[ARM]: Extend -mfpu options for half-precision and vfpv3xd

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

diff --git a/test/CodeGen/ARM/vld4.ll b/test/CodeGen/ARM/vld4.ll
index a616a98a25a2e878e1f933b6ef58e69ea49b424e..575e0fa717fb3c306babab271342e201c90d4879 100644 (file)
--- a/test/CodeGen/ARM/vld4.ll
+++ b/test/CodeGen/ARM/vld4.ll
@@ -1,4 +1,4 @@
-; RUN: llc < %s -march=arm -mattr=+neon | FileCheck %s
+; RUN: llc -mtriple=arm-eabi -mattr=+neon %s -o - | FileCheck %s
 
 %struct.__neon_int8x8x4_t = type { <8 x i8>,  <8 x i8>,  <8 x i8>, <8 x i8> }
 %struct.__neon_int16x4x4_t = type { <4 x i16>, <4 x i16>, <4 x i16>, <4 x i16> }
@@ -12,9 +12,9 @@
 %struct.__neon_float32x4x4_t = type { <4 x float>, <4 x float>, <4 x float>, <4 x float> }
 
 define <8 x i8> @vld4i8(i8* %A) nounwind {
-;CHECK: vld4i8:
+;CHECK-LABEL: vld4i8:
 ;Check the alignment value.  Max for this instruction is 256 bits:
-;CHECK: vld4.8 {d16, d17, d18, d19}, [r0, :64]
+;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
@@ -22,10 +22,24 @@ define <8 x i8> @vld4i8(i8* %A) nounwind {
        ret <8 x i8> %tmp4
 }
 
+;Check for a post-increment updating load with register increment.
+define <8 x i8> @vld4i8_update(i8** %ptr, i32 %inc) nounwind {
+;CHECK-LABEL: vld4i8_update:
+;CHECK: vld4.8 {d16, d17, d18, d19}, [r2:128], r1
+       %A = load i8*, i8** %ptr
+       %tmp1 = call %struct.__neon_int8x8x4_t @llvm.arm.neon.vld4.v8i8(i8* %A, i32 16)
+       %tmp2 = extractvalue %struct.__neon_int8x8x4_t %tmp1, 0
+       %tmp3 = extractvalue %struct.__neon_int8x8x4_t %tmp1, 2
+       %tmp4 = add <8 x i8> %tmp2, %tmp3
+       %tmp5 = getelementptr i8, i8* %A, i32 %inc
+       store i8* %tmp5, i8** %ptr
+       ret <8 x i8> %tmp4
+}
+
 define <4 x i16> @vld4i16(i16* %A) nounwind {
-;CHECK: vld4i16:
+;CHECK-LABEL: vld4i16:
 ;Check the alignment value.  Max for this instruction is 256 bits:
-;CHECK: vld4.16 {d16, d17, d18, d19}, [r0, :128]
+;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, i32 16)
         %tmp2 = extractvalue %struct.__neon_int16x4x4_t %tmp1, 0
@@ -35,9 +49,9 @@ define <4 x i16> @vld4i16(i16* %A) nounwind {
 }
 
 define <2 x i32> @vld4i32(i32* %A) nounwind {
-;CHECK: vld4i32:
+;CHECK-LABEL: vld4i32:
 ;Check the alignment value.  Max for this instruction is 256 bits:
-;CHECK: vld4.32 {d16, d17, d18, d19}, [r0, :256]
+;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, i32 32)
         %tmp2 = extractvalue %struct.__neon_int32x2x4_t %tmp1, 0
@@ -47,7 +61,7 @@ define <2 x i32> @vld4i32(i32* %A) nounwind {
 }
 
 define <2 x float> @vld4f(float* %A) nounwind {
-;CHECK: vld4f:
+;CHECK-LABEL: vld4f:
 ;CHECK: vld4.32
        %tmp0 = bitcast float* %A to i8*
        %tmp1 = call %struct.__neon_float32x2x4_t @llvm.arm.neon.vld4.v2f32(i8* %tmp0, i32 1)
@@ -58,9 +72,9 @@ define <2 x float> @vld4f(float* %A) nounwind {
 }
 
 define <1 x i64> @vld4i64(i64* %A) nounwind {
-;CHECK: vld4i64:
+;CHECK-LABEL: vld4i64:
 ;Check the alignment value.  Max for this instruction is 256 bits:
-;CHECK: vld1.64 {d16, d17, d18, d19}, [r0, :256]
+;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, i32 64)
         %tmp2 = extractvalue %struct.__neon_int64x1x4_t %tmp1, 0
@@ -69,11 +83,24 @@ define <1 x i64> @vld4i64(i64* %A) nounwind {
        ret <1 x i64> %tmp4
 }
 
