shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/llvm/test/Transforms/LoopVectorize
History
Roman Lebedev 8e37b53360
[X86] Rewrite getScalarizationOverhead() 
All of our insert/extract ops work on 128-bit lanes.

For `Insert`, we need to extract affected 128-bit lane,
unless it's being fully overwritten (FIXME: do we need to be
careful about legalization-induced padding that we obviously don't demand?),
perform insertions, and then insert the 128-bit lane back.

But hold on. If we are operating on an 256-bit legal vector,
and thus have two 128-bit subvectors, and are fully overwriting them both,
we don't actually need to insert *both* subvectors,
only the second one, into the implicitly-widened first one.

Also, `Insert` wasn't actually querying the costs,
but just assuming them to be `1`.

`getShuffleCost(TTI::SK_ExtractSubvector)` notes:
```
  // Note that in general, the insertion starting at the beginning of a vector
  // isn't free, because we need to preserve the rest of the wide vector.
```
... so as far as i can tell, we didn't account for that.

I was hoping this would allow vectorization at a higher VF at one case i looked at,
but the subvector insertion cost is still dis-advising that.

The change for `Extract` is NFC, and is for consistency only,
i wanted to get rid of of that weird explicit discounting of insertion of 0'th element,
since the general code should already deal with that.

Reviewed By: RKSimon

Differential Revision: https://reviews.llvm.org/D137913
2022-11-15 21:07:12 +03:00
..
AArch64
[VectorUtils] Skip interleave members with diff type and alloca sizes.
2022-11-13 22:06:20 +00:00
AMDGPU
fix tests for commit 9df0b254d24eca098
2022-07-23 22:32:30 +01:00
ARM
[ARM] Add a test demonstrating reductions with reused extend. NFC
2022-10-24 19:38:19 +01:00
Hexagon
[LV] Fix crash when computing max VF too early
2021-02-01 12:14:59 +00:00
PowerPC
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
RISCV
[RISCV] Prevent autovectorization using vscale with Zvl32b.
2022-11-02 13:55:21 -07:00
SystemZ
fix tests for commit 9df0b254d24eca098
2022-07-23 22:32:30 +01:00
VE
…
X86
[X86] Rewrite getScalarizationOverhead()
2022-11-15 21:07:12 +03:00
XCore
…
12-12-11-if-conv.ll
…
2012-10-20-infloop.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
2012-10-22-isconsec.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
2016-07-27-loop-vec.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
alias-set-with-uncomputable-bounds.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
align.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
assume.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
bsd_regex.ll
[NFC] Switch a few uses of undef to poison as placeholders for unreachable code
2022-07-23 21:50:11 +01:00
bzip_reverse_loops.ll
[Analysis] remove bogus smin/smax pattern detection
2022-03-09 17:50:34 -05:00
calloc.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
cast-induction.ll
[LV] Add check line to test interleaving only with induction cast.
2022-05-22 20:11:47 +01:00
check-prof-info.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
conditional-assignment.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
consec_no_gep.ll
…
consecutive-ptr-uniforms.ll
[NFC][1/n] Remove -enable-new-pm=0 flags from lit tests
2022-09-19 09:57:37 -07:00
control-flow.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
cpp-new-array.ll
…
create-induction-resume.ll
[LoopVectorize] Uninitialized phi node leads to a crash in SSAUpdater.
2022-06-22 10:49:27 +07:00
dbg.value.ll
…
dead_instructions.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
debugloc.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
demanded-bits-of-pointer-instruction.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
diag-missing-instr-debug-loc.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
diag-with-hotness-info-2.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
diag-with-hotness-info.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
disable_nonforced_enable.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
disable_nonforced.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
discriminator.ll
[VPlan] Add VPCanonicalIVPHIRecipe, partly retire createInductionVariable.
2022-01-05 10:46:06 +00:00
dont-fold-tail-for-const-TC.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
dont-fold-tail-for-divisible-TC.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
ee-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
epilog-vectorization-reductions.ll
[LV] Do not LoopSimplify/LCSSA after generating main vector loop.
2022-05-20 09:58:40 +01:00
exact.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
explicit_outer_detection.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
explicit_outer_nonuniform_inner.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
explicit_outer_uniform_diverg_branch.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
extract-last-veclane.ll
[VPlan] Model first exit values using VPLiveOut.
2022-05-21 16:01:38 +01:00
fcmp-vectorize.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
first-order-recurrence-chains-vplan.ll
[LV] Support fixed order recurrences.
