c7657cf7d1
If the vectorized GEP instruction can be still kept as a scalar GEP,
better to keep it as scalar instead of extractelement. In many cases it
is more profitable.
Metric: size..text
Program size..text
results results0 diff
test-suite :: SingleSource/Benchmarks/Misc/oourafft.test 18911.00 19695.00 4.1%
test-suite :: SingleSource/Benchmarks/Misc-C++-EH/spirit.test 59987.00 60707.00 1.2%
test-suite :: External/SPEC/CFP2017speed/638.imagick_s/638.imagick_s.test 1392209.00 1392753.00 0.0%
test-suite :: External/SPEC/CFP2017rate/538.imagick_r/538.imagick_r.test 1392209.00 1392753.00 0.0%
test-suite :: External/SPEC/CINT2006/400.perlbench/400.perlbench.test 1087996.00 1088236.00 0.0%
test-suite :: MultiSource/Benchmarks/Bullet/bullet.test 309310.00 309342.00 0.0%
test-suite :: External/SPEC/CINT2017rate/525.x264_r/525.x264_r.test 664661.00 664693.00 0.0%
test-suite :: External/SPEC/CINT2017speed/625.x264_s/625.x264_s.test 664661.00 664693.00 0.0%
test-suite :: External/SPEC/CFP2017rate/526.blender_r/526.blender_r.test 12354636.00 12354908.00 0.0%
test-suite :: External/SPEC/CFP2006/453.povray/453.povray.test 1152748.00 1152716.00 -0.0%
test-suite :: MultiSource/Applications/oggenc/oggenc.test 191787.00 191771.00 -0.0%
test-suite :: SingleSource/UnitTests/matrix-types-spec.test 480796.00 480476.00 -0.1%
Misc/oourafft - Extra code gets vectorized
Misc-C++-EH/spirit - same
CFP2017speed/638.imagick_s
CFP2017rate/538.imagick_r - same, extra code gets vectorized
CINT2006/400.perlbench - some extra 4 x ptr stores vectorized
Bullet/bullet - extra 4 x ptr store vectorized
CINT2017rate/525.x264_r
CINT2017speed/625.x264_s - same
CFP2017rate/526.blender_r - extra 8 x float stores (several), some extra
4 x ptr stores
CFP2006/453.povray - 2 x double loads/stores replaced by 4 x double
loads/stores
Applications/oggenc - extra code is vectorized
UnitTests/matrix-types-spec - extra code gets vectorized
Reviewers: RKSimon
Reviewed By: RKSimon
Pull Request: https://github.com/llvm/llvm-project/pull/88877
78 lines
3.3 KiB
LLVM
78 lines
3.3 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
|
|
; RUN: opt -S -passes=slp-vectorizer -mtriple=x86_64-apple-macosx -mcpu=haswell < %s | FileCheck %s
|
|
|
|
define void @test(ptr %r, ptr %p, ptr %q) #0 {
|
|
; CHECK-LABEL: @test(
|
|
; CHECK-NEXT: [[P0:%.*]] = getelementptr inbounds i64, ptr [[P:%.*]], i64 0
|
|
; CHECK-NEXT: [[Q0:%.*]] = getelementptr inbounds i64, ptr [[Q:%.*]], i64 0
|
|
; CHECK-NEXT: [[TMP1:%.*]] = load <4 x i64>, ptr [[P0]], align 2
|
|
; CHECK-NEXT: [[TMP2:%.*]] = load <4 x i64>, ptr [[Q0]], align 2
|
|
; CHECK-NEXT: [[TMP3:%.*]] = sub nsw <4 x i64> [[TMP1]], [[TMP2]]
|
|
; CHECK-NEXT: [[TMP4:%.*]] = extractelement <4 x i64> [[TMP3]], i32 0
|
|
; CHECK-NEXT: [[G0:%.*]] = getelementptr inbounds i32, ptr [[R:%.*]], i64 [[TMP4]]
|
