; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --check-globals none --version 5
; RUN: opt -passes='loop-vectorize,verify<loops>' -force-vector-width=8 -force-vector-interleave=1 -enable-early-exit-vectorization -S %s | FileCheck --check-prefixes=VF8UF1 %s
; RUN: opt -passes='loop-vectorize,verify<loops>' -force-vector-width=8 -force-vector-interleave=2 -enable-early-exit-vectorization -S %s | FileCheck --check-prefixes=VF8UF2 %s
; RUN: opt -passes='loop-vectorize,verify<loops>' -force-vector-width=16 -force-vector-interleave=1 -enable-early-exit-vectorization -S %s | FileCheck --check-prefixes=VF16UF1 %s

target datalayout = "e-m:o-i64:64-i128:128-n32:64-S128"

; Check if the vector loop condition can be simplified to true for a given
; VF/IC combination.
define i8 @test_early_exit_max_tc_less_than_16(ptr dereferenceable(16) %A) nosync nofree {
; VF8UF1-LABEL: define i8 @test_early_exit_max_tc_less_than_16(
; VF8UF1-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0:[0-9]+]] {
; VF8UF1-NEXT:  [[ENTRY:.*]]:
; VF8UF1-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF8UF1:       [[VECTOR_PH]]:
; VF8UF1-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF8UF1:       [[VECTOR_BODY]]:
; VF8UF1-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; VF8UF1-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[INDEX]]
; VF8UF1-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF8UF1-NEXT:    [[WIDE_LOAD:%.*]] = load <8 x i8>, ptr [[TMP2]], align 1
; VF8UF1-NEXT:    [[TMP3:%.*]] = icmp eq <8 x i8> [[WIDE_LOAD]], zeroinitializer
; VF8UF1-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8
; VF8UF1-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> [[TMP3]])
; VF8UF1-NEXT:    [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16
; VF8UF1-NEXT:    [[TMP6:%.*]] = or i1 [[TMP4]], [[TMP5]]
; VF8UF1-NEXT:    br i1 [[TMP6]], label %[[MIDDLE_SPLIT:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; VF8UF1:       [[MIDDLE_SPLIT]]:
; VF8UF1-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF8UF1:       [[MIDDLE_BLOCK]]:
; VF8UF1-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF8UF1:       [[VECTOR_EARLY_EXIT]]:
; VF8UF1-NEXT:    br label %[[EXIT]]
; VF8UF1:       [[SCALAR_PH]]:
; VF8UF1-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF8UF1-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF8UF1:       [[LOOP_HEADER]]:
; VF8UF1-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF8UF1-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF8UF1-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF8UF1-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF8UF1-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF8UF1:       [[LOOP_LATCH]]:
; VF8UF1-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF8UF1-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF8UF1-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP3:![0-9]+]]
; VF8UF1:       [[EXIT]]:
; VF8UF1-NEXT:    [[RES:%.*]] = phi i8 [ 0, %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ 0, %[[VECTOR_EARLY_EXIT]] ]
; VF8UF1-NEXT:    ret i8 [[RES]]
;
; VF8UF2-LABEL: define i8 @test_early_exit_max_tc_less_than_16(
; VF8UF2-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0:[0-9]+]] {
; VF8UF2-NEXT:  [[ENTRY:.*]]:
; VF8UF2-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF8UF2:       [[VECTOR_PH]]:
; VF8UF2-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF8UF2:       [[VECTOR_BODY]]:
; VF8UF2-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 0
