Re: [PATCH v9 06/21] KVM: x86: Avoid NTP frequency skew for KVM clock on 32-bit host

[Sean Christopherson <seanjc@google.com>]

On Mon, Aug 10, 2026, Sean Christopherson wrote:
> From: David Woodhouse <dwmw@amazon.co.uk>
> 
> Commit 53fafdbb8b21 ("KVM: x86: switch KVMCLOCK base to monotonic raw
> clock") did so only for 64-bit hosts, by capturing the boot offset from
> within the existing clocksource notifier update_pvclock_gtod().
> 
> That notifier was added in commit 16e8d74d2da9 ("KVM: x86: notifier for
> clocksource changes") but only on x86_64, because its original purpose
> was just to disable the "master clock" mode which is only supported on
> x86_64.
> 
> Now that the notifier is used for more than disabling master clock mode,
> enable it for the 32-bit build too so that get_kvmclock_base_ns() can be
> unaffected by NTP sync on 32-bit too.
> 
> Signed-off-by: David Woodhouse <dwmw@amazon.co.uk>
> Reviewed-by: Paul Durrant <paul@xen.org>
> [sean: rebase on top of ktime_mono_to_any() usage]
> Signed-off-by: Sean Christopherson <seanjc@google.com>
> ---

...

> @@ -7118,9 +7111,9 @@ int kvm_x86_vendor_init(struct kvm_x86_init_ops *ops)
>  
>  	if (pi_inject_timer == -1)
>  		pi_inject_timer = housekeeping_enabled(HK_TYPE_TIMER);
> -#ifdef CONFIG_X86_64
>  	pvclock_gtod_register_notifier(&pvclock_gtod_notifier);

From https://sashiko.dev/#/patchset/20260810225500.869288-1-seanjc%40google.com:

 : Does this add unnecessary overhead to the timekeeper update path on 32-bit
 : builds?
 : 
 : The commit message notes a rebase on top of ktime_mono_to_any() usage. Because
 : of that rebase, get_kvmclock_base_ns() now uses ktime_mono_to_any() directly
 : and no longer reads from pvclock_gtod_data.
 : 
 : Since all other readers of pvclock_gtod_data remain guarded by CONFIG_X86_64,
 : is pvclock_gtod_data now effectively write-only on 32-bit? This would mean
 : update_pvclock_gtod() runs on every host core timekeeping update just to
 : populate an unused struct.

Huh.  Indeed.  Now that "Compute kvmclock base without pvclock_gtod_data" will
land before this patch, there's no need to register KVM's notifier on 32-bit,
and this patch is simply:

diff --git arch/x86/kvm/x86.c arch/x86/kvm/x86.c
index 67c762b3bf28..edafe13f74cf 100644
--- arch/x86/kvm/x86.c
+++ arch/x86/kvm/x86.c
@@ -926,19 +926,13 @@ static void update_pvclock_gtod(struct timekeeper *tk)
 
        write_seqcount_end(&vdata->seq);
 }
+#endif
 
 static s64 get_kvmclock_base_ns(void)
 {
        /* Count up from boot time, but with the frequency of the raw clock.  */
        return ktime_to_ns(ktime_mono_to_any(ktime_get_raw(), TK_OFFS_BOOT));
 }
-#else
-static s64 get_kvmclock_base_ns(void)
-{
-       /* Master clock not used, so we can just use CLOCK_BOOTTIME.  */
-       return ktime_get_boottime_ns();
-}
-#endif
 
 static uint32_t div_frac(uint32_t dividend, uint32_t divisor)
 {

I'll post a v10, since this is a non-trivial change.