xen-blkfront.c 57.4 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
/*
 * blkfront.c
 *
 * XenLinux virtual block device driver.
 *
 * Copyright (c) 2003-2004, Keir Fraser & Steve Hand
 * Modifications by Mark A. Williamson are (c) Intel Research Cambridge
 * Copyright (c) 2004, Christian Limpach
 * Copyright (c) 2004, Andrew Warfield
 * Copyright (c) 2005, Christopher Clark
 * Copyright (c) 2005, XenSource Ltd
 *
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License version 2
 * as published by the Free Software Foundation; or, when distributed
 * separately from the Linux kernel or incorporated into other
 * software packages, subject to the following license:
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this source file (the "Software"), to deal in the Software without
 * restriction, including without limitation the rights to use, copy, modify,
 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
 * and to permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

#include <linux/interrupt.h>
#include <linux/blkdev.h>
40
#include <linux/blk-mq.h>
41
#include <linux/hdreg.h>
42
#include <linux/cdrom.h>
43
#include <linux/module.h>
44
#include <linux/slab.h>
45
#include <linux/mutex.h>
46
#include <linux/scatterlist.h>
47
#include <linux/bitmap.h>
48
#include <linux/list.h>
49

50
#include <xen/xen.h>
51 52 53 54
#include <xen/xenbus.h>
#include <xen/grant_table.h>
#include <xen/events.h>
#include <xen/page.h>
55
#include <xen/platform_pci.h>
56 57 58

#include <xen/interface/grant_table.h>
#include <xen/interface/io/blkif.h>
59
#include <xen/interface/io/protocols.h>
60 61 62 63 64 65 66 67 68

#include <asm/xen/hypervisor.h>

enum blkif_state {
	BLKIF_STATE_DISCONNECTED,
	BLKIF_STATE_CONNECTED,
	BLKIF_STATE_SUSPENDED,
};

69 70 71
struct grant {
	grant_ref_t gref;
	unsigned long pfn;
72
	struct list_head node;
73 74
};

75 76
struct blk_shadow {
	struct blkif_request req;
77
	struct request *request;
78 79
	struct grant **grants_used;
	struct grant **indirect_grants;
80
	struct scatterlist *sg;
81 82 83 84 85 86
};

struct split_bio {
	struct bio *bio;
	atomic_t pending;
	int err;
87 88
};

89
static DEFINE_MUTEX(blkfront_mutex);
90
static const struct block_device_operations xlvbd_block_fops;
91

92 93 94 95 96 97 98
/*
 * Maximum number of segments in indirect requests, the actual value used by
 * the frontend driver is the minimum of this value and the value provided
 * by the backend driver.
 */

static unsigned int xen_blkif_max_segments = 32;
99 100
module_param_named(max, xen_blkif_max_segments, int, S_IRUGO);
MODULE_PARM_DESC(max, "Maximum amount of segments in indirect requests (default is 32)");
101

102 103 104 105 106 107 108 109 110 111 112 113 114 115 116
/*
 * Maximum order of pages to be used for the shared ring between front and
 * backend, 4KB page granularity is used.
 */
static unsigned int xen_blkif_max_ring_order;
module_param_named(max_ring_page_order, xen_blkif_max_ring_order, int, S_IRUGO);
MODULE_PARM_DESC(max_ring_page_order, "Maximum order of pages to be used for the shared ring");

#define BLK_RING_SIZE(info) __CONST_RING_SIZE(blkif, PAGE_SIZE * (info)->nr_ring_pages)
#define BLK_MAX_RING_SIZE __CONST_RING_SIZE(blkif, PAGE_SIZE * XENBUS_MAX_RING_PAGES)
/*
 * ring-ref%i i=(-1UL) would take 11 characters + 'ring-ref' is 8, so 19
 * characters are enough. Define to 20 to keep consist with backend.
 */
#define RINGREF_NAME_LEN (20)
117 118 119 120 121 122 123 124

/*
 * We have one of these per vbd, whether ide, scsi or 'other'.  They
 * hang in private_data off the gendisk structure. We may end up
 * putting all kinds of interesting stuff here :-)
 */
struct blkfront_info
{
125
	spinlock_t io_lock;
126
	struct mutex mutex;
127 128 129 130 131
	struct xenbus_device *xbdev;
	struct gendisk *gd;
	int vdevice;
	blkif_vdev_t handle;
	enum blkif_state connected;
132 133
	int ring_ref[XENBUS_MAX_RING_PAGES];
	unsigned int nr_ring_pages;
134 135 136 137 138
	struct blkif_front_ring ring;
	unsigned int evtchn, irq;
	struct request_queue *rq;
	struct work_struct work;
	struct gnttab_free_callback callback;
139
	struct blk_shadow shadow[BLK_MAX_RING_SIZE];
140 141
	struct list_head grants;
	struct list_head indirect_pages;
142
	unsigned int persistent_gnts_c;
143
	unsigned long shadow_free;
144
	unsigned int feature_flush;
145 146
	unsigned int feature_discard:1;
	unsigned int feature_secdiscard:1;
147 148
	unsigned int discard_granularity;
	unsigned int discard_alignment;
149
	unsigned int feature_persistent:1;
150
	unsigned int max_indirect_segments;
151
	int is_ready;
152
	struct blk_mq_tag_set tag_set;
153 154
};

155 156 157 158
static unsigned int nr_minors;
static unsigned long *minors;
static DEFINE_SPINLOCK(minor_lock);

159 160 161
#define GRANT_INVALID_REF	0

#define PARTS_PER_DISK		16
162
#define PARTS_PER_EXT_DISK      256
163 164 165 166

#define BLKIF_MAJOR(dev) ((dev)>>8)
#define BLKIF_MINOR(dev) ((dev) & 0xff)

167 168 169 170
#define EXT_SHIFT 28
#define EXTENDED (1<<EXT_SHIFT)
#define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
#define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
171 172
#define EMULATED_HD_DISK_MINOR_OFFSET (0)
#define EMULATED_HD_DISK_NAME_OFFSET (EMULATED_HD_DISK_MINOR_OFFSET / 256)
173 174
#define EMULATED_SD_DISK_MINOR_OFFSET (0)
#define EMULATED_SD_DISK_NAME_OFFSET (EMULATED_SD_DISK_MINOR_OFFSET / 256)
175

176
#define DEV_NAME	"xvd"	/* name in /dev */
177

178
#define SEGS_PER_INDIRECT_FRAME \
179
	(PAGE_SIZE/sizeof(struct blkif_request_segment))
180 181 182 183
#define INDIRECT_GREFS(_segs) \
	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)

static int blkfront_setup_indirect(struct blkfront_info *info);
184
static int blkfront_gather_backend_features(struct blkfront_info *info);
185

186 187 188
static int get_id_from_freelist(struct blkfront_info *info)
{
	unsigned long free = info->shadow_free;
189
	BUG_ON(free >= BLK_RING_SIZE(info));
190 191
	info->shadow_free = info->shadow[free].req.u.rw.id;
	info->shadow[free].req.u.rw.id = 0x0fffffee; /* debug */
192 193 194
	return free;
}

195
static int add_id_to_freelist(struct blkfront_info *info,
196 197
			       unsigned long id)
{
198 199 200 201
	if (info->shadow[id].req.u.rw.id != id)
		return -EINVAL;
	if (info->shadow[id].request == NULL)
		return -EINVAL;
202
	info->shadow[id].req.u.rw.id  = info->shadow_free;
203
	info->shadow[id].request = NULL;
204
	info->shadow_free = id;
205
	return 0;
206 207
}

208 209 210 211 212 213 214 215 216 217 218
static int fill_grant_buffer(struct blkfront_info *info, int num)
{
	struct page *granted_page;
	struct grant *gnt_list_entry, *n;
	int i = 0;

	while(i < num) {
		gnt_list_entry = kzalloc(sizeof(struct grant), GFP_NOIO);
		if (!gnt_list_entry)
			goto out_of_memory;

219 220 221 222 223 224 225
		if (info->feature_persistent) {
			granted_page = alloc_page(GFP_NOIO);
			if (!granted_page) {
				kfree(gnt_list_entry);
				goto out_of_memory;
			}
			gnt_list_entry->pfn = page_to_pfn(granted_page);
226 227 228
		}

		gnt_list_entry->gref = GRANT_INVALID_REF;
229
		list_add(&gnt_list_entry->node, &info->grants);
230 231 232 233 234 235 236
		i++;
	}

	return 0;

out_of_memory:
	list_for_each_entry_safe(gnt_list_entry, n,
237
	                         &info->grants, node) {
238
		list_del(&gnt_list_entry->node);
239 240
		if (info->feature_persistent)
			__free_page(pfn_to_page(gnt_list_entry->pfn));
241 242 243 244 245 246 247 248
		kfree(gnt_list_entry);
		i--;
	}
	BUG_ON(i != 0);
	return -ENOMEM;
}

static struct grant *get_grant(grant_ref_t *gref_head,
249
                               unsigned long pfn,
250 251 252 253 254
                               struct blkfront_info *info)
{
	struct grant *gnt_list_entry;
	unsigned long buffer_mfn;

255 256
	BUG_ON(list_empty(&info->grants));
	gnt_list_entry = list_first_entry(&info->grants, struct grant,
257 258 259 260 261 262 263 264 265 266 267
	                                  node);
	list_del(&gnt_list_entry->node);

	if (gnt_list_entry->gref != GRANT_INVALID_REF) {
		info->persistent_gnts_c--;
		return gnt_list_entry;
	}

	/* Assign a gref to this page */
	gnt_list_entry->gref = gnttab_claim_grant_reference(gref_head);
	BUG_ON(gnt_list_entry->gref == -ENOSPC);
268 269 270 271
	if (!info->feature_persistent) {
		BUG_ON(!pfn);
		gnt_list_entry->pfn = pfn;
	}
272 273 274 275 276 277 278
	buffer_mfn = pfn_to_mfn(gnt_list_entry->pfn);
	gnttab_grant_foreign_access_ref(gnt_list_entry->gref,
	                                info->xbdev->otherend_id,
	                                buffer_mfn, 0);
	return gnt_list_entry;
}

