do not compare actual branch labels; this may fix llvm-gcc-x86_64-darwin10-cross...

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

diff --git a/test/CodeGen/ARM/vld4.ll b/test/CodeGen/ARM/vld4.ll
index d1bf957ebadcab10c9d5228a9f7c672bb90b4785..6330d4f79502c1aeab90beb3a882c86aceb138ef 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, i32 1)
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.8 {d0, d1, d2, d3}, [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 {d0, d1, d2, d3}, [r0, :128]
        %tmp0 = bitcast i16* %A to i8*
-       %tmp1 = call %struct.__neon_int16x4x4_t @llvm.arm.neon.vld4.v4i16(i8* %tmp0, i32 1)
+       %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 {d0, d1, d2, d3}, [r0, :256]
        %tmp0 = bitcast i32* %A to i8*
-       %tmp1 = call %struct.__neon_int32x2x4_t @llvm.arm.neon.vld4.v2i32(i8* %tmp0, i32 1)
+       %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
@@ -56,9 +59,10 @@ define <2 x float> @vld4f(float* %A) nounwind {
 
 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 {d0, d1, d2, d3}, [r0, :256]
        %tmp0 = bitcast i64* %A to i8*
-       %tmp1 = call %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8* %tmp0, i32 1)
+       %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, i32 1)
+;Check the alignment value.  Max for this instruction is 256 bits:
+;CHECK: vld4.8 {d0, d2, d4, d6}, [r0, :256]!
+;CHECK: vld4.8 {d1, d3, d5, d7}, [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
@@ -78,8 +83,9 @@ 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 {d0, d2, d4, d6}, [r0]!
+;CHECK: vld4.16 {d1, d3, d5, d7}, [r0]
        %tmp0 = bitcast i16* %A to i8*
        %tmp1 = call %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8* %tmp0, i32 1)
         %tmp2 = extractvalue %struct.__neon_int16x8x4_t %tmp1, 0