif_response *bret)
985 {
986 int i = 0;
987 struct scatterlist *sg;
988 char *bvec_data;
989 void *shared_data;
990 int nseg;
991
992 nseg = s->req.operation == BLKIF_OP_INDIRECT ?
993 s->req.u.indirect.nr_segments : s->req.u.rw.nr_segments;
994
995 if (bret->operation == BLKIF_OP_READ) {
996 /*
997 * Copy the data received from the backend into the bvec.
998 * Since bv_offset can be different than 0, and bv_len different
999 * than PAGE_SIZE, we have to keep track of the current offset,
1000 * to be sure we are copying the data from the right shared page.
1001 */
1002 for_each_sg(s->sg, sg, nseg, i) {
1003 BUG_ON(sg->offset + sg->length > PAGE_SIZE);
1004 shared_data = kmap_atomic(
1005 pfn_to_page(s->grants_used[i]->pfn));
1006 bvec_data = kmap_atomic(sg_page(sg));
1007 memcpy(bvec_data + sg->offset,
1008 shared_data + sg->offset,
1009 sg->length);
1010 kunmap_atomic(bvec_data);
1011 kunmap_atomic(shared_data);
1012 }
1013 }
1014 /* Add the persistent grant into the list of free grants */
1015 for (i = 0; i < nseg; i++) {
1016 list_add(&s->grants_used[i]->node, &info->persistent_gnts);
1017 info->persistent_gnts_c++;
1018 }
1019 if (s->req.operation == BLKIF_OP_INDIRECT) {
1020 for (i = 0; i < INDIRECT_GREFS(nseg); i++) {
1021 list_add(&s->indirect_grants[i]->node, &info->persistent_gnts);
1022 info->persistent_gnts_c++;
1023 }
1024 }
1025 }
1026
1027 static irqreturn_t blkif_interrupt(int irq, void *dev_id)
1028 {
1029 struct request *req;
1030 struct blkif_response *bret;
1031 RING_IDX i, rp;
1032 unsigned long flags;
1033 struct blkfront_info *info = (struct blkfront_info *)dev_id;
1034 int error;
1035
1036 spin_lock_irqsave(&info->io_lock, flags);
1037
1038 if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
1039 spin_unlock_irqrestore(&info->io_lock, flags);
1040 return IRQ_HANDLED;
1041 }
1042
1043 again:
1044 rp = info->ring.sring->rsp_prod;
1045 rmb(); /* Ensure we see queued responses up to 'rp'. */
1046
1047 for (i = info->ring.rsp_cons; i != rp; i++) {
1048 unsigned long id;
1049
1050 bret = RING_GET_RESPONSE(&info->ring, i);
1051 id = bret->id;
1052 /*
1053 * The backend has messed up and given us an id that we would
1054 * never have given to it (we stamp it up to BLK_RING_SIZE -
1055 * look in get_id_from_freelist.
1056 */
1057 if (id >= BLK_RING_SIZE) {
1058 WARN(1, "%s: response to %s has incorrect id (%ld)\n",
1059 info->gd->disk_name, op_name(bret->operation), id);
1060 /* We can't safely get the 'struct request' as
1061 * the id is busted. */
1062 continue;
1063 }
1064 req = info->shadow[id].request;
1065
1066 if (bret->operation != BLKIF_OP_DISCARD)
1067 blkif_completion(&info->shadow[id], info, bret);
1068
1069 if (add_id_to_freelist(info, id)) {
1070 WARN(1, "%s: response to %s (id %ld) couldn't be recycled!\n",
1071 info->gd->disk_name, op_name(bret->operation), id);
1072 continue;
1073 }
1074
1075 error = (bret->status == BLKIF_RSP_OKAY) ? 0 : -EIO;
1076 switch (bret->operation) {
1077 case BLKIF_OP_DISCARD:
1078 if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
1079 struct request_queue *rq = info->rq;
1080 printk(KERN_WARNING "blkfront: %s: %s op failed\n",
1081 info->gd->disk_name, op_name(bret->operation));
1082 error = -EOPNOTSUPP;
1083 info->feature_discard = 0;
1084 info->feature_secdiscard = 0;
1085 queue_flag_clear(QUEUE_FLAG_DISCARD, rq);
1086 queue_flag_clear(QUEUE_FLAG_SECDISCARD, rq);
1087 }
1088 __blk_end_request_all(req, error);
1089 break;
1090 case BLKIF_OP_FLUSH_DISKCACHE:
1091 case BLKIF_OP_WRITE_BARRIER:
1092 if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
1093 printk(KERN_WARNING "blkfront: %s: %s op failed\n",
1094 info->gd->disk_name, op_name(bret->operation));
1095 error = -EOPNOTSUPP;
1096 }
1097 if (unlikely(bret->status == BLKIF_RSP_ERROR &&
1098 info->shadow[id].req.u.rw.nr_segments == 0)) {
1099 printk(KERN_WARNING "blkfront: %s: empty %s op failed\n",
1100 info->gd->disk_name, op_name(bret->operation));
1101 error = -EOPNOTSUPP;
1102 }
