KVM: x86: Remove WARN sanity check on hypervisor timer vs. UNINITIALIZED vCPU
Sean Christopherson
maintainerPATCH 1/1 KVM: x86: Remove WARN sanity check on hypervisor timer vs. UNINITIALIZED vCPU
Drop the WARN in KVM_RUN that asserts that KVM isn't using the hypervisor timer, a.k.a. the VMX preemption timer, for a vCPU that is in the UNINITIALIZIED activity state. The intent of the WARN is to sanity check that KVM won't drop a timer interrupt due to an unexpected transition to UNINITIALIZED, but unfortunately userspace can use various ioctl()s to force the unexpected state.
Drop the sanity check instead of switching from the hypervisor timer to a software based timer, as the only reason to switch to a software timer when a vCPU is blocking is to ensure the timer interrupt wakes the vCPU, but said interrupt isn't a valid wake event for vCPUs in UNINITIALIZED state *and* the interrupt will be dropped in the end.
--- arch/x86/kvm/x86.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)
arch/x86/kvm/x86.c +9 -4
@@ -11091,12 +11091,17 @@ int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu) r = -EINTR ; goto out; } +/* -* It should be impossible for the hypervisor timer to be in -* use before KVM has ever run the vCPU. +* Don't bother switching APIC timer emulation from the +* hypervisor timer to the software timer, the only way for the +* APIC timer to be active is if userspace stuffed vCPU state, +* i.e. put the vCPU into a nonsensical state. Only an INIT +* will transition the vCPU out of UNINITIALIZED (without more +* state stuffing from userspace), which will reset the local +* APIC and thus smother the timer anyways, i.e. the APIC timer +* IRQ(s) will be dropped no matter what. */ -WARN_ON_ONCE (kvm_lapic_hv_timer_in_use (vcpu)); -kvm_vcpu_srcu_read_unlock (vcpu);kvm_vcpu_block (vcpu);kvm_vcpu_srcu_read_lock (vcpu);
base-commit: fdf0eaf11452d72945af31804e2a1048ee1b574c
Signature · 1 line
2.41.0.640.ga95def55d0-goog
arch/x86/kvm/x86.c int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu) · line 11102
Quoted 10 lines · Sean Christopherson
/*
arch/x86/kvm/x86.c +7
-2-* It should be impossible for the hypervisor timer to be in -* use before KVM has ever run the vCPU. +* Don't bother switching APIC timer emulation from the +* hypervisor timer to the software timer, the only way for the +* APIC timer to be active is if userspace stuffed vCPU state, +* i.e. put the vCPU into a nonsensical state. Only an INIT +* will transition the vCPU out of UNINITIALIZED (without more +* state stuffing from userspace), which will reset the local +* APIC and thus smother the timer anyways, i.e. the APIC timer
"Cancel" is probably more understandable to non-native speakers, though undoubtedly less poetic.
Paolo
Quoted 4 lines
arch/x86/kvm/x86.c +1
-2+* IRQ(s) will be dropped no matter what. */ -WARN_ON_ONCE (kvm_lapic_hv_timer_in_use (vcpu)); -
Sean Christopherson
maintainerQuoted 14 lines · Paolo Bonzini
arch/x86/kvm/x86.c int kvm_arch_vcpu_ioctl_run(struct kvm_vcpu *vcpu) · line 11102
Quoted 10 lines · Sean Christopherson
/*
arch/x86/kvm/x86.c +7
-2-* It should be impossible for the hypervisor timer to be in -* use before KVM has ever run the vCPU. +* Don't bother switching APIC timer emulation from the +* hypervisor timer to the software timer, the only way for the +* APIC timer to be active is if userspace stuffed vCPU state, +* i.e. put the vCPU into a nonsensical state. Only an INIT +* will transition the vCPU out of UNINITIALIZED (without more +* state stuffing from userspace), which will reset the local +* APIC and thus smother the timer anyways, i.e. the APIC timer "Cancel" is probably more understandable to non-native speakers, though undoubtedly less poetic.
I intentionally avoided "cancel" because there is no guaranteed the timer would actually be canceled (if KVM switched to the software timer). I am/was trying to call out that even if the timer expires and pends an interrupts, the interrupt will ultimately be dropped.
Maybe "squashed"?
Quoted 8 lines · Sean Christopherson
Drop the WARN in
KVM_RUNthat asserts that KVM isn't using the hypervisor timer, a.k.a. the VMX preemption timer, for a vCPU that is in the UNINITIALIZIED activity state. The intent of the WARN is to sanity check that KVM won't drop a timer interrupt due to an unexpected transition to UNINITIALIZED, but unfortunately userspace can use various ioctl()s to force the unexpected state.[...]
Applied to kvm-x86 misc, with a tweaked comment to avoid "smother".
[1/1] KVM: x86: Remove WARN sanity check on hypervisor timer vs. UNINITIALIZED vCPU
https://github.com/kvm-x86/linux/commit/7b0151caf73a
--
https://github.com/kvm-x86/linux/tree/next
https://github.com/kvm-x86/linux/tree/fixes