Skip to content

LLVM optimization on PR 160810 [bot] #71

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: pr-160810-before-1758872462
Choose a base branch
from
Open

LLVM optimization on PR 160810 [bot] #71

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 22 additions & 22 deletions results/MicroBenchmarks/ImageProcessing/Dither/CMakeFiles/Dither.dir/orderedDitherKernel.s
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -351,9 +351,8 @@ orderedDitherKernel: # @orderedDitherKernel
vreplgr2vr.w $vr1, $a5
lu32i.d $a5, 0
vrepli.w $vr2, 3
vrepli.b $vr3, 0
ori $t3, $zero, 12
vrepli.w $vr4, 255
vrepli.w $vr3, 255
b .LBB0_41
.p2align 4, , 16
.LBB0_40: # %._crit_edge.us151
Expand All @@ -379,38 +378,39 @@ orderedDitherKernel: # @orderedDitherKernel
# in Loop: Header=BB0_41 Depth=1
move $t5, $a3
move $t6, $a2
vori.b $vr5, $vr0, 0
vori.b $vr4, $vr0, 0
.p2align 4, , 16
.LBB0_44: # %vector.body269
# Parent Loop BB0_41 Depth=1
# => This Inner Loop Header: Depth=2
vmuh.wu $vr6, $vr5, $vr1
vsrli.w $vr6, $vr6, 1
vori.b $vr7, $vr5, 0
vmsub.w $vr7, $vr6, $vr2
vld $vr6, $t5, 0
vilvh.w $vr8, $vr3, $vr7
vilvl.w $vr7, $vr3, $vr7
vpickve2gr.d $t7, $vr7, 0
vmuh.wu $vr5, $vr4, $vr1
vsrli.w $vr5, $vr5, 1
vori.b $vr6, $vr4, 0
vmsub.w $vr6, $vr5, $vr2
vld $vr5, $t5, 0
vshuf4i.w $vr7, $vr6, 14
vsllwil.du.wu $vr7, $vr7, 0
vsllwil.du.wu $vr6, $vr6, 0
vpickve2gr.d $t7, $vr6, 0
mul.d $t7, $t7, $t3
vpickve2gr.d $t8, $vr7, 1
vpickve2gr.d $t8, $vr6, 1
mul.d $t8, $t8, $t3
vpickve2gr.d $fp, $vr8, 0
vpickve2gr.d $fp, $vr7, 0
mul.d $fp, $fp, $t3
vpickve2gr.d $s0, $vr8, 1
vpickve2gr.d $s0, $vr7, 1
mul.d $s0, $s0, $t3
ldx.w $t7, $t4, $t7
ldx.w $t8, $t4, $t8
ldx.w $fp, $t4, $fp
ldx.w $s0, $t4, $s0
vinsgr2vr.w $vr7, $t7, 0
vinsgr2vr.w $vr7, $t8, 1
vinsgr2vr.w $vr7, $fp, 2
vinsgr2vr.w $vr7, $s0, 3
vslt.w $vr6, $vr7, $vr6
vand.v $vr6, $vr6, $vr4
vst $vr6, $t5, 0
vaddi.wu $vr5, $vr5, 4
vinsgr2vr.w $vr6, $t7, 0
vinsgr2vr.w $vr6, $t8, 1
vinsgr2vr.w $vr6, $fp, 2
vinsgr2vr.w $vr6, $s0, 3
vslt.w $vr5, $vr6, $vr5
vand.v $vr5, $vr5, $vr3
vst $vr5, $t5, 0
vaddi.wu $vr4, $vr4, 4
addi.d $t6, $t6, -4
addi.d $t5, $t5, 16
bnez $t6, .LBB0_44
Expand Down
51 changes: 24 additions & 27 deletions ...pVectorization/CMakeFiles/LoopEpilogueVectorizationBenchmarks.dir/EpilogueVectorization.s
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2340,27 +2340,26 @@ _ZL24loopWithReductionAutoVecIhEmPT_S1_S1_i: # @_ZL24loopWithReductionAutoVecIhE
addi.d $a1, $a0, 2
move $a4, $a2
vori.b $vr1, $vr0, 0
vori.b $vr2, $vr0, 0
.p2align 4, , 16