2022-08-18 19:15:52 +01:00
first-order-recurrence-chains.ll
[IVDescriptors] Before moving an instruction in SinkAfter checking if it is target of other instructions
2022-10-03 18:47:51 +00:00
first-order-recurrence-complex.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
first-order-recurrence-multiply-recurrences.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
first-order-recurrence-sink-replicate-region.ll
[LV] Use safe-divisor lowering for fixed vectors if profitable
2022-09-08 09:15:54 -07:00
first-order-recurrence.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
fix-reduction-dbg.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
flags.ll
[Vectorizers][TTI] remove option to bypass creation of vector reduction intrinsics
2021-02-12 08:13:50 -05:00
float-induction.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
float-minmax-instruction-flag.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
float-reduction.ll
Enable support for floating-point division reductions
2020-11-23 20:00:58 -05:00
fneg.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
followup.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
forked-pointers.ll
[LAA] Add recursive IR walker for forked pointers
2022-07-18 12:06:17 +01:00
fpsat.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
funcall.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
gcc-examples.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
gep_with_bitcast.ll
…
global_alias.ll
[opt] Directly translate -O# to -passes='default<O#>'
2021-10-18 16:48:10 -07:00
hints-trans.ll
[SimplifyCFG] TryToSimplifyUncondBranchFromEmptyBlock() already knows how to preserve DomTree
2020-12-17 01:03:49 +03:00
hoist-loads.ll
…
i8-induction.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
icmp-uniforms.ll
[VPlan] Print predicates for widened cmp instructions (NFC).
2022-10-21 08:54:11 +01:00
if-conv-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
if-conversion-edgemasks.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
if-conversion-nest.ll
[InstCombine] Fold ult(add(x,-1),c) -> ule(x,c) iff x != 0 (PR57635)
2022-09-20 16:44:41 +01:00
if-conversion-reduction.ll
…
if-conversion.ll
[ConstExpr] Remove div/rem constant expressions
2022-07-06 10:11:34 +02:00
if-pred-non-void.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
if-pred-not-when-safe.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
if-pred-stores.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
if-reduction.ll
[LoopVectorize] Fix VPRecipeBuilder::createEdgeMask to correctly generate the mask
2021-02-14 21:12:34 +09:00
incorrect-dom-info.ll
Revert "[NFC] remove explicit default value for strboolattr attribute in tests"
2021-05-24 19:43:40 +02:00
increment.ll
…
induction_plus.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
induction-multiple-uses-in-same-instruction.ll
[VPlan] Replace remaining use of needsScalarIV.
2022-06-09 12:05:37 +01:00
induction-ptrcasts.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
induction-step.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
induction-unroll-novec.ll
[LV] Add test case for PR54427.
2022-04-07 23:21:21 +02:00
induction.ll
[InstCombine] Fold ult(add(x,-1),c) -> ule(x,c) iff x != 0 (PR57635)
2022-09-20 16:44:41 +01:00
infiniteloop.ll
…
instruction-only-used-outside-of-loop.ll
[LV] Fix test cases where vector loop never executed.
2022-08-31 13:24:49 +01:00
int_sideeffect.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
interleave-and-scalarize-only.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
interleaved-accesses-1.ll
[IR] Let IRBuilder's CreateVectorSplat/CreateShuffleVector use poison as placeholder
2020-12-30 04:21:04 +09:00
interleaved-accesses-2.ll
…
interleaved-accesses-3.ll
…
interleaved-accesses-alias.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
interleaved-accesses-masked-group.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
interleaved-accesses-pred-stores.ll
[VPlan] Use onlyFirstLaneUsed in sinkScalarOperands.
2022-10-29 19:45:19 +01:00
interleaved-accesses-uniform-load.ll
…
interleaved-accesses.ll
[test] Remove readonly from some parameters that are written through in tests.
2022-10-29 15:05:20 -07:00
interleaved-acess-with-remarks.ll
…
intrinsic.ll
Update @llvm.powi to handle different int sizes for the exponent
2021-06-17 09:38:28 +02:00
invariant-store-vectorization-2.ll
[opt] Stop treating alias analysis specially when translating legacy opt syntax
2022-10-07 11:50:43 -07:00
invariant-store-vectorization.ll
[opt] Stop treating alias analysis specially when translating legacy opt syntax
2022-10-07 11:50:43 -07:00
irregular_type.ll
[LoopVectorize] Refine hasIrregularType predicate
2021-03-17 17:03:47 +01:00
iv_outside_user.ll
[VPlan] Model first exit values using VPLiveOut.
2022-05-21 16:01:38 +01:00
lcssa-crashes.ll
[LCSSA] Update unreachable uses with poison.