; VF8UF2-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF8UF2-NEXT:    [[WIDE_LOAD:%.*]] = load <8 x i8>, ptr [[TMP2]], align 1
; VF8UF2-NEXT:    [[TMP3:%.*]] = icmp eq <8 x i8> [[WIDE_LOAD]], zeroinitializer
; VF8UF2-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> [[TMP3]])
; VF8UF2-NEXT:    br label %[[MIDDLE_SPLIT:.*]]
; VF8UF2:       [[MIDDLE_SPLIT]]:
; VF8UF2-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF8UF2:       [[MIDDLE_BLOCK]]:
; VF8UF2-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF8UF2:       [[VECTOR_EARLY_EXIT]]:
; VF8UF2-NEXT:    br label %[[EXIT]]
; VF8UF2:       [[SCALAR_PH]]:
; VF8UF2-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF8UF2-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF8UF2:       [[LOOP_HEADER]]:
; VF8UF2-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF8UF2-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF8UF2-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF8UF2-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF8UF2-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF8UF2:       [[LOOP_LATCH]]:
; VF8UF2-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF8UF2-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF8UF2-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP0:![0-9]+]]
; VF8UF2:       [[EXIT]]:
; VF8UF2-NEXT:    [[RES:%.*]] = phi i8 [ 0, %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ 0, %[[VECTOR_EARLY_EXIT]] ]
; VF8UF2-NEXT:    ret i8 [[RES]]
;
; VF16UF1-LABEL: define i8 @test_early_exit_max_tc_less_than_16(
; VF16UF1-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0:[0-9]+]] {
; VF16UF1-NEXT:  [[ENTRY:.*]]:
; VF16UF1-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF16UF1:       [[VECTOR_PH]]:
; VF16UF1-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF16UF1:       [[VECTOR_BODY]]:
; VF16UF1-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 0
; VF16UF1-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF16UF1-NEXT:    [[WIDE_LOAD:%.*]] = load <16 x i8>, ptr [[TMP2]], align 1
; VF16UF1-NEXT:    [[TMP3:%.*]] = icmp eq <16 x i8> [[WIDE_LOAD]], zeroinitializer
; VF16UF1-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v16i1(<16 x i1> [[TMP3]])
; VF16UF1-NEXT:    br label %[[MIDDLE_SPLIT:.*]]
; VF16UF1:       [[MIDDLE_SPLIT]]:
; VF16UF1-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF16UF1:       [[MIDDLE_BLOCK]]:
; VF16UF1-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF16UF1:       [[VECTOR_EARLY_EXIT]]:
; VF16UF1-NEXT:    br label %[[EXIT]]
; VF16UF1:       [[SCALAR_PH]]:
; VF16UF1-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF16UF1-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF16UF1:       [[LOOP_HEADER]]:
; VF16UF1-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF16UF1-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF16UF1-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF16UF1-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF16UF1-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF16UF1:       [[LOOP_LATCH]]:
; VF16UF1-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF16UF1-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF16UF1-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP0:![0-9]+]]
; VF16UF1:       [[EXIT]]:
; VF16UF1-NEXT:    [[RES:%.*]] = phi i8 [ 0, %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ 0, %[[VECTOR_EARLY_EXIT]] ]
; VF16UF1-NEXT:    ret i8 [[RES]]
;
entry:
  br label %loop.header

