[linux-2.6-block.git] / drivers / block / virtio_blk.c

//#define DEBUG
#include <linux/spinlock.h>
#include <linux/blkdev.h>
#include <linux/hdreg.h>
#include <linux/virtio.h>
#include <linux/virtio_blk.h>
#include <linux/scatterlist.h>

#define PART_BITS 4

static int major, index;

struct virtio_blk
{
	spinlock_t lock;

	struct virtio_device *vdev;
	struct virtqueue *vq;

	/* The disk structure for the kernel. */
	struct gendisk *disk;

	/* Request tracking. */
	struct list_head reqs;

	mempool_t *pool;

	/* What host tells us, plus 2 for header & tailer. */
	unsigned int sg_elems;

	/* Scatterlist: can be too big for stack. */
	struct scatterlist sg[/*sg_elems*/];
};

struct virtblk_req
{
	struct list_head list;
	struct request *req;
	struct virtio_blk_outhdr out_hdr;
	struct virtio_scsi_inhdr in_hdr;
	u8 status;
};

static void blk_done(struct virtqueue *vq)
{
	struct virtio_blk *vblk = vq->vdev->priv;
	struct virtblk_req *vbr;
	unsigned int len;
	unsigned long flags;

	spin_lock_irqsave(&vblk->lock, flags);
	while ((vbr = vblk->vq->vq_ops->get_buf(vblk->vq, &len)) != NULL) {
		int error;

		switch (vbr->status) {
		case VIRTIO_BLK_S_OK:
			error = 0;
			break;
		case VIRTIO_BLK_S_UNSUPP:
			error = -ENOTTY;
			break;
		default:
			error = -EIO;
			break;
		}

		if (blk_pc_request(vbr->req)) {
			vbr->req->resid_len = vbr->in_hdr.residual;
			vbr->req->sense_len = vbr->in_hdr.sense_len;
			vbr->req->errors = vbr->in_hdr.errors;
		}

		__blk_end_request_all(vbr->req, error);
		list_del(&vbr->list);
		mempool_free(vbr, vblk->pool);
	}
	/* In case queue is stopped waiting for more buffers. */
	blk_start_queue(vblk->disk->queue);
	spin_unlock_irqrestore(&vblk->lock, flags);
}

static bool do_req(struct request_queue *q, struct virtio_blk *vblk,
		   struct request *req)
{
	unsigned long num, out = 0, in = 0;
	struct virtblk_req *vbr;

	vbr = mempool_alloc(vblk->pool, GFP_ATOMIC);
	if (!vbr)
		/* When another request finishes we'll try again. */
		return false;

	vbr->req = req;
	if (blk_fs_request(vbr->req)) {
		vbr->out_hdr.type = 0;
		vbr->out_hdr.sector = blk_rq_pos(vbr->req);
		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
	} else if (blk_pc_request(vbr->req)) {
		vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
		vbr->out_hdr.sector = 0;
		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
	} else {
		/* We don't put anything else in the queue. */
		BUG();
	}

	if (blk_barrier_rq(vbr->req))
		vbr->out_hdr.type |= VIRTIO_BLK_T_BARRIER;

	sg_set_buf(&vblk->sg[out++], &vbr->out_hdr, sizeof(vbr->out_hdr));

	/*
	 * If this is a packet command we need a couple of additional headers.
	 * Behind the normal outhdr we put a segment with the scsi command
	 * block, and before the normal inhdr we put the sense data and the
	 * inhdr with additional status information before the normal inhdr.
	 */
	if (blk_pc_request(vbr->req))
		sg_set_buf(&vblk->sg[out++], vbr->req->cmd, vbr->req->cmd_len);

	num = blk_rq_map_sg(q, vbr->req, vblk->sg + out);

	if (blk_pc_request(vbr->req)) {
		sg_set_buf(&vblk->sg[num + out + in++], vbr->req->sense, 96);
		sg_set_buf(&vblk->sg[num + out + in++], &vbr->in_hdr,
			   sizeof(vbr->in_hdr));
	}

