Re: [GIT PULL] KVM (mostly s390) fixes for Linux 7.2-rc7

[Paolo Bonzini]

On Mon, Aug 10, 2026 at 11:30 PM Yosry Ahmed <yosry@xxxxxxxxxx> wrote:
> I assume the answer is yes, and something is broken. But if the answer
> is no, then we shouldn't be looking any further and the fix proposed
> by you/Sean is correct regardless.

Even if the answer is yes, the exact same code works on Intel just
with INVVPID instead of INVLPGA, so that's fishy.

We did ask AMD a few months ago, and again about a month ago after
confirming that it's happening even with recent microcode (and also
happens on Milan processors that would not be affected by other
INVLPGA errata), but haven't gotten any answer yet.

> My goal is to make sure we are not papering over an existing bug in
> KVM's Hyper-V code. The fact that hyperv_tlb_flush is flaky with NPT=0
> makes me suspect there's a bug somewhere in this code, and it might be
> manifesting differently for NPT=0/1.

I've meant to debug the test but didn't have time for it... Maybe
today. It seems to be a different issue since it's 100% reproducible
with npt=0 and takes hundreds of hours with npt=1.

Paolo