From: Stefan Hajnoczi <stefanha@redhat.com>
To: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Cc: qemu-block@nongnu.org, Kevin Wolf <kwolf@redhat.com>,
Hanna Reitz <hreitz@redhat.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>, Fam Zheng <fam@euphon.net>,
qemu-devel@nongnu.org
Subject: Re: [PATCH 1/2] virtio-scsi: replace VirtIOBlock dataplane_{start/starting/stopped} with enum
Date: Mon, 8 Aug 2022 11:43:11 -0400 [thread overview]
Message-ID: <YvEvD1X9/QSvyMGP@fedora> (raw)
In-Reply-To: <20220808094147.612472-2-eesposit@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 1734 bytes --]
On Mon, Aug 08, 2022 at 05:41:46AM -0400, Emanuele Giuseppe Esposito wrote:
> Simplify the various dataplane stages in dataplane_start/stop by using
> a single enum instead of having multiple flags.
>
> Read/write the enum atomically, as it can be read also by iothread
> callbacks.
What guarantees that these relaxed loads/stores always produce
DATAPLANE_STARTING/STARTED in virtio_scsi_defer_to_dataplane() and not
an older value? Are there implicit memory barriers?
>
> Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
> ---
> hw/scsi/virtio-scsi-dataplane.c | 21 +++++++++------------
> hw/scsi/virtio-scsi.c | 10 ++++++----
> include/hw/virtio/virtio-scsi.h | 5 ++---
> include/hw/virtio/virtio.h | 7 +++++++
> 4 files changed, 24 insertions(+), 19 deletions(-)
>
> diff --git a/hw/scsi/virtio-scsi-dataplane.c b/hw/scsi/virtio-scsi-dataplane.c
> index a575c3f0cd..9ad73e3e19 100644
> --- a/hw/scsi/virtio-scsi-dataplane.c
> +++ b/hw/scsi/virtio-scsi-dataplane.c
> @@ -106,13 +106,12 @@ int virtio_scsi_dataplane_start(VirtIODevice *vdev)
> VirtIOSCSICommon *vs = VIRTIO_SCSI_COMMON(vdev);
> VirtIOSCSI *s = VIRTIO_SCSI(vdev);
>
> - if (s->dataplane_started ||
> - s->dataplane_starting ||
> + if (qatomic_read(&s->dataplane_state) <= DATAPLANE_STARTED ||
It's not obvious that the STOPPING and STOPPED constants have a value
greater than STARTING and STARTED. It could be other way around too. It
would be safer to write the code so there are no assumptions about the
constants:
VirtIODataplaneStates state = qatomic_read(&s->dataplane_state);
if (state == DATAPLANE_STARTING || state == DATAPLANE_STARTED || ...)
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2022-08-08 15:44 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-08 9:41 [PATCH 0/2] virtio-blk and scsi: replace dataplane_{start/stopping/started} Emanuele Giuseppe Esposito
2022-08-08 9:41 ` [PATCH 1/2] virtio-scsi: replace VirtIOBlock dataplane_{start/starting/stopped} with enum Emanuele Giuseppe Esposito
2022-08-08 15:43 ` Stefan Hajnoczi [this message]
2022-08-08 9:41 ` [PATCH 2/2] virtio-blk: replace dataplane_start/stopping/started " Emanuele Giuseppe Esposito
2022-08-08 15:43 ` Stefan Hajnoczi
2022-08-08 15:52 ` [PATCH 0/2] virtio-blk and scsi: replace dataplane_{start/stopping/started} Stefan Hajnoczi
2022-08-08 16:29 ` Stefan Hajnoczi
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=YvEvD1X9/QSvyMGP@fedora \
--to=stefanha@redhat.com \
--cc=eesposit@redhat.com \
--cc=fam@euphon.net \
--cc=hreitz@redhat.com \
--cc=kwolf@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-block@nongnu.org \
--cc=qemu-devel@nongnu.org \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).