1103 if (unlikely(error)) {
1104 if (error == -EOPNOTSUPP)
1105 error = 0;
1106 info->feature_flush = 0;
1107 info->flush_op = 0;
1108 xlvbd_flush(info);
1109 }
1110 /* fall through */
1111 case BLKIF_OP_READ:
1112 case BLKIF_OP_WRITE:
1113 if (unlikely(bret->status != BLKIF_RSP_OKAY))
1114 dev_dbg(&info->xbdev->dev, "Bad return from blkdev data "
1115 "request: %x\n", bret->status);
1116
1117 __blk_end_request_all(req, error);
1118 break;
1119 default:
1120 BUG();
1121 }
1122 }
1123
1124 info->ring.rsp_cons = i;
1125
1126 if (i != info->ring.req_prod_pvt) {
1127 int more_to_do;
1128 RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
1129 if (more_to_do)
1130 goto again;
1131 } else
1132 info->ring.sring->rsp_event = i + 1;
1133
1134 kick_pending_request_queues(info);
1135
1136 spin_unlock_irqrestore(&info->io_lock, flags);
1137
1138 return IRQ_HANDLED;
1139 }
1140
1141
1142 static int setup_blkring(struct xenbus_device *dev,
1143 struct blkfront_info *info)
1144 {
1145 struct blkif_sring *sring;
1146 int err;
1147
1148 info->ring_ref = GRANT_INVALID_REF;
1149
1150 sring = (struct blkif_sring *)__get_free_page(GFP_NOIO | __GFP_HIGH);
1151 if (!sring) {
1152 xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
1153 return -ENOMEM;
1154 }
1155 SHARED_RING_INIT(sring);
1156 FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
1157
1158 err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring));
1159 if (err < 0) {
1160 free_page((unsigned long)sring);
1161 info->ring.sring = NULL;
1162 goto fail;
1163 }
1164 info->ring_ref = err;
1165
1166 err = xenbus_alloc_evtchn(dev, &info->evtchn);
1167 if (err)
1168 goto fail;
1169
1170 err = bind_evtchn_to_irqhandler(info->evtchn, blkif_interrupt, 0,
1171 "blkif", info);
1172 if (err <= 0) {
1173 xenbus_dev_fatal(dev, err,
1174 "bind_evtchn_to_irqhandler failed");
1175 goto fail;
1176 }
1177 info->irq = err;
1178
1179 return 0;
1180 fail:
1181 blkif_free(info, 0);
1182 return err;
1183 }
1184
1185
1186 /* Common code used when first setting up, and when resuming. */
1187 static int talk_to_blkback(struct xenbus_device *dev,
1188 struct blkfront_info *info)
1189 {
1190 const char *message = NULL;
1191 struct xenbus_transaction xbt;
1192 int err;
1193
1194 /* Create shared ring, alloc event channel. */
1195 err = setup_blkring(dev, info);
1196 if (err)
1197 goto out;
1198
1199 again:
1200 err = xenbus_transaction_start(&xbt);
1201 if (err) {
1202 xenbus_dev_fatal(dev, err, "starting transaction");
1203 goto destroy_blkring;
1204 }
1205
1206 err = xenbus_printf(xbt, dev->nodename,
1207 "ring-ref", "%u", info->ring_ref);
1208 if (err) {
1209 message = "writing ring-ref";
1210 goto abort_transaction;
1211 }
1212 err = xenbus_printf(xbt, dev->nodename,
1213 "event-channel", "%u", info->evtchn);
1214 if (err) {
1215 message = "writing event-channel";
1216 goto abort_transaction;
1217 }
1218 err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
1219 XEN_IO_PROTO_ABI_NATIVE);
1220 if (err) {
1221 message = "writing protocol";
1222 goto abort_transaction;
1223 }
1224 err = xenbus_printf(xbt, dev->nodename,
1225 "feature-persistent", "%u", 1);
1226 if (err)
1227 dev_warn(&dev->dev,
1228 "writing persistent grants feature to xenbus");
1229
1230 err = xenbus_transaction_end(xbt, 0);
1231 if (err) {
1232 if (err == -EAGAIN)
1233 goto again;
1234 xenbus_dev_fatal(dev, err, "completing transaction");
1235 goto destroy_blkring;
1236 }
1237
1238 xenbus_switch_state(dev, XenbusStateInitialised);
1239
1240 return 0;
1241
1242 abort_transaction:
1243 xenbus_transaction_end(xbt, 1);
1244 if (message)
1245 xenbus_dev_fatal(dev, err, "%s", message);
1246 destroy_blkring:
1247 blkif_free(info, 0);
1248 out:
1249 return err;
1250 }
1251
1252 /**
1253 * Entry point to this code when a new device is created. Allocate the basic
1254 * structures and the ring buffer for communication with the backend, and
1255 * inform the backend of the appropriate details for those. Switch to
1256 * Initialised state.
1257 */
1258 static int blkfront_probe(struct xenbus_device *dev,
1259 const struct xenbus_device_id *id)
1260 {
1261 int err, vdevice, i;
1262 struct blkfront_info *info;
1263
1264 /* FIXME: Use dynamic device id if this is not set. */
1265 err = xenbus_scanf(XBT_NIL, dev->nodename,
1266 "virtual-device", "%i", &vdevice);
1267 if (err != 1) {
1268 /* go looking in the extended area instead */
1269 err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext",
1270 "%i", &vdevice);
1271 if (err != 1) {
1272 xenbus_dev_fatal(dev, err, "reading virtual-device");
1273 return err;
1274 }
1275 }
1276
1277 if (xen_hvm_domain()) {
1278 char *type;
1279 int len;
1280 /* no unplug has been done: do not hook devices != xen vbds */
1281 if (xen_platform_pci_unplug & XEN_UNPLUG_UNNECESSARY) {
1282 int major;
1283
1284 if (!VDEV_IS_EXTENDED(vdevice))
1285 major = BLKIF_MAJOR(vdevice);
1286 else
1287 major = XENVBD_MAJOR;
1288
1289 if (major != XENVBD_MAJOR) {
1290 printk(KERN_INFO
1291 "%s: HVM does not support vbd %d as xen block device\n",
1292 __FUNCTION__, vdevice);
1293 return -ENODEV;
1294 }
1295 }
1296 /* do not create a PV cdrom device if we are an HVM guest */
1297 type = xenbus_read(XBT_NIL, dev->nodename, "device-type", &len);
1298 if (IS_ERR(type))
1299 return -ENODEV;
1300 if (strncmp(type, "cdrom", 5) == 0) {
1301 kfree(type);
1302 return -ENODEV;
1303 }
1304 kfree(type);
1305 }
1306 info = kzalloc(sizeof(*info), GFP_KERNEL);
1307 if (!info) {
1308 xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
1309 return -ENOMEM;
1310 }
1311
1312 mutex_init(&info->mutex);
1313 spin_lock_init(&info->io_lock);
1314 info->xbdev = dev;
1315 info->vdevice = vdevice;
1316 INIT_LIST_HEAD(&info->persistent_gnts);
1317 info->persistent_gnts_c = 0;
1318 info->connected = BLKIF_STATE_DISCONNECTED;
1319 INIT_WORK(&info->work, blkif_restart_queue);
1320
1321 for (i = 0; i < BLK_RING_SIZE; i++)
1322 info->shadow[i].req.u.rw.id = i+1;
1323 info->shadow[BLK_RING_SIZE-1].req.u.rw.id = 0x0fffffff;
1324
1325 /* Front end dir is a number, which is used as the id. */
1326 info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0);
1327 dev_set_drvdata(&dev->dev, info);
1328
1329 err = talk_to_blkback(dev, info);
1330 if (err) {
1331 kfree(info);
1332 dev_set_drvdata(&dev->dev, NULL);
1333 return err;
1334 }
1335
1336 return 0;
1337 }
1338
1339 /*
1340 * This is a clone of md_trim_bio, used to split a bio into smaller ones
1341 */
1342 static void trim_bio(struct bio *bio, int offset, int size)
1343 {
1344 /* 'bio' is a cloned bio which we need to trim to match
1345 * the given offset and size.
1346 * This requires adjusting bi_sector, bi_size, and bi_io_vec
1347 */
1348 int i;
1349 struct bio_vec *bvec;
1350 int sofar = 0;
1351
1352 size <<= 9;
1353 if (offset == 0 && size == bio->bi_size)
1354 return;
1355
1356 bio->bi_sector += offset;
1357 bio->bi_size = size;
1358 offset <<= 9;
1359 clear_bit(BIO_SEG_VALID, &bio->bi_flags);
1360
1361 while (bio->bi_idx < bio->bi_vcnt &&
1362 bio->bi_io_vec[bio->bi_idx].bv_len <= offset) {
1363 /* remove this whole bio_vec */
1364 offset -= bio->bi_io_vec[bio->bi_idx].bv_len;
1365 bio->bi_idx++;
1366 }
1367 if (bio->bi_idx < bio->bi_vcnt) {
1368 bio->bi_io_vec[bio->bi_idx].bv_offset += offset;
1369 bio->bi_io_vec[bio->bi_idx].bv_len -= offset;
1370 }
1371 /* avoid any complications with bi_idx being non-zero*/
1372 if (bio->bi_idx) {
1373 memmove(bio->bi_io_vec, bio->bi_io_vec+bio->bi_idx,
1374 (bio->bi_vcnt - bio->bi_idx) * sizeof(struct bio_vec));
1375 bio->bi_vcnt -= bio->bi_idx;
1376 bio->bi_idx = 0;
1377 }
1378 /* Make sure vcnt and last bv are not too big */
1379 bio_for_each_segment(bvec, bio, i) {
1380 if (sofar + bvec->bv_len > size)
1381 bvec->bv_len = size - sofar;
1382 if (bvec->bv_len == 0) {
1383 bio->bi_vcnt = i;
1384 break;
1385 }
1386 sofar += bvec->bv_len;
1387 }
1388 }
1389
1390 static void split_bio_end(struct bio *bio, int error)
1391 {
1392 struct split_bio *split_bio = bio->bi_private;
1393
1394 if (error)
1395 split_bio->err = error;
1396
1397 if (atomic_dec_and_test(&split_bio->pending)) {
1398 split_bio->bio->bi_phys_segments = 0;
1399 bio_endio(split_bio->bio, split_bio->err);
1400 kfree(split_bio);
1401 }
1402 bio_put(bio);
1403 }
1404
1405 static int blkif_recover(struct blkfront_info *info)
1406 {
1407 int i;
1408 struct request *req, *n;
1409 struct blk_shadow *copy;
1410 int rc;
1411 struct bio *bio, *cloned_bio;
1412 struct bio_list bio_list, merge_bio;
1413 unsigned int segs, offset;
1414 int pending, size;
1415 struct split_bio *split_bio;
1416 struct list_head requests;
1417
1418 /* Stage 1: Make a safe copy of the shadow state. */
1419 copy = kmemdup(info->shadow, sizeof(info->shadow),
1420 GFP_NOIO | __GFP_REPEAT | __GFP_HIGH);
1421 if (!copy)
1422 return -ENOMEM;
1423
1424 /* Stage 2: Set up free list. */
1425 memset(&info->shadow, 0, sizeof(info->shadow));
1426 for (i = 0; i < BLK_RING_SIZE; i++)
1427 info->shadow[i].req.u.rw.id = i+1;
1428 info->shadow_free = info->ring.req_prod_pvt;
1429 info->shadow[BLK_RING_SIZE-1].req.u.rw.id = 0x0fffffff;
1430
1431 rc = blkfront_setup_indirect(info);
1432 if (rc) {
1433 kfree(copy);
1434 return rc;
1435 }
1436
1437 segs = info->max_indirect_segments ? : BLKIF_MAX_SEGMENTS_PER_REQUEST;
1438 blk_queue_max_segments(info->rq, segs);
1439 bio_list_init(&bio_list);
1440 INIT_LIST_HEAD(&requests);
1441 for (i = 0; i < BLK_RING_SIZE; i++) {
1442 /* Not in use? */
1443 if (!copy[i].request)
1444 continue;
1445
1446 /*
1447 * Get the bios in the request so we can re-queue them.
1448 */
1449 if (copy[i].request->cmd_flags &
1450 (REQ_FLUSH | REQ_FUA | REQ_DISCARD | REQ_SECURE)) {
1451 /*
1452 * Flush operations don't contain bios, so
1453 * we need to requeue the whole request
1454 */
1455 list_add(&copy[i].request->queuelist, &requests);
1456 continue;
1457 }
1458 merge_bio.head = copy[i].request->bio;
1459 merge_bio.tail = copy[i].request->biotail;
1460 bio_list_merge(&bio_list, &merge_bio);
1461 copy[i].request->bio = NULL;
1462 blk_put_request(copy[i].request);
1463 }
1464
1465 kfree(copy);
1466
1467 /*
1468 * Empty the queue, this is important because we might have
1469 * requests in the queue with more segments than what we
1470 * can handle now.
