| ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --filter-out-after "^scalar.ph" --version 6 |
| ; RUN: opt < %s -S -p loop-vectorize -force-vector-width=4 -enable-interleaved-mem-accesses=true | FileCheck %s |
| |
| ; If the member holding the insert position is dead, make sure we choose a new member to insert at that dominates all other members. |
| define void @dead_member_order(ptr noalias %src, ptr noalias %dst1, ptr noalias %dst2, i32 %N) { |
| ; CHECK-LABEL: define void @dead_member_order( |
| ; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr noalias [[DST1:%.*]], ptr noalias [[DST2:%.*]], i32 [[N:%.*]]) { |
| ; CHECK-NEXT: [[ENTRY:.*:]] |
| ; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[N]], 4 |
| ; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_SCEVCHECK:.*]] |
| ; CHECK: [[VECTOR_SCEVCHECK]]: |
| ; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1 |
| ; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i32 [[TMP0]], 0 |
| ; CHECK-NEXT: br i1 [[TMP1]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]] |
| ; CHECK: [[VECTOR_PH]]: |
| ; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[N]], 4 |
| ; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[N]], [[N_MOD_VF]] |
| ; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] |
| ; CHECK: [[VECTOR_BODY]]: |
| ; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] |
| ; CHECK-NEXT: [[TMP2:%.*]] = getelementptr [3 x i8], ptr [[SRC]], i32 [[INDEX]], i32 1 |
| ; CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[TMP2]], i32 -1 |
| ; CHECK-NEXT: [[WIDE_VEC:%.*]] = load <12 x i8>, ptr [[TMP3]], align 1 |
| ; CHECK-NEXT: [[STRIDED_VEC:%.*]] = shufflevector <12 x i8> [[WIDE_VEC]], <12 x i8> poison, <4 x i32> <i32 0, i32 3, i32 6, i32 9> |
| ; CHECK-NEXT: [[STRIDED_VEC1:%.*]] = shufflevector <12 x i8> [[WIDE_VEC]], <12 x i8> poison, <4 x i32> <i32 1, i32 4, i32 7, i32 10> |
| ; CHECK-NEXT: [[STRIDED_VEC2:%.*]] = shufflevector <12 x i8> [[WIDE_VEC]], <12 x i8> poison, <4 x i32> <i32 2, i32 5, i32 8, i32 11> |
| ; CHECK-NEXT: [[TMP4:%.*]] = getelementptr i8, ptr [[DST1]], i32 [[INDEX]] |
| ; CHECK-NEXT: store <4 x i8> [[STRIDED_VEC1]], ptr [[TMP4]], align 1 |
| ; CHECK-NEXT: [[TMP5:%.*]] = add <4 x i8> [[STRIDED_VEC1]], [[STRIDED_VEC]] |
| ; CHECK-NEXT: [[TMP6:%.*]] = getelementptr i8, ptr [[DST2]], i32 [[INDEX]] |
| ; CHECK-NEXT: store <4 x i8> [[TMP5]], ptr [[TMP6]], align 1 |
| ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 |
| ; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]] |
| ; CHECK-NEXT: br i1 [[TMP7]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]] |
| ; CHECK: [[MIDDLE_BLOCK]]: |
| ; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[N]], [[N_VEC]] |
| ; CHECK-NEXT: br i1 [[CMP_N]], [[EXIT:label %.*]], label %[[SCALAR_PH]] |
| ; CHECK: [[SCALAR_PH]]: |
| ; |
| entry: |
| br label %loop |
| |
| loop: |
| %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ] |
| |
| %gep.x = getelementptr [3 x i8], ptr %src, i32 %iv, i32 2 |
| %x = load i8, ptr %gep.x ; dead |
| |
| %gep.y = getelementptr [3 x i8], ptr %src, i32 %iv, i32 1 |
| %y = load i8, ptr %gep.y |
| |
