Re: -mm merge plans for 2.6.21

[Doug Thompson]

--- Andi Kleen <andi@xxxxxxxxxxxxxx> wrote:
> > As to the size of the MCE code doing everything EDAC K8 does, that
> is
> > mostly true. BUT then the x86_64 MCE mechanism becomes the
> exception
> > path. 
> 
> Sorry you lost me on that one.
> 
> -Andi

In similiar manner of the original SATA driver model, the one that
didn't use the SCSI stack. Then the SATA model was enhanced to fit
under SCSI, then the SATA drivers ported to run under the SATA module.
At that point, the original SATA driver model was an exception to the
SATA model.

Yes, if all the world used the hardware MCE mechanism, then the MCE
device probably would satisify the requirements. But since linux runs
across so many architectures and processors, the export of a single
architecture's mechanism then implies that a second interface is
necessary anyway in order to provide for those "other" archs. 

At that point we then move the abstraction into userspace anyway.
Scripts will then need to "know" which arch they are on, in order to
pull ECC events from one or the other interface.

Thus, where is the line as to where the "portable" abstraction "fits"?

thanks for your comments

doug t


-
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/