[llvm] r343965 - [x86] add 16 missed hadd patterns (PR39195); NFC
Sanjay Patel via llvm-commits
llvm-commits at lists.llvm.org
Mon Oct 8 05:54:34 PDT 2018
Author: spatel
Date: Mon Oct 8 05:54:33 2018
New Revision: 343965
URL: http://llvm.org/viewvc/llvm-project?rev=343965&view=rev
Log:
[x86] add 16 missed hadd patterns (PR39195); NFC
Modified:
llvm/trunk/test/CodeGen/X86/haddsub-undef.ll
Modified: llvm/trunk/test/CodeGen/X86/haddsub-undef.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/haddsub-undef.ll?rev=343965&r1=343964&r2=343965&view=diff
==============================================================================
--- llvm/trunk/test/CodeGen/X86/haddsub-undef.ll (original)
+++ llvm/trunk/test/CodeGen/X86/haddsub-undef.ll Mon Oct 8 05:54:33 2018
@@ -449,3 +449,330 @@ define <8 x i32> @test17_undef(<8 x i32>
%vecinit4 = insertelement <8 x i32> %vecinit3, i32 %add4, i32 3
ret <8 x i32> %vecinit4
}
+
+define <2 x double> @add_pd_003(<2 x double> %a, <2 x double> %b) {
+; SSE-LABEL: add_pd_003:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: addpd %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_pd_003:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vaddpd %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %shuffle = shufflevector <2 x double> %b, <2 x double> undef, <2 x i32> <i32 undef, i32 0>
+ %add = fadd <2 x double> %shuffle, %b
+ ret <2 x double> %add
+}
+
+define <2 x double> @add_pd_005(<2 x double> %a, <2 x double> %b) {
+; SSE-LABEL: add_pd_005:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: addpd %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_pd_005:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vaddpd %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %vecinit2 = shufflevector <2 x double> %b, <2 x double> undef, <2 x i32> <i32 undef, i32 0>
+ %add = fadd <2 x double> %vecinit2, %b
+ ret <2 x double> %add
+}
+
+define <2 x double> @add_pd_010(<2 x double> %a, <2 x double> %b) {
+; SSE-LABEL: add_pd_010:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: addpd %xmm1, %xmm0
+; SSE-NEXT: unpckhpd {{.*#+}} xmm0 = xmm0[1,1]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_pd_010:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vaddpd %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
+; AVX-NEXT: retq
+ %shuffle = shufflevector <2 x double> %b, <2 x double> undef, <2 x i32> <i32 undef, i32 0>
+ %add = fadd <2 x double> %shuffle, %b
+ %shuffle2 = shufflevector <2 x double> %add, <2 x double> undef, <2 x i32> <i32 1, i32 undef>
+ ret <2 x double> %shuffle2
+}
+
+define <2 x double> @add_pd_012(<2 x double> %a, <2 x double> %b) {
+; SSE-LABEL: add_pd_012:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: addpd %xmm1, %xmm0
+; SSE-NEXT: unpckhpd {{.*#+}} xmm0 = xmm0[1,1]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_pd_012:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vaddpd %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
+; AVX-NEXT: retq
+ %vecinit2 = shufflevector <2 x double> %b, <2 x double> undef, <2 x i32> <i32 undef, i32 0>
+ %add = fadd <2 x double> %vecinit2, %b
+ %shuffle = shufflevector <2 x double> %add, <2 x double> undef, <2 x i32> <i32 1, i32 undef>
+ ret <2 x double> %shuffle
+}
+
+define <4 x float> @add_ps_007(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_007:
+; SSE: # %bb.0:
+; SSE-NEXT: movaps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,0,2]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_007:
+; AVX: # %bb.0:
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm1[0,1,0,2]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 2>
+ %shuffle1 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 3>
+ %add = fadd <4 x float> %shuffle, %shuffle1
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_013(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_013:
+; SSE: # %bb.0:
+; SSE-NEXT: movaps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,0,2]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_013:
+; AVX: # %bb.0:
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm1[0,1,0,2]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 2>
+ %vecinit15 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 3>
+ %add = fadd <4 x float> %vecinit6, %vecinit15
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_030(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_030:
+; SSE: # %bb.0:
+; SSE-NEXT: movaps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,0,2]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[3,2,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_030:
+; AVX: # %bb.0:
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm1[0,1,0,2]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,2,3]
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 2>
+ %shuffle1 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 3>
+ %add = fadd <4 x float> %shuffle, %shuffle1
+ %shuffle2 = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 3, i32 2, i32 undef, i32 undef>
+ ret <4 x float> %shuffle2
+}
+
+define <4 x float> @add_ps_036(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_036:
+; SSE: # %bb.0:
+; SSE-NEXT: movaps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,0,2]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[3,2,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_036:
+; AVX: # %bb.0:
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm1[0,1,0,2]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,2,2,3]
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 2>
+ %vecinit15 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 3>
+ %add = fadd <4 x float> %vecinit6, %vecinit15
+ %shuffle = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 3, i32 2, i32 undef, i32 undef>
+ ret <4 x float> %shuffle
+}
+
+define <4 x float> @add_ps_007_2(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_007_2:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_007_2:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 undef>
+ %shuffle1 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 undef>
+ %add = fadd <4 x float> %shuffle, %shuffle1
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_008(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_008:
+; SSE: # %bb.0:
+; SSE-NEXT: movsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_008:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 undef, i32 2>
+ %add = fadd <4 x float> %shuffle, %b
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_011(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_011:
+; SSE: # %bb.0:
+; SSE-NEXT: movddup {{.*#+}} xmm0 = xmm1[0,0]
+; SSE-NEXT: shufps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_011:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 undef>
+ %vecinit15 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 undef>
+ %add = fadd <4 x float> %vecinit6, %vecinit15
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_012(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_012:
+; SSE: # %bb.0:
+; SSE-NEXT: movsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_012:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 undef, i32 2>
+ %add = fadd <4 x float> %vecinit6, %b
+ ret <4 x float> %add
+}
+
+define <4 x float> @add_ps_017(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_017:
+; SSE: # %bb.0:
+; SSE-NEXT: movsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[3,1,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_017:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 undef, i32 2>
+ %add = fadd <4 x float> %shuffle, %b
+ %shuffle2 = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 3, i32 undef, i32 undef, i32 undef>
+ ret <4 x float> %shuffle2
+}
+
+define <4 x float> @add_ps_018(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_018:
+; SSE: # %bb.0:
+; SSE-NEXT: movapd %xmm1, %xmm0
+; SSE-NEXT: movddup {{.*#+}} xmm1 = xmm1[0,0]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_018:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; AVX-NEXT: retq
+ %shuffle = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 undef>
+ %shuffle1 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 undef>
+ %add = fadd <4 x float> %shuffle, %shuffle1
+ %shuffle2 = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 undef, i32 2, i32 undef, i32 undef>
+ ret <4 x float> %shuffle2
+}
+
+define <4 x float> @add_ps_021(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_021:
+; SSE: # %bb.0:
+; SSE-NEXT: movsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[3,1,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_021:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovsldup {{.*#+}} xmm0 = xmm1[0,0,2,2]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 undef, i32 2>
+ %add = fadd <4 x float> %vecinit6, %b
+ %shuffle = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 3, i32 undef, i32 undef, i32 undef>
+ ret <4 x float> %shuffle
+}
+
+define <4 x float> @add_ps_022(<4 x float> %a, <4 x float> %b) {
+; SSE-LABEL: add_ps_022:
+; SSE: # %bb.0:
+; SSE-NEXT: movapd %xmm1, %xmm0
+; SSE-NEXT: movddup {{.*#+}} xmm1 = xmm1[0,0]
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,1,1,3]
+; SSE-NEXT: addps %xmm1, %xmm0
+; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; SSE-NEXT: retq
+;
+; AVX-LABEL: add_ps_022:
+; AVX: # %bb.0:
+; AVX-NEXT: vmovddup {{.*#+}} xmm0 = xmm1[0,0]
+; AVX-NEXT: vpermilps {{.*#+}} xmm1 = xmm1[0,1,1,3]
+; AVX-NEXT: vaddps %xmm1, %xmm0, %xmm0
+; AVX-NEXT: vpermilps {{.*#+}} xmm0 = xmm0[0,2,2,3]
+; AVX-NEXT: retq
+ %vecinit6 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 0, i32 undef>
+ %vecinit15 = shufflevector <4 x float> %b, <4 x float> undef, <4 x i32> <i32 undef, i32 undef, i32 1, i32 undef>
+ %add = fadd <4 x float> %vecinit6, %vecinit15
+ %shuffle = shufflevector <4 x float> %add, <4 x float> undef, <4 x i32> <i32 undef, i32 2, i32 undef, i32 undef>
+ ret <4 x float> %shuffle
+}
+
More information about the llvm-commits
mailing list