On Sat, Jun 05, 2021 at 05:40:17PM -0500, michael.christie@oracle.com wrote:

On 6/3/21 5:16 PM, Mike Christie wrote:

quoted

On 6/3/21 9:37 AM, Stefan Hajnoczi wrote:

quoted

On Tue, May 25, 2021 at 01:05:51PM -0500, Mike Christie wrote:

quoted

The following patches apply over linus's tree or mst's vhost branch
and my cleanup patchset:

https://urldefense.com/v3/__https://lists.linuxfoundation.org/pipermail/virtualization/2021-May/054354.html__;!!GqivPVa7Brio!P55eslnMW_iZkoTUZckwhnSw_9Z35JBScgtSYRh4CMFTRkSsWwKOYqY0huUfBfIPM8BV$ 

These patches allow us to support multiple vhost workers per device. I
ended up just doing Stefan's original idea where userspace has the
kernel create a worker and we pass back the pid. This has the benefit
over the workqueue and userspace thread approach where we only have
one'ish code path in the kernel during setup to detect old tools. The
main IO paths and device/vq setup/teardown paths all use common code.

The kernel patches here allow us to then do N workers device and also
share workers across devices.

I've also included a patch for qemu so you can get an idea of how it
works. If we are ok with the kernel code then I'll break that up into
a patchset and send to qemu-devel.

It seems risky to allow userspace process A to "share" a vhost worker
thread with userspace process B based on a matching pid alone. Should
they have ptrace_may_access() or similar?

I'm not sure. I already made it a little restrictive in this posting, but

Made a mistake here. In this posting I did not make it restrictive and
I was allowing any old 2 processes to share. So we would need something
like ptrace_may_access if go this route.

If we restrict sharing workers with the same owner, then I'm not sure if
need anything.

Agreed.

Sharing between processes becomes most interesting when there is busy
polling (because it consumes CPU and we should consolidate polling onto
as few CPUs as possible). Without polling we can just share the threads
within a process.

Stefan