Two sets of changes. Sorry they are intermingled.

[oota-llvm.git] / test / CodeGen / ARM / vld4.ll

diff --git a/test/CodeGen/ARM/vld4.ll b/test/CodeGen/ARM/vld4.ll
index e800cb539aa4b06dc20356a95dd4aa5d44602d1b..a616a98a25a2e878e1f933b6ef58e69ea49b424e 100644 (file)
--- a/test/CodeGen/ARM/vld4.ll
+++ b/test/CodeGen/ARM/vld4.ll
@@ -13,8 +13,9 @@
 
 define <8 x i8> @vld4i8(i8* %A) nounwind {
 ;CHECK: vld4i8:
-;CHECK: vld4.8
-       %tmp1 = call %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4.v8i8(i8* %A)
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.8 {d16, d17, d18, d19}, [r0, :64]
+       %tmp1 = call %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4.v8i8(i8* %A, i32 8)
         %tmp2 = extractvalue %struct.__neon_int8x8x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int8x8x4_t %tmp1, 2
         %tmp4 = add <8 x i8> %tmp2, %tmp3
@@ -23,9 +24,10 @@ define <8 x i8> @vld4i8(i8* %A) nounwind {
 
 define <4 x i16> @vld4i16(i16* %A) nounwind {
 ;CHECK: vld4i16:
-;CHECK: vld4.16
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.16 {d16, d17, d18, d19}, [r0, :128]
        %tmp0 = bitcast i16* %A to i8*
-       %tmp1 = call %struct.__neon_int16x4x4_t @llvm.arm.neon.vld4.v4i16(i8* %tmp0)
+       %tmp1 = call %struct.__neon_int16x4x4_t @llvm.arm.neon.vld4.v4i16(i8* %tmp0, i32 16)
         %tmp2 = extractvalue %struct.__neon_int16x4x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int16x4x4_t %tmp1, 2
         %tmp4 = add <4 x i16> %tmp2, %tmp3
@@ -34,9 +36,10 @@ define <4 x i16> @vld4i16(i16* %A) nounwind {
 
 define <2 x i32> @vld4i32(i32* %A) nounwind {
 ;CHECK: vld4i32:
-;CHECK: vld4.32
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.32 {d16, d17, d18, d19}, [r0, :256]
        %tmp0 = bitcast i32* %A to i8*
-       %tmp1 = call %struct.__neon_int32x2x4_t @llvm.arm.neon.vld4.v2i32(i8* %tmp0)
+       %tmp1 = call %struct.__neon_int32x2x4_t @llvm.arm.neon.vld4.v2i32(i8* %tmp0, i32 32)
         %tmp2 = extractvalue %struct.__neon_int32x2x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int32x2x4_t %tmp1, 2
         %tmp4 = add <2 x i32> %tmp2, %tmp3
@@ -47,18 +50,19 @@ define <2 x float> @vld4f(float* %A) nounwind {
 ;CHECK: vld4f:
 ;CHECK: vld4.32
        %tmp0 = bitcast float* %A to i8*
-       %tmp1 = call %struct.__neon_float32x2x4_t @llvm.arm.neon.vld4.v2f32(i8* %tmp0)
+       %tmp1 = call %struct.__neon_float32x2x4_t @llvm.arm.neon.vld4.v2f32(i8* %tmp0, i32 1)
         %tmp2 = extractvalue %struct.__neon_float32x2x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_float32x2x4_t %tmp1, 2
-        %tmp4 = add <2 x float> %tmp2, %tmp3
+        %tmp4 = fadd <2 x float> %tmp2, %tmp3
        ret <2 x float> %tmp4
 }
 
 define <1 x i64> @vld4i64(i64* %A) nounwind {
 ;CHECK: vld4i64:
-;CHECK: vld1.64
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld1.64 {d16, d17, d18, d19}, [r0, :256]
        %tmp0 = bitcast i64* %A to i8*
-       %tmp1 = call %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8* %tmp0)
+       %tmp1 = call %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8* %tmp0, i32 64)
         %tmp2 = extractvalue %struct.__neon_int64x1x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int64x1x4_t %tmp1, 2
         %tmp4 = add <1 x i64> %tmp2, %tmp3
@@ -67,9 +71,10 @@ define <1 x i64> @vld4i64(i64* %A) nounwind {
 