2022-09-04 22:26:18 +01:00
libcall-remark.ll
…
lifetime.ll
[NFC] Remove redundant test in Transforms/LoopVectorize/lifetime.ll
2021-08-05 14:39:08 +01:00
loop-form.ll
[VPlan] Remove dead recipes across whole plan.
2022-06-23 13:36:02 +02:00
loop-legality-checks.ll
[NFC][LV] Use different braces in debug output
2022-03-07 19:32:37 +03:00
loop-scalars.ll
[NFC][1/n] Remove -enable-new-pm=0 flags from lit tests
2022-09-19 09:57:37 -07:00
loop-vect-memdep.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
loop-vect-option.ll
…
memdep-fold-tail.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
memdep.ll
…
memory-dep-remarks.ll
[LAA] Change to function analysis for new PM.
2022-10-01 15:44:27 +01:00
metadata-unroll.ll
…
metadata-width.ll
[LV] Enable scalable vectorization by default for SVE cores.
2021-12-20 16:23:29 +00:00
metadata.ll
…
middle-block-dbg.ll
…
miniters.ll
…
minmax_reduction.ll
[LV] Recognize intrinsic min/max reductions
2021-09-15 10:45:50 +01:00
mixed-precision-remarks.ll
[LV] Add analysis remark for mixed precision conversions
2021-02-17 21:37:08 -05:00
multi-use-reduction-bug.ll
…
multiple-address-spaces.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
multiple-exits-versioning.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
multiple-strides-vectorization.ll
[LoopUtils] Simplify addRuntimeCheck to return a single value.
2021-10-18 18:03:09 +01:00
no_array_bounds_scalable.ll
[SVE] Add support for scalable vectors with vectorize.scalable.enable loop attribute
2020-12-02 13:23:43 +00:00
no_array_bounds.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
no_idiv_reduction.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
no_int_induction.ll
…
no_outside_user.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
no_switch_disable_vectorization.ll
…
no_switch.ll
…
no-interleave-up-front.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
noalias-md-licm.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
noalias-md.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
noalias-scope-decl.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
nofloat-report.ll
…
nofloat.ll
…
non-const-n.ll
…
nontemporal.ll
…
nounroll.ll
[NFC][LV] Use different braces in debug output
2022-03-07 19:32:37 +03:00
novect-lcssa-cfg-invalidation.ll
[LAA] Change to function analysis for new PM.
2022-10-01 15:44:27 +01:00
nsw-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
nuw.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
opaque-ptr.ll
[LV] Keep track of cost-based ScalarAfterVec in VPWidenPointerInd.
2022-09-19 18:14:35 +01:00
opt.ll
…
optimal-epilog-vectorization-limitations.ll
[NFC][LV] Use different braces in debug output
2022-03-07 19:32:37 +03:00
optimal-epilog-vectorization-liveout.ll
[VPlan] Model first exit values using VPLiveOut.
2022-05-21 16:01:38 +01:00
optimal-epilog-vectorization-scalable.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
optimal-epilog-vectorization.ll
[LV] Do not LoopSimplify/LCSSA after generating main vector loop.
2022-05-20 09:58:40 +01:00
optsize.ll
[NFC][1/n] Remove -enable-new-pm=0 flags from lit tests
2022-09-19 09:57:37 -07:00
outer_loop_test1.ll
[LV] Mark increment of main vector loop induction variable as NUW.
2021-06-07 10:47:52 +01:00
outer_loop_test2.ll
[IRBuilder] Use PoisonValue in CreateMasked*
2022-09-19 11:01:41 -07:00
outer-loop-vec-phi-predecessor-order.ll
[IRBuilder] Use PoisonValue in CreateMasked*
2022-09-19 11:01:41 -07:00
partial-lcssa.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
phi-cost.ll
[NFC][1/n] Remove -enable-new-pm=0 flags from lit tests
2022-09-19 09:57:37 -07:00
phi-hang.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pointer-induction.ll
[VPlan] Remove dead recipes before sinking.
2022-10-12 12:49:42 +01:00
pointer-select-runtime-checks.ll
[VPlan] Only generate single instr for loads uniform across all parts.
2022-09-08 14:27:58 +01:00
pr25281.ll
[AST] Remove legacy AliasSetPrinter pass
2022-11-14 15:50:38 +01:00
pr28541.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr30654-phiscev-sext-trunc.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr30806-phi-scev.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
pr30806.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr31098.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr31190.ll
…
pr32859.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr33706.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr34681.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr35743.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr35773.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr36311.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr36983.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr37248.ll
…
pr37515.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
pr38697.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr38800.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr39099.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
pr39417-optsize-scevchecks.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr43166-fold-tail-by-masking.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr44488-predication.ll
[VPlan] Only generate single instr for unpredicated stores of varying value to invariant address
2022-09-22 08:53:46 -07:00
pr45259.ll
Recommit "[SCEV] Look through single value PHIs." (take 3)
2022-06-09 15:20:10 +01:00
pr45525.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr45679-fold-tail-by-masking.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
pr46525-expander-insertpoint.ll
[LV] Restructure isPredicatedInst and isScalarWithPredication (w/a fix for uniform mem ops)
2022-08-18 07:14:04 -07:00
pr47343-expander-lcssa-after-cfg-update.ll
[VPlan] Only generate single instr for unpredicated stores of varying value to invariant address
2022-09-22 08:53:46 -07:00
pr48832.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr50686.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr51614-fold-tail-by-masking.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
pr55100-expand-scev-predicate-used.ll
[VPlan] Remove dead recipes across whole plan.
2022-06-23 13:36:02 +02:00
pr55167-fold-tail-live-out.ll
[VPlan] Model first exit values using VPLiveOut.
2022-05-21 16:01:38 +01:00
preserve-dbg-loc-and-loop-metadata.ll
…
ptr_loops.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
ptr-induction.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
read-only.ll
…
reduction-inloop-cond.ll
[LV][IVDescriptors] Fix recurrence identity element for FMin and FMax reductions
2022-11-04 10:39:37 -04:00
reduction-inloop-pred.ll
[NFC] Switch a few uses of undef to poison as placeholders for unreachable code
2022-07-23 21:50:11 +01:00
reduction-inloop-uf4.ll
[NFC] Switch a few uses of undef to poison as placeholders for unreachable code
2022-07-23 21:50:11 +01:00
reduction-inloop.ll
[NFC] Switch a few uses of undef to poison as placeholders for unreachable code
2022-07-23 21:50:11 +01:00
reduction-odd-interleave-counts.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
reduction-order.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
reduction-predselect.ll
[InstCombine] reassociate bitwise logic chains based on uses
2022-08-21 09:42:14 -04:00
reduction-ptr.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
reduction-small-size.ll
[LV] Use safe-divisor lowering for fixed vectors if profitable
2022-09-08 09:15:54 -07:00
reduction-with-invariant-store.ll
[LoopVectorize] Add missing test for D133687
2022-10-03 14:54:17 +01:00
reduction.ll
[VPlan] Register recipe for instr if the simplified value is recipe.
2021-05-11 14:32:34 +01:00
remarks-multi-exit-loops.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
remove_metadata.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
reverse_induction.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
reverse_iter.ll
…
runtime-check-address-space.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
runtime-check-needed-but-empty.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
runtime-check-pointer-element-type.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
runtime-check-readonly-address-space.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
runtime-check-readonly.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
runtime-check-small-clamped-bounds.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
runtime-check.ll
[opt] Stop treating alias analysis specially when translating legacy opt syntax
2022-10-07 11:50:43 -07:00
runtime-checks-difference.ll
[LAA] Require AddRecs to be in the innermost loop for diff-checks.
2022-08-26 20:39:52 +01:00
runtime-drop-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
safegep.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
same-base-access.ll
…
scalable-assume.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scalable-first-order-recurrence.ll
[IR] Let ConstantVector::getSplat use poison instead of undef
2021-08-10 08:27:43 +01:00
scalable-inductions.ll
[LV] Create & use VPScalarIVSteps for all scalar users.
2022-03-13 17:15:24 +00:00
scalable-lifetime.ll
[LoopVectorize] Improve vectorisation of some intrinsics by treating them as uniform
2021-08-06 10:13:15 +01:00
scalable-loop-unpredicated-body-scalar-tail.ll
[LAA,LV] Add initial support for pointer-diff memory checks.
2022-05-16 15:27:22 +01:00
scalable-noalias-scope-decl.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scalable-reduction-inloop.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scalable-trunc-min-bitwidth.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scalable-vf-hint.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scalar_after_vectorization.ll
Revert "[NFC] IRBuilderBase::CreateAdd(): place constant onto RHS"
2021-10-27 22:21:37 +03:00
scalar-select.ll
…
scalarize-masked-call.ll
[NFC][LV] Scalarizing test for masked vector calls
2022-09-21 15:43:25 +01:00
scalarized-bitcast.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
scev-during-mutation.ll
[LV] Add basic check lines to test added in 00200dbda31674e59.