279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295
static const char *op_name(int op)
{
	static const char *const names[] = {
		[BLKIF_OP_READ] = "read",
		[BLKIF_OP_WRITE] = "write",
		[BLKIF_OP_WRITE_BARRIER] = "barrier",
		[BLKIF_OP_FLUSH_DISKCACHE] = "flush",
		[BLKIF_OP_DISCARD] = "discard" };

	if (op < 0 || op >= ARRAY_SIZE(names))
		return "unknown";

	if (!names[op])
		return "reserved";

	return names[op];
}
296 297 298 299 300 301 302 303
static int xlbd_reserve_minors(unsigned int minor, unsigned int nr)
{
	unsigned int end = minor + nr;
	int rc;

	if (end > nr_minors) {
		unsigned long *bitmap, *old;

304
		bitmap = kcalloc(BITS_TO_LONGS(end), sizeof(*bitmap),
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323
				 GFP_KERNEL);
		if (bitmap == NULL)
			return -ENOMEM;

		spin_lock(&minor_lock);
		if (end > nr_minors) {
			old = minors;
			memcpy(bitmap, minors,
			       BITS_TO_LONGS(nr_minors) * sizeof(*bitmap));
			minors = bitmap;
			nr_minors = BITS_TO_LONGS(end) * BITS_PER_LONG;
		} else
			old = bitmap;
		spin_unlock(&minor_lock);
		kfree(old);
	}

	spin_lock(&minor_lock);
	if (find_next_bit(minors, end, minor) >= end) {
324
		bitmap_set(minors, minor, nr);
325 326 327 328 329 330 331 332 333 334 335 336 337 338
		rc = 0;
	} else
		rc = -EBUSY;
	spin_unlock(&minor_lock);

	return rc;
}

static void xlbd_release_minors(unsigned int minor, unsigned int nr)
{
	unsigned int end = minor + nr;

	BUG_ON(end > nr_minors);
	spin_lock(&minor_lock);
339
	bitmap_clear(minors,  minor, nr);
340 341 342
	spin_unlock(&minor_lock);
}

343 344 345 346 347 348
static void blkif_restart_queue_callback(void *arg)
{
	struct blkfront_info *info = (struct blkfront_info *)arg;
	schedule_work(&info->work);
}

349
static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
350 351 352 353 354 355 356 357 358 359 360 361 362 363 364
{
	/* We don't have real geometry info, but let's at least return
	   values consistent with the size of the device */
	sector_t nsect = get_capacity(bd->bd_disk);
	sector_t cylinders = nsect;

	hg->heads = 0xff;
	hg->sectors = 0x3f;
	sector_div(cylinders, hg->heads * hg->sectors);
	hg->cylinders = cylinders;
	if ((sector_t)(hg->cylinders + 1) * hg->heads * hg->sectors < nsect)
		hg->cylinders = 0xffff;
	return 0;
}

Al Viro's avatar
Al Viro committed
365
static int blkif_ioctl(struct block_device *bdev, fmode_t mode,
366
		       unsigned command, unsigned long argument)
367
{
Al Viro's avatar
Al Viro committed
368
	struct blkfront_info *info = bdev->bd_disk->private_data;
369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397
	int i;

	dev_dbg(&info->xbdev->dev, "command: 0x%x, argument: 0x%lx\n",
		command, (long)argument);

	switch (command) {
	case CDROMMULTISESSION:
		dev_dbg(&info->xbdev->dev, "FIXME: support multisession CDs later\n");
		for (i = 0; i < sizeof(struct cdrom_multisession); i++)
			if (put_user(0, (char __user *)(argument + i)))
				return -EFAULT;
		return 0;

	case CDROM_GET_CAPABILITY: {
		struct gendisk *gd = info->gd;
		if (gd->flags & GENHD_FL_CD)
			return 0;
		return -EINVAL;
	}

	default:
		/*printk(KERN_ALERT "ioctl %08x not supported by Xen blkdev\n",
		  command);*/
		return -EINVAL; /* same return as native Linux */
	}

	return 0;
}

398
/*
399
 * Generate a Xen blkfront IO request from a blk layer request.  Reads
400
 * and writes are handled as expected.
401
 *
402
 * @req: a request struct
403 404 405 406 407 408 409
 */
static int blkif_queue_request(struct request *req)
{
	struct blkfront_info *info = req->rq_disk->private_data;
	struct blkif_request *ring_req;
	unsigned long id;
	unsigned int fsect, lsect;
410
	int i, ref, n;
411
	struct blkif_request_segment *segments = NULL;
412 413 414 415 416 417 418

	/*
	 * Used to store if we are able to queue the request by just using
	 * existing persistent grants, or if we have to get new grants,
	 * as there are not sufficiently many free.
	 */
	bool new_persistent_gnts;
419
	grant_ref_t gref_head;
420
	struct grant *gnt_list_entry = NULL;
421
	struct scatterlist *sg;
422
	int nseg, max_grefs;
423 424 425 426

	if (unlikely(info->connected != BLKIF_STATE_CONNECTED))
		return 1;

427 428 429 430 431 432 433
	max_grefs = req->nr_phys_segments;
	if (max_grefs > BLKIF_MAX_SEGMENTS_PER_REQUEST)
		/*
		 * If we are using indirect segments we need to account
		 * for the indirect grefs used in the request.
		 */
		max_grefs += INDIRECT_GREFS(req->nr_phys_segments);
434 435 436

	/* Check if we have enough grants to allocate a requests */
	if (info->persistent_gnts_c < max_grefs) {
437 438
		new_persistent_gnts = 1;
		if (gnttab_alloc_grant_references(
439
		    max_grefs - info->persistent_gnts_c,
440 441 442 443 444
		    &gref_head) < 0) {
			gnttab_request_free_callback(
				&info->callback,
				blkif_restart_queue_callback,
				info,
445
				max_grefs);
446 447 448 449
			return 1;
		}
	} else
		new_persistent_gnts = 0;
450 451 452 453

	/* Fill out a communications ring structure. */
	ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
	id = get_id_from_freelist(info);
454
	info->shadow[id].request = req;
455

456
	if (unlikely(req->cmd_flags & (REQ_DISCARD | REQ_SECURE))) {
457 458
		ring_req->operation = BLKIF_OP_DISCARD;
		ring_req->u.discard.nr_sectors = blk_rq_sectors(req);
459 460
		ring_req->u.discard.id = id;
		ring_req->u.discard.sector_number = (blkif_sector_t)blk_rq_pos(req);
461 462 463 464
		if ((req->cmd_flags & REQ_SECURE) && info->feature_secdiscard)
			ring_req->u.discard.flag = BLKIF_DISCARD_SECURE;
		else
			ring_req->u.discard.flag = 0;
465
	} else {
466 467 468 469
		BUG_ON(info->max_indirect_segments == 0 &&
		       req->nr_phys_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST);
		BUG_ON(info->max_indirect_segments &&
		       req->nr_phys_segments > info->max_indirect_segments);
470
		nseg = blk_rq_map_sg(req->q, req, info->shadow[id].sg);
471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
		ring_req->u.rw.id = id;
		if (nseg > BLKIF_MAX_SEGMENTS_PER_REQUEST) {
			/*
			 * The indirect operation can only be a BLKIF_OP_READ or
			 * BLKIF_OP_WRITE
			 */
			BUG_ON(req->cmd_flags & (REQ_FLUSH | REQ_FUA));
			ring_req->operation = BLKIF_OP_INDIRECT;
			ring_req->u.indirect.indirect_op = rq_data_dir(req) ?
				BLKIF_OP_WRITE : BLKIF_OP_READ;
			ring_req->u.indirect.sector_number = (blkif_sector_t)blk_rq_pos(req);
			ring_req->u.indirect.handle = info->handle;
			ring_req->u.indirect.nr_segments = nseg;
		} else {
			ring_req->u.rw.sector_number = (blkif_sector_t)blk_rq_pos(req);
			ring_req->u.rw.handle = info->handle;
			ring_req->operation = rq_data_dir(req) ?
				BLKIF_OP_WRITE : BLKIF_OP_READ;
			if (req->cmd_flags & (REQ_FLUSH | REQ_FUA)) {
				/*
				 * Ideally we can do an unordered flush-to-disk. In case the
				 * backend onlysupports barriers, use that. A barrier request
				 * a superset of FUA, so we can implement it the same
				 * way.  (It's also a FLUSH+FUA, since it is
				 * guaranteed ordered WRT previous writes.)
				 */
497 498 499 500 501 502 503 504 505 506 507 508 509
				switch (info->feature_flush &
					((REQ_FLUSH|REQ_FUA))) {
				case REQ_FLUSH|REQ_FUA:
					ring_req->operation =
						BLKIF_OP_WRITE_BARRIER;
					break;
				case REQ_FLUSH:
					ring_req->operation =
						BLKIF_OP_FLUSH_DISKCACHE;
					break;
				default:
					ring_req->operation = 0;
				}
510 511 512
			}
			ring_req->u.rw.nr_segments = nseg;
		}
513
		for_each_sg(info->shadow[id].sg, sg, nseg, i) {
514 515
			fsect = sg->offset >> 9;
			lsect = fsect + (sg->length >> 9) - 1;
516

517 518
			if ((ring_req->operation == BLKIF_OP_INDIRECT) &&
			    (i % SEGS_PER_INDIRECT_FRAME == 0)) {
519
				unsigned long uninitialized_var(pfn);
520

521 522 523 524
				if (segments)
					kunmap_atomic(segments);

				n = i / SEGS_PER_INDIRECT_FRAME;
525 526 527 528 529 530 531 532 533 534 535
				if (!info->feature_persistent) {
					struct page *indirect_page;

					/* Fetch a pre-allocated page to use for indirect grefs */
					BUG_ON(list_empty(&info->indirect_pages));
					indirect_page = list_first_entry(&info->indirect_pages,
					                                 struct page, lru);
					list_del(&indirect_page->lru);
					pfn = page_to_pfn(indirect_page);
				}
				gnt_list_entry = get_grant(&gref_head, pfn, info);
536 537 538 539 540
				info->shadow[id].indirect_grants[n] = gnt_list_entry;
				segments = kmap_atomic(pfn_to_page(gnt_list_entry->pfn));
				ring_req->u.indirect.indirect_grefs[n] = gnt_list_entry->gref;
			}

541
			gnt_list_entry = get_grant(&gref_head, page_to_pfn(sg_page(sg)), info);
542
			ref = gnt_list_entry->gref;
543 544 545

			info->shadow[id].grants_used[i] = gnt_list_entry;

546
			if (rq_data_dir(req) && info->feature_persistent) {
547 548 549 550 551
				char *bvec_data;
				void *shared_data;

				BUG_ON(sg->offset + sg->length > PAGE_SIZE);

552
				shared_data = kmap_atomic(pfn_to_page(gnt_list_entry->pfn));
553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570
				bvec_data = kmap_atomic(sg_page(sg));

				/*
				 * this does not wipe data stored outside the
				 * range sg->offset..sg->offset+sg->length.
				 * Therefore, blkback *could* see data from
				 * previous requests. This is OK as long as
				 * persistent grants are shared with just one
				 * domain. It may need refactoring if this
				 * changes
				 */
				memcpy(shared_data + sg->offset,
				       bvec_data   + sg->offset,
				       sg->length);

				kunmap_atomic(bvec_data);
				kunmap_atomic(shared_data);
			}
571 572 573 574 575 576 577 578 579
			if (ring_req->operation != BLKIF_OP_INDIRECT) {
				ring_req->u.rw.seg[i] =
						(struct blkif_request_segment) {
							.gref       = ref,
							.first_sect = fsect,
							.last_sect  = lsect };
			} else {
				n = i % SEGS_PER_INDIRECT_FRAME;
				segments[n] =
580
					(struct blkif_request_segment) {
581 582 583 584
							.gref       = ref,
							.first_sect = fsect,
							.last_sect  = lsect };
			}
585
		}
586 587
		if (segments)
			kunmap_atomic(segments);
588 589 590 591 592 593 594
	}

	info->ring.req_prod_pvt++;

	/* Keep a private copy so we can reissue requests when recovering. */
	info->shadow[id].req = *ring_req;

595 596
	if (new_persistent_gnts)
		gnttab_free_grant_references(gref_head);
597 598 599 600 601 602 603 604 605 606 607 608 609 610 611

	return 0;
}


static inline void flush_requests(struct blkfront_info *info)
{
	int notify;

	RING_PUSH_REQUESTS_AND_CHECK_NOTIFY(&info->ring, notify);

	if (notify)
		notify_remote_via_irq(info->irq);
}

612 613
static inline bool blkif_request_flush_invalid(struct request *req,
					       struct blkfront_info *info)
614 615
{
	return ((req->cmd_type != REQ_TYPE_FS) ||
616 617 618 619
		((req->cmd_flags & REQ_FLUSH) &&
		 !(info->feature_flush & REQ_FLUSH)) ||
		((req->cmd_flags & REQ_FUA) &&
		 !(info->feature_flush & REQ_FUA)));
620 621
}

622 623
static int blkif_queue_rq(struct blk_mq_hw_ctx *hctx,
			   const struct blk_mq_queue_data *qd)
624
{
625
	struct blkfront_info *info = qd->rq->rq_disk->private_data;
626

627 628 629 630
	blk_mq_start_request(qd->rq);
	spin_lock_irq(&info->io_lock);
	if (RING_FULL(&info->ring))
		goto out_busy;
631

632 633
	if (blkif_request_flush_invalid(qd->rq, info))
		goto out_err;
634

635 636
	if (blkif_queue_request(qd->rq))
		goto out_busy;
637

638 639 640
	flush_requests(info);
	spin_unlock_irq(&info->io_lock);
	return BLK_MQ_RQ_QUEUE_OK;
641

642 643 644
out_err:
	spin_unlock_irq(&info->io_lock);
	return BLK_MQ_RQ_QUEUE_ERROR;
645

646 647 648 649
out_busy:
	spin_unlock_irq(&info->io_lock);
	blk_mq_stop_hw_queue(hctx);
	return BLK_MQ_RQ_QUEUE_BUSY;
650 651
}

652 653 654 655 656
static struct blk_mq_ops blkfront_mq_ops = {
	.queue_rq = blkif_queue_rq,
	.map_queue = blk_mq_map_queue,
};

657
static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size,
658
				unsigned int physical_sector_size,
659
				unsigned int segments)
660
{
661
	struct request_queue *rq;
662
	struct blkfront_info *info = gd->private_data;
663

664 665 666 667 668 669 670 671 672 673
	memset(&info->tag_set, 0, sizeof(info->tag_set));
	info->tag_set.ops = &blkfront_mq_ops;
	info->tag_set.nr_hw_queues = 1;
	info->tag_set.queue_depth =  BLK_RING_SIZE(info);
	info->tag_set.numa_node = NUMA_NO_NODE;
	info->tag_set.flags = BLK_MQ_F_SHOULD_MERGE | BLK_MQ_F_SG_MERGE;
	info->tag_set.cmd_size = 0;
	info->tag_set.driver_data = info;

	if (blk_mq_alloc_tag_set(&info->tag_set))
674
		return -1;
675 676 677 678 679
	rq = blk_mq_init_queue(&info->tag_set);
	if (IS_ERR(rq)) {
		blk_mq_free_tag_set(&info->tag_set);
		return -1;
	}
680

681
	queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq);
682

683 684 685 686 687
	if (info->feature_discard) {
		queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, rq);
		blk_queue_max_discard_sectors(rq, get_capacity(gd));
		rq->limits.discard_granularity = info->discard_granularity;
		rq->limits.discard_alignment = info->discard_alignment;
688 689
		if (info->feature_secdiscard)
			queue_flag_set_unlocked(QUEUE_FLAG_SECDISCARD, rq);
690 691
	}

692
	/* Hard sector size and max sectors impersonate the equiv. hardware. */
693
	blk_queue_logical_block_size(rq, sector_size);
694
	blk_queue_physical_block_size(rq, physical_sector_size);
695
	blk_queue_max_hw_sectors(rq, (segments * PAGE_SIZE) / 512);
696 697 698 699 700 701

	/* Each segment in a request is up to an aligned page in size. */
	blk_queue_segment_boundary(rq, PAGE_SIZE - 1);
	blk_queue_max_segment_size(rq, PAGE_SIZE);

	/* Ensure a merged request will fit in a single I/O ring slot. */
702
	blk_queue_max_segments(rq, segments);
703 704 705 706

	/* Make sure buffer addresses are sector-aligned. */
	blk_queue_dma_alignment(rq, 511);

707 708 709
	/* Make sure we don't use bounce buffers. */
	blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);

710 711 712 713 714
	gd->queue = rq;

	return 0;
}

715 716 717 718 719 720 721 722 723 724 725
static const char *flush_info(unsigned int feature_flush)
{
	switch (feature_flush & ((REQ_FLUSH | REQ_FUA))) {
	case REQ_FLUSH|REQ_FUA:
		return "barrier: enabled;";
	case REQ_FLUSH:
		return "flush diskcache: enabled;";
	default:
		return "barrier or flush: disabled;";
	}
}
726

727
static void xlvbd_flush(struct blkfront_info *info)
728
{
729
	blk_queue_flush(info->rq, info->feature_flush);
730 731 732 733 734
	pr_info("blkfront: %s: %s %s %s %s %s\n",
		info->gd->disk_name, flush_info(info->feature_flush),
		"persistent grants:", info->feature_persistent ?
		"enabled;" : "disabled;", "indirect descriptors:",
		info->max_indirect_segments ? "enabled;" : "disabled;");
735 736
}

737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795
static int xen_translate_vdev(int vdevice, int *minor, unsigned int *offset)
{
	int major;
	major = BLKIF_MAJOR(vdevice);
	*minor = BLKIF_MINOR(vdevice);
	switch (major) {
		case XEN_IDE0_MAJOR:
			*offset = (*minor / 64) + EMULATED_HD_DISK_NAME_OFFSET;
			*minor = ((*minor / 64) * PARTS_PER_DISK) +
				EMULATED_HD_DISK_MINOR_OFFSET;
			break;
		case XEN_IDE1_MAJOR:
			*offset = (*minor / 64) + 2 + EMULATED_HD_DISK_NAME_OFFSET;
			*minor = (((*minor / 64) + 2) * PARTS_PER_DISK) +
				EMULATED_HD_DISK_MINOR_OFFSET;
			break;
		case XEN_SCSI_DISK0_MAJOR:
			*offset = (*minor / PARTS_PER_DISK) + EMULATED_SD_DISK_NAME_OFFSET;
			*minor = *minor + EMULATED_SD_DISK_MINOR_OFFSET;
			break;
		case XEN_SCSI_DISK1_MAJOR:
		case XEN_SCSI_DISK2_MAJOR:
		case XEN_SCSI_DISK3_MAJOR:
		case XEN_SCSI_DISK4_MAJOR:
		case XEN_SCSI_DISK5_MAJOR:
		case XEN_SCSI_DISK6_MAJOR:
		case XEN_SCSI_DISK7_MAJOR:
			*offset = (*minor / PARTS_PER_DISK) + 
				((major - XEN_SCSI_DISK1_MAJOR + 1) * 16) +
				EMULATED_SD_DISK_NAME_OFFSET;
			*minor = *minor +
				((major - XEN_SCSI_DISK1_MAJOR + 1) * 16 * PARTS_PER_DISK) +
				EMULATED_SD_DISK_MINOR_OFFSET;
			break;
		case XEN_SCSI_DISK8_MAJOR:
		case XEN_SCSI_DISK9_MAJOR:
		case XEN_SCSI_DISK10_MAJOR:
		case XEN_SCSI_DISK11_MAJOR:
		case XEN_SCSI_DISK12_MAJOR:
		case XEN_SCSI_DISK13_MAJOR:
		case XEN_SCSI_DISK14_MAJOR:
		case XEN_SCSI_DISK15_MAJOR:
			*offset = (*minor / PARTS_PER_DISK) + 
				((major - XEN_SCSI_DISK8_MAJOR + 8) * 16) +
				EMULATED_SD_DISK_NAME_OFFSET;
			*minor = *minor +
				((major - XEN_SCSI_DISK8_MAJOR + 8) * 16 * PARTS_PER_DISK) +
				EMULATED_SD_DISK_MINOR_OFFSET;
			break;
		case XENVBD_MAJOR:
			*offset = *minor / PARTS_PER_DISK;
			break;
		default:
			printk(KERN_WARNING "blkfront: your disk configuration is "
					"incorrect, please use an xvd device instead\n");
			return -ENODEV;
	}
	return 0;
}
796

797 798 799 800 801 802 803 804
static char *encode_disk_name(char *ptr, unsigned int n)
{
	if (n >= 26)
		ptr = encode_disk_name(ptr, n / 26 - 1);
	*ptr = 'a' + n % 26;
	return ptr + 1;
}

805 806
static int xlvbd_alloc_gendisk(blkif_sector_t capacity,
			       struct blkfront_info *info,
807 808
			       u16 vdisk_info, u16 sector_size,
			       unsigned int physical_sector_size)
809 810 811
{
	struct gendisk *gd;
	int nr_minors = 1;
812
	int err;
813 814 815
	unsigned int offset;
	int minor;
	int nr_parts;
816
	char *ptr;
817 818 819 820

	BUG_ON(info->gd != NULL);
	BUG_ON(info->rq != NULL);

821 822 823 824 825 826 827
	if ((info->vdevice>>EXT_SHIFT) > 1) {
		/* this is above the extended range; something is wrong */
		printk(KERN_WARNING "blkfront: vdevice 0x%x is above the extended range; ignoring\n", info->vdevice);
		return -ENODEV;
	}

	if (!VDEV_IS_EXTENDED(info->vdevice)) {
828 829 830 831
		err = xen_translate_vdev(info->vdevice, &minor, &offset);
		if (err)
			return err;		
 		nr_parts = PARTS_PER_DISK;
832 833 834
	} else {
		minor = BLKIF_MINOR_EXT(info->vdevice);
		nr_parts = PARTS_PER_EXT_DISK;
835
		offset = minor / nr_parts;
836
		if (xen_hvm_domain() && offset < EMULATED_HD_DISK_NAME_OFFSET + 4)
837 838 839
			printk(KERN_WARNING "blkfront: vdevice 0x%x might conflict with "
					"emulated IDE disks,\n\t choose an xvd device name"
					"from xvde on\n", info->vdevice);
840
	}
841 842 843 844 845
	if (minor >> MINORBITS) {
		pr_warn("blkfront: %#x's minor (%#x) out of range; ignoring\n",
			info->vdevice, minor);
		return -ENODEV;
	}
846 847 848

	if ((minor % nr_parts) == 0)
		nr_minors = nr_parts;
849

850 851 852 853 854
	err = xlbd_reserve_minors(minor, nr_minors);
	if (err)
		goto out;
	err = -ENODEV;

855 856
	gd = alloc_disk(nr_minors);
	if (gd == NULL)
857
		goto release;
858

859 860 861 862 863 864 865 866
	strcpy(gd->disk_name, DEV_NAME);
	ptr = encode_disk_name(gd->disk_name + sizeof(DEV_NAME) - 1, offset);
	BUG_ON(ptr >= gd->disk_name + DISK_NAME_LEN);
	if (nr_minors > 1)
		*ptr = 0;
	else
		snprintf(ptr, gd->disk_name + DISK_NAME_LEN - ptr,
			 "%d", minor & (nr_parts - 1));
867 868 869 870 871 872 873 874

	gd->major = XENVBD_MAJOR;
	gd->first_minor = minor;
	gd->fops = &xlvbd_block_fops;
	gd->private_data = info;
	gd->driverfs_dev = &(info->xbdev->dev);
	set_capacity(gd, capacity);

875
	if (xlvbd_init_blk_queue(gd, sector_size, physical_sector_size,
876 877
				 info->max_indirect_segments ? :
				 BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
878
		del_gendisk(gd);
879
		goto release;
880 881 882 883 884
	}

	info->rq = gd->queue;
	info->gd = gd;

885
	xlvbd_flush(info);
886 887 888 889 890 891 892 893 894 895 896 897

	if (vdisk_info & VDISK_READONLY)
		set_disk_ro(gd, 1);

	if (vdisk_info & VDISK_REMOVABLE)
		gd->flags |= GENHD_FL_REMOVABLE;

	if (vdisk_info & VDISK_CDROM)
		gd->flags |= GENHD_FL_CD;

	return 0;

898 899
 release:
	xlbd_release_minors(minor, nr_minors);
900 901 902 903
 out:
	return err;
}

904 905 906 907 908 909 910 911
static void xlvbd_release_gendisk(struct blkfront_info *info)
{
	unsigned int minor, nr_minors;

	if (info->rq == NULL)
		return;

	/* No more blkif_request(). */
912
	blk_mq_stop_hw_queues(info->rq);
913 914 915 916 917

	/* No more gnttab callback work. */
	gnttab_cancel_free_callback(&info->callback);

	/* Flush gnttab callback work. Must be done with no locks held. */
918
	flush_work(&info->work);
919 920 921 922 923 924 925 926

	del_gendisk(info->gd);

	minor = info->gd->first_minor;
	nr_minors = info->gd->minors;
	xlbd_release_minors(minor, nr_minors);

	blk_cleanup_queue(info->rq);
927
	blk_mq_free_tag_set(&info->tag_set);
928 929 930 931 932 933
	info->rq = NULL;

	put_disk(info->gd);
	info->gd = NULL;
}

934
/* Must be called with io_lock holded */
935 936
static void kick_pending_request_queues(struct blkfront_info *info)
{
937 938
	if (!RING_FULL(&info->ring))
		blk_mq_start_stopped_hw_queues(info->rq, true);
939 940 941 942 943 944
}

static void blkif_restart_queue(struct work_struct *work)
{
	struct blkfront_info *info = container_of(work, struct blkfront_info, work);

945
	spin_lock_irq(&info->io_lock);
946 947
	if (info->connected == BLKIF_STATE_CONNECTED)
		kick_pending_request_queues(info);
948
	spin_unlock_irq(&info->io_lock);
949 950 951 952
}

static void blkif_free(struct blkfront_info *info, int suspend)
{
953 954
	struct grant *persistent_gnt;
	struct grant *n;
955
	int i, j, segs;
956

957
	/* Prevent new requests being issued until we fix things up. */
958
	spin_lock_irq(&info->io_lock);
959 960 961 962
	info->connected = suspend ?
		BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
	/* No more blkif_request(). */
	if (info->rq)
963
		blk_mq_stop_hw_queues(info->rq);
964 965

	/* Remove all persistent grants */
966
	if (!list_empty(&info->grants)) {
967
		list_for_each_entry_safe(persistent_gnt, n,
968
		                         &info->grants, node) {
969
			list_del(&persistent_gnt->node);
970 971 972 973 974
			if (persistent_gnt->gref != GRANT_INVALID_REF) {
				gnttab_end_foreign_access(persistent_gnt->gref,
				                          0, 0UL);
				info->persistent_gnts_c--;
			}
975 976
			if (info->feature_persistent)
				__free_page(pfn_to_page(persistent_gnt->pfn));
977
			kfree(persistent_gnt);
978 979
		}
	}
980
	BUG_ON(info->persistent_gnts_c != 0);
981

982 983 984 985 986 987 988 989 990 991 992 993 994 995
	/*
	 * Remove indirect pages, this only happens when using indirect
	 * descriptors but not persistent grants
	 */
	if (!list_empty(&info->indirect_pages)) {
		struct page *indirect_page, *n;

		BUG_ON(info->feature_persistent);
		list_for_each_entry_safe(indirect_page, n, &info->indirect_pages, lru) {
			list_del(&indirect_page->lru);
			__free_page(indirect_page);
		}
	}

996
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009
		/*
		 * Clear persistent grants present in requests already
		 * on the shared ring
		 */
		if (!info->shadow[i].request)
			goto free_shadow;

		segs = info->shadow[i].req.operation == BLKIF_OP_INDIRECT ?
		       info->shadow[i].req.u.indirect.nr_segments :
		       info->shadow[i].req.u.rw.nr_segments;
		for (j = 0; j < segs; j++) {
			persistent_gnt = info->shadow[i].grants_used[j];
			gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL);
1010 1011
			if (info->feature_persistent)
				__free_page(pfn_to_page(persistent_gnt->pfn));
1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033
			kfree(persistent_gnt);
		}

		if (info->shadow[i].req.operation != BLKIF_OP_INDIRECT)
			/*
			 * If this is not an indirect operation don't try to
			 * free indirect segments
			 */
			goto free_shadow;

		for (j = 0; j < INDIRECT_GREFS(segs); j++) {
			persistent_gnt = info->shadow[i].indirect_grants[j];
			gnttab_end_foreign_access(persistent_gnt->gref, 0, 0UL);
			__free_page(pfn_to_page(persistent_gnt->pfn));
			kfree(persistent_gnt);
		}

free_shadow:
		kfree(info->shadow[i].grants_used);
		info->shadow[i].grants_used = NULL;
		kfree(info->shadow[i].indirect_grants);
		info->shadow[i].indirect_grants = NULL;
1034 1035
		kfree(info->shadow[i].sg);
		info->shadow[i].sg = NULL;
1036 1037
	}

1038 1039
	/* No more gnttab callback work. */
	gnttab_cancel_free_callback(&info->callback);
1040
	spin_unlock_irq(&info->io_lock);
1041 1042

