[git pull] drm fixes
airlied at gmail.com
Tue Dec 6 08:36:33 PST 2011
> What the heck is up?
Well I do care about kexec but only due to being forced into caring
about it for a certain enterprise distro that uses it a lot, so maybe
I was a bit biased in this case, and I dislike random memory
corruptions due to my subsystem even in the kexec case. Writing a
random 0 dword somewhere in memory isn't that pretty and no fun to
track down, when the kexec looks like it succeeds.
But I've no problem leaving it sitting around until -next.
The kzalloc->kcalloc ones are partly better integer overflow defence
where userspace passes in a number of objects but I just need to make
sure they are that and not cleanups, but I haven't gotten around to it
Cool I'll resend just the Intel and obvious radeon one and stick the
others into -next.
More information about the dri-devel