summaryrefslogtreecommitdiffstats
path: root/llvm/test/CodeGen/ARM/vstlane.ll
diff options
context:
space:
mode:
Diffstat (limited to 'llvm/test/CodeGen/ARM/vstlane.ll')
-rw-r--r--llvm/test/CodeGen/ARM/vstlane.ll90
1 files changed, 45 insertions, 45 deletions
diff --git a/llvm/test/CodeGen/ARM/vstlane.ll b/llvm/test/CodeGen/ARM/vstlane.ll
index a4575417bce..7e130ea01b6 100644
--- a/llvm/test/CodeGen/ARM/vstlane.ll
+++ b/llvm/test/CodeGen/ARM/vstlane.ll
@@ -110,7 +110,7 @@ define void @vst2lanei8(i8* %A, <8 x i8>* %B) nounwind {
;Check the alignment value. Max for this instruction is 16 bits:
;CHECK: vst2.8 {d16[1], d17[1]}, [r0:16]
%tmp1 = load <8 x i8>, <8 x i8>* %B
- call void @llvm.arm.neon.vst2lane.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 4)
+ call void @llvm.arm.neon.vst2lane.p0i8.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 4)
ret void
}
@@ -120,7 +120,7 @@ define void @vst2lanei16(i16* %A, <4 x i16>* %B) nounwind {
;CHECK: vst2.16 {d16[1], d17[1]}, [r0:32]
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <4 x i16>, <4 x i16>* %B
- call void @llvm.arm.neon.vst2lane.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 8)
+ call void @llvm.arm.neon.vst2lane.p0i8.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 8)
ret void
}
@@ -131,7 +131,7 @@ define void @vst2lanei16_update(i16** %ptr, <4 x i16>* %B, i32 %inc) nounwind {
%A = load i16*, i16** %ptr
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <4 x i16>, <4 x i16>* %B
- call void @llvm.arm.neon.vst2lane.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 2)
+ call void @llvm.arm.neon.vst2lane.p0i8.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 2)
%tmp2 = getelementptr i16, i16* %A, i32 %inc
store i16* %tmp2, i16** %ptr
ret void
@@ -142,7 +142,7 @@ define void @vst2lanei32(i32* %A, <2 x i32>* %B) nounwind {
;CHECK: vst2.32
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <2 x i32>, <2 x i32>* %B
- call void @llvm.arm.neon.vst2lane.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst2lane.p0i8.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 1)
ret void
}
@@ -151,7 +151,7 @@ define void @vst2lanef(float* %A, <2 x float>* %B) nounwind {
;CHECK: vst2.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <2 x float>, <2 x float>* %B
- call void @llvm.arm.neon.vst2lane.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst2lane.p0i8.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
ret void
}
@@ -161,7 +161,7 @@ define void @vst2laneQi16(i16* %A, <8 x i16>* %B) nounwind {
;CHECK: vst2.16 {d17[1], d19[1]}, [r0]
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <8 x i16>, <8 x i16>* %B
- call void @llvm.arm.neon.vst2lane.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 5, i32 1)
+ call void @llvm.arm.neon.vst2lane.p0i8.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 5, i32 1)
ret void
}
@@ -171,7 +171,7 @@ define void @vst2laneQi32(i32* %A, <4 x i32>* %B) nounwind {
;CHECK: vst2.32 {d17[0], d19[0]}, [r0:64]
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <4 x i32>, <4 x i32>* %B
- call void @llvm.arm.neon.vst2lane.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 2, i32 16)
+ call void @llvm.arm.neon.vst2lane.p0i8.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 2, i32 16)
ret void
}
@@ -180,24 +180,24 @@ define void @vst2laneQf(float* %A, <4 x float>* %B) nounwind {
;CHECK: vst2.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <4 x float>, <4 x float>* %B
- call void @llvm.arm.neon.vst2lane.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, i32 3, i32 1)
+ call void @llvm.arm.neon.vst2lane.p0i8.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, i32 3, i32 1)
ret void
}
-declare void @llvm.arm.neon.vst2lane.v8i8(i8*, <8 x i8>, <8 x i8>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v4i16(i8*, <4 x i16>, <4 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v2i32(i8*, <2 x i32>, <2 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v2f32(i8*, <2 x float>, <2 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v8i8(i8*, <8 x i8>, <8 x i8>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v4i16(i8*, <4 x i16>, <4 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v2i32(i8*, <2 x i32>, <2 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v2f32(i8*, <2 x float>, <2 x float>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v8i16(i8*, <8 x i16>, <8 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v4i32(i8*, <4 x i32>, <4 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst2lane.v4f32(i8*, <4 x float>, <4 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v8i16(i8*, <8 x i16>, <8 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v4i32(i8*, <4 x i32>, <4 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst2lane.p0i8.v4f32(i8*, <4 x float>, <4 x float>, i32, i32) nounwind
