xen-blkfront.c 57.3 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>
B
Bob Liu 已提交
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
};

struct split_bio {
	struct bio *bio;
	atomic_t pending;
86 87
};

88
static DEFINE_MUTEX(blkfront_mutex);
89
static const struct block_device_operations xlvbd_block_fops;
90

91 92 93 94 95 96 97
/*
 * 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;
98 99
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)");
100

B
Bob Liu 已提交
101 102 103 104 105 106 107 108 109 110 111 112 113 114 115
/*
 * 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)
116 117 118 119 120 121 122 123

/*
 * 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
{
124
	spinlock_t io_lock;
125
	struct mutex mutex;
126 127 128 129 130
	struct xenbus_device *xbdev;
	struct gendisk *gd;
	int vdevice;
	blkif_vdev_t handle;
	enum blkif_state connected;
B
Bob Liu 已提交
131 132
	int ring_ref[XENBUS_MAX_RING_PAGES];
	unsigned int nr_ring_pages;
133 134 135 136 137
	struct blkif_front_ring ring;
	unsigned int evtchn, irq;
	struct request_queue *rq;
	struct work_struct work;
	struct gnttab_free_callback callback;
B
Bob Liu 已提交
138
	struct blk_shadow shadow[BLK_MAX_RING_SIZE];
139 140
	struct list_head grants;
	struct list_head indirect_pages;
141
	unsigned int persistent_gnts_c;
142
	unsigned long shadow_free;
143
	unsigned int feature_flush;
144 145
	unsigned int feature_discard:1;
	unsigned int feature_secdiscard:1;
146 147
	unsigned int discard_granularity;
	unsigned int discard_alignment;
148
	unsigned int feature_persistent:1;
149
	unsigned int max_indirect_segments;
150
	int is_ready;
B
Bob Liu 已提交
151
	struct blk_mq_tag_set tag_set;
152 153
};

154 155 156 157
static unsigned int nr_minors;
static unsigned long *minors;
static DEFINE_SPINLOCK(minor_lock);

158 159 160
#define GRANT_INVALID_REF	0

#define PARTS_PER_DISK		16
161
#define PARTS_PER_EXT_DISK      256
162 163 164 165

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

166 167 168 169
#define EXT_SHIFT 28
#define EXTENDED (1<<EXT_SHIFT)
#define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
#define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
170 171
#define EMULATED_HD_DISK_MINOR_OFFSET (0)
#define EMULATED_HD_DISK_NAME_OFFSET (EMULATED_HD_DISK_MINOR_OFFSET / 256)
172 173
#define EMULATED_SD_DISK_MINOR_OFFSET (0)
#define EMULATED_SD_DISK_NAME_OFFSET (EMULATED_SD_DISK_MINOR_OFFSET / 256)
174

175
#define DEV_NAME	"xvd"	/* name in /dev */
176

177
#define SEGS_PER_INDIRECT_FRAME \
178
	(PAGE_SIZE/sizeof(struct blkif_request_segment))
179 180 181 182
#define INDIRECT_GREFS(_segs) \
	((_segs + SEGS_PER_INDIRECT_FRAME - 1)/SEGS_PER_INDIRECT_FRAME)

static int blkfront_setup_indirect(struct blkfront_info *info);
183
static int blkfront_gather_backend_features(struct blkfront_info *info);
184

185 186 187
static int get_id_from_freelist(struct blkfront_info *info)
{
	unsigned long free = info->shadow_free;
B
Bob Liu 已提交
188
	BUG_ON(free >= BLK_RING_SIZE(info));
189 190
	info->shadow_free = info->shadow[free].req.u.rw.id;
	info->shadow[free].req.u.rw.id = 0x0fffffee; /* debug */
191 192 193
	return free;
}

194
static int add_id_to_freelist(struct blkfront_info *info,
195 196
			       unsigned long id)
{
197 198 199 200
	if (info->shadow[id].req.u.rw.id != id)
		return -EINVAL;
	if (info->shadow[id].request == NULL)
		return -EINVAL;
201
	info->shadow[id].req.u.rw.id  = info->shadow_free;
202
	info->shadow[id].request = NULL;
203
	info->shadow_free = id;
204
	return 0;
205 206
}

207 208 209 210 211 212 213 214 215 216 217
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;

218 219 220 221 222 223 224
		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);
225 226 227
		}

		gnt_list_entry->gref = GRANT_INVALID_REF;
228
		list_add(&gnt_list_entry->node, &info->grants);
229 230 231 232 233 234 235
		i++;
	}

	return 0;

out_of_memory:
	list_for_each_entry_safe(gnt_list_entry, n,
236
	                         &info->grants, node) {
237
		list_del(&gnt_list_entry->node);
238 239
		if (info->feature_persistent)
			__free_page(pfn_to_page(gnt_list_entry->pfn));
240 241 242 243 244 245 246 247
		kfree(gnt_list_entry);
		i--;
	}
	BUG_ON(i != 0);
	return -ENOMEM;
}

static struct grant *get_grant(grant_ref_t *gref_head,
248
                               unsigned long pfn,
249 250 251
                               struct blkfront_info *info)
{
	struct grant *gnt_list_entry;
252
	unsigned long buffer_gfn;
253

254 255
	BUG_ON(list_empty(&info->grants));
	gnt_list_entry = list_first_entry(&info->grants, struct grant,
256 257 258 259 260 261 262 263 264 265 266
	                                  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);
267 268 269 270
	if (!info->feature_persistent) {
		BUG_ON(!pfn);
		gnt_list_entry->pfn = pfn;
	}
271
	buffer_gfn = pfn_to_gfn(gnt_list_entry->pfn);
272 273
	gnttab_grant_foreign_access_ref(gnt_list_entry->gref,
	                                info->xbdev->otherend_id,
274
	                                buffer_gfn, 0);
275 276 277
	return gnt_list_entry;
}

278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294
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];
}
295 296 297 298 299 300 301 302
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;

303
		bitmap = kcalloc(BITS_TO_LONGS(end), sizeof(*bitmap),
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322
				 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) {
323
		bitmap_set(minors, minor, nr);
324 325 326 327 328 329 330 331 332 333 334 335 336 337
		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);
338
	bitmap_clear(minors,  minor, nr);
339 340 341
	spin_unlock(&minor_lock);
}

342 343 344 345 346 347
static void blkif_restart_queue_callback(void *arg)
{
	struct blkfront_info *info = (struct blkfront_info *)arg;
	schedule_work(&info->work);
}

H
Harvey Harrison 已提交
348
static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
349 350 351 352 353 354 355 356 357 358 359 360 361 362 363
{
	/* 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;
}

A
Al Viro 已提交
364
static int blkif_ioctl(struct block_device *bdev, fmode_t mode,
365
		       unsigned command, unsigned long argument)
366
{
A
Al Viro 已提交
367
	struct blkfront_info *info = bdev->bd_disk->private_data;
368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396
	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;
}

397
/*
398
 * Generate a Xen blkfront IO request from a blk layer request.  Reads
399
 * and writes are handled as expected.
400
 *
401
 * @req: a request struct
402 403 404 405 406 407 408
 */
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;
409
	int i, ref, n;
410
	struct blkif_request_segment *segments = NULL;
411 412 413 414 415 416 417

	/*
	 * 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;
418
	grant_ref_t gref_head;
419
	struct grant *gnt_list_entry = NULL;
420
	struct scatterlist *sg;
421
	int nseg, max_grefs;
422 423 424 425

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

426 427 428 429 430 431 432
	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);
433 434 435

	/* Check if we have enough grants to allocate a requests */
	if (info->persistent_gnts_c < max_grefs) {
436 437
		new_persistent_gnts = 1;
		if (gnttab_alloc_grant_references(
438
		    max_grefs - info->persistent_gnts_c,
439 440 441 442 443
		    &gref_head) < 0) {
			gnttab_request_free_callback(
				&info->callback,
				blkif_restart_queue_callback,
				info,
444
				max_grefs);
445 446 447 448
			return 1;
		}
	} else
		new_persistent_gnts = 0;
449 450 451 452

	/* Fill out a communications ring structure. */
	ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
	id = get_id_from_freelist(info);
453
	info->shadow[id].request = req;
454

455
	if (unlikely(req->cmd_flags & (REQ_DISCARD | REQ_SECURE))) {
456 457
		ring_req->operation = BLKIF_OP_DISCARD;
		ring_req->u.discard.nr_sectors = blk_rq_sectors(req);
458 459
		ring_req->u.discard.id = id;
		ring_req->u.discard.sector_number = (blkif_sector_t)blk_rq_pos(req);
460 461 462 463
		if ((req->cmd_flags & REQ_SECURE) && info->feature_secdiscard)
			ring_req->u.discard.flag = BLKIF_DISCARD_SECURE;
		else
			ring_req->u.discard.flag = 0;
464
	} else {
465 466 467 468
		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);
469
		nseg = blk_rq_map_sg(req->q, req, info->shadow[id].sg);
470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
		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.)
				 */
496 497 498 499 500 501 502 503 504 505 506 507 508
				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;
				}
509 510 511
			}
			ring_req->u.rw.nr_segments = nseg;
		}
512
		for_each_sg(info->shadow[id].sg, sg, nseg, i) {
513 514
			fsect = sg->offset >> 9;
			lsect = fsect + (sg->length >> 9) - 1;
515