 define <16 x i8> @vld4Qi8(i8* %A) nounwind {
 ;CHECK: vld4Qi8:
-;CHECK: vld4.8
-;CHECK: vld4.8
-       %tmp1 = call %struct.__neon_int8x16x4_t @llvm.arm.neon.vld4.v16i8(i8* %A)
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.8 {d16, d18, d20, d22}, [r0, :256]!
+;CHECK: vld4.8 {d17, d19, d21, d23}, [r0, :256]
+       %tmp1 = call %struct.__neon_int8x16x4_t @llvm.arm.neon.vld4.v16i8(i8* %A, i32 64)
         %tmp2 = extractvalue %struct.__neon_int8x16x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int8x16x4_t %tmp1, 2
         %tmp4 = add <16 x i8> %tmp2, %tmp3
@@ -78,10 +83,11 @@ define <16 x i8> @vld4Qi8(i8* %A) nounwind {
 
 define <8 x i16> @vld4Qi16(i16* %A) nounwind {
 ;CHECK: vld4Qi16:
-;CHECK: vld4.16
-;CHECK: vld4.16
+;Check for no alignment specifier.
+;CHECK: vld4.16 {d16, d18, d20, d22}, [r0]!
+;CHECK: vld4.16 {d17, d19, d21, d23}, [r0]
        %tmp0 = bitcast i16* %A to i8*
-       %tmp1 = call %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8* %tmp0)
+       %tmp1 = call %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8* %tmp0, i32 1)
         %tmp2 = extractvalue %struct.__neon_int16x8x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int16x8x4_t %tmp1, 2
         %tmp4 = add <8 x i16> %tmp2, %tmp3
@@ -93,7 +99,7 @@ define <4 x i32> @vld4Qi32(i32* %A) nounwind {
 ;CHECK: vld4.32
 ;CHECK: vld4.32
        %tmp0 = bitcast i32* %A to i8*
-       %tmp1 = call %struct.__neon_int32x4x4_t @llvm.arm.neon.vld4.v4i32(i8* %tmp0)
+       %tmp1 = call %struct.__neon_int32x4x4_t @llvm.arm.neon.vld4.v4i32(i8* %tmp0, i32 1)
         %tmp2 = extractvalue %struct.__neon_int32x4x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_int32x4x4_t %tmp1, 2
         %tmp4 = add <4 x i32> %tmp2, %tmp3
@@ -105,20 +111,20 @@ define <4 x float> @vld4Qf(float* %A) nounwind {
 ;CHECK: vld4.32
 ;CHECK: vld4.32
        %tmp0 = bitcast float* %A to i8*
-       %tmp1 = call %struct.__neon_float32x4x4_t @llvm.arm.neon.vld4.v4f32(i8* %tmp0)
+       %tmp1 = call %struct.__neon_float32x4x4_t @llvm.arm.neon.vld4.v4f32(i8* %tmp0, i32 1)
         %tmp2 = extractvalue %struct.__neon_float32x4x4_t %tmp1, 0
         %tmp3 = extractvalue %struct.__neon_float32x4x4_t %tmp1, 2
-        %tmp4 = add <4 x float> %tmp2, %tmp3
+        %tmp4 = fadd <4 x float> %tmp2, %tmp3
        ret <4 x float> %tmp4
 }
 
-declare %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4.v8i8(i8*) nounwind readonly
-declare %struct.__neon_int16x4x4_t @llvm.arm.neon.vld4.v4i16(i8*) nounwind readonly
-declare %struct.__neon_int32x2x4_t @llvm.arm.neon.vld4.v2i32(i8*) nounwind readonly
-declare %struct.__neon_float32x2x4_t @llvm.arm.neon.vld4.v2f32(i8*) nounwind readonly
-declare %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8*) nounwind readonly
+declare %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4.v8i8(i8*, i32) nounwind readonly
+declare %struct.__neon_int16x4x4_t @llvm.arm.neon.vld4.v4i16(i8*, i32) nounwind readonly
+declare %struct.__neon_int32x2x4_t @llvm.arm.neon.vld4.v2i32(i8*, i32) nounwind readonly
+declare %struct.__neon_float32x2x4_t @llvm.arm.neon.vld4.v2f32(i8*, i32) nounwind readonly
+declare %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8*, i32) nounwind readonly
 
-declare %struct.__neon_int8x16x4_t @llvm.arm.neon.vld4.v16i8(i8*) nounwind readonly
-declare %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8*) nounwind readonly
-declare %struct.__neon_int32x4x4_t @llvm.arm.neon.vld4.v4i32(i8*) nounwind readonly
-declare %struct.__neon_float32x4x4_t @llvm.arm.neon.vld4.v4f32(i8*) nounwind readonly
+declare %struct.__neon_int8x16x4_t @llvm.arm.neon.vld4.v16i8(i8*, i32) nounwind readonly
+declare %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8*, i32) nounwind readonly
+declare %struct.__neon_int32x4x4_t @llvm.arm.neon.vld4.v4i32(i8*, i32) nounwind readonly
+declare %struct.__neon_float32x4x4_t @llvm.arm.neon.vld4.v4f32(i8*, i32) nounwind readonly