1 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx | FileCheck %s
4 define <4 x double> @andpd256(<4 x double> %y, <4 x double> %x) nounwind uwtable readnone ssp {
6 %0 = bitcast <4 x double> %x to <4 x i64>
7 %1 = bitcast <4 x double> %y to <4 x i64>
8 %and.i = and <4 x i64> %0, %1
9 %2 = bitcast <4 x i64> %and.i to <4 x double>
13 ; CHECK: vandpd LCP{{.*}}(%rip)
14 define <4 x double> @andpd256fold(<4 x double> %y) nounwind uwtable readnone ssp {
16 %0 = bitcast <4 x double> %y to <4 x i64>
17 %and.i = and <4 x i64> %0, <i64 4616752568008179712, i64 4614838538166547251, i64 4612361558371493478, i64 4608083138725491507>
18 %1 = bitcast <4 x i64> %and.i to <4 x double>
23 define <8 x float> @andps256(<8 x float> %y, <8 x float> %x) nounwind uwtable readnone ssp {
25 %0 = bitcast <8 x float> %x to <8 x i32>
26 %1 = bitcast <8 x float> %y to <8 x i32>
27 %and.i = and <8 x i32> %0, %1
28 %2 = bitcast <8 x i32> %and.i to <8 x float>
32 ; CHECK: vandps LCP{{.*}}(%rip)
33 define <8 x float> @andps256fold(<8 x float> %y) nounwind uwtable readnone ssp {
35 %0 = bitcast <8 x float> %y to <8 x i32>
36 %and.i = and <8 x i32> %0, <i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938, i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938>
37 %1 = bitcast <8 x i32> %and.i to <8 x float>
42 define <4 x double> @xorpd256(<4 x double> %y, <4 x double> %x) nounwind uwtable readnone ssp {
44 %0 = bitcast <4 x double> %x to <4 x i64>
45 %1 = bitcast <4 x double> %y to <4 x i64>
46 %xor.i = xor <4 x i64> %0, %1
47 %2 = bitcast <4 x i64> %xor.i to <4 x double>
51 ; CHECK: vxorpd LCP{{.*}}(%rip)
52 define <4 x double> @xorpd256fold(<4 x double> %y) nounwind uwtable readnone ssp {
54 %0 = bitcast <4 x double> %y to <4 x i64>
55 %xor.i = xor <4 x i64> %0, <i64 4616752568008179712, i64 4614838538166547251, i64 4612361558371493478, i64 4608083138725491507>
56 %1 = bitcast <4 x i64> %xor.i to <4 x double>
61 define <8 x float> @xorps256(<8 x float> %y, <8 x float> %x) nounwind uwtable readnone ssp {
63 %0 = bitcast <8 x float> %x to <8 x i32>
64 %1 = bitcast <8 x float> %y to <8 x i32>
65 %xor.i = xor <8 x i32> %0, %1
66 %2 = bitcast <8 x i32> %xor.i to <8 x float>
70 ; CHECK: vxorps LCP{{.*}}(%rip)
71 define <8 x float> @xorps256fold(<8 x float> %y) nounwind uwtable readnone ssp {
73 %0 = bitcast <8 x float> %y to <8 x i32>
74 %xor.i = xor <8 x i32> %0, <i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938, i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938>
75 %1 = bitcast <8 x i32> %xor.i to <8 x float>
80 define <4 x double> @orpd256(<4 x double> %y, <4 x double> %x) nounwind uwtable readnone ssp {
82 %0 = bitcast <4 x double> %x to <4 x i64>
83 %1 = bitcast <4 x double> %y to <4 x i64>
84 %or.i = or <4 x i64> %0, %1
85 %2 = bitcast <4 x i64> %or.i to <4 x double>
89 ; CHECK: vorpd LCP{{.*}}(%rip)
90 define <4 x double> @orpd256fold(<4 x double> %y) nounwind uwtable readnone ssp {
92 %0 = bitcast <4 x double> %y to <4 x i64>
93 %or.i = or <4 x i64> %0, <i64 4616752568008179712, i64 4614838538166547251, i64 4612361558371493478, i64 4608083138725491507>
94 %1 = bitcast <4 x i64> %or.i to <4 x double>
99 define <8 x float> @orps256(<8 x float> %y, <8 x float> %x) nounwind uwtable readnone ssp {
101 %0 = bitcast <8 x float> %x to <8 x i32>
102 %1 = bitcast <8 x float> %y to <8 x i32>
103 %or.i = or <8 x i32> %0, %1
104 %2 = bitcast <8 x i32> %or.i to <8 x float>
108 ; CHECK: vorps LCP{{.*}}(%rip)
109 define <8 x float> @orps256fold(<8 x float> %y) nounwind uwtable readnone ssp {
111 %0 = bitcast <8 x float> %y to <8 x i32>
112 %or.i = or <8 x i32> %0, <i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938, i32 1083179008, i32 1079613850, i32 1075000115, i32 1067030938>
113 %1 = bitcast <8 x i32> %or.i to <8 x float>
118 define <4 x double> @andnotpd256(<4 x double> %y, <4 x double> %x) nounwind uwtable readnone ssp {
120 %0 = bitcast <4 x double> %x to <4 x i64>
121 %neg.i = xor <4 x i64> %0, <i64 -1, i64 -1, i64 -1, i64 -1>
122 %1 = bitcast <4 x double> %y to <4 x i64>
123 %and.i = and <4 x i64> %1, %neg.i
124 %2 = bitcast <4 x i64> %and.i to <4 x double>
129 define <4 x double> @andnotpd256fold(<4 x double> %y, <4 x double>* nocapture %x) nounwind uwtable readonly ssp {
131 %tmp2 = load <4 x double>* %x, align 32
132 %0 = bitcast <4 x double> %y to <4 x i64>
133 %neg.i = xor <4 x i64> %0, <i64 -1, i64 -1, i64 -1, i64 -1>
134 %1 = bitcast <4 x double> %tmp2 to <4 x i64>
135 %and.i = and <4 x i64> %1, %neg.i
136 %2 = bitcast <4 x i64> %and.i to <4 x double>
141 define <8 x float> @andnotps256(<8 x float> %y, <8 x float> %x) nounwind uwtable readnone ssp {
143 %0 = bitcast <8 x float> %x to <8 x i32>
144 %neg.i = xor <8 x i32> %0, <i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1>
145 %1 = bitcast <8 x float> %y to <8 x i32>
146 %and.i = and <8 x i32> %1, %neg.i
147 %2 = bitcast <8 x i32> %and.i to <8 x float>
152 define <8 x float> @andnotps256fold(<8 x float> %y, <8 x float>* nocapture %x) nounwind uwtable readonly ssp {
154 %tmp2 = load <8 x float>* %x, align 32
155 %0 = bitcast <8 x float> %y to <8 x i32>
156 %neg.i = xor <8 x i32> %0, <i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1>
157 %1 = bitcast <8 x float> %tmp2 to <8 x i32>
158 %and.i = and <8 x i32> %1, %neg.i
159 %2 = bitcast <8 x i32> %and.i to <8 x float>
163 ;;; Test that basic 2 x i64 logic use the integer version on AVX
166 define <2 x i64> @vpandn(<2 x i64> %a, <2 x i64> %b) nounwind uwtable readnone ssp {
168 ; Force the execution domain with an add.
169 %a2 = add <2 x i64> %a, <i64 1, i64 1>
170 %y = xor <2 x i64> %a2, <i64 -1, i64 -1>
171 %x = and <2 x i64> %a, %y
176 define <2 x i64> @vpand(<2 x i64> %a, <2 x i64> %b) nounwind uwtable readnone ssp {
178 ; Force the execution domain with an add.
179 %a2 = add <2 x i64> %a, <i64 1, i64 1>
180 %x = and <2 x i64> %a2, %b