Two sets of changes. Sorry they are intermingled.

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

diff --git a/test/CodeGen/ARM/vqadd.ll b/test/CodeGen/ARM/vqadd.ll
index c9e235995360a71703e4f2674317a9590c08553c..a1669b60ab5643ce8f799efe171553d6be083df0 100644 (file)
--- a/test/CodeGen/ARM/vqadd.ll
+++ b/test/CodeGen/ARM/vqadd.ll
@@ -1,14 +1,8 @@
-; RUN: llvm-as < %s | llc -march=arm -mattr=+neon > %t
-; RUN: grep {vqadd\\.s8} %t | count 2
-; RUN: grep {vqadd\\.s16} %t | count 2
-; RUN: grep {vqadd\\.s32} %t | count 2
-; RUN: grep {vqadd\\.s64} %t | count 2
-; RUN: grep {vqadd\\.u8} %t | count 2
-; RUN: grep {vqadd\\.u16} %t | count 2
-; RUN: grep {vqadd\\.u32} %t | count 2
-; RUN: grep {vqadd\\.u64} %t | count 2
+; RUN: llc < %s -march=arm -mattr=+neon | FileCheck %s
 
 define <8 x i8> @vqadds8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
+;CHECK: vqadds8:
+;CHECK: vqadd.s8
        %tmp1 = load <8 x i8>* %A
        %tmp2 = load <8 x i8>* %B
        %tmp3 = call <8 x i8> @llvm.arm.neon.vqadds.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
@@ -16,6 +10,8 @@ define <8 x i8> @vqadds8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
 }
 
 define <4 x i16> @vqadds16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
+;CHECK: vqadds16:
+;CHECK: vqadd.s16
        %tmp1 = load <4 x i16>* %A
        %tmp2 = load <4 x i16>* %B
        %tmp3 = call <4 x i16> @llvm.arm.neon.vqadds.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
@@ -23,6 +19,8 @@ define <4 x i16> @vqadds16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
 }
 
 define <2 x i32> @vqadds32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
+;CHECK: vqadds32:
+;CHECK: vqadd.s32
        %tmp1 = load <2 x i32>* %A
        %tmp2 = load <2 x i32>* %B
        %tmp3 = call <2 x i32> @llvm.arm.neon.vqadds.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
@@ -30,6 +28,8 @@ define <2 x i32> @vqadds32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
 }
 
 define <1 x i64> @vqadds64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
+;CHECK: vqadds64:
+;CHECK: vqadd.s64
        %tmp1 = load <1 x i64>* %A
        %tmp2 = load <1 x i64>* %B
        %tmp3 = call <1 x i64> @llvm.arm.neon.vqadds.v1i64(<1 x i64> %tmp1, <1 x i64> %tmp2)
@@ -37,6 +37,8 @@ define <1 x i64> @vqadds64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
 }
 
 define <8 x i8> @vqaddu8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
+;CHECK: vqaddu8:
+;CHECK: vqadd.u8
        %tmp1 = load <8 x i8>* %A
        %tmp2 = load <8 x i8>* %B
        %tmp3 = call <8 x i8> @llvm.arm.neon.vqaddu.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
@@ -44,6 +46,8 @@ define <8 x i8> @vqaddu8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
 }
 
 define <4 x i16> @vqaddu16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
+;CHECK: vqaddu16:
+;CHECK: vqadd.u16
        %tmp1 = load <4 x i16>* %A
        %tmp2 = load <4 x i16>* %B
        %tmp3 = call <4 x i16> @llvm.arm.neon.vqaddu.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
@@ -51,6 +55,8 @@ define <4 x i16> @vqaddu16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
 }
 
 define <2 x i32> @vqaddu32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
+;CHECK: vqaddu32:
+;CHECK: vqadd.u32
        %tmp1 = load <2 x i32>* %A
        %tmp2 = load <2 x i32>* %B
        %tmp3 = call <2 x i32> @llvm.arm.neon.vqaddu.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
@@ -58,6 +64,8 @@ define <2 x i32> @vqaddu32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
 }
 