	sg_set_buf(&vblk->sg[num + out + in++], &vbr->status,
		   sizeof(vbr->status));

	if (num) {
		if (rq_data_dir(vbr->req) == WRITE) {
			vbr->out_hdr.type |= VIRTIO_BLK_T_OUT;
			out += num;
		} else {
			vbr->out_hdr.type |= VIRTIO_BLK_T_IN;
			in += num;
		}
	}

	if (vblk->vq->vq_ops->add_buf(vblk->vq, vblk->sg, out, in, vbr)) {
		mempool_free(vbr, vblk->pool);
		return false;
	}

	list_add_tail(&vbr->list, &vblk->reqs);
	return true;
}

static void do_virtblk_request(struct request_queue *q)
{
	struct virtio_blk *vblk = q->queuedata;
	struct request *req;
	unsigned int issued = 0;

	while ((req = blk_peek_request(q)) != NULL) {
		BUG_ON(req->nr_phys_segments + 2 > vblk->sg_elems);

		/* If this request fails, stop queue and wait for something to
		   finish to restart it. */
		if (!do_req(q, vblk, req)) {
			blk_stop_queue(q);
			break;
		}
		blk_start_request(req);
		issued++;
	}

	if (issued)
		vblk->vq->vq_ops->kick(vblk->vq);
}

static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
			 unsigned cmd, unsigned long data)
{
	struct gendisk *disk = bdev->bd_disk;
	struct virtio_blk *vblk = disk->private_data;

	/*
	 * Only allow the generic SCSI ioctls if the host can support it.
	 */
	if (!virtio_has_feature(vblk->vdev, VIRTIO_BLK_F_SCSI))
		return -ENOIOCTLCMD;

	return scsi_cmd_ioctl(disk->queue, disk, mode, cmd,
			      (void __user *)data);
}

/* We provide getgeo only to please some old bootloader/partitioning tools */
static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo)
{
	struct virtio_blk *vblk = bd->bd_disk->private_data;
	struct virtio_blk_geometry vgeo;
	int err;

	/* see if the host passed in geometry config */
	err = virtio_config_val(vblk->vdev, VIRTIO_BLK_F_GEOMETRY,
				offsetof(struct virtio_blk_config, geometry),
				&vgeo);

	if (!err) {
		geo->heads = vgeo.heads;
		geo->sectors = vgeo.sectors;
		geo->cylinders = vgeo.cylinders;
	} else {
		/* some standard values, similar to sd */
		geo->heads = 1 << 6;
		geo->sectors = 1 << 5;
		geo->cylinders = get_capacity(bd->bd_disk) >> 11;
	}
	return 0;
}

static struct block_device_operations virtblk_fops = {
	.locked_ioctl = virtblk_ioctl,
	.owner  = THIS_MODULE,
	.getgeo = virtblk_getgeo,
};

static int index_to_minor(int index)
{
	return index << PART_BITS;
}

static int virtblk_probe(struct virtio_device *vdev)
{
	struct virtio_blk *vblk;
	int err;
	u64 cap;
	u32 v;
	u32 blk_size, sg_elems;

	if (index_to_minor(index) >= 1 << MINORBITS)
		return -ENOSPC;

	/* We need to know how many segments before we allocate. */
	err = virtio_config_val(vdev, VIRTIO_BLK_F_SEG_MAX,
				offsetof(struct virtio_blk_config, seg_max),
				&sg_elems);
	if (err)
		sg_elems = 1;

	/* We need an extra sg elements at head and tail. */
	sg_elems += 2;
	vdev->priv = vblk = kmalloc(sizeof(*vblk) +
				    sizeof(vblk->sg[0]) * sg_elems, GFP_KERNEL);
	if (!vblk) {
		err = -ENOMEM;
		goto out;
	}

	INIT_LIST_HEAD(&vblk->reqs);
	spin_lock_init(&vblk->lock);
	vblk->vdev = vdev;
	vblk->sg_elems = sg_elems;
	sg_init_table(vblk->sg, vblk->sg_elems);

