linux-kernel - Re: udelay function delays the wrong time interval in multiprocessor system, if ARCH_HAS_READ_CURRENT_TIMER is not defined and on current timer is used.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20130318152514.GG4977@n2100.arm.linux.org.uk>
Date:	Mon, 18 Mar 2013 15:25:14 +0000
From:	Russell King - ARM Linux <linux@....linux.org.uk>
To:	chpoph <chpoph@...il.com>
Cc:	Will Deacon <will.deacon@....com>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"stable@...r.kernel.org" <stable@...r.kernel.org>,
	Viresh Kumar <viresh.kumar@...aro.org>,
	Nicolas Pitre <nico@...aro.org>,
	Liviu Dudau <Liviu.Dudau@....com>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: Re: udelay function delays the wrong time interval in
	multiprocessor system, if ARCH_HAS_READ_CURRENT_TIMER is not
	defined and on current timer is used.

On Mon, Mar 18, 2013 at 10:40:56PM +0800, chpoph wrote:
>  On Sun, Mar 17, 2013 at 08:05:43PM +0000, Will Deacon wrote:
> >Do you actually have an ARM platform that can scale the CPU frequencies
> independently?
> 
> Yes,  my smart phone use Qualcomm's 8x25 and 8064 platform, which can
> scale the CPU frequencies independently. I test the delay loop, the
> phone can't get accurate delayed time intervals.

That is expected.  udelay() is only approximate.  Sometimes, it will give
you a slightly shorter delay than asked for, or if preempted, it can give
you a much longer delay - because it has no idea how long it has been
preempted for.

This is why in a SMP system it is much better to use a timer-based udelay()
implementation which is more independent of these effects.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/