 define <1 x i64> @vqaddu64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
+;CHECK: vqaddu64:
+;CHECK: vqadd.u64
        %tmp1 = load <1 x i64>* %A
        %tmp2 = load <1 x i64>* %B
        %tmp3 = call <1 x i64> @llvm.arm.neon.vqaddu.v1i64(<1 x i64> %tmp1, <1 x i64> %tmp2)
@@ -65,6 +73,8 @@ define <1 x i64> @vqaddu64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
 }
 
 define <16 x i8> @vqaddQs8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
+;CHECK: vqaddQs8:
+;CHECK: vqadd.s8
        %tmp1 = load <16 x i8>* %A
        %tmp2 = load <16 x i8>* %B
        %tmp3 = call <16 x i8> @llvm.arm.neon.vqadds.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
@@ -72,6 +82,8 @@ define <16 x i8> @vqaddQs8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
 }
 
 define <8 x i16> @vqaddQs16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
+;CHECK: vqaddQs16:
+;CHECK: vqadd.s16
        %tmp1 = load <8 x i16>* %A
        %tmp2 = load <8 x i16>* %B
        %tmp3 = call <8 x i16> @llvm.arm.neon.vqadds.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
@@ -79,6 +91,8 @@ define <8 x i16> @vqaddQs16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
 }
 
 define <4 x i32> @vqaddQs32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
+;CHECK: vqaddQs32:
+;CHECK: vqadd.s32
        %tmp1 = load <4 x i32>* %A
        %tmp2 = load <4 x i32>* %B
        %tmp3 = call <4 x i32> @llvm.arm.neon.vqadds.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
@@ -86,6 +100,8 @@ define <4 x i32> @vqaddQs32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
 }
 
 define <2 x i64> @vqaddQs64(<2 x i64>* %A, <2 x i64>* %B) nounwind {
+;CHECK: vqaddQs64:
+;CHECK: vqadd.s64
        %tmp1 = load <2 x i64>* %A
        %tmp2 = load <2 x i64>* %B
        %tmp3 = call <2 x i64> @llvm.arm.neon.vqadds.v2i64(<2 x i64> %tmp1, <2 x i64> %tmp2)
@@ -93,6 +109,8 @@ define <2 x i64> @vqaddQs64(<2 x i64>* %A, <2 x i64>* %B) nounwind {
 }
 
 define <16 x i8> @vqaddQu8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
+;CHECK: vqaddQu8:
+;CHECK: vqadd.u8
        %tmp1 = load <16 x i8>* %A
        %tmp2 = load <16 x i8>* %B
        %tmp3 = call <16 x i8> @llvm.arm.neon.vqaddu.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
@@ -100,6 +118,8 @@ define <16 x i8> @vqaddQu8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
 }
 
 define <8 x i16> @vqaddQu16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
+;CHECK: vqaddQu16:
+;CHECK: vqadd.u16
        %tmp1 = load <8 x i16>* %A
        %tmp2 = load <8 x i16>* %B
        %tmp3 = call <8 x i16> @llvm.arm.neon.vqaddu.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
@@ -107,6 +127,8 @@ define <8 x i16> @vqaddQu16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
 }
 
 define <4 x i32> @vqaddQu32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
+;CHECK: vqaddQu32:
+;CHECK: vqadd.u32
        %tmp1 = load <4 x i32>* %A
        %tmp2 = load <4 x i32>* %B
        %tmp3 = call <4 x i32> @llvm.arm.neon.vqaddu.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
@@ -114,6 +136,8 @@ define <4 x i32> @vqaddQu32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
 }
 
 define <2 x i64> @vqaddQu64(<2 x i64>* %A, <2 x i64>* %B) nounwind {
+;CHECK: vqaddQu64:
+;CHECK: vqadd.u64
        %tmp1 = load <2 x i64>* %A
        %tmp2 = load <2 x i64>* %B
        %tmp3 = call <2 x i64> @llvm.arm.neon.vqaddu.v2i64(<2 x i64> %tmp1, <2 x i64> %tmp2)