	/* Flush gnttab callback work. Must be done with no locks held. */
1043
	flush_work(&info->work);
1044 1045

	/* Free resources associated with old device channel. */
1046 1047 1048 1049 1050
	for (i = 0; i < info->nr_ring_pages; i++) {
		if (info->ring_ref[i] != GRANT_INVALID_REF) {
			gnttab_end_foreign_access(info->ring_ref[i], 0, 0);
			info->ring_ref[i] = GRANT_INVALID_REF;
		}
1051
	}
1052 1053 1054
	free_pages((unsigned long)info->ring.sring, get_order(info->nr_ring_pages * PAGE_SIZE));
	info->ring.sring = NULL;

1055 1056 1057 1058 1059 1060
	if (info->irq)
		unbind_from_irqhandler(info->irq, info);
	info->evtchn = info->irq = 0;

}

1061 1062
static void blkif_completion(struct blk_shadow *s, struct blkfront_info *info,
			     struct blkif_response *bret)
1063
{
1064
	int i = 0;
1065
	struct scatterlist *sg;
1066 1067
	char *bvec_data;
	void *shared_data;
1068 1069 1070 1071
	int nseg;

	nseg = s->req.operation == BLKIF_OP_INDIRECT ?
		s->req.u.indirect.nr_segments : s->req.u.rw.nr_segments;
1072

1073
	if (bret->operation == BLKIF_OP_READ && info->feature_persistent) {
1074 1075
		for_each_sg(s->sg, sg, nseg, i) {
			BUG_ON(sg->offset + sg->length > PAGE_SIZE);
1076 1077
			shared_data = kmap_atomic(
				pfn_to_page(s->grants_used[i]->pfn));
1078 1079 1080 1081 1082
			bvec_data = kmap_atomic(sg_page(sg));
			memcpy(bvec_data   + sg->offset,
			       shared_data + sg->offset,
			       sg->length);
			kunmap_atomic(bvec_data);
1083 1084 1085 1086
			kunmap_atomic(shared_data);
		}
	}
	/* Add the persistent grant into the list of free grants */
1087
	for (i = 0; i < nseg; i++) {
1088 1089 1090 1091 1092 1093 1094
		if (gnttab_query_foreign_access(s->grants_used[i]->gref)) {
			/*
			 * If the grant is still mapped by the backend (the
			 * backend has chosen to make this grant persistent)
			 * we add it at the head of the list, so it will be
			 * reused first.
			 */
1095 1096 1097 1098
			if (!info->feature_persistent)
				pr_alert_ratelimited("backed has not unmapped grant: %u\n",
						     s->grants_used[i]->gref);
			list_add(&s->grants_used[i]->node, &info->grants);
1099 1100 1101 1102 1103 1104 1105 1106 1107 1108
			info->persistent_gnts_c++;
		} else {
			/*
			 * If the grant is not mapped by the backend we end the
			 * foreign access and add it to the tail of the list,
			 * so it will not be picked again unless we run out of
			 * persistent grants.
			 */
			gnttab_end_foreign_access(s->grants_used[i]->gref, 0, 0UL);
			s->grants_used[i]->gref = GRANT_INVALID_REF;
1109
			list_add_tail(&s->grants_used[i]->node, &info->grants);
1110
		}
1111
	}
1112 1113
	if (s->req.operation == BLKIF_OP_INDIRECT) {
		for (i = 0; i < INDIRECT_GREFS(nseg); i++) {
1114
			if (gnttab_query_foreign_access(s->indirect_grants[i]->gref)) {
1115 1116 1117 1118
				if (!info->feature_persistent)
					pr_alert_ratelimited("backed has not unmapped grant: %u\n",
							     s->indirect_grants[i]->gref);
				list_add(&s->indirect_grants[i]->node, &info->grants);
1119 1120
				info->persistent_gnts_c++;
			} else {
1121 1122
				struct page *indirect_page;

1123
				gnttab_end_foreign_access(s->indirect_grants[i]->gref, 0, 0UL);
1124 1125 1126 1127
				/*
				 * Add the used indirect page back to the list of
				 * available pages for indirect grefs.
				 */
1128 1129 1130 1131
				if (!info->feature_persistent) {
					indirect_page = pfn_to_page(s->indirect_grants[i]->pfn);
					list_add(&indirect_page->lru, &info->indirect_pages);
				}
1132
				s->indirect_grants[i]->gref = GRANT_INVALID_REF;
1133
				list_add_tail(&s->indirect_grants[i]->node, &info->grants);
1134
			}
1135 1136
		}
	}
1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
}

static irqreturn_t blkif_interrupt(int irq, void *dev_id)
{
	struct request *req;
	struct blkif_response *bret;
	RING_IDX i, rp;
	unsigned long flags;
	struct blkfront_info *info = (struct blkfront_info *)dev_id;

1147
	spin_lock_irqsave(&info->io_lock, flags);
1148 1149

	if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
1150
		spin_unlock_irqrestore(&info->io_lock, flags);
1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162
		return IRQ_HANDLED;
	}

 again:
	rp = info->ring.sring->rsp_prod;
	rmb(); /* Ensure we see queued responses up to 'rp'. */

	for (i = info->ring.rsp_cons; i != rp; i++) {
		unsigned long id;

		bret = RING_GET_RESPONSE(&info->ring, i);
		id   = bret->id;
1163 1164 1165 1166 1167
		/*
		 * The backend has messed up and given us an id that we would
		 * never have given to it (we stamp it up to BLK_RING_SIZE -
		 * look in get_id_from_freelist.
		 */
1168
		if (id >= BLK_RING_SIZE(info)) {
1169 1170 1171 1172 1173 1174
			WARN(1, "%s: response to %s has incorrect id (%ld)\n",
			     info->gd->disk_name, op_name(bret->operation), id);
			/* We can't safely get the 'struct request' as
			 * the id is busted. */
			continue;
		}
1175
		req  = info->shadow[id].request;
1176

1177
		if (bret->operation != BLKIF_OP_DISCARD)
1178
			blkif_completion(&info->shadow[id], info, bret);
1179

1180 1181 1182 1183 1184
		if (add_id_to_freelist(info, id)) {
			WARN(1, "%s: response to %s (id %ld) couldn't be recycled!\n",
			     info->gd->disk_name, op_name(bret->operation), id);
			continue;
		}
1185

1186
		req->errors = (bret->status == BLKIF_RSP_OKAY) ? 0 : -EIO;
1187
		switch (bret->operation) {
1188 1189 1190
		case BLKIF_OP_DISCARD:
			if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
				struct request_queue *rq = info->rq;
1191 1192
				printk(KERN_WARNING "blkfront: %s: %s op failed\n",
					   info->gd->disk_name, op_name(bret->operation));
1193
				req->errors = -EOPNOTSUPP;
1194
				info->feature_discard = 0;
1195
				info->feature_secdiscard = 0;
1196
				queue_flag_clear(QUEUE_FLAG_DISCARD, rq);
1197
				queue_flag_clear(QUEUE_FLAG_SECDISCARD, rq);
1198
			}
1199
			blk_mq_complete_request(req);
1200
			break;
1201
		case BLKIF_OP_FLUSH_DISKCACHE:
1202 1203
		case BLKIF_OP_WRITE_BARRIER:
			if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
1204 1205
				printk(KERN_WARNING "blkfront: %s: %s op failed\n",
				       info->gd->disk_name, op_name(bret->operation));
1206
				req->errors = -EOPNOTSUPP;
1207 1208
			}
			if (unlikely(bret->status == BLKIF_RSP_ERROR &&
1209
				     info->shadow[id].req.u.rw.nr_segments == 0)) {
1210 1211
				printk(KERN_WARNING "blkfront: %s: empty %s op failed\n",
				       info->gd->disk_name, op_name(bret->operation));
1212
				req->errors = -EOPNOTSUPP;
1213
			}
1214 1215 1216
			if (unlikely(req->errors)) {
				if (req->errors == -EOPNOTSUPP)
					req->errors = 0;
1217 1218
				info->feature_flush = 0;
				xlvbd_flush(info);
1219 1220 1221 1222 1223 1224 1225 1226
			}
			/* fall through */
		case BLKIF_OP_READ:
		case BLKIF_OP_WRITE:
			if (unlikely(bret->status != BLKIF_RSP_OKAY))
				dev_dbg(&info->xbdev->dev, "Bad return from blkdev data "
					"request: %x\n", bret->status);

1227
			blk_mq_complete_request(req);
1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245
			break;
		default:
			BUG();
		}
	}

	info->ring.rsp_cons = i;

	if (i != info->ring.req_prod_pvt) {
		int more_to_do;
		RING_FINAL_CHECK_FOR_RESPONSES(&info->ring, more_to_do);
		if (more_to_do)
			goto again;
	} else
		info->ring.sring->rsp_event = i + 1;

	kick_pending_request_queues(info);

1246
	spin_unlock_irqrestore(&info->io_lock, flags);
1247 1248 1249 1250 1251 1252 1253 1254 1255

	return IRQ_HANDLED;
}


static int setup_blkring(struct xenbus_device *dev,
			 struct blkfront_info *info)
{
	struct blkif_sring *sring;
1256 1257 1258
	int err, i;
	unsigned long ring_size = info->nr_ring_pages * PAGE_SIZE;
	grant_ref_t gref[XENBUS_MAX_RING_PAGES];
1259

1260 1261
	for (i = 0; i < info->nr_ring_pages; i++)
		info->ring_ref[i] = GRANT_INVALID_REF;
1262

1263 1264
	sring = (struct blkif_sring *)__get_free_pages(GFP_NOIO | __GFP_HIGH,
						       get_order(ring_size));
1265 1266 1267 1268 1269
	if (!sring) {
		xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
		return -ENOMEM;
	}
	SHARED_RING_INIT(sring);
1270
	FRONT_RING_INIT(&info->ring, sring, ring_size);
1271

1272
	err = xenbus_grant_ring(dev, info->ring.sring, info->nr_ring_pages, gref);
1273
	if (err < 0) {
1274
		free_pages((unsigned long)sring, get_order(ring_size));
1275 1276 1277
		info->ring.sring = NULL;
		goto fail;
	}
1278 1279
	for (i = 0; i < info->nr_ring_pages; i++)
		info->ring_ref[i] = gref[i];
1280 1281 1282 1283 1284

	err = xenbus_alloc_evtchn(dev, &info->evtchn);
	if (err)
		goto fail;

1285 1286
	err = bind_evtchn_to_irqhandler(info->evtchn, blkif_interrupt, 0,
					"blkif", info);
1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301
	if (err <= 0) {
		xenbus_dev_fatal(dev, err,
				 "bind_evtchn_to_irqhandler failed");
		goto fail;
	}
	info->irq = err;

	return 0;
fail:
	blkif_free(info, 0);
	return err;
}


/* Common code used when first setting up, and when resuming. */
1302
static int talk_to_blkback(struct xenbus_device *dev,
1303 1304 1305 1306
			   struct blkfront_info *info)
{
	const char *message = NULL;
	struct xenbus_transaction xbt;
1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318
	int err, i;
	unsigned int max_page_order = 0;
	unsigned int ring_page_order = 0;

	err = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
			   "max-ring-page-order", "%u", &max_page_order);
	if (err != 1)
		info->nr_ring_pages = 1;
	else {
		ring_page_order = min(xen_blkif_max_ring_order, max_page_order);
		info->nr_ring_pages = 1 << ring_page_order;
	}
1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331

	/* Create shared ring, alloc event channel. */
	err = setup_blkring(dev, info);
	if (err)
		goto out;

again:
	err = xenbus_transaction_start(&xbt);
	if (err) {
		xenbus_dev_fatal(dev, err, "starting transaction");
		goto destroy_blkring;
	}

1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357
	if (info->nr_ring_pages == 1) {
		err = xenbus_printf(xbt, dev->nodename,
				    "ring-ref", "%u", info->ring_ref[0]);
		if (err) {
			message = "writing ring-ref";
			goto abort_transaction;
		}
	} else {
		err = xenbus_printf(xbt, dev->nodename,
				    "ring-page-order", "%u", ring_page_order);
		if (err) {
			message = "writing ring-page-order";
			goto abort_transaction;
		}

		for (i = 0; i < info->nr_ring_pages; i++) {
			char ring_ref_name[RINGREF_NAME_LEN];

			snprintf(ring_ref_name, RINGREF_NAME_LEN, "ring-ref%u", i);
			err = xenbus_printf(xbt, dev->nodename, ring_ref_name,
					    "%u", info->ring_ref[i]);
			if (err) {
				message = "writing ring-ref";
				goto abort_transaction;
			}
		}
1358 1359 1360 1361 1362 1363 1364
	}
	err = xenbus_printf(xbt, dev->nodename,
			    "event-channel", "%u", info->evtchn);
	if (err) {
		message = "writing event-channel";
		goto abort_transaction;
	}
1365 1366 1367 1368 1369 1370
	err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
			    XEN_IO_PROTO_ABI_NATIVE);
	if (err) {
		message = "writing protocol";
		goto abort_transaction;
	}
1371
	err = xenbus_printf(xbt, dev->nodename,
1372
			    "feature-persistent", "%u", 1);
1373 1374 1375
	if (err)
		dev_warn(&dev->dev,
			 "writing persistent grants feature to xenbus");
1376 1377 1378 1379 1380 1381 1382 1383 1384

	err = xenbus_transaction_end(xbt, 0);
	if (err) {
		if (err == -EAGAIN)
			goto again;
		xenbus_dev_fatal(dev, err, "completing transaction");
		goto destroy_blkring;
	}

1385 1386 1387
	for (i = 0; i < BLK_RING_SIZE(info); i++)
		info->shadow[i].req.u.rw.id = i+1;
	info->shadow[BLK_RING_SIZE(info)-1].req.u.rw.id = 0x0fffffff;
1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410
	xenbus_switch_state(dev, XenbusStateInitialised);

	return 0;

 abort_transaction:
	xenbus_transaction_end(xbt, 1);
	if (message)
		xenbus_dev_fatal(dev, err, "%s", message);
 destroy_blkring:
	blkif_free(info, 0);
 out:
	return err;
}

/**
 * Entry point to this code when a new device is created.  Allocate the basic
 * structures and the ring buffer for communication with the backend, and
 * inform the backend of the appropriate details for those.  Switch to
 * Initialised state.
 */
static int blkfront_probe(struct xenbus_device *dev,
			  const struct xenbus_device_id *id)
{
1411
	int err, vdevice;
1412 1413 1414 1415 1416 1417
	struct blkfront_info *info;

	/* FIXME: Use dynamic device id if this is not set. */
	err = xenbus_scanf(XBT_NIL, dev->nodename,
			   "virtual-device", "%i", &vdevice);
	if (err != 1) {
1418 1419 1420 1421 1422 1423 1424
		/* go looking in the extended area instead */
		err = xenbus_scanf(XBT_NIL, dev->nodename, "virtual-device-ext",
				   "%i", &vdevice);
		if (err != 1) {
			xenbus_dev_fatal(dev, err, "reading virtual-device");
			return err;
		}
1425 1426
	}

1427 1428 1429 1430
	if (xen_hvm_domain()) {
		char *type;
		int len;
		/* no unplug has been done: do not hook devices != xen vbds */
1431
		if (xen_has_pv_and_legacy_disk_devices()) {
1432 1433 1434 1435 1436 1437 1438 1439 1440 1441
			int major;

			if (!VDEV_IS_EXTENDED(vdevice))
				major = BLKIF_MAJOR(vdevice);
			else
				major = XENVBD_MAJOR;

			if (major != XENVBD_MAJOR) {
				printk(KERN_INFO
						"%s: HVM does not support vbd %d as xen block device\n",
1442
						__func__, vdevice);
1443 1444 1445 1446 1447 1448 1449 1450 1451
				return -ENODEV;
			}
		}
		/* do not create a PV cdrom device if we are an HVM guest */
		type = xenbus_read(XBT_NIL, dev->nodename, "device-type", &len);
		if (IS_ERR(type))
			return -ENODEV;
		if (strncmp(type, "cdrom", 5) == 0) {
			kfree(type);
1452 1453
			return -ENODEV;
		}
1454
		kfree(type);
1455
	}
1456 1457 1458 1459 1460 1461
	info = kzalloc(sizeof(*info), GFP_KERNEL);
	if (!info) {
		xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
		return -ENOMEM;
	}

1462
	mutex_init(&info->mutex);
1463
	spin_lock_init(&info->io_lock);
1464 1465
	info->xbdev = dev;
	info->vdevice = vdevice;
1466 1467
	INIT_LIST_HEAD(&info->grants);
	INIT_LIST_HEAD(&info->indirect_pages);
1468
	info->persistent_gnts_c = 0;
1469 1470 1471 1472 1473
	info->connected = BLKIF_STATE_DISCONNECTED;
	INIT_WORK(&info->work, blkif_restart_queue);

	/* Front end dir is a number, which is used as the id. */
	info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0);
1474
	dev_set_drvdata(&dev->dev, info);
1475 1476 1477 1478

	return 0;
}

1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492
static void split_bio_end(struct bio *bio, int error)
{
	struct split_bio *split_bio = bio->bi_private;

	if (error)
		split_bio->err = error;

	if (atomic_dec_and_test(&split_bio->pending)) {
		split_bio->bio->bi_phys_segments = 0;
		bio_endio(split_bio->bio, split_bio->err);
		kfree(split_bio);
	}
	bio_put(bio);
}
1493 1494 1495 1496

static int blkif_recover(struct blkfront_info *info)
{
	int i;
1497
	struct request *req, *n;
1498
	struct blk_shadow *copy;
1499 1500 1501 1502 1503 1504 1505
	int rc;
	struct bio *bio, *cloned_bio;
	struct bio_list bio_list, merge_bio;
	unsigned int segs, offset;
	int pending, size;
	struct split_bio *split_bio;
	struct list_head requests;
1506 1507

	/* Stage 1: Make a safe copy of the shadow state. */
1508
	copy = kmemdup(info->shadow, sizeof(info->shadow),
1509
		       GFP_NOIO | __GFP_REPEAT | __GFP_HIGH);
1510 1511 1512 1513 1514
	if (!copy)
		return -ENOMEM;

	/* Stage 2: Set up free list. */
	memset(&info->shadow, 0, sizeof(info->shadow));
1515
	for (i = 0; i < BLK_RING_SIZE(info); i++)
1516
		info->shadow[i].req.u.rw.id = i+1;
1517
	info->shadow_free = info->ring.req_prod_pvt;
1518
	info->shadow[BLK_RING_SIZE(info)-1].req.u.rw.id = 0x0fffffff;
1519

1520
	rc = blkfront_gather_backend_features(info);
1521 1522 1523 1524 1525 1526 1527 1528 1529
	if (rc) {
		kfree(copy);
		return rc;
	}

	segs = info->max_indirect_segments ? : BLKIF_MAX_SEGMENTS_PER_REQUEST;
	blk_queue_max_segments(info->rq, segs);
	bio_list_init(&bio_list);
	INIT_LIST_HEAD(&requests);
1530
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1531
		/* Not in use? */
1532
		if (!copy[i].request)
1533 1534
			continue;

1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545
		/*
		 * Get the bios in the request so we can re-queue them.
		 */
		if (copy[i].request->cmd_flags &
		    (REQ_FLUSH | REQ_FUA | REQ_DISCARD | REQ_SECURE)) {
			/*
			 * Flush operations don't contain bios, so
			 * we need to requeue the whole request
			 */
			list_add(&copy[i].request->queuelist, &requests);
			continue;
1546
		}
1547 1548 1549 1550
		merge_bio.head = copy[i].request->bio;
		merge_bio.tail = copy[i].request->biotail;
		bio_list_merge(&bio_list, &merge_bio);
		copy[i].request->bio = NULL;
1551
		blk_end_request_all(copy[i].request, 0);
1552 1553 1554 1555 1556 1557
	}

	kfree(copy);

	xenbus_switch_state(info->xbdev, XenbusStateConnected);

1558
	spin_lock_irq(&info->io_lock);
1559 1560 1561 1562 1563 1564 1565

	/* Now safe for us to use the shared ring */
	info->connected = BLKIF_STATE_CONNECTED;

	/* Kick any other new requests queued since we resumed */
	kick_pending_request_queues(info);

1566 1567 1568 1569
	list_for_each_entry_safe(req, n, &requests, queuelist) {
		/* Requeue pending requests (flush or discard) */
		list_del_init(&req->queuelist);
		BUG_ON(req->nr_phys_segments > segs);
1570
		blk_mq_requeue_request(req);
1571
	}
1572
	spin_unlock_irq(&info->io_lock);
1573
	blk_mq_kick_requeue_list(info->rq);
1574

1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589
	while ((bio = bio_list_pop(&bio_list)) != NULL) {
		/* Traverse the list of pending bios and re-queue them */
		if (bio_segments(bio) > segs) {
			/*
			 * This bio has more segments than what we can
			 * handle, we have to split it.
			 */
			pending = (bio_segments(bio) + segs - 1) / segs;
			split_bio = kzalloc(sizeof(*split_bio), GFP_NOIO);
			BUG_ON(split_bio == NULL);
			atomic_set(&split_bio->pending, pending);
			split_bio->bio = bio;
			for (i = 0; i < pending; i++) {
				offset = (i * segs * PAGE_SIZE) >> 9;
				size = min((unsigned int)(segs * PAGE_SIZE) >> 9,
1590
					   (unsigned int)bio_sectors(bio) - offset);
1591 1592
				cloned_bio = bio_clone(bio, GFP_NOIO);
				BUG_ON(cloned_bio == NULL);
1593
				bio_trim(cloned_bio, offset, size);
1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607
				cloned_bio->bi_private = split_bio;
				cloned_bio->bi_end_io = split_bio_end;
				submit_bio(cloned_bio->bi_rw, cloned_bio);
			}
			/*
			 * Now we have to wait for all those smaller bios to
			 * end, so we can also end the "parent" bio.
			 */
			continue;
		}
		/* We don't need to split this bio */
		submit_bio(bio->bi_rw, bio);
	}

1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618
	return 0;
}

/**
 * We are reconnecting to the backend, due to a suspend/resume, or a backend
 * driver restart.  We tear down our blkif structure and recreate it, but
 * leave the device-layer structures intact so that this is transparent to the
 * rest of the kernel.
 */
static int blkfront_resume(struct xenbus_device *dev)
{
1619
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1620 1621 1622 1623 1624 1625
	int err;

	dev_dbg(&dev->dev, "blkfront_resume: %s\n", dev->nodename);

	blkif_free(info, info->connected == BLKIF_STATE_CONNECTED);

1626
	err = talk_to_blkback(dev, info);
1627 1628 1629 1630 1631 1632

	/*
	 * We have to wait for the backend to switch to
	 * connected state, since we want to read which
	 * features it supports.
	 */
1633 1634 1635 1636

	return err;
}

1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661
static void
blkfront_closing(struct blkfront_info *info)
{
	struct xenbus_device *xbdev = info->xbdev;
	struct block_device *bdev = NULL;

	mutex_lock(&info->mutex);

	if (xbdev->state == XenbusStateClosing) {
		mutex_unlock(&info->mutex);
		return;
	}

	if (info->gd)
		bdev = bdget_disk(info->gd, 0);

	mutex_unlock(&info->mutex);

	if (!bdev) {
		xenbus_frontend_closed(xbdev);
		return;
	}

	mutex_lock(&bdev->bd_mutex);

1662
	if (bdev->bd_openers) {
1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673
		xenbus_dev_error(xbdev, -EBUSY,
				 "Device in use; refusing to close");
		xenbus_switch_state(xbdev, XenbusStateClosing);
	} else {
		xlvbd_release_gendisk(info);
		xenbus_frontend_closed(xbdev);
	}

	mutex_unlock(&bdev->bd_mutex);
	bdput(bdev);
}
1674

1675 1676 1677 1678 1679
static void blkfront_setup_discard(struct blkfront_info *info)
{
	int err;
	unsigned int discard_granularity;
	unsigned int discard_alignment;
1680
	unsigned int discard_secure;
1681

1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695
	info->feature_discard = 1;
	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
		"discard-granularity", "%u", &discard_granularity,
		"discard-alignment", "%u", &discard_alignment,
		NULL);
	if (!err) {
		info->discard_granularity = discard_granularity;
		info->discard_alignment = discard_alignment;
	}
	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
		    "discard-secure", "%d", &discard_secure,
		    NULL);
	if (!err)
		info->feature_secdiscard = !!discard_secure;
1696 1697
}

1698 1699
static int blkfront_setup_indirect(struct blkfront_info *info)
{
1700
	unsigned int segs;
1701 1702
	int err, i;

1703
	if (info->max_indirect_segments == 0)
1704
		segs = BLKIF_MAX_SEGMENTS_PER_REQUEST;
1705
	else
1706 1707
		segs = info->max_indirect_segments;

1708
	err = fill_grant_buffer(info, (segs + INDIRECT_GREFS(segs)) * BLK_RING_SIZE(info));
1709 1710 1711
	if (err)
		goto out_of_memory;

1712 1713 1714 1715 1716 1717
	if (!info->feature_persistent && info->max_indirect_segments) {
		/*
		 * We are using indirect descriptors but not persistent
		 * grants, we need to allocate a set of pages that can be
		 * used for mapping indirect grefs
		 */
1718
		int num = INDIRECT_GREFS(segs) * BLK_RING_SIZE(info);
1719 1720 1721 1722 1723 1724 1725 1726 1727 1728

		BUG_ON(!list_empty(&info->indirect_pages));
		for (i = 0; i < num; i++) {
			struct page *indirect_page = alloc_page(GFP_NOIO);
			if (!indirect_page)
				goto out_of_memory;
			list_add(&indirect_page->lru, &info->indirect_pages);
		}
	}

1729
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1730 1731 1732
		info->shadow[i].grants_used = kzalloc(
			sizeof(info->shadow[i].grants_used[0]) * segs,
			GFP_NOIO);
1733
		info->shadow[i].sg = kzalloc(sizeof(info->shadow[i].sg[0]) * segs, GFP_NOIO);
1734 1735 1736 1737 1738 1739
		if (info->max_indirect_segments)
			info->shadow[i].indirect_grants = kzalloc(
				sizeof(info->shadow[i].indirect_grants[0]) *
				INDIRECT_GREFS(segs),
				GFP_NOIO);
		if ((info->shadow[i].grants_used == NULL) ||
1740
			(info->shadow[i].sg == NULL) ||
1741 1742 1743
		     (info->max_indirect_segments &&
		     (info->shadow[i].indirect_grants == NULL)))
			goto out_of_memory;
1744
		sg_init_table(info->shadow[i].sg, segs);
1745 1746 1747 1748 1749 1750
	}


	return 0;

out_of_memory:
1751
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1752 1753
		kfree(info->shadow[i].grants_used);
		info->shadow[i].grants_used = NULL;
1754 1755
		kfree(info->shadow[i].sg);
		info->shadow[i].sg = NULL;
1756 1757 1758
		kfree(info->shadow[i].indirect_grants);
		info->shadow[i].indirect_grants = NULL;
	}
1759 1760 1761 1762 1763 1764 1765
	if (!list_empty(&info->indirect_pages)) {
		struct page *indirect_page, *n;
		list_for_each_entry_safe(indirect_page, n, &info->indirect_pages, lru) {
			list_del(&indirect_page->lru);
			__free_page(indirect_page);
		}
	}
1766 1767 1768
	return -ENOMEM;
}

1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830
/*
 * Gather all backend feature-*
 */
static int blkfront_gather_backend_features(struct blkfront_info *info)
{
	int err;
	int barrier, flush, discard, persistent;
	unsigned int indirect_segments;

	info->feature_flush = 0;

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			"feature-barrier", "%d", &barrier,
			NULL);

	/*
	 * If there's no "feature-barrier" defined, then it means
	 * we're dealing with a very old backend which writes
	 * synchronously; nothing to do.
	 *
	 * If there are barriers, then we use flush.
	 */
	if (!err && barrier)
		info->feature_flush = REQ_FLUSH | REQ_FUA;
	/*
	 * And if there is "feature-flush-cache" use that above
	 * barriers.
	 */
	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			"feature-flush-cache", "%d", &flush,
			NULL);

	if (!err && flush)
		info->feature_flush = REQ_FLUSH;

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			"feature-discard", "%d", &discard,
			NULL);

	if (!err && discard)
		blkfront_setup_discard(info);

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			"feature-persistent", "%u", &persistent,
			NULL);
	if (err)
		info->feature_persistent = 0;
	else
		info->feature_persistent = persistent;

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			    "feature-max-indirect-segments", "%u", &indirect_segments,
			    NULL);
	if (err)
		info->max_indirect_segments = 0;
	else
		info->max_indirect_segments = min(indirect_segments,
						  xen_blkif_max_segments);

	return blkfront_setup_indirect(info);
}

1831 1832 1833 1834 1835 1836 1837 1838
/*
 * Invoked when the backend is finally 'ready' (and has told produced
 * the details about the physical device - #sectors, size, etc).
 */
static void blkfront_connect(struct blkfront_info *info)
{
	unsigned long long sectors;
	unsigned long sector_size;
1839
	unsigned int physical_sector_size;
1840 1841 1842
	unsigned int binfo;
	int err;

1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855
	switch (info->connected) {
	case BLKIF_STATE_CONNECTED:
		/*
		 * Potentially, the back-end may be signalling
		 * a capacity change; update the capacity.
		 */
		err = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
				   "sectors", "%Lu", &sectors);
		if (XENBUS_EXIST_ERR(err))
			return;
		printk(KERN_INFO "Setting capacity to %Lu\n",
		       sectors);
		set_capacity(info->gd, sectors);
1856
		revalidate_disk(info->gd);
1857