2021-11-18 14:08:57 +00:00
scev-exitlim-crash.ll
[IR] Remove support for float binop constant expressions
2022-07-12 09:40:49 +02:00
select-cmp-predicated.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
select-cmp.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
select-min-index.ll
[LV] Add tests for vectorizing select of minimum idx idiom.
2022-08-14 17:44:11 +01:00
select-reduction.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
simple-unroll.ll
…
single-value-blend-phis.ll
[VPlan] Replace remaining use of needsScalarIV.
2022-06-09 12:05:37 +01:00
skeleton-lcssa-crash.ll
[SCEVExpander] Move LCSSA fixup to ::expand.
2022-09-29 20:49:56 +01:00
skip-iterations.ll
…
small-loop.ll
…
start-non-zero.ll
…
store-shuffle-bug.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
struct_access.ll
…
tail-folding-counting-down.ll
…
tail-folding-vectorization-factor-1.ll
[LV] Use buildScalarSteps to also handle VF = 1. (NFCI)
2022-10-20 14:30:01 +01:00
tbaa-nodep.ll
[opt] Stop treating alias analysis specially when translating legacy opt syntax
2022-10-07 11:50:43 -07:00
tripcount.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
trunc-reductions.ll
[NFC] Switch a few uses of undef to poison as placeholders for unreachable code
2022-07-23 21:50:11 +01:00
undef-inst-bug.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
uniform-blend.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unroll_nonlatch.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unroll_novec.ll
[LV] Mark increment of main vector loop induction variable as NUW.
2021-06-07 10:47:52 +01:00
unroll-novec-memcheck-metadata.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unroll.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unsafe-dep-remark.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unsafe-vf-hint-remark.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
unsized-pointee-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
use-scalar-epilogue-if-tp-fails.ll
[LV] Refresh autogen tests to reflect naming changes [nfc]
2022-08-29 14:16:54 -07:00
value-ptr-bug.ll
…
vect-phiscev-sext-trunc.ll
[VPlan] Replace remaining use of needsScalarIV.
2022-06-09 12:05:37 +01:00
vect.omp.persistence.ll
[NFC][LV] Use different braces in debug output
2022-03-07 19:32:37 +03:00
vect.stats.ll
…
vector-freeze.ll
[LV] Widen freeze instead of scalarizing it
2022-05-19 12:28:01 +08:00
vector-geps.ll
[instcombine] Canonicalize constant index type to i64 for extractelement/insertelement
2021-12-13 16:56:22 -08:00
vector-intrinsic-call-cost.ll
[LV] Make some tests more robust by adding missing users.
2022-01-24 13:04:09 +00:00
vector-loop-backedge-elimination.ll
[LV] Add test case where SCEV is needed to remove vector backedge.
2022-08-31 14:01:42 +01:00
vectorize-once.ll
[LV] Use variables instead of hard-coded metadata IDs in tests.
2022-08-16 12:21:49 +01:00
vectorize-pointer-phis.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00
vectorizeVFone.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
version-mem-access.ll
[opt] Don't translate legacy -analysis flag to require<analysis>
2022-10-07 14:54:34 -07:00
vplan_hcfg_stress_test.ll
…
vplan-dot-printing.ll
[VPlan] Add field to track if intrinsic should be used for call. (NFC)
2022-09-01 13:14:40 +01:00
vplan-outer-loop-uncomputable-trip-count.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
vplan-printing-outer-loop.ll
[VPlan] Replace CondBit with BranchOnCond VPInstruction.
2022-06-03 11:48:31 +01:00
vplan-printing.ll
[VPlan] Print predicates for widened cmp instructions (NFC).
2022-10-21 08:54:11 +01:00
vplan-sink-scalars-and-merge-vf1.ll
[VPlan] Replace CondBit with BranchOnCond VPInstruction.
2022-06-03 11:48:31 +01:00
vplan-sink-scalars-and-merge.ll
[VPlan] Print predicates for widened cmp instructions (NFC).
2022-10-21 08:54:11 +01:00
vplan-stress-test-no-explict-vf.ll
…
vplan-vectorize-inner-loop-reduction.ll
[IRBuilder] Use PoisonValue in CreateMasked*
2022-09-19 11:01:41 -07:00
vplan-widen-call-instruction.ll
[IRBuilder] Use PoisonValue in CreateMasked*
2022-09-19 11:01:41 -07:00
vplan-widen-select-instruction.ll
Revert "[NFCI] Regenerate SROA/LoopVectorize test checks"
2022-04-04 01:15:30 +02:00
write-only.ll
…
zero-sized-pointee-crash.ll
[test][NewPM] Use -passes=loop-vectorize instead of -loop-vectorize
2022-04-28 16:46:00 +02:00