Re: [PATCH v2] KVM: arm64: timers: Convert per-vcpu virtual offset to a global value

[Oliver Upton <oliver.upton@linux.dev>]

On Tue, Feb 28, 2023 at 07:18:38PM +0000, Marc Zyngier wrote:
> On Tue, 28 Feb 2023 11:26:07 +0000,
> Joey Gouly <joey.gouly@arm.com> wrote:
> > 
> > Hi Marc,
> > 
> > On Fri, Feb 24, 2023 at 07:16:40PM +0000, Marc Zyngier wrote:
> > > Having a per-vcpu virtual offset is a pain. It needs to be synchronized
> > > on each update, and expands badly to a setup where different timers can
> > > have different offsets, or have composite offsets (as with NV).
> > > 
> > > So let's start by replacing the use of the CNTVOFF_EL2 shadow register
> > > (which we want to reclaim for NV anyway), and make the virtual timer
> > > carry a pointer to a VM-wide offset.
> > > 
> > > This simplifies the code significantly. It also addresses two terrible bugs:
> > > 
> > > - The use of CNTVOFF_EL2 leads to some nice offset corruption
> > >   when the sysreg gets reset, as reported by Joey.
> > > 
> > > - The kvm mutex is taken from a vcpu ioctl, which goes against
> > >   the locking rules...
> > > 
> > > Reported-by: Joey Gouly <joey.gouly@arm.com>
> > > Reviewed-by: Reiji Watanabe <reijiw@google.com>
> > > Signed-off-by: Marc Zyngier <maz@kernel.org>
> > > Link: https://lore.kernel.org/r/20230224173915.GA17407@e124191.cambridge.arm.com
> > 
> > Fixes my mismatched timer offset issues.
> > 
> > Tested-by: Joey Gouly <joey.gouly@arm.com>
> 
> Thanks for having given it a go. Hopefully Oliver will be able to send
> this to as a fix shortly.

Absolutely, I have this queued up. Thanks for testing the patch Joey.

Want to see if I can get push access to the kvmarm repository in time,
otherwise I'll send a pull to you Marc.

-- 
Thanks,
Oliver