Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [PATCH 2/2] KVM: LAPIC: micro-optimize fixed mode ipi delivery
To:     Wanpeng Li <kernellwp@gmail.com>, linux-kernel@vger.kernel.org,
        kvm@vger.kernel.org
Cc:     =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= <rkrcmar@redhat.com>,
        Sean Christopherson <sean.j.christopherson@intel.com>,
        Vitaly Kuznetsov <vkuznets@redhat.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>
References: <1573283135-5502-1-git-send-email-wanpengli@tencent.com>
 <1573283135-5502-2-git-send-email-wanpengli@tencent.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Openpgp: preference=signencrypt
Message-ID: <7a526814-c44e-c188-fba4-c6fb97b88b71@redhat.com>
Date:   Mon, 11 Nov 2019 22:59:08 +0100
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101
 Thunderbird/60.8.0
MIME-Version: 1.0
In-Reply-To: <1573283135-5502-2-git-send-email-wanpengli@tencent.com>
Content-Language: en-US
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 09/11/19 08:05, Wanpeng Li wrote:
> From: Wanpeng Li <wanpengli@tencent.com>
>=20
> After disabling mwait/halt/pause vmexits, RESCHEDULE_VECTOR and
> CALL_FUNCTION_SINGLE_VECTOR etc IPI is one of the main remaining
> cause of vmexits observed in product environment which can't be
> optimized by PV IPIs. This patch is the follow-up on commit
> 0e6d242eccdb (KVM: LAPIC: Micro optimize IPI latency), to optimize
> redundancy logic before fixed mode ipi is delivered in the fast
> path.
>=20
> - broadcast handling needs to go slow path, so the delivery mode repair
>   can be delayed to before slow path.

I agree with this part, but is the cost of the irq->shorthand check
really measurable?

Paolo

> - self-IPI will not be intervened by hypervisor any more after APICv is
>   introduced and the boxes support APICv are popular now. In addition,
>   kvm_apic_map_get_dest_lapic() can handle the self-IPI, so there is no
>   need a shortcut for the non-APICv case.
>=20
> Signed-off-by: Wanpeng Li <wanpengli@tencent.com>
> ---
>  arch/x86/kvm/irq_comm.c | 6 +++---
>  arch/x86/kvm/lapic.c    | 5 -----
>  2 files changed, 3 insertions(+), 8 deletions(-)
>=20
> diff --git a/arch/x86/kvm/irq_comm.c b/arch/x86/kvm/irq_comm.c
> index 8ecd48d..aa88156 100644
> --- a/arch/x86/kvm/irq_comm.c
> +++ b/arch/x86/kvm/irq_comm.c
> @@ -52,15 +52,15 @@ int kvm_irq_delivery_to_apic(struct kvm *kvm, struct =
kvm_lapic *src,
>  =09unsigned long dest_vcpu_bitmap[BITS_TO_LONGS(KVM_MAX_VCPUS)];
>  =09unsigned int dest_vcpus =3D 0;
> =20
> +=09if (kvm_irq_delivery_to_apic_fast(kvm, src, irq, &r, dest_map))
> +=09=09return r;
> +
>  =09if (irq->dest_mode =3D=3D 0 && irq->dest_id =3D=3D 0xff &&
>  =09=09=09kvm_lowest_prio_delivery(irq)) {
>  =09=09printk(KERN_INFO "kvm: apic: phys broadcast and lowest prio\n");
>  =09=09irq->delivery_mode =3D APIC_DM_FIXED;
>  =09}
> =20
> -=09if (kvm_irq_delivery_to_apic_fast(kvm, src, irq, &r, dest_map))
> -=09=09return r;
> -
>  =09memset(dest_vcpu_bitmap, 0, sizeof(dest_vcpu_bitmap));
> =20
>  =09kvm_for_each_vcpu(i, vcpu, kvm) {
> diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c
> index b29d00b..ea936fa 100644
> --- a/arch/x86/kvm/lapic.c
> +++ b/arch/x86/kvm/lapic.c
> @@ -951,11 +951,6 @@ bool kvm_irq_delivery_to_apic_fast(struct kvm *kvm, =
struct kvm_lapic *src,
> =20
>  =09*r =3D -1;
> =20
> -=09if (irq->shorthand =3D=3D APIC_DEST_SELF) {
> -=09=09*r =3D kvm_apic_set_irq(src->vcpu, irq, dest_map);
> -=09=09return true;
> -=09}
> -
>  =09rcu_read_lock();
>  =09map =3D rcu_dereference(kvm->arch.apic_map);
> =20
>=20