xen-blkfront.c 31.2 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/hdreg.h>
41
#include <linux/cdrom.h>
42
#include <linux/module.h>
43
#include <linux/slab.h>
44
#include <linux/mutex.h>
45
#include <linux/scatterlist.h>
46

47
#include <xen/xen.h>
48 49 50 51
#include <xen/xenbus.h>
#include <xen/grant_table.h>
#include <xen/events.h>
#include <xen/page.h>
52
#include <xen/platform_pci.h>
53 54 55

#include <xen/interface/grant_table.h>
#include <xen/interface/io/blkif.h>
56
#include <xen/interface/io/protocols.h>
57 58 59 60 61 62 63 64 65 66 67

#include <asm/xen/hypervisor.h>

enum blkif_state {
	BLKIF_STATE_DISCONNECTED,
	BLKIF_STATE_CONNECTED,
	BLKIF_STATE_SUSPENDED,
};

struct blk_shadow {
	struct blkif_request req;
68
	struct request *request;
69 70 71
	unsigned long frame[BLKIF_MAX_SEGMENTS_PER_REQUEST];
};

72
static DEFINE_MUTEX(blkfront_mutex);
73
static const struct block_device_operations xlvbd_block_fops;
74 75 76 77 78 79 80 81 82 83

#define BLK_RING_SIZE __RING_SIZE((struct blkif_sring *)0, PAGE_SIZE)

/*
 * 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
{
84
	struct mutex mutex;
85 86 87 88 89 90 91
	struct xenbus_device *xbdev;
	struct gendisk *gd;
	int vdevice;
	blkif_vdev_t handle;
	enum blkif_state connected;
	int ring_ref;
	struct blkif_front_ring ring;
92
	struct scatterlist sg[BLKIF_MAX_SEGMENTS_PER_REQUEST];
93 94 95 96 97 98
	unsigned int evtchn, irq;
	struct request_queue *rq;
	struct work_struct work;
	struct gnttab_free_callback callback;
	struct blk_shadow shadow[BLK_RING_SIZE];
	unsigned long shadow_free;
99
	unsigned int feature_flush;
100
	int is_ready;
101 102 103 104
};

static DEFINE_SPINLOCK(blkif_io_lock);

105 106 107 108
static unsigned int nr_minors;
static unsigned long *minors;
static DEFINE_SPINLOCK(minor_lock);

109 110 111 112 113
#define MAXIMUM_OUTSTANDING_BLOCK_REQS \
	(BLKIF_MAX_SEGMENTS_PER_REQUEST * BLK_RING_SIZE)
#define GRANT_INVALID_REF	0

#define PARTS_PER_DISK		16
114
#define PARTS_PER_EXT_DISK      256
115 116 117 118

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

119 120 121 122
#define EXT_SHIFT 28
#define EXTENDED (1<<EXT_SHIFT)
#define VDEV_IS_EXTENDED(dev) ((dev)&(EXTENDED))
#define BLKIF_MINOR_EXT(dev) ((dev)&(~EXTENDED))
123

124
#define DEV_NAME	"xvd"	/* name in /dev */
125 126 127 128

static int get_id_from_freelist(struct blkfront_info *info)
{
	unsigned long free = info->shadow_free;
129
	BUG_ON(free >= BLK_RING_SIZE);
130 131 132 133 134 135 136 137 138
	info->shadow_free = info->shadow[free].req.id;
	info->shadow[free].req.id = 0x0fffffee; /* debug */
	return free;
}

static void add_id_to_freelist(struct blkfront_info *info,
			       unsigned long id)
{
	info->shadow[id].req.id  = info->shadow_free;
139
	info->shadow[id].request = NULL;
140 141 142
	info->shadow_free = id;
}

143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191
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;

		bitmap = kzalloc(BITS_TO_LONGS(end) * sizeof(*bitmap),
				 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) {
		for (; minor < end; ++minor)
			__set_bit(minor, minors);
		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);
	for (; minor < end; ++minor)
		__clear_bit(minor, minors);
	spin_unlock(&minor_lock);
}