	/* We expect one virtqueue, for output. */
	vblk->vq = vdev->config->find_vq(vdev, 0, blk_done);
	if (IS_ERR(vblk->vq)) {
		err = PTR_ERR(vblk->vq);
		goto out_free_vblk;
	}

	vblk->pool = mempool_create_kmalloc_pool(1,sizeof(struct virtblk_req));
	if (!vblk->pool) {
		err = -ENOMEM;
		goto out_free_vq;
	}

	/* FIXME: How many partitions?  How long is a piece of string? */
	vblk->disk = alloc_disk(1 << PART_BITS);
	if (!vblk->disk) {
		err = -ENOMEM;
		goto out_mempool;
	}

	vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock);
	if (!vblk->disk->queue) {
		err = -ENOMEM;
		goto out_put_disk;
	}

	vblk->disk->queue->queuedata = vblk;
	queue_flag_set_unlocked(QUEUE_FLAG_VIRT, vblk->disk->queue);

	if (index < 26) {
		sprintf(vblk->disk->disk_name, "vd%c", 'a' + index % 26);
	} else if (index < (26 + 1) * 26) {
		sprintf(vblk->disk->disk_name, "vd%c%c",
			'a' + index / 26 - 1, 'a' + index % 26);
	} else {
		const unsigned int m1 = (index / 26 - 1) / 26 - 1;
		const unsigned int m2 = (index / 26 - 1) % 26;
		const unsigned int m3 =  index % 26;
		sprintf(vblk->disk->disk_name, "vd%c%c%c",
			'a' + m1, 'a' + m2, 'a' + m3);
	}

	vblk->disk->major = major;
	vblk->disk->first_minor = index_to_minor(index);
	vblk->disk->private_data = vblk;
	vblk->disk->fops = &virtblk_fops;
	vblk->disk->driverfs_dev = &vdev->dev;
	index++;

	/* If barriers are supported, tell block layer that queue is ordered */
	if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER))
		blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_TAG, NULL);

	/* If disk is read-only in the host, the guest should obey */
	if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO))
		set_disk_ro(vblk->disk, 1);

	/* Host must always specify the capacity. */
	vdev->config->get(vdev, offsetof(struct virtio_blk_config, capacity),
			  &cap, sizeof(cap));

	/* If capacity is too big, truncate with warning. */
	if ((sector_t)cap != cap) {
		dev_warn(&vdev->dev, "Capacity %llu too large: truncating\n",
			 (unsigned long long)cap);
		cap = (sector_t)-1;
	}
	set_capacity(vblk->disk, cap);

	/* We can handle whatever the host told us to handle. */
	blk_queue_max_phys_segments(vblk->disk->queue, vblk->sg_elems-2);
	blk_queue_max_hw_segments(vblk->disk->queue, vblk->sg_elems-2);

	/* No real sector limit. */
	blk_queue_max_sectors(vblk->disk->queue, -1U);

	/* Host can optionally specify maximum segment size and number of
	 * segments. */
	err = virtio_config_val(vdev, VIRTIO_BLK_F_SIZE_MAX,
				offsetof(struct virtio_blk_config, size_max),
				&v);
	if (!err)
		blk_queue_max_segment_size(vblk->disk->queue, v);
	else
		blk_queue_max_segment_size(vblk->disk->queue, -1U);

	/* Host can optionally specify the block size of the device */
	err = virtio_config_val(vdev, VIRTIO_BLK_F_BLK_SIZE,
				offsetof(struct virtio_blk_config, blk_size),
				&blk_size);
	if (!err)
		blk_queue_hardsect_size(vblk->disk->queue, blk_size);

	add_disk(vblk->disk);
	return 0;

out_put_disk:
	put_disk(vblk->disk);
out_mempool:
	mempool_destroy(vblk->pool);
out_free_vq:
	vdev->config->del_vq(vblk->vq);
out_free_vblk:
	kfree(vblk);
out:
	return err;
}

static void virtblk_remove(struct virtio_device *vdev)
{
	struct virtio_blk *vblk = vdev->priv;

	/* Nothing should be pending. */
	BUG_ON(!list_empty(&vblk->reqs));

	/* Stop all the virtqueues. */
	vdev->config->reset(vdev);

	del_gendisk(vblk->disk);
	blk_cleanup_queue(vblk->disk->queue);
	put_disk(vblk->disk);
	mempool_destroy(vblk->pool);
	vdev->config->del_vq(vblk->vq);
	kfree(vblk);
}

static struct virtio_device_id id_table[] = {
	{ VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID },
	{ 0 },
};

static unsigned int features[] = {
	VIRTIO_BLK_F_BARRIER, VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX,
	VIRTIO_BLK_F_GEOMETRY, VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
	VIRTIO_BLK_F_SCSI,
};

static struct virtio_driver virtio_blk = {
	.feature_table = features,
	.feature_table_size = ARRAY_SIZE(features),
	.driver.name =	KBUILD_MODNAME,
	.driver.owner =	THIS_MODULE,
	.id_table =	id_table,
	.probe =	virtblk_probe,
	.remove =	__devexit_p(virtblk_remove),
};

static int __init init(void)
{
	major = register_blkdev(0, "virtblk");
	if (major < 0)
		return major;
	return register_virtio_driver(&virtio_blk);
}

static void __exit fini(void)
{
	unregister_blkdev(major, "virtblk");
	unregister_virtio_driver(&virtio_blk);
}
module_init(init);
module_exit(fini);

