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 RFC v3 1/6] x86/paravirt: Add pv_idle_ops to paravirt ops
To:     Wanpeng Li <kernellwp@gmail.com>
Cc:     Juergen Gross <jgross@suse.com>, Quan Xu <quan.xu03@gmail.com>,
        kvm <kvm@vger.kernel.org>, linux-doc@vger.kernel.org,
        "open list:FILESYSTEMS (VFS and infrastructure)" 
        <linux-fsdevel@vger.kernel.org>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        virtualization@lists.linux-foundation.org,
        the arch/x86 maintainers <x86@kernel.org>,
        xen-devel <xen-devel@lists.xenproject.org>,
        Yang Zhang <yang.zhang.wz@gmail.com>,
        Alok Kataria <akataria@vmware.com>,
        Rusty Russell <rusty@rustcorp.com.au>,
        Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>,
        "H. Peter Anvin" <hpa@zytor.com>
References: <1510567565-5118-1-git-send-email-quan.xu0@gmail.com>
 <1510567565-5118-2-git-send-email-quan.xu0@gmail.com>
 <07fac696-e3d4-8f35-8f3d-764d7ab41204@suse.com>
 <902da704-1e4f-583b-91c3-1a62ccd6e73d@gmail.com>
 <CANRm+CxirO0aSAQnOfPG0+KPCOnXf3SEASwBiprLmtncZE53pA@mail.gmail.com>
From:   Quan Xu <quan.xu0@gmail.com>
Message-ID: <d233b179-9815-0134-ef04-c492f9523121@gmail.com>
Date:   Tue, 14 Nov 2017 16:15:39 +0800
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101
 Thunderbird/52.4.0
MIME-Version: 1.0
In-Reply-To: <CANRm+CxirO0aSAQnOfPG0+KPCOnXf3SEASwBiprLmtncZE53pA@mail.gmail.com>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Content-Language: en-US
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


On 2017/11/14 15:12, Wanpeng Li wrote:
> 2017-11-14 15:02 GMT+08:00 Quan Xu <quan.xu0@gmail.com>:
>>
>> On 2017/11/13 18:53, Juergen Gross wrote:
>>> On 13/11/17 11:06, Quan Xu wrote:
>>>> From: Quan Xu <quan.xu0@gmail.com>
>>>>
>>>> So far, pv_idle_ops.poll is the only ops for pv_idle. .poll is called
>>>> in idle path which will poll for a while before we enter the real idle
>>>> state.
>>>>
>>>> In virtualization, idle path includes several heavy operations
>>>> includes timer access(LAPIC timer or TSC deadline timer) which will
>>>> hurt performance especially for latency intensive workload like message
>>>> passing task. The cost is mainly from the vmexit which is a hardware
>>>> context switch between virtual machine and hypervisor. Our solution is
>>>> to poll for a while and do not enter real idle path if we can get the
>>>> schedule event during polling.
>>>>
>>>> Poll may cause the CPU waste so we adopt a smart polling mechanism to
>>>> reduce the useless poll.
>>>>
>>>> Signed-off-by: Yang Zhang <yang.zhang.wz@gmail.com>
>>>> Signed-off-by: Quan Xu <quan.xu0@gmail.com>
>>>> Cc: Juergen Gross <jgross@suse.com>
>>>> Cc: Alok Kataria <akataria@vmware.com>
>>>> Cc: Rusty Russell <rusty@rustcorp.com.au>
>>>> Cc: Thomas Gleixner <tglx@linutronix.de>
>>>> Cc: Ingo Molnar <mingo@redhat.com>
>>>> Cc: "H. Peter Anvin" <hpa@zytor.com>
>>>> Cc: x86@kernel.org
>>>> Cc: virtualization@lists.linux-foundation.org
>>>> Cc: linux-kernel@vger.kernel.org
>>>> Cc: xen-devel@lists.xenproject.org
>>> Hmm, is the idle entry path really so critical to performance that a new
>>> pvops function is necessary?
>> Juergen, Here is the data we get when running benchmark netperf:
>>   1. w/o patch and disable kvm dynamic poll (halt_poll_ns=0):
>>      29031.6 bit/s -- 76.1 %CPU
>>
>>   2. w/ patch and disable kvm dynamic poll (halt_poll_ns=0):
>>      35787.7 bit/s -- 129.4 %CPU
>>
>>   3. w/ kvm dynamic poll:
>>      35735.6 bit/s -- 200.0 %CPU
> Actually we can reduce the CPU utilization by sleeping a period of
> time as what has already been done in the poll logic of IO subsystem,
> then we can improve the algorithm in kvm instead of introduing another
> duplicate one in the kvm guest.
We really appreciate upstream's kvm dynamic poll mechanism, which is
really helpful for a lot of scenario..

However, as description said, in virtualization, idle path includes
several heavy operations includes timer access (LAPIC timer or TSC
deadline timer) which will hurt performance especially for latency
intensive workload like message passing task. The cost is mainly from
the vmexit which is a hardware context switch between virtual machine
and hypervisor.

for upstream's kvm dynamic poll mechanism, even you could provide a
better algorism, how could you bypass timer access (LAPIC timer or TSC
deadline timer), or a hardware context switch between virtual machine
and hypervisor. I know these is a tradeoff.

Furthermore, here is the data we get when running benchmark contextswitch
to measure the latency(lower is better):

1. w/o patch and disable kvm dynamic poll (halt_poll_ns=0):
   3402.9 ns/ctxsw -- 199.8 %CPU

2. w/ patch and disable kvm dynamic poll:
   1163.5 ns/ctxsw -- 205.5 %CPU

3. w/ kvm dynamic poll:
   2280.6 ns/ctxsw -- 199.5 %CPU

so, these tow solution are quite similar, but not duplicate..

that's also why to add a generic idle poll before enter real idle path.
When a reschedule event is pending, we can bypass the real idle path.


Quan
Alibaba Cloud


> Regards,
> Wanpeng Li
>
>>   4. w/patch and w/ kvm dynamic poll:
>>      42225.3 bit/s -- 198.7 %CPU
>>
>>   5. idle=poll
>>      37081.7 bit/s -- 998.1 %CPU
>>
>>
>>
>>   w/ this patch, we will improve performance by 23%.. even we could improve
>>   performance by 45.4%, if we use w/patch and w/ kvm dynamic poll. also the
>>   cost of CPU is much lower than 'idle=poll' case..
>>
>>> Wouldn't a function pointer, maybe guarded
>>> by a static key, be enough? A further advantage would be that this would
>>> work on other architectures, too.
>>
>> I assume this feature will be ported to other archs.. a new pvops makes code
>> clean and easy to maintain. also I tried to add it into existed pvops, but
>> it
>> doesn't match.
>>
>>
>>
>> Quan
>> Alibaba Cloud
>>>
>>> Juergen
>>>


From 1584027504485254943@xxx Tue Nov 14 07:59:07 +0000 2017
X-GM-THRID: 1583947963447418004
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread