On Thu, 2024-08-01 at 20:54 +0200, Thomas Gleixner wrote:

On Thu, Aug 01 2024 at 16:14, Michael Kelley wrote:

quoted

From: Thomas Gleixner <redacted> Sent: Thursday, August
1, 2024 7:21 AM
FWIW, in Hyper-V guests with the Hyper-V quirk removed, tglx's new
sequence does *not* stop the PIT. But this sequence does:

outb_p(0x30, PIT_MODE);
outb_p(0xff, PIT_CH0);
outb_p(0xff, PIT_CH0);

outb_p(0x30, PIT_MODE);
outb_p(0xff, PIT_CH0);

That works on bare metal too

What about writing *just* the MODE? The datasheet definitely seems to
be saying that should work too. And I believe it works on Hyper-V;
that's what the "workaround" has been doing, because Hyper-V actually
got it right.

quoted

I don't have a convenient way to test my sequence on KVM.

But still fails in KVM

By KVM you mean the in-kernel one that we want to kill because everyone
should be using userspace IRQ chips these days?

I don't know that we care; the failure mode is basically harmless so
let's just let it get fixed in KVM rather than pandering to it?