David Sherwood f575b18fdc [LV] Add support for partial reductions without a binary op (#133922) ...

Consider IR such as this:

for.body:
  %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ]
  %accum = phi i32 [ 0, %entry ], [ %add, %for.body ]
  %gep.a = getelementptr i8, ptr %a, i64 %iv
  %load.a = load i8, ptr %gep.a, align 1
  %ext.a = zext i8 %load.a to i32
  %add = add i32 %ext.a, %accum
  %iv.next = add i64 %iv, 1
  %exitcond.not = icmp eq i64 %iv.next, 1025
  br i1 %exitcond.not, label %for.exit, label %for.body

Conceptually we can vectorise this using partial reductions too,
although the current loop vectoriser implementation requires the
accumulation of a multiply. For AArch64 this is easily done with
a udot or sdot with an identity operand, i.e. a vector of (i16 1).

In order to do this I had to teach getScaledReductions that the
accumulated value may come from a unary op, hence there is only
one extension to consider. Similarly, I updated the vplan and
AArch64 TTI cost model to understand the possible unary op.

---------

Co-authored-by: Matt Devereau <matthew.devereau@arm.com>

2025-07-02 13:05:51 +01:00

..

AArch64

[LV] Add support for partial reductions without a binary op (#133922)

2025-07-02 13:05:51 +01:00

AMDGPU

[AMDGPU] Add custom MachineValueType entries for buffer fat poiners (#127692)

2025-03-04 17:19:06 -06:00

ARM

[VPlan] Add VPExpressionRecipe, replacing extended reduction recipes. (#144281)

2025-07-01 20:44:50 +01:00

Hexagon

[LV] Bail out early if BTC+1 wraps.

2025-01-14 22:07:38 +00:00

LoongArch

[LV] Move VPlan-based calculateRegisterUsage to VPlanAnalysis (NFC). (#135673)

2025-06-02 17:40:50 +01:00

PowerPC

[VPlan] Add ReductionStartVector VPInstruction. (#142290)

2025-06-09 20:59:12 +01:00

RISCV

[VPlan] Emit VPVectorEndPointerRecipe for reverse interleave pointer adjustment (#144864)

2025-07-02 18:16:02 +08:00

SystemZ

[VPlan] Introduce constant folder in simplifyRecipe (#125365)

2025-05-20 14:16:01 +01:00

VE

…

WebAssembly

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

X86

[VPlan] Truncate/Extend ComputeReductionResult at construction (NFC). (#141860)

2025-06-30 22:39:17 +01:00

XCore

…

12-12-11-if-conv.ll

…

2012-10-20-infloop.ll

…

2012-10-22-isconsec.ll

…

2016-07-27-loop-vec.ll

…

alias-set-with-uncomputable-bounds.ll

…

align.ll

…

as_cast.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

assume.ll

…

blend-in-header.ll

[VPlan] Simplify BLEND %a, %b, NOT(%m) -> BLEND %b, %a, %m. (#128375)

2025-02-27 17:43:24 +00:00

branch-weights.ll

[LV][NFC] Add branch weight test showing incorrect behaviour (#144682)

2025-06-19 10:49:40 +01:00

bsd_regex.ll

…

bzip_reverse_loops.ll

…

calloc.ll

…

cast-induction.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

check-no-vectorize.ll

…

check-prof-info.ll

[LV] Use vscale for tuning when updating profile information (#143690)

2025-06-16 10:02:38 +01:00

conditional-assignment.ll

…

consec_no_gep.ll

…

consecutive-ptr-uniforms.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

constantfolder-infer-correct-gepty.ll

[VPlan] Introduce constant folder in simplifyRecipe (#125365)

2025-05-20 14:16:01 +01:00

constantfolder.ll

[VPlan] Use InstSimplifyFolder instead of TargetFolder (#141222)

2025-05-28 11:00:14 +02:00

control-flow.ll

…

cpp-new-array.ll

…

create-induction-resume.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

dbg-outer-loop-vect.ll

[VPlan] Preserve dbg location on canonical IVs in native path.

