projects / platform / upstream / mesa.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: e088390) | patch
gallivm: Use 8 wide AoS sampling on AVX2.
authorRoland Scheidegger <sroland@vmware.com>
Thu, 21 Nov 2013 09:14:47 +0000 (09:14 +0000)
committerJose Fonseca <jfonseca@vmware.com>
Tue, 4 Oct 2016 22:36:20 +0000 (23:36 +0100)
commitbc80741d7ab201cdffd2af4fbcfc4237dcce5f7e
tree46ea4901d1943f2d5288c64149fc0271b3fbb80ftree | snapshot
parente088390c7ddce4e64559a5dad6235ffc430ac736commit | diff
gallivm: Use 8 wide AoS sampling on AVX2.

v2: Make sure that with num_lods > 1 and min_filter != mag_filter we
still enter the splitting path. So this case would still use 4-wide aos
path (as a side note, the 4-wide aos sampling path could actually be
improved quite a bit if we have avx2, by just doing the filtering with
256bit vectors).

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
src/gallium/auxiliary/gallivm/lp_bld_sample_soa.c diff | blob | history
Domain: Graphics System / GL;