linux-kernel - Re: clockevent driver as source for pretimout watchdog interrupt

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>] [day] [month] [year] [list]

Date:	Thu, 21 Apr 2016 10:45:38 +0200 (CEST)
From:	Thomas Gleixner <tglx@...utronix.de>
To:	"fixed-term.Oleksij.Rempel" <fixed-term.Oleksij.Rempel@...bosch.com>
cc:	Daniel Lezcano <daniel.lezcano@...aro.org>,
	linux-kernel@...r.kernel.org,
	"Dirk.Behme" <Dirk.Behme@...bosch.com>
Subject: Re: clockevent driver as source for pretimout watchdog interrupt

On Thu, 21 Apr 2016, fixed-term.Oleksij.Rempel wrote:
> i'm doing research on proper implementation for watchdog pretimout
> interrupt for devices which do not provide this functionality.
> 
> Some watchdog controllers like Freescale iMX6 can generate an interrupt
> for some configured time before watchdog will reset the system.
> Suddenly most watchdog controllers don't have this functionality.
> 
> My idea is to reuse some count down (one shot?) timer:
> - configure it to generate an interrupt for a time less then watchdog
> deadline.
> - reset it each time we get a ping, before an interrupt is generated.
> - interrupt should executed one of governors from pretimout patchset,
> see the link:
> https://www.mail-archive.com/linux-watchdog@vger.kernel.org/msg02394.html
> 
> I assume clocksource framework is kind of suitable for this purpose? Is
> it correct assumption?
> Any other ideas? Suggestions?

What's wrong with simply using a hrtimer for this? The extra clockevent is not
buying you anything AFAICT.

Thanks,

	tglx