2007-10-21 21:03:38 -04:00
|
|
|
//#define DEBUG
|
|
|
|
#include <linux/spinlock.h>
|
|
|
|
#include <linux/blkdev.h>
|
|
|
|
#include <linux/hdreg.h>
|
|
|
|
#include <linux/virtio.h>
|
|
|
|
#include <linux/virtio_blk.h>
|
2007-10-24 07:21:21 -04:00
|
|
|
#include <linux/scatterlist.h>
|
|
|
|
|
|
|
|
#define VIRTIO_MAX_SG (3+MAX_PHYS_SEGMENTS)
|
2008-01-31 09:53:53 -05:00
|
|
|
#define PART_BITS 4
|
2007-10-21 21:03:38 -04:00
|
|
|
|
2008-02-01 03:05:00 -05:00
|
|
|
static int major, index;
|
2008-01-31 09:53:53 -05:00
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
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;
|
|
|
|
|
|
|
|
/* Scatterlist: can be too big for stack. */
|
2007-10-24 07:21:21 -04:00
|
|
|
struct scatterlist sg[VIRTIO_MAX_SG];
|
2007-10-21 21:03:38 -04:00
|
|
|
};
|
|
|
|
|
|
|
|
struct virtblk_req
|
|
|
|
{
|
|
|
|
struct list_head list;
|
|
|
|
struct request *req;
|
|
|
|
struct virtio_blk_outhdr out_hdr;
|
2008-05-02 22:50:45 -04:00
|
|
|
u8 status;
|
2007-10-21 21:03:38 -04:00
|
|
|
};
|
|
|
|
|
2008-02-04 23:49:57 -05:00
|
|
|
static void blk_done(struct virtqueue *vq)
|
2007-10-21 21:03:38 -04:00
|
|
|
{
|
|
|
|
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) {
|
2008-10-01 10:11:20 -04:00
|
|
|
int error;
|
2008-05-02 22:50:45 -04:00
|
|
|
switch (vbr->status) {
|
2007-10-21 21:03:38 -04:00
|
|
|
case VIRTIO_BLK_S_OK:
|
2008-10-01 10:11:20 -04:00
|
|
|
error = 0;
|
2007-10-21 21:03:38 -04:00
|
|
|
break;
|
|
|
|
case VIRTIO_BLK_S_UNSUPP:
|
2008-10-01 10:11:20 -04:00
|
|
|
error = -ENOTTY;
|
2007-10-21 21:03:38 -04:00
|
|
|
break;
|
|
|
|
default:
|
2008-10-01 10:11:20 -04:00
|
|
|
error = -EIO;
|
2007-10-21 21:03:38 -04:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2008-10-01 10:11:20 -04:00
|
|
|
__blk_end_request(vbr->req, error, blk_rq_bytes(vbr->req));
|
2007-10-21 21:03:38 -04:00
|
|
|
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, in;
|
|
|
|
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 = vbr->req->sector;
|
2008-08-14 03:59:13 -04:00
|
|
|
vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
|
2007-10-21 21:03:38 -04:00
|
|
|
} else if (blk_pc_request(vbr->req)) {
|
|
|
|
vbr->out_hdr.type = VIRTIO_BLK_T_SCSI_CMD;
|
|
|
|
vbr->out_hdr.sector = 0;
|
2008-08-14 03:59:13 -04:00
|
|
|
vbr->out_hdr.ioprio = req_get_ioprio(vbr->req);
|
2007-10-21 21:03:38 -04:00
|
|
|
} 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;
|
|
|
|
|
2007-10-24 07:21:21 -04:00
|
|
|
/* This init could be done at vblk creation time */
|
|
|
|
sg_init_table(vblk->sg, VIRTIO_MAX_SG);
|
2007-10-21 21:03:38 -04:00
|
|
|
sg_set_buf(&vblk->sg[0], &vbr->out_hdr, sizeof(vbr->out_hdr));
|
|
|
|
num = blk_rq_map_sg(q, vbr->req, vblk->sg+1);
|
2008-05-02 22:50:45 -04:00
|
|
|
sg_set_buf(&vblk->sg[num+1], &vbr->status, sizeof(vbr->status));
|
2007-10-21 21:03:38 -04:00
|
|
|
|
|
|
|
if (rq_data_dir(vbr->req) == WRITE) {
|
|
|
|
vbr->out_hdr.type |= VIRTIO_BLK_T_OUT;
|
|
|
|
out = 1 + num;
|
|
|
|
in = 1;
|
|
|
|
} else {
|
|
|
|
vbr->out_hdr.type |= VIRTIO_BLK_T_IN;
|
|
|
|
out = 1;
|
|
|
|
in = 1 + 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 = NULL;
|
|
|
|
struct request *req;
|
|
|
|
unsigned int issued = 0;
|
|
|
|
|
|
|
|
while ((req = elv_next_request(q)) != NULL) {