define void @vst3lanei8(i8* %A, <8 x i8>* %B) nounwind {
;CHECK-LABEL: vst3lanei8:
;CHECK: vst3.8
%tmp1 = load <8 x i8>, <8 x i8>* %B
- call void @llvm.arm.neon.vst3lane.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 1)
ret void
}
@@ -207,7 +207,7 @@ define void @vst3lanei16(i16* %A, <4 x i16>* %B) nounwind {
;CHECK: vst3.16 {d16[1], d17[1], d18[1]}, [r0]
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <4 x i16>, <4 x i16>* %B
- call void @llvm.arm.neon.vst3lane.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 8)
+ call void @llvm.arm.neon.vst3lane.p0i8.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 8)
ret void
}
@@ -216,7 +216,7 @@ define void @vst3lanei32(i32* %A, <2 x i32>* %B) nounwind {
;CHECK: vst3.32
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <2 x i32>, <2 x i32>* %B
- call void @llvm.arm.neon.vst3lane.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 1)
ret void
}
@@ -225,7 +225,7 @@ define void @vst3lanef(float* %A, <2 x float>* %B) nounwind {
;CHECK: vst3.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <2 x float>, <2 x float>* %B
- call void @llvm.arm.neon.vst3lane.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
ret void
}
@@ -235,7 +235,7 @@ define void @vst3laneQi16(i16* %A, <8 x i16>* %B) nounwind {
;CHECK: vst3.16 {d17[2], d19[2], d21[2]}, [r0]
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <8 x i16>, <8 x i16>* %B
- call void @llvm.arm.neon.vst3lane.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 6, i32 8)
+ call void @llvm.arm.neon.vst3lane.p0i8.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 6, i32 8)
ret void
}
@@ -244,7 +244,7 @@ define void @vst3laneQi32(i32* %A, <4 x i32>* %B) nounwind {
;CHECK: vst3.32
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <4 x i32>, <4 x i32>* %B
- call void @llvm.arm.neon.vst3lane.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 0, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 0, i32 1)
ret void
}
@@ -255,7 +255,7 @@ define void @vst3laneQi32_update(i32** %ptr, <4 x i32>* %B) nounwind {
%A = load i32*, i32** %ptr
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <4 x i32>, <4 x i32>* %B
- call void @llvm.arm.neon.vst3lane.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 0, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 0, i32 1)
%tmp2 = getelementptr i32, i32* %A, i32 3
store i32* %tmp2, i32** %ptr
ret void
@@ -266,18 +266,18 @@ define void @vst3laneQf(float* %A, <4 x float>* %B) nounwind {
;CHECK: vst3.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <4 x float>, <4 x float>* %B
- call void @llvm.arm.neon.vst3lane.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst3lane.p0i8.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, i32 1, i32 1)
ret void
}
-declare void @llvm.arm.neon.vst3lane.v8i8(i8*, <8 x i8>, <8 x i8>, <8 x i8>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v4i16(i8*, <4 x i16>, <4 x i16>, <4 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v2i32(i8*, <2 x i32>, <2 x i32>, <2 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v2f32(i8*, <2 x float>, <2 x float>, <2 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v8i8(i8*, <8 x i8>, <8 x i8>, <8 x i8>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v4i16(i8*, <4 x i16>, <4 x i16>, <4 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v2i32(i8*, <2 x i32>, <2 x i32>, <2 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v2f32(i8*, <2 x float>, <2 x float>, <2 x float>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v8i16(i8*, <8 x i16>, <8 x i16>, <8 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v4i32(i8*, <4 x i32>, <4 x i32>, <4 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst3lane.v4f32(i8*, <4 x float>, <4 x float>, <4 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v8i16(i8*, <8 x i16>, <8 x i16>, <8 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v4i32(i8*, <4 x i32>, <4 x i32>, <4 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst3lane.p0i8.v4f32(i8*, <4 x float>, <4 x float>, <4 x float>, i32, i32) nounwind
define void @vst4lanei8(i8* %A, <8 x i8>* %B) nounwind {
@@ -285,7 +285,7 @@ define void @vst4lanei8(i8* %A, <8 x i8>* %B) nounwind {
;Check the alignment value. Max for this instruction is 32 bits:
;CHECK: vst4.8 {d16[1], d17[1], d18[1], d19[1]}, [r0:32]
%tmp1 = load <8 x i8>, <8 x i8>* %B
- call void @llvm.arm.neon.vst4lane.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 8)
+ call void @llvm.arm.neon.vst4lane.p0i8.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 8)
ret void
}
@@ -295,7 +295,7 @@ define void @vst4lanei8_update(i8** %ptr, <8 x i8>* %B) nounwind {
;CHECK: vst4.8 {d16[1], d17[1], d18[1], d19[1]}, [r1:32]!
