Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1501399

Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry

Show all headers | View raw

On Fri, Oct 14, 2016 at 08:21:27PM +0200, Paolo Bonzini wrote:
> On some benchmarks (e.g. netperf with ioeventfd disabled), APICv
> posted interrupts turn out to be slower than interrupt injection via
> KVM_REQ_EVENT.
> 
> This patch optimizes a bit the IRR update, avoiding expensive atomic
> operations in the common case where PI.ON=0 at vmentry or the PIR vector
> is mostly zero.  This saves at least 20 cycles (1%) per vmexit, as
> measured by kvm-unit-tests' inl_from_qemu test (20 runs):
> 
>               | enable_apicv=1  |  enable_apicv=0
>               | mean     stdev  |  mean     stdev
>     ----------|-----------------|------------------
>     before    | 5826     32.65  |  5765     47.09
>     after     | 5809     43.42  |  5777     77.02
> 
> Of course, any change in the right column is just placebo effect. :)
> The savings are bigger if interrupts are frequent.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
>  arch/x86/kvm/lapic.c | 6 ++++--
>  arch/x86/kvm/vmx.c   | 9 ++++++++-
>  2 files changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index 23b99f305382..63a442aefc12 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -342,9 +342,11 @@ void __kvm_apic_update_irr(u32 *pir, void *regs)
>  	u32 i, pir_val;
>  
>  	for (i = 0; i <= 7; i++) {
> -		pir_val = xchg(&pir[i], 0);
> -		if (pir_val)
> +		pir_val = READ_ONCE(pir[i]);
> +		if (pir_val) {
> +			pir_val = xchg(&pir[i], 0);
>  			*((u32 *)(regs + APIC_IRR + i * 0x10)) |= pir_val;
> +		}
>  	}
>  }
>  EXPORT_SYMBOL_GPL(__kvm_apic_update_irr);

gcc doesn't seem to unroll this loop and it's
probably worth unrolling it

The following seems to do the trick for me on upstream - I didn't
benchmark it though. Is there a kvm unit test for interrupts?

--->

kvm: unroll the loop in __kvm_apic_update_irr.

This is hot data path in interrupt-rich workloads, worth unrolling.

Signed-off-by: Michael S. Tsirkin <mst@redhat.com>


diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
index b62c852..0c3462c 100644
--- a/arch/x86/kvm/lapic.c
+++ b/arch/x86/kvm/lapic.c
@@ -337,7 +337,8 @@ static u8 count_vectors(void *bitmap)
 	return count;
 }
 
-void __kvm_apic_update_irr(u32 *pir, void *regs)
+void __attribute__((optimize("unroll-loops")))
+__kvm_apic_update_irr(u32 *pir, void *regs)
 {
 	u32 i, pir_val;
 

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 0/5] KVM: x86: cleanup and minimal speedup for APICv Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 20:30 +0200
  [PATCH 3/5] KVM: x86: do not use KVM_REQ_EVENT for APICv interrupt injection Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 20:30 +0200
    Re: [PATCH 3/5] KVM: x86: do not use KVM_REQ_EVENT for APICv  interrupt injection Radim Krčmář <rkrcmar@redhat.com> - 2016-10-26 22:10 +0200
  [PATCH 2/5] KVM: x86: do not scan IRR twice on APICv vmentry Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 20:30 +0200
    Re: [PATCH 2/5] KVM: x86: do not scan IRR twice on APICv vmentry Wanpeng Li <kernellwp@gmail.com> - 2016-10-18 08:10 +0200
    Re: [PATCH 2/5] KVM: x86: do not scan IRR twice on APICv vmentry Radim Krčmář <rkrcmar@redhat.com> - 2016-10-26 22:00 +0200
  [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 20:30 +0200
    Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 21:00 +0200
      Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Nadav Amit <nadav.amit@gmail.com> - 2016-10-14 21:50 +0200
        Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Paolo Bonzini <pbonzini@redhat.com> - 2016-10-15 09:50 +0200
          Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry "Michael S. Tsirkin" <mst@redhat.com> - 2016-10-16 04:40 +0200
            Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-10-19 20:00 +0200
              Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry "Michael S. Tsirkin" <mst@redhat.com> - 2016-10-27 00:00 +0200
    Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Nadav Amit <nadav.amit@gmail.com> - 2016-10-14 21:00 +0200
    Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry "Michael S. Tsirkin" <mst@redhat.com> - 2016-10-16 05:30 +0200
      Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Paolo Bonzini <pbonzini@redhat.com> - 2016-10-17 13:10 +0200
    Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry Radim Krčmář <rkrcmar@redhat.com> - 2016-10-26 22:00 +0200
      Re: [PATCH 1/5] KVM: x86: avoid atomic operations on APICv vmentry "Michael S. Tsirkin" <mst@redhat.com> - 2016-10-26 23:50 +0200
  [PATCH 5/5] KVM: vmx: clear pending interrupts on KVM_SET_LAPIC Paolo Bonzini <pbonzini@redhat.com> - 2016-10-14 20:30 +0200
    Re: [PATCH 5/5] KVM: vmx: clear pending interrupts on KVM_SET_LAPIC Radim Krčmář <rkrcmar@redhat.com> - 2016-10-26 22:10 +0200
  Re: [PATCH 0/5] KVM: x86: cleanup and minimal speedup for APICv "Michael S. Tsirkin" <mst@redhat.com> - 2016-10-27 00:00 +0200

csiph-web