|
|
|
|
vblk = req->rq_disk->private_data;
|
|
|
|
BUG_ON(req->nr_phys_segments > ARRAY_SIZE(vblk->sg));
|
|
|
|
|
|
|
|
/* 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;
|
|
|
|
}
|
|
|
|
blkdev_dequeue_request(req);
|
|
|
|
issued++;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (issued)
|
|
|
|
vblk->vq->vq_ops->kick(vblk->vq);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int virtblk_ioctl(struct inode *inode, struct file *filp,
|
|
|
|
unsigned cmd, unsigned long data)
|
|
|
|
{
|
|
|
|
return scsi_cmd_ioctl(filp, inode->i_bdev->bd_disk->queue,
|
|
|
|
inode->i_bdev->bd_disk, cmd,
|
|
|
|
(void __user *)data);
|
|
|
|
}
|
|
|
|
|
2008-01-23 11:56:50 -05:00
|
|
|
/* We provide getgeo only to please some old bootloader/partitioning tools */
|
|
|
|
static int virtblk_getgeo(struct block_device *bd, struct hd_geometry *geo)
|
|
|
|
{
|
2008-04-16 14:56:37 -04:00
|
|
|
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;
|
|
|
|
}
|
2008-01-23 11:56:50 -05:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
static struct block_device_operations virtblk_fops = {
|
2008-01-23 11:56:50 -05:00
|
|
|
.ioctl = virtblk_ioctl,
|
|
|
|
.owner = THIS_MODULE,
|
|
|
|
.getgeo = virtblk_getgeo,
|
2007-10-21 21:03:38 -04:00
|
|
|
};
|
|
|
|
|
2008-02-01 03:05:00 -05:00
|
|
|
static int index_to_minor(int index)
|
|
|
|
{
|
|
|
|
return index << PART_BITS;
|
|
|
|
}
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
static int virtblk_probe(struct virtio_device *vdev)
|
|
|
|
{
|
|
|
|
struct virtio_blk *vblk;
|
2008-01-31 09:53:53 -05:00
|
|
|
int err;
|
2007-10-21 21:03:38 -04:00
|
|
|
u64 cap;
|
|
|
|
u32 v;
|
2008-05-29 05:08:26 -04:00
|
|
|
u32 blk_size;
|
2007-10-21 21:03:38 -04:00
|
|
|
|
2008-02-01 03:05:00 -05:00
|
|
|
if (index_to_minor(index) >= 1 << MINORBITS)
|
2008-01-31 09:53:53 -05:00
|
|
|
return -ENOSPC;
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
vdev->priv = vblk = kmalloc(sizeof(*vblk), GFP_KERNEL);
|
|
|
|
if (!vblk) {
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
INIT_LIST_HEAD(&vblk->reqs);
|
|
|
|
spin_lock_init(&vblk->lock);
|
|
|
|
vblk->vdev = vdev;
|
|
|
|
|
|
|
|
/* We expect one virtqueue, for output. */
|
2008-02-04 23:49:56 -05:00
|
|
|
vblk->vq = vdev->config->find_vq(vdev, 0, blk_done);
|
2007-10-21 21:03:38 -04:00
|
|
|
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? */
|
2008-01-31 09:53:53 -05:00
|
|
|
vblk->disk = alloc_disk(1 << PART_BITS);
|
2007-10-21 21:03:38 -04:00
|
|
|
if (!vblk->disk) {
|
|
|
|
err = -ENOMEM;
|
2008-01-31 09:53:53 -05:00
|
|
|
goto out_mempool;
|
2007-10-21 21:03:38 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
vblk->disk->queue = blk_init_queue(do_virtblk_request, &vblk->lock);
|
|
|
|
if (!vblk->disk->queue) {
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto out_put_disk;
|
|
|
|
}
|
|
|
|
|
2008-02-01 03:05:00 -05:00
|
|
|
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);
|
|
|
|
}
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
vblk->disk->major = major;
|
2008-02-01 03:05:00 -05:00
|
|
|
vblk->disk->first_minor = index_to_minor(index);
|
2007-10-21 21:03:38 -04:00
|
|
|
vblk->disk->private_data = vblk;
|
|
|
|
vblk->disk->fops = &virtblk_fops;
|
2008-03-02 17:00:15 -05:00
|
|
|
