llvm/test/Transforms/LoopVectorize/opaque-ptr.ll

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -S -passes=loop-vectorize -force-vector-width=2 < %s | FileCheck %s

define void @test_ptr_iv_no_inbounds(ptr %p1.start, ptr %p2.start, ptr %p1.end) {
; CHECK-LABEL: @test_ptr_iv_no_inbounds(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[P1_END1:%.*]] = ptrtoaddr ptr [[P1_END:%.*]] to i64
; CHECK-NEXT:    [[TMP8:%.*]] = ptrtoaddr ptr [[P1_START:%.*]] to i64
; CHECK-NEXT:    [[TMP0:%.*]] = add i64 [[TMP8]], -4
; CHECK-NEXT:    [[TMP1:%.*]] = sub i64 [[TMP0]], [[P1_END1]]
; CHECK-NEXT:    [[TMP2:%.*]] = lshr i64 [[TMP1]], 2
; CHECK-NEXT:    [[TMP3:%.*]] = add nuw nsw i64 [[TMP2]], 1
; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP3]], 2
; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; CHECK:       vector.scevcheck:
; CHECK-NEXT:    [[TMP4:%.*]] = trunc i64 [[TMP8]] to i2
; CHECK-NEXT:    [[TMP5:%.*]] = trunc i64 [[P1_END1]] to i2
; CHECK-NEXT:    [[TMP6:%.*]] = sub i2 [[TMP4]], [[TMP5]]
; CHECK-NEXT:    [[TMP7:%.*]] = zext i2 [[TMP6]] to i64
; CHECK-NEXT:    [[IDENT_CHECK:%.*]] = icmp ne i64 [[TMP7]], 0
; CHECK-NEXT:    br i1 [[IDENT_CHECK]], label [[SCALAR_PH]], label [[VECTOR_MEMCHECK:%.*]]
; CHECK:       vector.memcheck:
; CHECK-NEXT:    [[TMP11:%.*]] = and i64 [[TMP1]], -4
; CHECK-NEXT:    [[TMP12:%.*]] = add i64 [[TMP11]], 4
; CHECK-NEXT:    [[SCEVGEP:%.*]] = getelementptr i8, ptr [[P1_END]], i64 [[TMP12]]
; CHECK-NEXT:    [[SCEVGEP5:%.*]] = getelementptr i8, ptr [[P2_START:%.*]], i64 [[TMP12]]
; CHECK-NEXT:    [[BOUND0:%.*]] = icmp ult ptr [[P1_END]], [[SCEVGEP5]]
; CHECK-NEXT:    [[BOUND1:%.*]] = icmp ult ptr [[P2_START]], [[SCEVGEP]]
; CHECK-NEXT:    [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; CHECK-NEXT:    br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    [[N_MOD_VF:%.*]] = and i64 [[TMP3]], 1
; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP3]], [[N_MOD_VF]]
; CHECK-NEXT:    [[TMP13:%.*]] = shl i64 [[N_VEC]], 2
; CHECK-NEXT:    [[TMP16:%.*]] = getelementptr i8, ptr [[P1_END]], i64 [[TMP13]]
; CHECK-NEXT:    [[IND_END8:%.*]] = getelementptr i8, ptr [[P2_START]], i64 [[TMP13]]
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = shl i64 [[INDEX]], 2
; CHECK-NEXT:    [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[P1_END]], i64 [[OFFSET_IDX]]
; CHECK-NEXT:    [[NEXT_GEP8:%.*]] = getelementptr i8, ptr [[P2_START]], i64 [[OFFSET_IDX]]
; CHECK-NEXT:    [[WIDE_LOAD:%.*]] = load <2 x float>, ptr [[NEXT_GEP]], align 4, !alias.scope [[META0:![0-9]+]], !noalias [[META3:![0-9]+]]
; CHECK-NEXT:    [[WIDE_LOAD12:%.*]] = load <2 x float>, ptr [[NEXT_GEP8]], align 4, !alias.scope [[META3]]
; CHECK-NEXT:    [[TMP19:%.*]] = fadd <2 x float> [[WIDE_LOAD]], [[WIDE_LOAD12]]
; CHECK-NEXT:    store <2 x float> [[TMP19]], ptr [[NEXT_GEP]], align 4, !alias.scope [[META0]], !noalias [[META3]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2
; CHECK-NEXT:    [[TMP20:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[TMP20]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK:       scalar.ph:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi ptr [ [[TMP16]], [[MIDDLE_BLOCK]] ], [ [[P1_END]], [[ENTRY:%.*]] ], [ [[P1_END]], [[VECTOR_SCEVCHECK]] ], [ [[P1_END]], [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT:    [[BC_RESUME_VAL9:%.*]] = phi ptr [ [[IND_END8]], [[MIDDLE_BLOCK]] ], [ [[P2_START]], [[ENTRY]] ], [ [[P2_START]], [[VECTOR_SCEVCHECK]] ], [ [[P2_START]], [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[P1:%.*]] = phi ptr [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[P1_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT:    [[P2:%.*]] = phi ptr [ [[BC_RESUME_VAL9]], [[SCALAR_PH]] ], [ [[P2_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT:    [[P1_VAL:%.*]] = load float, ptr [[P1]], align 4
; CHECK-NEXT:    [[P2_VAL:%.*]] = load float, ptr [[P2]], align 4
; CHECK-NEXT:    [[SUM:%.*]] = fadd float [[P1_VAL]], [[P2_VAL]]
; CHECK-NEXT:    store float [[SUM]], ptr [[P1]], align 4
; CHECK-NEXT:    [[P1_NEXT]] = getelementptr float, ptr [[P1]], i64 1
; CHECK-NEXT:    [[P2_NEXT]] = getelementptr float, ptr [[P2]], i64 1
; CHECK-NEXT:    [[C:%.*]] = icmp ne ptr [[P1_NEXT]], [[P1_START]]
; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[EXIT]], !llvm.loop [[LOOP8:![0-9]+]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop

loop:
  %p1 = phi ptr [ %p1.start, %entry ], [ %p1.next, %loop ]
  %p2 = phi ptr [ %p2.start, %entry ], [ %p2.next, %loop ]
  %p1.val = load float, ptr %p1
  %p2.val = load float, ptr %p2
  %sum = fadd float %p1.val, %p2.val
  store float %sum, ptr %p1
  %p1.next = getelementptr float, ptr %p1, i64 1
  %p2.next = getelementptr float, ptr %p2, i64 1
  %c = icmp ne ptr %p1.next, %p1.end
  br i1 %c, label %loop, label %exit

exit:
  ret void
}

define void @test_ptr_iv_with_inbounds(ptr %p1.start, ptr %p2.start, ptr %p1.end) {
; CHECK-LABEL: @test_ptr_iv_with_inbounds(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[P1_END1:%.*]] = ptrtoaddr ptr [[P1_END:%.*]] to i64
; CHECK-NEXT:    [[TMP4:%.*]] = ptrtoaddr ptr [[P1_START:%.*]] to i64
; CHECK-NEXT:    [[TMP5:%.*]] = add i64 [[P1_END1]], -4
; CHECK-NEXT:    [[TMP1:%.*]] = sub i64 [[TMP5]], [[TMP4]]
; CHECK-NEXT:    [[TMP2:%.*]] = lshr i64 [[TMP1]], 2
; CHECK-NEXT:    [[TMP3:%.*]] = add nuw nsw i64 [[TMP2]], 1
; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP3]], 2
; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_MEMCHECK:%.*]]
; CHECK:       vector.memcheck:
; CHECK-NEXT:    [[TMP7:%.*]] = and i64 [[TMP1]], -4
; CHECK-NEXT:    [[TMP8:%.*]] = add i64 [[TMP7]], 4
; CHECK-NEXT:    [[SCEVGEP:%.*]] = getelementptr i8, ptr [[P1_START]], i64 [[TMP8]]
; CHECK-NEXT:    [[SCEVGEP3:%.*]] = getelementptr i8, ptr [[P2_START:%.*]], i64 [[TMP8]]
; CHECK-NEXT:    [[BOUND0:%.*]] = icmp ult ptr [[P1_START]], [[SCEVGEP3]]
; CHECK-NEXT:    [[BOUND1:%.*]] = icmp ult ptr [[P2_START]], [[SCEVGEP]]
; CHECK-NEXT:    [[FOUND_CONFLICT:%.*]] = and i1 [[BOUND0]], [[BOUND1]]
; CHECK-NEXT:    br i1 [[FOUND_CONFLICT]], label [[SCALAR_PH]], label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    [[N_MOD_VF:%.*]] = and i64 [[TMP3]], 1
; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP3]], [[N_MOD_VF]]
; CHECK-NEXT:    [[TMP9:%.*]] = shl i64 [[N_VEC]], 2
; CHECK-NEXT:    [[TMP14:%.*]] = getelementptr i8, ptr [[P1_START]], i64 [[TMP9]]
; CHECK-NEXT:    [[IND_END6:%.*]] = getelementptr i8, ptr [[P2_START]], i64 [[TMP9]]
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[OFFSET_IDX:%.*]] = shl i64 [[INDEX]], 2
; CHECK-NEXT:    [[NEXT_GEP:%.*]] = getelementptr i8, ptr [[P1_START]], i64 [[OFFSET_IDX]]
; CHECK-NEXT:    [[NEXT_GEP6:%.*]] = getelementptr i8, ptr [[P2_START]], i64 [[OFFSET_IDX]]
; CHECK-NEXT:    [[WIDE_LOAD:%.*]] = load <2 x float>, ptr [[NEXT_GEP]], align 4, !alias.scope [[META9:![0-9]+]], !noalias [[META12:![0-9]+]]
; CHECK-NEXT:    [[WIDE_LOAD10:%.*]] = load <2 x float>, ptr [[NEXT_GEP6]], align 4, !alias.scope [[META12]]
; CHECK-NEXT:    [[TMP15:%.*]] = fadd <2 x float> [[WIDE_LOAD]], [[WIDE_LOAD10]]
; CHECK-NEXT:    store <2 x float> [[TMP15]], ptr [[NEXT_GEP]], align 4, !alias.scope [[META9]], !noalias [[META12]]
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2
; CHECK-NEXT:    [[TMP16:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[TMP16]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP14:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK:       scalar.ph:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi ptr [ [[TMP14]], [[MIDDLE_BLOCK]] ], [ [[P1_START]], [[ENTRY:%.*]] ], [ [[P1_START]], [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT:    [[BC_RESUME_VAL7:%.*]] = phi ptr [ [[IND_END6]], [[MIDDLE_BLOCK]] ], [ [[P2_START]], [[ENTRY]] ], [ [[P2_START]], [[VECTOR_MEMCHECK]] ]
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[P1:%.*]] = phi ptr [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[P1_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT:    [[P2:%.*]] = phi ptr [ [[BC_RESUME_VAL7]], [[SCALAR_PH]] ], [ [[P2_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT:    [[P1_VAL:%.*]] = load float, ptr [[P1]], align 4
; CHECK-NEXT:    [[P2_VAL:%.*]] = load float, ptr [[P2]], align 4
; CHECK-NEXT:    [[SUM:%.*]] = fadd float [[P1_VAL]], [[P2_VAL]]
; CHECK-NEXT:    store float [[SUM]], ptr [[P1]], align 4
; CHECK-NEXT:    [[P1_NEXT]] = getelementptr inbounds float, ptr [[P1]], i64 1
; CHECK-NEXT:    [[P2_NEXT]] = getelementptr inbounds float, ptr [[P2]], i64 1
; CHECK-NEXT:    [[C:%.*]] = icmp ne ptr [[P1_NEXT]], [[P1_END]]
; CHECK-NEXT:    br i1 [[C]], label [[LOOP]], label [[EXIT]], !llvm.loop [[LOOP15:![0-9]+]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop

loop:
  %p1 = phi ptr [ %p1.start, %entry ], [ %p1.next, %loop ]
  %p2 = phi ptr [ %p2.start, %entry ], [ %p2.next, %loop ]
  %p1.val = load float, ptr %p1
  %p2.val = load float, ptr %p2
  %sum = fadd float %p1.val, %p2.val
  store float %sum, ptr %p1
  %p1.next = getelementptr inbounds float, ptr %p1, i64 1
  %p2.next = getelementptr inbounds float, ptr %p2, i64 1
  %c = icmp ne ptr %p1.next, %p1.end
  br i1 %c, label %loop, label %exit

exit:
  ret void
}

define void @store_pointer_induction(ptr %start, ptr %end) {
; CHECK-LABEL: @store_pointer_induction(
; CHECK-NEXT:  entry:
; CHECK-NEXT:    [[START2:%.*]] = ptrtoaddr ptr [[START:%.*]] to i64
; CHECK-NEXT:    [[END1:%.*]] = ptrtoaddr ptr [[END:%.*]] to i64
; CHECK-NEXT:    [[TMP9:%.*]] = add i64 [[START2]], -8
; CHECK-NEXT:    [[TMP1:%.*]] = sub i64 [[TMP9]], [[END1]]
; CHECK-NEXT:    [[TMP2:%.*]] = lshr i64 [[TMP1]], 3
; CHECK-NEXT:    [[TMP3:%.*]] = add nuw nsw i64 [[TMP2]], 1
; CHECK-NEXT:    [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[TMP3]], 2
; CHECK-NEXT:    br i1 [[MIN_ITERS_CHECK]], label [[SCALAR_PH:%.*]], label [[VECTOR_PH:%.*]]
; CHECK:       vector.ph:
; CHECK-NEXT:    [[N_MOD_VF:%.*]] = and i64 [[TMP3]], 1
; CHECK-NEXT:    [[N_VEC:%.*]] = sub i64 [[TMP3]], [[N_MOD_VF]]
; CHECK-NEXT:    [[TMP4:%.*]] = shl i64 [[N_VEC]], 3
; CHECK-NEXT:    [[TMP7:%.*]] = getelementptr i8, ptr [[END]], i64 [[TMP4]]
; CHECK-NEXT:    br label [[VECTOR_BODY:%.*]]
; CHECK:       vector.body:
; CHECK-NEXT:    [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[POINTER_PHI:%.*]] = phi ptr [ [[END]], [[VECTOR_PH]] ], [ [[PTR_IND:%.*]], [[VECTOR_BODY]] ]
; CHECK-NEXT:    [[TMP5:%.*]] = getelementptr i8, ptr [[POINTER_PHI]], <2 x i64> <i64 0, i64 8>
; CHECK-NEXT:    [[TMP6:%.*]] = extractelement <2 x ptr> [[TMP5]], i64 0
; CHECK-NEXT:    store <2 x ptr> [[TMP5]], ptr [[TMP6]], align 4
; CHECK-NEXT:    [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 2
; CHECK-NEXT:    [[PTR_IND]] = getelementptr i8, ptr [[POINTER_PHI]], i64 16
; CHECK-NEXT:    [[TMP8:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[TMP8]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP16:![0-9]+]]
; CHECK:       middle.block:
; CHECK-NEXT:    [[CMP_N:%.*]] = icmp eq i64 [[TMP3]], [[N_VEC]]
; CHECK-NEXT:    br i1 [[CMP_N]], label [[EXIT:%.*]], label [[SCALAR_PH]]
; CHECK:       scalar.ph:
; CHECK-NEXT:    [[BC_RESUME_VAL:%.*]] = phi ptr [ [[TMP7]], [[MIDDLE_BLOCK]] ], [ [[END]], [[ENTRY:%.*]] ]
; CHECK-NEXT:    br label [[LOOP:%.*]]
; CHECK:       loop:
; CHECK-NEXT:    [[IV:%.*]] = phi ptr [ [[BC_RESUME_VAL]], [[SCALAR_PH]] ], [ [[IV_NEXT:%.*]], [[LOOP]] ]
; CHECK-NEXT:    store ptr [[IV]], ptr [[IV]], align 4
; CHECK-NEXT:    [[IV_NEXT]] = getelementptr inbounds ptr, ptr [[IV]], i32 1
; CHECK-NEXT:    [[EXITCOND:%.*]] = icmp eq ptr [[IV_NEXT]], [[START]]
; CHECK-NEXT:    br i1 [[EXITCOND]], label [[EXIT]], label [[LOOP]], !llvm.loop [[LOOP17:![0-9]+]]
; CHECK:       exit:
; CHECK-NEXT:    ret void
;
entry:
  br label %loop

loop:
  %iv = phi ptr [ %start, %entry ], [ %iv.next, %loop ]
  store ptr %iv, ptr %iv, align 4
  %iv.next = getelementptr inbounds ptr, ptr %iv, i32 1
  %exitcond = icmp eq ptr %iv.next, %end
  br i1 %exitcond, label %exit, label %loop

exit:
  ret void
}