192 193 194 195 196 197
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
198
static int blkif_getgeo(struct block_device *bd, struct hd_geometry *hg)
199 200 201 202 203 204 205 206 207 208 209 210 211 212 213
{
	/* 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
214
static int blkif_ioctl(struct block_device *bdev, fmode_t mode,
215
		       unsigned command, unsigned long argument)
216
{
A
Al Viro 已提交
217
	struct blkfront_info *info = bdev->bd_disk->private_data;
218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246
	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;
}

247
/*
248 249 250
 * Generate a Xen blkfront IO request from a blk layer request.  Reads
 * and writes are handled as expected.  Since we lack a loose flush
 * request, we map flushes into a full ordered barrier.
251
 *
252
 * @req: a request struct
253 254 255 256 257 258 259 260
 */
static int blkif_queue_request(struct request *req)
{
	struct blkfront_info *info = req->rq_disk->private_data;
	unsigned long buffer_mfn;
	struct blkif_request *ring_req;
	unsigned long id;
	unsigned int fsect, lsect;
261
	int i, ref;
262
	grant_ref_t gref_head;
263
	struct scatterlist *sg;
264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280

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

	if (gnttab_alloc_grant_references(
		BLKIF_MAX_SEGMENTS_PER_REQUEST, &gref_head) < 0) {
		gnttab_request_free_callback(
			&info->callback,
			blkif_restart_queue_callback,
			info,
			BLKIF_MAX_SEGMENTS_PER_REQUEST);
		return 1;
	}

	/* Fill out a communications ring structure. */
	ring_req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
	id = get_id_from_freelist(info);
281
	info->shadow[id].request = req;
282 283

	ring_req->id = id;
284
	ring_req->sector_number = (blkif_sector_t)blk_rq_pos(req);
285 286 287 288
	ring_req->handle = info->handle;

	ring_req->operation = rq_data_dir(req) ?
		BLKIF_OP_WRITE : BLKIF_OP_READ;
289 290 291 292 293 294 295 296 297 298

	if (req->cmd_flags & (REQ_FLUSH | REQ_FUA)) {
		/*
		 * Ideally we could just do an unordered
		 * flush-to-disk, but all we have is a full write
		 * barrier at the moment.  However, a barrier write is
		 * 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.)
		 */
299
		ring_req->operation = BLKIF_OP_WRITE_BARRIER;
300
	}
301

302 303 304 305 306 307 308
	ring_req->nr_segments = blk_rq_map_sg(req->q, req, info->sg);
	BUG_ON(ring_req->nr_segments > BLKIF_MAX_SEGMENTS_PER_REQUEST);

	for_each_sg(info->sg, sg, ring_req->nr_segments, i) {
		buffer_mfn = pfn_to_mfn(page_to_pfn(sg_page(sg)));
		fsect = sg->offset >> 9;
		lsect = fsect + (sg->length >> 9) - 1;
309 310 311 312 313
		/* install a grant reference. */
		ref = gnttab_claim_grant_reference(&gref_head);
		BUG_ON(ref == -ENOSPC);

		gnttab_grant_foreign_access_ref(
314 315 316 317 318
				ref,
				info->xbdev->otherend_id,
				buffer_mfn,
				rq_data_dir(req) );

319 320
		info->shadow[id].frame[i] = mfn_to_pfn(buffer_mfn);
		ring_req->seg[i] =
321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351
				(struct blkif_request_segment) {
					.gref       = ref,
					.first_sect = fsect,
					.last_sect  = lsect };
	}

	info->ring.req_prod_pvt++;

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

	gnttab_free_grant_references(gref_head);

	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);
}

/*
 * do_blkif_request
 *  read a block; request is in a request queue
 */
352
static void do_blkif_request(struct request_queue *rq)
353 354 355 356 357 358 359 360 361
{
	struct blkfront_info *info = NULL;
	struct request *req;
	int queued;

	pr_debug("Entered do_blkif_request\n");

	queued = 0;

362
	while ((req = blk_peek_request(rq)) != NULL) {
363 364 365 366 367
		info = req->rq_disk->private_data;

		if (RING_FULL(&info->ring))
			goto wait;