vblk->disk->driverfs_dev = &vdev->dev;
|
2008-02-01 03:05:00 -05:00
|
|
|
index++;
|
2008-01-31 09:53:53 -05:00
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
/* If barriers are supported, tell block layer that queue is ordered */
|
2008-05-02 22:50:50 -04:00
|
|
|
if (virtio_has_feature(vdev, VIRTIO_BLK_F_BARRIER))
|
2007-10-21 21:03:38 -04:00
|
|
|
blk_queue_ordered(vblk->disk->queue, QUEUE_ORDERED_TAG, NULL);
|
|
|
|
|
2008-05-16 05:17:03 -04:00
|
|
|
/* 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);
|
|
|
|
|
2008-02-04 23:49:56 -05:00
|
|
|
/* Host must always specify the capacity. */
|
2008-05-02 22:50:49 -04:00
|
|
|
vdev->config->get(vdev, offsetof(struct virtio_blk_config, capacity),
|
|
|
|
&cap, sizeof(cap));
|
2007-10-21 21:03:38 -04:00
|
|
|
|
|
|
|
/* 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);
|
|
|
|
|
2008-02-04 23:49:56 -05:00
|
|
|
/* 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);
|
2007-10-21 21:03:38 -04:00
|
|
|
if (!err)
|
|
|
|
blk_queue_max_segment_size(vblk->disk->queue, v);
|
|
|
|
|
2008-02-04 23:49:56 -05:00
|
|
|
err = virtio_config_val(vdev, VIRTIO_BLK_F_SEG_MAX,
|
|
|
|
offsetof(struct virtio_blk_config, seg_max),
|
|
|
|
&v);
|
2007-10-21 21:03:38 -04:00
|
|
|
if (!err)
|
|
|
|
blk_queue_max_hw_segments(vblk->disk->queue, v);
|
|
|
|
|
2008-05-29 05:08:26 -04:00
|
|
|
/* 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);
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
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;
|
|
|
|
|
2008-02-04 23:50:03 -05:00
|
|
|
/* Nothing should be pending. */
|
2007-10-21 21:03:38 -04:00
|
|
|
BUG_ON(!list_empty(&vblk->reqs));
|
2008-02-04 23:50:03 -05:00
|
|
|
|
|
|
|
/* Stop all the virtqueues. */
|
|
|
|
vdev->config->reset(vdev);
|
|
|
|
|
2008-05-30 16:09:41 -04:00
|
|
|
del_gendisk(vblk->disk);
|
2007-10-21 21:03:38 -04:00
|
|
|
blk_cleanup_queue(vblk->disk->queue);
|
|
|
|
put_disk(vblk->disk);
|
|
|
|
mempool_destroy(vblk->pool);
|
2007-11-19 11:20:42 -05:00
|
|
|
vdev->config->del_vq(vblk->vq);
|
2007-10-21 21:03:38 -04:00
|
|
|
kfree(vblk);
|
|
|
|
}
|
|
|
|
|
|
|
|
static struct virtio_device_id id_table[] = {
|
|
|
|
{ VIRTIO_ID_BLOCK, VIRTIO_DEV_ANY_ID },
|
|
|
|
{ 0 },
|
|
|
|
};
|
|
|
|
|
2008-05-02 22:50:50 -04:00
|
|
|
static unsigned int features[] = {
|
|
|
|
VIRTIO_BLK_F_BARRIER, VIRTIO_BLK_F_SEG_MAX, VIRTIO_BLK_F_SIZE_MAX,
|
2008-05-29 05:08:26 -04:00
|
|
|
VIRTIO_BLK_F_GEOMETRY, VIRTIO_BLK_F_RO, VIRTIO_BLK_F_BLK_SIZE,
|
2008-05-02 22:50:50 -04:00
|
|
|
};
|
|
|
|
|
2007-10-21 21:03:38 -04:00
|
|
|
static struct virtio_driver virtio_blk = {
|
2008-05-02 22:50:50 -04:00
|
|
|
.feature_table = features,
|
|
|
|
.feature_table_size = ARRAY_SIZE(features),
|
2007-10-21 21:03:38 -04:00
|
|
|
.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)
|
|
|
|
{
|
2008-01-31 09:53:53 -05:00
|
|
|
major = register_blkdev(0, "virtblk");
|
|
|
|
if (major < 0)
|
|
|
|
return major;
|
2007-10-21 21:03:38 -04:00
|
|
|
return register_virtio_driver(&virtio_blk);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __exit fini(void)
|
|
|
|
{
|
2008-01-31 09:53:53 -05:00
|
|
|
unregister_blkdev(major, "virtblk");
|
2007-10-21 21:03:38 -04:00
|
|
|
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");
|