Re: 2.6.39-rc5-git2 boot crashs

[Borislav Petkov]

On Wed, May 04, 2011 at 09:38:54AM +0200, Ingo Molnar wrote:
> 
> * Borislav Petkov <bp@xxxxxxxxx> wrote:
> 
> > On Tue, May 03, 2011 at 02:45:10PM -0700, Linus Torvalds wrote:
> > > Now, Werner is actually _running_ on an AMD Opteron (or whatever
> > > family 15 is), I think.
> > 
> > http://marc.info/?l=linux-kernel&m=130438580705332 says it's a AMD
> > Athlon(tm) 64 X2 Dual Core Processor 4800+, i.e. a good old K8 desktop.
> > The X86_ELAN Kconfig option, however, says:
> > 
> > config X86_ELAN
> > 	bool "AMD Elan"
> > 	depends on X86_32
> > 	depends on X86_EXTENDED_PLATFORM
> > 	---help---
> > 	  Select this for an AMD Elan processor.
> > 
> > 	  Do not use this option for K6/Athlon/Opteron processors!
> > 
> > 	  If unsure, choose "PC-compatible" instead.
> > 
> > so if the sentence before last used to mean anything, this could be a
> > problem. Quick search about it gives http://lkml.org/lkml/2004/1/12/239
> > which introduces that different compiler arch for ELAN: -march=i486,
> > which could conflict with the generic selection?
> 
> Well, but CONFIG_X86_ELAN=y always worked/booted fine on generic hardware - 
> including later AMD CPUs. I have booted it on an AMD Athlon64 CPU today:
> 
>  config-Wed_May__4_09_26_50_CEST_2011.good:CONFIG_X86_32_SMP=y
>  config-Wed_May__4_09_26_50_CEST_2011.good:CONFIG_MELAN=y
>  config-Wed_May__4_09_26_50_CEST_2011.good:CONFIG_SMP=y
> 
> When it comes to regressions it's past behavior that counts, not the Kconfig 
> help text.

Right, I was referring to the odd problem Linus was talking about:

"He had CONFIG_X86_EXTENDED_PLATTFORM and CONFIG_X86_ELAN on, and that
has apparently worked for him (and a lot of other people - he does a
distribution) up until 2.6.38. But as of 2.6.39-rc1 it causes some
really odd problems under IO (his test-case is "unzip", but that's
probably fairly random). The problem seems to show up as a bogus IO list
for SATA, causing a big WARN_ON() or oops and then a dead machine due to
IO problems."

So booting might've not triggered it. But reportedly .38 was fine so
yeah, the Kconfig help text might not even mean anything anymore.

Maybe I should run the same .config and test case on a K8 box here to
see what happens.

@Werner: can you send me the exact .config and the testcase that
triggers the issue?

Thanks.

-- 
Regards/Gruss,
    Boris.
--
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/