Messages in this thread | | | Date | Tue, 18 Feb 2020 21:34:04 +0100 | From | Peter Zijlstra <> | Subject | Re: [RFC] #MC mess |
| |
On Tue, Feb 18, 2020 at 08:11:10PM +0000, Luck, Tony wrote: > > Then please rewrite the #MC entry code to deal with nested exceptions > > unmasking the MCE, very similr to NMI. > > #MC doesn't work like NMI. It isn't enabled by IRET. Nested #MC cause an > immediate reset. Detection of nested case is by IA32_MCG_STATUS.MCIP. > We don't clear MCG_STATUS until we are ready to return from the machine > check handler.
Ooh, excellent! That saves a whole heap of problems.
Then I suppose we should work at getting in_nmi() set. Let me have another look at things.
| |