1; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx | FileCheck %s 2 3; Check that we perform a scalar XOR on i32. 4 5; CHECK: pull_bitcast 6; CHECK: xorl 7; CHECK: ret 8define void @pull_bitcast (<4 x i8>* %pA, <4 x i8>* %pB) { 9 %A = load <4 x i8>* %pA 10 %B = load <4 x i8>* %pB 11 %C = xor <4 x i8> %A, %B 12 store <4 x i8> %C, <4 x i8>* %pA 13 ret void 14} 15 16; CHECK: multi_use_swizzle 17; CHECK: mov 18; CHECK-NEXT: shuf 19; CHECK-NEXT: shuf 20; CHECK-NEXT: shuf 21; CHECK-NEXT: xor 22; CHECK-NEXT: ret 23define <4 x i32> @multi_use_swizzle (<4 x i32>* %pA, <4 x i32>* %pB) { 24 %A = load <4 x i32>* %pA 25 %B = load <4 x i32>* %pB 26 %S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 1, i32 5, i32 6> 27 %S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 1, i32 3, i32 2, i32 2> 28 %S2 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 2, i32 1, i32 0, i32 2> 29 %R = xor <4 x i32> %S1, %S2 30 ret <4 x i32> %R 31} 32 33; CHECK: pull_bitcast2 34; CHECK: xorl 35; CHECK: ret 36define <4 x i8> @pull_bitcast2 (<4 x i8>* %pA, <4 x i8>* %pB, <4 x i8>* %pC) { 37 %A = load <4 x i8>* %pA 38 store <4 x i8> %A, <4 x i8>* %pC 39 %B = load <4 x i8>* %pB 40 %C = xor <4 x i8> %A, %B 41 store <4 x i8> %C, <4 x i8>* %pA 42 ret <4 x i8> %C 43} 44 45 46 47; CHECK: reverse_1 48; CHECK-NOT: shuf 49; CHECK: ret 50define <4 x i32> @reverse_1 (<4 x i32>* %pA, <4 x i32>* %pB) { 51 %A = load <4 x i32>* %pA 52 %B = load <4 x i32>* %pB 53 %S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 0, i32 3, i32 2> 54 %S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 1, i32 0, i32 3, i32 2> 55 ret <4 x i32> %S1 56} 57 58 59; CHECK: no_reverse_shuff 60; CHECK: shuf 61; CHECK: ret 62define <4 x i32> @no_reverse_shuff (<4 x i32>* %pA, <4 x i32>* %pB) { 63 %A = load <4 x i32>* %pA 64 %B = load <4 x i32>* %pB 65 %S = shufflevector <4 x i32> %A, <4 x i32> %B, <4 x i32> <i32 1, i32 0, i32 3, i32 2> 66 %S1 = shufflevector <4 x i32> %S, <4 x i32> undef, <4 x i32> <i32 3, i32 2, i32 3, i32 2> 67 ret <4 x i32> %S1 68} 69