2025-04-29 21:40:42 +01:00

dbg.value.ll

Reapply "Verifier: Add check for DICompositeType elements being null"

2025-01-23 22:29:30 +00:00

dead_instructions.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

debugloc-optimize-vfuf-term.ll

[VPlan] Introduce constant folder in simplifyRecipe (#125365)

2025-05-20 14:16:01 +01:00

debugloc.ll

Reapply "[VPlan] Fold NOT into predicate of wide compares." (#130347)

2025-04-28 20:01:37 +01:00

demanded-bits-of-pointer-instruction.ll

…

dereferenceable-info-from-assumption-constant-size.ll

[LV] Don't mark ptrs as safe to speculate if fed by UB/poison op. (#143204)

2025-06-20 13:05:19 +01:00

diag-missing-instr-debug-loc.ll

…

diag-with-hotness-info-2.ll

…

diag-with-hotness-info.ll

…

disable_nonforced_enable.ll

…

disable_nonforced.ll

…

discriminator.ll

[Verifier] Require that dbg.declare variable is a ptr (#134355)

2025-04-04 15:34:45 +02:00

dont-fold-tail-for-const-TC.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

dont-fold-tail-for-divisible-TC.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

early_exit_legality.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

early_exit_store_legality.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

ee-crash.ll

…

epilog-iv-select-cmp.ll

[VPlan] Add ReductionStartVector VPInstruction. (#142290)

2025-06-09 20:59:12 +01:00

epilog-vectorization-any-of-reductions.ll

[VPlan] Expand VPWidenIntOrFpInductionRecipe into separate recipes (#118638)

2025-06-17 18:24:07 +01:00

epilog-vectorization-reductions.ll

[VPlan] Truncate/Extend ComputeReductionResult at construction (NFC). (#141860)

2025-06-30 22:39:17 +01:00

epilog-vectorization-trunc-induction-steps.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

epilog-vectorization-vector-trip-count-zero.ll

[LV] Bail out early if BTC+1 wraps.

2025-01-14 22:07:38 +00:00

exact.ll

…

expand-scev-after-invoke.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

explicit_outer_detection.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

explicit_outer_nonuniform_inner.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

explicit_outer_uniform_diverg_branch.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

extract-from-end-vector-constant.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

extract-last-veclane.ll

…

fcmp-vectorize.ll

…

first-order-recurrence-chains-vplan.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

first-order-recurrence-chains.ll

[LV] Add users to some first-order recurrence tests.

2025-06-15 21:24:10 +01:00

first-order-recurrence-complex.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

first-order-recurrence-dead-instructions.ll

[VPlan] Simplify ExtractLastElement(Broadcast(A)) -> A.

2025-06-20 21:08:14 +01:00

first-order-recurrence-interleave-only.ll

[VPlan] Mark VPFirstOrderRecurrencePHI as not reading/writing memory.

2025-06-15 22:00:47 +01:00

first-order-recurrence-multiply-recurrences.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

first-order-recurrence-scalable-vf1.ll

[LoopVectorize] Vectorize fixed-order recurrence with vscale x 1. (#142772)

2025-06-18 16:03:20 +08:00

first-order-recurrence-sink-replicate-region.ll

[LV] Replace redundant ExtractLastElement of reduction result (NFC).

2025-06-24 21:48:58 +01:00

first-order-recurrence.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

fix-reduction-dbg.ll

…

fixed-order-recurrences-memory-instructions.ll

…

flags.ll

…

float-induction.ll

[VPlan] Also duplicated scalar-steps when it enables sinking scalars. (#136021)

2025-04-21 18:36:43 +01:00

float-minmax-instruction-flag.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

float-reduction.ll

…

fmax-without-fast-math-flags-interleave.ll

[LV] Add tests with fmax reductions without fast-math flags.

2025-06-22 20:35:38 +01:00

fmax-without-fast-math-flags.ll

[LV] Add tests with fmax reductions without fast-math flags.

2025-06-22 20:35:38 +01:00

fneg.ll

…

followup.ll

…

forked-pointers.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

fpsat.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

funcall.ll

…

gcc-examples.ll

…

gep_with_bitcast.ll

…

global_alias.ll

…

hints-trans.ll

…

histograms.ll

…

hoist-loads.ll

…

i8-induction.ll

[RemoveDIs] Remove "try-debuginfo-iterators..." test flags (#130298)

2025-03-14 15:50:49 +00:00

icmp-uniforms.ll

[VPlan] Connect Entry to scalar preheader during initial construction. (#140132)

2025-05-27 16:07:56 +01:00

if-conv-crash.ll

…

if-conversion-edgemasks.ll

…

if-conversion-nest.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

if-conversion-reduction.ll

…

if-conversion.ll

[VPlan] Simplify BLEND %a, %b, NOT(%m) -> BLEND %b, %a, %m. (#128375)

2025-02-27 17:43:24 +00:00

if-pred-non-void.ll

[VPlan] Set debug location for recipes in VPBB::executeRecipes.

2025-04-05 16:22:47 +01:00

if-pred-not-when-safe.ll

…

if-pred-stores.ll

[VPlan] Add ReductionStartVector VPInstruction. (#142290)

2025-06-09 20:59:12 +01:00

if-reduction.ll

[LV] Re-org tests; introduce iv-select-cmp-decreasing.ll (#141769)

2025-06-03 18:13:51 +01:00

incorrect-dom-info.ll

…

increment.ll

…

induction_plus.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

induction-multiple-uses-in-same-instruction.ll

…

induction-ptrcasts.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

induction-step.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

induction-unroll-novec.ll

[LV] Regen a test with UTC (#134076)

2025-04-02 17:23:00 +01:00

induction.ll

[VPlan] Add ReductionStartVector VPInstruction. (#142290)

2025-06-09 20:59:12 +01:00

infiniteloop.ll

…

instruction-only-used-outside-of-loop.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

int_sideeffect.ll

…

interleave-and-scalarize-only.ll

[VPlan] Use EMIT-SCALAR when printing casts.

2025-06-21 10:23:53 +01:00

interleave-with-i65-induction.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

interleaved-accesses-1.ll

…

interleaved-accesses-2.ll

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

interleaved-accesses-3.ll

[VPlan] Add incoming values for all predecessor to ResumePHI (NFCI).

2025-02-09 11:20:20 +00:00

interleaved-accesses-alias.ll

…

interleaved-accesses-different-insert-position.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

interleaved-accesses-masked-group.ll

…

interleaved-accesses-max-dependences.ll

…

interleaved-accesses-pred-stores.ll

…

interleaved-accesses-uniform-load.ll

…

interleaved-accesses.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

interleaved-acess-with-remarks.ll

…

intrinsic.ll

…

invalidate-scev-at-scope-after-vectorization.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

invariant-store-vectorization-2.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

invariant-store-vectorization.ll

[VPlan] Introduce explicit broadcasts for live-ins. (#124644)

2025-02-26 13:57:51 +00:00

irregular_type.ll

…

is_fpclass.ll

[VPlan] Simplify select c, x, x -> x (#133731)

2025-04-02 10:26:48 +01:00

iv_outside_user.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

iv-select-cmp-blend.ll

…

iv-select-cmp-decreasing.ll

[LV] Add support for cmp reductions with decreasing IVs. (#140451)

2025-06-29 11:17:03 +01:00

iv-select-cmp-nested-loop.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

iv-select-cmp-no-wrap.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

iv-select-cmp-non-const-iv-start.ll

…

iv-select-cmp-trunc.ll

[LV] Extend FindLastIV to unsigned case (#141752)

2025-06-23 15:27:49 +01:00

iv-select-cmp.ll

[LV] Extend FindLastIV to unsigned case (#141752)

2025-06-23 15:27:49 +01:00

lcssa-crashes.ll

…

libcall-remark.ll

…

lifetime.ll

…

load-deref-pred-align.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

load-deref-pred-neg-off.ll

[VPlan] Use State.get to extract lane mask for BranchOnMask.

