[net-next-2.6.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;
	switch (req->cmd_type) {
	case REQ_TYPE_FS:
		vbr->out_hdr.type = 0;
		vbr->out_hdr.sector = blk_rq_pos(vbr->req);
		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
		break;
	case REQ_TYPE_BLOCK_PC:
		vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
		vbr->out_hdr.sector = 0;
		vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
		break;
	case REQ_TYPE_LINUX_BLOCK:
		if (req->cmd[0] == REQ_LB_OP_FLUSH) {
			vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH;
			vbr->out_hdr.sector = 0;
			vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
			break;
		}
		/*FALLTHRU*/
	default:
		/* 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) < 0) {
		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 void virtblk_prepare_flush(struct request_queue *q, struct request *req)
{
	req->cmd_type = REQ_TYPE_LINUX_BLOCK;
	req->cmd[0] = REQ_LB_OP_FLUSH;
}

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 -ENOTTY;

	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 const 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 __devinit 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 = virtio_find_single_vq(vdev, blk_done, "requests");
	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;

	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_FLUSH))
		blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_DRAIN_FLUSH,
				  virtblk_prepare_flush);
	else 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 need to bounce any requests */
	blk_queue_bounce_limit(vblk->disk->queue, BLK_BOUNCE_ANY);

	/* 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_logical_block_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_vqs(vdev);
out_free_vblk:
	kfree(vblk);
out:
	return err;
}

static void __devexit 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_vqs(vdev);
	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, VIRTIO_BLK_F_FLUSH
};

/*
 * virtio_blk causes spurious section mismatch warning by
 * simultaneously referring to a __devinit and a __devexit function.
 * Use __refdata to avoid this warning.
 */
static struct virtio_driver __refdata 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;
f1b0ef06 CH	94	switch (req->cmd_type) {
f1b0ef06 CH	95	case REQ_TYPE_FS:
e467cde2	96	vbr->out_hdr.type = 0;
83096ebf	97	vbr->out_hdr.sector = blk_rq_pos(vbr->req);
766ca442	98	vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
f1b0ef06 CH	99	break;
f1b0ef06 CH	100	case REQ_TYPE_BLOCK_PC:
e467cde2 RR	101	vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
e467cde2 RR	102	vbr->out_hdr.sector = 0;
766ca442	103	vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
f1b0ef06 CH	104	break;
	105	case REQ_TYPE_LINUX_BLOCK:
	106	if (req->cmd[0] == REQ_LB_OP_FLUSH) {
	107	vbr->out_hdr.type = VIRTIO_BLK_T_FLUSH;
	108	vbr->out_hdr.sector = 0;
	109	vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
	110	break;
	111	}
	112	/FALLTHRU/
	113	default:
e467cde2 RR	114	/* We don't put anything else in the queue. */
	115	BUG();
	116	}
	117
	118	if (blk_barrier_rq(vbr->req))
	119	vbr->out_hdr.type \|= VIRTIO_BLK_T_BARRIER;
	120
1cde26f9	121	sg_set_buf(&vblk->sg[out++], &vbr->out_hdr, sizeof(vbr->out_hdr));
e467cde2	122
1cde26f9 HR	123	/*
	124	* If this is a packet command we need a couple of additional headers.
	125	* Behind the normal outhdr we put a segment with the scsi command
	126	* block, and before the normal inhdr we put the sense data and the
	127	* inhdr with additional status information before the normal inhdr.
	128	*/
	129	if (blk_pc_request(vbr->req))
	130	sg_set_buf(&vblk->sg[out++], vbr->req->cmd, vbr->req->cmd_len);
	131
	132	num = blk_rq_map_sg(q, vbr->req, vblk->sg + out);
	133
	134	if (blk_pc_request(vbr->req)) {
	135	sg_set_buf(&vblk->sg[num + out + in++], vbr->req->sense, 96);
	136	sg_set_buf(&vblk->sg[num + out + in++], &vbr->in_hdr,
	137	sizeof(vbr->in_hdr));
	138	}
	139
	140	sg_set_buf(&vblk->sg[num + out + in++], &vbr->status,
	141	sizeof(vbr->status));
	142
	143	if (num) {
	144	if (rq_data_dir(vbr->req) == WRITE) {
	145	vbr->out_hdr.type \|= VIRTIO_BLK_T_OUT;
	146	out += num;
	147	} else {
	148	vbr->out_hdr.type \|= VIRTIO_BLK_T_IN;
	149	in += num;
	150	}
e467cde2 RR	151	}
e467cde2 RR	152
3c1b27d5	153	if (vblk->vq->vq_ops->add_buf(vblk->vq, vblk->sg, out, in, vbr) < 0) {
e467cde2 RR	154	mempool_free(vbr, vblk->pool);
	155	return false;
	156	}
	157
	158	list_add_tail(&vbr->list, &vblk->reqs);
	159	return true;
	160	}
	161
	162	static void do_virtblk_request(struct request_queue *q)
	163	{
6c3b46f7	164	struct virtio_blk *vblk = q->queuedata;
e467cde2 RR	165	struct request *req;
	166	unsigned int issued = 0;
	167
9934c8c0	168	while ((req = blk_peek_request(q)) != NULL) {
0864b79a	169	BUG_ON(req->nr_phys_segments + 2 > vblk->sg_elems);
e467cde2 RR	170
	171	/* If this request fails, stop queue and wait for something to
	172	finish to restart it. */
	173	if (!do_req(q, vblk, req)) {
	174	blk_stop_queue(q);
	175	break;
	176	}
9934c8c0	177	blk_start_request(req);
e467cde2 RR	178	issued++;
	179	}
	180
	181	if (issued)
	182	vblk->vq->vq_ops->kick(vblk->vq);
	183	}
	184
f1b0ef06 CH	185	static void virtblk_prepare_flush(struct request_queue q, struct request req)
	186	{
	187	req->cmd_type = REQ_TYPE_LINUX_BLOCK;
	188	req->cmd[0] = REQ_LB_OP_FLUSH;
	189	}
	190
4e109852	191	static int virtblk_ioctl(struct block_device *bdev, fmode_t mode,
e467cde2 RR	192	unsigned cmd, unsigned long data)
e467cde2 RR	193	{
1cde26f9 HR	194	struct gendisk *disk = bdev->bd_disk;
	195	struct virtio_blk *vblk = disk->private_data;
	196
	197	/*
	198	* Only allow the generic SCSI ioctls if the host can support it.
	199	*/
	200	if (!virtio_has_feature(vblk->vdev, VIRTIO_BLK_F_SCSI))
d9ecdea7	201	return -ENOTTY;
1cde26f9	202
3225beab RR	203	return scsi_cmd_ioctl(disk->queue, disk, mode, cmd,
3225beab RR	204	(void __user *)data);
e467cde2 RR	205	}
e467cde2 RR	206
135da0b0 CB	207	/* We provide getgeo only to please some old bootloader/partitioning tools */
	208	static int virtblk_getgeo(struct block_device bd, struct hd_geometry geo)
	209	{
48e4043d RH	210	struct virtio_blk *vblk = bd->bd_disk->private_data;
	211	struct virtio_blk_geometry vgeo;
	212	int err;
	213
	214	/* see if the host passed in geometry config */
	215	err = virtio_config_val(vblk->vdev, VIRTIO_BLK_F_GEOMETRY,
	216	offsetof(struct virtio_blk_config, geometry),
	217	&vgeo);
	218
	219	if (!err) {
	220	geo->heads = vgeo.heads;
	221	geo->sectors = vgeo.sectors;
	222	geo->cylinders = vgeo.cylinders;
	223	} else {
	224	/* some standard values, similar to sd */
	225	geo->heads = 1 << 6;
	226	geo->sectors = 1 << 5;
	227	geo->cylinders = get_capacity(bd->bd_disk) >> 11;
	228	}
135da0b0 CB	229	return 0;
	230	}
	231
83d5cde4	232	static const struct block_device_operations virtblk_fops = {
4e109852	233	.locked_ioctl = virtblk_ioctl,
135da0b0 CB	234	.owner = THIS_MODULE,
135da0b0 CB	235	.getgeo = virtblk_getgeo,
e467cde2 RR	236	};
e467cde2 RR	237
d50ed907 CB	238	static int index_to_minor(int index)
	239	{
	240	return index << PART_BITS;
	241	}
	242
98e94444	243	static int __devinit virtblk_probe(struct virtio_device *vdev)
e467cde2 RR	244	{
e467cde2 RR	245	struct virtio_blk *vblk;
4f3bf19c	246	int err;
e467cde2 RR	247	u64 cap;
e467cde2 RR	248	u32 v;
0864b79a	249	u32 blk_size, sg_elems;
e467cde2	250
d50ed907	251	if (index_to_minor(index) >= 1 << MINORBITS)
4f3bf19c CB	252	return -ENOSPC;
4f3bf19c CB	253
0864b79a RR	254	/* We need to know how many segments before we allocate. */
	255	err = virtio_config_val(vdev, VIRTIO_BLK_F_SEG_MAX,
	256	offsetof(struct virtio_blk_config, seg_max),
	257	&sg_elems);
	258	if (err)
	259	sg_elems = 1;
	260
	261	/* We need an extra sg elements at head and tail. */
	262	sg_elems += 2;
	263	vdev->priv = vblk = kmalloc(sizeof(*vblk) +
	264	sizeof(vblk->sg[0]) * sg_elems, GFP_KERNEL);
e467cde2 RR	265	if (!vblk) {
	266	err = -ENOMEM;
	267	goto out;
	268	}
	269
	270	INIT_LIST_HEAD(&vblk->reqs);
	271	spin_lock_init(&vblk->lock);
	272	vblk->vdev = vdev;
0864b79a RR	273	vblk->sg_elems = sg_elems;
0864b79a RR	274	sg_init_table(vblk->sg, vblk->sg_elems);
e467cde2 RR	275
e467cde2 RR	276	/* We expect one virtqueue, for output. */
d2a7ddda	277	vblk->vq = virtio_find_single_vq(vdev, blk_done, "requests");
e467cde2 RR	278	if (IS_ERR(vblk->vq)) {
	279	err = PTR_ERR(vblk->vq);
	280	goto out_free_vblk;
	281	}
	282
	283	vblk->pool = mempool_create_kmalloc_pool(1,sizeof(struct virtblk_req));
	284	if (!vblk->pool) {
	285	err = -ENOMEM;
	286	goto out_free_vq;
	287	}
	288
e467cde2	289	/* FIXME: How many partitions? How long is a piece of string? */
4f3bf19c	290	vblk->disk = alloc_disk(1 << PART_BITS);
e467cde2 RR	291	if (!vblk->disk) {
e467cde2 RR	292	err = -ENOMEM;
4f3bf19c	293	goto out_mempool;
e467cde2 RR	294	}
	295
	296	vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock);
	297	if (!vblk->disk->queue) {
	298	err = -ENOMEM;
	299	goto out_put_disk;
	300	}
	301
6c3b46f7	302	vblk->disk->queue->queuedata = vblk;
7d116b62	303
d50ed907 CB	304	if (index < 26) {
	305	sprintf(vblk->disk->disk_name, "vd%c", 'a' + index % 26);
	306	} else if (index < (26 + 1) * 26) {
	307	sprintf(vblk->disk->disk_name, "vd%c%c",
	308	'a' + index / 26 - 1, 'a' + index % 26);
	309	} else {
	310	const unsigned int m1 = (index / 26 - 1) / 26 - 1;
	311	const unsigned int m2 = (index / 26 - 1) % 26;
	312	const unsigned int m3 = index % 26;
	313	sprintf(vblk->disk->disk_name, "vd%c%c%c",
	314	'a' + m1, 'a' + m2, 'a' + m3);
	315	}
	316
e467cde2	317	vblk->disk->major = major;
d50ed907	318	vblk->disk->first_minor = index_to_minor(index);
e467cde2 RR	319	vblk->disk->private_data = vblk;
e467cde2 RR	320	vblk->disk->fops = &virtblk_fops;
c4839346	321	vblk->disk->driverfs_dev = &vdev->dev;
d50ed907	322	index++;
4f3bf19c	323
e467cde2	324	/* If barriers are supported, tell block layer that queue is ordered */
f1b0ef06 CH	325	if (virtio_has_feature(vdev, VIRTIO_BLK_F_FLUSH))
	326	blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_DRAIN_FLUSH,
	327	virtblk_prepare_flush);
	328	else if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER))
e467cde2 RR	329	blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_TAG, NULL);
e467cde2 RR	330
3ef53609 CB	331	/* If disk is read-only in the host, the guest should obey */
	332	if (virtio_has_feature(vdev, VIRTIO_BLK_F_RO))
	333	set_disk_ro(vblk->disk, 1);
	334
a586d4f6	335	/* Host must always specify the capacity. */
72e61eb4 RR	336	vdev->config->get(vdev, offsetof(struct virtio_blk_config, capacity),
72e61eb4 RR	337	&cap, sizeof(cap));
e467cde2 RR	338
	339	/* If capacity is too big, truncate with warning. */
	340	if ((sector_t)cap != cap) {
	341	dev_warn(&vdev->dev, "Capacity %llu too large: truncating\n",
	342	(unsigned long long)cap);
	343	cap = (sector_t)-1;
	344	}
	345	set_capacity(vblk->disk, cap);