368
		blk_start_request(req);
369

370
		if (req->cmd_type != REQ_TYPE_FS) {
371 372 373 374
			__blk_end_request_all(req, -EIO);
			continue;
		}

375
		pr_debug("do_blk_req %p: cmd %p, sec %lx, "
376 377 378 379
			 "(%u/%u) buffer:%p [%s]\n",
			 req, req->cmd, (unsigned long)blk_rq_pos(req),
			 blk_rq_cur_sectors(req), blk_rq_sectors(req),
			 req->buffer, rq_data_dir(req) ? "write" : "read");
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397

		if (blkif_queue_request(req)) {
			blk_requeue_request(rq, req);
wait:
			/* Avoid pointless unplugs. */
			blk_stop_queue(rq);
			break;
		}

		queued++;
	}

	if (queued != 0)
		flush_requests(info);
}

static int xlvbd_init_blk_queue(struct gendisk *gd, u16 sector_size)
{
398
	struct request_queue *rq;
399 400 401 402 403

	rq = blk_init_queue(do_blkif_request, &blkif_io_lock);
	if (rq == NULL)
		return -1;

404
	queue_flag_set_unlocked(QUEUE_FLAG_VIRT, rq);
405 406

	/* Hard sector size and max sectors impersonate the equiv. hardware. */
407
	blk_queue_logical_block_size(rq, sector_size);
408
	blk_queue_max_hw_sectors(rq, 512);
409 410 411 412 413 414

	/* 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. */
415
	blk_queue_max_segments(rq, BLKIF_MAX_SEGMENTS_PER_REQUEST);
416 417 418 419

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

420 421 422
	/* Make sure we don't use bounce buffers. */
	blk_queue_bounce_limit(rq, BLK_BOUNCE_ANY);

423 424 425 426 427 428
	gd->queue = rq;

	return 0;
}


429
static void xlvbd_flush(struct blkfront_info *info)
430
{
431
	blk_queue_flush(info->rq, info->feature_flush);
432
	printk(KERN_INFO "blkfront: %s: barriers %s\n",
433 434
	       info->gd->disk_name,
	       info->feature_flush ? "enabled" : "disabled");
435 436 437
}


438 439 440
static int xlvbd_alloc_gendisk(blkif_sector_t capacity,
			       struct blkfront_info *info,
			       u16 vdisk_info, u16 sector_size)
441 442 443 444
{
	struct gendisk *gd;
	int nr_minors = 1;
	int err = -ENODEV;
445 446 447
	unsigned int offset;
	int minor;
	int nr_parts;
448 449 450 451

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

452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467
	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)) {
		minor = BLKIF_MINOR(info->vdevice);
		nr_parts = PARTS_PER_DISK;
	} else {
		minor = BLKIF_MINOR_EXT(info->vdevice);
		nr_parts = PARTS_PER_EXT_DISK;
	}

	if ((minor % nr_parts) == 0)
		nr_minors = nr_parts;
468

469 470 471 472 473
	err = xlbd_reserve_minors(minor, nr_minors);
	if (err)
		goto out;
	err = -ENODEV;

474 475
	gd = alloc_disk(nr_minors);
	if (gd == NULL)
476
		goto release;
477

478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496
	offset = minor / nr_parts;

	if (nr_minors > 1) {
		if (offset < 26)
			sprintf(gd->disk_name, "%s%c", DEV_NAME, 'a' + offset);
		else
			sprintf(gd->disk_name, "%s%c%c", DEV_NAME,
				'a' + ((offset / 26)-1), 'a' + (offset % 26));
	} else {
		if (offset < 26)
			sprintf(gd->disk_name, "%s%c%d", DEV_NAME,
				'a' + offset,
				minor & (nr_parts - 1));
		else
			sprintf(gd->disk_name, "%s%c%c%d", DEV_NAME,
				'a' + ((offset / 26) - 1),
				'a' + (offset % 26),
				minor & (nr_parts - 1));
	}
497 498 499 500 501 502 503 504 505 506

	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);

	if (xlvbd_init_blk_queue(gd, sector_size)) {
		del_gendisk(gd);
507
		goto release;
508 509 510 511 512
	}

