09:28 karolherbst: uhm...
09:28 karolherbst: imirkin: I think we waste a lot of memory in our mm_slab code ...
09:29 karolherbst: nouveau_mm_allocate called with size == 0x200000, but we allocate a 0x400000 big slab
09:34 karolherbst: ehh wait, no that's fine
09:34 karolherbst: the code is just confusing
10:00 karolherbst: *sigh*, two threads mapping the same bo
10:00 karolherbst: lol
10:01 HdkR: Seems fine to me :)
10:01 karolherbst: not if they race on the VM address
10:01 karolherbst: race around drm_mmap :/
10:02 HdkR: Time for a mutex?
10:02 karolherbst: not at this level
10:02 karolherbst: I am actually wondering why we even need nouveau_bo_map...
10:02 karolherbst: because.. I think all bos are already mapped anyway
10:05 karolherbst: huh actually.. I can lock a different lock earlier
16:19 fling: No crases yet!
16:19 fling: lua hack in awesome helped
16:19 fling: llvmpipe rocks!