From: David Woodhouse <dwmw2@infradead.org>
To: Sean Christopherson <seanjc@google.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>,
Jonathan Corbet <corbet@lwn.net>,
Shuah Khan <skhan@linuxfoundation.org>,
Thomas Gleixner <tglx@kernel.org>, Ingo Molnar <mingo@redhat.com>,
Borislav Petkov <bp@alien8.de>,
Dave Hansen <dave.hansen@linux.intel.com>,
x86@kernel.org, "H. Peter Anvin" <hpa@zytor.com>,
Vitaly Kuznetsov <vkuznets@redhat.com>,
Juergen Gross <jgross@suse.com>,
Boris Ostrovsky <boris.ostrovsky@oracle.com>,
Paul Durrant <paul@xen.org>, Jonathan Cameron <jic23@kernel.org>,
Sascha Bischoff <Sascha.Bischoff@arm.com>,
Marc Zyngier <maz@kernel.org>, Joey Gouly <joey.gouly@arm.com>,
Jack Allister <jalliste@amazon.com>,
Dongli Zhang <dongli.zhang@oracle.com>,
joe.jin@oracle.com, kvm@vger.kernel.org,
linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org,
xen-devel@lists.xenproject.org, linux-kselftest@vger.kernel.org
Subject: Re: [PATCH v7 17/36] KVM: x86: Allow KVM master clock mode when TSCs are offset from each other
Date: Tue, 11 Aug 2026 16:05:03 +0100 [thread overview]
Message-ID: <e256fa4af96e916ba30019cbba501fa896025fb0.camel@infradead.org> (raw)
In-Reply-To: <ansywxh0VX5rtfWc@google.com>
[-- Attachment #1: Type: text/plain, Size: 3633 bytes --]
On Tue, 2026-08-11 at 07:33 -0700, Sean Christopherson wrote:
>
> Actually, why are KVM_{G,S}ET_CLOCK_GUEST vCPU-scoped? Per the documentation,
> the API "Sets the KVM clock (for the whole VM) in terms of the vCPU TSC". If
> the APIs are VM-scoped instead of vCPU-scoped, then KVM can simply save/restore
> what's in the per-VM masterclock state, no?
They're vCPU-scoped because they need to be tied to a guest TSC (on
live migration, neither ka->master_cycle_now nor ka->master_kernel_ns
are useful — those are the "per-VM masterclock state").
Theoretically, guest TSCs can be different on each vCPU (different
offset, different *rate* even. Not that we allow KVM_[GS]ET_CLOCK_GUEST
at different rates, I concede).
So they operate in the context of a given vCPU, and *its* TSC.
And I think I'm going to defend that 'theoretical they can be
different', because I *would* like to eliminate the ways that a *guest*
can force non-masterclock mode, and that does mean allowing the offset-
TSC case.
FWIW in my local tree I've just extended the pvclock_migration_test to
test precisely the thing you were concerned about: three vCPUs with
divergent TSC offsets, migrated by setting each vCPU's TSC and then
invoking KVM_SET_CLOCK_GUEST once, through vCPU0. Masterclock stays
active, TSC_STABLE_BIT is correctly clear, and all three vCPUs'
pvclocks (and KVM_GET_CLOCK) agree to within a nanosecond afterwards.
I'll include that in the next spin.
> That would also help address my concerns about sanity checking the TSC frequency
> against the kvmclock frequency, as the APIs are much more blatantly about saving
> and restoring masterclock state. For whatever reason, it feels more natural for
> me to say that KVM_SET_CLOCK_GUEST will fail if the target frequency doesn't
> (fuzzily?) match the frequency at which the masterclock is already configured.
> Probably because use_master_clock directly gates that information? Whereas the
> vCPU's frequency is independently configured but obviously influences masterclock
> mode.
I am perfectly happy to say that the *existing* check as I have coded
it, is matching against the frequency at which the masterclock is
configured. Because we can't *get* there if the guest is not in
masterclock mode, and it can't be in masterclock mode unless all its
vCPUs are running at the same rate, which *is* the master clock rate.
:)
I guess I could even concede to change the actual code rather than just
the comment... (untested)
Still needs the *offset* of the vCPU though.
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -3595,15 +3595,15 @@ static int kvm_vcpu_ioctl_set_clock_guest(struct kvm_vcpu *v, void __user *argp
if (kvm_caps.has_tsc_control)
curr_tsc_hz = kvm_scale_tsc(curr_tsc_hz,
- v->arch.l1_tsc_scaling_ratio);
+ ka->master_tsc_scaling_ratio);
/*
* The mul/shift in the provided pvclock structure encode the guest
* TSC frequency at which it was generated. Sanity-check that it is
- * consistent with this vCPU's effective TSC frequency, allowing a
- * discrepancy of 1 kHz either way since independently calibrated
- * hosts will not measure precisely the same value even for the
- * same nominal frequency.