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

513
	xlvbd_flush(info);
514 515 516 517 518 519 520 521 522 523 524 525

	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;

526 527
 release:
	xlbd_release_minors(minor, nr_minors);
528 529 530 531
 out:
	return err;
}

D
Daniel Stodden 已提交
532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564
static void xlvbd_release_gendisk(struct blkfront_info *info)
{
	unsigned int minor, nr_minors;
	unsigned long flags;

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

	spin_lock_irqsave(&blkif_io_lock, flags);

	/* No more blkif_request(). */
	blk_stop_queue(info->rq);

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

	/* Flush gnttab callback work. Must be done with no locks held. */
	flush_scheduled_work();

	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);
	info->rq = NULL;

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

565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627
static void kick_pending_request_queues(struct blkfront_info *info)
{
	if (!RING_FULL(&info->ring)) {
		/* Re-enable calldowns. */
		blk_start_queue(info->rq);
		/* Kick things off immediately. */
		do_blkif_request(info->rq);
	}
}

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

	spin_lock_irq(&blkif_io_lock);
	if (info->connected == BLKIF_STATE_CONNECTED)
		kick_pending_request_queues(info);
	spin_unlock_irq(&blkif_io_lock);
}

static void blkif_free(struct blkfront_info *info, int suspend)
{
	/* Prevent new requests being issued until we fix things up. */
	spin_lock_irq(&blkif_io_lock);
	info->connected = suspend ?
		BLKIF_STATE_SUSPENDED : BLKIF_STATE_DISCONNECTED;
	/* No more blkif_request(). */
	if (info->rq)
		blk_stop_queue(info->rq);
	/* No more gnttab callback work. */
	gnttab_cancel_free_callback(&info->callback);
	spin_unlock_irq(&blkif_io_lock);

	/* Flush gnttab callback work. Must be done with no locks held. */
	flush_scheduled_work();

	/* Free resources associated with old device channel. */
	if (info->ring_ref != GRANT_INVALID_REF) {
		gnttab_end_foreign_access(info->ring_ref, 0,
					  (unsigned long)info->ring.sring);
		info->ring_ref = GRANT_INVALID_REF;
		info->ring.sring = NULL;
	}
	if (info->irq)
		unbind_from_irqhandler(info->irq, info);
	info->evtchn = info->irq = 0;

}

static void blkif_completion(struct blk_shadow *s)
{
	int i;
	for (i = 0; i < s->req.nr_segments; i++)
		gnttab_end_foreign_access(s->req.seg[i].gref, 0, 0UL);
}

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;
628
	int error;
629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645

	spin_lock_irqsave(&blkif_io_lock, flags);

	if (unlikely(info->connected != BLKIF_STATE_CONNECTED)) {
		spin_unlock_irqrestore(&blkif_io_lock, flags);
		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;
646
		req  = info->shadow[id].request;
647 648 649 650 651

		blkif_completion(&info->shadow[id]);

		add_id_to_freelist(info, id);

652
		error = (bret->status == BLKIF_RSP_OKAY) ? 0 : -EIO;
653 654 655 656 657
		switch (bret->operation) {
		case BLKIF_OP_WRITE_BARRIER:
			if (unlikely(bret->status == BLKIF_RSP_EOPNOTSUPP)) {
				printk(KERN_WARNING "blkfront: %s: write barrier op failed\n",
				       info->gd->disk_name);
658
				error = -EOPNOTSUPP;
659 660
				info->feature_flush = 0;
				xlvbd_flush(info);
661 662 663 664 665 666 667 668
			}
			/* 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);

669
			__blk_end_request_all(req, error);
670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701
			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);

	spin_unlock_irqrestore(&blkif_io_lock, flags);

	return IRQ_HANDLED;
}


static int setup_blkring(struct xenbus_device *dev,
			 struct blkfront_info *info)
{
	struct blkif_sring *sring;
	int err;

	info->ring_ref = GRANT_INVALID_REF;

