AVX-512: store <4 x i1> and <2 x i1> values in memory

[oota-llvm.git] / test / CodeGen / X86 / x86-setcc-int-to-fp-combine.ll

diff --git a/test/CodeGen/X86/x86-setcc-int-to-fp-combine.ll b/test/CodeGen/X86/x86-setcc-int-to-fp-combine.ll
index 4fe6c664df53bb7b6412a316a20476c131923ecf..248a9202e99795b5cf83698b53c926d9d7cbf280 100644 (file)
--- a/test/CodeGen/X86/x86-setcc-int-to-fp-combine.ll
+++ b/test/CodeGen/X86/x86-setcc-int-to-fp-combine.ll
@@ -2,10 +2,10 @@
 
 define <4 x float> @foo(<4 x float> %val, <4 x float> %test) nounwind {
 ; CHECK-LABEL: LCPI0_0:
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
 ; CHECK-LABEL: foo:
 ; CHECK: cmpeqps %xmm1, %xmm0
 ; CHECK-NEXT: andps LCPI0_0(%rip), %xmm0
@@ -27,12 +27,8 @@ define void @foo1(<4 x float> %val, <4 x float> %test, <4 x double>* %p) nounwin
 ; CHECK-NEXT: .long 1                       ## 0x1
 ; CHECK-NEXT: .long 1                       ## 0x1
 ; CHECK-LABEL: foo1:
-;   FIXME: The operation gets scalarized. If/when the compiler learns to better
-;          use [V]CVTDQ2PD, this will need updated.
-; CHECK: cvtsi2sdq
-; CHECK: cvtsi2sdq
-; CHECK: cvtsi2sdq
-; CHECK: cvtsi2sdq
+; CHECK: cvtdq2pd
+; CHECK: cvtdq2pd
   %cmp = fcmp oeq <4 x float> %val, %test
   %ext = zext <4 x i1> %cmp to <4 x i32>
   %result = sitofp <4 x i32> %ext to <4 x double>
@@ -59,10 +55,10 @@ define void @foo2(<4 x float>* noalias %result) nounwind {
 ; scalar value like what the zext creates.
 define <4 x float> @foo3(<4 x float> %val, <4 x float> %test) nounwind {
 ; CHECK-LABEL: LCPI3_0:
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
-; CHECK-NEXT: .long 0                       ## float 0.000000e+00
-; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
-; CHECK-NEXT: .long 0                       ## float 0.000000e+00
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
+; CHECK-NEXT: .long 0                       ## 0x0
+; CHECK-NEXT: .long 1065353216              ## 0x3f800000
+; CHECK-NEXT: .long 0                       ## 0x0
 ; CHECK-LABEL: foo3:
 ; CHECK: cmpeqps %xmm1, %xmm0
 ; CHECK-NEXT: andps LCPI3_0(%rip), %xmm0
@@ -72,3 +68,18 @@ define <4 x float> @foo3(<4 x float> %val, <4 x float> %test) nounwind {
   %result = sitofp <4 x i32> %and to <4 x float>
   ret <4 x float> %result
 }
+
+; Test the general purpose constant folding of uint->fp.
+define void @foo4(<4 x float>* noalias %result) nounwind {
+; CHECK-LABEL: LCPI4_0:
+; CHECK-NEXT: .long 1065353216              ## float 1.000000e+00
+; CHECK-NEXT: .long 1123942400              ## float 1.270000e+02
+; CHECK-NEXT: .long 1124073472              ## float 1.280000e+02
+; CHECK-NEXT: .long 1132396544              ## float 2.550000e+02
+; CHECK-LABEL: foo4:
+; CHECK:  movaps LCPI4_0(%rip), %xmm0
+
+  %val = uitofp <4 x i8> <i8 1, i8 127, i8 -128, i8 -1> to <4 x float>
+  store <4 x float> %val, <4 x float>* %result
+  ret void
+}