2025-01-27 21:35:36 +00:00

load-deref-pred-poison-ub-ops-feeding-pointer.ll

[LV] Don't mark ptrs as safe to speculate if fed by UB/poison op. (#143204)

2025-06-20 13:05:19 +01:00

load-of-struct-deref-pred.ll

[VPlan] Also duplicated scalar-steps when it enables sinking scalars. (#136021)

2025-04-21 18:36:43 +01:00

loop-form.ll

Reapply "[VPlan] Fold NOT into predicate of wide compares." (#130347)

2025-04-28 20:01:37 +01:00

loop-legality-checks.ll

…

loop-scalars.ll

…

loop-vect-memdep.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

loop-vect-option.ll

…

loop-with-constant-exit-condition.ll

[VPlan] Remove stale assertion in HCFG builder.

2025-02-20 17:01:49 +01:00

make-followup-loop-id.ll

[clang][CodeGen] Generate follow-up metadata for loops in correct format (#131985)

2025-03-27 20:29:37 +09:00

memdep-fold-tail.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

memdep.ll

…

memory-dep-remarks.ll

…

metadata-unroll.ll

…

metadata-width.ll

…

metadata.ll

[LV] Extend metadata.ll tests to also check interleaving.

2025-04-14 21:51:15 +02:00

middle-block-dbg.ll

…

min-trip-count-known-via-scev.ll

[SCEV] Add additional test coverage for loop-guards reasoning.

2025-06-01 22:39:37 +01:00

minimumnum-maximumnum-reductions.ll

[IVDescriptors] Support reductions with minimumnum/maximumnum. (#137335)

2025-04-28 11:16:36 +01:00

miniters.ll

…

minmax_reduction.ll

…

mixed-precision-remarks.ll

…

multi_early_exit_live_outs.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

multi_early_exit.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

multi-use-reduction-bug.ll

…

multiple-address-spaces.ll

…

multiple-exits-versioning.ll

…

multiple-result-intrinsics.ll

[LV] Teach the vectorizer to cost and vectorize modf and sincospi intrinsics (#129064)

2025-02-28 12:56:12 +00:00

multiple-strides-vectorization.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

narrow-to-single-scalar.ll

[VPlan] Support VPWidenSelectRecipe in narrowToSingleScalar.

2025-06-27 15:42:42 +01:00

no_array_bounds_scalable.ll

…

no_array_bounds.ll

…

no_idiv_reduction.ll

…

no_int_induction.ll

…

no_outside_user.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

no_switch_disable_vectorization.ll

…

no_switch.ll

…

no-fold-tail-by-masking-iv-external-uses.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

no-interleave-up-front.ll

…

noalias-md-licm.ll

…

noalias-md.ll

…

noalias-scope-decl.ll

…

nofloat-report.ll

…

nofloat.ll

…

non-const-n.ll

…

nontemporal.ll

…

nounroll.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

novect-lcssa-cfg-invalidation.ll

…

nsw-crash.ll

…

nuw.ll

…

opaque-ptr.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

opt.ll

…

optimal-epilog-vectorization-limitations.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

optimal-epilog-vectorization-liveout.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

optimal-epilog-vectorization-scalable.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

optimal-epilog-vectorization.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

optsize.ll

[VPlan] Mark VPFirstOrderRecurrencePHI as not reading/writing memory.