516 517
			if ((ring_req->operation == BLKIF_OP_INDIRECT) &&
			    (i % SEGS_PER_INDIRECT_FRAME == 0)) {
518
				unsigned long uninitialized_var(pfn);
519

520 521 522 523
				if (segments)
					kunmap_atomic(segments);

				n = i / SEGS_PER_INDIRECT_FRAME;
524 525 526 527 528 529 530 531 532 533 534
				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);
535 536 537 538 539
				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;
			}

540
			gnt_list_entry = get_grant(&gref_head, page_to_pfn(sg_page(sg)), info);
541
			ref = gnt_list_entry->gref;
542 543 544

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

545
			if (rq_data_dir(req) && info->feature_persistent) {
546 547 548 549 550
				char *bvec_data;
				void *shared_data;

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

551
				shared_data = kmap_atomic(pfn_to_page(gnt_list_entry->pfn));
552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569
				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);
			}
570 571 572 573 574 575 576 577 578
			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] =
579
					(struct blkif_request_segment) {
580 581 582 583
							.gref       = ref,
							.first_sect = fsect,
							.last_sect  = lsect };
			}
584
		}
585 586
		if (segments)
			kunmap_atomic(segments);
587 588 589 590 591 592 593
	}

	info->ring.req_prod_pvt++;

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

594 595
	if (new_persistent_gnts)
		gnttab_free_grant_references(gref_head);
596 597 598 599 600 601 602 603 604 605 606 607 608 609 610

	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);
}

611 612
static inline bool blkif_request_flush_invalid(struct request *req,
					       struct blkfront_info *info)
613 614
{
	return ((req->cmd_type != REQ_TYPE_FS) ||
615 616 617 618
		((req->cmd_flags & REQ_FLUSH) &&
		 !(info->feature_flush & REQ_FLUSH)) ||
		((req->cmd_flags & REQ_FUA) &&
		 !(info->feature_flush & REQ_FUA)));
619 620
}

B
Bob Liu 已提交
621 622
static int blkif_queue_rq(struct blk_mq_hw_ctx *hctx,
			   const struct blk_mq_queue_data *qd)
623
{
B
Bob Liu 已提交
624
	struct blkfront_info *info = qd->rq->rq_disk->private_data;
625

B
Bob Liu 已提交
626 627 628 629
	blk_mq_start_request(qd->rq);
	spin_lock_irq(&info->io_lock);
	if (RING_FULL(&info->ring))
		goto out_busy;
630

B
Bob Liu 已提交
631 632
	if (blkif_request_flush_invalid(qd->rq, info))
		goto out_err;
633

B
Bob Liu 已提交
634 635
	if (blkif_queue_request(qd->rq))
		goto out_busy;
636

B
Bob Liu 已提交
637 638 639
	flush_requests(info);
	spin_unlock_irq(&info->io_lock);
	return BLK_MQ_RQ_QUEUE_OK;
640

B
Bob Liu 已提交
641 642 643
out_err:
	spin_unlock_irq(&info->io_lock);
	return BLK_MQ_RQ_QUEUE_ERROR;
644

B
Bob Liu 已提交
645 646 647 648
out_busy:
	spin_unlock_irq(&info->io_lock);
	blk_mq_stop_hw_queue(hctx);
	return BLK_MQ_RQ_QUEUE_BUSY;
649 650
}

B
Bob Liu 已提交
651 652 653 654 655
static struct blk_mq_ops blkfront_mq_ops = {
	.queue_rq = blkif_queue_rq,
	.map_queue = blk_mq_map_queue,
};

656
static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size,
657
				unsigned int physical_sector_size,
658
				unsigned int segments)
659
{
660
	struct request_queue *rq;
661
	struct blkfront_info *info = gd->private_data;
662

B
Bob Liu 已提交
663 664 665 666 667 668 669 670 671 672
	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))
673
		return -1;
B
Bob Liu 已提交
674 675 676 677 678
	rq = blk_mq_init_queue(&info->tag_set);
	if (IS_ERR(rq)) {
		blk_mq_free_tag_set(&info->tag_set);
		return -1;
	}
679

680
	queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq);
681

682 683 684 685 686
	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;
687 688
		if (info->feature_secdiscard)
			queue_flag_set_unlocked(QUEUE_FLAG_SECDISCARD, rq);
689 690
	}

691
	/* Hard sector size and max sectors impersonate the equiv. hardware. */
692
	blk_queue_logical_block_size(rq, sector_size);
693
	blk_queue_physical_block_size(rq, physical_sector_size);
694
	blk_queue_max_hw_sectors(rq, (segments * PAGE_SIZE) / 512);
695 696 697 698 699 700

	/* 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. */
701
	blk_queue_max_segments(rq, segments);
702 703 704 705

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

706 707 708
	/* Make sure we don't use bounce buffers. */
	blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);

709 710 711 712 713
	gd->queue = rq;

	return 0;
}

714 715 716 717 718 719 720 721 722 723 724
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;";
	}
}
725

726
static void xlvbd_flush(struct blkfront_info *info)
727
{
728
	blk_queue_flush(info->rq, info->feature_flush);
729 730 731 732 733
	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;");
734 735
}

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
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;
}
795

796 797 798 799 800 801 802 803
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;
}

804 805
static int xlvbd_alloc_gendisk(blkif_sector_t capacity,
			       struct blkfront_info *info,
806 807
			       u16 vdisk_info, u16 sector_size,
			       unsigned int physical_sector_size)
808 809 810
{
	struct gendisk *gd;
	int nr_minors = 1;
811
	int err;
812 813 814
	unsigned int offset;
	int minor;
	int nr_parts;
815
	char *ptr;
816 817 818 819

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

820 821 822 823 824 825 826
	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)) {
827 828 829 830
		err = xen_translate_vdev(info->vdevice, &minor, &offset);
		if (err)
			return err;		
 		nr_parts = PARTS_PER_DISK;
831 832 833
	} else {
		minor = BLKIF_MINOR_EXT(info->vdevice);
		nr_parts = PARTS_PER_EXT_DISK;
834
		offset = minor / nr_parts;
835
		if (xen_hvm_domain() && offset < EMULATED_HD_DISK_NAME_OFFSET + 4)
836 837 838
			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);
839
	}
840 841 842 843 844
	if (minor >> MINORBITS) {
		pr_warn("blkfront: %#x's minor (%#x) out of range; ignoring\n",
			info->vdevice, minor);
		return -ENODEV;
	}
845 846 847

	if ((minor % nr_parts) == 0)
		nr_minors = nr_parts;
848

849 850 851 852 853
	err = xlbd_reserve_minors(minor, nr_minors);
	if (err)
		goto out;
	err = -ENODEV;

854 855
	gd = alloc_disk(nr_minors);
	if (gd == NULL)
856
		goto release;
857

858 859 860 861 862 863 864 865
	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));
866 867 868 869 870 871 872 873

	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);

874
	if (xlvbd_init_blk_queue(gd, sector_size, physical_sector_size,
875 876
				 info->max_indirect_segments ? :
				 BLKIF_MAX_SEGMENTS_PER_REQUEST)) {
877
		del_gendisk(gd);
878
		goto release;
879 880 881 882 883
	}

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

884
	xlvbd_flush(info);
885 886 887 888 889 890 891 892 893 894 895 896

	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;

897 898
 release:
	xlbd_release_minors(minor, nr_minors);
899 900 901 902
 out:
	return err;
}

D
Daniel Stodden 已提交
903 904 905 906 907 908 909 910
static void xlvbd_release_gendisk(struct blkfront_info *info)
{
	unsigned int minor, nr_minors;

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

	/* No more blkif_request(). */
B
Bob Liu 已提交
911
	blk_mq_stop_hw_queues(info->rq);
D
Daniel Stodden 已提交
912 913 914 915 916

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

	/* Flush gnttab callback work. Must be done with no locks held. */
917
	flush_work(&info->work);
D
Daniel Stodden 已提交
918 919 920 921 922 923 924 925

	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);
B
Bob Liu 已提交
926
	blk_mq_free_tag_set(&info->tag_set);
D
Daniel Stodden 已提交
927 928 929 930 931 932
	info->rq = NULL;

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

B
Bob Liu 已提交
933
/* Must be called with io_lock holded */
934 935
static void kick_pending_request_queues(struct blkfront_info *info)
{
B
Bob Liu 已提交
936 937
	if (!RING_FULL(&info->ring))
		blk_mq_start_stopped_hw_queues(info->rq, true);
938 939 940 941 942 943
}

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

944
	spin_lock_irq(&info->io_lock);
945 946
	if (info->connected == BLKIF_STATE_CONNECTED)
		kick_pending_request_queues(info);
947
	spin_unlock_irq(&info->io_lock);
948 949 950 951
}

static void blkif_free(struct blkfront_info *info, int suspend)
{
952 953
	struct grant *persistent_gnt;
	struct grant *n;
954
	int i, j, segs;
955

956
	/* Prevent new requests being issued until we fix things up. */
957
	spin_lock_irq(&info->io_lock);
958 959 960 961
	info->connected = suspend ?
		BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
	/* No more blkif_request(). */
	if (info->rq)
B
Bob Liu 已提交
962
		blk_mq_stop_hw_queues(info->rq);
963 964

	/* Remove all persistent grants */
965
	if (!list_empty(&info->grants)) {
966
		list_for_each_entry_safe(persistent_gnt, n,
967
		                         &info->grants, node) {
968
			list_del(&persistent_gnt->node);
969 970 971 972 973
			if (persistent_gnt->gref != GRANT_INVALID_REF) {
				gnttab_end_foreign_access(persistent_gnt->gref,
				                          0, 0UL);
				info->persistent_gnts_c--;
			}
974 975
			if (info->feature_persistent)
				__free_page(pfn_to_page(persistent_gnt->pfn));
976
			kfree(persistent_gnt);
977 978
		}
	}
979
	BUG_ON(info->persistent_gnts_c != 0);
980

981 982 983 984 985 986 987 988 989 990 991 992 993 994
	/*
	 * 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);
		}
	}

B
Bob Liu 已提交
995
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008
		/*
		 * 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);
1009 1010
			if (info->feature_persistent)
				__free_page(pfn_to_page(persistent_gnt->pfn));
1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032
			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;
1033 1034
		kfree(info->shadow[i].sg);
		info->shadow[i].sg = NULL;
1035 1036
	}

1037 1038
	/* No more gnttab callback work. */
	gnttab_cancel_free_callback(&info->callback);
1039
	spin_unlock_irq(&info->io_lock);
1040 1041

	/* Flush gnttab callback work. Must be done with no locks held. */
1042
	flush_work(&info->work);
1043 1044

	/* Free resources associated with old device channel. */
B
Bob Liu 已提交
1045 1046 1047 1048 1049
	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;
		}
1050
	}
B
Bob Liu 已提交
1051 1052 1053
	free_pages((unsigned long)info->ring.sring, get_order(info->nr_ring_pages * PAGE_SIZE));
	info->ring.sring = NULL;

1054 1055 1056 1057 1058 1059
	if (info->irq)
		unbind_from_irqhandler(info->irq, info);
	info->evtchn = info->irq = 0;

}

1060 1061
static void blkif_completion(struct blk_shadow *s, struct blkfront_info *info,
			     struct blkif_response *bret)
1062
{
1063
	int i = 0;
1064
	struct scatterlist *sg;
1065 1066
	char *bvec_data;
	void *shared_data;
1067 1068 1069 1070
	int nseg;

	nseg = s->req.operation == BLKIF_OP_INDIRECT ?
		s->req.u.indirect.nr_segments : s->req.u.rw.nr_segments;
1071

1072
	if (bret->operation == BLKIF_OP_READ && info->feature_persistent) {
1073 1074
		for_each_sg(s->sg, sg, nseg, i) {
			BUG_ON(sg->offset + sg->length > PAGE_SIZE);
1075 1076
			shared_data = kmap_atomic(
				pfn_to_page(s->grants_used[i]->pfn));
1077 1078 1079 1080 1081
			bvec_data = kmap_atomic(sg_page(sg));
			memcpy(bvec_data   + sg->offset,
			       shared_data + sg->offset,
			       sg->length);
			kunmap_atomic(bvec_data);
1082 1083 1084 1085
			kunmap_atomic(shared_data);
		}
	}
	/* Add the persistent grant into the list of free grants */
1086
	for (i = 0; i < nseg; i++) {
1087 1088 1089 1090 1091 1092 1093
		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.
			 */
1094 1095 1096 1097
			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);
1098 1099 1100 1101 1102 1103 1104 1105 1106 1107
			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;
1108
			list_add_tail(&s->grants_used[i]->node, &info->grants);
1109
		}
1110
	}
1111 1112
	if (s->req.operation == BLKIF_OP_INDIRECT) {
		for (i = 0; i < INDIRECT_GREFS(nseg); i++) {
1113
			if (gnttab_query_foreign_access(s->indirect_grants[i]->gref)) {
1114 1115 1116 1117
				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);
1118 1119
				info->persistent_gnts_c++;
			} else {
1120 1121
				struct page *indirect_page;

1122
				gnttab_end_foreign_access(s->indirect_grants[i]->gref, 0, 0UL);
1123 1124 1125 1126
				/*
				 * Add the used indirect page back to the list of
				 * available pages for indirect grefs.
				 */
1127 1128 1129 1130
				if (!info->feature_persistent) {
					indirect_page = pfn_to_page(s->indirect_grants[i]->pfn);
					list_add(&indirect_page->lru, &info->indirect_pages);
				}
1131
				s->indirect_grants[i]->gref = GRANT_INVALID_REF;
1132
				list_add_tail(&s->indirect_grants[i]->node, &info->grants);
1133
			}
1134 1135
		}
	}
1136 1137 1138 1139 1140 1141 1142 1143 1144
}

