Messages in this thread | | | Date | Sat, 01 Feb 2014 17:25:18 -0800 | From | "H. Peter Anvin" <> | Subject | Re: [PATCH] Make math_state_restore() save and restore the interrupt flag |
| |
On 02/01/2014 05:19 PM, George Spelvin wrote: >> OK, let's circle back for a bit. We have an active bug, and we clearly >> have a lot of restructuring that could/should be done. We need to fix >> the bug first; if we're going to a bunch of restructuring then that >> ought to be separate. The first bit is how we fix the immediate bug. > > Well, that's what the [PATCH] that started this thread,s for which seems > to greatly reduce the incidence. > > The issue being discussed here is the fact that it's far from clear to > people that the result of applying that patch is actually bug-free. > > Given how long it's been in shipping kernels (2.6.26 or 3.10, depending > on what you think causes it to trigger) and how few complaints there > have been, I'm happy to take a while to think about it. >
Yes, I would agree with that. I guess I should have said "how do we fix the immediate bug(s) properly."
-hpa
| |