From: Vasileios Almpanis <vasileios.almpanis@virtuozzo.com>
Subject: Re: [Devel] [PATCH VZ10 2/5] drivers/vhost/blk: report correct used-ring lengths
Date: Thu, 20 Aug 2026 14:11:04 +0000 [thread overview]
Message-ID: <178723506422.906560.13876408914701409678.b4-review@b4> (raw)
In-Reply-To: <20260818150920.13123-3-andrey.zhadchenko@virtuozzo.com>
> We are expected to return amount of bytes written to the guest,
> which also includes status.
>
> https://virtuozzo.atlassian.net/browse/VSTOR-138640
> Fixes: 40a5928ec730 ("drivers/vhost: vhost-blk accelerator for virtio-blk guests")
> Signed-off-by: Andrey Zhadchenko <andrey.zhadchenko@virtuozzo.com>
>
> diff --git a/drivers/vhost/blk.c b/drivers/vhost/blk.c
> index 3def988cdf18..ae6b916f96e1 100644
> --- a/drivers/vhost/blk.c
> +++ b/drivers/vhost/blk.c
> @@ -439,12 +439,14 @@ static int vhost_blk_req_submit(struct vhost_blk_req *req)
>
> static int vhost_blk_req_handle(struct vhost_virtqueue *vq,
> struct virtio_blk_outhdr *hdr,
> - u16 head, u16 total_iov_nr)
> + u16 head, u16 out, u16 in)
> {
> struct vhost_blk *blk = container_of(vq->dev, struct vhost_blk, dev);
> struct vhost_blk_vq *blk_vq = container_of(vq, struct vhost_blk_vq, vq);
> struct vhost_blk_req *req;
> + u16 total_iov_nr = out + in;
> struct iov_iter iter;
> + size_t in_len;
> int ret, len;
> u8 status;
>
> @@ -455,8 +457,9 @@ static int vhost_blk_req_handle(struct vhost_virtqueue *vq,
> req->sector = hdr->sector;
> req->iov = blk_vq->iov;
> req->bio_err = 0;
> + in_len = iov_length(vq->iov + out, in);
>
> - if (iov_length(vq->iov, total_iov_nr) < sizeof(status))
> + if (in_len < sizeof(status) || in_len > INT_MAX)
> return -EINVAL;
>
> req->len = iov_length(vq->iov, total_iov_nr) - sizeof(status);
> @@ -492,14 +495,14 @@ static int vhost_blk_req_handle(struct vhost_virtqueue *vq,
> ret = vhost_blk_set_status(req, status);
> if (ret)
> break;
> - vhost_add_used_and_signal(&blk->dev, vq, head, len);
> + vhost_add_used_and_signal(&blk->dev, vq, head, in_len);
> break;
> default:
> status = VIRTIO_BLK_S_UNSUPP;
> ret = vhost_blk_set_status(req, status);
> if (ret)
> break;
> - vhost_add_used_and_signal(&blk->dev, vq, head, 0);
> + vhost_add_used_and_signal(&blk->dev, vq, head, sizeof(status));
> }
>
> return ret;
> @@ -555,7 +558,7 @@ static void vhost_blk_handle_guest_kick(struct vhost_work *work)
> break;
> }
>
> - ret = vhost_blk_req_handle(vq, &hdr, head, out + in);
> + ret = vhost_blk_req_handle(vq, &hdr, head, out, in);
> if (ret == -EAGAIN || ret == -ENOMEM) {
> vhost_discard_vq_desc(vq, 1);
> vhost_poll_queue(&vq->poll);
> @@ -604,7 +607,7 @@ static void vhost_blk_handle_host_kick(struct vhost_work *work)
> if (vhost_blk_set_status(req, status)) {
> vhostblk_vq_err(blk, vq, "Failed to write status");
> } else {
> - vhost_add_used(vq, req->head, req->len);
> + vhost_add_used(vq, req->head, req->len + sizeof(status));
Is req->len + sizeof(status) the right thing to use for write requests?
We should only write the status byte not status + payload length
--
Vasileios Almpanis <vasileios.almpanis@virtuozzo.com>
next prev parent reply other threads:[~2026-08-20 14:11 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-08-18 15:09 [Devel] [PATCH VZ10 0/5] vhost-blk: fix protocol handling and backend setup Andrey Zhadchenko
2026-08-18 15:09 ` [Devel] [PATCH VZ10 1/5] drivers/vhost/blk: harden get_id command Andrey Zhadchenko
2026-08-18 15:09 ` [Devel] [PATCH VZ10 2/5] drivers/vhost/blk: report correct used-ring lengths Andrey Zhadchenko
2026-08-20 14:11 ` Vasileios Almpanis [this message]
2026-08-24 15:04 ` Andrey Zhadchenko
2026-08-18 15:09 ` [Devel] [PATCH VZ10 3/5] drivers/vhost/blk: fix flush support Andrey Zhadchenko
2026-08-18 15:09 ` [Devel] [PATCH VZ10 4/5] drivers/vhost/blk: fix sector alignment calculation Andrey Zhadchenko
2026-08-18 22:57 ` Andrey Zhadchenko
2026-08-18 15:09 ` [Devel] [PATCH VZ10 5/5] drivers/vhost/blk: rework queue/backend setup Andrey Zhadchenko
2026-08-20 14:11 ` Vasileios Almpanis
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=178723506422.906560.13876408914701409678.b4-review@b4 \
--to=vasileios.almpanis@virtuozzo.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.