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;
1145
	int error;
1146

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.
		 */
B
Bob Liu 已提交
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
		error = (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
				error = -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, error);
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
				error = -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
				error = -EOPNOTSUPP;
1213
			}
1214 1215 1216
			if (unlikely(error)) {
				if (error == -EOPNOTSUPP)
					error = 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, error);
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;
B
Bob Liu 已提交
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

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

B
Bob Liu 已提交
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);
B
Bob Liu 已提交
1270
	FRONT_RING_INIT(&info->ring, sring, ring_size);
1271

B
Bob Liu 已提交
1272
	err = xenbus_grant_ring(dev, info->ring.sring, info->nr_ring_pages, gref);
1273
	if (err < 0) {
B
Bob Liu 已提交
1274
		free_pages((unsigned long)sring, get_order(ring_size));
1275 1276 1277
		info->ring.sring = NULL;
		goto fail;
	}
B
Bob Liu 已提交
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;
B
Bob Liu 已提交
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;
	}

B
Bob Liu 已提交
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;
	}

B
Bob Liu 已提交
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)
{
B
Bob Liu 已提交
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
static void split_bio_end(struct bio *bio)
1480 1481 1482 1483 1484
{
	struct split_bio *split_bio = bio->bi_private;

	if (atomic_dec_and_test(&split_bio->pending)) {
		split_bio->bio->bi_phys_segments = 0;
1485 1486
		split_bio->bio->bi_error = bio->bi_error;
		bio_endio(split_bio->bio);
1487 1488 1489 1490
		kfree(split_bio);
	}
	bio_put(bio);
}
1491 1492 1493 1494

static int blkif_recover(struct blkfront_info *info)
{
	int i;
1495
	struct request *req, *n;
1496
	struct blk_shadow *copy;
1497 1498 1499 1500 1501 1502 1503
	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;
1504 1505

	/* Stage 1: Make a safe copy of the shadow state. */
1506
	copy = kmemdup(info->shadow, sizeof(info->shadow),
1507
		       GFP_NOIO | __GFP_REPEAT | __GFP_HIGH);
1508 1509 1510 1511 1512
	if (!copy)
		return -ENOMEM;

	/* Stage 2: Set up free list. */
	memset(&info->shadow, 0, sizeof(info->shadow));
B
Bob Liu 已提交
1513
	for (i = 0; i < BLK_RING_SIZE(info); i++)
1514
		info->shadow[i].req.u.rw.id = i+1;
1515
	info->shadow_free = info->ring.req_prod_pvt;
B
Bob Liu 已提交
1516
	info->shadow[BLK_RING_SIZE(info)-1].req.u.rw.id = 0x0fffffff;
1517

1518
	rc = blkfront_gather_backend_features(info);
1519 1520 1521 1522 1523 1524 1525 1526 1527
	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);
B
Bob Liu 已提交
1528
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1529
		/* Not in use? */
1530
		if (!copy[i].request)
1531 1532
			continue;

1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543
		/*
		 * 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;
1544
		}
1545 1546 1547 1548
		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;
1549
		blk_end_request_all(copy[i].request, 0);
1550 1551 1552 1553 1554 1555
	}

	kfree(copy);

	xenbus_switch_state(info->xbdev, XenbusStateConnected);

1556
	spin_lock_irq(&info->io_lock);
1557 1558 1559 1560 1561 1562 1563

	/* 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);

1564 1565 1566 1567
	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);
B
Bob Liu 已提交
1568
		blk_mq_requeue_request(req);
1569
	}
1570
	spin_unlock_irq(&info->io_lock);
B
Bob Liu 已提交
1571
	blk_mq_kick_requeue_list(info->rq);
1572

1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587
	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,
1588
					   (unsigned int)bio_sectors(bio) - offset);
1589 1590
				cloned_bio = bio_clone(bio, GFP_NOIO);
				BUG_ON(cloned_bio == NULL);
1591
				bio_trim(cloned_bio, offset, size);
1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605
				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);
	}

1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616
	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)
{
1617
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1618 1619 1620 1621 1622 1623
	int err;

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

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

1624
	err = talk_to_blkback(dev, info);
1625 1626 1627 1628 1629 1630

	/*
	 * We have to wait for the backend to switch to
	 * connected state, since we want to read which
	 * features it supports.
	 */
1631 1632 1633 1634

	return err;
}

1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659
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);

1660
	if (bdev->bd_openers) {
1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671
		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);
}
1672

1673 1674 1675 1676 1677
static void blkfront_setup_discard(struct blkfront_info *info)
{
	int err;
	unsigned int discard_granularity;
	unsigned int discard_alignment;
1678
	unsigned int discard_secure;
1679

1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693
	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;
1694 1695
}

1696 1697
static int blkfront_setup_indirect(struct blkfront_info *info)
{
1698
	unsigned int segs;
1699 1700
	int err, i;

1701
	if (info->max_indirect_segments == 0)
1702
		segs = BLKIF_MAX_SEGMENTS_PER_REQUEST;
1703
	else
1704 1705
		segs = info->max_indirect_segments;

B
Bob Liu 已提交
1706
	err = fill_grant_buffer(info, (segs + INDIRECT_GREFS(segs)) * BLK_RING_SIZE(info));
1707 1708 1709
	if (err)
		goto out_of_memory;

1710 1711 1712 1713 1714 1715
	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
		 */
B
Bob Liu 已提交
1716
		int num = INDIRECT_GREFS(segs) * BLK_RING_SIZE(info);
1717 1718 1719 1720 1721 1722 1723 1724 1725 1726

		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);
		}
	}

B
Bob Liu 已提交
1727
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1728 1729 1730
		info->shadow[i].grants_used = kzalloc(
			sizeof(info->shadow[i].grants_used[0]) * segs,
			GFP_NOIO);
1731
		info->shadow[i].sg = kzalloc(sizeof(info->shadow[i].sg[0]) * segs, GFP_NOIO);
1732 1733 1734 1735 1736 1737
		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) ||
1738
			(info->shadow[i].sg == NULL) ||
1739 1740 1741
		     (info->max_indirect_segments &&
		     (info->shadow[i].indirect_grants == NULL)))
			goto out_of_memory;
1742
		sg_init_table(info->shadow[i].sg, segs);
1743 1744 1745 1746 1747 1748
	}


	return 0;

out_of_memory:
B
Bob Liu 已提交
1749
	for (i = 0; i < BLK_RING_SIZE(info); i++) {
1750 1751
		kfree(info->shadow[i].grants_used);
		info->shadow[i].grants_used = NULL;
1752 1753
		kfree(info->shadow[i].sg);
		info->shadow[i].sg = NULL;
1754 1755 1756
		kfree(info->shadow[i].indirect_grants);
		info->shadow[i].indirect_grants = NULL;
	}
1757 1758 1759 1760 1761 1762 1763
	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);
		}
	}
1764 1765 1766
	return -ENOMEM;
}

1767 1768 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
/*
 * 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);
}

1829 1830 1831 1832 1833 1834 1835 1836
/*
 * 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;
1837
	unsigned int physical_sector_size;
1838 1839 1840
	unsigned int binfo;
	int err;

1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853
	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);
1854
		revalidate_disk(info->gd);
1855

1856
		return;
1857
	case BLKIF_STATE_SUSPENDED:
1858 1859 1860 1861 1862 1863 1864
		/*
		 * 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);
1865 1866
		return;

1867 1868
	default:
		break;
1869
	}
1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885

	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;
	}

1886 1887 1888 1889 1890 1891 1892 1893 1894 1895
	/*
	 * 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;

1896
	err = blkfront_gather_backend_features(info);
1897 1898 1899 1900 1901 1902
	if (err) {
		xenbus_dev_fatal(info->xbdev, err, "setup_indirect at %s",
				 info->xbdev->otherend);
		return;
	}

1903 1904
	err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size,
				  physical_sector_size);
1905 1906 1907 1908 1909 1910 1911 1912 1913
	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. */
1914
	spin_lock_irq(&info->io_lock);
1915 1916
	info->connected = BLKIF_STATE_CONNECTED;
	kick_pending_request_queues(info);
1917
	spin_unlock_irq(&info->io_lock);
1918 1919

	add_disk(info->gd);
1920 1921

	info->is_ready = 1;
1922 1923 1924 1925 1926
}

/**
 * Callback received when the backend's state changes.
 */
1927
static void blkback_changed(struct xenbus_device *dev,
1928 1929
			    enum xenbus_state backend_state)
{
1930
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1931

1932
	dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state);
1933 1934 1935

	switch (backend_state) {
	case XenbusStateInitWait:
1936 1937
		if (dev->state != XenbusStateInitialising)
			break;
1938 1939 1940 1941 1942 1943
		if (talk_to_blkback(dev, info)) {
			kfree(info);
			dev_set_drvdata(&dev->dev, NULL);
			break;
		}
	case XenbusStateInitialising:
1944
	case XenbusStateInitialised:
1945 1946
	case XenbusStateReconfiguring:
	case XenbusStateReconfigured:
1947 1948 1949 1950 1951 1952 1953
	case XenbusStateUnknown:
		break;

	case XenbusStateConnected:
		blkfront_connect(info);
		break;

1954 1955 1956 1957
	case XenbusStateClosed:
		if (dev->state == XenbusStateClosed)
			break;
		/* Missed the backend's Closing state -- fallthrough */
1958
	case XenbusStateClosing:
1959
		blkfront_closing(info);
1960 1961 1962 1963
		break;
	}
}

1964
static int blkfront_remove(struct xenbus_device *xbdev)
1965
{
1966 1967 1968
	struct blkfront_info *info = dev_get_drvdata(&xbdev->dev);
	struct block_device *bdev = NULL;
	struct gendisk *disk;
1969

1970
	dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename);
1971 1972 1973

	blkif_free(info, 0);

1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996
	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;

1997 1998 1999 2000
	dev_warn(disk_to_dev(disk),
		 "%s was hot-unplugged, %d stale handles\n",
		 xbdev->nodename, bdev->bd_openers);

2001
	if (info && !bdev->bd_openers) {
2002 2003
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
2004
		kfree(info);
2005 2006 2007 2008
	}

	mutex_unlock(&bdev->bd_mutex);
	bdput(bdev);
2009 2010 2011 2012

	return 0;
}

2013 2014
static int blkfront_is_ready(struct xenbus_device *dev)
{
2015
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
2016

2017
	return info->is_ready && info->xbdev;
2018 2019
}

A
Al Viro 已提交
2020
static int blkif_open(struct block_device *bdev, fmode_t mode)
2021
{
2022 2023 2024
	struct gendisk *disk = bdev->bd_disk;
	struct blkfront_info *info;
	int err = 0;
2025

2026
	mutex_lock(&blkfront_mutex);
2027

2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043
	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:
2044
	mutex_unlock(&blkfront_mutex);
2045
	return err;
2046 2047
}

2048
static void blkif_release(struct gendisk *disk, fmode_t mode)
2049
{
A
Al Viro 已提交
2050
	struct blkfront_info *info = disk->private_data;
2051 2052 2053
	struct block_device *bdev;
	struct xenbus_device *xbdev;

2054
	mutex_lock(&blkfront_mutex);
2055 2056 2057

	bdev = bdget_disk(disk, 0);

2058 2059 2060 2061
	if (!bdev) {
		WARN(1, "Block device %s yanked out from us!\n", disk->disk_name);
		goto out_mutex;
	}
2062 2063 2064
	if (bdev->bd_openers)
		goto out;

2065 2066 2067 2068 2069 2070 2071 2072 2073 2074
	/*
	 * 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 */
2075
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
2076 2077 2078 2079 2080 2081 2082 2083
		xlvbd_release_gendisk(info);
		xenbus_frontend_closed(info->xbdev);
 	}

	mutex_unlock(&info->mutex);

	if (!xbdev) {
		/* sudden device removal */
2084
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
2085 2086 2087
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
		kfree(info);
2088
	}
2089

J
Jens Axboe 已提交
2090
out:
2091
	bdput(bdev);
2092
out_mutex:
2093
	mutex_unlock(&blkfront_mutex);
2094 2095
}

2096
static const struct block_device_operations xlvbd_block_fops =
2097 2098
{
	.owner = THIS_MODULE,
A
Al Viro 已提交
2099 2100
	.open = blkif_open,
	.release = blkif_release,
2101
	.getgeo = blkif_getgeo,
2102
	.ioctl = blkif_ioctl,
2103 2104 2105
};


2106
static const struct xenbus_device_id blkfront_ids[] = {
2107 2108 2109 2110
	{ "vbd" },
	{ "" }
};

2111 2112
static struct xenbus_driver blkfront_driver = {
	.ids  = blkfront_ids,
2113 2114 2115
	.probe = blkfront_probe,
	.remove = blkfront_remove,
	.resume = blkfront_resume,
2116
	.otherend_changed = blkback_changed,
2117
	.is_ready = blkfront_is_ready,
2118
};
2119 2120 2121

static int __init xlblk_init(void)
{
2122 2123
	int ret;

2124
	if (!xen_domain())
2125 2126
		return -ENODEV;

B
Bob Liu 已提交
2127 2128 2129 2130 2131 2132
	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;
	}

2133
	if (!xen_has_pv_disk_devices())
2134 2135
		return -ENODEV;

2136 2137 2138 2139 2140 2141
	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;
	}

2142
	ret = xenbus_register_frontend(&blkfront_driver);
2143 2144 2145 2146 2147 2148
	if (ret) {
		unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
		return ret;
	}

	return 0;
2149 2150 2151 2152
}
module_init(xlblk_init);


2153
static void __exit xlblk_exit(void)
2154
{
2155 2156 2157
	xenbus_unregister_driver(&blkfront_driver);
	unregister_blkdev(XENVBD_MAJOR, DEV_NAME);
	kfree(minors);
2158 2159 2160 2161 2162 2163
}
module_exit(xlblk_exit);

MODULE_DESCRIPTION("Xen virtual block device frontend");
MODULE_LICENSE("GPL");
MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
2164
MODULE_ALIAS("xen:vbd");
2165
MODULE_ALIAS("xenblk");