1858
		return;
1859
	case BLKIF_STATE_SUSPENDED:
1860 1861 1862 1863 1864 1865 1866
		/*
		 * If we are recovering from suspension, we need to wait
		 * for the backend to announce it's features before
		 * reconnecting, at least we need to know if the backend
		 * supports indirect descriptors, and how many.
		 */
		blkif_recover(info);
1867 1868
		return;

1869 1870
	default:
		break;
1871
	}
1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887

	dev_dbg(&info->xbdev->dev, "%s:%s.\n",
		__func__, info->xbdev->otherend);

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
			    "sectors", "%llu", &sectors,
			    "info", "%u", &binfo,
			    "sector-size", "%lu", &sector_size,
			    NULL);
	if (err) {
		xenbus_dev_fatal(info->xbdev, err,
				 "reading backend fields at %s",
				 info->xbdev->otherend);
		return;
	}

1888 1889 1890 1891 1892 1893 1894 1895 1896 1897
	/*
	 * physcial-sector-size is a newer field, so old backends may not
	 * provide this. Assume physical sector size to be the same as
	 * sector_size in that case.
	 */
	err = xenbus_scanf(XBT_NIL, info->xbdev->otherend,
			   "physical-sector-size", "%u", &physical_sector_size);
	if (err != 1)
		physical_sector_size = sector_size;

1898
	err = blkfront_gather_backend_features(info);
1899 1900 1901 1902 1903 1904
	if (err) {
		xenbus_dev_fatal(info->xbdev, err, "setup_indirect at %s",
				 info->xbdev->otherend);
		return;
	}

1905 1906
	err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size,
				  physical_sector_size);
1907 1908 1909 1910 1911 1912 1913 1914 1915
	if (err) {
		xenbus_dev_fatal(info->xbdev, err, "xlvbd_add at %s",
				 info->xbdev->otherend);
		return;
	}

	xenbus_switch_state(info->xbdev, XenbusStateConnected);

	/* Kick pending requests. */
1916
	spin_lock_irq(&info->io_lock);
1917 1918
	info->connected = BLKIF_STATE_CONNECTED;
	kick_pending_request_queues(info);
1919
	spin_unlock_irq(&info->io_lock);
1920 1921

	add_disk(info->gd);
1922 1923

	info->is_ready = 1;
1924 1925 1926 1927 1928
}

/**
 * Callback received when the backend's state changes.
 */
1929
static void blkback_changed(struct xenbus_device *dev,
1930 1931
			    enum xenbus_state backend_state)
{
1932
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1933

1934
	dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state);
1935 1936 1937

	switch (backend_state) {
	case XenbusStateInitWait:
1938 1939
		if (dev->state != XenbusStateInitialising)
			break;
1940 1941 1942 1943 1944 1945
		if (talk_to_blkback(dev, info)) {
			kfree(info);
			dev_set_drvdata(&dev->dev, NULL);
			break;
		}
	case XenbusStateInitialising:
1946
	case XenbusStateInitialised:
1947 1948
	case XenbusStateReconfiguring:
	case XenbusStateReconfigured:
1949 1950 1951 1952 1953 1954 1955
	case XenbusStateUnknown:
		break;

	case XenbusStateConnected:
		blkfront_connect(info);
		break;

1956 1957 1958 1959
	case XenbusStateClosed:
		if (dev->state == XenbusStateClosed)
			break;
		/* Missed the backend's Closing state -- fallthrough */
1960
	case XenbusStateClosing:
1961
		blkfront_closing(info);
1962 1963 1964 1965
		break;
	}
}

1966
static int blkfront_remove(struct xenbus_device *xbdev)
1967
{
1968 1969 1970
	struct blkfront_info *info = dev_get_drvdata(&xbdev->dev);
	struct block_device *bdev = NULL;
	struct gendisk *disk;
1971

1972
	dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename);
1973 1974 1975

	blkif_free(info, 0);

1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
	mutex_lock(&info->mutex);

	disk = info->gd;
	if (disk)
		bdev = bdget_disk(disk, 0);

	info->xbdev = NULL;
	mutex_unlock(&info->mutex);

	if (!bdev) {
		kfree(info);
		return 0;
	}

	/*
	 * The xbdev was removed before we reached the Closed
	 * state. See if it's safe to remove the disk. If the bdev
	 * isn't closed yet, we let release take care of it.
	 */

	mutex_lock(&bdev->bd_mutex);
	info = disk->private_data;

1999 2000 2001 2002
	dev_warn(disk_to_dev(disk),
		 "%s was hot-unplugged, %d stale handles\n",
		 xbdev->nodename, bdev->bd_openers);

2003
	if (info && !bdev->bd_openers) {
2004 2005
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
2006
		kfree(info);
2007 2008 2009 2010
	}

	mutex_unlock(&bdev->bd_mutex);
	bdput(bdev);
2011 2012 2013 2014

	return 0;
}

2015 2016
static int blkfront_is_ready(struct xenbus_device *dev)
{
2017
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
2018

2019
	return info->is_ready && info->xbdev;
2020 2021
}

Al Viro's avatar
Al Viro committed
2022
static int blkif_open(struct block_device *bdev, fmode_t mode)
2023
{
2024 2025 2026
	struct gendisk *disk = bdev->bd_disk;
	struct blkfront_info *info;
	int err = 0;
2027

2028
	mutex_lock(&blkfront_mutex);
2029

2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045
	info = disk->private_data;
	if (!info) {
		/* xbdev gone */
		err = -ERESTARTSYS;
		goto out;
	}

	mutex_lock(&info->mutex);

	if (!info->gd)
		/* xbdev is closed */
		err = -ERESTARTSYS;

	mutex_unlock(&info->mutex);

out:
2046
	mutex_unlock(&blkfront_mutex);
2047
	return err;
2048 2049
}

2050
static void blkif_release(struct gendisk *disk, fmode_t mode)
2051
{
Al Viro's avatar
Al Viro committed
2052
	struct blkfront_info *info = disk->private_data;
2053 2054 2055
	struct block_device *bdev;
	struct xenbus_device *xbdev;

2056
	mutex_lock(&blkfront_mutex);
2057 2058 2059

	bdev = bdget_disk(disk, 0);

2060 2061 2062 2063
	if (!bdev) {
		WARN(1, "Block device %s yanked out from us!\n", disk->disk_name);
		goto out_mutex;
	}
2064 2065 2066
	if (bdev->bd_openers)
		goto out;

2067 2068 2069 2070 2071 2072 2073 2074 2075 2076
	/*
	 * Check if we have been instructed to close. We will have
	 * deferred this request, because the bdev was still open.
	 */

	mutex_lock(&info->mutex);
	xbdev = info->xbdev;

	if (xbdev && xbdev->state == XenbusStateClosing) {
		/* pending switch to state closed */
2077
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
2078 2079 2080 2081 2082 2083 2084 2085
		xlvbd_release_gendisk(info);
		xenbus_frontend_closed(info->xbdev);
 	}

	mutex_unlock(&info->mutex);

	if (!xbdev) {
		/* sudden device removal */
2086
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
2087 2088 2089
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
		kfree(info);
2090
	}
2091

2092
out:
2093
	bdput(bdev);
2094
out_mutex:
2095
	mutex_unlock(&blkfront_mutex);
2096 2097
}

2098
static const struct block_device_operations xlvbd_block_fops =
2099 2100
{
	.owner = THIS_MODULE,
Al Viro's avatar
Al Viro committed
2101 2102
	.open = blkif_open,
	.release = blkif_release,
2103
	.getgeo = blkif_getgeo,
2104
	.ioctl = blkif_ioctl,
2105 2106 2107
};


2108
static const struct xenbus_device_id blkfront_ids[] = {
2109 2110 2111 2112
	{ "vbd" },
	{ "" }
};

2113 2114
static struct xenbus_driver blkfront_driver = {
	.ids  = blkfront_ids,
2115 2116 2117
	.probe = blkfront_probe,
	.remove = blkfront_remove,
	.resume = blkfront_resume,
2118
	.otherend_changed = blkback_changed,
2119
	.is_ready = blkfront_is_ready,
2120
};
2121 2122 2123

static int __init xlblk_init(void)
{
2124 2125
	int ret;

2126
	if (!xen_domain())
2127 2128
		return -ENODEV;

2129 2130 2131 2132 2133 2134
	if (xen_blkif_max_ring_order > XENBUS_MAX_RING_PAGE_ORDER) {
		pr_info("Invalid max_ring_order (%d), will use default max: %d.\n",
			xen_blkif_max_ring_order, XENBUS_MAX_RING_PAGE_ORDER);
		xen_blkif_max_ring_order = 0;
	}

2135
	if (!xen_has_pv_disk_devices())
2136 2137
		return -ENODEV;

2138 2139 2140 2141 2142 2143
	if (register_blkdev(XENVBD_MAJOR, DEV_NAME)) {
		printk(KERN_WARNING "xen_blk: can't get major %d with name %s\n",
		       XENVBD_MAJOR, DEV_NAME);
		return -ENODEV;
	}

2144
	ret = xenbus_register_frontend(&blkfront_driver);
2145 2146 2147 2148 2149 2150
	if (ret) {
		unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
		return ret;
	}

	return 0;
2151 2152 2153 2154
}
module_init(xlblk_init);


2155
static void __exit xlblk_exit(void)
2156
{
2157 2158 2159
	xenbus_unregister_driver(&blkfront_driver);
	unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
	kfree(minors);
2160 2161 2162 2163 2164 2165
}
module_exit(xlblk_exit);

MODULE_DESCRIPTION("Xen virtual block device frontend");
MODULE_LICENSE("GPL");
MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
2166
MODULE_ALIAS("xen:vbd");
2167
MODULE_ALIAS("xenblk");