+define <1 x i64> @vld4i64_update(i64** %ptr, i64* %A) nounwind {
+;CHECK-LABEL: vld4i64_update:
+;CHECK: vld1.64 {d16, d17, d18, d19}, [r1:256]!
+        %tmp0 = bitcast i64* %A to i8*
+        %tmp1 = call %struct.__neon_int64x1x4_t @llvm.arm.neon.vld4.v1i64(i8* %tmp0, i32 64)
+        %tmp5 = getelementptr i64, i64* %A, i32 4
+        store i64* %tmp5, i64** %ptr
+        %tmp2 = extractvalue %struct.__neon_int64x1x4_t %tmp1, 0
+        %tmp3 = extractvalue %struct.__neon_int64x1x4_t %tmp1, 2
+        %tmp4 = add <1 x i64> %tmp2, %tmp3
+        ret <1 x i64> %tmp4
+}
+
 define <16 x i8> @vld4Qi8(i8* %A) nounwind {
-;CHECK: vld4Qi8:
+;CHECK-LABEL: vld4Qi8:
 ;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]
+;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
@@ -82,7 +109,7 @@ define <16 x i8> @vld4Qi8(i8* %A) nounwind {
 }
 
 define <8 x i16> @vld4Qi16(i16* %A) nounwind {
-;CHECK: vld4Qi16:
+;CHECK-LABEL: vld4Qi16:
 ;Check for no alignment specifier.
 ;CHECK: vld4.16 {d16, d18, d20, d22}, [r0]!
 ;CHECK: vld4.16 {d17, d19, d21, d23}, [r0]
@@ -94,8 +121,24 @@ define <8 x i16> @vld4Qi16(i16* %A) nounwind {
        ret <8 x i16> %tmp4
 }
 
+;Check for a post-increment updating load. 
+define <8 x i16> @vld4Qi16_update(i16** %ptr) nounwind {
+;CHECK-LABEL: vld4Qi16_update:
+;CHECK: vld4.16 {d16, d18, d20, d22}, [r1:64]!
+;CHECK: vld4.16 {d17, d19, d21, d23}, [r1:64]!
+       %A = load i16*, i16** %ptr
+       %tmp0 = bitcast i16* %A to i8*
+       %tmp1 = call %struct.__neon_int16x8x4_t @llvm.arm.neon.vld4.v8i16(i8* %tmp0, i32 8)
+       %tmp2 = extractvalue %struct.__neon_int16x8x4_t %tmp1, 0
+       %tmp3 = extractvalue %struct.__neon_int16x8x4_t %tmp1, 2
+       %tmp4 = add <8 x i16> %tmp2, %tmp3
+       %tmp5 = getelementptr i16, i16* %A, i32 32
+       store i16* %tmp5, i16** %ptr
+       ret <8 x i16> %tmp4
+}
+
 define <4 x i32> @vld4Qi32(i32* %A) nounwind {
-;CHECK: vld4Qi32:
+;CHECK-LABEL: vld4Qi32:
 ;CHECK: vld4.32
 ;CHECK: vld4.32
        %tmp0 = bitcast i32* %A to i8*
@@ -107,7 +150,7 @@ define <4 x i32> @vld4Qi32(i32* %A) nounwind {
 }
 
 define <4 x float> @vld4Qf(float* %A) nounwind {
-;CHECK: vld4Qf:
+;CHECK-LABEL: vld4Qf:
 ;CHECK: vld4.32
 ;CHECK: vld4.32
        %tmp0 = bitcast float* %A to i8*