From: Marc Zyngier <maz@kernel.org>
To: He Ying <heying24@huawei.com>
Cc: <vincent.guittot@linaro.org>, <Valentin.Schneider@arm.com>,
<andrew@lunn.ch>, <catalin.marinas@arm.com>,
<f.fainelli@gmail.com>, <gregory.clement@bootlin.com>,
<jason@lakedaemon.net>, <kernel-team@android.com>,
<linux-arm-kernel@lists.infradead.org>,
<linux-kernel@vger.kernel.org>, <linux@arm.linux.org.uk>,
<saravanak@google.com>, <sumit.garg@linaro.org>,
<tglx@linutronix.de>, <will@kernel.org>
Subject: Re: [PATCH v3 03/16] arm64: Allow IPIs to be handled as normal interrupts
Date: Thu, 06 May 2021 12:44:21 +0100 [thread overview]
Message-ID: <87pmy4qe7e.wl-maz@kernel.org> (raw)
In-Reply-To: <c66367b0-e8a0-2b7b-13c3-c9413462357c@huawei.com>
On Thu, 06 May 2021 08:50:42 +0100,
He Ying <heying24@huawei.com> wrote:
>
> Hello Marc,
>
> We have faced a performance regression for handling ipis since this
> commit. I think it's the same issue reported by Vincent.
Can you share more details on what regression you have observed?
What's the workload, the system, the performance drop?
> I found you pointed out the possible two causes:
>
> (1) irq_enter/exit on the rescheduling IPI means we reschedule much
> more often.
It turned out to be a red herring. We don't reschedule more often, but
we instead suffer from the overhead of irq_enter()/irq_exit().
However, this only matters for silly benchmarks, and no real-life
workload showed any significant regression. Have you identified such
realistic workload?
> (2) irq_domain lookups add some overhead.
While this is also a potential source of overhead, it turned out not
to be the case.
> But I don't see any following patches in mainline. So, are you still
> working on this issue? Looking forward to your reply.
See [1]. However, there is probably better things to do than this
low-level specialisation of IPIs, and Thomas outlined what needs to be
done (see v1 of the patch series).
Thanks,
M.
[1] https://lore.kernel.org/lkml/20201124141449.572446-1-maz@kernel.org/
--
Without deviation from the norm, progress is not possible.
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
WARNING: multiple messages have this Message-ID (diff)
From: Marc Zyngier <maz@kernel.org>
To: He Ying <heying24@huawei.com>
Cc: <vincent.guittot@linaro.org>, <Valentin.Schneider@arm.com>,
<andrew@lunn.ch>, <catalin.marinas@arm.com>,
<f.fainelli@gmail.com>, <gregory.clement@bootlin.com>,
<jason@lakedaemon.net>, <kernel-team@android.com>,
<linux-arm-kernel@lists.infradead.org>,
<linux-kernel@vger.kernel.org>, <linux@arm.linux.org.uk>,
<saravanak@google.com>, <sumit.garg@linaro.org>,
<tglx@linutronix.de>, <will@kernel.org>
Subject: Re: [PATCH v3 03/16] arm64: Allow IPIs to be handled as normal interrupts
Date: Thu, 06 May 2021 12:44:21 +0100 [thread overview]
Message-ID: <87pmy4qe7e.wl-maz@kernel.org> (raw)
In-Reply-To: <c66367b0-e8a0-2b7b-13c3-c9413462357c@huawei.com>
On Thu, 06 May 2021 08:50:42 +0100,
He Ying <heying24@huawei.com> wrote:
>
> Hello Marc,
>
> We have faced a performance regression for handling ipis since this
> commit. I think it's the same issue reported by Vincent.
Can you share more details on what regression you have observed?
What's the workload, the system, the performance drop?
> I found you pointed out the possible two causes:
>
> (1) irq_enter/exit on the rescheduling IPI means we reschedule much
> more often.
It turned out to be a red herring. We don't reschedule more often, but
we instead suffer from the overhead of irq_enter()/irq_exit().
However, this only matters for silly benchmarks, and no real-life
workload showed any significant regression. Have you identified such
realistic workload?
> (2) irq_domain lookups add some overhead.
While this is also a potential source of overhead, it turned out not
to be the case.
> But I don't see any following patches in mainline. So, are you still
> working on this issue? Looking forward to your reply.
See [1]. However, there is probably better things to do than this
low-level specialisation of IPIs, and Thomas outlined what needs to be
done (see v1 of the patch series).
Thanks,
M.
