On Tue, May 22, 2018 at 08:38:39AM -0500, Alex G. wrote:
> > It looks like the *real* reason for this change is that you
> > re-introduce ghes_severity() as a different function in the second
> > patch.
>
> /me holds fist at Borislav
That was a misunderstanding with Rafael and me - we fixed it on IRC.
But this is not the real problem with your approach - it is the marking
of all PCIe errors as recoverable, regardless of the signature. That's a
no-no, IMO.
--
Regards/Gruss,
Boris.
Good mailing practices for 400: avoid top-posting and trim the reply.
