From 2a6e6161422154ea44f9e653e2d0bbfc3942af2b Mon Sep 17 00:00:00 2001 From: Bob Wilson Date: Thu, 23 Sep 2010 23:42:37 +0000 Subject: Set alignment operand for NEON VST instructions. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@114709 91177308-0d34-0410-b5e6-96231b3b80d8 --- test/CodeGen/ARM/vst4.ll | 32 +++++++++++++++++++------------- 1 file changed, 19 insertions(+), 13 deletions(-) (limited to 'test/CodeGen/ARM/vst4.ll') diff --git a/test/CodeGen/ARM/vst4.ll b/test/CodeGen/ARM/vst4.ll index d302f097fc..a339192b17 100644 --- a/test/CodeGen/ARM/vst4.ll +++ b/test/CodeGen/ARM/vst4.ll @@ -2,27 +2,30 @@ define void @vst4i8(i8* %A, <8 x i8>* %B) nounwind { ;CHECK: vst4i8: -;CHECK: vst4.8 +;Check the alignment value. Max for this instruction is 256 bits: +;CHECK: vst4.8 {d0, d1, d2, d3}, [r0, :64] %tmp1 = load <8 x i8>* %B - call void @llvm.arm.neon.vst4.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 1) + call void @llvm.arm.neon.vst4.v8i8(i8* %A, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, <8 x i8> %tmp1, i32 8) ret void } define void @vst4i16(i16* %A, <4 x i16>* %B) nounwind { ;CHECK: vst4i16: -;CHECK: vst4.16 +;Check the alignment value. Max for this instruction is 256 bits: +;CHECK: vst4.16 {d0, d1, d2, d3}, [r0, :128] %tmp0 = bitcast i16* %A to i8* %tmp1 = load <4 x i16>* %B - call void @llvm.arm.neon.vst4.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 1) + call void @llvm.arm.neon.vst4.v4i16(i8* %tmp0, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, <4 x i16> %tmp1, i32 16) ret void } define void @vst4i32(i32* %A, <2 x i32>* %B) nounwind { ;CHECK: vst4i32: -;CHECK: vst4.32 +;Check the alignment value. Max for this instruction is 256 bits: +;CHECK: vst4.32 {d0, d1, d2, d3}, [r0, :256] %tmp0 = bitcast i32* %A to i8* %tmp1 = load <2 x i32>* %B - call void @llvm.arm.neon.vst4.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 1) + call void @llvm.arm.neon.vst4.v2i32(i8* %tmp0, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, <2 x i32> %tmp1, i32 32) ret void } @@ -37,26 +40,29 @@ define void @vst4f(float* %A, <2 x float>* %B) nounwind { define void @vst4i64(i64* %A, <1 x i64>* %B) nounwind { ;CHECK: vst4i64: -;CHECK: vst1.64 +;Check the alignment value. Max for this instruction is 256 bits: +;CHECK: vst1.64 {d0, d1, d2, d3}, [r0, :256] %tmp0 = bitcast i64* %A to i8* %tmp1 = load <1 x i64>* %B - call void @llvm.arm.neon.vst4.v1i64(i8* %tmp0, <1 x i64> %tmp1, <1 x i64> %tmp1, <1 x i64> %tmp1, <1 x i64> %tmp1, i32 1) + call void @llvm.arm.neon.vst4.v1i64(i8* %tmp0, <1 x i64> %tmp1, <1 x i64> %tmp1, <1 x i64> %tmp1, <1 x i64> %tmp1, i32 64) ret void } define void @vst4Qi8(i8* %A, <16 x i8>* %B) nounwind { ;CHECK: vst4Qi8: -;CHECK: vst4.8 -;CHECK: vst4.8 +;Check the alignment value. Max for this instruction is 256 bits: +;CHECK: vst4.8 {d0, d2, d4, d6}, [r0, :256]! +;CHECK: vst4.8 {d1, d3, d5, d7}, [r0, :256] %tmp1 = load <16 x i8>* %B - call void @llvm.arm.neon.vst4.v16i8(i8* %A, <16 x i8> %tmp1, <16 x i8> %tmp1, <16 x i8> %tmp1, <16 x i8> %tmp1, i32 1) + call void @llvm.arm.neon.vst4.v16i8(i8* %A, <16 x i8> %tmp1, <16 x i8> %tmp1, <16 x i8> %tmp1, <16 x i8> %tmp1, i32 64) ret void } define void @vst4Qi16(i16* %A, <8 x i16>* %B) nounwind { ;CHECK: vst4Qi16: -;CHECK: vst4.16 -;CHECK: vst4.16 +;Check for no alignment specifier. +;CHECK: vst4.16 {d0, d2, d4, d6}, [r0]! +;CHECK: vst4.16 {d1, d3, d5, d7}, [r0] %tmp0 = bitcast i16* %A to i8* %tmp1 = load <8 x i16>* %B call void @llvm.arm.neon.vst4.v8i16(i8* %tmp0, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, <8 x i16> %tmp1, i32 1) -- cgit v1.2.3