Re: [PATCH 07/14] mce3: pass mce info to EDAC for decoding

[Doug Thompson <norsk5@yahoo.com>]

--- On Mon, 7/20/09, Andi Kleen <andi@firstfloor.org> wrote:

> From: Andi Kleen <andi@firstfloor.org>
> Subject: Re: [PATCH 07/14] mce3: pass mce info to EDAC for decoding
> To: "Doug Thompson" <norsk5@yahoo.com>
> Cc: "Borislav Petkov" <borislav.petkov@amd.com>, "Andi Kleen" <andi@firstfloor.org>, mingo@elte.hu, hpa@zytor.com, tglx@linutronix.de, aris@redhat.com, linux-kernel@vger.kernel.org, x86@kernel.org
> Date: Monday, July 20, 2009, 1:22 PM
> > The goal is to have a default
> handler within the kernel and a more specific handler in a
> module after it has loaded. Using the weak symbol as a
> mechanism to do that. I haven't used that, so I don't know
> if it works.
> 
> weak symbols don't work over kernel modules.
> 
> -Andi

ok, thanks. That is one thing I was questioning. Unloading the module and the "fixup" was and what the dynamic linker did is the item I had doubts on. Bottom line: It does not re-fix it back to the original.

Boris, we need to change that.

edac_stub.c is a stub section of code that resides IN the kernel at all times (unless EDAC is disabled) and is a vector point currently used for some information.

My inclination was to add to that file for the jump from MCE to EDAC module space.

My initial code was to place a function in there to call and have MCE always call it.  Then have a registration api that the EDAC Module would call to register a callback routine.

When EDAC is disabled, then the function call from MCE became a macro that was reduced to an empty string.

doug t

> 
> -- 
> ak@linux.intel.com
> -- Speaking for myself only.
>