[PATCH] Avoid use of uninitialised spinlock in EEH.

[linas]

Hi,

It seems that I missed the begining of this conversation;
Let me jump in ayway.

On Wed, Nov 16, 2005 at 01:38:24PM +0000, David Woodhouse was heard to remark:
> If we were to attempt to fix that, we'd have to handle other platforms;
> we'd have to take into account FW_FEATURE_{PMAC,MAPLE,CELL}_ALWAYS when
> setting FW_FEATURE_ALWAYS. And even if we did that, we'd still probably
> want to set the FW_FEATURE_EEH bit _only_ if EEH-capable devices were
> found, 

There already is a global flag, computed at boot time, that indicates if
there's EEH hardware on the system: its called "eeh_subsystem_enabled".

Starting with power5, if there is EEH hardware, then it *must* be turned
on, else the hardware won't work.  Thus the flag could be renamed to
"eeh_hardware_is_present".

> Paulus, the reason your G5 doesn't die with CONFIG_EEH set is because
> you haven't got spinlock debugging. Try turning that on too :)

? What is the specific problem? EEH uses very few spinlocks; I'm having
trouble imagining what the problem is.

--linas