| %gep.dst1 = getelementptr i8, ptr %dst1, i32 %iv |
| store i8 %y, ptr %gep.dst1 |
| |
| %gep.z = getelementptr [3 x i8], ptr %src, i32 %iv, i32 0 |
| %z = load i8, ptr %gep.z |
| |
| %res = add i8 %y, %z |
| %gep.dst2 = getelementptr i8, ptr %dst2, i32 %iv |
| store i8 %res, ptr %gep.dst2 |
| |
| %iv.next = add i32 %iv, 1 |
| %ec = icmp eq i32 %iv.next, %N |
| br i1 %ec, label %exit, label %loop |
| |
| exit: |
| ret void |
| } |
| |
| ; Make sure we intersect the fact that the dead load doesn't have !invariant.load |
| define void @dead_member_metadata(ptr noalias %src, ptr noalias %dst, i32 %N) { |
| ; CHECK-LABEL: define void @dead_member_metadata( |
| ; CHECK-SAME: ptr noalias [[SRC:%.*]], ptr noalias [[DST:%.*]], i32 [[N:%.*]]) { |
| ; CHECK-NEXT: [[ENTRY:.*:]] |
| ; CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i32 [[N]], 4 |
| ; CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label %[[SCALAR_PH:.*]], label %[[VECTOR_SCEVCHECK:.*]] |
| ; CHECK: [[VECTOR_SCEVCHECK]]: |
| ; CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1 |
| ; CHECK-NEXT: [[TMP1:%.*]] = icmp slt i32 [[TMP0]], 0 |
| ; CHECK-NEXT: br i1 [[TMP1]], label %[[SCALAR_PH]], label %[[VECTOR_PH:.*]] |
| ; CHECK: [[VECTOR_PH]]: |
| ; CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i32 [[N]], 4 |
| ; CHECK-NEXT: [[N_VEC:%.*]] = sub i32 [[N]], [[N_MOD_VF]] |
| ; CHECK-NEXT: br label %[[VECTOR_BODY:.*]] |
| ; CHECK: [[VECTOR_BODY]]: |
| ; CHECK-NEXT: [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ] |
| ; CHECK-NEXT: [[TMP2:%.*]] = getelementptr [2 x i8], ptr [[SRC]], i32 [[INDEX]], i32 1 |
| ; CHECK-NEXT: [[TMP5:%.*]] = getelementptr i8, ptr [[TMP2]], i32 -1 |
| ; CHECK-NEXT: [[WIDE_VEC:%.*]] = load <8 x i8>, ptr [[TMP5]], align 1 |
| ; CHECK-NEXT: [[STRIDED_VEC:%.*]] = shufflevector <8 x i8> [[WIDE_VEC]], <8 x i8> poison, <4 x i32> <i32 0, i32 2, i32 4, i32 6> |
| ; CHECK-NEXT: [[STRIDED_VEC1:%.*]] = shufflevector <8 x i8> [[WIDE_VEC]], <8 x i8> poison, <4 x i32> <i32 1, i32 3, i32 5, i32 7> |
| ; CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[DST]], i32 [[INDEX]] |
| ; CHECK-NEXT: store <4 x i8> [[STRIDED_VEC1]], ptr [[TMP3]], align 1 |
| ; CHECK-NEXT: [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 4 |
| ; CHECK-NEXT: [[TMP4:%.*]] = icmp eq i32 [[INDEX_NEXT]], [[N_VEC]] |
| ; CHECK-NEXT: br i1 [[TMP4]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]] |
| ; CHECK: [[MIDDLE_BLOCK]]: |
| ; CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i32 [[N]], [[N_VEC]] |
| ; CHECK-NEXT: br i1 [[CMP_N]], [[EXIT:label %.*]], label %[[SCALAR_PH]] |
| ; CHECK: [[SCALAR_PH]]: |
| ; |
| entry: |
| br label %loop |
| |
| loop: |
| %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ] |
| |
| %gep.y = getelementptr [2 x i8], ptr %src, i32 %iv, i32 1 |
| %y = load i8, ptr %gep.y, !invariant.load !{} |
| %gep.dst = getelementptr i8, ptr %dst, i32 %iv |
| store i8 %y, ptr %gep.dst |
| |
| %gep.x = getelementptr [2 x i8], ptr %src, i32 %iv, i32 0 |
| %x = load i8, ptr %gep.x ; dead |
| |
| %iv.next = add i32 %iv, 1 |
| %ec = icmp eq i32 %iv.next, %N |
| br i1 %ec, label %exit, label %loop |
| |
| exit: |
| ret void |
| } |