MODULE_DEVICE_TABLE(virtio, id_table);
MODULE_DESCRIPTION("Virtio block driver");
MODULE_LICENSE("GPL");
Commit	Line	Data
e467cde2 RR	1	//#define DEBUG
	2	#include <linux/spinlock.h>
	3	#include <linux/blkdev.h>
	4	#include <linux/hdreg.h>
	5	#include <linux/virtio.h>
	6	#include <linux/virtio_blk.h>
3d1266c7 JA	7	#include <linux/scatterlist.h>
3d1266c7 JA	8
4f3bf19c	9	#define PART_BITS 4
e467cde2	10
d50ed907	11	static int major, index;
4f3bf19c	12
e467cde2 RR	13	struct virtio_blk
	14	{
	15	spinlock_t lock;
	16
	17	struct virtio_device *vdev;
	18	struct virtqueue *vq;
	19
	20	/* The disk structure for the kernel. */
	21	struct gendisk *disk;
	22
	23	/* Request tracking. */
	24	struct list_head reqs;
	25
	26	mempool_t *pool;
	27
0864b79a RR	28	/* What host tells us, plus 2 for header & tailer. */
	29	unsigned int sg_elems;
	30
e467cde2	31	/* Scatterlist: can be too big for stack. */
0864b79a	32	struct scatterlist sg[/sg_elems/];
e467cde2 RR	33	};
	34
	35	struct virtblk_req
	36	{
	37	struct list_head list;
	38	struct request *req;
	39	struct virtio_blk_outhdr out_hdr;
1cde26f9	40	struct virtio_scsi_inhdr in_hdr;
cb38fa23	41	u8 status;
e467cde2 RR	42	};
e467cde2 RR	43
18445c4d	44	static void blk_done(struct virtqueue *vq)
e467cde2 RR	45	{
	46	struct virtio_blk *vblk = vq->vdev->priv;
	47	struct virtblk_req *vbr;
	48	unsigned int len;
	49	unsigned long flags;
	50
	51	spin_lock_irqsave(&vblk->lock, flags);
	52	while ((vbr = vblk->vq->vq_ops->get_buf(vblk->vq, &len)) != NULL) {
8316982a	53	int error;
1cde26f9	54
cb38fa23	55	switch (vbr->status) {
e467cde2	56	case VIRTIO_BLK_S_OK:
8316982a	57	error = 0;
e467cde2 RR	58	break;
e467cde2 RR	59	case VIRTIO_BLK_S_UNSUPP:
8316982a	60	error = -ENOTTY;
e467cde2 RR	61	break;
e467cde2 RR	62	default:
8316982a	63	error = -EIO;
e467cde2 RR	64	break;
	65	}
	66
1cde26f9 HR	67	if (blk_pc_request(vbr->req)) {
	68	vbr->req->resid_len = vbr->in_hdr.residual;
	69	vbr->req->sense_len = vbr->in_hdr.sense_len;
	70	vbr->req->errors = vbr->in_hdr.errors;
	71	}
	72
40cbbb78	73	__blk_end_request_all(vbr->req, error);
e467cde2 RR	74	list_del(&vbr->list);
	75	mempool_free(vbr, vblk->pool);
	76	}
	77	/* In case queue is stopped waiting for more buffers. */
	78	blk_start_queue(vblk->disk->queue);
	79	spin_unlock_irqrestore(&vblk->lock, flags);
e467cde2 RR	80	}
	81
	82	static bool do_req(struct request_queue q, struct virtio_blk vblk,
	83	struct request *req)
	84	{
1cde26f9	85	unsigned long num, out = 0, in = 0;
e467cde2 RR	86	struct virtblk_req *vbr;
	87
	88	vbr = mempool_alloc(vblk->pool, GFP_ATOMIC);
	89	if (!vbr)
	90	/* When another request finishes we'll try again. */
	91	return false;
	92
	93	vbr->req = req;
	94	if (blk_fs_request(vbr->req)) {
	95	vbr->out_hdr.type = 0;
83096ebf	96	vbr->out_hdr.sector = blk_rq_pos(vbr->req);
766ca442	97	vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
e467cde2 RR	98	} else if (blk_pc_request(vbr->req)) {
	99	vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
	100	vbr->out_hdr.sector = 0;
766ca442	101	vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
e467cde2 RR	102	} else {
	103	/* We don't put anything else in the queue. */
	104	BUG();
	105	}
	106
	107	if (blk_barrier_rq(vbr->req))
	108	vbr->out_hdr.type \|= VIRTIO_BLK_T_BARRIER;
	109
1cde26f9	110	sg_set_buf(&vblk->sg[out++], &vbr->out_hdr, sizeof(vbr->out_hdr));
e467cde2	111
1cde26f9 HR	112	/*
	113	* If this is a packet command we need a couple of additional headers.
