vmx_vcpu_run and svm_vcpu_run are large functions, and this can actually make a substantial cycle difference by keeping the fast path contiguous in memory. Without it, the retpoline guest/retpoline host case is about 50 cycles slower.
Cc: [email protected] Cc: Radim Krčmář <[email protected]> Cc: KarimAllah Ahmed <[email protected]> Cc: David Woodhouse <[email protected]> Cc: Jim Mattson <[email protected]> Cc: Thomas Gleixner <[email protected]> Cc: Ingo Molnar <[email protected]> Cc: [email protected] Reviewed-by: Jim Mattson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]> --- arch/x86/kvm/svm.c | 2 +- arch/x86/kvm/vmx.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 1598beeda11c..24c9521ebc24 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -5465,7 +5465,7 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu) * If the L02 MSR bitmap does not intercept the MSR, then we need to * save it. */ - if (!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL)) + if (unlikely(!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL))) svm->spec_ctrl = native_read_msr(MSR_IA32_SPEC_CTRL); if (svm->spec_ctrl) diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index af89d377681d..e13fd2a833c4 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -9589,7 +9589,7 @@ static void __noclone vmx_vcpu_run(struct kvm_vcpu *vcpu) * If the L02 MSR bitmap does not intercept the MSR, then we need to * save it. */ - if (!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL)) + if (unlikely(!msr_write_intercepted(vcpu, MSR_IA32_SPEC_CTRL))) vmx->spec_ctrl = native_read_msr(MSR_IA32_SPEC_CTRL); if (vmx->spec_ctrl) -- 1.8.3.1
