Re: [PATCH] clocksource/drivers/arm_global_timer: Always use {readl|writel}_relaxed

Path	csiph.com!news.mixmin.net!eternal-september.org!feeder.eternal-september.org!aioe.org!bofh.it!news.nic.it!robomod
From	Arnd Bergmann <arnd@arndb.de>
Newsgroups	linux.kernel
Subject	Re: [PATCH] clocksource/drivers/arm_global_timer: Always use {readl\|writel}_relaxed
Date	Fri, 13 Nov 2015 13:40:02 +0100
Message-ID	<qulO2-6pf-9@gated-at.bofh.it> (permalink)
References	<qui3M-41Z-5@gated-at.bofh.it> <qujVU-5dq-19@gated-at.bofh.it> <qulEm-6kY-17@gated-at.bofh.it>
X-Original-To	linux-arm-kernel@lists.infradead.org
User-Agent	KMail/4.11.5 (Linux/3.16.0-10-generic; KDE/4.11.5; x86_64; ; )
MIME-Version	1.0
Content-Transfer-Encoding	7Bit
Content-Type	text/plain; charset="us-ascii"
X-Provags-ID	V03:K0:uPL6DC5kyuEGDrT2hmWCmLSTB+in/FTdPt+SkKNZxet7HMJIVjc gOIoYyiwo1wUe6wxPmPy9YYa3MPIcLKclvKvxeaqr62oZcZ19L1tM+Uya39td6kOQx+f3Dr pinCzW7orT4U+etyKOyspr/kxmVvUJ/7eirn5pbKKqn+79Ck7QYA+a2HQwI3cj4YfKczLiu yu5nn/znWccayW6delnNw==
X-Ui-Out-Filterresults	notjunk:1;V01:K0:rSBokLK/DKk=:auBNnaJZGUixINASvnjeEL qkEjzO/Au7V72b6VFgXKwED2Y8xoSWsSGZ8/GhYBgzGaOR1D0LSAveSKtYYvX+xjOTbYjF8Bb CNs55IgSm+aIjD4rqiXQvtdwA2GPp5wOCwBV0Uywv7sz58r3fjlkQyXQmSkr5GvHWTXd8gzdq Em8pkcV96e+0qOf+KLhMqVcJhSnC1dwYFTCGuGZVkV3lql4x4M91U1STohBeVHJ7GZWvgKADU l+4csQ84L5aKZV4N/FqjnwxOu0Jp77t4sA9sxlc0eDtBNoo/Nb13How1w5foaMmvh87qm8+IK MnLkogGLHJMj2OkD6kqaJUbgUq5B4nYcXxdaL6BveseYO2XAf5ZAYfxR5ap+iys4KsD18fHCO BkR0SS98D1qSLuRPEbfh6XlYclqcwzAFFSiIrEOim8TClw1FZwLuv2FbcA8a9r8WT9CAevgeX 9XRIKOb1f2nbswUcGlcofjhq+geityXhcqGqzDOVUcYqGD/3VQwGeC85MSzSSRaC3RVrsTpZf IU06FaBtdPeTT11DspQaL+2Zpfo4/+I1tfuEqd5gYQD0tG3Mh3Po819eVJ8YWrTE8tTlwWy56 o0+m6rFhGHJg3eVPgM2RQzh1Z3X3WZrIErn6lqZNzk0+H9+QVRbQem+hfWgnsYKTb9Hf0q8xQ /ZEpht72WW517sJqfOlvnM1gMbscNfcdMzeiPaw6xhhbOBtY8XKg7e3xrrEcYk25P6stTkaz1 L1ND/RKLvOmEmkKw
Sender	robomod@news.nic.it
List-ID	<linux-kernel.vger.kernel.org>
X-Mailing-List	linux-kernel@vger.kernel.org
Approved	robomod@news.nic.it
Lines	34
Organization	linux.* mail to news gateway
X-Original-Cc	Jisheng Zhang <jszhang@marvell.com>, kernel@stlinux.com, srinivas.kandagatla@gmail.com, daniel.lezcano@linaro.org, linux-kernel@vger.kernel.org, patrice.chotard@st.com, tglx@linutronix.de, maxime.coquelin@st.com
X-Original-Date	Fri, 13 Nov 2015 13:37:24 +0100
X-Original-Message-ID	<53950580.slDxBxJnNC@wuerfel>
X-Original-References	<1447403678-7217-1-git-send-email-jszhang@marvell.com> <25602407.8sIohphlWH@wuerfel> <20151113202001.5933ae54@xhacker>
X-Original-Sender	linux-kernel-owner@vger.kernel.org
Xref	csiph.com linux.kernel:1268911

Show key headers only | View raw

On Friday 13 November 2015 20:20:01 Jisheng Zhang wrote:
> 
> > for outer_cache.sync(). The Aurora outer cache sync has a different method
> > and also doesn't use l2x0_lock. Finally, tauros3 doesn't need a cache sync
> > at all.
> > 
> > Did you look at an older kernel version? We used to do a loop in the
> 
> oops, yes. The kernel version in product still needs the spinlock in sync.
> I didn't check the L2 cache code for about 1 year, sorry for that.
> If we upgrade to newer kernel version, yes, the bit performance bottleneck -- 
> spinlock contention won't exist anymore. Thanks for pointing out this.

If you still see lock contention on the l2x0 lock with your patch applied,
you might want to backport the optimizations to your product kernel, even
more so for the aurora controller in the Armada 370 that had some extra
optimizations.

> But I think we may still see trivial system performance improvement in 500-1000
> times/s of clockevent programming case due to the mb() in writel.

Yes, I think it's fine. Just try to put your best estimate of the
overhead in the patch description when you do the new version.

Unfortunately, it is not easy to measure what the actual overhead is
because low-level benchmarks of outer_cache.sync will show a much lower
overhead than doing it occasionally with an active cache.

	Arnd
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

[PATCH] clocksource/drivers/arm_global_timer: Always use {readl|writel}_relaxed Jisheng Zhang <jszhang@marvell.com> - 2015-11-13 09:40 +0100
  Re: [PATCH] clocksource/drivers/arm_global_timer: Always use  {readl|writel}_relaxed Jisheng Zhang <jszhang@marvell.com> - 2015-11-13 09:50 +0100
    Re: [PATCH] clocksource/drivers/arm_global_timer: Always use {readl|writel}_relaxed Arnd Bergmann <arnd@arndb.de> - 2015-11-13 10:30 +0100
      Re: [PATCH] clocksource/drivers/arm_global_timer: Always use  {readl|writel}_relaxed Jisheng Zhang <jszhang@marvell.com> - 2015-11-13 11:10 +0100
        Re: [PATCH] clocksource/drivers/arm_global_timer: Always use {readl|writel}_relaxed Arnd Bergmann <arnd@arndb.de> - 2015-11-13 11:40 +0100
          Re: [PATCH] clocksource/drivers/arm_global_timer: Always use  {readl|writel}_relaxed Jisheng Zhang <jszhang@marvell.com> - 2015-11-13 13:30 +0100
            Re: [PATCH] clocksource/drivers/arm_global_timer: Always use {readl|writel}_relaxed Arnd Bergmann <arnd@arndb.de> - 2015-11-13 13:40 +0100

csiph-web