[PATCH]: PCI Error Recovery Implementation

Linas Vepstas linas at austin.ibm.com
Wed Jun 1 07:56:28 EST 2005


On Tue, May 31, 2005 at 03:52:07PM -0500, John Rose was heard to remark:
> Hi Linas/Greg/Everyone-
> 
> +int handle_eeh_events (struct notifier_block *self,
> +                       unsigned long reason, void *ev)
> 
> At the risk of sounding like a broken record, I don't think that this
> belongs in the RPA PCI Hotplug driver.  This bit of code _uses_ PCI
> Hotplug rather than implementing it, and thus stands out from the rest
> of the module.  Not to mention that it uses EEH-specific stuff that
> doesn't belong here.  I'm in the midst of reducing the codebase of this
> module significantly, and this adds more unrelated stuff to an already
> cluttered module.
> 
> I think the PCI hotplug driver could register enable/disable functions
> with eeh.c, and that the handle_events() code should reside there.  If
> someone has a good technical explanation of why this is bad, please
> chime in.

What would be the correct way of forcing the rpaphp.o module to get
loaded, and what would be the correct way of invoking functions in that
module?

--linas




More information about the Linuxppc64-dev mailing list