linux-kernel - Re: [RFC 02/18] power/reset: vexpress: Use sched

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

Message-ID: <1389196863.23721.20.camel@hornet>
Date:	Wed, 08 Jan 2014 16:01:03 +0000
From:	Pawel Moll <pawel.moll@....com>
To:	John Stultz <john.stultz@...aro.org>
Cc:	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
	linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
	Samuel Ortiz <sameo@...ux.intel.com>,
	Arnd Bergmann <arnd@...db.de>, Jon Medhurst <tixy@...aro.org>,
	"arm@...nel.org" <arm@...nel.org>, Olof Johansson <olof@...om.net>,
	Anton Vorontsov <anton@...msg.org>,
	David Woodhouse <dwmw2@...radead.org>
Subject: Re: [RFC 02/18] power/reset: vexpress: Use sched_clock as the time
 source

On Mon, 2013-12-23 at 19:28 +0000, John Stultz wrote:
> On Mon, Dec 23, 2013 at 8:23 AM, Pawel Moll <pawel.moll@....com> wrote:
> > At this stage of system shutdown procedure the jiffies may
> > not be updated anymore, so have to base on raw sched_clock
> > values.
> >
> > Signed-off-by: Pawel Moll <pawel.moll@....com>
> > Cc: Anton Vorontsov <anton@...msg.org>
> > Cc: David Woodhouse <dwmw2@...radead.org>
> > ---
> >  drivers/power/reset/vexpress-poweroff.c | 7 ++++---
> >  1 file changed, 4 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/power/reset/vexpress-poweroff.c b/drivers/power/reset/vexpress-poweroff.c
> > index 476aa49..d752233 100644
> > --- a/drivers/power/reset/vexpress-poweroff.c
> > +++ b/drivers/power/reset/vexpress-poweroff.c
> > @@ -15,6 +15,7 @@
> >  #include <linux/of.h>
> >  #include <linux/of_device.h>
> >  #include <linux/platform_device.h>
> > +#include <linux/sched.h>
> >  #include <linux/stat.h>
> >  #include <linux/vexpress.h>
> >
> > @@ -27,12 +28,12 @@ static void vexpress_reset_do(struct device *dev, const char *what)
> >                         vexpress_config_func_get_by_dev(dev);
> >
> >         if (func) {
> > -               unsigned long timeout;
> > +               unsigned long long timeout_ns;
> >
> >                 err = vexpress_config_write(func, 0, 0);
> >
> > -               timeout = jiffies + HZ;
> > -               while (time_before(jiffies, timeout))
> > +               timeout_ns = sched_clock() + 50000000;
> > +               while (!err && time_before64(sched_clock(), timeout_ns))
> >                         cpu_relax();
> >         }
> 
> So this may not be a problem in this particular case, but sched_clock
> could be backed by jiffies on some hardware, causing the same problem
> to appear.

Uh, right. As you guessed, on vexpress sched_clock, once registered, is
always available. But of course I shouldn't fully rely on this.

> Might udelay/mdelay be a better fit for this sort of case (since
> udelay may be counter backed, but may also be loop backed on hardware
> without continuous counters)?

I'm sure I though about udelay, but for some reason decided against it.
The reason may be already invalid, so I'll check it again :-)

Thanks!

Paweł

--
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/