loop.header:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop.latch ]
  %p.src = getelementptr inbounds i8, ptr %A, i64 %iv
  %l = load i8, ptr %p.src, align 1
  %c = icmp eq i8 %l, 0
  br i1 %c, label %exit, label %loop.latch

loop.latch:
  %iv.next = add nsw i64 %iv, 1
  %cmp = icmp eq i64 %iv.next, 16
  br i1 %cmp, label %exit, label %loop.header

exit:
  %res = phi i8 [ 0, %loop.header ], [ 1, %loop.latch ]
  ret i8 %res
}

define i64 @test_early_exit_max_tc_less_than_16_with_iv_used_outside(ptr dereferenceable(16) %A) nosync nofree {
; VF8UF1-LABEL: define i64 @test_early_exit_max_tc_less_than_16_with_iv_used_outside(
; VF8UF1-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0]] {
; VF8UF1-NEXT:  [[ENTRY:.*]]:
; VF8UF1-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF8UF1:       [[VECTOR_PH]]:
; VF8UF1-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF8UF1:       [[VECTOR_BODY]]:
; VF8UF1-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, %[[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], %[[VECTOR_BODY]] ]
; VF8UF1-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[INDEX]]
; VF8UF1-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF8UF1-NEXT:    [[WIDE_LOAD:%.*]] = load <8 x i8>, ptr [[TMP2]], align 1
; VF8UF1-NEXT:    [[TMP3:%.*]] = icmp eq <8 x i8> [[WIDE_LOAD]], zeroinitializer
; VF8UF1-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 8
; VF8UF1-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> [[TMP3]])
; VF8UF1-NEXT:    [[TMP5:%.*]] = icmp eq i64 [[INDEX_NEXT]], 16
; VF8UF1-NEXT:    [[TMP6:%.*]] = or i1 [[TMP4]], [[TMP5]]
; VF8UF1-NEXT:    br i1 [[TMP6]], label %[[MIDDLE_SPLIT:.*]], label %[[VECTOR_BODY]], !llvm.loop [[LOOP4:![0-9]+]]
; VF8UF1:       [[MIDDLE_SPLIT]]:
; VF8UF1-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF8UF1:       [[MIDDLE_BLOCK]]:
; VF8UF1-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF8UF1:       [[VECTOR_EARLY_EXIT]]:
; VF8UF1-NEXT:    [[FIRST_ACTIVE_LANE:%.*]] = call i64 @llvm.experimental.cttz.elts.i64.v8i1(<8 x i1> [[TMP3]], i1 true)
; VF8UF1-NEXT:    [[TMP8:%.*]] = add i64 [[INDEX]], [[FIRST_ACTIVE_LANE]]
; VF8UF1-NEXT:    br label %[[EXIT]]
; VF8UF1:       [[SCALAR_PH]]:
; VF8UF1-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF8UF1-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF8UF1:       [[LOOP_HEADER]]:
; VF8UF1-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF8UF1-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF8UF1-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF8UF1-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF8UF1-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF8UF1:       [[LOOP_LATCH]]:
; VF8UF1-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF8UF1-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF8UF1-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP5:![0-9]+]]
; VF8UF1:       [[EXIT]]:
; VF8UF1-NEXT:    [[RES:%.*]] = phi i64 [ [[IV1]], %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ [[TMP8]], %[[VECTOR_EARLY_EXIT]] ]
; VF8UF1-NEXT:    ret i64 [[RES]]
;
; VF8UF2-LABEL: define i64 @test_early_exit_max_tc_less_than_16_with_iv_used_outside(
; VF8UF2-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0]] {
; VF8UF2-NEXT:  [[ENTRY:.*]]:
; VF8UF2-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF8UF2:       [[VECTOR_PH]]:
; VF8UF2-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF8UF2:       [[VECTOR_BODY]]:
; VF8UF2-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 0
; VF8UF2-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF8UF2-NEXT:    [[WIDE_LOAD:%.*]] = load <8 x i8>, ptr [[TMP2]], align 1
; VF8UF2-NEXT:    [[TMP3:%.*]] = icmp eq <8 x i8> [[WIDE_LOAD]], zeroinitializer
; VF8UF2-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v8i1(<8 x i1> [[TMP3]])
; VF8UF2-NEXT:    br label %[[MIDDLE_SPLIT:.*]]
; VF8UF2:       [[MIDDLE_SPLIT]]:
; VF8UF2-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF8UF2:       [[MIDDLE_BLOCK]]:
; VF8UF2-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF8UF2:       [[VECTOR_EARLY_EXIT]]:
