nir: add options::vectorize_vec2_16bit to limit vectorization to vec2 16
authorMarek Olšák <marek.olsak@amd.com>
Mon, 4 May 2020 00:10:57 +0000 (20:10 -0400)
committerMarge Bot <eric+marge@anholt.net>
Tue, 2 Jun 2020 20:01:18 +0000 (20:01 +0000)
commit116e006693dc37245b7c0823e1a394ad9c2fb770
tree9212afb1e97dbc1fcb9a4afc2a67771f667d95c3
parenta6916d1ce823439d7de752b2f2013e9b3e434d82
nir: add options::vectorize_vec2_16bit to limit vectorization to vec2 16

for hardware that is scalar but can do 2 16-bit operations on low and high
16 bits of registers at once.

Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>
src/compiler/nir/nir.h
src/compiler/nir/nir_opt_vectorize.c