1471 */
1472 spin_lock_irq(&info->io_lock);
1473 while ((req = blk_fetch_request(info->rq)) != NULL) {
1474 if (req->cmd_flags &
1475 (REQ_FLUSH | REQ_FUA | REQ_DISCARD | REQ_SECURE)) {
1476 list_add(&req->queuelist, &requests);
1477 continue;
1478 }
1479 merge_bio.head = req->bio;
1480 merge_bio.tail = req->biotail;
1481 bio_list_merge(&bio_list, &merge_bio);
1482 req->bio = NULL;
1483 if (req->cmd_flags & (REQ_FLUSH | REQ_FUA))
1484 pr_alert("diskcache flush request found!\n");
1485 __blk_put_request(info->rq, req);
1486 }
1487 spin_unlock_irq(&info->io_lock);
1488
1489 xenbus_switch_state(info->xbdev, XenbusStateConnected);
1490
1491 spin_lock_irq(&info->io_lock);
1492
1493 /* Now safe for us to use the shared ring */
1494 info->connected = BLKIF_STATE_CONNECTED;
1495
1496 /* Kick any other new requests queued since we resumed */
1497 kick_pending_request_queues(info);
1498
1499 list_for_each_entry_safe(req, n, &requests, queuelist) {
1500 /* Requeue pending requests (flush or discard) */
1501 list_del_init(&req->queuelist);
1502 BUG_ON(req->nr_phys_segments > segs);
1503 blk_requeue_request(info->rq, req);
1504 }
1505 spin_unlock_irq(&info->io_lock);
1506
1507 while ((bio = bio_list_pop(&bio_list)) != NULL) {
1508 /* Traverse the list of pending bios and re-queue them */
1509 if (bio_segments(bio) > segs) {
1510 /*
1511 * This bio has more segments than what we can
1512 * handle, we have to split it.
1513 */
1514 pending = (bio_segments(bio) + segs - 1) / segs;
1515 split_bio = kzalloc(sizeof(*split_bio), GFP_NOIO);
1516 BUG_ON(split_bio == NULL);
1517 atomic_set(&split_bio->pending, pending);
1518 split_bio->bio = bio;
1519 for (i = 0; i < pending; i++) {
1520 offset = (i * segs * PAGE_SIZE) >> 9;
1521 size = min((unsigned int)(segs * PAGE_SIZE) >> 9,
1522 (unsigned int)(bio->bi_size >> 9) - offset);
1523 cloned_bio = bio_clone(bio, GFP_NOIO);
1524 BUG_ON(cloned_bio == NULL);
1525 trim_bio(cloned_bio, offset, size);
1526 cloned_bio->bi_private = split_bio;
1527 cloned_bio->bi_end_io = split_bio_end;
1528 submit_bio(cloned_bio->bi_rw, cloned_bio);
1529 }
1530 /*
1531 * Now we have to wait for all those smaller bios to
1532 * end, so we can also end the "parent" bio.
1533 */
1534 continue;
1535 }
1536 /* We don't need to split this bio */
1537 submit_bio(bio->bi_rw, bio);
1538 }
1539
1540 return 0;
1541 }
1542
1543 /**
1544 * We are reconnecting to the backend, due to a suspend/resume, or a backend
1545 * driver restart. We tear down our blkif structure and recreate it, but
1546 * leave the device-layer structures intact so that this is transparent to the
1547 * rest of the kernel.
1548 */
1549 static int blkfront_resume(struct xenbus_device *dev)
1550 {
1551 struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1552 int err;
1553
1554 dev_dbg(&dev->dev, "blkfront_resume: %s\n", dev->nodename);
1555
1556 blkif_free(info, info->connected == BLKIF_STATE_CONNECTED);
1557
1558 err = talk_to_blkback(dev, info);
1559
1560 /*
1561 * We have to wait for the backend to switch to
1562 * connected state, since we want to read which
1563 * features it supports.
1564 */
1565
1566 return err;
1567 }
1568
1569 static void
1570 blkfront_closing(struct blkfront_info *info)
1571 {
1572 struct xenbus_device *xbdev = info->xbdev;
1573 struct block_device *bdev = NULL;
1574
1575 mutex_lock(&info->mutex);
1576
1577 if (xbdev->state == XenbusStateClosing) {
1578 mutex_unlock(&info->mutex);
1579 return;
1580 }
1581
1582 if (info->gd)
1583 bdev = bdget_disk(info->gd, 0);
1584
1585 mutex_unlock(&info->mutex);
1586
1587 if (!bdev) {
1588 xenbus_frontend_closed(xbdev);
1589 return;
1590 }
1591
1592 mutex_lock(&bdev->bd_mutex);
1593
1594 if (bdev->bd_openers) {
1595 xenbus_dev_error(xbdev, -EBUSY,
1596 "Device in use; refusing to close");
1597 xenbus_switch_state(xbdev, XenbusStateClosing);
1598 } else {
1599 xlvbd_release_gendisk(info);
1600 xenbus_frontend_closed(xbdev);
1601 }
1602
1603 mutex_unlock(&bdev->bd_mutex);