+ * consistent with the frequency at which the masterclock is
+ * configured, allowing a discrepancy of 1 kHz either way since
+ * independently calibrated hosts will not measure precisely the
+ * same value even for the same nominal frequency.
*/
if (user_tsc_hz < curr_tsc_hz - 1000 ||
user_tsc_hz > curr_tsc_hz + 1000) {
[-- Attachment #2: smime.p7s --]
[-- Type: application/pkcs7-signature, Size: 6179 bytes --]
next prev parent reply other threads:[~2026-08-11 15:05 UTC|newest]
Thread overview: 74+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-28 14:39 [PATCH v7 00/36] Cleaning up the KVM clock mess David Woodhouse
2026-07-28 14:39 ` [PATCH v7 01/36] KVM: x86: Improve accuracy of KVM clock when TSC scaling is in force David Woodhouse
2026-07-28 14:39 ` [PATCH v7 02/36] KVM: x86: Explicitly disable TSC scaling without CONSTANT_TSC David Woodhouse
2026-07-28 14:39 ` [PATCH v7 03/36] KVM: x86: Activate master clock immediately on vCPU creation David Woodhouse
2026-07-28 14:39 ` [PATCH v7 04/36] KVM: x86: Avoid NTP frequency skew for KVM clock on 32-bit host David Woodhouse
2026-07-28 14:39 ` [PATCH v7 05/36] KVM: x86: Fold __get_kvmclock() into get_kvmclock() David Woodhouse
2026-07-28 23:06 ` Sean Christopherson
2026-07-28 14:39 ` [PATCH v7 06/36] KVM: x86: Drop CPU pinning in get_kvmclock() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 07/36] KVM: x86: Restructure get_kvmclock() David Woodhouse
2026-07-28 23:04 ` Sean Christopherson
2026-07-28 14:39 ` [PATCH v7 08/36] KVM: x86: Fix KVM clock precision in get_kvmclock() with TSC scaling David Woodhouse
2026-07-28 14:39 ` [PATCH v7 09/36] KVM: x86: Use get_kvmclock() in kvm_get_wall_clock_epoch() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 10/36] KVM: x86: Fix compute_guest_tsc() to handle negative time deltas David Woodhouse
2026-07-28 14:39 ` [PATCH v7 11/36] KVM: x86: Restructure kvm_guest_time_update() for TSC upscaling David Woodhouse
2026-08-04 1:26 ` Sean Christopherson
2026-07-28 14:39 ` [PATCH v7 12/36] KVM: x86: Simplify and comment kvm_get_time_scale() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 13/36] KVM: x86: Remove implicit rdtsc() from kvm_compute_l1_tsc_offset() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 14/36] KVM: x86: Improve synchronization in kvm_synchronize_tsc() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 15/36] KVM: x86: Kill last_tsc_{nsec,write,offset} fields David Woodhouse
2026-07-28 14:39 ` [PATCH v7 16/36] KVM: x86: Replace nr_vcpus_matched_tsc count with all_vcpus_matched_tsc bool David Woodhouse
2026-07-28 14:39 ` [PATCH v7 17/36] KVM: x86: Allow KVM master clock mode when TSCs are offset from each other David Woodhouse
2026-08-10 17:47 ` Sean Christopherson
2026-08-10 18:20 ` David Woodhouse
2026-08-10 20:56 ` Sean Christopherson
2026-08-10 21:05 ` David Woodhouse
2026-08-11 14:16 ` Sean Christopherson
2026-08-11 14:33 ` Sean Christopherson
2026-08-11 15:05 ` David Woodhouse [this message]
2026-08-11 16:40 ` Sean Christopherson
2026-08-11 17:18 ` David Woodhouse
2026-08-11 17:28 ` Sean Christopherson
2026-08-11 17:32 ` David Woodhouse
2026-08-11 18:41 ` Sean Christopherson
2026-08-11 21:00 ` David Woodhouse
2026-08-11 21:46 ` Sean Christopherson
2026-08-11 21:56 ` David Woodhouse
2026-08-11 23:14 ` Sean Christopherson
2026-08-14 7:01 ` David Woodhouse
2026-08-14 17:54 ` Sean Christopherson
2026-08-15 7:02 ` David Woodhouse
2026-08-13 8:44 ` David Woodhouse
2026-07-28 14:39 ` [PATCH v7 18/36] KVM: x86: Factor out kvm_use_master_clock() David Woodhouse
