; SSE-NEXT: movapd %xmm2, %xmm1
; SSE-NEXT: retq
;
-; AVX1-LABEL: avx_vhadd_pd_test:
-; AVX1: # %bb.0:
-; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
-; AVX1-NEXT: vhaddpd %xmm2, %xmm1, %xmm1
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
-; AVX1-NEXT: vhaddpd %xmm2, %xmm0, %xmm0
-; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-NEXT: retq
-;
-; AVX2-LABEL: avx_vhadd_pd_test:
-; AVX2: # %bb.0:
-; AVX2-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
-; AVX2-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX2-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
-; AVX2-NEXT: retq
+; AVX-LABEL: avx_vhadd_pd_test:
+; AVX: # %bb.0:
+; AVX-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
+; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
+; AVX-NEXT: retq
%vecext = extractelement <4 x double> %A, i32 0
%vecext1 = extractelement <4 x double> %A, i32 1
%add = fadd double %vecext, %vecext1
; SSE-NEXT: movapd %xmm2, %xmm1
; SSE-NEXT: retq
;
-; AVX1-LABEL: avx_vhsub_pd_test:
-; AVX1: # %bb.0:
-; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
-; AVX1-NEXT: vhsubpd %xmm2, %xmm1, %xmm1
-; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm2
-; AVX1-NEXT: vhsubpd %xmm2, %xmm0, %xmm0
-; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-NEXT: retq
-;
-; AVX2-LABEL: avx_vhsub_pd_test:
-; AVX2: # %bb.0:
-; AVX2-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
-; AVX2-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX2-NEXT: vhsubpd %ymm2, %ymm0, %ymm0
-; AVX2-NEXT: retq
+; AVX-LABEL: avx_vhsub_pd_test:
+; AVX: # %bb.0:
+; AVX-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
+; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX-NEXT: vhsubpd %ymm2, %ymm0, %ymm0
+; AVX-NEXT: retq
%vecext = extractelement <4 x double> %A, i32 0
%vecext1 = extractelement <4 x double> %A, i32 1
%sub = fsub double %vecext, %vecext1
;
; AVX1-FAST-LABEL: add_pd_011:
; AVX1-FAST: # %bb.0:
-; AVX1-FAST-NEXT: vhaddpd %xmm0, %xmm1, %xmm2
-; AVX1-FAST-NEXT: vhaddpd %xmm1, %xmm0, %xmm0
-; AVX1-FAST-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-FAST-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm2
+; AVX1-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-FAST-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
; AVX1-FAST-NEXT: retq
;
; AVX512-LABEL: add_pd_011:
; AVX-SLOW-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-SLOW-NEXT: retq
;
-; AVX1-FAST-LABEL: PR34724_add_v4f64_u123:
-; AVX1-FAST: # %bb.0:
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm1, %xmm2
-; AVX1-FAST-NEXT: vhaddpd %xmm2, %xmm1, %xmm1
-; AVX1-FAST-NEXT: vhaddpd %xmm0, %xmm0, %xmm0
-; AVX1-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-FAST-NEXT: retq
-;
-; AVX512-FAST-LABEL: PR34724_add_v4f64_u123:
-; AVX512-FAST: # %bb.0:
-; AVX512-FAST-NEXT: vextractf128 $1, %ymm0, %xmm0
-; AVX512-FAST-NEXT: vblendpd {{.*#+}} ymm2 = ymm0[0,1],ymm1[2,3]
-; AVX512-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512-FAST-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
-; AVX512-FAST-NEXT: retq
+; AVX-FAST-LABEL: PR34724_add_v4f64_u123:
+; AVX-FAST: # %bb.0:
+; AVX-FAST-NEXT: vextractf128 $1, %ymm0, %xmm0
+; AVX-FAST-NEXT: vblendpd {{.*#+}} ymm2 = ymm0[0,1],ymm1[2,3]
+; AVX-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX-FAST-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
+; AVX-FAST-NEXT: retq
%3 = shufflevector <4 x double> %0, <4 x double> %1, <2 x i32> <i32 2, i32 4>
%4 = shufflevector <4 x double> %0, <4 x double> %1, <2 x i32> <i32 3, i32 5>
%5 = fadd <2 x double> %3, %4
; AVX-SLOW-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-SLOW-NEXT: retq
;
-; AVX1-FAST-LABEL: PR34724_add_v4f64_0u23:
-; AVX1-FAST: # %bb.0:
-; AVX1-FAST-NEXT: vhaddpd %xmm1, %xmm0, %xmm0
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm1, %xmm2
-; AVX1-FAST-NEXT: vhaddpd %xmm2, %xmm1, %xmm1
-; AVX1-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-FAST-NEXT: retq
-;
-; AVX512-FAST-LABEL: PR34724_add_v4f64_0u23:
-; AVX512-FAST: # %bb.0:
-; AVX512-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512-FAST-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
-; AVX512-FAST-NEXT: retq
+; AVX-FAST-LABEL: PR34724_add_v4f64_0u23:
+; AVX-FAST: # %bb.0:
+; AVX-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX-FAST-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
+; AVX-FAST-NEXT: retq
%3 = shufflevector <4 x double> %0, <4 x double> %1, <2 x i32> <i32 0, i32 4>
%4 = shufflevector <4 x double> %0, <4 x double> %1, <2 x i32> <i32 1, i32 5>
%5 = fadd <2 x double> %3, %4
; AVX-SLOW-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-SLOW-NEXT: retq
;
-; AVX1-FAST-LABEL: PR34724_add_v4f64_01u3:
-; AVX1-FAST: # %bb.0:
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm0, %xmm2
-; AVX1-FAST-NEXT: vhaddpd %xmm2, %xmm0, %xmm0
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm1, %xmm1
-; AVX1-FAST-NEXT: vhaddpd %xmm1, %xmm1, %xmm1
-; AVX1-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-FAST-NEXT: retq
-;
-; AVX512-FAST-LABEL: PR34724_add_v4f64_01u3:
-; AVX512-FAST: # %bb.0:
-; AVX512-FAST-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
-; AVX512-FAST-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
-; AVX512-FAST-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
-; AVX512-FAST-NEXT: retq
+; AVX-FAST-LABEL: PR34724_add_v4f64_01u3:
+; AVX-FAST: # %bb.0:
+; AVX-FAST-NEXT: vperm2f128 {{.*#+}} ymm2 = ymm0[2,3],ymm1[2,3]
+; AVX-FAST-NEXT: vblendpd {{.*#+}} ymm0 = ymm0[0,1],ymm1[2,3]
+; AVX-FAST-NEXT: vhaddpd %ymm2, %ymm0, %ymm0
+; AVX-FAST-NEXT: retq
%3 = shufflevector <4 x double> %0, <4 x double> undef, <2 x i32> <i32 0, i32 2>
%4 = shufflevector <4 x double> %0, <4 x double> undef, <2 x i32> <i32 1, i32 3>
%5 = fadd <2 x double> %3, %4
; AVX-SLOW-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-SLOW-NEXT: retq
;
-; AVX1-FAST-LABEL: PR34724_add_v4f64_012u:
-; AVX1-FAST: # %bb.0:
-; AVX1-FAST-NEXT: vextractf128 $1, %ymm0, %xmm2
-; AVX1-FAST-NEXT: vhaddpd %xmm2, %xmm0, %xmm0
-; AVX1-FAST-NEXT: vhaddpd %xmm1, %xmm1, %xmm1
-; AVX1-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX1-FAST-NEXT: retq
-;
-; AVX512-FAST-LABEL: PR34724_add_v4f64_012u:
-; AVX512-FAST: # %bb.0:
-; AVX512-FAST-NEXT: vextractf128 $1, %ymm0, %xmm2
-; AVX512-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
-; AVX512-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm2, %ymm1
-; AVX512-FAST-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
-; AVX512-FAST-NEXT: retq
+; AVX-FAST-LABEL: PR34724_add_v4f64_012u:
+; AVX-FAST: # %bb.0:
+; AVX-FAST-NEXT: vextractf128 $1, %ymm0, %xmm2
+; AVX-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX-FAST-NEXT: vinsertf128 $1, %xmm1, %ymm2, %ymm1
+; AVX-FAST-NEXT: vhaddpd %ymm1, %ymm0, %ymm0
+; AVX-FAST-NEXT: retq
%3 = shufflevector <4 x double> %0, <4 x double> undef, <2 x i32> <i32 0, i32 2>
%4 = shufflevector <4 x double> %0, <4 x double> undef, <2 x i32> <i32 1, i32 3>
%5 = fadd <2 x double> %3, %4