mbox series

[v2,0/3] revert 9pfs reply truncation, wait for free room to reply

Message ID alpine.DEB.2.21.2005211225030.27502@sstabellini-ThinkPad-T480s (mailing list archive)
Headers show
Series revert 9pfs reply truncation, wait for free room to reply | expand

Message

Stefano Stabellini May 21, 2020, 7:26 p.m. UTC 
Hi all,

This short series reverts commit 16724a173049ac29c7b5ade741da93a0f46edff
becauses it is the cause for https://bugs.launchpad.net/bugs/1877688.

The original issue addressed by 16724a173049ac29c7b5ade741da93a0f46edff
is solved differently in this series by using qemu_coroutine_yield() to
wait for the client to free more data from the ring before sending the
reply.

Cheers,

Stefano

Changes in v2:
- add comments on barriers
- add patch to increase ring size to the max allowed

Comments

Greg Kurz May 25, 2020, 1:48 p.m. UTC | #1 
On Thu, 21 May 2020 12:26:18 -0700 (PDT)
Stefano Stabellini <sstabellini@kernel.org> wrote:

> Hi all,
> 
> This short series reverts commit 16724a173049ac29c7b5ade741da93a0f46edff
> becauses it is the cause for https://bugs.launchpad.net/bugs/1877688.
> 
> The original issue addressed by 16724a173049ac29c7b5ade741da93a0f46edff
> is solved differently in this series by using qemu_coroutine_yield() to
> wait for the client to free more data from the ring before sending the
> reply.
> 
> Cheers,
> 
> Stefano
> 

I wasn't really involved in the review of these patches, but I've
done some testing with virtio-9p which show no regression. I
assume you have done some testing on the Xen side as well. So
I've pushed the series to 9p-next. I shall send a PR in a day
or so.

Cheers,

--
Greg

> Changes in v2:
> - add comments on barriers
> - add patch to increase ring size to the max allowed