From: Sean Christopherson <seanjc@google.com>
To: Vitaly Kuznetsov <vkuznets@redhat.com>
Cc: Anton Romanov <romanton@google.com>,
kvm@vger.kernel.org, pbonzini@redhat.com
Subject: Re: [PATCH] KVM: x86: Use current rather than snapshotted TSC frequency if it is constant
Date: Tue, 19 Apr 2022 16:13:49 +0000 [thread overview]
Message-ID: <Yl7fvbDz+ckj/psQ@google.com> (raw)
In-Reply-To: <87o80x3vkx.fsf@redhat.com>
On Tue, Apr 19, 2022, Vitaly Kuznetsov wrote:
> Sean Christopherson <seanjc@google.com> writes:
>
> > On Tue, Apr 19, 2022, Vitaly Kuznetsov wrote:
> >> Sean Christopherson <seanjc@google.com> writes:
> >> > The Hyper-V guest code also sets cpu_tsc_khz, should we WARN if that notifier is
> >> > invoked and Hyper-V told us there's a constant TSC?
...
> >> (apologies for the delayed reply)
> >>
> >> No, I think Hyper-V's "Reenlightenment" feature overrides (re-defines?)
> >> X86_FEATURE_CONSTANT_TSC. E.g. I've checked a VM on E5-2667 v4
> >> (Broadwell) CPU with no TSC scaling. This VM has 'constant_tsc' and will
> >> certainly get reenlightenment irq on migration.
> >
> > Ooh, so that a VM with a constant TSC be live migrated to another system with a
> > constant, but different, TSC. Does the below look correct as fixup for this patch?
> >
> > diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> > index ab336f7c82e4..a944e4ba5532 100644
> > --- a/arch/x86/kvm/x86.c
> > +++ b/arch/x86/kvm/x86.c
> > @@ -8708,10 +8708,18 @@ static void kvm_hyperv_tsc_notifier(void)
> > /* no guest entries from this point */
> > hyperv_stop_tsc_emulation();
> >
> > - /* TSC frequency always matches when on Hyper-V */
> > - for_each_present_cpu(cpu)
> > - per_cpu(cpu_tsc_khz, cpu) = tsc_khz;
> > - kvm_max_guest_tsc_khz = tsc_khz;
> > + /*
> > + * TSC frequency always matches when on Hyper-V. Skip the updates if
> > + * the TSC is "officially" constant, in which case KVM doesn't use the
> > + * per-CPU and max variables. Note, the notifier can still fire with
> > + * a constant TSC, e.g. if this VM (KVM is a Hyper-V guest) is migrated
> > + * to a system with a different TSC frequency.
> > + */
> > + if (!boot_cpu_has(X86_FEATURE_CONSTANT_TSC)) {
> > + for_each_present_cpu(cpu)
> > + per_cpu(cpu_tsc_khz, cpu) = tsc_khz;
> > + kvm_max_guest_tsc_khz = tsc_khz;
> > + }
>
> Looks good for cpu_tsc_khz but I'm not particularly sure about
> kvm_max_guest_tsc_khz.
Doh, ignore that, I got kvm_max_guest_tsc_khz confused with max_tsc_khz.
prev parent reply other threads:[~2022-04-19 16:13 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-04-14 18:31 [PATCH] KVM: x86: Use current rather than snapshotted TSC frequency if it is constant Anton Romanov
2022-04-14 19:33 ` Sean Christopherson
2022-04-14 20:42 ` Anton Romanov
2022-04-14 22:22 ` Sean Christopherson
2022-04-19 7:46 ` Vitaly Kuznetsov
2022-04-19 15:39 ` Sean Christopherson
2022-04-19 16:07 ` Vitaly Kuznetsov
2022-04-19 16:13 ` Sean Christopherson [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Yl7fvbDz+ckj/psQ@google.com \
--to=seanjc@google.com \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=romanton@google.com \
--cc=vkuznets@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox