Re: [PATCH v2 1/1] io: make zerocopy fallback accounting more accurate

["Daniel P. Berrangé" <berrange@redhat.com>]

On Mon, Mar 09, 2026 at 05:42:08PM +0000, Tejus GK wrote:
> 
> 
> > On 9 Mar 2026, at 10:47 PM, Daniel P. Berrangé <berrange@redhat.com> wrote:
> > 
> > !-------------------------------------------------------------------|
> >  CAUTION: External Email
> > 
> > |-------------------------------------------------------------------!
> > 
> > On Mon, Mar 09, 2026 at 12:59:44PM -0400, Peter Xu wrote:
> >> On Mon, Mar 09, 2026 at 04:48:37PM +0000, Daniel P. Berrangé wrote:
> >>>> @@ -881,8 +881,8 @@ static int qio_channel_socket_flush_internal(QIOChannel *ioc,
> >>>>         sioc->zero_copy_sent += serr->ee_data - serr->ee_info + 1;
> >>>> 
> >>>>         /* If any sendmsg() succeeded using zero copy, mark zerocopy success */
> >>>> -        if (serr->ee_code != SO_EE_CODE_ZEROCOPY_COPIED) {
> >>>> -            sioc->new_zero_copy_sent_success = true;
> >>>> +        if (serr->ee_code == SO_EE_CODE_ZEROCOPY_COPIED) {
> >>>> +            sioc->zero_copy_fallback++;
> >>> 
> >>> ...this is counting the number of MSG_ERRQUEUE items, which is not
> >>> the same as the number of IO requests. That's why we only used it
> >>> as a boolean marker originally, rather than making it a counter.
> >> 
> >> Would the logic still work and better than before?  Say, it's a counter of
> >> "messages" rather than "IOs" then.
> > 
> > IIUC it is a counter of processing notifications which is not directly
> > correlated to any action by QEMU - neither bytes nor syscalls.
> 
> Please correct me if I'm wrong about this, isn’t each notification an information 
> about what happened to an individual IO?

If userspace hasn't read a queued notification yet, the kernel will
merge new notifications with the existing queued one.

The line above your change

  serr->ee_data - serr->ee_info + 1;

records how many notifications were merged, so we now how many
syscalls were processed.

If ee_code is  SO_EE_CODE_ZEROCOPY_COPIED though it means at least
one syscall resulted in a copy, but that doesn't imply that *all*
syscalls resulted in a copy.

AFAICT, it could be 1 out of a 1000 syscalls resulted in a copy,
or it could be 1000 out of 1000 resulted in a copy. We don't know.

IIUC the kernel's merging of notifications appears lossy wrt this
information. It could be partially mitigated by doing a flush for
notifications really really frequently but that feels like it would
have its own downsides


> >> The problem with the old code was we may report fallback=0 even if there
> >> can have fallback happened, as we mask that fact as long as one zerocopy
> >> happened in the whole batch between two flushes.  So it seems this (even if
> >> the counter is not per-IO) is still better.
> > 
> > Better for what purpose though ?
> > 
> > If we enabled zero-copy, it is useful to know if /something/ managed
> > to benefit from zero-copy. ie if /always/ fails to zero-copy then
> > we can diagnose that the NIC driver isn't capable of it, or there
> > is some other limitation.  If something manages to zero-copy, then
> > we know the feature is functionally working.
> > 
> > What will we do with a count of notificaitons ?
> 
> I was wondering if it can be useful for debugging live migration issues where zerocopy is 
> enabled. For instance, let’s say a zerocopy write failed due to the socket error queue being
> full. Now it could be either due to the out of order processing we had seen before 
> (https://github.com/qemu/qemu/commit/84005f4a2b8745e5934f955c045a0b4311cd0992) or 
> due to it getting filled up because some copies getting deferred. For the latter, this stat can be
> worthwile as a debugging stat. 



With regards,
Daniel
-- 
|: https://berrange.com       ~~        https://hachyderm.io/@berrange :|
|: https://libvirt.org          ~~          https://entangle-photo.org :|
|: https://pixelfed.art/berrange   ~~    https://fstop138.berrange.com :|