.LBB10_5: # %vector.body
# =>This Inner Loop Header: Depth=1
ld.h $a5, $a1, -2
ld.h $a6, $a1, 0
vinsgr2vr.h $vr3, $a5, 0
vinsgr2vr.h $vr4, $a6, 0
vilvl.b $vr3, $vr0, $vr3
vilvl.h $vr3, $vr0, $vr3
vilvl.w $vr3, $vr0, $vr3
vilvl.b $vr4, $vr0, $vr4
vilvl.h $vr4, $vr0, $vr4
vilvl.w $vr4, $vr0, $vr4
vinsgr2vr.h $vr2, $a5, 0
vinsgr2vr.h $vr3, $a6, 0
vsllwil.hu.bu $vr2, $vr2, 0
vsllwil.wu.hu $vr2, $vr2, 0
vsllwil.du.wu $vr2, $vr2, 0
vsllwil.hu.bu $vr3, $vr3, 0
vsllwil.wu.hu $vr3, $vr3, 0
vsllwil.du.wu $vr3, $vr3, 0
vadd.d $vr0, $vr0, $vr2
vadd.d $vr1, $vr1, $vr3
vadd.d $vr2, $vr2, $vr4
addi.d $a4, $a4, -4
addi.d $a1, $a1, 4
bnez $a4, .LBB10_5
# %bb.6: # %middle.block
vadd.d $vr0, $vr2, $vr1
vadd.d $vr0, $vr1, $vr0
vhaddw.q.d $vr0, $vr0, $vr0
vpickve2gr.d $a1, $vr0, 0
beq $a2, $a3, .LBB10_9
Expand Down Expand Up @@ -2628,25 +2627,24 @@ _ZL24loopWithReductionAutoVecItEmPT_S1_S1_i: # @_ZL24loopWithReductionAutoVecItE
addi.d $a1, $a0, 4
move $a4, $a2
vori.b $vr1, $vr0, 0
vori.b $vr2, $vr0, 0
.p2align 4, , 16
.LBB14_5: # %vector.body
# =>This Inner Loop Header: Depth=1
ld.w $a5, $a1, -4
ld.w $a6, $a1, 0
vinsgr2vr.w $vr3, $a5, 0
vinsgr2vr.w $vr4, $a6, 0
vilvl.h $vr3, $vr0, $vr3
vilvl.w $vr3, $vr0, $vr3
vilvl.h $vr4, $vr0, $vr4
vilvl.w $vr4, $vr0, $vr4
vinsgr2vr.w $vr2, $a5, 0
vinsgr2vr.w $vr3, $a6, 0
vsllwil.wu.hu $vr2, $vr2, 0
vsllwil.du.wu $vr2, $vr2, 0
vsllwil.wu.hu $vr3, $vr3, 0
vsllwil.du.wu $vr3, $vr3, 0
vadd.d $vr0, $vr0, $vr2
vadd.d $vr1, $vr1, $vr3
vadd.d $vr2, $vr2, $vr4
addi.d $a4, $a4, -4
addi.d $a1, $a1, 8
bnez $a4, .LBB14_5
# %bb.6: # %middle.block
vadd.d $vr0, $vr2, $vr1
vadd.d $vr0, $vr1, $vr0
vhaddw.q.d $vr0, $vr0, $vr0
vpickve2gr.d $a1, $vr0, 0
beq $a2, $a3, .LBB14_9
Expand Down Expand Up @@ -2894,23 +2892,22 @@ _ZL24loopWithReductionAutoVecIjEmPT_S1_S1_i: # @_ZL24loopWithReductionAutoVecIjE
addi.d $a1, $a0, 8
move $a4, $a2
vori.b $vr1, $vr0, 0
vori.b $vr2, $vr0, 0
.p2align 4, , 16
.LBB18_5: # %vector.body
# =>This Inner Loop Header: Depth=1
ld.d $a5, $a1, -8
ld.d $a6, $a1, 0
vinsgr2vr.d $vr3, $a5, 0
vinsgr2vr.d $vr4, $a6, 0
vilvl.w $vr3, $vr0, $vr3
vilvl.w $vr4, $vr0, $vr4
vinsgr2vr.d $vr2, $a5, 0
vinsgr2vr.d $vr3, $a6, 0
vsllwil.du.wu $vr2, $vr2, 0
vsllwil.du.wu $vr3, $vr3, 0
vadd.d $vr0, $vr0, $vr2
vadd.d $vr1, $vr1, $vr3
vadd.d $vr2, $vr2, $vr4
addi.d $a4, $a4, -4
addi.d $a1, $a1, 16
bnez $a4, .LBB18_5
# %bb.6: # %middle.block
vadd.d $vr0, $vr2, $vr1
vadd.d $vr0, $vr1, $vr0
vhaddw.q.d $vr0, $vr0, $vr0
vpickve2gr.d $a1, $vr0, 0
beq $a2, $a3, .LBB18_9
Expand Down
Loading