caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
ec056f54589d937ae3f74d795b94de886cbb7ec7
clang-p2996/llvm/test/Transforms/LoopVectorize/PowerPC/optimal-epilog-vectorization.ll
Florian Hahn 32c4493d5f [VPlan] Add incoming values for all predecessor to ResumePHI (NFCI). 
Follow-up as discussed when using VPInstruction::ResumePhi for all resume
values (#112147). This patch explicitly adds incoming values for each
predecessor in VPlan. This simplifies codegen and allows transformations
adjusting the predecessors of blocks with

NFC modulo incoming block order in phis.
2025-02-09 11:20:20 +00:00

658 lines
52 KiB
LLVM
Raw Blame History

; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 3
; RUN: opt < %s -passes='loop-vectorize' -enable-epilogue-vectorization -epilogue-vectorization-force-VF=2 -S | FileCheck %s --check-prefix VF-TWO-CHECK
; RUN: opt < %s -passes='loop-vectorize' -enable-epilogue-vectorization -epilogue-vectorization-force-VF=4 -S | FileCheck %s --check-prefix VF-FOUR-CHECK
target datalayout = "e-m:e-i64:64-n32:64"
target triple = "powerpc64le-unknown-linux-gnu"
; Function Attrs: nounwind
define dso_local void @f1(ptr noalias %aa, ptr noalias %bb, ptr noalias %cc, i32 signext %N) #0 {
; VF-TWO-CHECK-LABEL: define dso_local void @f1(
; VF-TWO-CHECK-SAME: ptr noalias [[AA:%.*]], ptr noalias [[BB:%.*]], ptr noalias [[CC:%.*]], i32 signext [[N:%.*]]) #[[ATTR0:[0-9]+]] {
; VF-TWO-CHECK-NEXT: entry:
; VF-TWO-CHECK-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[N]], 0
; VF-TWO-CHECK-NEXT: br i1 [[CMP1]], label [[ITER_CHECK:%.*]], label [[FOR_END:%.*]]
; VF-TWO-CHECK: iter.check:
; VF-TWO-CHECK-NEXT: [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[N]] to i64
; VF-TWO-CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 2
; VF-TWO-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH:%.*]], label [[VECTOR_MAIN_LOOP_ITER_CHECK:%.*]]
; VF-TWO-CHECK: vector.main.loop.iter.check:
; VF-TWO-CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 32
; VF-TWO-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; VF-TWO-CHECK: vector.ph:
; VF-TWO-CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 32
; VF-TWO-CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF]]
; VF-TWO-CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; VF-TWO-CHECK: vector.body:
; VF-TWO-CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; VF-TWO-CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[TMP0]]
; VF-TWO-CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP17:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 4
; VF-TWO-CHECK-NEXT: [[TMP18:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 8
; VF-TWO-CHECK-NEXT: [[TMP19:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 12
; VF-TWO-CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 16
; VF-TWO-CHECK-NEXT: [[TMP21:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 20
; VF-TWO-CHECK-NEXT: [[TMP22:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 24
; VF-TWO-CHECK-NEXT: [[TMP23:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 28
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP16]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <4 x float>, ptr [[TMP17]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD3:%.*]] = load <4 x float>, ptr [[TMP18]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD4:%.*]] = load <4 x float>, ptr [[TMP19]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD5:%.*]] = load <4 x float>, ptr [[TMP20]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD6:%.*]] = load <4 x float>, ptr [[TMP21]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD7:%.*]] = load <4 x float>, ptr [[TMP22]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD8:%.*]] = load <4 x float>, ptr [[TMP23]], align 4
; VF-TWO-CHECK-NEXT: [[TMP24:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[TMP0]]
; VF-TWO-CHECK-NEXT: [[TMP32:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP33:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 4
; VF-TWO-CHECK-NEXT: [[TMP34:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 8
; VF-TWO-CHECK-NEXT: [[TMP35:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 12
; VF-TWO-CHECK-NEXT: [[TMP36:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 16
; VF-TWO-CHECK-NEXT: [[TMP37:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 20
; VF-TWO-CHECK-NEXT: [[TMP38:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 24
; VF-TWO-CHECK-NEXT: [[TMP39:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 28
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD9:%.*]] = load <4 x float>, ptr [[TMP32]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD10:%.*]] = load <4 x float>, ptr [[TMP33]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD11:%.*]] = load <4 x float>, ptr [[TMP34]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD12:%.*]] = load <4 x float>, ptr [[TMP35]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD13:%.*]] = load <4 x float>, ptr [[TMP36]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD14:%.*]] = load <4 x float>, ptr [[TMP37]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD15:%.*]] = load <4 x float>, ptr [[TMP38]], align 4
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD16:%.*]] = load <4 x float>, ptr [[TMP39]], align 4
; VF-TWO-CHECK-NEXT: [[TMP40:%.*]] = fadd fast <4 x float> [[WIDE_LOAD]], [[WIDE_LOAD9]]
; VF-TWO-CHECK-NEXT: [[TMP41:%.*]] = fadd fast <4 x float> [[WIDE_LOAD2]], [[WIDE_LOAD10]]
; VF-TWO-CHECK-NEXT: [[TMP42:%.*]] = fadd fast <4 x float> [[WIDE_LOAD3]], [[WIDE_LOAD11]]
; VF-TWO-CHECK-NEXT: [[TMP43:%.*]] = fadd fast <4 x float> [[WIDE_LOAD4]], [[WIDE_LOAD12]]
; VF-TWO-CHECK-NEXT: [[TMP44:%.*]] = fadd fast <4 x float> [[WIDE_LOAD5]], [[WIDE_LOAD13]]
; VF-TWO-CHECK-NEXT: [[TMP45:%.*]] = fadd fast <4 x float> [[WIDE_LOAD6]], [[WIDE_LOAD14]]
; VF-TWO-CHECK-NEXT: [[TMP46:%.*]] = fadd fast <4 x float> [[WIDE_LOAD7]], [[WIDE_LOAD15]]
; VF-TWO-CHECK-NEXT: [[TMP47:%.*]] = fadd fast <4 x float> [[WIDE_LOAD8]], [[WIDE_LOAD16]]
; VF-TWO-CHECK-NEXT: [[TMP48:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[TMP0]]
; VF-TWO-CHECK-NEXT: [[TMP56:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 4
; VF-TWO-CHECK-NEXT: [[TMP58:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 8
; VF-TWO-CHECK-NEXT: [[TMP59:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 12
; VF-TWO-CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 16
; VF-TWO-CHECK-NEXT: [[TMP61:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 20
; VF-TWO-CHECK-NEXT: [[TMP62:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 24
; VF-TWO-CHECK-NEXT: [[TMP63:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 28
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP40]], ptr [[TMP56]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP41]], ptr [[TMP57]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP42]], ptr [[TMP58]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP43]], ptr [[TMP59]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP44]], ptr [[TMP60]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP45]], ptr [[TMP61]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP46]], ptr [[TMP62]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP47]], ptr [[TMP63]], align 4
; VF-TWO-CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32
; VF-TWO-CHECK-NEXT: [[TMP64:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: br i1 [[TMP64]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; VF-TWO-CHECK: middle.block:
; VF-TWO-CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
; VF-TWO-CHECK: vec.epilog.iter.check:
; VF-TWO-CHECK-NEXT: [[N_VEC_REMAINING:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: [[MIN_EPILOG_ITERS_CHECK:%.*]] = icmp ult i64 [[N_VEC_REMAINING]], 2
; VF-TWO-CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; VF-TWO-CHECK: vec.epilog.ph:
; VF-TWO-CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; VF-TWO-CHECK-NEXT: [[N_MOD_VF17:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 2
; VF-TWO-CHECK-NEXT: [[N_VEC18:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF17]]
; VF-TWO-CHECK-NEXT: br label [[VEC_EPILOG_VECTOR_BODY:%.*]]
; VF-TWO-CHECK: vec.epilog.vector.body:
; VF-TWO-CHECK-NEXT: [[INDEX20:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT23:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[TMP65:%.*]] = add i64 [[INDEX20]], 0
; VF-TWO-CHECK-NEXT: [[TMP66:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[TMP65]]
; VF-TWO-CHECK-NEXT: [[TMP67:%.*]] = getelementptr inbounds float, ptr [[TMP66]], i32 0
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD21:%.*]] = load <2 x float>, ptr [[TMP67]], align 4
; VF-TWO-CHECK-NEXT: [[TMP68:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[TMP65]]
; VF-TWO-CHECK-NEXT: [[TMP69:%.*]] = getelementptr inbounds float, ptr [[TMP68]], i32 0
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD22:%.*]] = load <2 x float>, ptr [[TMP69]], align 4
; VF-TWO-CHECK-NEXT: [[TMP70:%.*]] = fadd fast <2 x float> [[WIDE_LOAD21]], [[WIDE_LOAD22]]
; VF-TWO-CHECK-NEXT: [[TMP71:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[TMP65]]
; VF-TWO-CHECK-NEXT: [[TMP72:%.*]] = getelementptr inbounds float, ptr [[TMP71]], i32 0
; VF-TWO-CHECK-NEXT: store <2 x float> [[TMP70]], ptr [[TMP72]], align 4
; VF-TWO-CHECK-NEXT: [[INDEX_NEXT23]] = add nuw i64 [[INDEX20]], 2
; VF-TWO-CHECK-NEXT: [[TMP73:%.*]] = icmp eq i64 [[INDEX_NEXT23]], [[N_VEC18]]
; VF-TWO-CHECK-NEXT: br i1 [[TMP73]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; VF-TWO-CHECK: vec.epilog.middle.block:
; VF-TWO-CHECK-NEXT: [[CMP_N19:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC18]]
; VF-TWO-CHECK-NEXT: br i1 [[CMP_N19]], label [[FOR_END_LOOPEXIT]], label [[VEC_EPILOG_SCALAR_PH]]
; VF-TWO-CHECK: vec.epilog.scalar.ph:
; VF-TWO-CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC18]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-TWO-CHECK-NEXT: br label [[FOR_BODY:%.*]]
; VF-TWO-CHECK: for.body:
; VF-TWO-CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[INDVARS_IV]]
; VF-TWO-CHECK-NEXT: [[TMP74:%.*]] = load float, ptr [[ARRAYIDX]], align 4
; VF-TWO-CHECK-NEXT: [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[INDVARS_IV]]
; VF-TWO-CHECK-NEXT: [[TMP75:%.*]] = load float, ptr [[ARRAYIDX2]], align 4
; VF-TWO-CHECK-NEXT: [[ADD:%.*]] = fadd fast float [[TMP74]], [[TMP75]]
; VF-TWO-CHECK-NEXT: [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[INDVARS_IV]]
; VF-TWO-CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX4]], align 4
; VF-TWO-CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
; VF-TWO-CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
; VF-TWO-CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT]], !llvm.loop [[LOOP4:![0-9]+]]
; VF-TWO-CHECK: for.end.loopexit:
; VF-TWO-CHECK-NEXT: br label [[FOR_END]]
; VF-TWO-CHECK: for.end:
; VF-TWO-CHECK-NEXT: ret void
;
; VF-FOUR-CHECK-LABEL: define dso_local void @f1(
; VF-FOUR-CHECK-SAME: ptr noalias [[AA:%.*]], ptr noalias [[BB:%.*]], ptr noalias [[CC:%.*]], i32 signext [[N:%.*]]) #[[ATTR0:[0-9]+]] {
; VF-FOUR-CHECK-NEXT: entry:
; VF-FOUR-CHECK-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[N]], 0
; VF-FOUR-CHECK-NEXT: br i1 [[CMP1]], label [[ITER_CHECK:%.*]], label [[FOR_END:%.*]]
; VF-FOUR-CHECK: iter.check:
; VF-FOUR-CHECK-NEXT: [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[N]] to i64
; VF-FOUR-CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 4
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH:%.*]], label [[VECTOR_MAIN_LOOP_ITER_CHECK:%.*]]
; VF-FOUR-CHECK: vector.main.loop.iter.check:
; VF-FOUR-CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 32
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; VF-FOUR-CHECK: vector.ph:
; VF-FOUR-CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 32
; VF-FOUR-CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF]]
; VF-FOUR-CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; VF-FOUR-CHECK: vector.body:
; VF-FOUR-CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[TMP0:%.*]] = add i64 [[INDEX]], 0
; VF-FOUR-CHECK-NEXT: [[TMP8:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[TMP0]]
; VF-FOUR-CHECK-NEXT: [[TMP16:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP17:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 4
; VF-FOUR-CHECK-NEXT: [[TMP18:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 8
; VF-FOUR-CHECK-NEXT: [[TMP19:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 12
; VF-FOUR-CHECK-NEXT: [[TMP20:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 16
; VF-FOUR-CHECK-NEXT: [[TMP21:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 20
; VF-FOUR-CHECK-NEXT: [[TMP22:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 24
; VF-FOUR-CHECK-NEXT: [[TMP23:%.*]] = getelementptr inbounds float, ptr [[TMP8]], i32 28
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP16]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <4 x float>, ptr [[TMP17]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD3:%.*]] = load <4 x float>, ptr [[TMP18]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD4:%.*]] = load <4 x float>, ptr [[TMP19]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD5:%.*]] = load <4 x float>, ptr [[TMP20]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD6:%.*]] = load <4 x float>, ptr [[TMP21]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD7:%.*]] = load <4 x float>, ptr [[TMP22]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD8:%.*]] = load <4 x float>, ptr [[TMP23]], align 4
; VF-FOUR-CHECK-NEXT: [[TMP24:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[TMP0]]
; VF-FOUR-CHECK-NEXT: [[TMP32:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP33:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 4
; VF-FOUR-CHECK-NEXT: [[TMP34:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 8
; VF-FOUR-CHECK-NEXT: [[TMP35:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 12
; VF-FOUR-CHECK-NEXT: [[TMP36:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 16
; VF-FOUR-CHECK-NEXT: [[TMP37:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 20
; VF-FOUR-CHECK-NEXT: [[TMP38:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 24
; VF-FOUR-CHECK-NEXT: [[TMP39:%.*]] = getelementptr inbounds float, ptr [[TMP24]], i32 28
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD9:%.*]] = load <4 x float>, ptr [[TMP32]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD10:%.*]] = load <4 x float>, ptr [[TMP33]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD11:%.*]] = load <4 x float>, ptr [[TMP34]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD12:%.*]] = load <4 x float>, ptr [[TMP35]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD13:%.*]] = load <4 x float>, ptr [[TMP36]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD14:%.*]] = load <4 x float>, ptr [[TMP37]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD15:%.*]] = load <4 x float>, ptr [[TMP38]], align 4
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD16:%.*]] = load <4 x float>, ptr [[TMP39]], align 4
; VF-FOUR-CHECK-NEXT: [[TMP40:%.*]] = fadd fast <4 x float> [[WIDE_LOAD]], [[WIDE_LOAD9]]
; VF-FOUR-CHECK-NEXT: [[TMP41:%.*]] = fadd fast <4 x float> [[WIDE_LOAD2]], [[WIDE_LOAD10]]
; VF-FOUR-CHECK-NEXT: [[TMP42:%.*]] = fadd fast <4 x float> [[WIDE_LOAD3]], [[WIDE_LOAD11]]
; VF-FOUR-CHECK-NEXT: [[TMP43:%.*]] = fadd fast <4 x float> [[WIDE_LOAD4]], [[WIDE_LOAD12]]
; VF-FOUR-CHECK-NEXT: [[TMP44:%.*]] = fadd fast <4 x float> [[WIDE_LOAD5]], [[WIDE_LOAD13]]
; VF-FOUR-CHECK-NEXT: [[TMP45:%.*]] = fadd fast <4 x float> [[WIDE_LOAD6]], [[WIDE_LOAD14]]
; VF-FOUR-CHECK-NEXT: [[TMP46:%.*]] = fadd fast <4 x float> [[WIDE_LOAD7]], [[WIDE_LOAD15]]
; VF-FOUR-CHECK-NEXT: [[TMP47:%.*]] = fadd fast <4 x float> [[WIDE_LOAD8]], [[WIDE_LOAD16]]
; VF-FOUR-CHECK-NEXT: [[TMP48:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[TMP0]]
; VF-FOUR-CHECK-NEXT: [[TMP56:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 4
; VF-FOUR-CHECK-NEXT: [[TMP58:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 8
; VF-FOUR-CHECK-NEXT: [[TMP59:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 12
; VF-FOUR-CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 16
; VF-FOUR-CHECK-NEXT: [[TMP61:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 20
; VF-FOUR-CHECK-NEXT: [[TMP62:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 24
; VF-FOUR-CHECK-NEXT: [[TMP63:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 28
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP40]], ptr [[TMP56]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP41]], ptr [[TMP57]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP42]], ptr [[TMP58]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP43]], ptr [[TMP59]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP44]], ptr [[TMP60]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP45]], ptr [[TMP61]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP46]], ptr [[TMP62]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP47]], ptr [[TMP63]], align 4
; VF-FOUR-CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32
; VF-FOUR-CHECK-NEXT: [[TMP64:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: br i1 [[TMP64]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP0:![0-9]+]]
; VF-FOUR-CHECK: middle.block:
; VF-FOUR-CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
; VF-FOUR-CHECK: vec.epilog.iter.check:
; VF-FOUR-CHECK-NEXT: [[N_VEC_REMAINING:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: [[MIN_EPILOG_ITERS_CHECK:%.*]] = icmp ult i64 [[N_VEC_REMAINING]], 4
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; VF-FOUR-CHECK: vec.epilog.ph:
; VF-FOUR-CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; VF-FOUR-CHECK-NEXT: [[N_MOD_VF17:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 4
; VF-FOUR-CHECK-NEXT: [[N_VEC18:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF17]]
; VF-FOUR-CHECK-NEXT: br label [[VEC_EPILOG_VECTOR_BODY:%.*]]
; VF-FOUR-CHECK: vec.epilog.vector.body:
; VF-FOUR-CHECK-NEXT: [[INDEX20:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT23:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[TMP65:%.*]] = add i64 [[INDEX20]], 0
