1; RUN: llc < %s -march=arm -mattr=+neon | FileCheck %s 2 3define <8 x i8> @vzipi8(<8 x i8>* %A, <8 x i8>* %B) nounwind { 4;CHECK: vzipi8: 5;CHECK: vzip.8 6;CHECK-NEXT: vadd.i8 7 %tmp1 = load <8 x i8>* %A 8 %tmp2 = load <8 x i8>* %B 9 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> 10 %tmp4 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> 11 %tmp5 = add <8 x i8> %tmp3, %tmp4 12 ret <8 x i8> %tmp5 13} 14 15define <4 x i16> @vzipi16(<4 x i16>* %A, <4 x i16>* %B) nounwind { 16;CHECK: vzipi16: 17;CHECK: vzip.16 18;CHECK-NEXT: vadd.i16 19 %tmp1 = load <4 x i16>* %A 20 %tmp2 = load <4 x i16>* %B 21 %tmp3 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5> 22 %tmp4 = shufflevector <4 x i16> %tmp1, <4 x i16> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7> 23 %tmp5 = add <4 x i16> %tmp3, %tmp4 24 ret <4 x i16> %tmp5 25} 26 27; VZIP.32 is equivalent to VTRN.32 for 64-bit vectors. 28 29define <16 x i8> @vzipQi8(<16 x i8>* %A, <16 x i8>* %B) nounwind { 30;CHECK: vzipQi8: 31;CHECK: vzip.8 32;CHECK-NEXT: vadd.i8 33 %tmp1 = load <16 x i8>* %A 34 %tmp2 = load <16 x i8>* %B 35 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 0, i32 16, i32 1, i32 17, i32 2, i32 18, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> 36 %tmp4 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 8, i32 24, i32 9, i32 25, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 29, i32 14, i32 30, i32 15, i32 31> 37 %tmp5 = add <16 x i8> %tmp3, %tmp4 38 ret <16 x i8> %tmp5 39} 40 41define <8 x i16> @vzipQi16(<8 x i16>* %A, <8 x i16>* %B) nounwind { 42;CHECK: vzipQi16: 43;CHECK: vzip.16 44;CHECK-NEXT: vadd.i16 45 %tmp1 = load <8 x i16>* %A 46 %tmp2 = load <8 x i16>* %B 47 %tmp3 = shufflevector <8 x i16> %tmp1, <8 x i16> %tmp2, <8 x i32> <i32 0, i32 8, i32 1, i32 9, i32 2, i32 10, i32 3, i32 11> 48 %tmp4 = shufflevector <8 x i16> %tmp1, <8 x i16> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 14, i32 7, i32 15> 49 %tmp5 = add <8 x i16> %tmp3, %tmp4 50 ret <8 x i16> %tmp5 51} 52 53define <4 x i32> @vzipQi32(<4 x i32>* %A, <4 x i32>* %B) nounwind { 54;CHECK: vzipQi32: 55;CHECK: vzip.32 56;CHECK-NEXT: vadd.i32 57 %tmp1 = load <4 x i32>* %A 58 %tmp2 = load <4 x i32>* %B 59 %tmp3 = shufflevector <4 x i32> %tmp1, <4 x i32> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5> 60 %tmp4 = shufflevector <4 x i32> %tmp1, <4 x i32> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7> 61 %tmp5 = add <4 x i32> %tmp3, %tmp4 62 ret <4 x i32> %tmp5 63} 64 65define <4 x float> @vzipQf(<4 x float>* %A, <4 x float>* %B) nounwind { 66;CHECK: vzipQf: 67;CHECK: vzip.32 68;CHECK-NEXT: vadd.f32 69 %tmp1 = load <4 x float>* %A 70 %tmp2 = load <4 x float>* %B 71 %tmp3 = shufflevector <4 x float> %tmp1, <4 x float> %tmp2, <4 x i32> <i32 0, i32 4, i32 1, i32 5> 72 %tmp4 = shufflevector <4 x float> %tmp1, <4 x float> %tmp2, <4 x i32> <i32 2, i32 6, i32 3, i32 7> 73 %tmp5 = fadd <4 x float> %tmp3, %tmp4 74 ret <4 x float> %tmp5 75} 76 77; Undef shuffle indices should not prevent matching to VZIP: 78 79define <8 x i8> @vzipi8_undef(<8 x i8>* %A, <8 x i8>* %B) nounwind { 80;CHECK: vzipi8_undef: 81;CHECK: vzip.8 82;CHECK-NEXT: vadd.i8 83 %tmp1 = load <8 x i8>* %A 84 %tmp2 = load <8 x i8>* %B 85 %tmp3 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 0, i32 undef, i32 1, i32 9, i32 undef, i32 10, i32 3, i32 11> 86 %tmp4 = shufflevector <8 x i8> %tmp1, <8 x i8> %tmp2, <8 x i32> <i32 4, i32 12, i32 5, i32 13, i32 6, i32 undef, i32 undef, i32 15> 87 %tmp5 = add <8 x i8> %tmp3, %tmp4 88 ret <8 x i8> %tmp5 89} 90 91define <16 x i8> @vzipQi8_undef(<16 x i8>* %A, <16 x i8>* %B) nounwind { 92;CHECK: vzipQi8_undef: 93;CHECK: vzip.8 94;CHECK-NEXT: vadd.i8 95 %tmp1 = load <16 x i8>* %A 96 %tmp2 = load <16 x i8>* %B 97 %tmp3 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 0, i32 16, i32 1, i32 undef, i32 undef, i32 undef, i32 3, i32 19, i32 4, i32 20, i32 5, i32 21, i32 6, i32 22, i32 7, i32 23> 98 %tmp4 = shufflevector <16 x i8> %tmp1, <16 x i8> %tmp2, <16 x i32> <i32 8, i32 24, i32 9, i32 undef, i32 10, i32 26, i32 11, i32 27, i32 12, i32 28, i32 13, i32 undef, i32 14, i32 30, i32 undef, i32 31> 99 %tmp5 = add <16 x i8> %tmp3, %tmp4 100 ret <16 x i8> %tmp5 101} 102 103