llvm/test/Transforms/LoopVectorize/constant-fold-commutative-and.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 6
; RUN: opt -passes=loop-vectorize -force-vector-width=2 \
; RUN:   -tail-folding-policy=prefer-fold-tail -S %s \
; RUN:   | FileCheck %s

; Test that we constant fold logical and with swapped operand order.

define void @constant_fold_commutative_and(ptr %ptr.n, ptr noalias %p, i1 %cond) {
; CHECK-LABEL: define void @constant_fold_commutative_and(
; CHECK-SAME: ptr [[PTR_N:%.*]], ptr noalias [[P:%.*]], i1 [[COND:%.*]]) {
; CHECK-NEXT:  [[ENTRY:.*:]]
; CHECK-NEXT:    br label %[[VECTOR_PH:.*]]
; CHECK:       [[VECTOR_PH]]:
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT:%.*]] = insertelement <2 x i1> poison, i1 [[COND]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT:%.*]] = shufflevector <2 x i1> [[BROADCAST_SPLATINSERT]], <2 x i1> poison, <2 x i32> zeroinitializer
; CHECK-NEXT:    [[TMP0:%.*]] = xor <2 x i1> [[BROADCAST_SPLAT]], splat (i1 true)
; CHECK-NEXT:    br label %[[VECTOR_BODY:.*]]
; CHECK:       [[VECTOR_BODY]]:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[PRED_STORE_CONTINUE5:.*]] ]
; CHECK-NEXT:    [[VEC_IND:%.*]] = phi <2 x i8> [ <i8 0, i8 1>, %[[VECTOR_PH]] ], [ [[VEC_IND_NEXT:%.*]], %[[PRED_STORE_CONTINUE5]] ]
; CHECK-NEXT:    [[TMP1:%.*]] = icmp ule <2 x i8> [[VEC_IND]], splat (i8 16)
; CHECK-NEXT:    [[TMP2:%.*]] = load i64, ptr [[PTR_N]], align 4
; CHECK-NEXT:    [[BROADCAST_SPLATINSERT1:%.*]] = insertelement <2 x i64> poison, i64 [[TMP2]], i64 0
; CHECK-NEXT:    [[BROADCAST_SPLAT2:%.*]] = shufflevector <2 x i64> [[BROADCAST_SPLATINSERT1]], <2 x i64> poison, <2 x i32> zeroinitializer
; CHECK-NEXT:    [[TMP3:%.*]] = icmp ult <2 x i64> [[BROADCAST_SPLAT2]], splat (i64 4)
; CHECK-NEXT:    [[TMP4:%.*]] = icmp ult <2 x i64> [[BROADCAST_SPLAT2]], splat (i64 7)
; CHECK-NEXT:    [[TMP5:%.*]] = select <2 x i1> [[TMP0]], <2 x i1> [[TMP3]], <2 x i1> zeroinitializer
; CHECK-NEXT:    [[PREDPHI:%.*]] = select <2 x i1> [[TMP5]], <2 x i1> splat (i1 true), <2 x i1> [[TMP4]]
; CHECK-NEXT:    [[PREDPHI3:%.*]] = select i1 [[COND]], <2 x i1> zeroinitializer, <2 x i1> [[PREDPHI]]
; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x i1> [[TMP1]], i64 0
; CHECK-NEXT:    br i1 [[TMP6]], label %[[PRED_STORE_IF:.*]], label %[[PRED_STORE_CONTINUE:.*]]
; CHECK:       [[PRED_STORE_IF]]:
; CHECK-NEXT:    [[TMP8:%.*]] = getelementptr i8, ptr [[P]], i64 [[INDEX]]
; CHECK-NEXT:    [[TMP9:%.*]] = extractelement <2 x i1> [[PREDPHI3]], i64 0
; CHECK-NEXT:    store i1 [[TMP9]], ptr [[TMP8]], align 1
; CHECK-NEXT:    br label %[[PRED_STORE_CONTINUE]]
; CHECK:       [[PRED_STORE_CONTINUE]]:
; CHECK-NEXT:    [[TMP10:%.*]] = extractelement <2 x i1> [[TMP1]], i64 1
; CHECK-NEXT:    br i1 [[TMP10]], label %[[PRED_STORE_IF4:.*]], label %[[PRED_STORE_CONTINUE5]]
; CHECK:       [[PRED_STORE_IF4]]:
; CHECK-NEXT:    [[TMP11:%.*]] = add i64 [[INDEX]], 1
; CHECK-NEXT:    [[TMP20:%.*]] = getelementptr i8, ptr [[P]], i64 [[TMP11]]
; CHECK-NEXT:    [[TMP21:%.*]] = extractelement <2 x i1> [[PREDPHI3]], i64 1
; CHECK-NEXT:    store i1 [[TMP21]], ptr [[TMP20]], align 1
; CHECK-NEXT:    br label %[[PRED_STORE_CONTINUE5]]
; CHECK:       [[PRED_STORE_CONTINUE5]]:
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2
; CHECK-NEXT:    [[VEC_IND_NEXT]] = add nuw <2 x i8> [[VEC_IND]], splat (i8 2)
; CHECK-NEXT:    [[TMP15:%.*]] = icmp eq i64 [[INDEX_NEXT]], 18
; CHECK-NEXT:    br i1 [[TMP15]], label %[[MIDDLE_BLOCK:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; CHECK:       [[MIDDLE_BLOCK]]:
; CHECK-NEXT:    br label %[[EXIT:.*]]
; CHECK:       [[EXIT]]:
; CHECK-NEXT:    ret void
;
entry:
  br label %ph

ph:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %latch ]
  %n = load i64, ptr %ptr.n
  br i1 %cond, label %latch, label %pred.1

pred.1:
  %cmp = icmp ult i64 %n, 4
  br i1 %cmp, label %latch, label %pred.2

pred.2:
  %cmp.2 = icmp ult i64 %n, 7
  br label %latch

latch:
  %merge.val = phi i1 [ 0, %ph ], [ 1, %pred.1 ], [ %cmp.2, %pred.2 ]
  %gep = getelementptr i8, ptr %p, i64 %iv
  store i1 %merge.val, ptr %gep
  %iv.next = add i64 %iv, 1
  %ec = icmp eq i64 %iv, 16
  br i1 %ec, label %exit, label %ph

exit:
  ret void
}