diff mbox series

[RFC,v2,09/10] vdpa_sim_blk: implement ramdisk behaviour

Message ID 20210128144127.113245-10-sgarzare@redhat.com (mailing list archive)
State New, archived
Headers show
Series vdpa: add vdpa simulator for block device | expand

Commit Message

Stefano Garzarella Jan. 28, 2021, 2:41 p.m. UTC 
The previous implementation wrote only the status of each request.
This patch implements a more accurate block device simulator,
providing a ramdisk-like behavior.

Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
---
v2:
- used %zd %zx to print size_t and ssize_t variables in dev_err()
- removed unnecessary new line [Jason]
- moved VIRTIO_BLK_T_GET_ID in another patch [Jason]
- used push/pull instead of write/read terminology
- added vdpasim_blk_check_range() to avoid overflows [Stefan]
- use vdpasim*_to_cpu instead of le*_to_cpu
- used vringh_kiov_length() helper [Jason]
---
 drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 164 ++++++++++++++++++++++++---
 1 file changed, 146 insertions(+), 18 deletions(-)

Comments

Jason Wang Feb. 1, 2021, 6:03 a.m. UTC | #1 
On 2021/1/28 下午10:41, Stefano Garzarella wrote:
> The previous implementation wrote only the status of each request.
> This patch implements a more accurate block device simulator,
> providing a ramdisk-like behavior.
>
> Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
> ---
> v2:
> - used %zd %zx to print size_t and ssize_t variables in dev_err()
> - removed unnecessary new line [Jason]
> - moved VIRTIO_BLK_T_GET_ID in another patch [Jason]
> - used push/pull instead of write/read terminology
> - added vdpasim_blk_check_range() to avoid overflows [Stefan]
> - use vdpasim*_to_cpu instead of le*_to_cpu
> - used vringh_kiov_length() helper [Jason]


