From: Stefan Hajnoczi <stefanha@redhat.com>
To: Roman Kiryanov <rkir@google.com>
Cc: virtio-dev@lists.oasis-open.org, JP Cottin <jpcottin@google.com>,
Jorge Moreira Broche <jemoreira@google.com>
Subject: Re: [virtio-dev] virtio-snd and snapshots (e.g. in QEMU) when audio is active
Date: Thu, 20 Jul 2023 11:40:34 -0400 [thread overview]
Message-ID: <20230720154034.GE184015@fedora> (raw)
In-Reply-To: <CAOGAQeoLSGR2RtRHX0kyS5dxK=wa+mDZ8BRyZX5oDKaL16LdSg@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 1769 bytes --]
On Wed, Jul 19, 2023 at 04:21:08PM -0700, Roman Kiryanov wrote:
> Hi,
>
> I work in Android Studio Emulator and we use virtio-snd (implemented
> ourselves) for audio output/input. According to the spec (1.2), the
> device has one TX virtqueue for all output streams and one RX
> virtqueue for all input streams. Each stream may and usually have more
> than one period (I request 4 periods).
>
> Because virtqueues are shared between streams (if there are more than
> one stream in the same direction), I cannot fetch vq messages when a
> stream needs one. I fetch vq messages (and put them into my own buffer
> to process them later) when the kernel puts them into a vq. I hope
> this is correct. I think I tried processing them immediately (at least
> for TX) but the kernel was not happy with this because I was draining
> the buffer too fast causing XRUN.
>
> If a snapshot request comes when audio streams are active I may have
> several unprocessed messages for several streams for both TX and RX.
> In my case messages are VirtQueueElement* which I don't think can be
> saved directly.
>
> Could you please advise what a device is expected to do in this case?
Do you mean QEMU's VirtQueueElement? There are devices in QEMU that
save/load in-flight VirtQueueElements. See qemu_put_virtqueue_element().
Often devices quiesce (e.g. by draining in-flight I/O requests) when the
VM is stopped before the device state is saved. That makes life simpler.
Stefan
>
> Thank you.
>
> Regards,
> Roman.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: virtio-dev-unsubscribe@lists.oasis-open.org
> For additional commands, e-mail: virtio-dev-help@lists.oasis-open.org
>
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
next prev parent reply other threads:[~2023-07-20 15:41 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-07-19 23:21 [virtio-dev] virtio-snd and snapshots (e.g. in QEMU) when audio is active Roman Kiryanov
2023-07-20 15:40 ` Stefan Hajnoczi [this message]
2023-07-22 5:15 ` Roman Kiryanov
2023-07-21 7:44 ` Matias Ezequiel Vara Larsen
2023-07-22 5:19 ` Roman Kiryanov
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=20230720154034.GE184015@fedora \
--to=stefanha@redhat.com \
--cc=jemoreira@google.com \
--cc=jpcottin@google.com \
--cc=rkir@google.com \
--cc=virtio-dev@lists.oasis-open.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