	114	* Behind the normal outhdr we put a segment with the scsi command
	115	* block, and before the normal inhdr we put the sense data and the
	116	* inhdr with additional status information before the normal inhdr.
	117	*/
	118	if (blk_pc_request(vbr->req))
	119	sg_set_buf(&vblk->sg[out++], vbr->req->cmd, vbr->req->cmd_len);
	120
	121	num = blk_rq_map_sg(q, vbr->req, vblk->sg + out);
	122
	123	if (blk_pc_request(vbr->req)) {
	124	sg_set_buf(&vblk->sg[num + out + in++], vbr->req->sense, 96);
	125	sg_set_buf(&vblk->sg[num + out + in++], &vbr->in_hdr,
	126	sizeof(vbr->in_hdr));
	127	}
	128
	129	sg_set_buf(&vblk->sg[num + out + in++], &vbr->status,
	130	sizeof(vbr->status));
	131
	132	if (num) {
	133	if (rq_data_dir(vbr->req) == WRITE) {
	134	vbr->out_hdr.type \|= VIRTIO_BLK_T_OUT;
	135	out += num;
	136	} else {
	137	vbr->out_hdr.type \|= VIRTIO_BLK_T_IN;
	138	in += num;
	139	}
e467cde2 RR	140	}
	141
	142	if (vblk->vq->vq_ops->add_buf(vblk->vq, vblk->sg, out, in, vbr)) {
	143	mempool_free(vbr, vblk->pool);
	144	return false;
	145	}
	146
	147	list_add_tail(&vbr->list, &vblk->reqs);
	148	return true;
	149	}
	150
	151	static void do_virtblk_request(struct request_queue *q)
	152	{
6c3b46f7	153	struct virtio_blk *vblk = q->queuedata;
e467cde2 RR	154	struct request *req;
	155	unsigned int issued = 0;
	156
9934c8c0	157	while ((req = blk_peek_request(q)) != NULL) {
0864b79a	158	BUG_ON(req->nr_phys_segments + 2 > vblk->sg_elems);
e467cde2 RR	159
	160	/* If this request fails, stop queue and wait for something to
	161	finish to restart it. */
	162	if (!do_req(q, vblk, req)) {
	163	blk_stop_queue(q);
	164	break;
	165	}
9934c8c0	166	blk_start_request(req);
e467cde2 RR	167	issued++;
	168	}
	169
	170	if (issued)
	171	vblk->vq->vq_ops->kick(vblk->vq);
	172	}
	173
4e109852	174	static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
e467cde2 RR	175	unsigned cmd, unsigned long data)
e467cde2 RR	176	{
1cde26f9 HR	177	struct gendisk *disk = bdev->bd_disk;
	178	struct virtio_blk *vblk = disk->private_data;
	179
	180	/*
	181	* Only allow the generic SCSI ioctls if the host can support it.
	182	*/
	183	if (!virtio_has_feature(vblk->vdev, VIRTIO_BLK_F_SCSI))
	184	return -ENOIOCTLCMD;
	185
	186	return scsi_cmd_ioctl(disk->queue, disk, mode, cmd,
e467cde2 RR	187	(void __user *)data);
	188	}
	189
135da0b0 CB	190	/* We provide getgeo only to please some old bootloader/partitioning tools */
	191	static int virtblk_getgeo(struct block_device bd, struct hd_geometry geo)
	192	{
48e4043d RH	193	struct virtio_blk *vblk = bd->bd_disk->private_data;
	194	struct virtio_blk_geometry vgeo;
	195	int err;
	196
	197	/* see if the host passed in geometry config */
	198	err = virtio_config_val(vblk->vdev, VIRTIO_BLK_F_GEOMETRY,
	199	offsetof(struct virtio_blk_config, geometry),
	200	&vgeo);
	201
	202	if (!err) {
	203	geo->heads = vgeo.heads;
	204	geo->sectors = vgeo.sectors;
	205	geo->cylinders = vgeo.cylinders;
	206	} else {
	207	/* some standard values, similar to sd */
	208	geo->heads = 1 << 6;
	209	geo->sectors = 1 << 5;
	210	geo->cylinders = get_capacity(bd->bd_disk) >> 11;
	211	}
135da0b0 CB	212	return 0;
	213	}
	214
e467cde2	215	static struct block_device_operations virtblk_fops = {
4e109852	216	.locked_ioctl = virtblk_ioctl,
135da0b0 CB	217	.owner = THIS_MODULE,
135da0b0 CB	218	.getgeo = virtblk_getgeo,
e467cde2 RR	219	};
e467cde2 RR	220
d50ed907 CB	221	static int index_to_minor(int index)
	222	{
	223	return index << PART_BITS;
	224	}
	225
e467cde2 RR	226	static int virtblk_probe(struct virtio_device *vdev)
	227	{
	228	struct virtio_blk *vblk;
4f3bf19c	229	int err;
e467cde2 RR	230	u64 cap;
e467cde2 RR	231	u32 v;
0864b79a	232	u32 blk_size, sg_elems;
e467cde2	233
d50ed907	234	if (index_to_minor(index) >= 1 << MINORBITS)
4f3bf19c CB	235	return -ENOSPC;
4f3bf19c CB	236
0864b79a RR	237	/* We need to know how many segments before we allocate. */
	238	err = virtio_config_val(vdev, VIRTIO_BLK_F_SEG_MAX,
	239	offsetof(struct virtio_blk_config, seg_max),
	240	&sg_elems);
	241	if (err)
	242	sg_elems = 1;
	243
	244	/* We need an extra sg elements at head and tail. */
	245	sg_elems += 2;
	246	vdev->priv = vblk = kmalloc(sizeof(*vblk) +
	247	sizeof(vblk->sg[0]) * sg_elems, GFP_KERNEL);
e467cde2 RR	248	if (!vblk) {
	249	err = -ENOMEM;
	250	goto out;
	251	}
	252
	253	INIT_LIST_HEAD(&vblk->reqs);
	254	spin_lock_init(&vblk->lock);
	255	vblk->vdev = vdev;
0864b79a RR	256	vblk->sg_elems = sg_elems;
0864b79a RR	257	sg_init_table(vblk->sg, vblk->sg_elems);
e467cde2 RR	258
e467cde2 RR	259	/* We expect one virtqueue, for output. */
a586d4f6	260	vblk->vq = vdev->config->find_vq(vdev, 0, blk_done);
e467cde2 RR	261	if (IS_ERR(vblk->vq)) {
	262	err = PTR_ERR(vblk->vq);
	263	goto out_free_vblk;
	264	}
	265
	266	vblk->pool = mempool_create_kmalloc_pool(1,sizeof(struct virtblk_req));
	267	if (!vblk->pool) {
	268	err = -ENOMEM;
	269	goto out_free_vq;
	270	}
	271
e467cde2	272	/* FIXME: How many partitions? How long is a piece of string? */
4f3bf19c	273	vblk->disk = alloc_disk(1 << PART_BITS);
e467cde2 RR	274	if (!vblk->disk) {
e467cde2 RR	275	err = -ENOMEM;
4f3bf19c	276	goto out_mempool;
e467cde2 RR	277	}
	278
	279	vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock);
	280	if (!vblk->disk->queue) {
	281	err = -ENOMEM;
	282	goto out_put_disk;
	283	}
	284
6c3b46f7	285	vblk->disk->queue->queuedata = vblk;
7d116b62 FLVC	286	queue_flag_set_unlocked(QUEUE_FLAG_VIRT, vblk->disk->queue);
7d116b62 FLVC	287
d50ed907 CB	288	if (index < 26) {
	289	sprintf(vblk->disk->disk_name, "vd%c", 'a' + index % 26);
	290	} else if (index < (26 + 1) * 26) {
	291	sprintf(vblk->disk->disk_name, "vd%c%c",
	292	'a' + index / 26 - 1, 'a' + index % 26);
	293	} else {
	294	const unsigned int m1 = (index / 26 - 1) / 26 - 1;
	295	const unsigned int m2 = (index / 26 - 1) % 26;
	296	const unsigned int m3 = index % 26;
	297	sprintf(vblk->disk->disk_name, "vd%c%c%c",
	298	'a' + m1, 'a' + m2, 'a' + m3);
	299	}
	300
e467cde2	301	vblk->disk->major = major;
d50ed907	302	vblk->disk->first_minor = index_to_minor(index);
e467cde2 RR	303	vblk->disk->private_data = vblk;
e467cde2 RR	304	vblk->disk->fops = &virtblk_fops;
c4839346	305	vblk->disk->driverfs_dev = &vdev->dev;
d50ed907	306	index++;
4f3bf19c	307
e467cde2	308	/* If barriers are supported, tell block layer that queue is ordered */
c45a6816	309	if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER))
e467cde2 RR	310	blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_TAG, NULL);
e467cde2 RR	311
3ef53609 CB	312	/* If disk is read-only in the host, the guest should obey */
	313	if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO))
	314	set_disk_ro(vblk->disk, 1);
	315
a586d4f6	316	/* Host must always specify the capacity. */
72e61eb4 RR	317	vdev->config->get(vdev, offsetof(struct virtio_blk_config, capacity),
72e61eb4 RR	318	&cap, sizeof(cap));
e467cde2 RR	319
