Re: [PATCH 4/7] KVM: x86/xen: Remove unnecessary irqsave from GPC lock usage in xen.c

[David Woodhouse]

On Mon, 2026-05-11 at 10:48 -0700, Sean Christopherson wrote:
> On Mon, May 11, 2026, David Woodhouse wrote:
> > On Mon, 2026-05-11 at 09:51 -0700, Sean Christopherson wrote:
> > > On Fri, May 08, 2026, David Woodhouse wrote:
> > > > From: David Woodhouse <dwmw@xxxxxxxxxxxx>
> > > > 
> > > > Now that the hardirq path (xen_timer_callback and set_evtchn_fast) uses
> > > > read_trylock() instead of read_lock_irqsave(), the remaining GPC lock
> > > > users in xen.c are only called from process context (vcpu_run, ioctls).
> > > > There is no need to disable interrupts to prevent concurrent access from
> > > > a hardirq user, since the hardirq path no longer takes the lock.
> > > 
> > > No longer _waits_ on the lock, correct?  I.e. the hardirq path can still take the
> > > lock, but only ever does so using trylock.
> > 
> > Indeed so. Want me to fix that?
> 
> Nah, at least not yet.  I'm in the process of wrapping my head around all of the
> in-flight gpc series (namely yours and Fred's), just wanted to make sure I wasn't
> missing something.

I think Fred has a few performance fixes and is due to send another
version once he can find the cycles to do so.

My assumption (without actually *talking* to Fred about it) was that he
would be able to rebase fairly trivially on top of this series.

Attachment: smime.p7s
Description: S/MIME cryptographic signature