	346
0864b79a RR	347	/* We can handle whatever the host told us to handle. */
	348	blk_queue_max_phys_segments(vblk->disk->queue, vblk->sg_elems-2);
	349	blk_queue_max_hw_segments(vblk->disk->queue, vblk->sg_elems-2);
	350
4eff3cae CH	351	/* No need to bounce any requests */
	352	blk_queue_bounce_limit(vblk->disk->queue, BLK_BOUNCE_ANY);
	353
4b7f7e20 RR	354	/* No real sector limit. */
	355	blk_queue_max_sectors(vblk->disk->queue, -1U);
	356
a586d4f6 RR	357	/* Host can optionally specify maximum segment size and number of
	358	* segments. */
	359	err = virtio_config_val(vdev, VIRTIO_BLK_F_SIZE_MAX,
	360	offsetof(struct virtio_blk_config, size_max),
	361	&v);
e467cde2 RR	362	if (!err)
e467cde2 RR	363	blk_queue_max_segment_size(vblk->disk->queue, v);
4b7f7e20	364	else
b194aee9	365	blk_queue_max_segment_size(vblk->disk->queue, -1U);
e467cde2	366
066f4d82 CB	367	/* Host can optionally specify the block size of the device */
	368	err = virtio_config_val(vdev, VIRTIO_BLK_F_BLK_SIZE,
	369	offsetof(struct virtio_blk_config, blk_size),
	370	&blk_size);
	371	if (!err)
e1defc4f	372	blk_queue_logical_block_size(vblk->disk->queue, blk_size);
066f4d82	373
e467cde2 RR	374	add_disk(vblk->disk);
	375	return 0;
	376
	377	out_put_disk:
	378	put_disk(vblk->disk);
e467cde2 RR	379	out_mempool:
	380	mempool_destroy(vblk->pool);
	381	out_free_vq:
d2a7ddda	382	vdev->config->del_vqs(vdev);
e467cde2 RR	383	out_free_vblk:
	384	kfree(vblk);
	385	out:
	386	return err;
	387	}
	388
98e94444	389	static void __devexit virtblk_remove(struct virtio_device *vdev)
e467cde2 RR	390	{
e467cde2 RR	391	struct virtio_blk *vblk = vdev->priv;
e467cde2	392
6e5aa7ef	393	/* Nothing should be pending. */
e467cde2	394	BUG_ON(!list_empty(&vblk->reqs));
6e5aa7ef RR	395
	396	/* Stop all the virtqueues. */
	397	vdev->config->reset(vdev);
	398
ac9d463a	399	del_gendisk(vblk->disk);
e467cde2 RR	400	blk_cleanup_queue(vblk->disk->queue);
e467cde2 RR	401	put_disk(vblk->disk);
e467cde2	402	mempool_destroy(vblk->pool);
d2a7ddda	403	vdev->config->del_vqs(vdev);
e467cde2 RR	404	kfree(vblk);
	405	}
	406
	407	static struct virtio_device_id id_table[] = {
	408	{ VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID },
	409	{ 0 },
	410	};
	411
c45a6816 RR	412	static unsigned int features[] = {
c45a6816 RR	413	VIRTIO_BLK_F_BARRIER, VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX,
066f4d82	414	VIRTIO_BLK_F_GEOMETRY, VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
3225beab	415	VIRTIO_BLK_F_SCSI, VIRTIO_BLK_F_FLUSH
c45a6816 RR	416	};
c45a6816 RR	417
4fbfff76 RM	418	/*
	419	* virtio_blk causes spurious section mismatch warning by
	420	* simultaneously referring to a __devinit and a __devexit function.
	421	* Use __refdata to avoid this warning.
	422	*/
	423	static struct virtio_driver __refdata virtio_blk = {
c45a6816 RR	424	.feature_table = features,
c45a6816 RR	425	.feature_table_size = ARRAY_SIZE(features),
e467cde2 RR	426	.driver.name = KBUILD_MODNAME,
	427	.driver.owner = THIS_MODULE,
	428	.id_table = id_table,
	429	.probe = virtblk_probe,
	430	.remove = __devexit_p(virtblk_remove),
	431	};
	432
	433	static int __init init(void)
	434	{
4f3bf19c CB	435	major = register_blkdev(0, "virtblk");
	436	if (major < 0)
	437	return major;
e467cde2 RR	438	return register_virtio_driver(&virtio_blk);
	439	}
	440
	441	static void __exit fini(void)
	442	{
4f3bf19c	443	unregister_blkdev(major, "virtblk");
e467cde2 RR	444	unregister_virtio_driver(&virtio_blk);
	445	}
	446	module_init(init);
	447	module_exit(fini);
	448
	449	MODULE_DEVICE_TABLE(virtio, id_table);
	450	MODULE_DESCRIPTION("Virtio block driver");
	451	MODULE_LICENSE("GPL");