%A = load i8*, i8** %ptr
%tmp1 = load <8 x i8>, <8 x i8>* %B
- call void @llvm.arm.neon.vst4lane.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 8)
+ call void @llvm.arm.neon.vst4lane.p0i8.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1, i32 8)
%tmp2 = getelementptr i8, i8* %A, i32 4
store i8* %tmp2, i8** %ptr
ret void
@@ -306,7 +306,7 @@ define void @vst4lanei16(i16* %A, <4 x i16>* %B) nounwind {
;CHECK: vst4.16
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <4 x i16>, <4 x i16>* %B
- call void @llvm.arm.neon.vst4lane.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst4lane.p0i8.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1, i32 1)
ret void
}
@@ -316,7 +316,7 @@ define void @vst4lanei32(i32* %A, <2 x i32>* %B) nounwind {
;CHECK: vst4.32 {d16[1], d17[1], d18[1], d19[1]}, [r0:128]
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <2 x i32>, <2 x i32>* %B
- call void @llvm.arm.neon.vst4lane.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 16)
+ call void @llvm.arm.neon.vst4lane.p0i8.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1, i32 16)
ret void
}
@@ -325,7 +325,7 @@ define void @vst4lanef(float* %A, <2 x float>* %B) nounwind {
;CHECK: vst4.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <2 x float>, <2 x float>* %B
- call void @llvm.arm.neon.vst4lane.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst4lane.p0i8.v2f32(i8* %tmp0, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, <2 x float> %tmp1, i32 1, i32 1)
ret void
}
@@ -335,7 +335,7 @@ define void @vst4laneQi16(i16* %A, <8 x i16>* %B) nounwind {
;CHECK: vst4.16 {d17[3], d19[3], d21[3], d23[3]}, [r0:64]
%tmp0 = bitcast i16* %A to i8*
%tmp1 = load <8 x i16>, <8 x i16>* %B
- call void @llvm.arm.neon.vst4lane.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 7, i32 16)
+ call void @llvm.arm.neon.vst4lane.p0i8.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 7, i32 16)
ret void
}
@@ -345,7 +345,7 @@ define void @vst4laneQi32(i32* %A, <4 x i32>* %B) nounwind {
;CHECK: vst4.32 {d17[0], d19[0], d21[0], d23[0]}, [r0]
%tmp0 = bitcast i32* %A to i8*
%tmp1 = load <4 x i32>, <4 x i32>* %B
- call void @llvm.arm.neon.vst4lane.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 2, i32 1)
+ call void @llvm.arm.neon.vst4lane.p0i8.v4i32(i8* %tmp0, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, <4 x i32> %tmp1, i32 2, i32 1)
ret void
}
@@ -354,7 +354,7 @@ define void @vst4laneQf(float* %A, <4 x float>* %B) nounwind {
;CHECK: vst4.32
%tmp0 = bitcast float* %A to i8*
%tmp1 = load <4 x float>, <4 x float>* %B
- call void @llvm.arm.neon.vst4lane.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, i32 1, i32 1)
+ call void @llvm.arm.neon.vst4lane.p0i8.v4f32(i8* %tmp0, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, <4 x float> %tmp1, i32 1, i32 1)
ret void
}
@@ -365,11 +365,11 @@ define <8 x i16> @variable_insertelement(<8 x i16> %a, i16 %b, i32 %c) nounwind
ret <8 x i16> %r
}
-declare void @llvm.arm.neon.vst4lane.v8i8(i8*, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v4i16(i8*, <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v2i32(i8*, <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v2f32(i8*, <2 x float>, <2 x float>, <2 x float>, <2 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v8i8(i8*, <8 x i8>, <8 x i8>, <8 x i8>, <8 x i8>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v4i16(i8*, <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v2i32(i8*, <2 x i32>, <2 x i32>, <2 x i32>, <2 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v2f32(i8*, <2 x float>, <2 x float>, <2 x float>, <2 x float>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v8i16(i8*, <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v4i32(i8*, <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32>, i32, i32) nounwind
-declare void @llvm.arm.neon.vst4lane.v4f32(i8*, <4 x float>, <4 x float>, <4 x float>, <4 x float>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v8i16(i8*, <8 x i16>, <8 x i16>, <8 x i16>, <8 x i16>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v4i32(i8*, <4 x i32>, <4 x i32>, <4 x i32>, <4 x i32>, i32, i32) nounwind
+declare void @llvm.arm.neon.vst4lane.p0i8.v4f32(i8*, <4 x float>, <4 x float>, <4 x float>, <4 x float>, i32, i32) nounwind
OpenPOWER on IntegriCloud