RE: [PATCH] x86/mcheck/therm_throt.c: Don't log power limit andpackage level thermal throttle event in mce log

From: Yu, Fenghua
Date: Tue Dec 06 2011 - 12:52:27 EST


> -----Original Message-----
> From: Borislav Petkov [mailto:bp@xxxxxxxxx]
> Sent: Tuesday, December 06, 2011 7:31 AM
> To: Luck, Tony
> Cc: Yu, Fenghua; H Peter Anvin; Thomas Gleixner; Ingo Molnar; Andrew
> Morton; Brown, Len; linux-kernel; x86
> Subject: Re: [PATCH] x86/mcheck/therm_throt.c: Don't log power limit
> and package level thermal throttle event in mce log
>
> On Mon, Dec 05, 2011 at 02:18:26PM -0800, Luck, Tony wrote:
> > This looks like a sane improvement, Tony I'm assuming you're handling
> this?
> >
> > I can push this - but it would be nice to have some other comments
> from
> > outside of Intel (we've been beaten up written-by Intel, acked-by:
> Intel
> > too many times).
> >
> > Can I count that as an "Acked-by:"?
>
> Well, from looking at Udo's feedback here
>
> http://marc.info/?l=linux-kernel&m=132206108724885
>
> this patch looks like it helps in his case partially. I'm still not
> sure
> about all the printks in therm_throt_process(), do you guys feel the
> need to notify the user that the CPU clock was throttled/power limits
> were applied, or rather have something more subtle...?

The printk is one way to notify users about the power limit and thermal throttle. The printk only dumps the events in an interval (300*HZ).

Another way is to count the events in /sys/devices/system/cpu/cpu#/thermal_throttle. In this way, kernel logs every interrupt on any cpu into respective counters. User application can poll the counters and get more accurate and timely information for the events.

As explained in this patch, core level thermal throttle is still logged in mcelog for legacy reason after this patch is applied.

Thanks.

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