; VF-FOUR-CHECK-NEXT: [[TMP66:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[TMP65]]
; VF-FOUR-CHECK-NEXT: [[TMP67:%.*]] = getelementptr inbounds float, ptr [[TMP66]], i32 0
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD21:%.*]] = load <4 x float>, ptr [[TMP67]], align 4
; VF-FOUR-CHECK-NEXT: [[TMP68:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[TMP65]]
; VF-FOUR-CHECK-NEXT: [[TMP69:%.*]] = getelementptr inbounds float, ptr [[TMP68]], i32 0
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD22:%.*]] = load <4 x float>, ptr [[TMP69]], align 4
; VF-FOUR-CHECK-NEXT: [[TMP70:%.*]] = fadd fast <4 x float> [[WIDE_LOAD21]], [[WIDE_LOAD22]]
; VF-FOUR-CHECK-NEXT: [[TMP71:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[TMP65]]
; VF-FOUR-CHECK-NEXT: [[TMP72:%.*]] = getelementptr inbounds float, ptr [[TMP71]], i32 0
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP70]], ptr [[TMP72]], align 4
; VF-FOUR-CHECK-NEXT: [[INDEX_NEXT23]] = add nuw i64 [[INDEX20]], 4
; VF-FOUR-CHECK-NEXT: [[TMP73:%.*]] = icmp eq i64 [[INDEX_NEXT23]], [[N_VEC18]]
; VF-FOUR-CHECK-NEXT: br i1 [[TMP73]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP3:![0-9]+]]
; VF-FOUR-CHECK: vec.epilog.middle.block:
; VF-FOUR-CHECK-NEXT: [[CMP_N19:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC18]]
; VF-FOUR-CHECK-NEXT: br i1 [[CMP_N19]], label [[FOR_END_LOOPEXIT]], label [[VEC_EPILOG_SCALAR_PH]]
; VF-FOUR-CHECK: vec.epilog.scalar.ph:
; VF-FOUR-CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC18]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-FOUR-CHECK-NEXT: br label [[FOR_BODY:%.*]]
; VF-FOUR-CHECK: for.body:
; VF-FOUR-CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[BB]], i64 [[INDVARS_IV]]
; VF-FOUR-CHECK-NEXT: [[TMP74:%.*]] = load float, ptr [[ARRAYIDX]], align 4
; VF-FOUR-CHECK-NEXT: [[ARRAYIDX2:%.*]] = getelementptr inbounds float, ptr [[CC]], i64 [[INDVARS_IV]]
; VF-FOUR-CHECK-NEXT: [[TMP75:%.*]] = load float, ptr [[ARRAYIDX2]], align 4
; VF-FOUR-CHECK-NEXT: [[ADD:%.*]] = fadd fast float [[TMP74]], [[TMP75]]
; VF-FOUR-CHECK-NEXT: [[ARRAYIDX4:%.*]] = getelementptr inbounds float, ptr [[AA]], i64 [[INDVARS_IV]]
; VF-FOUR-CHECK-NEXT: store float [[ADD]], ptr [[ARRAYIDX4]], align 4
; VF-FOUR-CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
; VF-FOUR-CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
; VF-FOUR-CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT]], !llvm.loop [[LOOP4:![0-9]+]]
; VF-FOUR-CHECK: for.end.loopexit:
; VF-FOUR-CHECK-NEXT: br label [[FOR_END]]
; VF-FOUR-CHECK: for.end:
; VF-FOUR-CHECK-NEXT: ret void
;
entry:
%cmp1 = icmp sgt i32 %N, 0
br i1 %cmp1, label %for.body.preheader, label %for.end
for.body.preheader: ; preds = %entry
%wide.trip.count = zext i32 %N to i64
br label %for.body
for.body: ; preds = %for.body.preheader, %for.body
%indvars.iv = phi i64 [ 0, %for.body.preheader ], [ %indvars.iv.next, %for.body ]
%arrayidx = getelementptr inbounds float, ptr %bb, i64 %indvars.iv
%0 = load float, ptr %arrayidx, align 4
%arrayidx2 = getelementptr inbounds float, ptr %cc, i64 %indvars.iv
%1 = load float, ptr %arrayidx2, align 4
%add = fadd fast float %0, %1
%arrayidx4 = getelementptr inbounds float, ptr %aa, i64 %indvars.iv
store float %add, ptr %arrayidx4, align 4
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
%exitcond = icmp ne i64 %indvars.iv.next, %wide.trip.count
br i1 %exitcond, label %for.body, label %for.end.loopexit
for.end.loopexit: ; preds = %for.body
br label %for.end
for.end: ; preds = %for.end.loopexit, %entry
ret void
}
define dso_local signext i32 @f2(ptr noalias %A, ptr noalias %B, i32 signext %n) #0 {
; VF-TWO-CHECK-LABEL: define dso_local signext i32 @f2(
; VF-TWO-CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i32 signext [[N:%.*]]) #[[ATTR0]] {
; VF-TWO-CHECK-NEXT: entry:
; VF-TWO-CHECK-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[N]], 1
; VF-TWO-CHECK-NEXT: br i1 [[CMP1]], label [[ITER_CHECK:%.*]], label [[FOR_END:%.*]]
; VF-TWO-CHECK: iter.check:
; VF-TWO-CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1
; VF-TWO-CHECK-NEXT: [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[TMP0]] to i64
; VF-TWO-CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 2
; VF-TWO-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; VF-TWO-CHECK: vector.scevcheck:
; VF-TWO-CHECK-NEXT: [[TMP1:%.*]] = add nsw i64 [[WIDE_TRIP_COUNT]], -1
; VF-TWO-CHECK-NEXT: [[TMP2:%.*]] = trunc i64 [[TMP1]] to i32
; VF-TWO-CHECK-NEXT: [[MUL:%.*]] = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 1, i32 [[TMP2]])
; VF-TWO-CHECK-NEXT: [[MUL_RESULT:%.*]] = extractvalue { i32, i1 } [[MUL]], 0
; VF-TWO-CHECK-NEXT: [[MUL_OVERFLOW:%.*]] = extractvalue { i32, i1 } [[MUL]], 1
; VF-TWO-CHECK-NEXT: [[TMP3:%.*]] = sub i32 [[TMP0]], [[MUL_RESULT]]
; VF-TWO-CHECK-NEXT: [[TMP4:%.*]] = icmp sgt i32 [[TMP3]], [[TMP0]]
; VF-TWO-CHECK-NEXT: [[TMP5:%.*]] = or i1 [[TMP4]], [[MUL_OVERFLOW]]
; VF-TWO-CHECK-NEXT: [[TMP6:%.*]] = icmp ugt i64 [[TMP1]], 4294967295
; VF-TWO-CHECK-NEXT: [[TMP7:%.*]] = or i1 [[TMP5]], [[TMP6]]
; VF-TWO-CHECK-NEXT: br i1 [[TMP7]], label [[VEC_EPILOG_SCALAR_PH]], label [[VECTOR_MAIN_LOOP_ITER_CHECK:%.*]]
; VF-TWO-CHECK: vector.main.loop.iter.check:
; VF-TWO-CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 32
; VF-TWO-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; VF-TWO-CHECK: vector.ph:
; VF-TWO-CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 32
; VF-TWO-CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF]]
; VF-TWO-CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; VF-TWO-CHECK: vector.body:
; VF-TWO-CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 0
; VF-TWO-CHECK-NEXT: [[OFFSET_IDX:%.*]] = trunc i64 [[INDEX]] to i32
; VF-TWO-CHECK-NEXT: [[TMP16:%.*]] = add i32 [[OFFSET_IDX]], 0
; VF-TWO-CHECK-NEXT: [[TMP24:%.*]] = xor i32 [[TMP16]], -1
; VF-TWO-CHECK-NEXT: [[TMP32:%.*]] = add i32 [[TMP24]], [[N]]
; VF-TWO-CHECK-NEXT: [[TMP40:%.*]] = sext i32 [[TMP32]] to i64
; VF-TWO-CHECK-NEXT: [[TMP48:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[TMP40]]
