drm/nv50: improve vram page table construction
authorBen Skeggs <bskeggs@redhat.com>
Thu, 11 Feb 2010 01:31:44 +0000 (11:31 +1000)
committerBen Skeggs <bskeggs@redhat.com>
Tue, 23 Feb 2010 03:50:03 +0000 (13:50 +1000)
This commit changes nouveau to construct PTEs which look very much like
the ones the binary driver creates.

I presume that filling multiple PTEs identically with length flags and
the physical address of the start of a block of VRAM is a hint to the
memory controller that it need not perform additional page table lookups
for that range of addresses.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

No differences found