	320	/* If capacity is too big, truncate with warning. */
	321	if ((sector_t)cap != cap) {
	322	dev_warn(&vdev->dev, "Capacity %llu too large: truncating\n",
	323	(unsigned long long)cap);
	324	cap = (sector_t)-1;
	325	}
	326	set_capacity(vblk->disk, cap);
	327
0864b79a RR	328	/* We can handle whatever the host told us to handle. */
	329	blk_queue_max_phys_segments(vblk->disk->queue, vblk->sg_elems-2);
	330	blk_queue_max_hw_segments(vblk->disk->queue, vblk->sg_elems-2);
	331
4b7f7e20 RR	332	/* No real sector limit. */
	333	blk_queue_max_sectors(vblk->disk->queue, -1U);
	334
a586d4f6 RR	335	/* Host can optionally specify maximum segment size and number of
	336	* segments. */
	337	err = virtio_config_val(vdev, VIRTIO_BLK_F_SIZE_MAX,
	338	offsetof(struct virtio_blk_config, size_max),
	339	&v);
e467cde2 RR	340	if (!err)
e467cde2 RR	341	blk_queue_max_segment_size(vblk->disk->queue, v);
4b7f7e20	342	else
b194aee9	343	blk_queue_max_segment_size(vblk->disk->queue, -1U);
e467cde2	344
066f4d82 CB	345	/* Host can optionally specify the block size of the device */
	346	err = virtio_config_val(vdev, VIRTIO_BLK_F_BLK_SIZE,
	347	offsetof(struct virtio_blk_config, blk_size),
	348	&blk_size);
	349	if (!err)
	350	blk_queue_hardsect_size(vblk->disk->queue, blk_size);
	351
e467cde2 RR	352	add_disk(vblk->disk);
	353	return 0;
	354
	355	out_put_disk:
	356	put_disk(vblk->disk);
e467cde2 RR	357	out_mempool:
	358	mempool_destroy(vblk->pool);
	359	out_free_vq:
	360	vdev->config->del_vq(vblk->vq);
	361	out_free_vblk:
	362	kfree(vblk);
	363	out:
	364	return err;
	365	}
	366
	367	static void virtblk_remove(struct virtio_device *vdev)
	368	{
	369	struct virtio_blk *vblk = vdev->priv;
e467cde2	370
6e5aa7ef	371	/* Nothing should be pending. */
e467cde2	372	BUG_ON(!list_empty(&vblk->reqs));
6e5aa7ef RR	373
	374	/* Stop all the virtqueues. */
	375	vdev->config->reset(vdev);
	376
ac9d463a	377	del_gendisk(vblk->disk);
e467cde2 RR	378	blk_cleanup_queue(vblk->disk->queue);
e467cde2 RR	379	put_disk(vblk->disk);
e467cde2	380	mempool_destroy(vblk->pool);
74b2553f	381	vdev->config->del_vq(vblk->vq);
e467cde2 RR	382	kfree(vblk);
	383	}
	384
	385	static struct virtio_device_id id_table[] = {
	386	{ VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID },
	387	{ 0 },
	388	};
	389
c45a6816 RR	390	static unsigned int features[] = {
c45a6816 RR	391	VIRTIO_BLK_F_BARRIER, VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX,
066f4d82	392	VIRTIO_BLK_F_GEOMETRY, VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
1cde26f9	393	VIRTIO_BLK_F_SCSI,
c45a6816 RR	394	};
c45a6816 RR	395
e467cde2	396	static struct virtio_driver virtio_blk = {
c45a6816 RR	397	.feature_table = features,
c45a6816 RR	398	.feature_table_size = ARRAY_SIZE(features),
e467cde2 RR	399	.driver.name = KBUILD_MODNAME,
	400	.driver.owner = THIS_MODULE,
	401	.id_table = id_table,
	402	.probe = virtblk_probe,
	403	.remove = __devexit_p(virtblk_remove),
	404	};
	405
	406	static int __init init(void)
	407	{
4f3bf19c CB	408	major = register_blkdev(0, "virtblk");
	409	if (major < 0)
	410	return major;
e467cde2 RR	411	return register_virtio_driver(&virtio_blk);
	412	}
	413
	414	static void __exit fini(void)
	415	{
4f3bf19c	416	unregister_blkdev(major, "virtblk");
e467cde2 RR	417	unregister_virtio_driver(&virtio_blk);
	418	}
	419	module_init(init);
	420	module_exit(fini);
	421
	422	MODULE_DEVICE_TABLE(virtio, id_table);
	423	MODULE_DESCRIPTION("Virtio block driver");
	424	MODULE_LICENSE("GPL");