[openib-general] [linux-pm] 2.6.19-rc4: known unfixed regressions (v3)

Len Brown len.brown at intel.com
Mon Nov 6 21:41:24 PST 2006


On Monday 06 November 2006 21:17, Ernst Herzberg wrote:
> On Sunday 05 November 2006 07:48, Adrian Bunk wrote:
> > ...
> > Subject    : ThinkPad R50p: boot fail with (lapic && on_battery)
> > References : http://lkml.org/lkml/2006/10/31/333
> > Submitter  : Ernst Herzberg <earny at net4u.de>
> > Status     : problem is being debugged
> 
> Update:
> 
> 2.6.19-rc4-git11 does _not_ fix the problem.
> 
> But now it doesn't matter if Cardbus/PCMCIA is compiled in or not.
> 
> What also matters is a setting in the BIOS:
> 
> If i set instead of
> 
> Power -> 
>    SpeedStep ->
>       Mode for AC -    "Max Performance"
>       Mode for Battery "Max Battery"
> 
> to
> 
> Power -> 
>    SpeedStep ->
>       Mode for AC -    "Max Battery"
>       Mode for Battery "Max Performance" 
> 
> he _does_ boot from battery but also on AC...

so with the new settings it boots properly in either AC or battery mode with forced "lapic"?
Strange, one would expect this to refer to APM settings, but who knows...

Please test if booting with "processor.max_cstate=1" makes any difference

Please test if building with CONFIG_CPU_FREQ=n makes any difference.

Also, please make sure that booting with "apm=off" makes no difference --
there is a bug where the APM code is not currently disabled in ACPI mode,
and who knows what effect that may have...

> Also the problem is now(?) not very reliable. Sometimes the boot is successful 
> even on battery and then the laptop works without a glitch.
> This makes it not easier to isolate the problem.
> 

> Is somewhere a howto-revert-patches-in-kernel-git-for-raw-beginners?

Documentation/git-bisect.txt is what you want -- though if you can't
reliably reproduce the failure it may not easily lead you to the failure.

Also, it may be a good use of time to make darn sure that booting 2.6.18
with forced "lapic" when on battery does not fail.  ie. is this really
a regression -- of did forcing the lapic on this box never work reliably?

-Len




More information about the general mailing list