Messages in this thread | | | Date | Wed, 1 Dec 2010 22:30:34 +0100 | From | Andi Kleen <> | Subject | Re: [PATCH 4/9] x86, NMI: Remove DIE_NMI_IPI and add priorties to handlers |
| |
On Wed, Dec 01, 2010 at 09:41:28PM +0300, Cyrill Gorcunov wrote: > On Tue, Nov 30, 2010 at 05:27:25PM -0500, Don Zickus wrote: > > When re-ordering how the NMI handles its callbacks, a conversation started > > asking what DIE_NMI_IPI meant. No one could answer it. > > It should have came from commit > > | commit c4b2bffee2a4115fed2825530f2b906ee2f17bd7 > | Author: Andi Kleen <ak@suse.de> > | Date: Fri Jan 23 18:46:40 2004 -0800 > | > | [PATCH] x86-64 merge > | > | Mainly lots of bug fixes and a few minor features. One change is that > | it uses drivers/Kconfig now like i386. This requires a few minor changes in > | outside Kconfig files which I am sending separately. > ... > > Andi do you remember what the initial idea was? Didn't find any user of it > even in this old commit. Just curious.
The original die names were pretty much a 1:1 conversion of the hooks used by both the external KDB and KGDB patchkits floating around at that time.
IIRC DIE_NMI_IPI was the one that was early in the NMI handler and DIE_NMI late when everything else failed.
So you could use NMI_IPI when you just wanted to stop all CPUs with a broadcast NMI and can check that reliable through some memory location, and NMI when you wanted to drop into the debugger as a last resort.
-Andi
-- ak@linux.intel.com -- Speaking for myself only.
| |