702
	sring = (struct blkif_sring *)__get_free_page(GFP_NOIO | __GFP_HIGH);
703 704 705 706 707 708
	if (!sring) {
		xenbus_dev_fatal(dev, -ENOMEM, "allocating shared ring");
		return -ENOMEM;
	}
	SHARED_RING_INIT(sring);
	FRONT_RING_INIT(&info->ring, sring, PAGE_SIZE);
709 710

	sg_init_table(info->sg, BLKIF_MAX_SEGMENTS_PER_REQUEST);
711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741

	err = xenbus_grant_ring(dev, virt_to_mfn(info->ring.sring));
	if (err < 0) {
		free_page((unsigned long)sring);
		info->ring.sring = NULL;
		goto fail;
	}
	info->ring_ref = err;

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

	err = bind_evtchn_to_irqhandler(info->evtchn,
					blkif_interrupt,
					IRQF_SAMPLE_RANDOM, "blkif", info);
	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. */
742
static int talk_to_blkback(struct xenbus_device *dev,
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
			   struct blkfront_info *info)
{
	const char *message = NULL;
	struct xenbus_transaction xbt;
	int err;

	/* 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;
	}

	err = xenbus_printf(xbt, dev->nodename,
			    "ring-ref", "%u", info->ring_ref);
	if (err) {
		message = "writing ring-ref";
		goto abort_transaction;
	}
	err = xenbus_printf(xbt, dev->nodename,
			    "event-channel", "%u", info->evtchn);
	if (err) {
		message = "writing event-channel";
		goto abort_transaction;
	}
773 774 775 776 777 778
	err = xenbus_printf(xbt, dev->nodename, "protocol", "%s",
			    XEN_IO_PROTO_ABI_NATIVE);
	if (err) {
		message = "writing protocol";
		goto abort_transaction;
	}
779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817

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

	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)
{
	int err, vdevice, i;
	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) {
818 819 820 821 822 823 824
		/* 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;
		}
825 826
	}

827 828 829 830
	if (xen_hvm_domain()) {
		char *type;
		int len;
		/* no unplug has been done: do not hook devices != xen vbds */
831
		if (xen_platform_pci_unplug & XEN_UNPLUG_UNNECESSARY) {
832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851
			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",
						__FUNCTION__, vdevice);
				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);
852 853
			return -ENODEV;
		}
854
		kfree(type);
855
	}
856 857 858 859 860 861
	info = kzalloc(sizeof(*info), GFP_KERNEL);
	if (!info) {
		xenbus_dev_fatal(dev, -ENOMEM, "allocating info structure");
		return -ENOMEM;
	}

862
	mutex_init(&info->mutex);
863 864 865 866 867 868 869 870 871 872 873
	info->xbdev = dev;
	info->vdevice = vdevice;
	info->connected = BLKIF_STATE_DISCONNECTED;
	INIT_WORK(&info->work, blkif_restart_queue);

	for (i = 0; i < BLK_RING_SIZE; i++)
		info->shadow[i].req.id = i+1;
	info->shadow[BLK_RING_SIZE-1].req.id = 0x0fffffff;

	/* Front end dir is a number, which is used as the id. */
	info->handle = simple_strtoul(strrchr(dev->nodename, '/')+1, NULL, 0);
874
	dev_set_drvdata(&dev->dev, info);
875

876
	err = talk_to_blkback(dev, info);
877 878
	if (err) {
		kfree(info);
879
		dev_set_drvdata(&dev->dev, NULL);
880 881 882 883 884 885 886 887 888 889 890 891 892 893 894
		return err;
	}

	return 0;
}


static int blkif_recover(struct blkfront_info *info)
{
	int i;
	struct blkif_request *req;
	struct blk_shadow *copy;
	int j;

	/* Stage 1: Make a safe copy of the shadow state. */
895 896
	copy = kmalloc(sizeof(info->shadow),
		       GFP_NOIO | __GFP_REPEAT | __GFP_HIGH);
897 898 899 900 901 902 903 904 905 906 907 908 909 910
	if (!copy)
		return -ENOMEM;
	memcpy(copy, info->shadow, sizeof(info->shadow));

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

	/* Stage 3: Find pending requests and requeue them. */
	for (i = 0; i < BLK_RING_SIZE; i++) {
		/* Not in use? */
911
		if (!copy[i].request)
912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927
			continue;