Acked-by: Jason Wang <jasowang@redhat.com>


> ---
>   drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 164 ++++++++++++++++++++++++---
>   1 file changed, 146 insertions(+), 18 deletions(-)
>
> diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> index 999f9ca0b628..fc47e8320358 100644
> --- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> +++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
> @@ -3,6 +3,7 @@
>    * VDPA simulator for block device.
>    *
>    * Copyright (c) 2020, Mellanox Technologies. All rights reserved.
> + * Copyright (c) 2021, Red Hat Inc. All rights reserved.
>    *
>    */
>   
> @@ -13,6 +14,7 @@
>   #include <linux/sched.h>
>   #include <linux/vringh.h>
>   #include <linux/vdpa.h>
> +#include <linux/blkdev.h>
>   #include <uapi/linux/virtio_blk.h>
>   
>   #include "vdpa_sim.h"
> @@ -36,10 +38,151 @@
>   
>   static struct vdpasim *vdpasim_blk_dev;
>   
> +static bool vdpasim_blk_check_range(u64 start_sector, size_t range_size)
> +{
> +	u64 range_sectors = range_size >> SECTOR_SHIFT;
> +
> +	if (range_size > VDPASIM_BLK_SIZE_MAX * VDPASIM_BLK_SEG_MAX)
> +		return false;
> +
> +	if (start_sector > VDPASIM_BLK_CAPACITY)
> +		return false;
> +
> +	if (range_sectors > VDPASIM_BLK_CAPACITY - start_sector)
> +		return false;
> +
> +	return true;
> +}
> +
> +/* Returns 'true' if the request is handled (with or without an I/O error)
> + * and the status is correctly written in the last byte of the 'in iov',
> + * 'false' otherwise.
> + */
> +static bool vdpasim_blk_handle_req(struct vdpasim *vdpasim,
> +				   struct vdpasim_virtqueue *vq)
> +{
> +	size_t pushed = 0, to_pull, to_push;
> +	struct virtio_blk_outhdr hdr;
> +	ssize_t bytes;
> +	loff_t offset;
> +	u64 sector;
> +	u8 status;
> +	u32 type;
> +	int ret;
> +
> +	ret = vringh_getdesc_iotlb(&vq->vring, &vq->out_iov, &vq->in_iov,
> +				   &vq->head, GFP_ATOMIC);
> +	if (ret != 1)
> +		return false;
> +
> +	if (vq->out_iov.used < 1 || vq->in_iov.used < 1) {
> +		dev_err(&vdpasim->vdpa.dev, "missing headers - out_iov: %u in_iov %u\n",
> +			vq->out_iov.used, vq->in_iov.used);
> +		return false;
> +	}
> +
> +	if (vq->in_iov.iov[vq->in_iov.used - 1].iov_len < 1) {
> +		dev_err(&vdpasim->vdpa.dev, "request in header too short\n");
> +		return false;
> +	}
> +
> +	/* The last byte is the status and we checked if the last iov has
> +	 * enough room for it.
> +	 */
> +	to_push = vringh_kiov_length(&vq->in_iov) - 1;
> +
> +	to_pull = vringh_kiov_length(&vq->out_iov);
> +
> +	bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov, &hdr,
> +				      sizeof(hdr));
> +	if (bytes != sizeof(hdr)) {
> +		dev_err(&vdpasim->vdpa.dev, "request out header too short\n");
> +		return false;
> +	}
> +
> +	to_pull -= bytes;
> +
> +	type = vdpasim32_to_cpu(vdpasim, hdr.type);
> +	sector = vdpasim64_to_cpu(vdpasim, hdr.sector);
> +	offset = sector << SECTOR_SHIFT;
> +	status = VIRTIO_BLK_S_OK;
> +
> +	switch (type) {
> +	case VIRTIO_BLK_T_IN:
> +		if (!vdpasim_blk_check_range(sector, to_push)) {
> +			dev_err(&vdpasim->vdpa.dev,
> +				"reading over the capacity - offset: 0x%llx len: 0x%zx\n",
> +				offset, to_push);
> +			status = VIRTIO_BLK_S_IOERR;
> +			break;
> +		}
> +
> +		bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
> +					      vdpasim->buffer + offset,
> +					      to_push);
> +		if (bytes < 0) {
> +			dev_err(&vdpasim->vdpa.dev,
> +				"vringh_iov_push_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
> +				bytes, offset, to_push);
> +			status = VIRTIO_BLK_S_IOERR;
> +			break;
> +		}
> +
> +		pushed += bytes;
> +		break;
> +
> +	case VIRTIO_BLK_T_OUT:
> +		if (!vdpasim_blk_check_range(sector, to_pull)) {
> +			dev_err(&vdpasim->vdpa.dev,
> +				"writing over the capacity - offset: 0x%llx len: 0x%zx\n",
> +				offset, to_pull);
> +			status = VIRTIO_BLK_S_IOERR;
> +			break;
> +		}
> +
> +		bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov,
> +					      vdpasim->buffer + offset,
> +					      to_pull);
> +		if (bytes < 0) {
> +			dev_err(&vdpasim->vdpa.dev,
> +				"vringh_iov_pull_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
> +				bytes, offset, to_pull);
> +			status = VIRTIO_BLK_S_IOERR;
> +			break;
> +		}
> +		break;
> +
> +	default:
> +		dev_warn(&vdpasim->vdpa.dev,
> +			 "Unsupported request type %d\n", type);
> +		status = VIRTIO_BLK_S_IOERR;
> +		break;
> +	}
> +
> +	/* If some operations fail, we need to skip the remaining bytes
> +	 * to put the status in the last byte
> +	 */
> +	if (to_push - pushed > 0)
> +		vringh_kiov_advance(&vq->in_iov, to_push - pushed);
> +
> +	/* Last byte is the status */
> +	bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov, &status, 1);
> +	if (bytes != 1)
> +		return false;
> +
> +	pushed += bytes;
> +
> +	/* Make sure data is wrote before advancing index */
> +	smp_wmb();
> +
> +	vringh_complete_iotlb(&vq->vring, vq->head, pushed);
> +
> +	return true;
> +}
> +
>   static void vdpasim_blk_work(struct work_struct *work)
>   {
>   	struct vdpasim *vdpasim = container_of(work, struct vdpasim, work);
> -	u8 status = VIRTIO_BLK_S_OK;
>   	int i;
>   
>   	spin_lock(&vdpasim->lock);
> @@ -53,22 +196,7 @@ static void vdpasim_blk_work(struct work_struct *work)
>   		if (!vq->ready)
>   			continue;
>   
> -		while (vringh_getdesc_iotlb(&vq->vring, &vq->out_iov,
> -					    &vq->in_iov, &vq->head,
> -					    GFP_ATOMIC) > 0) {
> -			int write;
> -
> -			vq->in_iov.i = vq->in_iov.used - 1;
> -			write = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
> -						      &status, 1);
> -			if (write <= 0)
> -				break;
> -
> -			/* Make sure data is wrote before advancing index */
> -			smp_wmb();
> -
> -			vringh_complete_iotlb(&vq->vring, vq->head, write);
> -
> +		while (vdpasim_blk_handle_req(vdpasim, vq)) {
>   			/* Make sure used is visible before rasing the interrupt. */
>   			smp_wmb();
>   
> @@ -109,7 +237,7 @@ static int __init vdpasim_blk_init(void)
>   	dev_attr.config_size = sizeof(struct virtio_blk_config);
>   	dev_attr.get_config = vdpasim_blk_get_config;
>   	dev_attr.work_fn = vdpasim_blk_work;
> -	dev_attr.buffer_size = PAGE_SIZE;
> +	dev_attr.buffer_size = VDPASIM_BLK_CAPACITY << SECTOR_SHIFT;
>   
>   	vdpasim_blk_dev = vdpasim_create(&dev_attr);
>   	if (IS_ERR(vdpasim_blk_dev)) {
Stefan Hajnoczi Feb. 2, 2021, 9:41 a.m. UTC | #2 
On Thu, Jan 28, 2021 at 03:41:26PM +0100, Stefano Garzarella wrote:
> The previous implementation wrote only the status of each request.
> This patch implements a more accurate block device simulator,
> providing a ramdisk-like behavior.
> 
> Signed-off-by: Stefano Garzarella <sgarzare@redhat.com>
> ---
> v2:
> - used %zd %zx to print size_t and ssize_t variables in dev_err()
> - removed unnecessary new line [Jason]
> - moved VIRTIO_BLK_T_GET_ID in another patch [Jason]
> - used push/pull instead of write/read terminology
> - added vdpasim_blk_check_range() to avoid overflows [Stefan]
> - use vdpasim*_to_cpu instead of le*_to_cpu
> - used vringh_kiov_length() helper [Jason]
> ---
>  drivers/vdpa/vdpa_sim/vdpa_sim_blk.c | 164 ++++++++++++++++++++++++---
>  1 file changed, 146 insertions(+), 18 deletions(-)

Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
diff mbox series

Patch

diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
index 999f9ca0b628..fc47e8320358 100644
--- a/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
+++ b/drivers/vdpa/vdpa_sim/vdpa_sim_blk.c
@@ -3,6 +3,7 @@ 
  * VDPA simulator for block device.
  *
  * Copyright (c) 2020, Mellanox Technologies. All rights reserved.
+ * Copyright (c) 2021, Red Hat Inc. All rights reserved.
  *
  */
 
@@ -13,6 +14,7 @@ 
 #include <linux/sched.h>
 #include <linux/vringh.h>
 #include <linux/vdpa.h>
+#include <linux/blkdev.h>
 #include <uapi/linux/virtio_blk.h>
 
 #include "vdpa_sim.h"
@@ -36,10 +38,151 @@ 
 
 static struct vdpasim *vdpasim_blk_dev;
 
+static bool vdpasim_blk_check_range(u64 start_sector, size_t range_size)
+{
+	u64 range_sectors = range_size >> SECTOR_SHIFT;
+
+	if (range_size > VDPASIM_BLK_SIZE_MAX * VDPASIM_BLK_SEG_MAX)
+		return false;
+
+	if (start_sector > VDPASIM_BLK_CAPACITY)
+		return false;
+
+	if (range_sectors > VDPASIM_BLK_CAPACITY - start_sector)
+		return false;
+
+	return true;
+}
+
+/* Returns 'true' if the request is handled (with or without an I/O error)
+ * and the status is correctly written in the last byte of the 'in iov',
+ * 'false' otherwise.
+ */
+static bool vdpasim_blk_handle_req(struct vdpasim *vdpasim,
+				   struct vdpasim_virtqueue *vq)
+{
+	size_t pushed = 0, to_pull, to_push;
+	struct virtio_blk_outhdr hdr;
+	ssize_t bytes;
+	loff_t offset;
+	u64 sector;
+	u8 status;
+	u32 type;
+	int ret;
+
+	ret = vringh_getdesc_iotlb(&vq->vring, &vq->out_iov, &vq->in_iov,
+				   &vq->head, GFP_ATOMIC);
+	if (ret != 1)
+		return false;
+
+	if (vq->out_iov.used < 1 || vq->in_iov.used < 1) {
+		dev_err(&vdpasim->vdpa.dev, "missing headers - out_iov: %u in_iov %u\n",
+			vq->out_iov.used, vq->in_iov.used);
+		return false;
+	}
+
+	if (vq->in_iov.iov[vq->in_iov.used - 1].iov_len < 1) {
+		dev_err(&vdpasim->vdpa.dev, "request in header too short\n");
+		return false;
+	}
+
+	/* The last byte is the status and we checked if the last iov has
+	 * enough room for it.
+	 */
+	to_push = vringh_kiov_length(&vq->in_iov) - 1;
+
+	to_pull = vringh_kiov_length(&vq->out_iov);
+
+	bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov, &hdr,
+				      sizeof(hdr));
+	if (bytes != sizeof(hdr)) {
+		dev_err(&vdpasim->vdpa.dev, "request out header too short\n");
+		return false;
+	}
+
+	to_pull -= bytes;
+
+	type = vdpasim32_to_cpu(vdpasim, hdr.type);
+	sector = vdpasim64_to_cpu(vdpasim, hdr.sector);
+	offset = sector << SECTOR_SHIFT;
+	status = VIRTIO_BLK_S_OK;
+
+	switch (type) {
+	case VIRTIO_BLK_T_IN:
+		if (!vdpasim_blk_check_range(sector, to_push)) {
+			dev_err(&vdpasim->vdpa.dev,
+				"reading over the capacity - offset: 0x%llx len: 0x%zx\n",
+				offset, to_push);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
+					      vdpasim->buffer + offset,
+					      to_push);
+		if (bytes < 0) {
+			dev_err(&vdpasim->vdpa.dev,
+				"vringh_iov_push_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
+				bytes, offset, to_push);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		pushed += bytes;
+		break;
+
+	case VIRTIO_BLK_T_OUT:
+		if (!vdpasim_blk_check_range(sector, to_pull)) {
+			dev_err(&vdpasim->vdpa.dev,
+				"writing over the capacity - offset: 0x%llx len: 0x%zx\n",
+				offset, to_pull);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+
+		bytes = vringh_iov_pull_iotlb(&vq->vring, &vq->out_iov,
+					      vdpasim->buffer + offset,
+					      to_pull);
+		if (bytes < 0) {
+			dev_err(&vdpasim->vdpa.dev,
+				"vringh_iov_pull_iotlb() error: %zd offset: 0x%llx len: 0x%zx\n",
+				bytes, offset, to_pull);
+			status = VIRTIO_BLK_S_IOERR;
+			break;
+		}
+		break;
+
+	default:
+		dev_warn(&vdpasim->vdpa.dev,
+			 "Unsupported request type %d\n", type);
+		status = VIRTIO_BLK_S_IOERR;
+		break;
+	}
+
+	/* If some operations fail, we need to skip the remaining bytes
+	 * to put the status in the last byte
+	 */
+	if (to_push - pushed > 0)
+		vringh_kiov_advance(&vq->in_iov, to_push - pushed);
+
+	/* Last byte is the status */
+	bytes = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov, &status, 1);
+	if (bytes != 1)
+		return false;
+
+	pushed += bytes;
+
+	/* Make sure data is wrote before advancing index */
+	smp_wmb();
+
+	vringh_complete_iotlb(&vq->vring, vq->head, pushed);
+
+	return true;
+}
+
 static void vdpasim_blk_work(struct work_struct *work)
 {
 	struct vdpasim *vdpasim = container_of(work, struct vdpasim, work);
-	u8 status = VIRTIO_BLK_S_OK;
 	int i;
 
 	spin_lock(&vdpasim->lock);
@@ -53,22 +196,7 @@  static void vdpasim_blk_work(struct work_struct *work)
 		if (!vq->ready)
 			continue;
 
-		while (vringh_getdesc_iotlb(&vq->vring, &vq->out_iov,
-					    &vq->in_iov, &vq->head,
-					    GFP_ATOMIC) > 0) {
-			int write;
-
-			vq->in_iov.i = vq->in_iov.used - 1;
-			write = vringh_iov_push_iotlb(&vq->vring, &vq->in_iov,
-						      &status, 1);
-			if (write <= 0)
-				break;
-
-			/* Make sure data is wrote before advancing index */
-			smp_wmb();
-
-			vringh_complete_iotlb(&vq->vring, vq->head, write);
-
+		while (vdpasim_blk_handle_req(vdpasim, vq)) {
 			/* Make sure used is visible before rasing the interrupt. */
 			smp_wmb();
 
@@ -109,7 +237,7 @@  static int __init vdpasim_blk_init(void)
 	dev_attr.config_size = sizeof(struct virtio_blk_config);
 	dev_attr.get_config = vdpasim_blk_get_config;
 	dev_attr.work_fn = vdpasim_blk_work;
-	dev_attr.buffer_size = PAGE_SIZE;
+	dev_attr.buffer_size = VDPASIM_BLK_CAPACITY << SECTOR_SHIFT;
 
 	vdpasim_blk_dev = vdpasim_create(&dev_attr);
 	if (IS_ERR(vdpasim_blk_dev)) {