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;
MIME-Version: 1.0
References: <1559178307-6835-1-git-send-email-wanpengli@tencent.com>
 <20190610143420.GA6594@flask> <20190611011100.GB24835@linux.intel.com>
 <CANRm+Cwv5jqxBW=Ss5nkX7kZM3_Y-Ucs66yx5+wN09=W4pUdzA@mail.gmail.com> <F136E492-5350-49EE-A856-FBAEDB12FF99@gmail.com>
In-Reply-To: <F136E492-5350-49EE-A856-FBAEDB12FF99@gmail.com>
From:   Wanpeng Li <kernellwp@gmail.com>
Date:   Tue, 11 Jun 2019 18:02:25 +0800
Message-ID: <CANRm+CyZcvuT80ixp9f0FNmjN+rTUtw8MshtBG0Uk4L1B1UjDw@mail.gmail.com>
Subject: Re: [PATCH v3 0/3] KVM: Yield to IPI target if necessary
To:     Nadav Amit <nadav.amit@gmail.com>
Cc:     Sean Christopherson <sean.j.christopherson@intel.com>,
        =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= <rkrcmar@redhat.com>,
        LKML <linux-kernel@vger.kernel.org>, kvm <kvm@vger.kernel.org>,
        Paolo Bonzini <pbonzini@redhat.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, 11 Jun 2019 at 09:48, Nadav Amit <nadav.amit@gmail.com> wrote:
>
> > On Jun 10, 2019, at 6:45 PM, Wanpeng Li <kernellwp@gmail.com> wrote:
> >
> > On Tue, 11 Jun 2019 at 09:11, Sean Christopherson
> > <sean.j.christopherson@intel.com> wrote:
> >> On Mon, Jun 10, 2019 at 04:34:20PM +0200, Radim Kr=C4=8Dm=C3=A1=C5=99 =
wrote:
> >>> 2019-05-30 09:05+0800, Wanpeng Li:
> >>>> The idea is from Xen, when sending a call-function IPI-many to vCPUs=
,
> >>>> yield if any of the IPI target vCPUs was preempted. 17% performance
> >>>> increasement of ebizzy benchmark can be observed in an over-subscrib=
e
> >>>> environment. (w/ kvm-pv-tlb disabled, testing TLB flush call-functio=
n
> >>>> IPI-many since call-function is not easy to be trigged by userspace
> >>>> workload).
> >>>
> >>> Have you checked if we could gain performance by having the yield as =
an
> >>> extension to our PV IPI call?
> >>>
> >>> It would allow us to skip the VM entry/exit overhead on the caller.
> >>> (The benefit of that might be negligible and it also poses a
> >>> complication when splitting the target mask into several PV IPI
> >>> hypercalls.)
> >>
> >> Tangetially related to splitting PV IPI hypercalls, are there any majo=
r
> >> hurdles to supporting shorthand?  Not having to generate the mask for
> >> ->send_IPI_allbutself and ->kvm_send_ipi_all seems like an easy to way
> >> shave cycles for affected flows.
> >
> > Not sure why shorthand is not used for native x2apic mode.
>
> Why do you say so? native_send_call_func_ipi() checks if allbutself
> shorthand should be used and does so (even though the check can be more
> efficient - I=E2=80=99m looking at that code right now=E2=80=A6)

Please continue to follow the apic/x2apic driver. Just apic_flat set
APIC_DEST_ALLBUT/APIC_DEST_ALLINC to ICR.

Regards,
Wanpeng Li