On Tue, Feb 14, 2012 at 2:18 AM, Eric Anholt eric@anholt.net wrote:
On Thu, 9 Feb 2012 00:19:31 +0100, Ben Widawsky ben@bwidawsk.net wrote:
Mostly copied from i915 gtt mmaps, this will properly fault in pages as the user tries to use them. The only thing of note are that no prefaulting occurs, so perhaps some kind of madvise will happen later if needed.
The only other thing missing right not is shrinker support, which will come next after I figure out if locking is actually required right now. Hmm, and now that I think about it, mremap, and munmap may not work either.
I still think the mmap ioctl should be fixed to just get the mmap offset, and you use the normal mmap() syscall later. The do_mmap-in-the-ioctl in the original GEM implementation was a bad idea.
Yes not accepting an mmap inside things anymore, shouldn't have accepted the one in GEM.
Dave.