Page MenuHomePhabricator
Differential D113782 Diff 391036 llvm/test/Transforms/SLPVectorizer/X86/cmp_commute.ll

Changeset View

Standalone View

View Options

llvm/test/Transforms/SLPVectorizer/X86/cmp_commute.ll

Show First 20 Lines • Show All 230 Lines • ▼ Show 20 Lines;
%d2 = insertelement <4 x i1> %d1, i1 %c2, i32 2 %d2 = insertelement <4 x i1> %d1, i1 %c2, i32 2
%d3 = insertelement <4 x i1> %d2, i1 %c3, i32 3 %d3 = insertelement <4 x i1> %d2, i1 %c3, i32 3
%r = sext <4 x i1> %d3 to <4 x i32> %r = sext <4 x i1> %d3 to <4 x i32>
ret <4 x i32> %r ret <4 x i32> %r
} }
define <4 x i32> @fcmp_ord_uno_v4i32(<4 x float> %a, float* %b) { define <4 x i32> @fcmp_ord_uno_v4i32(<4 x float> %a, float* %b) {
; CHECK-LABEL: @fcmp_ord_uno_v4i32( ; CHECK-LABEL: @fcmp_ord_uno_v4i32(
; CHECK-NEXT: [[A0:%.*]] = extractelement <4 x float> [[A:%.*]], i32 0
; CHECK-NEXT: [[A3:%.*]] = extractelement <4 x float> [[A]], i32 3
; CHECK-NEXT: [[P1:%.*]] = getelementptr inbounds float, float* [[B:%.*]], i64 1 ; CHECK-NEXT: [[P1:%.*]] = getelementptr inbounds float, float* [[B:%.*]], i64 1
; CHECK-NEXT: [[P3:%.*]] = getelementptr inbounds float, float* [[B]], i64 3 ; CHECK-NEXT: [[P3:%.*]] = getelementptr inbounds float, float* [[B]], i64 3
; CHECK-NEXT: [[B0:%.*]] = load float, float* [[B]], align 4 ; CHECK-NEXT: [[B0:%.*]] = load float, float* [[B]], align 4
; CHECK-NEXT: [[TMP1:%.*]] = bitcast float* [[P1]] to <2 x float>* ; CHECK-NEXT: [[TMP1:%.*]] = bitcast float* [[P1]] to <2 x float>*
; CHECK-NEXT: [[TMP2:%.*]] = load <2 x float>, <2 x float>* [[TMP1]], align 4 ; CHECK-NEXT: [[TMP2:%.*]] = load <2 x float>, <2 x float>* [[TMP1]], align 4
; CHECK-NEXT: [[B3:%.*]] = load float, float* [[P3]], align 4 ; CHECK-NEXT: [[B3:%.*]] = load float, float* [[P3]], align 4
; CHECK-NEXT: [[C0:%.*]] = fcmp ord float [[A0]], [[B0]] ; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x float> [[A:%.*]], <4 x float> undef, <2 x i32> <i32 1, i32 2>
; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x float> [[A]], <4 x float> undef, <2 x i32> <i32 1, i32 2>
; CHECK-NEXT: [[TMP4:%.*]] = fcmp uno <2 x float> [[TMP2]], [[TMP3]] ; CHECK-NEXT: [[TMP4:%.*]] = fcmp uno <2 x float> [[TMP2]], [[TMP3]]
; CHECK-NEXT: [[C3:%.*]] = fcmp ord float [[A3]], [[B3]] ; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <4 x float> [[A]], <4 x float> undef, <2 x i32> <i32 3, i32 0>
; CHECK-NEXT: [[D0:%.*]] = insertelement <4 x i1> undef, i1 [[C0]], i32 0 ; CHECK-NEXT: [[TMP6:%.*]] = insertelement <2 x float> poison, float [[B3]], i32 0
; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <2 x i1> [[TMP4]], <2 x i1> poison, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef> ; CHECK-NEXT: [[TMP7:%.*]] = insertelement <2 x float> [[TMP6]], float [[B0]], i32 1
; CHECK-NEXT: [[D21:%.*]] = shufflevector <4 x i1> [[D0]], <4 x i1> [[TMP5]], <4 x i32> <i32 0, i32 4, i32 5, i32 undef> ; CHECK-NEXT: [[TMP8:%.*]] = fcmp ord <2 x float> [[TMP5]], [[TMP7]]
; CHECK-NEXT: [[D3:%.*]] = insertelement <4 x i1> [[D21]], i1 [[C3]], i32 3 ; CHECK-NEXT: [[TMP9:%.*]] = shufflevector <2 x i1> [[TMP8]], <2 x i1> poison, <4 x i32> <i32 0, i32 undef, i32 undef, i32 undef>
; CHECK-NEXT: [[D0:%.*]] = shufflevector <2 x i1> [[TMP8]], <2 x i1> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
; CHECK-NEXT: [[TMP10:%.*]] = shufflevector <2 x i1> [[TMP4]], <2 x i1> poison, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
; CHECK-NEXT: [[D21:%.*]] = shufflevector <4 x i1> [[D0]], <4 x i1> [[TMP10]], <4 x i32> <i32 0, i32 4, i32 5, i32 undef>
; CHECK-NEXT: [[D3:%.*]] = shufflevector <4 x i1> [[D21]], <4 x i1> [[TMP9]], <4 x i32> <i32 0, i32 1, i32 2, i32 4>
; CHECK-NEXT: [[R:%.*]] = sext <4 x i1> [[D3]] to <4 x i32> ; CHECK-NEXT: [[R:%.*]] = sext <4 x i1> [[D3]] to <4 x i32>
; CHECK-NEXT: ret <4 x i32> [[R]] ; CHECK-NEXT: ret <4 x i32> [[R]]
; ;
%a0 = extractelement <4 x float> %a, i32 0 %a0 = extractelement <4 x float> %a, i32 0
%a1 = extractelement <4 x float> %a, i32 1 %a1 = extractelement <4 x float> %a, i32 1
%a2 = extractelement <4 x float> %a, i32 2 %a2 = extractelement <4 x float> %a, i32 2
%a3 = extractelement <4 x float> %a, i32 3 %a3 = extractelement <4 x float> %a, i32 3
%p0 = getelementptr inbounds float, float* %b, i32 0 %p0 = getelementptr inbounds float, float* %b, i32 0
Show All 18 Lines