[1] https://lore.kernel.org/lkml/20201124141449.572446-1-maz@kernel.org/
--
Without deviation from the norm, progress is not possible.
next prev parent reply other threads:[~2021-05-06 11:47 UTC|newest]
Thread overview: 170+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-01 14:43 [PATCH v3 00/16] arm/arm64: Turning IPIs into normal interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 01/16] genirq: Add fasteoi IPI flow Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 02/16] genirq: Allow interrupts to be excluded from /proc/interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 03/16] arm64: Allow IPIs to be handled as normal interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-11 15:05 ` Catalin Marinas
2020-09-11 15:05 ` Catalin Marinas
2020-10-19 12:42 ` Vincent Guittot
2020-10-19 12:42 ` Vincent Guittot
2020-10-19 13:04 ` Marc Zyngier
2020-10-19 13:04 ` Marc Zyngier
2020-10-19 15:43 ` Vincent Guittot
2020-10-19 15:43 ` Vincent Guittot
2020-10-19 16:00 ` Valentin Schneider
2020-10-19 16:00 ` Valentin Schneider
2020-10-27 10:12 ` Vincent Guittot
2020-10-27 10:12 ` Vincent Guittot
2020-10-27 10:37 ` Marc Zyngier
2020-10-27 10:37 ` Marc Zyngier
2020-10-27 10:50 ` Vincent Guittot
2020-10-27 10:50 ` Vincent Guittot
2020-10-27 11:21 ` Vincent Guittot
2020-10-27 11:21 ` Vincent Guittot
2020-10-27 12:06 ` Marc Zyngier
2020-10-27 12:06 ` Marc Zyngier
2020-10-27 13:17 ` Vincent Guittot
2020-10-27 13:17 ` Vincent Guittot
[not found] ` <c66367b0-e8a0-2b7b-13c3-c9413462357c@huawei.com>
2021-05-06 11:44 ` Marc Zyngier [this message]
2021-05-06 11:44 ` Marc Zyngier
2021-05-07 7:30 ` He Ying
2021-05-07 7:30 ` He Ying
2021-05-07 8:56 ` Marc Zyngier
2021-05-07 8:56 ` Marc Zyngier
2021-05-07 9:31 ` He Ying
2021-05-07 9:31 ` He Ying
2020-09-01 14:43 ` [PATCH v3 04/16] ARM: " Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 05/16] irqchip/gic-v3: Describe the SGI range Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 06/16] irqchip/gic-v3: Configure SGIs as standard interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 07/16] irqchip/gic: Refactor SMP configuration Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 08/16] irqchip/gic: Configure SGIs as standard interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-14 13:06 ` Marek Szyprowski
2020-09-14 13:06 ` Marek Szyprowski
2020-09-14 13:13 ` Marc Zyngier
2020-09-14 13:13 ` Marc Zyngier
2020-09-14 13:26 ` Marek Szyprowski
2020-09-14 13:26 ` Marek Szyprowski
2020-09-14 15:09 ` Marc Zyngier
2020-09-14 15:09 ` Marc Zyngier
2020-09-15 6:48 ` Marek Szyprowski
2020-09-15 6:48 ` Marek Szyprowski
2020-09-15 8:07 ` Marc Zyngier
2020-09-15 8:07 ` Marc Zyngier
2020-09-15 8:35 ` Marek Szyprowski
2020-09-15 8:35 ` Marek Szyprowski
2020-09-15 9:48 ` Marc Zyngier
2020-09-15 9:48 ` Marc Zyngier
2020-09-16 14:16 ` Jon Hunter
2020-09-16 14:16 ` Jon Hunter
2020-09-16 15:10 ` Marc Zyngier
2020-09-16 15:10 ` Marc Zyngier
2020-09-16 15:46 ` Jon Hunter
2020-09-16 15:46 ` Jon Hunter
2020-09-16 15:55 ` Marc Zyngier
2020-09-16 15:55 ` Marc Zyngier
2020-09-16 15:58 ` Jon Hunter
2020-09-16 15:58 ` Jon Hunter
2020-09-16 16:22 ` Marc Zyngier
2020-09-16 16:22 ` Marc Zyngier
2020-09-16 16:28 ` Marc Zyngier
2020-09-16 16:28 ` Marc Zyngier
2020-09-16 19:08 ` Jon Hunter
2020-09-16 19:08 ` Jon Hunter
2020-09-16 19:06 ` Jon Hunter
2020-09-16 19:06 ` Jon Hunter
2020-09-16 19:26 ` Mikko Perttunen
2020-09-16 19:26 ` Mikko Perttunen
2020-09-16 19:39 ` Jon Hunter
2020-09-16 19:39 ` Jon Hunter
2020-09-17 7:40 ` Linus Walleij
2020-09-17 7:40 ` Linus Walleij
2020-09-17 7:50 ` Marc Zyngier
2020-09-17 7:50 ` Marc Zyngier
2020-09-17 7:54 ` Jon Hunter
