mirrors/mesa
0
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2024-12-01 06:04:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

agx: Implement extract_[ui]16

Browse Source 
Instead of lowering to bitwise ops. Yet another way of subdividing in NIR.
Probably insignificant but makes it easy to check that the pass ordering from the
previous pass is right. It does let us get much better codegen for
unpacksnorm2x16, whatever that's worth.

No shader-db changes.

Signed-off-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/21674>
This commit is contained in:
Alyssa Rosenzweig 2023-03-03 00:29:49 -05:00 committed by Marge Bot
parent 706815488e
commit 8a53050d7d
2 changed files with 11 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
src/asahi/compiler/agx_compile.h
View File

@ -206,7 +206,6 @@ static const nir_shader_compiler_options agx_nir_options = {
.lower_pack_half_2x16 = true,
.lower_unpack_half_2x16 = true,
.lower_extract_byte = true,
.lower_extract_word = true,
.lower_insert_byte = true,
.lower_insert_word = true,
.lower_cs_local_index_to_id = true,

13
src/asahi/compiler/agx_nir_algebraic.py
View File

@ -21,12 +21,21 @@ for s in [8, 16, 32, 64]:
lower_sm5_shift += [((shift, f'a@{s}', b),
(shift, a, ('iand', b, s - 1)))]
lower_half_pack = [
lower_pack = [
(('pack_half_2x16_split', a, b),
('pack_32_2x16_split', ('f2f16', a), ('f2f16', b))),
(('unpack_half_2x16_split_x', a), ('f2f32', ('unpack_32_2x16_split_x', a))),
(('unpack_half_2x16_split_y', a), ('f2f32', ('unpack_32_2x16_split_y', a))),
(('extract_u16', 'a@32', 0), ('u2u32', ('unpack_32_2x16_split_x', a))),
(('extract_u16', 'a@32', 1), ('u2u32', ('unpack_32_2x16_split_y', a))),
(('extract_i16', 'a@32', 0), ('i2i32', ('unpack_32_2x16_split_x', a))),
(('extract_i16', 'a@32', 1), ('i2i32', ('unpack_32_2x16_split_y', a))),
# For optimizing extract->convert sequences for unpack/pack norm
(('u2f32', ('u2u32', a)), ('u2f32', a)),
(('i2f32', ('i2i32', a)), ('i2f32', a)),
]
def main():
@ -42,7 +51,7 @@ def run():
print('#include "agx_nir.h"')
print(nir_algebraic.AlgebraicPass("agx_nir_lower_algebraic_late",
lower_sm5_shift + lower_half_pack).render())
lower_sm5_shift + lower_pack).render())
if __name__ == '__main__':