1604 bdput(bdev);
1605 }
1606
1607 static void blkfront_setup_discard(struct blkfront_info *info)
1608 {
1609 int err;
1610 char *type;
1611 unsigned int discard_granularity;
1612 unsigned int discard_alignment;
1613 unsigned int discard_secure;
1614
1615 type = xenbus_read(XBT_NIL, info->xbdev->otherend, "type", NULL);
1616 if (IS_ERR(type))
1617 return;
1618
1619 info->feature_secdiscard = 0;
1620 if (strncmp(type, "phy", 3) == 0) {
1621 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1622 "discard-granularity", "%u", &discard_granularity,
1623 "discard-alignment", "%u", &discard_alignment,
1624 NULL);
1625 if (!err) {
1626 info->feature_discard = 1;
1627 info->discard_granularity = discard_granularity;
1628 info->discard_alignment = discard_alignment;
1629 }
1630 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1631 "discard-secure", "%d", &discard_secure,
1632 NULL);
1633 if (!err)
1634 info->feature_secdiscard = discard_secure;
1635
1636 } else if (strncmp(type, "file", 4) == 0)
1637 info->feature_discard = 1;
1638
1639 kfree(type);
1640 }
1641
1642 static int blkfront_setup_indirect(struct blkfront_info *info)
1643 {
1644 unsigned int indirect_segments, segs;
1645 int err, i;
1646
1647 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1648 "feature-max-indirect-segments", "%u", &indirect_segments,
1649 NULL);
1650 if (err) {
1651 info->max_indirect_segments = 0;
1652 segs = BLKIF_MAX_SEGMENTS_PER_REQUEST;
1653 } else {
1654 info->max_indirect_segments = min(indirect_segments,
1655 xen_blkif_max_segments);
1656 segs = info->max_indirect_segments;
1657 }
1658
1659 err = fill_grant_buffer(info, (segs + INDIRECT_GREFS(segs)) * BLK_RING_SIZE);
1660 if (err)
1661 goto out_of_memory;
1662
1663 for (i = 0; i < BLK_RING_SIZE; i++) {
1664 info->shadow[i].grants_used = kzalloc(
1665 sizeof(info->shadow[i].grants_used[0]) * segs,
1666 GFP_NOIO);
1667 info->shadow[i].sg = kzalloc(sizeof(info->shadow[i].sg[0]) * segs, GFP_NOIO);
1668 if (info->max_indirect_segments)
1669 info->shadow[i].indirect_grants = kzalloc(
1670 sizeof(info->shadow[i].indirect_grants[0]) *
1671 INDIRECT_GREFS(segs),
1672 GFP_NOIO);
1673 if ((info->shadow[i].grants_used == NULL) ||
1674 (info->shadow[i].sg == NULL) ||
1675 (info->max_indirect_segments &&
1676 (info->shadow[i].indirect_grants == NULL)))
1677 goto out_of_memory;
1678 sg_init_table(info->shadow[i].sg, segs);
1679 }
1680
1681
1682 return 0;
1683
1684 out_of_memory:
1685 for (i = 0; i < BLK_RING_SIZE; i++) {
1686 kfree(info->shadow[i].grants_used);
1687 info->shadow[i].grants_used = NULL;
1688 kfree(info->shadow[i].sg);
1689 info->shadow[i].sg = NULL;
1690 kfree(info->shadow[i].indirect_grants);
1691 info->shadow[i].indirect_grants = NULL;
1692 }
1693 return -ENOMEM;
1694 }
1695
1696 /*
1697 * Invoked when the backend is finally 'ready' (and has told produced
1698 * the details about the physical device - #sectors, size, etc).
1699 */
1700 static void blkfront_connect(struct blkfront_info *info)
1701 {
1702 unsigned long long sectors;
1703 unsigned long sector_size;
1704 unsigned int physical_sector_size;
1705 unsigned int binfo;
1706 int err;
1707 int barrier, flush, discard, persistent;
1708
1709 switch (info->connected) {
1710 case BLKIF_STATE_CONNECTED:
1711 /*
1712 * Potentially, the back-end may be signalling
1713 * a capacity change; update the capacity.
1714 */
1715 err = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
1716 "sectors", "%Lu", &sectors);
1717 if (XENBUS_EXIST_ERR(err))
1718 return;
1719 printk(KERN_INFO "Setting capacity to %Lu\n",
1720 sectors);
1721 set_capacity(info->gd, sectors);
1722 revalidate_disk(info->gd);
1723
1724 return;
1725 case BLKIF_STATE_SUSPENDED:
1726 /*
1727 * If we are recovering from suspension, we need to wait
1728 * for the backend to announce it's features before
1729 * reconnecting, at least we need to know if the backend
1730 * supports indirect descriptors, and how many.