; VF8UF2-NEXT:    [[FIRST_ACTIVE_LANE:%.*]] = call i64 @llvm.experimental.cttz.elts.i64.v8i1(<8 x i1> [[TMP3]], i1 true)
; VF8UF2-NEXT:    [[TMP5:%.*]] = add i64 0, [[FIRST_ACTIVE_LANE]]
; VF8UF2-NEXT:    br label %[[EXIT]]
; VF8UF2:       [[SCALAR_PH]]:
; VF8UF2-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF8UF2-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF8UF2:       [[LOOP_HEADER]]:
; VF8UF2-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF8UF2-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF8UF2-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF8UF2-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF8UF2-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF8UF2:       [[LOOP_LATCH]]:
; VF8UF2-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF8UF2-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF8UF2-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP3:![0-9]+]]
; VF8UF2:       [[EXIT]]:
; VF8UF2-NEXT:    [[RES:%.*]] = phi i64 [ [[IV1]], %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ [[TMP5]], %[[VECTOR_EARLY_EXIT]] ]
; VF8UF2-NEXT:    ret i64 [[RES]]
;
; VF16UF1-LABEL: define i64 @test_early_exit_max_tc_less_than_16_with_iv_used_outside(
; VF16UF1-SAME: ptr dereferenceable(16) [[A:%.*]]) #[[ATTR0]] {
; VF16UF1-NEXT:  [[ENTRY:.*]]:
; VF16UF1-NEXT:    br i1 false, label %[[SCALAR_PH:.*]], label %[[VECTOR_PH:.*]]
; VF16UF1:       [[VECTOR_PH]]:
; VF16UF1-NEXT:    br label %[[VECTOR_BODY:.*]]
; VF16UF1:       [[VECTOR_BODY]]:
; VF16UF1-NEXT:    [[P_SRC:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 0
; VF16UF1-NEXT:    [[TMP2:%.*]] = getelementptr inbounds i8, ptr [[P_SRC]], i32 0
; VF16UF1-NEXT:    [[WIDE_LOAD:%.*]] = load <16 x i8>, ptr [[TMP2]], align 1
; VF16UF1-NEXT:    [[TMP3:%.*]] = icmp eq <16 x i8> [[WIDE_LOAD]], zeroinitializer
; VF16UF1-NEXT:    [[TMP4:%.*]] = call i1 @llvm.vector.reduce.or.v16i1(<16 x i1> [[TMP3]])
; VF16UF1-NEXT:    br label %[[MIDDLE_SPLIT:.*]]
; VF16UF1:       [[MIDDLE_SPLIT]]:
; VF16UF1-NEXT:    br i1 [[TMP4]], label %[[VECTOR_EARLY_EXIT:.*]], label %[[MIDDLE_BLOCK:.*]]
; VF16UF1:       [[MIDDLE_BLOCK]]:
; VF16UF1-NEXT:    br i1 true, label %[[EXIT:.*]], label %[[SCALAR_PH]]
; VF16UF1:       [[VECTOR_EARLY_EXIT]]:
; VF16UF1-NEXT:    [[FIRST_ACTIVE_LANE:%.*]] = call i64 @llvm.experimental.cttz.elts.i64.v16i1(<16 x i1> [[TMP3]], i1 true)
; VF16UF1-NEXT:    [[TMP5:%.*]] = add i64 0, [[FIRST_ACTIVE_LANE]]
; VF16UF1-NEXT:    br label %[[EXIT]]
; VF16UF1:       [[SCALAR_PH]]:
; VF16UF1-NEXT:    [[BC_RESUME_VAL:%.*]] = phi i64 [ 16, %[[MIDDLE_BLOCK]] ], [ 0, %[[ENTRY]] ]
; VF16UF1-NEXT:    br label %[[LOOP_HEADER:.*]]
; VF16UF1:       [[LOOP_HEADER]]:
; VF16UF1-NEXT:    [[IV1:%.*]] = phi i64 [ [[BC_RESUME_VAL]], %[[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], %[[LOOP_LATCH:.*]] ]
; VF16UF1-NEXT:    [[P_SRC1:%.*]] = getelementptr inbounds i8, ptr [[A]], i64 [[IV1]]
; VF16UF1-NEXT:    [[L:%.*]] = load i8, ptr [[P_SRC1]], align 1
; VF16UF1-NEXT:    [[C:%.*]] = icmp eq i8 [[L]], 0
; VF16UF1-NEXT:    br i1 [[C]], label %[[EXIT]], label %[[LOOP_LATCH]]
; VF16UF1:       [[LOOP_LATCH]]:
; VF16UF1-NEXT:    [[IV_NEXT]] = add nsw i64 [[IV1]], 1
; VF16UF1-NEXT:    [[CMP:%.*]] = icmp eq i64 [[IV_NEXT]], 16
; VF16UF1-NEXT:    br i1 [[CMP]], label %[[EXIT]], label %[[LOOP_HEADER]], !llvm.loop [[LOOP3:![0-9]+]]
; VF16UF1:       [[EXIT]]:
; VF16UF1-NEXT:    [[RES:%.*]] = phi i64 [ [[IV1]], %[[LOOP_HEADER]] ], [ 1, %[[LOOP_LATCH]] ], [ 1, %[[MIDDLE_BLOCK]] ], [ [[TMP5]], %[[VECTOR_EARLY_EXIT]] ]
; VF16UF1-NEXT:    ret i64 [[RES]]
;
entry:
  br label %loop.header

loop.header:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %loop.latch ]
  %p.src = getelementptr inbounds i8, ptr %A, i64 %iv
  %l = load i8, ptr %p.src, align 1
  %c = icmp eq i8 %l, 0
  br i1 %c, label %exit, label %loop.latch

loop.latch:
  %iv.next = add nsw i64 %iv, 1
  %cmp = icmp eq i64 %iv.next, 16
  br i1 %cmp, label %exit, label %loop.header

exit:
  %res = phi i64 [ %iv, %loop.header ], [ 1, %loop.latch ]
  ret i64 %res
}