		/* Grab a request slot and copy shadow state into it. */
		req = RING_GET_REQUEST(&info->ring, info->ring.req_prod_pvt);
		*req = copy[i].req;

		/* We get a new request id, and must reset the shadow state. */
		req->id = get_id_from_freelist(info);
		memcpy(&info->shadow[req->id], &copy[i], sizeof(copy[i]));

		/* Rewrite any grant references invalidated by susp/resume. */
		for (j = 0; j < req->nr_segments; j++)
			gnttab_grant_foreign_access_ref(
				req->seg[j].gref,
				info->xbdev->otherend_id,
				pfn_to_mfn(info->shadow[req->id].frame[j]),
928
				rq_data_dir(info->shadow[req->id].request));
929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961
		info->shadow[req->id].req = *req;

		info->ring.req_prod_pvt++;
	}

	kfree(copy);

	xenbus_switch_state(info->xbdev, XenbusStateConnected);

	spin_lock_irq(&blkif_io_lock);

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

	/* Send off requeued requests */
	flush_requests(info);

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

	spin_unlock_irq(&blkif_io_lock);

	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)
{
962
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
963 964 965 966 967 968
	int err;

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

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

969
	err = talk_to_blkback(dev, info);
970 971 972 973 974 975
	if (info->connected == BLKIF_STATE_SUSPENDED && !err)
		err = blkif_recover(info);

	return err;
}

976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000
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);

1001
	if (bdev->bd_openers) {
1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012
		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);
}
1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023

/*
 * 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;
	unsigned int binfo;
	int err;
1024
	int barrier;
1025

1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038
	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);
1039
		revalidate_disk(info->gd);
1040 1041 1042

		/* fall through */
	case BLKIF_STATE_SUSPENDED:
1043 1044
		return;

1045 1046
	default:
		break;
1047
	}
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064

	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;
	}

	err = xenbus_gather(XBT_NIL, info->xbdev->otherend,
1065
			    "feature-barrier", "%lu", &barrier,
1066
			    NULL);
1067 1068 1069 1070

	/*
	 * If there's no "feature-barrier" defined, then it means
	 * we're dealing with a very old backend which writes
1071
	 * synchronously; nothing to do.
1072
	 *
T
Tejun Heo 已提交
1073
	 * If there are barriers, then we use flush.
1074
	 */
1075
	info->feature_flush = 0;
1076

1077
	if (!err && barrier)
1078
		info->feature_flush = REQ_FLUSH | REQ_FUA;
1079

1080
	err = xlvbd_alloc_gendisk(sectors, info, binfo, sector_size);
1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095
	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. */
	spin_lock_irq(&blkif_io_lock);
	info->connected = BLKIF_STATE_CONNECTED;
	kick_pending_request_queues(info);
	spin_unlock_irq(&blkif_io_lock);

	add_disk(info->gd);
1096 1097

	info->is_ready = 1;
1098 1099 1100 1101 1102
}

/**
 * Callback received when the backend's state changes.
 */
1103
static void blkback_changed(struct xenbus_device *dev,
1104 1105
			    enum xenbus_state backend_state)
{
1106
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1107

1108
	dev_dbg(&dev->dev, "blkfront:blkback_changed to state %d.\n", backend_state);
1109 1110 1111 1112 1113

	switch (backend_state) {
	case XenbusStateInitialising:
	case XenbusStateInitWait:
	case XenbusStateInitialised:
1114 1115
	case XenbusStateReconfiguring:
	case XenbusStateReconfigured:
1116 1117 1118 1119 1120 1121 1122 1123 1124
	case XenbusStateUnknown:
	case XenbusStateClosed:
		break;

	case XenbusStateConnected:
		blkfront_connect(info);
		break;

	case XenbusStateClosing:
1125
		blkfront_closing(info);
1126 1127 1128 1129
		break;
	}
}

1130
static int blkfront_remove(struct xenbus_device *xbdev)
1131
{
1132 1133 1134
	struct blkfront_info *info = dev_get_drvdata(&xbdev->dev);
	struct block_device *bdev = NULL;
	struct gendisk *disk;
1135

1136
	dev_dbg(&xbdev->dev, "%s removed", xbdev->nodename);
1137 1138 1139

	blkif_free(info, 0);

1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162
	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;

1163 1164 1165 1166
	dev_warn(disk_to_dev(disk),
		 "%s was hot-unplugged, %d stale handles\n",
		 xbdev->nodename, bdev->bd_openers);

1167
	if (info && !bdev->bd_openers) {
1168 1169
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
1170
		kfree(info);
1171 1172 1173 1174
	}

	mutex_unlock(&bdev->bd_mutex);
	bdput(bdev);
1175 1176 1177 1178

	return 0;
}

1179 1180
static int blkfront_is_ready(struct xenbus_device *dev)
{
1181
	struct blkfront_info *info = dev_get_drvdata(&dev->dev);
1182

1183
	return info->is_ready && info->xbdev;
1184 1185
}

A
Al Viro 已提交
1186
static int blkif_open(struct block_device *bdev, fmode_t mode)
1187
{
1188 1189 1190
	struct gendisk *disk = bdev->bd_disk;
	struct blkfront_info *info;
	int err = 0;
1191

1192
	mutex_lock(&blkfront_mutex);
1193

1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209
	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:
1210
	mutex_unlock(&blkfront_mutex);
1211
	return err;
1212 1213
}

A
Al Viro 已提交
1214
static int blkif_release(struct gendisk *disk, fmode_t mode)
1215
{
A
Al Viro 已提交
1216
	struct blkfront_info *info = disk->private_data;
1217 1218 1219
	struct block_device *bdev;
	struct xenbus_device *xbdev;

1220
	mutex_lock(&blkfront_mutex);
1221 1222 1223 1224

	bdev = bdget_disk(disk, 0);
	bdput(bdev);

1225 1226 1227
	if (bdev->bd_openers)
		goto out;

1228 1229 1230 1231 1232 1233 1234 1235 1236 1237
	/*
	 * 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 */
1238
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
1239 1240 1241 1242 1243 1244 1245 1246
		xlvbd_release_gendisk(info);
		xenbus_frontend_closed(info->xbdev);
 	}

	mutex_unlock(&info->mutex);

	if (!xbdev) {
		/* sudden device removal */
1247
		dev_info(disk_to_dev(bdev->bd_disk), "releasing disk\n");
1248 1249 1250
		xlvbd_release_gendisk(info);
		disk->private_data = NULL;
		kfree(info);
1251
	}
1252

J
Jens Axboe 已提交
1253
out:
1254
	mutex_unlock(&blkfront_mutex);
1255 1256 1257
	return 0;
}

1258
static const struct block_device_operations xlvbd_block_fops =
1259 1260
{
	.owner = THIS_MODULE,
A
Al Viro 已提交
1261 1262
	.open = blkif_open,
	.release = blkif_release,
1263
	.getgeo = blkif_getgeo,
1264
	.ioctl = blkif_ioctl,
1265 1266 1267
};


1268
static const struct xenbus_device_id blkfront_ids[] = {
1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279
	{ "vbd" },
	{ "" }
};

static struct xenbus_driver blkfront = {
	.name = "vbd",
	.owner = THIS_MODULE,
	.ids = blkfront_ids,
	.probe = blkfront_probe,
	.remove = blkfront_remove,
	.resume = blkfront_resume,
1280
	.otherend_changed = blkback_changed,
1281
	.is_ready = blkfront_is_ready,
1282 1283 1284 1285
};

static int __init xlblk_init(void)
{
1286
	if (!xen_domain())
1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299
		return -ENODEV;

	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;
	}

	return xenbus_register_frontend(&blkfront);
}
module_init(xlblk_init);


1300
static void __exit xlblk_exit(void)
1301 1302 1303 1304 1305 1306 1307 1308
{
	return xenbus_unregister_driver(&blkfront);
}
module_exit(xlblk_exit);

MODULE_DESCRIPTION("Xen virtual block device frontend");
MODULE_LICENSE("GPL");
MODULE_ALIAS_BLOCKDEV_MAJOR(XENVBD_MAJOR);
1309
MODULE_ALIAS("xen:vbd");
1310
MODULE_ALIAS("xenblk");