2025-06-15 22:00:47 +01:00

outer_loop_early_exit.ll

…

outer_loop_hcfg_construction.ll

[VPlan] Refactor VPlan creation, add transform introducing region (NFC). (#128419)

2025-03-09 15:05:35 +00:00

outer_loop_scalable.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

outer_loop_test1.ll

…

outer_loop_test2.ll

…

outer-loop-inner-latch-successors.ll

Reapply "[VPlan] Fold NOT into predicate of wide compares." (#130347)

2025-04-28 20:01:37 +01:00

outer-loop-vec-phi-predecessor-order.ll

…

outer-loop-wide-phis.ll

[VPlan] Use VPlan predecessors in VPWidenPHIRecipe (NFC). (#126388)

2025-02-17 16:40:37 +01:00

partial-lcssa.ll

…

phi-cost.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

phi-hang.ll

…

pointer-induction-index-width-smaller-than-iv-width.ll

[VPlan] Truncate VFxUF if needed in VPWidenPointerInduction::execute.

2025-03-16 11:37:58 +00:00

pointer-induction-unroll.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

pointer-induction.ll

[VPlan] Expand VPWidenIntOrFpInductionRecipe into separate recipes (#118638)

2025-06-17 18:24:07 +01:00

pointer-select-runtime-checks.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

pr25281.ll

…

pr28541.ll

…

pr30654-phiscev-sext-trunc.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pr30806.ll

…

pr31098.ll

…

pr31190.ll

…

pr32859.ll

…

pr33706.ll

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

pr34681.ll

…

pr35743.ll

…

pr35773.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pr36311.ll

…

pr36983.ll

…

pr37248.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pr37515.ll

…

pr38697.ll

…

pr38800.ll

…

pr39099.ll

…

pr39417-optsize-scevchecks.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pr43166-fold-tail-by-masking.ll

…

pr44488-predication.ll

…

pr45259.ll

[VPlan] Only execute VPExpandSCEVRecipes once and remove them (NFC).

2025-03-23 09:06:01 +00:00

pr45525.ll

…

pr45679-fold-tail-by-masking.ll

[LoopVectorize] Don't replicate blocks with optsize (#129265)

2025-04-17 11:50:49 +01:00

pr46525-expander-insertpoint.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

pr47343-expander-lcssa-after-cfg-update.ll

[VPlan] Add incoming values for all predecessor to ResumePHI (NFCI).

2025-02-09 11:20:20 +00:00

pr48832.ll

…

pr50686.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pr51614-fold-tail-by-masking.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

pr55100-expand-scev-predicate-used.ll

…

pr55167-fold-tail-live-out.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

pr58811-scev-expansion.ll

…

pr59319-loop-access-info-invalidation.ll

[VPlan] Create IR basic block for middle.block in VPlan.

2025-02-15 21:54:16 +01:00

pr66616.ll

[VPlan] Create IR basic block for middle.block in VPlan.

2025-02-15 21:54:16 +01:00

pr87407-trunc-with-intrinsic.ll

[LV] Fix MinBWs in WidenIntrinsic case (#137005)

2025-04-29 09:47:38 +01:00

pr125278.ll

[LV] Strip unmaintainable MinBWs assert (#136858)

2025-05-08 11:49:54 +01:00

predicate-switch.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

predicatedinst-loop-invariant.ll

[LV] Stengthen loop-invariance checks in isPredicatedInst (#140744)

2025-06-20 14:01:48 +01:00

preserve-dbg-loc-and-loop-metadata.ll

[VPlan] Don't access canonical IV in VPWidenPointerInduction::execute.

2025-03-15 21:32:48 +00:00

preserve-dbg-loc-reduction-inloop.ll

[VPlan] Set debug location for VPReduction/VPWidenIntrinsicRecipe. (#120054)

2024-12-27 10:37:21 +08:00

preserve-or-disjoint.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

pseudoprobe.ll

…

ptr_loops.ll

…

ptr-induction.ll

…

read-only.ll

…

reduction-align.ll

[VPlan] Add incoming values for all predecessor to ResumePHI (NFCI).

2025-02-09 11:20:20 +00:00

reduction-inloop-cond.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

reduction-inloop-min-max.ll

[IR] Convert from nocapture to captures(none) (#123181)

2025-01-29 16:56:47 +01:00

reduction-inloop-pred.ll

[VPlan] Expand VPWidenIntOrFpInductionRecipe into separate recipes (#118638)

2025-06-17 18:24:07 +01:00

reduction-inloop-uf4.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

reduction-inloop.ll

[VPlan] Implement VPlan-based cost model for VPReduction, VPExtendedReduction and VPMulAccumulateReduction. (#113903)

2025-05-29 11:15:16 +08:00

reduction-odd-interleave-counts.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

reduction-order.ll

…

reduction-predselect.ll

[InstCombine] Handle "add like" in ADD+GEP->GEP+GEP rewrites (#135156)

2025-04-14 17:11:13 +02:00

reduction-ptr.ll

…

reduction-small-size.ll

[VPlan] Truncate/Extend ComputeReductionResult at construction (NFC). (#141860)

2025-06-30 22:39:17 +01:00

reduction-with-invariant-store.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

reduction.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

remarks-multi-exit-loops.ll

…

remarks-reduction-inloop.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

remove_metadata.ll

…

reverse_induction.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

reverse_iter.ll

…

runtime-check-address-space.ll

…

runtime-check-needed-but-empty.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

runtime-check-pointer-element-type.ll

…

runtime-check-readonly-address-space.ll

…

runtime-check-readonly.ll

…

runtime-check-small-clamped-bounds.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

runtime-check.ll

[LV] Exercise type-mismatch with RT-check conflict rdx (#130295)

2025-04-02 17:22:40 +01:00

runtime-checks-difference-simplifications.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

runtime-checks-difference.ll

[LV] Exercise type-mismatch with RT-check conflict rdx (#130295)

2025-04-02 17:22:40 +01:00

runtime-checks-hoist.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

runtime-drop-crash.ll

…

safegep.ll

…

same-base-access.ll

…

scalable-assume.ll

Recommit "[VPlan] Try to narrow wide and replicating recipes to uniform recipes."

2025-01-12 20:10:28 +00:00

scalable-first-order-recurrence.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-inductions.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-iv-outside-user.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-lifetime.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-loop-unpredicated-body-scalar-tail.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-noalias-scope-decl.ll

…

scalable-predication.ll

[LV] Consider whether vscale is a known power of two for iteration check (#144963)

2025-06-20 11:37:27 -07:00

scalable-reduction-inloop.ll

[VPlan] Truncate/Extend ComputeReductionResult at construction (NFC). (#141860)

2025-06-30 22:39:17 +01:00

scalable-trunc-min-bitwidth.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

scalable-vf-hint.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

scalar_after_vectorization.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

scalar-select.ll

…

scalarize-masked-call.ll

…

scalarized-bitcast.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

scev-during-mutation.ll

…

scev-exit-phi-invalidation.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

scev-exitlim-crash.ll

…

scev-invalidation.ll

…

scev-predicate-reasoning.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

select-cmp-multiuse.ll

[VPlan] Also duplicated scalar-steps when it enables sinking scalars. (#136021)

2025-04-21 18:36:43 +01:00

select-cmp-predicated.ll

[VPlan] Also duplicated scalar-steps when it enables sinking scalars. (#136021)

2025-04-21 18:36:43 +01:00

select-cmp.ll

Reapply "[VPlan] Fold NOT into predicate of wide compares." (#130347)

2025-04-28 20:01:37 +01:00

select-index-interleaving.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-reduction-start-value-may-be-undef-or-poison.ll

Reapply "[VPlan] Fold NOT into predicate of wide compares." (#130347)

2025-04-28 20:01:37 +01:00

select-reduction.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

select-smax-last-index.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-smin-last-index.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-umax-last-index.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-umin-first-index.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-umin-last-index.ll

[LV] Reorganize index select tests (NFC).

