mbox series

[v5,0/2] virtio-blk: support polling I/O and mq_ops->queue_rqs()

Message ID 20220405150924.147021-1-suwan.kim027@gmail.com (mailing list archive)
Headers show
Series virtio-blk: support polling I/O and mq_ops->queue_rqs() | expand

Message

Suwan Kim April 5, 2022, 3:09 p.m. UTC
This patch serise adds support for polling I/O and mq_ops->queue_rqs()
to virtio-blk driver.

Changes

v4 -> v5
    - patch1 : virtblk_poll
        - Replace "req_done" with "found" in virtblk_poll()
        - Split for loop into two distinct for loop in init_vq()
          that sets callback function for each default/poll queues
        - Replace "if (i == HCTX_TYPE_DEFAULT)" with "i != HCTX_TYPE_POLL"
          in virtblk_map_queues()
        - Replace "virtblk_unmap_data(req, vbr);" with
          "virtblk_unmap_data(req, blk_mq_rq_to_pdu(req);"
          in virtblk_complete_batch()
    
    - patch2 : virtio_queue_rqs
        - Instead of using vbr.sg_num field, use vbr->sg_table.nents.
          So, remove sg_num field in struct virtblk_req
        - Drop the unnecessary argument of virtblk_add_req() because it
          doens't need "data_sg" and "have_data". It can be derived from "vbr"
          argument.
        - Add Reviewed-by tag from Stefan

v3 -> v4
    - patch1 : virtblk_poll
        - Add print the number of default/read/poll queues in init_vq()
        - Add blk_mq_start_stopped_hw_queues() to virtblk_poll()
              virtblk_poll()
                  ...
                  if (req_done)
                                   blk_mq_start_stopped_hw_queues(vblk->disk->queue, true);
                  ...

    - patch2 : virtio_queue_rqs
        - Modify virtio_queue_rqs() to hold lock only once when it adds
          requests to virtqueue just before virtqueue notify.
          It will guarantee that virtio_queue_rqs() will not use
          previous req again.

v2 -> v3
        - Fix warning by kernel test robot
          
            static int virtblk_poll()
                ...
                if (!blk_mq_add_to_batch(req, iob, virtblk_result(vbr),
                                                   -> vbr->status,

v1 -> v2
        - To receive the number of poll queues from user,
          use module parameter instead of QEMU uapi change.

        - Add the comment about virtblk_map_queues().

        - Add support for mq_ops->queue_rqs() to implement submit side
          batch.

Suwan Kim (2):
  virtio-blk: support polling I/O
  virtio-blk: support mq_ops->queue_rqs()

 drivers/block/virtio_blk.c | 229 +++++++++++++++++++++++++++++++++----
 1 file changed, 206 insertions(+), 23 deletions(-)

Comments

Stefan Hajnoczi April 5, 2022, 3:46 p.m. UTC | #1
On Wed, Apr 06, 2022 at 12:09:22AM +0900, Suwan Kim wrote:
> This patch serise adds support for polling I/O and mq_ops->queue_rqs()
> to virtio-blk driver.
> 
> Changes
> 
> v4 -> v5
>     - patch1 : virtblk_poll
>         - Replace "req_done" with "found" in virtblk_poll()
>         - Split for loop into two distinct for loop in init_vq()
>           that sets callback function for each default/poll queues
>         - Replace "if (i == HCTX_TYPE_DEFAULT)" with "i != HCTX_TYPE_POLL"
>           in virtblk_map_queues()
>         - Replace "virtblk_unmap_data(req, vbr);" with
>           "virtblk_unmap_data(req, blk_mq_rq_to_pdu(req);"
>           in virtblk_complete_batch()
>     
>     - patch2 : virtio_queue_rqs
>         - Instead of using vbr.sg_num field, use vbr->sg_table.nents.
>           So, remove sg_num field in struct virtblk_req
>         - Drop the unnecessary argument of virtblk_add_req() because it
>           doens't need "data_sg" and "have_data". It can be derived from "vbr"
>           argument.
>         - Add Reviewed-by tag from Stefan
> 
> v3 -> v4
>     - patch1 : virtblk_poll
>         - Add print the number of default/read/poll queues in init_vq()
>         - Add blk_mq_start_stopped_hw_queues() to virtblk_poll()
>               virtblk_poll()
>                   ...
>                   if (req_done)
>                                    blk_mq_start_stopped_hw_queues(vblk->disk->queue, true);
>                   ...
> 
>     - patch2 : virtio_queue_rqs
>         - Modify virtio_queue_rqs() to hold lock only once when it adds
>           requests to virtqueue just before virtqueue notify.
>           It will guarantee that virtio_queue_rqs() will not use
>           previous req again.
> 
> v2 -> v3
>         - Fix warning by kernel test robot
>           
>             static int virtblk_poll()
>                 ...
>                 if (!blk_mq_add_to_batch(req, iob, virtblk_result(vbr),
>                                                    -> vbr->status,
> 
> v1 -> v2
>         - To receive the number of poll queues from user,
>           use module parameter instead of QEMU uapi change.
> 
>         - Add the comment about virtblk_map_queues().
> 
>         - Add support for mq_ops->queue_rqs() to implement submit side
>           batch.
> 
> Suwan Kim (2):
>   virtio-blk: support polling I/O
>   virtio-blk: support mq_ops->queue_rqs()
> 
>  drivers/block/virtio_blk.c | 229 +++++++++++++++++++++++++++++++++----
>  1 file changed, 206 insertions(+), 23 deletions(-)
> 
> -- 
> 2.26.3
> 

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