2026-07-28 14:39 ` [PATCH v7 19/36] KVM: x86: Avoid gratuitous global clock updates David Woodhouse
2026-07-28 14:40 ` [PATCH v7 20/36] KVM: x86/xen: Prevent runstate times from becoming negative David Woodhouse
2026-07-28 14:40 ` [PATCH v7 21/36] KVM: x86: Avoid redundant masterclock updates from multiple vCPUs David Woodhouse
2026-07-28 14:40 ` [PATCH v7 22/36] KVM: x86: Remove runtime Xen TSC frequency CPUID update David Woodhouse
2026-07-28 14:40 ` [PATCH v7 23/36] KVM: x86: Re-synchronize TSC after KVM_SET_TSC_KHZ David Woodhouse
2026-07-28 14:40 ` [PATCH v7 24/36] KVM: x86: Use ktime_get_snapshot_id() for master clock David Woodhouse
2026-07-28 14:40 ` [PATCH v7 25/36] KVM: x86: Compute kvmclock base without pvclock_gtod_data David Woodhouse
2026-07-28 14:40 ` [PATCH v7 26/36] KVM: x86: Cache host vclock_mode for masterclock eligibility checks David Woodhouse
2026-07-28 14:40 ` [PATCH v7 27/36] KVM: x86: Remove pvclock_gtod_data and private timekeeping code David Woodhouse
2026-07-28 14:40 ` [PATCH v7 28/36] KVM: x86: Activate master clock from kvm_arch_init_vm() David Woodhouse
2026-07-28 14:40 ` [PATCH v7 29/36] UAPI: x86: Move pvclock-abi to UAPI for x86 platforms David Woodhouse
2026-07-28 14:40 ` [PATCH v7 30/36] KVM: selftests: Use UAPI pvclock-abi.h in xen_shinfo_test David Woodhouse
2026-07-28 14:40 ` [PATCH v7 31/36] KVM: x86: Add KVM_[GS]ET_CLOCK_GUEST for accurate KVM clock migration David Woodhouse
2026-07-31 23:24 ` Sean Christopherson
2026-08-01 8:07 ` David Woodhouse
2026-08-04 23:38 ` Sean Christopherson
2026-08-05 9:26 ` David Woodhouse
2026-08-11 23:40 ` Sean Christopherson
2026-08-11 23:47 ` David Woodhouse
2026-08-14 7:27 ` David Woodhouse
2026-08-28 23:26 ` David Woodhouse
2026-07-28 14:40 ` [PATCH v7 32/36] KVM: x86: Add KVM_VCPU_TSC_SCALE and fix the documentation on TSC migration David Woodhouse
2026-07-28 14:40 ` [PATCH v7 33/36] KVM: selftests: Add KVM/PV clock selftest to prove timer correction David Woodhouse
2026-07-31 23:32 ` Sean Christopherson
2026-07-28 14:40 ` [PATCH v7 34/36] KVM: selftests: Add master clock offset test David Woodhouse
2026-07-31 23:38 ` Sean Christopherson
2026-07-28 14:40 ` [PATCH v7 35/36] KVM: selftests: Add Xen/generic CPUID timing leaf test David Woodhouse
2026-07-28 14:40 ` [PATCH v7 36/36] KVM: selftests: Add Xen runstate migration test David Woodhouse
2026-07-28 23:18 ` [PATCH v7 00/36] Cleaning up the KVM clock mess Sean Christopherson
2026-07-29 10:42 ` David Woodhouse
2026-08-10 16:42 ` Sean Christopherson
2026-08-10 16:56 ` David Woodhouse
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=e256fa4af96e916ba30019cbba501fa896025fb0.camel@infradead.org \
--to=dwmw2@infradead.org \
--cc=Sascha.Bischoff@arm.com \
--cc=boris.ostrovsky@oracle.com \
--cc=bp@alien8.de \
--cc=corbet@lwn.net \
--cc=dave.hansen@linux.intel.com \
--cc=dongli.zhang@oracle.com \
--cc=hpa@zytor.com \
--cc=jalliste@amazon.com \
--cc=jgross@suse.com \
--cc=jic23@kernel.org \
--cc=joe.jin@oracle.com \
--cc=joey.gouly@arm.com \
--cc=kvm@vger.kernel.org \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-kselftest@vger.kernel.org \
--cc=maz@kernel.org \
--cc=mingo@redhat.com \
--cc=paul@xen.org \
--cc=pbonzini@redhat.com \
--cc=seanjc@google.com \
--cc=skhan@linuxfoundation.org \
--cc=tglx@kernel.org \
--cc=vkuznets@redhat.com \
--cc=x86@kernel.org \
--cc=xen-devel@lists.xenproject.org \
/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
all inboxes | Powered by JetHome®