LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* p4-clockmod N60 errata workaround.
@ 2006-10-28 19:53 Dave Jones
       [not found] ` <598fb2280702150837h33b73f3u2c5efe7838f7d420@mail.gmail.com>
  0 siblings, 1 reply; 3+ messages in thread
From: Dave Jones @ 2006-10-28 19:53 UTC (permalink / raw)
  To: Ben Collins; +Cc: Linux Kernel, Pallipadi, Venkatesh

Ben,
 For the best part of a year since that N60 errata workaround
went in, I've had floods of complaints from users of that driver
about this driver becoming even more useless than it was before
"I had 8 frequencies, now I have 2" being the common complaint.
which was to be expected given that the intention of the errata
workaround was to cripple frequencies <2GHz.

The point worth noting however, is that none of these users ever
noticed any problems when we didn't have the workaround in place,
so they were somewhat miffed when it stopped working.

The actual errata states..

"If a system de-asserts STPCLK# at a 12.5% duty cycle, the processor
 is running below 2 GHz, and the processor thermal control circuit (TCC)
 on-demand clock modulation is active, the processor may hang.
 This erratum does not occur under the automatic mode of the TCC."

I believe the reason we never saw any problems is that we _are_ using
the TCC by default.  See the code in arch/i386/kernel/cpu/mcheck/p4.c
intel_init_thermal() and friends.

So my current feeling is that we're working around an errata that
can never happen, and crippling functionality in the process for
no good reason.  I'm leaning towards just removing this workaround.

	Dave

-- 
http://www.codemonkey.org.uk

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: p4-clockmod N60 errata workaround.
       [not found] ` <598fb2280702150837h33b73f3u2c5efe7838f7d420@mail.gmail.com>
@ 2007-02-15 16:56   ` Ben Collins
  0 siblings, 0 replies; 3+ messages in thread
From: Ben Collins @ 2007-02-15 16:56 UTC (permalink / raw)
  To: Oscar Pearce; +Cc: davej, linux-kernel

On Thu, 2007-02-15 at 10:37 -0600, Oscar Pearce wrote:
> On 10/28/06, Dave Jones <davej@redhat.com> wrote:
>         Ben,
>         For the best part of a year since that N60 errata workaround
>         went in, I've had floods of complaints from users of that
>         driver
>         about this driver becoming even more useless than it was
>         before
>         "I had 8 frequencies, now I have 2" being the common
>         complaint. 
>         which was to be expected given that the intention of the
>         errata
>         workaround was to cripple frequencies <2GHz.

We've since removed that patch from our current tree. It might remain in
some older kernels. I'll definitely check into getting it pulled from
there next update we do for them.

Thanks.

> 
-- 
Ubuntu:    http://www.ubuntu.com/
Linux1394: http://www.linux1394.org/


^ permalink raw reply	[flat|nested] 3+ messages in thread

* p4-clockmod N60 errata workaround.
@ 2006-10-28 19:54 Dave Jones
  0 siblings, 0 replies; 3+ messages in thread
From: Dave Jones @ 2006-10-28 19:54 UTC (permalink / raw)
  To: Ben Collins; +Cc: Linux Kernel, Pallipadi, Venkatesh

Ben,
 For the best part of a year since that N60 errata workaround
went in, I've had floods of complaints from users of that driver
about this driver becoming even more useless than it was before
"I had 8 frequencies, now I have 2" being the common complaint.
which was to be expected given that the intention of the errata
workaround was to cripple frequencies <2GHz.

The point worth noting however, is that none of these users ever
noticed any problems when we didn't have the workaround in place,
so they were somewhat miffed when it stopped working.

The actual errata states..

"If a system de-asserts STPCLK# at a 12.5% duty cycle, the processor
 is running below 2 GHz, and the processor thermal control circuit (TCC)
 on-demand clock modulation is active, the processor may hang.
 This erratum does not occur under the automatic mode of the TCC."

I believe the reason we never saw any problems is that we _are_ using
the TCC by default.  See the code in arch/i386/kernel/cpu/mcheck/p4.c
intel_init_thermal() and friends.

So my current feeling is that we're working around an errata that
can never happen, and crippling functionality in the process for
no good reason.  I'm leaning towards just removing this workaround.

	Dave

-- 
http://www.codemonkey.org.uk

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2007-02-15 16:57 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-10-28 19:53 p4-clockmod N60 errata workaround Dave Jones
     [not found] ` <598fb2280702150837h33b73f3u2c5efe7838f7d420@mail.gmail.com>
2007-02-15 16:56   ` Ben Collins
2006-10-28 19:54 Dave Jones

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).