; VF-TWO-CHECK-NEXT: [[TMP56:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds float, ptr [[TMP56]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP58:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -4
; VF-TWO-CHECK-NEXT: [[TMP59:%.*]] = getelementptr inbounds float, ptr [[TMP58]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -8
; VF-TWO-CHECK-NEXT: [[TMP61:%.*]] = getelementptr inbounds float, ptr [[TMP60]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP62:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -12
; VF-TWO-CHECK-NEXT: [[TMP63:%.*]] = getelementptr inbounds float, ptr [[TMP62]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP64:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -16
; VF-TWO-CHECK-NEXT: [[TMP65:%.*]] = getelementptr inbounds float, ptr [[TMP64]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP66:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -20
; VF-TWO-CHECK-NEXT: [[TMP67:%.*]] = getelementptr inbounds float, ptr [[TMP66]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP68:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -24
; VF-TWO-CHECK-NEXT: [[TMP69:%.*]] = getelementptr inbounds float, ptr [[TMP68]], i32 -3
; VF-TWO-CHECK-NEXT: [[TMP70:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -28
; VF-TWO-CHECK-NEXT: [[TMP71:%.*]] = getelementptr inbounds float, ptr [[TMP70]], i32 -3
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP57]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE:%.*]] = shufflevector <4 x float> [[WIDE_LOAD]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <4 x float>, ptr [[TMP59]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE3:%.*]] = shufflevector <4 x float> [[WIDE_LOAD2]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD4:%.*]] = load <4 x float>, ptr [[TMP61]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE5:%.*]] = shufflevector <4 x float> [[WIDE_LOAD4]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD6:%.*]] = load <4 x float>, ptr [[TMP63]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE7:%.*]] = shufflevector <4 x float> [[WIDE_LOAD6]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD8:%.*]] = load <4 x float>, ptr [[TMP65]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE9:%.*]] = shufflevector <4 x float> [[WIDE_LOAD8]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD10:%.*]] = load <4 x float>, ptr [[TMP67]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE11:%.*]] = shufflevector <4 x float> [[WIDE_LOAD10]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD12:%.*]] = load <4 x float>, ptr [[TMP69]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE13:%.*]] = shufflevector <4 x float> [[WIDE_LOAD12]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD14:%.*]] = load <4 x float>, ptr [[TMP71]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE15:%.*]] = shufflevector <4 x float> [[WIDE_LOAD14]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[TMP72:%.*]] = fadd fast <4 x float> [[REVERSE]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP73:%.*]] = fadd fast <4 x float> [[REVERSE3]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP74:%.*]] = fadd fast <4 x float> [[REVERSE5]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP75:%.*]] = fadd fast <4 x float> [[REVERSE7]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP76:%.*]] = fadd fast <4 x float> [[REVERSE9]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP77:%.*]] = fadd fast <4 x float> [[REVERSE11]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP78:%.*]] = fadd fast <4 x float> [[REVERSE13]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP79:%.*]] = fadd fast <4 x float> [[REVERSE15]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP80:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[TMP8]]
; VF-TWO-CHECK-NEXT: [[TMP88:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP89:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 4
; VF-TWO-CHECK-NEXT: [[TMP90:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 8
; VF-TWO-CHECK-NEXT: [[TMP91:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 12
; VF-TWO-CHECK-NEXT: [[TMP92:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 16
; VF-TWO-CHECK-NEXT: [[TMP93:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 20
; VF-TWO-CHECK-NEXT: [[TMP94:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 24
; VF-TWO-CHECK-NEXT: [[TMP95:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 28
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP72]], ptr [[TMP88]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP73]], ptr [[TMP89]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP74]], ptr [[TMP90]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP75]], ptr [[TMP91]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP76]], ptr [[TMP92]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP77]], ptr [[TMP93]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP78]], ptr [[TMP94]], align 4
; VF-TWO-CHECK-NEXT: store <4 x float> [[TMP79]], ptr [[TMP95]], align 4
; VF-TWO-CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32
; VF-TWO-CHECK-NEXT: [[TMP96:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: br i1 [[TMP96]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; VF-TWO-CHECK: middle.block:
; VF-TWO-CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
; VF-TWO-CHECK: vec.epilog.iter.check:
; VF-TWO-CHECK-NEXT: [[IND_END18:%.*]] = trunc i64 [[N_VEC]] to i32
; VF-TWO-CHECK-NEXT: [[N_VEC_REMAINING:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-TWO-CHECK-NEXT: [[MIN_EPILOG_ITERS_CHECK:%.*]] = icmp ult i64 [[N_VEC_REMAINING]], 2
; VF-TWO-CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; VF-TWO-CHECK: vec.epilog.ph:
; VF-TWO-CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; VF-TWO-CHECK-NEXT: [[N_MOD_VF16:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 2
; VF-TWO-CHECK-NEXT: [[N_VEC17:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF16]]
; VF-TWO-CHECK-NEXT: [[IND_END:%.*]] = trunc i64 [[N_VEC17]] to i32
; VF-TWO-CHECK-NEXT: br label [[VEC_EPILOG_VECTOR_BODY:%.*]]
; VF-TWO-CHECK: vec.epilog.vector.body:
; VF-TWO-CHECK-NEXT: [[INDEX21:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT25:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[TMP97:%.*]] = add i64 [[INDEX21]], 0
; VF-TWO-CHECK-NEXT: [[OFFSET_IDX22:%.*]] = trunc i64 [[INDEX21]] to i32
; VF-TWO-CHECK-NEXT: [[TMP98:%.*]] = add i32 [[OFFSET_IDX22]], 0
; VF-TWO-CHECK-NEXT: [[TMP99:%.*]] = xor i32 [[TMP98]], -1
; VF-TWO-CHECK-NEXT: [[TMP100:%.*]] = add i32 [[TMP99]], [[N]]
; VF-TWO-CHECK-NEXT: [[TMP101:%.*]] = sext i32 [[TMP100]] to i64
; VF-TWO-CHECK-NEXT: [[TMP102:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[TMP101]]
; VF-TWO-CHECK-NEXT: [[TMP103:%.*]] = getelementptr inbounds float, ptr [[TMP102]], i32 0
; VF-TWO-CHECK-NEXT: [[TMP104:%.*]] = getelementptr inbounds float, ptr [[TMP103]], i32 -1
; VF-TWO-CHECK-NEXT: [[WIDE_LOAD23:%.*]] = load <2 x float>, ptr [[TMP104]], align 4
; VF-TWO-CHECK-NEXT: [[REVERSE24:%.*]] = shufflevector <2 x float> [[WIDE_LOAD23]], <2 x float> poison, <2 x i32> <i32 1, i32 0>
; VF-TWO-CHECK-NEXT: [[TMP105:%.*]] = fadd fast <2 x float> [[REVERSE24]], splat (float 1.000000e+00)
; VF-TWO-CHECK-NEXT: [[TMP106:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[TMP97]]
; VF-TWO-CHECK-NEXT: [[TMP107:%.*]] = getelementptr inbounds float, ptr [[TMP106]], i32 0
; VF-TWO-CHECK-NEXT: store <2 x float> [[TMP105]], ptr [[TMP107]], align 4
; VF-TWO-CHECK-NEXT: [[INDEX_NEXT25]] = add nuw i64 [[INDEX21]], 2
; VF-TWO-CHECK-NEXT: [[TMP108:%.*]] = icmp eq i64 [[INDEX_NEXT25]], [[N_VEC17]]
; VF-TWO-CHECK-NEXT: br i1 [[TMP108]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; VF-TWO-CHECK: vec.epilog.middle.block:
; VF-TWO-CHECK-NEXT: [[CMP_N20:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC17]]
; VF-TWO-CHECK-NEXT: br i1 [[CMP_N20]], label [[FOR_END_LOOPEXIT]], label [[VEC_EPILOG_SCALAR_PH]]
; VF-TWO-CHECK: vec.epilog.scalar.ph:
; VF-TWO-CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC17]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-TWO-CHECK-NEXT: [[BC_RESUME_VAL19:%.*]] = phi i32 [ [[IND_END]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END18]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-TWO-CHECK-NEXT: br label [[FOR_BODY:%.*]]
; VF-TWO-CHECK: for.body:
; VF-TWO-CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[I_014:%.*]] = phi i32 [ [[BC_RESUME_VAL19]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY]] ]
; VF-TWO-CHECK-NEXT: [[TMP109:%.*]] = xor i32 [[I_014]], -1
; VF-TWO-CHECK-NEXT: [[SUB2:%.*]] = add i32 [[TMP109]], [[N]]
; VF-TWO-CHECK-NEXT: [[IDXPROM:%.*]] = sext i32 [[SUB2]] to i64
; VF-TWO-CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[IDXPROM]]
; VF-TWO-CHECK-NEXT: [[TMP110:%.*]] = load float, ptr [[ARRAYIDX]], align 4
; VF-TWO-CHECK-NEXT: [[CONV3:%.*]] = fadd fast float [[TMP110]], 1.000000e+00
; VF-TWO-CHECK-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[INDVARS_IV]]
; VF-TWO-CHECK-NEXT: store float [[CONV3]], ptr [[ARRAYIDX5]], align 4
; VF-TWO-CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
; VF-TWO-CHECK-NEXT: [[INC]] = add nuw nsw i32 [[I_014]], 1
; VF-TWO-CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
; VF-TWO-CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT]], !llvm.loop [[LOOP7:![0-9]+]]
; VF-TWO-CHECK: for.end.loopexit:
; VF-TWO-CHECK-NEXT: br label [[FOR_END]]
; VF-TWO-CHECK: for.end:
; VF-TWO-CHECK-NEXT: ret i32 0
;
; VF-FOUR-CHECK-LABEL: define dso_local signext i32 @f2(
; VF-FOUR-CHECK-SAME: ptr noalias [[A:%.*]], ptr noalias [[B:%.*]], i32 signext [[N:%.*]]) #[[ATTR0]] {
; VF-FOUR-CHECK-NEXT: entry:
; VF-FOUR-CHECK-NEXT: [[CMP1:%.*]] = icmp sgt i32 [[N]], 1
; VF-FOUR-CHECK-NEXT: br i1 [[CMP1]], label [[ITER_CHECK:%.*]], label [[FOR_END:%.*]]
; VF-FOUR-CHECK: iter.check:
; VF-FOUR-CHECK-NEXT: [[TMP0:%.*]] = add i32 [[N]], -1
; VF-FOUR-CHECK-NEXT: [[WIDE_TRIP_COUNT:%.*]] = zext i32 [[TMP0]] to i64
; VF-FOUR-CHECK-NEXT: [[MIN_ITERS_CHECK:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 4
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH:%.*]], label [[VECTOR_SCEVCHECK:%.*]]
; VF-FOUR-CHECK: vector.scevcheck:
; VF-FOUR-CHECK-NEXT: [[TMP1:%.*]] = add nsw i64 [[WIDE_TRIP_COUNT]], -1
; VF-FOUR-CHECK-NEXT: [[TMP2:%.*]] = trunc i64 [[TMP1]] to i32
; VF-FOUR-CHECK-NEXT: [[MUL:%.*]] = call { i32, i1 } @llvm.umul.with.overflow.i32(i32 1, i32 [[TMP2]])
; VF-FOUR-CHECK-NEXT: [[MUL_RESULT:%.*]] = extractvalue { i32, i1 } [[MUL]], 0
; VF-FOUR-CHECK-NEXT: [[MUL_OVERFLOW:%.*]] = extractvalue { i32, i1 } [[MUL]], 1
; VF-FOUR-CHECK-NEXT: [[TMP3:%.*]] = sub i32 [[TMP0]], [[MUL_RESULT]]
; VF-FOUR-CHECK-NEXT: [[TMP4:%.*]] = icmp sgt i32 [[TMP3]], [[TMP0]]
; VF-FOUR-CHECK-NEXT: [[TMP5:%.*]] = or i1 [[TMP4]], [[MUL_OVERFLOW]]
; VF-FOUR-CHECK-NEXT: [[TMP6:%.*]] = icmp ugt i64 [[TMP1]], 4294967295
; VF-FOUR-CHECK-NEXT: [[TMP7:%.*]] = or i1 [[TMP5]], [[TMP6]]
; VF-FOUR-CHECK-NEXT: br i1 [[TMP7]], label [[VEC_EPILOG_SCALAR_PH]], label [[VECTOR_MAIN_LOOP_ITER_CHECK:%.*]]
; VF-FOUR-CHECK: vector.main.loop.iter.check:
; VF-FOUR-CHECK-NEXT: [[MIN_ITERS_CHECK1:%.*]] = icmp ult i64 [[WIDE_TRIP_COUNT]], 32
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_ITERS_CHECK1]], label [[VEC_EPILOG_PH:%.*]], label [[VECTOR_PH:%.*]]
; VF-FOUR-CHECK: vector.ph:
; VF-FOUR-CHECK-NEXT: [[N_MOD_VF:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 32
; VF-FOUR-CHECK-NEXT: [[N_VEC:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF]]
; VF-FOUR-CHECK-NEXT: br label [[VECTOR_BODY:%.*]]
; VF-FOUR-CHECK: vector.body:
; VF-FOUR-CHECK-NEXT: [[INDEX:%.*]] = phi i64 [ 0, [[VECTOR_PH]] ], [ [[INDEX_NEXT:%.*]], [[VECTOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[TMP8:%.*]] = add i64 [[INDEX]], 0
; VF-FOUR-CHECK-NEXT: [[OFFSET_IDX:%.*]] = trunc i64 [[INDEX]] to i32
; VF-FOUR-CHECK-NEXT: [[TMP16:%.*]] = add i32 [[OFFSET_IDX]], 0
; VF-FOUR-CHECK-NEXT: [[TMP24:%.*]] = xor i32 [[TMP16]], -1
; VF-FOUR-CHECK-NEXT: [[TMP32:%.*]] = add i32 [[TMP24]], [[N]]
; VF-FOUR-CHECK-NEXT: [[TMP40:%.*]] = sext i32 [[TMP32]] to i64
; VF-FOUR-CHECK-NEXT: [[TMP48:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[TMP40]]
; VF-FOUR-CHECK-NEXT: [[TMP56:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP57:%.*]] = getelementptr inbounds float, ptr [[TMP56]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP58:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -4
; VF-FOUR-CHECK-NEXT: [[TMP59:%.*]] = getelementptr inbounds float, ptr [[TMP58]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP60:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -8
; VF-FOUR-CHECK-NEXT: [[TMP61:%.*]] = getelementptr inbounds float, ptr [[TMP60]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP62:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -12
; VF-FOUR-CHECK-NEXT: [[TMP63:%.*]] = getelementptr inbounds float, ptr [[TMP62]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP64:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -16
; VF-FOUR-CHECK-NEXT: [[TMP65:%.*]] = getelementptr inbounds float, ptr [[TMP64]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP66:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -20
; VF-FOUR-CHECK-NEXT: [[TMP67:%.*]] = getelementptr inbounds float, ptr [[TMP66]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP68:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -24
; VF-FOUR-CHECK-NEXT: [[TMP69:%.*]] = getelementptr inbounds float, ptr [[TMP68]], i32 -3
; VF-FOUR-CHECK-NEXT: [[TMP70:%.*]] = getelementptr inbounds float, ptr [[TMP48]], i32 -28
; VF-FOUR-CHECK-NEXT: [[TMP71:%.*]] = getelementptr inbounds float, ptr [[TMP70]], i32 -3
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD:%.*]] = load <4 x float>, ptr [[TMP57]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE:%.*]] = shufflevector <4 x float> [[WIDE_LOAD]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD2:%.*]] = load <4 x float>, ptr [[TMP59]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE3:%.*]] = shufflevector <4 x float> [[WIDE_LOAD2]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD4:%.*]] = load <4 x float>, ptr [[TMP61]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE5:%.*]] = shufflevector <4 x float> [[WIDE_LOAD4]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD6:%.*]] = load <4 x float>, ptr [[TMP63]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE7:%.*]] = shufflevector <4 x float> [[WIDE_LOAD6]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD8:%.*]] = load <4 x float>, ptr [[TMP65]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE9:%.*]] = shufflevector <4 x float> [[WIDE_LOAD8]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD10:%.*]] = load <4 x float>, ptr [[TMP67]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE11:%.*]] = shufflevector <4 x float> [[WIDE_LOAD10]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD12:%.*]] = load <4 x float>, ptr [[TMP69]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE13:%.*]] = shufflevector <4 x float> [[WIDE_LOAD12]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD14:%.*]] = load <4 x float>, ptr [[TMP71]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE15:%.*]] = shufflevector <4 x float> [[WIDE_LOAD14]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[TMP72:%.*]] = fadd fast <4 x float> [[REVERSE]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP73:%.*]] = fadd fast <4 x float> [[REVERSE3]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP74:%.*]] = fadd fast <4 x float> [[REVERSE5]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP75:%.*]] = fadd fast <4 x float> [[REVERSE7]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP76:%.*]] = fadd fast <4 x float> [[REVERSE9]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP77:%.*]] = fadd fast <4 x float> [[REVERSE11]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP78:%.*]] = fadd fast <4 x float> [[REVERSE13]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP79:%.*]] = fadd fast <4 x float> [[REVERSE15]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP80:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[TMP8]]
; VF-FOUR-CHECK-NEXT: [[TMP88:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP89:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 4
; VF-FOUR-CHECK-NEXT: [[TMP90:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 8
; VF-FOUR-CHECK-NEXT: [[TMP91:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 12
; VF-FOUR-CHECK-NEXT: [[TMP92:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 16
; VF-FOUR-CHECK-NEXT: [[TMP93:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 20
; VF-FOUR-CHECK-NEXT: [[TMP94:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 24
; VF-FOUR-CHECK-NEXT: [[TMP95:%.*]] = getelementptr inbounds float, ptr [[TMP80]], i32 28
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP72]], ptr [[TMP88]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP73]], ptr [[TMP89]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP74]], ptr [[TMP90]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP75]], ptr [[TMP91]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP76]], ptr [[TMP92]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP77]], ptr [[TMP93]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP78]], ptr [[TMP94]], align 4
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP79]], ptr [[TMP95]], align 4
; VF-FOUR-CHECK-NEXT: [[INDEX_NEXT]] = add nuw i64 [[INDEX]], 32
; VF-FOUR-CHECK-NEXT: [[TMP96:%.*]] = icmp eq i64 [[INDEX_NEXT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: br i1 [[TMP96]], label [[MIDDLE_BLOCK:%.*]], label [[VECTOR_BODY]], !llvm.loop [[LOOP5:![0-9]+]]
; VF-FOUR-CHECK: middle.block:
; VF-FOUR-CHECK-NEXT: [[CMP_N:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: br i1 [[CMP_N]], label [[FOR_END_LOOPEXIT:%.*]], label [[VEC_EPILOG_ITER_CHECK:%.*]]
; VF-FOUR-CHECK: vec.epilog.iter.check:
; VF-FOUR-CHECK-NEXT: [[IND_END18:%.*]] = trunc i64 [[N_VEC]] to i32
; VF-FOUR-CHECK-NEXT: [[N_VEC_REMAINING:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_VEC]]
