Re: [BUG] Guest OSes die simultaneously (bisected)
From: Breno Leitao
Date: Thu Jan 04 2024 - 05:01:17 EST
On Wed, Jan 03, 2024 at 02:22:23PM -0800, Paul E. McKenney wrote:
> I was
> therefore forced to bisect among the commits backported to the internal
> v5.19-based kernel, which fingered the backported version of the patch
> called out above.
Just to add some context to these backport, this commit (c59a1f106f5c)
was backported to the internal v5.19-based kernel in order to easily
backport these two fixes.
a16eb25b09c02a54c ("KVM: x86: Mask LVTPC when handling a PMI")
73554b29bd70546c1 ("KVM: x86/pmu: Synthesize at most one PMI per VM-exit")
They are required to solve the softlockup problem reported here:
https://lore.kernel.org/all/ZRwcpki67uhpAUKi@xxxxxxxxx/