llvm/test/Transforms/LoopVectorize/AArch64/mul-simplification.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 5
; RUN: opt -p loop-vectorize -S %s | FileCheck %s

target triple = "arm64-apple-macosx"

; Test case for https://github.com/llvm/llvm-project/issues/107015.
define i64 @mul_select_operand_known_1_via_scev() {
; CHECK-LABEL: define i64 @mul_select_operand_known_1_via_scev() {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[RED:%.*]] = phi i64 [ 12, %[[ENTRY]] ], [ [[RED_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[IV:%.*]] = phi i32 [ 0, %[[ENTRY]] ], [ [[IV_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[TMP1:%.*]] = and i32 [[IV]], 1
; CHECK-NEXT:    [[CMP1_I:%.*]] = icmp eq i32 [[TMP1]], 0
; CHECK-NEXT:    [[NARROW_I:%.*]] = select i1 [[CMP1_I]], i32 1, i32 [[IV]]
; CHECK-NEXT:    [[MUL:%.*]] = zext nneg i32 [[NARROW_I]] to i64
; CHECK-NEXT:    [[RED_NEXT]] = mul nsw i64 [[RED]], [[MUL]]
; CHECK-NEXT:    [[IV_NEXT]] = add nuw nsw i32 [[IV]], 1
; CHECK-NEXT:    [[EC:%.*]] = icmp eq i32 [[IV]], 1
; CHECK-NEXT:    br i1 [[EC]], label %[[EXIT:.*]], label %[[LOOP]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[RES:%.*]] = phi i64 [ [[RED_NEXT]], %[[LOOP]] ]
; CHECK-NEXT:    ret i64 [[RES]]
;
entry:
  br label %loop

loop:
  %red = phi i64 [ 12, %entry ], [ %red.next, %loop ]
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %0 = and i32 %iv, 1
  %cmp1.i = icmp eq i32 %0, 0
  %narrow.i = select i1 %cmp1.i, i32 1, i32 %iv
  %mul = zext nneg i32 %narrow.i to i64
  %red.next = mul nsw i64 %red, %mul
  %iv.next = add nuw nsw i32 %iv, 1
  %ec = icmp eq i32 %iv, 1
  br i1 %ec, label %exit, label %loop

exit:
  %res = phi i64 [ %red.next, %loop ]
  ret i64 %res
}

define i32 @add_reduction_select_operand_constant_but_non_uniform() {
; CHECK-LABEL: define i32 @add_reduction_select_operand_constant_but_non_uniform() {
; CHECK-NEXT:  [[ENTRY:.*]]:
; CHECK-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
; CHECK:       [[VECTOR_BODY]]:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i32 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_PHI:%.*]] = phi <4 x i32> [ <i32 42, i32 0, i32 0, i32 0>, %[[VECTOR_PH]] ], [ [[TMP2:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[VEC_PHI1:%.*]] = phi <4 x i32> [ zeroinitializer, %[[VECTOR_PH]] ], [ [[TMP1:%.*]], %[[VECTOR_BODY]] ]
; CHECK-NEXT:    [[TMP2]] = add <4 x i32> zeroinitializer, [[VEC_PHI]]
; CHECK-NEXT:    [[TMP1]] = add <4 x i32> zeroinitializer, [[VEC_PHI1]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i32 [[INDEX]], 8
; CHECK-NEXT:    [[TMP0:%.*]] = icmp eq i32 [[INDEX_NEXT]], 64
; CHECK-NEXT:    br i1 [[TMP0]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    [[BIN_RDX:%.*]] = add <4 x i32> [[TMP1]], [[TMP2]]
; CHECK-NEXT:    [[TMP3:%.*]] = call i32 @llvm.vector.reduce.add.v4i32(<4 x i32> [[BIN_RDX]])
; CHECK-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; CHECK:       [[SCALAR_PH]]:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i32 [ 64, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; CHECK-NEXT:    [[BC_MERGE_RDX:%.*]] = phi i32 [ [[TMP3]], %[[MIDDLE_BLOCK]] ], [ 42, %[[ENTRY]] ]
; CHECK-NEXT:    br label %[[LOOP:.*]]
; CHECK:       [[LOOP]]:
; CHECK-NEXT:    [[IV:%.*]] = phi i32 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[ADD2_REASS:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[RDX:%.*]] = phi i32 [ [[BC_MERGE_RDX]], %[[SCALAR_PH]] ], [ [[RDX_NEXT:%.*]], %[[LOOP]] ]
; CHECK-NEXT:    [[ADD2_REASS]] = add i32 [[IV]], 1
; CHECK-NEXT:    [[RDX_NEXT]] = add i32 0, [[RDX]]
; CHECK-NEXT:    [[CMP:%.*]] = icmp ult i32 [[ADD2_REASS]], 64
; CHECK-NEXT:    br i1 [[CMP]], label %[[LOOP]], label %[[EXIT]], !llvm.loop [[LOOP3:![0-9]+]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    [[ADD_LCSSA:%.*]] = phi i32 [ [[RDX_NEXT]], %[[LOOP]] ], [ [[TMP3]], %[[MIDDLE_BLOCK]] ]
; CHECK-NEXT:    ret i32 [[ADD_LCSSA]]
;
entry:
  br label %loop

loop:
  %iv = phi i32 [ 0, %entry ], [ %iv.next, %loop ]
  %rdx = phi i32 [ 42, %entry ], [ %rdx.next, %loop ]

  %iv.next = add i32 %iv, 1
  %rdx.next = add i32 0, %rdx

  %cmp = icmp ult i32 %iv.next, 64
  br i1 %cmp, label %loop, label %exit

exit:
  ret i32 %rdx.next
}
;.
; CHECK: [[LOOP0]] = distinct !{[[LOOP0]], [[META1:![0-9]+]], [[META2:![0-9]+]]}
; CHECK: [[META1]] = !{!"llvm.loop.isvectorized", i32 1}
; CHECK: [[META2]] = !{!"llvm.loop.unroll.runtime.disable"}
; CHECK: [[LOOP3]] = distinct !{[[LOOP3]], [[META2]], [[META1]]}
;.