2025-06-08 22:34:07 +01:00

select-with-fastflags.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

simple-unroll.ll

…

single_early_exit_live_outs.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

single_early_exit_unsafe_ptrs.ll

…

single_early_exit_with_outer_loop.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

single_early_exit.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

single-early-exit-interleave-hint.ll

[LV] Fix test issue caused by #145877 (#146041)

2025-06-27 10:02:57 +01:00

single-early-exit-interleave.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

single-scalar-cast-minbw.ll

[LV] Pre-commit test case for support VPWidenCastRecipe in isSingleScalar. nfc (#143498)

2025-06-12 21:06:58 +08:00

single-value-blend-phis.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

skeleton-lcssa-crash.ll

[VPlan] Use InstSimplifyFolder instead of TargetFolder (#141222)

2025-05-28 11:00:14 +02:00

skip-iterations.ll

…

small-loop.ll

…

smax-idx.ll

…

start-non-zero.ll

…

store-reduction-results-in-tail-folded-loop.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

store-shuffle-bug.ll

…

strict-fadd-interleave-only.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

strided-accesses-interleave-only.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

struct_access.ll

…

struct-return-replicate.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

struct-return.ll

[LV] Add additional tests for replicating calls returning structs.

2025-06-22 13:48:25 +01:00

tail-folding-alloca-in-loop.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

tail-folding-counting-down.ll

[VPlan] Use VPlan operand order for VPBlendRecipes. (#139475)

2025-05-14 14:56:35 +01:00

tail-folding-optimize-vector-induction-width.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

tail-folding-switch.ll

[VPlan] Add initial CFG simplification, removing BranchOnCond true. (#106748)

2025-04-04 15:44:26 +01:00

tail-folding-vectorization-factor-1.ll

[VPlan] Remove ILV::sinkScalarOperands. (#136023)

2025-04-24 08:37:49 +01:00

tbaa-nodep.ll

…

trip-count-expansion-may-introduce-ub.ll

[VPlan] Also duplicated scalar-steps when it enables sinking scalars. (#136021)

2025-04-21 18:36:43 +01:00

tripcount.ll

…

trunc-extended-icmps.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

trunc-loads-p16.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

trunc-reductions.ll

…

trunc-shifts.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

uitofp-preserve-nneg.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

uncountable-early-exit-vplan.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

uncountable-single-exit-loops.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

undef-inst-bug.ll

…

uniform_across_vf_induction1_and.ll

[LV] Regenerate uniform_across_vf* check lines.

2025-06-20 19:10:26 +01:00

uniform_across_vf_induction1_div_urem.ll

[LV] Regenerate uniform_across_vf* check lines.

2025-06-20 19:10:26 +01:00

uniform_across_vf_induction1_lshr.ll

[LV] Regenerate uniform_across_vf* check lines.

2025-06-20 19:10:26 +01:00

uniform_across_vf_induction1.ll

[LV] Regenerate uniform_across_vf* check lines.

2025-06-20 19:10:26 +01:00

uniform_across_vf_induction2.ll

[LV] Regenerate uniform_across_vf* check lines.

2025-06-20 19:10:26 +01:00

uniform-args-call-variants.ll

…

uniform-blend.ll

[VPlan] Unroll VPReplicateRecipe by VF. (#142433)

2025-06-26 11:19:09 +01:00

unroll_nonlatch.ll

[VPlan] Generalize SCALAR-STEPS removal to any unroll factor.

2025-03-26 21:03:50 +00:00

unroll_novec.ll

…

unroll-novec-memcheck-metadata.ll

…

unroll.ll

…

unsafe-dep-remark.ll

…

unsafe-vf-hint-remark.ll

…

unsupported_early_exit.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

unused-blend-mask-for-first-operand.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

use-iv-start-value.ll

…

use-scalar-epilogue-if-tp-fails.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

value-ptr-bug.ll

…

vect-phiscev-sext-trunc.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

vect.omp.persistence.ll

…

vect.stats.ll

[LoopVectorize][NFC] Disable output for tests that don't need it (#124747)

2025-01-29 08:09:50 +00:00

vector-freeze.ll

…

vector-geps.ll

…

vector-intrinsic-call-cost.ll

[VPlan] Remove no-op SCALAR-STEPS after unrolling. (#123655)

2025-03-25 12:57:24 +00:00

vector-loop-backedge-elimination-branch-weights.ll

[LV] Add test case with branch weights.

