On 17/08/2022 19:19, Borislav Petkov wrote: > On Wed, Aug 17, 2022 at 07:09:26PM -0300, Guilherme G. Piccoli wrote: >> Again - a matter of a trade-off, a good compromise must be agreed by all >> parties (kdump maintainers are usually extremely afraid of taking risks >> to not break kdump). > > Right, and logging hw errors from a panic notifier feels simply weird. > > x86 has its own, special notifier exactly for that and it is independent > from the panic path but it gets run right after the exception raised due > to the hw error is done. > > Dunno if ARM has such facilities tho... > > Thx. > Yeah, MCE stuff right? Not sure for ARM and other architectures as well. Cheers, Guilherme