|
; CHECK-NEXT: [[TMP5:%.*]] = extractelement <4 x i64> [[TMP3]], i32 1
|
|
; CHECK-NEXT: [[G1:%.*]] = getelementptr inbounds i32, ptr [[R]], i64 [[TMP5]]
|
|
; CHECK-NEXT: [[TMP6:%.*]] = extractelement <4 x i64> [[TMP3]], i32 2
|
|
; CHECK-NEXT: [[G2:%.*]] = getelementptr inbounds i32, ptr [[R]], i64 [[TMP6]]
|
|
; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i64> [[TMP3]], i32 3
|
|
; CHECK-NEXT: [[G3:%.*]] = getelementptr inbounds i32, ptr [[R]], i64 [[TMP7]]
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
%p0 = getelementptr inbounds i64, ptr %p, i64 0
|
|
%p1 = getelementptr inbounds i64, ptr %p, i64 1
|
|
%p2 = getelementptr inbounds i64, ptr %p, i64 2
|
|
%p3 = getelementptr inbounds i64, ptr %p, i64 3
|
|
|
|
%q0 = getelementptr inbounds i64, ptr %q, i64 0
|
|
%q1 = getelementptr inbounds i64, ptr %q, i64 1
|
|
%q2 = getelementptr inbounds i64, ptr %q, i64 2
|
|
%q3 = getelementptr inbounds i64, ptr %q, i64 3
|
|
|
|
%x0 = load i64, ptr %p0, align 2
|
|
%x1 = load i64, ptr %p1, align 2
|
|
%x2 = load i64, ptr %p2, align 2
|
|
%x3 = load i64, ptr %p3, align 2
|
|
|
|
%y0 = load i64, ptr %q0, align 2
|
|
%y1 = load i64, ptr %q1, align 2
|
|
%y2 = load i64, ptr %q2, align 2
|
|
%y3 = load i64, ptr %q3, align 2
|
|
|
|
%sub0 = sub nsw i64 %x0, %y0
|
|
%sub1 = sub nsw i64 %x1, %y1
|
|
%sub2 = sub nsw i64 %x2, %y2
|
|
%sub3 = sub nsw i64 %x3, %y3
|
|
|
|
%g0 = getelementptr inbounds i32, ptr %r, i64 %sub0
|
|
%g1 = getelementptr inbounds i32, ptr %r, i64 %sub1
|
|
%g2 = getelementptr inbounds i32, ptr %r, i64 %sub2
|
|
%g3 = getelementptr inbounds i32, ptr %r, i64 %sub3
|
|
ret void
|
|
}
|
|
|
|
define void @test2(ptr %a, ptr %b) {
|
|
; CHECK-LABEL: @test2(
|
|
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <2 x ptr> poison, ptr [[A:%.*]], i32 0
|
|
; CHECK-NEXT: [[TMP2:%.*]] = insertelement <2 x ptr> [[TMP1]], ptr [[B:%.*]], i32 1
|
|
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr i64, <2 x ptr> [[TMP2]], <2 x i64> <i64 1, i64 3>
|
|
; CHECK-NEXT: [[A1:%.*]] = getelementptr inbounds i64, ptr [[A]], i64 1
|
|
; CHECK-NEXT: [[TMP4:%.*]] = ptrtoint <2 x ptr> [[TMP3]] to <2 x i64>
|
|
; CHECK-NEXT: [[TMP5:%.*]] = load <2 x i64>, ptr [[A1]], align 8
|
|
; CHECK-NEXT: [[TMP6:%.*]] = add <2 x i64> [[TMP4]], [[TMP5]]
|
|
; CHECK-NEXT: store <2 x i64> [[TMP6]], ptr [[A1]], align 8
|
|
; CHECK-NEXT: ret void
|
|
;
|
|
%a1 = getelementptr inbounds i64, ptr %a, i64 1
|
|
%a2 = getelementptr inbounds i64, ptr %a, i64 2
|
|
%i1 = ptrtoint ptr %a1 to i64
|
|
%b3 = getelementptr inbounds i64, ptr %b, i64 3
|
|
%i2 = ptrtoint ptr %b3 to i64
|
|
%v1 = load i64, ptr %a1, align 8
|
|
%v2 = load i64, ptr %a2, align 8
|
|
%add1 = add i64 %i1, %v1
|
|
%add2 = add i64 %i2, %v2
|
|
store i64 %add1, ptr %a1, align 8
|
|
store i64 %add2, ptr %a2, align 8
|
|
ret void
|
|
}
|