fdo-mirrors/mesa
1
0
Fork 0
mirror of https://gitlab.freedesktop.org/mesa/mesa.git synced 2026-05-06 20:18:12 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions 2

nir/opt_algebraic: create ldexp from exp2
Some checks are pending
macOS-CI / macOS-CI (dri) (push) Waiting to run
Details
macOS-CI / macOS-CI (xlib) (push) Waiting to run
Details

Browse source 
ldexp uses the full width VALU path, exp2 the transcendental SIMD8.

Foz-DB Navi21:
Totals from 729 (0.64% of 114627) affected shaders:
MaxWaves: 20071 -> 20103 (+0.16%); split: +0.18%, -0.02%
Instrs: 869129 -> 867654 (-0.17%); split: -0.17%, +0.00%
CodeSize: 4709000 -> 4708460 (-0.01%); split: -0.02%, +0.00%
VGPRs: 31184 -> 31128 (-0.18%); split: -0.23%, +0.05%
Latency: 7610726 -> 7597238 (-0.18%); split: -0.18%, +0.00%
InvThroughput: 1822323 -> 1819815 (-0.14%); split: -0.14%, +0.00%
VClause: 22494 -> 22493 (-0.00%); split: -0.03%, +0.02%
SClause: 20520 -> 20509 (-0.05%)
Copies: 72025 -> 72024 (-0.00%); split: -0.01%, +0.01%
Branches: 22028 -> 22029 (+0.00%)
PreVGPRs: 21601 -> 21602 (+0.00%)
VALU: 604821 -> 603339 (-0.25%); split: -0.25%, +0.00%
SALU: 114258 -> 114262 (+0.00%); split: -0.00%, +0.01%

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/33900>
This commit is contained in:
Georg Lehmann 2025-03-05 17:59:21 +01:00 • committed by Marge Bot
parent ec331cc48a
commit 6cfe6eaa79
1 changed files with 7 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

7
src/compiler/nir/nir_opt_algebraic.py
View file

@ -3167,6 +3167,10 @@ optimizations += [
(('ldexp@16', 'x', 'exp'), ldexp('x', 'exp', 16), '!options->has_ldexp', TestStatus.UNSUPPORTED), # All test inputs skipped.
(('ldexp@32', 'x', 'exp'), ldexp('x', 'exp', 32), '!options->has_ldexp', TestStatus.UNSUPPORTED), # All test inputs skipped.
(('ldexp@64', 'x', 'exp'), ldexp('x', 'exp', 64), '!options->has_ldexp', TestStatus.UNSUPPORTED), # All test inputs skipped.
(('fexp2(contract)', ('i2f', 'a@8')), ('ldexp', 1.0, ('i2i32', a)), 'options->has_ldexp'),
(('fexp2(contract)', ('i2f', 'a@16')), ('ldexp', 1.0, ('i2i32', a)), 'options->has_ldexp'),
(('fexp2(contract)', ('i2f', 'a@32')), ('ldexp', 1.0, a), 'options->has_ldexp'),
(('fexp2(contract,ninf)', ('u2f', a)), ('ldexp', 1.0, ('u2u32', a)), 'options->has_ldexp'),
]
# XCOM 2 (OpenGL) open-codes bitfieldReverse()
@ -3825,6 +3829,9 @@ late_optimizations.extend([
(('~fadd', ('fneg', ('ffmaz(is_used_once)', a, b, ('ffmaz', c, d, ('fmulz(is_used_once)', 'e(is_not_const_and_not_fsign)', 'f(is_not_const_and_not_fsign)')))), 'g(is_not_const)'),
('ffmaz', ('fneg', a), b, ('ffmaz', ('fneg', c), d, ('ffmaz', ('fneg', e), 'f', 'g'))), '(info->stage != MESA_SHADER_VERTEX && info->stage != MESA_SHADER_GEOMETRY) && !options->intel_vec4'),
(('fmul(contract)', a, ('ldexp(is_used_once)', 1.0, b)), ('ldexp', a, b), 'options->has_ldexp'),
(('frcp(contract,ninf)', ('ldexp', 1.0, b)), ('ldexp', 1.0, ('ineg', b)), 'options->has_ldexp'),
# Section 8.8 (Integer Functions) of the GLSL 4.60 spec says:
#
# If bits is zero, the result will be zero.