index
:
mesa/mesa
10.0
10.1
10.2
10.3
10.4
10.5
10.6
11.0
11.1
11.2
12.0
13.0
17.0
17.1
17.2
17.3
18.0
18.1
18.2
18.3
19.0
19.1
19.2
19.3
20.0
20.1
20.2
20.3
21.0
21.1
21.2
21.3
22.0
22.1
22.2
22.3
23.0
23.1
23.2
23.3
24.0
24.1
7.10
7.11
7.8
7.8-gles
7.9
8.0
9.0
9.1
9.2
a7xx-gmem
amber
elima/radv-video-encode-caps-maxbitrate
explicit-sync
main
powervr-mesa-next-wayland
review/fragment_shader_barycentric
staging/23.2
staging/23.3
staging/24.0
staging/24.1
uav-counter-meta
vk-no-nir-android
zink-stablefix
The Mesa 3D Graphics Library (mirrored from https://gitlab.freedesktop.org/mesa/mesa)
brianp
summary
refs
log
tree
commit
diff
log msg
author
committer
range
path:
root
/
src
/
amd
/
compiler
/
aco_optimizer.cpp
Age
Commit message (
Expand
)
Author
Files
Lines
2022-04-13
aco: handle read2st64/write2st64 in optimizer
Rhys Perry
1
-9
/
+15
2022-04-13
aco: remove occurences of VCC hint
Daniel Schürmann
1
-17
/
+0
2022-04-13
aco: use vcc for 64-bit vgpr addition
Rhys Perry
1
-2
/
+4
2022-04-04
aco: don't use v_mad_mix on GFX9 if 16-bit denormals must be preserved
Rhys Perry
1
-0
/
+4
2022-04-01
aco/optimizer: check recursively if we can eliminate s_and exec
Daniel Schürmann
1
-12
/
+27
2022-03-25
aco: remove 'high' parameter from can_use_opsel()
Daniel Schürmann
1
-2
/
+2
2022-03-25
aco/optimizer: fix call to can_use_opsel() in apply_insert()
Daniel Schürmann
1
-1
/
+1
2022-03-24
aco: Fix an MSVC warning
Boris Brezillon
1
-1
/
+1
2022-03-17
aco: use v_fma_mix to combine mul/add/fma output conversions
Rhys Perry
1
-4
/
+48
2022-03-17
aco: apply clamp to v_fma_mix
Rhys Perry
1
-3
/
+8
2022-03-17
aco: combine add/mul as v_fma_mix into fma
Rhys Perry
1
-34
/
+103
2022-03-17
aco: use v_fma_mix to combine mul/add/fma input conversions
Rhys Perry
1
-5
/
+153
2022-03-17
aco: refactor selection of mad/fma
Rhys Perry
1
-16
/
+16
2022-02-03
aco: fix neg(mul)/abs(mul) optimization with different bit-size
Rhys Perry
1
-0
/
+2
2022-02-03
aco: don't combine add/mul of different bit-size
Rhys Perry
1
-0
/
+3
2022-02-03
aco: don't apply omod/clamp of different bit-size
Rhys Perry
1
-0
/
+3
2022-02-03
aco: don't combine fneg/fabs of different bit-size
Rhys Perry
1
-5
/
+8
2022-01-31
aco: preserve pass_flags during format conversions
Rhys Perry
1
-0
/
+2
2022-01-31
aco: optimize abs(mul(a, b))
Rhys Perry
1
-5
/
+5
2022-01-31
aco: fix neg(abs(mul(a, b))) if the mul is not VOP3
Rhys Perry
1
-4
/
+8
2022-01-20
aco: consider legacy multiplications in optimizer
Rhys Perry
1
-7
/
+24
2021-12-31
aco: support DPP8
Tatsuyuki Ishi
1
-30
/
+59
2021-12-31
aco/optimizer: apply extract from subdword p_split_vector
Daniel Schürmann
1
-0
/
+9
2021-12-31
aco/optimizer: optimize extract(extract())
Daniel Schürmann
1
-3
/
+37
2021-12-31
aco/optimizer: apply extract from p_extract_vector
Daniel Schürmann
1
-5
/
+16
2021-12-31
aco/optimizer: keep instr_mod_labels after applying extract
Daniel Schürmann
1
-2
/
+2
2021-12-21
aco/optimizer: propagate and fold inline constants on VOP3P instructions
Daniel Schürmann
1
-2
/
+72
2021-12-21
aco: change fneg for VOP3P to use fmul with +1.0
Daniel Schürmann
1
-3
/
+3
2021-12-21
aco/optimizer: fix fneg modifier propagation on VOP3P
Daniel Schürmann
1
-4
/
+6
2021-12-17
aco: remove SMEM constant/addition combining out of the loop
Rhys Perry
1
-53
/
+57
2021-12-17
aco: skip &-4 before SMEM
Rhys Perry
1
-0
/
+34
2021-12-17
aco: disallow SMEM offsets that are not multiples of 4
Rhys Perry
1
-1
/
+2
2021-12-17
aco: fix check_vop3_operands() for f16vec2 ffma fneg combine
Rhys Perry
1
-2
/
+5
2021-12-13
radv,aco: implement nir_op_ffma
Rhys Perry
1
-3
/
+1
2021-12-13
aco: swap operands if necessary to create v_madak/v_fmaak
Rhys Perry
1
-22
/
+38
2021-12-13
aco: create v_fmamk_f32/v_fmaak_f32 from nir_op_ffma
Rhys Perry
1
-0
/
+7
2021-12-13
aco: use more predictable tiebreaker when forming MADs
Rhys Perry
1
-1
/
+8
2021-11-29
aco: properly update use counts if a extract is still used
Rhys Perry
1
-1
/
+3
2021-11-25
aco: fix right shift of exponent 32 detected by UBSAN
Samuel Pitoiset
1
-1
/
+1
2021-10-20
aco: disable mul(cndmask(0, 1, b), a) optimization sometimes
Rhys Perry
1
-1
/
+2
2021-10-12
aco/optimizer: Skip SDWA on v_lshlrev when unnecessary in apply_extract.
Timur Kristóf
1
-4
/
+11
2021-09-28
aco: clang-format
Daniel Schürmann
1
-4
/
+3
2021-09-28
aco: Omit p_extract after ds_read with matching bit size.
Timur Kristóf
1
-0
/
+61
2021-09-20
aco: Add ability to optimize v_lshl + v_sub into v_mad_i32_i24.
Timur Kristóf
1
-17
/
+41
2021-09-18
aco: cleanup assignment of unique_ptrs
Filip Gawin
1
-5
/
+5
2021-09-02
aco: remove explicit dst_preserve flag
Daniel Schürmann
1
-1
/
+0
2021-09-02
aco: rewrite SDWA selector
Daniel Schürmann
1
-41
/
+42
2021-08-31
aco: remove DPP when applying constants/literals/sgprs
Rhys Perry
1
-3
/
+10
2021-08-31
aco: don't constant propagate to DPP instructions
Rhys Perry
1
-1
/
+1
2021-08-30
aco: Unset 16 and 24-bit flags from operands in apply_extract.
Timur Kristóf
1
-0
/
+3
[next]