2025-06-14 16:48:44 +01:00

vector-loop-backedge-elimination-early-exit.ll

[LV] Enable auto-vectorisation of loops with uncountable exits (#133099)

2025-06-27 10:39:33 +01:00

vector-loop-backedge-elimination-outside-iv-users.ll

[VPlan] Use InstSimplifyFolder instead of TargetFolder (#141222)

2025-05-28 11:00:14 +02:00

vector-loop-backedge-elimination.ll

[VPlan] Use InstSimplifyFolder instead of TargetFolder (#141222)

2025-05-28 11:00:14 +02:00

vector-no-scevcheck.ll

…

vector-to-scalar-cast.ll

…

vectorize-force-tail-with-evl.ll

[LLVM][IRBuilder] Use NUW arithmetic for Create{ElementCount,TypeSize}. (#143532)

2025-06-19 13:24:39 +01:00

vectorize-once.ll

…

vectorize-pointer-phis.ll

…

vectorizeVFone.ll

…

version-mem-access.ll

…

version-stride-with-integer-casts.ll

[VPlan] Use InstSimplifyFolder instead of TargetFolder (#141222)

2025-05-28 11:00:14 +02:00

vplan_hcfg_stress_test.ll

[VPlan] Move plain CFG construction to VPlanConstruction. (NFC)

2025-04-18 21:52:05 +01:00

vplan-dot-printing.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-force-tail-with-evl.ll

…

vplan-incomplete-cases.ll

…

vplan-iv-transforms.ll

[VPlan] Use createScalarZExtOrTrunc when expanding expandVPWidenIntOrFpInduction

2025-06-20 19:18:49 +01:00

vplan-native-path-inner-loop-with-runtime-checks.ll

[LV] Don't crash on inner loops with RT checks in VPlan-native path.

2025-03-10 20:28:56 +00:00

vplan-outer-loop-uncomputable-trip-count.ll

…

vplan-predicate-switch.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-printing-before-execute.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-printing-outer-loop.ll

[VPlan] Add printPhiOperands to VPPhiAccessors, use for wide phis.

2025-05-08 20:56:48 +01:00

vplan-printing-reductions.ll

[VPlan] Add VPExpressionRecipe, replacing extended reduction recipes. (#144281)

2025-07-01 20:44:50 +01:00

vplan-printing.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-sink-scalars-and-merge-vf1.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-sink-scalars-and-merge.ll

[VPlan] Use regular phi printing for resume phis.

2025-06-07 21:49:54 +01:00

vplan-stress-test-no-explict-vf.ll

…

vplan-unused-interleave-group.ll

[VPlan] Connect Entry to scalar preheader during initial construction. (#140132)

2025-05-27 16:07:56 +01:00

vplan-vectorize-inner-loop-reduction.ll

…

vplan-widen-call-instruction.ll

[VPlan] Bail out on non-intrinsic calls in VPlanNativePath.

2025-03-19 21:35:15 +00:00

vplan-widen-select-instruction.ll

[VPlan] Introduce explicit broadcasts for live-ins. (#124644)

2025-02-26 13:57:51 +00:00

vplan-widen-struct-return.ll

[VPlan] Connect Entry to scalar preheader during initial construction. (#140132)

2025-05-27 16:07:56 +01:00

widen-gep-all-indices-invariant.ll

[VPlan] VPWidenGEPRecipe uses first lane of invariant indices (NFC)

2025-05-24 17:32:05 +01:00

widen-intrinsic.ll

[VectorUtils][VPlan] Consolidate VPWidenIntrinsicRecipe::onlyFirstLaneUsed and isVectorIntrinsicWithScalarOpAtArg (#137497)

2025-05-01 01:25:41 +08:00

write-only.ll

…