Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <1598578508-14134-1-git-send-email-wanpengli@tencent.com> <87a6ybx9pv.fsf@vitty.brq.redhat.com>
In-Reply-To: <87a6ybx9pv.fsf@vitty.brq.redhat.com>
From:   Wanpeng Li <kernellwp@gmail.com>
Date:   Tue, 1 Sep 2020 11:25:01 +0800
Message-ID: <CANRm+CwHiZjh3w94Xdd=ZQXP6XWysz87OG+LFR2ekQn5A2P7Dw@mail.gmail.com>
Subject: Re: [PATCH] KVM: LAPIC: Reset timer_advance_ns if timer mode switch
To:     Vitaly Kuznetsov <vkuznets@redhat.com>
Cc:     LKML <linux-kernel@vger.kernel.org>, kvm <kvm@vger.kernel.org>,
        Paolo Bonzini <pbonzini@redhat.com>,
        Sean Christopherson <sean.j.christopherson@intel.com>,
        Wanpeng Li <wanpengli@tencent.com>,
        Jim Mattson <jmattson@google.com>,
        Joerg Roedel <joro@8bytes.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, 31 Aug 2020 at 20:48, Vitaly Kuznetsov <vkuznets@redhat.com> wrote:
>
> Wanpeng Li <kernellwp@gmail.com> writes:
>
> > From: Wanpeng Li <wanpengli@tencent.com>
> >
> > per-vCPU timer_advance_ns should be set to 0 if timer mode is not tscdeadline
> > otherwise we waste cpu cycles in the function lapic_timer_int_injected(),
>
> lapic_timer_int_injected is just a test, kvm_wait_lapic_expire()
> (__kvm_wait_lapic_expire()) maybe?

Both the check in lapic_timer_int_injected(), the check in
__kvm_wait_lapic_expire(), and these function calls, we can observe
~1.3% world switch time reduce w/ this patch by
kvm-unit-tests/vmexit.flat vmcall testing on AMD server. In addition,
I think we should set apic->lapic_timer.expired_tscdeadline to 0 when
switching between tscdeadline mode and other modes on Intel in order
that we will not waste cpu cycles to tune advance value in
adjust_lapic_timer_advance() for one time.

Wanpeng