; VF-FOUR-CHECK-NEXT: [[MIN_EPILOG_ITERS_CHECK:%.*]] = icmp ult i64 [[N_VEC_REMAINING]], 4
; VF-FOUR-CHECK-NEXT: br i1 [[MIN_EPILOG_ITERS_CHECK]], label [[VEC_EPILOG_SCALAR_PH]], label [[VEC_EPILOG_PH]]
; VF-FOUR-CHECK: vec.epilog.ph:
; VF-FOUR-CHECK-NEXT: [[VEC_EPILOG_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_MAIN_LOOP_ITER_CHECK]] ]
; VF-FOUR-CHECK-NEXT: [[N_MOD_VF16:%.*]] = urem i64 [[WIDE_TRIP_COUNT]], 4
; VF-FOUR-CHECK-NEXT: [[N_VEC17:%.*]] = sub i64 [[WIDE_TRIP_COUNT]], [[N_MOD_VF16]]
; VF-FOUR-CHECK-NEXT: [[IND_END:%.*]] = trunc i64 [[N_VEC17]] to i32
; VF-FOUR-CHECK-NEXT: br label [[VEC_EPILOG_VECTOR_BODY:%.*]]
; VF-FOUR-CHECK: vec.epilog.vector.body:
; VF-FOUR-CHECK-NEXT: [[INDEX21:%.*]] = phi i64 [ [[VEC_EPILOG_RESUME_VAL]], [[VEC_EPILOG_PH]] ], [ [[INDEX_NEXT25:%.*]], [[VEC_EPILOG_VECTOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[TMP97:%.*]] = add i64 [[INDEX21]], 0
; VF-FOUR-CHECK-NEXT: [[OFFSET_IDX22:%.*]] = trunc i64 [[INDEX21]] to i32
; VF-FOUR-CHECK-NEXT: [[TMP98:%.*]] = add i32 [[OFFSET_IDX22]], 0
; VF-FOUR-CHECK-NEXT: [[TMP99:%.*]] = xor i32 [[TMP98]], -1
; VF-FOUR-CHECK-NEXT: [[TMP100:%.*]] = add i32 [[TMP99]], [[N]]
; VF-FOUR-CHECK-NEXT: [[TMP101:%.*]] = sext i32 [[TMP100]] to i64
; VF-FOUR-CHECK-NEXT: [[TMP102:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[TMP101]]
; VF-FOUR-CHECK-NEXT: [[TMP103:%.*]] = getelementptr inbounds float, ptr [[TMP102]], i32 0
; VF-FOUR-CHECK-NEXT: [[TMP104:%.*]] = getelementptr inbounds float, ptr [[TMP103]], i32 -3
; VF-FOUR-CHECK-NEXT: [[WIDE_LOAD23:%.*]] = load <4 x float>, ptr [[TMP104]], align 4
; VF-FOUR-CHECK-NEXT: [[REVERSE24:%.*]] = shufflevector <4 x float> [[WIDE_LOAD23]], <4 x float> poison, <4 x i32> <i32 3, i32 2, i32 1, i32 0>
; VF-FOUR-CHECK-NEXT: [[TMP105:%.*]] = fadd fast <4 x float> [[REVERSE24]], splat (float 1.000000e+00)
; VF-FOUR-CHECK-NEXT: [[TMP106:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[TMP97]]
; VF-FOUR-CHECK-NEXT: [[TMP107:%.*]] = getelementptr inbounds float, ptr [[TMP106]], i32 0
; VF-FOUR-CHECK-NEXT: store <4 x float> [[TMP105]], ptr [[TMP107]], align 4
; VF-FOUR-CHECK-NEXT: [[INDEX_NEXT25]] = add nuw i64 [[INDEX21]], 4
; VF-FOUR-CHECK-NEXT: [[TMP108:%.*]] = icmp eq i64 [[INDEX_NEXT25]], [[N_VEC17]]
; VF-FOUR-CHECK-NEXT: br i1 [[TMP108]], label [[VEC_EPILOG_MIDDLE_BLOCK:%.*]], label [[VEC_EPILOG_VECTOR_BODY]], !llvm.loop [[LOOP6:![0-9]+]]
; VF-FOUR-CHECK: vec.epilog.middle.block:
; VF-FOUR-CHECK-NEXT: [[CMP_N20:%.*]] = icmp eq i64 [[WIDE_TRIP_COUNT]], [[N_VEC17]]
; VF-FOUR-CHECK-NEXT: br i1 [[CMP_N20]], label [[FOR_END_LOOPEXIT]], label [[VEC_EPILOG_SCALAR_PH]]
; VF-FOUR-CHECK: vec.epilog.scalar.ph:
; VF-FOUR-CHECK-NEXT: [[BC_RESUME_VAL:%.*]] = phi i64 [ [[N_VEC17]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[N_VEC]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-FOUR-CHECK-NEXT: [[BC_RESUME_VAL19:%.*]] = phi i32 [ [[IND_END]], [[VEC_EPILOG_MIDDLE_BLOCK]] ], [ [[IND_END18]], [[VEC_EPILOG_ITER_CHECK]] ], [ 0, [[VECTOR_SCEVCHECK]] ], [ 0, [[ITER_CHECK]] ]
; VF-FOUR-CHECK-NEXT: br label [[FOR_BODY:%.*]]
; VF-FOUR-CHECK: for.body:
; VF-FOUR-CHECK-NEXT: [[INDVARS_IV:%.*]] = phi i64 [ [[BC_RESUME_VAL]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[I_014:%.*]] = phi i32 [ [[BC_RESUME_VAL19]], [[VEC_EPILOG_SCALAR_PH]] ], [ [[INC:%.*]], [[FOR_BODY]] ]
; VF-FOUR-CHECK-NEXT: [[TMP109:%.*]] = xor i32 [[I_014]], -1
; VF-FOUR-CHECK-NEXT: [[SUB2:%.*]] = add i32 [[TMP109]], [[N]]
; VF-FOUR-CHECK-NEXT: [[IDXPROM:%.*]] = sext i32 [[SUB2]] to i64
; VF-FOUR-CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds float, ptr [[B]], i64 [[IDXPROM]]
; VF-FOUR-CHECK-NEXT: [[TMP110:%.*]] = load float, ptr [[ARRAYIDX]], align 4
; VF-FOUR-CHECK-NEXT: [[CONV3:%.*]] = fadd fast float [[TMP110]], 1.000000e+00
; VF-FOUR-CHECK-NEXT: [[ARRAYIDX5:%.*]] = getelementptr inbounds float, ptr [[A]], i64 [[INDVARS_IV]]
; VF-FOUR-CHECK-NEXT: store float [[CONV3]], ptr [[ARRAYIDX5]], align 4
; VF-FOUR-CHECK-NEXT: [[INDVARS_IV_NEXT]] = add nuw nsw i64 [[INDVARS_IV]], 1
; VF-FOUR-CHECK-NEXT: [[INC]] = add nuw nsw i32 [[I_014]], 1
; VF-FOUR-CHECK-NEXT: [[EXITCOND:%.*]] = icmp ne i64 [[INDVARS_IV_NEXT]], [[WIDE_TRIP_COUNT]]
; VF-FOUR-CHECK-NEXT: br i1 [[EXITCOND]], label [[FOR_BODY]], label [[FOR_END_LOOPEXIT]], !llvm.loop [[LOOP7:![0-9]+]]
; VF-FOUR-CHECK: for.end.loopexit:
; VF-FOUR-CHECK-NEXT: br label [[FOR_END]]
; VF-FOUR-CHECK: for.end:
; VF-FOUR-CHECK-NEXT: ret i32 0
;
entry:
%cmp1 = icmp sgt i32 %n, 1
br i1 %cmp1, label %for.body.preheader, label %for.end
for.body.preheader: ; preds = %entry
%0 = add i32 %n, -1
%wide.trip.count = zext i32 %0 to i64
br label %for.body
for.body: ; preds = %for.body.preheader, %for.body
%indvars.iv = phi i64 [ 0, %for.body.preheader ], [ %indvars.iv.next, %for.body ]
%i.014 = phi i32 [ 0, %for.body.preheader ], [ %inc, %for.body ]
%1 = xor i32 %i.014, -1
%sub2 = add i32 %1, %n
%idxprom = sext i32 %sub2 to i64
%arrayidx = getelementptr inbounds float, ptr %B, i64 %idxprom
%2 = load float, ptr %arrayidx, align 4
%conv3 = fadd fast float %2, 1.000000e+00
%arrayidx5 = getelementptr inbounds float, ptr %A, i64 %indvars.iv
store float %conv3, ptr %arrayidx5, align 4
%indvars.iv.next = add nuw nsw i64 %indvars.iv, 1
%inc = add nuw nsw i32 %i.014, 1
%exitcond = icmp ne i64 %indvars.iv.next, %wide.trip.count
br i1 %exitcond, label %for.body, label %for.end.loopexit
for.end.loopexit: ; preds = %for.body
br label %for.end
for.end: ; preds = %for.end.loopexit, %entry
ret i32 0
}
;
;
attributes #0 = { nounwind "correctly-rounded-divide-sqrt-fp-math"="false" "disable-tail-calls"="false" "frame-pointer"="none" "less-precise-fpmad"="false" "min-legal-vector-width"="0" "no-infs-fp-math"="true" "no-jump-tables"="false" "no-nans-fp-math"="true" "no-signed-zeros-fp-math"="true" "no-trapping-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="ppc64le" "target-features"="+altivec,+bpermd,+crypto,+direct-move,+extdiv,+htm,+power8-vector,+vsx,-power9-vector,-spe" "unsafe-fp-math"="true" "use-soft-float"="false" }
Reference in New Issue View Git Blame Copy Permalink