fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2025-12-24 08:50:13 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

nir/algebraic: Optimize some extract_* expressions

Browse source 
v2: Add missing '!options->lower_extract_byte' to the last two
patterns. Every driver except Asahi sets both or neither.

shader-db:

All Intel platforms had similar results. (DG2 shown)
total instructions in shared programs: 19659360 -> 19659356 (<.01%)
instructions in affected programs: 44 -> 40 (-9.09%)
helped: 2 / HURT: 0

total cycles in shared programs: 823432524 -> 823432520 (<.01%)
cycles in affected programs: 1722 -> 1718 (-0.23%)
helped: 2 / HURT: 0

fossil-db:

All Intel platforms had similar results. (DG2 shown)
Totals:
Instrs: 153989787 -> 153989617 (-0.00%)
Cycle count: 17562079230 -> 17562079493 (+0.00%); split: -0.00%, +0.00%

Totals from 24 (0.00% of 631369) affected shaders:
Instrs: 13733 -> 13563 (-1.24%)
Cycle count: 341392 -> 341655 (+0.08%); split: -0.25%, +0.33%

Reviewed-by: Caio Oliveira <caio.oliveira@intel.com> [v1]
Acked-by: Alyssa Rosenzweig <alyssa@rosenzweig.io>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/27891>
This commit is contained in:
Ian Romanick 2024-04-29 12:38:55 -07:00
parent 0fa17962d6
commit 1b8cf06fc7
1 changed files with 17 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
src/compiler/nir/nir_opt_algebraic.py
View file

@ -1698,6 +1698,8 @@ optimizations.extend([
(('ishr', 'a@32', 24), ('extract_i8', a, 3), '!options->lower_extract_byte'),
(('ishr', 'a@64', 56), ('extract_i8', a, 7), '!options->lower_extract_byte'),
(('iand', 0xff, a), ('extract_u8', a, 0), '!options->lower_extract_byte'),
(('ishr', ('iand', a, 0x0000ff00), 8), ('extract_u8', a, 1), '!options->lower_extract_byte'),
(('ishr', ('iand', a, 0x00ff0000), 16), ('extract_u8', a, 2), '!options->lower_extract_byte'),
# Common pattern in many Vulkan CTS tests that read 8-bit integers from a
# storage buffer.
@ -1720,6 +1722,16 @@ optimizations.extend([
(('extract_u8', ('extract_i8', a, b), 0), ('extract_u8', a, b)),
(('extract_u8', ('extract_u8', a, b), 0), ('extract_u8', a, b)),
# The extract_X8(a & 0xff) patterns aren't included because the iand will
# already be converted to extract_u8.
(('extract_i8', ('iand', a, 0x0000ff00), 1), ('extract_i8', a, 1)),
(('extract_i8', ('iand', a, 0x00ff0000), 2), ('extract_i8', a, 2)),
(('extract_i8', ('iand', a, 0xff000000), 3), ('extract_i8', a, 3)),
(('extract_u8', ('iand', a, 0x0000ff00), 1), ('extract_u8', a, 1)),
(('extract_u8', ('iand', a, 0x00ff0000), 2), ('extract_u8', a, 2)),
(('extract_u8', ('iand', a, 0xff000000), 3), ('extract_u8', a, 3)),
# Word extraction
(('ushr', ('ishl', 'a@32', 16), 16), ('extract_u16', a, 0), '!options->lower_extract_word'),
(('ushr', 'a@32', 16), ('extract_u16', a, 1), '!options->lower_extract_word'),
@ -1739,6 +1751,11 @@ optimizations.extend([
(('extract_u16', ('extract_i16', a, b), 0), ('extract_u16', a, b)),
(('extract_u16', ('extract_u16', a, b), 0), ('extract_u16', a, b)),
# The extract_X16(a & 0xff) patterns aren't included because the iand will
# already be converted to extract_u8.
(('extract_i16', ('iand', a, 0x00ff0000), 1), ('extract_u8', a, 2), '!options->lower_extract_byte'), # extract_u8 is correct
(('extract_u16', ('iand', a, 0x00ff0000), 1), ('extract_u8', a, 2), '!options->lower_extract_byte'),
# Lower pack/unpack
(('pack_64_2x32_split', a, b), ('ior', ('u2u64', a), ('ishl', ('u2u64', b), 32)), 'options->lower_pack_64_2x32_split'),
(('pack_32_2x16_split', a, b), ('ior', ('u2u32', a), ('ishl', ('u2u32', b), 16)), 'options->lower_pack_32_2x16_split || options->lower_pack_split'),