2020-09-17 7:54 ` Jon Hunter
2020-09-17 8:45 ` Marc Zyngier
2020-09-17 8:45 ` Marc Zyngier
2020-09-17 8:49 ` Jon Hunter
2020-09-17 8:49 ` Jon Hunter
2020-09-17 8:54 ` Marek Szyprowski
2020-09-17 8:54 ` Marek Szyprowski
2020-09-17 9:09 ` Jon Hunter
2020-09-17 9:09 ` Jon Hunter
2020-09-17 9:13 ` Marek Szyprowski
2020-09-17 9:13 ` Marek Szyprowski
2020-09-17 9:29 ` Marc Zyngier
2020-09-17 9:29 ` Marc Zyngier
2020-09-17 14:53 ` Jon Hunter
2020-09-17 14:53 ` Jon Hunter
2020-09-17 18:24 ` Jon Hunter
2020-09-17 18:24 ` Jon Hunter
2020-09-18 8:24 ` Marc Zyngier
2020-09-18 8:24 ` Marc Zyngier
2020-09-17 8:56 ` Marc Zyngier
2020-09-17 8:56 ` Marc Zyngier
2020-09-17 10:11 ` Linus Walleij
2020-09-17 10:11 ` Linus Walleij
2020-09-16 14:03 ` Linus Walleij
2020-09-16 14:03 ` Linus Walleij
2020-09-16 14:14 ` Marc Zyngier
2020-09-16 14:14 ` Marc Zyngier
2020-09-18 9:58 ` James Morse
2020-09-18 9:58 ` James Morse
2020-09-18 10:21 ` Marc Zyngier
2020-09-18 10:21 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 09/16] irqchip/gic-common: Don't enable SGIs by default Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 10/16] irqchip/bcm2836: Configure mailbox interrupts as standard interrupts Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-14 14:32 ` Marek Szyprowski
2020-09-14 14:32 ` Marek Szyprowski
2020-09-14 16:10 ` Marc Zyngier
2020-09-14 16:10 ` Marc Zyngier
2020-09-14 19:13 ` Marek Szyprowski
2020-09-14 19:13 ` Marek Szyprowski
2020-09-01 14:43 ` [PATCH v3 11/16] irqchip/hip04: Configure IPIs " Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 12/16] irqchip/armada-370-xp: " Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 13/16] arm64: Kill __smp_cross_call and co Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-11 15:06 ` Catalin Marinas
2020-09-11 15:06 ` Catalin Marinas
2020-09-01 14:43 ` [PATCH v3 14/16] arm64: Remove custom IRQ stat accounting Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-11 15:06 ` Catalin Marinas
2020-09-11 15:06 ` Catalin Marinas
2020-09-01 14:43 ` [PATCH v3 15/16] ARM: Kill __smp_cross_call and co Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-01 14:43 ` [PATCH v3 16/16] ARM: Remove custom IRQ stat accounting Marc Zyngier
2020-09-01 14:43 ` Marc Zyngier
2020-09-02 7:41 ` kernel test robot
2020-09-02 7:41 ` kernel test robot
2020-09-02 7:41 ` kernel test robot
2020-09-02 20:20 ` Marc Zyngier
2020-09-02 20:20 ` Marc Zyngier
2020-09-02 20:20 ` Marc Zyngier
2020-09-24 9:00 ` Guillaume Tucker
2020-09-24 9:00 ` Guillaume Tucker
2020-09-24 9:29 ` Marc Zyngier
2020-09-24 9:29 ` Marc Zyngier
2020-09-24 13:09 ` Guillaume Tucker
2020-09-24 13:09 ` Guillaume Tucker
2020-09-28 9:00 ` Guillaume Tucker
2020-09-28 9:00 ` Guillaume Tucker
2020-09-24 13:34 ` Fabio Estevam
2020-09-24 13:34 ` Fabio Estevam
2020-09-24 14:19 ` Guillaume Tucker
2020-09-24 14:19 ` Guillaume Tucker
2020-09-07 6:06 ` [PATCH v3 00/16] arm/arm64: Turning IPIs into normal interrupts hasegawa-hitomi
2020-09-07 6:06 ` hasegawa-hitomi
2020-09-16 16:54 ` Florian Fainelli
2020-09-16 16:54 ` Florian Fainelli
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87pmy4qe7e.wl-maz@kernel.org \
--to=maz@kernel.org \
--cc=Valentin.Schneider@arm.com \
--cc=andrew@lunn.ch \
--cc=catalin.marinas@arm.com \
--cc=f.fainelli@gmail.com \
--cc=gregory.clement@bootlin.com \
--cc=heying24@huawei.com \
--cc=jason@lakedaemon.net \
--cc=kernel-team@android.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux@arm.linux.org.uk \
--cc=saravanak@google.com \
--cc=sumit.garg@linaro.org \
--cc=tglx@linutronix.de \
--cc=vincent.guittot@linaro.org \
--cc=will@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.