1731 */
1732 blkif_recover(info);
1733 return;
1734
1735 default:
1736 break;
1737 }
1738
1739 dev_dbg(&info->xbdev->dev, "%s:%s.\n",
1740 __func__, info->xbdev->otherend);
1741
1742 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1743 "sectors", "%llu", &sectors,
1744 "info", "%u", &binfo,
1745 "sector-size", "%lu", &sector_size,
1746 NULL);
1747 if (err) {
1748 xenbus_dev_fatal(info->xbdev, err,
1749 "reading backend fields at %s",
1750 info->xbdev->otherend);
1751 return;
1752 }
1753
1754 /*
1755 * physcial-sector-size is a newer field, so old backends may not
1756 * provide this. Assume physical sector size to be the same as
1757 * sector_size in that case.
1758 */
1759 err = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
1760 "physical-sector-size", "%u", &physical_sector_size);
1761 if (err != 1)
1762 physical_sector_size = sector_size;
1763
1764 info->feature_flush = 0;
1765 info->flush_op = 0;
1766
1767 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1768 "feature-barrier", "%d", &barrier,
1769 NULL);
1770
1771 /*
1772 * If there's no "feature-barrier" defined, then it means
1773 * we're dealing with a very old backend which writes
1774 * synchronously; nothing to do.
1775 *
1776 * If there are barriers, then we use flush.
1777 */
1778 if (!err && barrier) {
1779 info->feature_flush = REQ_FLUSH | REQ_FUA;
1780 info->flush_op = BLKIF_OP_WRITE_BARRIER;
1781 }
1782 /*
1783 * And if there is "feature-flush-cache" use that above
1784 * barriers.
1785 */
1786 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1787 "feature-flush-cache", "%d", &flush,
1788 NULL);
1789
1790 if (!err && flush) {
1791 info->feature_flush = REQ_FLUSH;
1792 info->flush_op = BLKIF_OP_FLUSH_DISKCACHE;
1793 }
1794
1795 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1796 "feature-discard", "%d", &discard,
1797 NULL);
1798
1799 if (!err && discard)
1800 blkfront_setup_discard(info);
1801
1802 err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1803 "feature-persistent", "%u", &persistent,
1804 NULL);
1805 if (err)
1806 info->feature_persistent = 0;
1807 else
1808 info->feature_persistent = persistent;
1809
1810 err = blkfront_setup_indirect(info);
1811 if (err) {
1812 xenbus_dev_fatal(info->xbdev, err, "setup_indirect at %s",
1813 info->xbdev->otherend);
1814 return;
1815 }
1816
1817 err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size,
1818 physical_sector_size);
1819 if (err) {
1820 xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s",
1821 info->xbdev->otherend);
1822 return;
1823 }
1824
1825 xenbus_switch_state(info->xbdev, XenbusStateConnected);
1826
1827 /* Kick pending requests. */
1828 spin_lock_irq(&info->io_lock);
1829 info->connected = BLKIF_STATE_CONNECTED;
1830 kick_pending_request_queues(info);
1831 spin_unlock_irq(&info->io_lock);
1832
1833 add_disk(info->gd);
1834
1835 info->is_ready = 1;
1836 }
1837
1838 /**
1839 * Callback received when the backend's state changes.
1840 */
1841 static void blkback_changed(struct xenbus_device *dev,
1842 enum xenbus_state backend_state)
1843 {
1844 struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1845
1846 dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state);
1847
1848 switch (backend_state) {
1849 case XenbusStateInitialising:
1850 case XenbusStateInitWait:
1851 case XenbusStateInitialised:
1852 case XenbusStateReconfiguring:
1853 case XenbusStateReconfigured:
1854 case XenbusStateUnknown:
1855 case XenbusStateClosed:
1856 break;
1857
1858 case XenbusStateConnected:
1859 blkfront_connect(info);
1860 break;
1861
1862 case XenbusStateClosing:
1863 blkfront_closing(info);
1864 break;
1865 }
1866 }
1867
1868 static int blkfront_remove(struct xenbus_device *xbdev)
1869 {
1870 struct blkfront_info *info = dev_get_drvdata(&xbdev->dev);
1871 struct block_device *bdev = NULL;
1872 struct gendisk *disk;
1873
1874 dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename);
1875
1876 blkif_free(info, 0);
1877
1878 mutex_lock(&info->mutex);
1879
1880 disk = info->gd;
1881 if (disk)
1882 bdev = bdget_disk(disk, 0);
1883
1884 info->xbdev = NULL;
1885 mutex_unlock(&info->mutex);
1886
1887 if (!bdev) {
1888 kfree(info);
1889 return 0;
1890 }
1891
1892 /*
1893 * The xbdev was removed before we reached the Closed
1894 * state. See if it's safe to remove the disk. If the bdev
1895 * isn't closed yet, we let release take care of it.
1896 */
1897
1898 mutex_lock(&bdev->bd_mutex);
1899 info = disk->private_data;
1900
1901 dev_warn(disk_to_dev(disk),
1902 "%s was hot-unplugged, %d stale handles\n",
1903 xbdev->nodename, bdev->bd_openers);
1904
1905 if (info && !bdev->bd_openers) {
1906 xlvbd_release_gendisk(info);
1907 disk->private_data = NULL;
1908 kfree(info);
1909 }
1910
1911 mutex_unlock(&bdev->bd_mutex);
1912 bdput(bdev);
1913
1914 return 0;
1915 }
1916
1917 static int blkfront_is_ready(struct xenbus_device *dev)
1918 {
1919 struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1920
1921 return info->is_ready && info->xbdev;
1922 }
1923
1924 static int blkif_open(struct block_device *bdev, fmode_t mode)
1925 {
1926 struct gendisk *disk = bdev->bd_disk;
1927 struct blkfront_info *info;
1928 int err = 0;
1929
1930 mutex_lock(&blkfront_mutex);
1931
1932 info = disk->private_data;
1933 if (!info) {
1934 /* xbdev gone */
1935 err = -ERESTARTSYS;
1936 goto out;
1937 }
1938
1939 mutex_lock(&info->mutex);
1940
1941 if (!info->gd)
1942 /* xbdev is closed */
1943 err = -ERESTARTSYS;
1944
1945 mutex_unlock(&info->mutex);
1946
1947 out:
1948 mutex_unlock(&blkfront_mutex);
1949 return err;
1950 }
1951
1952 static void blkif_release(struct gendisk *disk, fmode_t mode)
1953 {
1954 struct blkfront_info *info = disk->private_data;
1955 struct block_device *bdev;
1956 struct xenbus_device *xbdev;
1957
1958 mutex_lock(&blkfront_mutex);
1959
1960 bdev = bdget_disk(disk, 0);
1961
1962 if (bdev->bd_openers)
1963 goto out;
1964
1965 /*
1966 * Check if we have been instructed to close. We will have
1967 * deferred this request, because the bdev was still open.
1968 */
1969
1970 mutex_lock(&info->mutex);
1971 xbdev = info->xbdev;
1972
1973 if (xbdev && xbdev->state == XenbusStateClosing) {
1974 /* pending switch to state closed */
1975 dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
1976 xlvbd_release_gendisk(info);
1977 xenbus_frontend_closed(info->xbdev);
1978 }
1979
1980 mutex_unlock(&info->mutex);
1981
1982 if (!xbdev) {
1983 /* sudden device removal */
1984 dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
1985 xlvbd_release_gendisk(info);
1986 disk->private_data = NULL;
1987 kfree(info);
1988 }
1989
1990 out:
1991 bdput(bdev);
1992 mutex_unlock(&blkfront_mutex);
1993 }
1994
1995 static const struct block_device_operations xlvbd_block_fops =
1996 {
1997 .owner = THIS_MODULE,
1998 .open = blkif_open,
1999 .release = blkif_release,
2000 .getgeo = blkif_getgeo,
2001 .ioctl = blkif_ioctl,
2002 };
2003
2004
2005 static const struct xenbus_device_id blkfront_ids[] = {
2006 { "vbd" },
2007 { "" }
2008 };
2009
2010 static DEFINE_XENBUS_DRIVER(blkfront, ,
2011 .probe = blkfront_probe,
2012 .remove = blkfront_remove,
2013 .resume = blkfront_resume,
2014 .otherend_changed = blkback_changed,
2015 .is_ready = blkfront_is_ready,
2016 );
2017
2018 static int __init xlblk_init(void)
2019 {
2020 int ret;
2021
2022 if (!xen_domain())
2023 return -ENODEV;
2024
2025 if (xen_hvm_domain() && !xen_platform_pci_unplug)
2026 return -ENODEV;
2027
2028 if (register_blkdev(XENVBD_MAJOR, DEV_NAME)) {
2029 printk(KERN_WARNING "xen_blk: can't get major %d with name %s\n",
2030 XENVBD_MAJOR, DEV_NAME);
2031 return -ENODEV;
2032 }
2033
2034 ret = xenbus_register_frontend(&blkfront_driver);
2035 if (ret) {
2036 unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
2037 return ret;
2038 }
2039
2040 return 0;
2041 }
2042 module_init(xlblk_init);
2043
2044
2045 static void __exit xlblk_exit(void)
2046 {
2047 xenbus_unregister_driver(&blkfront_driver);
2048 unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
2049 kfree(minors);
2050 }
2051 module_exit(xlblk_exit);
2052
2053 MODULE_DESCRIPTION("Xen virtual block device frontend");
2054 MODULE_LICENSE("GPL");
2055 MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
2056 MODULE_ALIAS("xen:vbd");
2057 MODULE_ALIAS("xenblk");
Linus' kernel tree