On Wed, Jul 18, 2012 at 6:36 PM, Adam Jackson <ajax at redhat.com> wrote: > On Wed, 2012-07-18 at 09:23 -0700, Ian Romanick wrote: >> On 07/18/2012 02:20 AM, Daniel Vetter wrote: >> > A few things: >> > - I agree with Chris that reset_in_progress should go, if userspace can >> > sneak in and witness a reset event, we have a bug in the kernel. Since >> > very recently, we actually have a few bugs less in that area ;-) >> >> I'm operating under the assumption that, from user space's perspective, >> resets are not instantaneous. If resets are instantaneous, that may >> change things. > > Do they need to be instantaneous, or do they merely need to be > atomic-and-reasonably-quick? Could just block new ioctl submission > until the reset completes. That's actually what we already do, that "block ioctls until reset completes thing" ;-) -Daniel -- Daniel Vetter daniel.vetter at ffwll.ch - +41 (0) 79 364 57 48 - http://blog.ffwll.ch
