3 files changed, 81 insertions, 3 deletions
diff --git a/llvm/test/CodeGen/ARM/O3-pipeline.ll b/llvm/test/CodeGen/ARM/O3-pipeline.ll
index ec96f055a05..c528f5d0cee 100644
--- a/llvm/test/CodeGen/ARM/O3-pipeline.ll
+++ b/llvm/test/CodeGen/ARM/O3-pipeline.ll
@@ -37,8 +37,7 @@
 ; CHECK-NEXT:      Scalar Evolution Analysis
 ; CHECK-NEXT:      Basic Alias Analysis (stateless AA impl)
 ; CHECK-NEXT:      Function Alias Analysis Results
-; CHECK-NEXT:      Loop Pass Manager
-; CHECK-NEXT:        Transform loops to use DSP intrinsics
+; CHECK-NEXT:      Transform functions to use DSP intrinsics
 ; CHECK-NEXT:      Interleaved Access Pass
 ; CHECK-NEXT:      ARM IR optimizations
 ; CHECK-NEXT:      Dominator Tree Construction
diff --git a/llvm/test/CodeGen/ARM/ParallelDSP/blocks.ll b/llvm/test/CodeGen/ARM/ParallelDSP/blocks.ll
new file mode 100644
index 00000000000..8d26f61eb6c
--- /dev/null
+++ b/llvm/test/CodeGen/ARM/ParallelDSP/blocks.ll
@@ -0,0 +1,79 @@
+; RUN: opt -arm-parallel-dsp -mtriple=armv7-a -S %s -o - | FileCheck %s
+
+; CHECK-LABEL: single_block
+; CHECK: [[CAST_A:%[^ ]+]] = bitcast i16* %a to i32*
+; CHECK: [[A:%[^ ]+]] = load i32, i32* [[CAST_A]]
+; CHECK: [[CAST_B:%[^ ]+]] = bitcast i16* %b to i32*
+; CHECK: [[B:%[^ ]+]] = load i32, i32* [[CAST_B]]
+; CHECK  call i32 @llvm.arm.smlad(i32 [[A]], i32 [[B]], i32 %acc)
+define i32 @single_block(i16* %a, i16* %b, i32 %acc) {
+entry:
+  %ld.a.0 = load i16, i16* %a
+  %sext.a.0 = sext i16 %ld.a.0 to i32
+  %ld.b.0 = load i16, i16* %b
+  %sext.b.0 = sext i16 %ld.b.0 to i32
+  %mul.0 = mul i32 %sext.a.0, %sext.b.0
+  %addr.a.1 = getelementptr i16, i16* %a, i32 1
+  %addr.b.1 = getelementptr i16, i16* %b, i32 1
+  %ld.a.1 = load i16, i16* %addr.a.1
+  %sext.a.1 = sext i16 %ld.a.1 to i32
+  %ld.b.1 = load i16, i16* %addr.b.1
+  %sext.b.1 = sext i16 %ld.b.1 to i32
+  %mul.1 = mul i32 %sext.a.1, %sext.b.1
+  %add = add i32 %mul.0, %mul.1
+  %res = add i32 %add, %acc
+  ret i32 %res
+}
+
+; CHECK-LABEL: multi_block
+; CHECK: [[CAST_A:%[^ ]+]] = bitcast i16* %a to i32*
+; CHECK: [[A:%[^ ]+]] = load i32, i32* [[CAST_A]]
+; CHECK: [[CAST_B:%[^ ]+]] = bitcast i16* %b to i32*
+; CHECK: [[B:%[^ ]+]] = load i32, i32* [[CAST_B]]
+; CHECK  call i32 @llvm.arm.smlad(i32 [[A]], i32 [[B]], i32 0)
+define i32 @multi_block(i16* %a, i16* %b, i32 %acc) {
+entry:
+  %ld.a.0 = load i16, i16* %a
+  %sext.a.0 = sext i16 %ld.a.0 to i32
+  %ld.b.0 = load i16, i16* %b
+  %sext.b.0 = sext i16 %ld.b.0 to i32
+  %mul.0 = mul i32 %sext.a.0, %sext.b.0
+  %addr.a.1 = getelementptr i16, i16* %a, i32 1
+  %addr.b.1 = getelementptr i16, i16* %b, i32 1
+  %ld.a.1 = load i16, i16* %addr.a.1
+  %sext.a.1 = sext i16 %ld.a.1 to i32
+  %ld.b.1 = load i16, i16* %addr.b.1
+  %sext.b.1 = sext i16 %ld.b.1 to i32
+  %mul.1 = mul i32 %sext.a.1, %sext.b.1
+  %add = add i32 %mul.0, %mul.1
+  br label %bb.1
+
+bb.1:
+  %res = add i32 %add, %acc
+  ret i32 %res
+}
+
+; CHECK-LABEL: multi_block_1
+; CHECK-NOT: call i32 @llvm.arm.smlad
+define i32 @multi_block_1(i16* %a, i16* %b, i32 %acc) {
+entry:
+  %ld.a.0 = load i16, i16* %a
+  %sext.a.0 = sext i16 %ld.a.0 to i32
+  %ld.b.0 = load i16, i16* %b
+  %sext.b.0 = sext i16 %ld.b.0 to i32
+  %mul.0 = mul i32 %sext.a.0, %sext.b.0
+  br label %bb.1
+
+bb.1:
+  %addr.a.1 = getelementptr i16, i16* %a, i32 1
+  %addr.b.1 = getelementptr i16, i16* %b, i32 1
+  %ld.a.1 = load i16, i16* %addr.a.1
+  %sext.a.1 = sext i16 %ld.a.1 to i32
+  %ld.b.1 = load i16, i16* %addr.b.1
+  %sext.b.1 = sext i16 %ld.b.1 to i32
+  %mul.1 = mul i32 %sext.a.1, %sext.b.1
+  %add = add i32 %mul.0, %mul.1
+  %res = add i32 %add, %acc
+  ret i32 %res
+}
+
diff --git a/llvm/test/CodeGen/ARM/ParallelDSP/smlad12.ll b/llvm/test/CodeGen/ARM/ParallelDSP/smlad12.ll
index d4e09ca3fbb..637fc3d3704 100644
--- a/llvm/test/CodeGen/ARM/ParallelDSP/smlad12.ll
+++ b/llvm/test/CodeGen/ARM/ParallelDSP/smlad12.ll
@@ -2,7 +2,7 @@
 ;
 ; The loop header is not the loop latch.
 ;
-; CHECK-NOT:  call i32 @llvm.arm.smlad
+; CHECK:  call i32 @llvm.arm.smlad
 ;
 define dso_local i32 @test(i32 %arg, i32* nocapture readnone %arg1, i16* nocapture readonly %arg2, i16* nocapture readonly %arg3) {
 entry: