caio/clang-p2996
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
a78861fc55d18046989ff4d624a037e9181da170
clang-p2996/llvm/test/Transforms/SLPVectorizer/X86
History
Matthias Braun 054c23d78f X86: Improve cost model of fp16 conversion (#113195) 
Improve cost-modeling for x86 __fp16 conversions so the SLPVectorizer
transforms the patterns:

- Override `X86TTIImpl::getStoreMinimumVF` to report a minimum VF of 4 (SSE
  register can hold 4xfloat converted/stored to 4xf16) this is necessary as
  fp16 stores are neither modeled as trunc-stores nor can we mark direct Xxfp16
  stores as legal as we generally expand fp16 operations).
- Add missing cost entries to `X86TTIImpl::getCastInstrCost`
  conversion from/to fp16. Note that conversion from f64 to f16 is not
  supported by an X86 instruction.
2024-10-25 16:22:24 -07:00
..
aggregate.ll
…
align.ll
…
alternate-calls-inseltpoison.ll
[SLP]Fix the cost of the adjusted extracts in per-register analysis.
2024-06-28 14:33:08 -07:00
alternate-calls.ll
[SLP]Fix the cost of the adjusted extracts in per-register analysis.
2024-06-28 14:33:08 -07:00
alternate-cast-inseltpoison.ll
…
alternate-cast.ll
…
alternate-cmp-swapped-pred.ll
…
alternate-fp-inseltpoison.ll
…
alternate-fp.ll
…
alternate-int-inseltpoison.ll
…
alternate-int.ll
…
arith-abs.ll
…
arith-add-load.ll
…
arith-add-saddo.ll
…
arith-add-ssat.ll
[CostModel][X86] Add add/sat sat intrinsic costs
2024-07-25 14:35:39 +01:00
arith-add-uaddo.ll
…
arith-add-usat.ll
[CostModel][X86] Add add/sat sat intrinsic costs
2024-07-25 14:35:39 +01:00
arith-add.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-and-const-load.ll
…
arith-div.ll
…
arith-fix.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-fp-call.ll
[CostModel][X86] Treat lrint/llrint as fptosi calls (#90883)
2024-05-03 18:06:50 +01:00
arith-fp-inseltpoison.ll
…
arith-fp.ll
…
arith-fshl-rot.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-fshl.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-fshr-rot.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-fshr.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-max-cost.ll
…
arith-mul-load.ll
…
arith-mul-smulo.ll
…
arith-mul-umulo.ll
…
arith-mul.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-scmp.ll
[SLP] Add vectorization support for [u|s]cmp (#106747)
2024-09-02 17:06:07 +08:00
arith-smax.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-smin.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-sub-ssat.ll
[CostModel][X86] Add add/sat sat intrinsic costs
2024-07-25 14:35:39 +01:00
arith-sub-ssubo.ll
…
arith-sub-usat.ll
[CostModel][X86] Add add/sat sat intrinsic costs
2024-07-25 14:35:39 +01:00
arith-sub-usubo.ll
…
arith-sub.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-ucmp.ll
[SLP] Add vectorization support for [u|s]cmp (#106747)
2024-09-02 17:06:07 +08:00
arith-umax.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
arith-umin.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
atomics.ll
…
bad_types.ll
Remove the x86_mmx IR type. (#98505)
2024-07-25 09:19:22 -04:00
bad-reduction.ll
…
barriercall.ll
…
bitreverse.ll
…
blending-shuffle-inseltpoison.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
blending-shuffle.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
bool-mask.ll
…
bottom-to-top-reorder.ll
…
broadcast_long.ll
…
broadcast.ll
…
bswap.ll
…
buildvector_splat_extractvalue.ll
…
buildvector-float-and-extract-lane1.ll
…
buildvector-postpone-for-dependency.ll
[SLP]Add subvector vectorization for non-load nodes
2024-09-25 10:23:41 -04:00
buildvector-same-lane-insert.ll
…
buildvector-shuffle.ll
…
buildvector-with-reuses.ll
…
c-ray.ll
…
call.ll
[SLPVectorizer] Support SLPVectorizer cases of tan across all backends (#95517)
2024-06-27 15:15:13 -04:00
cast-operand-extracted.ll
[SLP]Try to keep operand of external casts as scalars, if profitable
2024-10-01 13:35:42 -04:00
cast.ll
…
cmp_commute-inseltpoison.ll
[InstCombine] Remove some of the complexity-based canonicalization (#91185)
2024-08-21 12:02:54 +02:00
cmp_commute.ll
[InstCombine] Remove some of the complexity-based canonicalization (#91185)
2024-08-21 12:02:54 +02:00
cmp_sel.ll
…
cmp-after-intrinsic-call-minbitwidth.ll
[SLP]Improve minbitwidth analysis for abs/smin/smax/umin/umax intrinsics.
2024-04-08 08:32:35 -07:00
cmp-as-alternate-ops.ll
…
cmp-diff-sized.ll
[SLP]Better sorting of cmp instructions by comparing type sizes.
2024-08-06 11:03:36 -04:00
cmp-in-bv-node-type-cost.ll
[SLP]Fix a crash when requestin the cost for buildvector cmp nodes types.
2024-08-29 03:53:28 -07:00
combined-stores-chains.ll
…
commutativity.ll
…
compare-reduce.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
const-in-different-functions.ll
Reland "[Support] Assert that DomTree nodes share parent" (#102782)
2024-08-13 11:56:02 +02:00
const-reduced-vals-resized.ll
[SLP]Use stored signedness after minbitwidth analysis.
2024-07-10 03:58:00 -07:00
continue_vectorizing.ll
…
control-dependence.ll
…
conversion-fp16.ll
X86: Improve cost model of fp16 conversion (#113195)
2024-10-25 16:22:24 -07:00
crash_7zip.ll
…
crash_binaryop.ll
…
crash_bullet3.ll
…
crash_bullet.ll
…
crash_clear_undefs.ll
…
crash_cmpop.ll
…
crash_dequeue.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
crash_flop7.ll
…
crash_gep.ll
…
crash_lencod-inseltpoison.ll
…
crash_lencod.ll
…
crash_mandeltext.ll
…
crash_netbsd_decompress.ll
…
crash_reordering_undefs.ll
…
crash_scheduling-inseltpoison.ll
…
crash_scheduling.ll
…
crash_sim4b1.ll
…
crash_smallpt.ll
[SLP][NFC]Replace more unreachable terminators by rets, NFC
2024-10-14 07:50:07 -07:00
crash_undef_mask_elem.ll
…
crash_vectorizeTree.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
cross_block_slp.ll
…
cse_extractelement.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
cse.ll
…
ctlz.ll
[X86] Allow speculative BSR/BSF instructions on targets with CMOV (#102885)
2024-08-22 11:11:00 +01:00
ctpop.ll
…
cttz.ll
…
cycle_dup.ll
…
debug_info.ll
[RemoveDIs] Print IR with debug records by default (#91724)
2024-06-14 15:07:27 +01:00
debug-counter.ll
[SLP]Add debug counter support
2024-10-02 11:14:34 -07:00
debug-info-salvage.ll
[SLP]Do not salvage debug info from instructions, marked for deletion already.
2024-07-12 08:08:50 -07:00
delayed-gather-emission.ll
…
diamond_broadcast_extra_shuffle.ll
…
diamond.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
different-vec-widths.ll
…
dot-product.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
entries-different-vf.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
external_user_jumbled_load-inseltpoison.ll
…
external_user_jumbled_load.ll
…
external_user.ll
…
external-used-across-reductions.ll
[SLP]Remove ExtraArgs from reductions.
2024-07-29 13:23:56 -04:00
extract_in_tree_user.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
extract-scalar-from-undef.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
extract-shuffle-inseltpoison.ll
…
extract-shuffle.ll
…
extract-with-undefs-shuffle.ll
…
extract.ll
…
extractcost.ll
…
extractelement-multi-register-use.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
extractelement-multiple-uses.ll
…
extractelement-phi-in-landingpad.ll
[SLP]Fix PR102187: do not insert extractelement before landingpad instruction.
2024-08-06 12:33:13 -07:00
extractelement-single-use-many-nodes.ll
[SLP][NFC]Replace more unreachable terminators by rets, NFC
2024-10-14 07:50:07 -07:00
extractelement-vecop-vectorized.ll
[SLP]Fix PR88123: use vectorized operands consistently.
2024-04-09 08:42:57 -07:00
extractelement.ll
…
extractelements-vector-ops-shuffle.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
fabs-cost-softfp.ll
…
fabs.ll
…
fcopysign.ll
…
flag.ll
…
float-min-max.ll
…
fma.ll
…
fmaximum-fminimum.ll
…
fmaxnum.ll
…
fminnum.ll
…
fmuladd.ll
[SLP] Enable optimization of freeze instructions (#102217)
2024-08-07 15:01:37 -04:00
fptosi-inseltpoison.ll
…
fptosi.ll
…
fptoui.ll
…
fround.ll
…
full-non-schedulable-overlap.ll
[SLP]Fix PR104637: do not create new nodes for fully overlapped non-schedulable nodes
2024-08-16 13:49:44 -07:00
funclet.ll
…
gather-extractelements-different-bbs.ll
[SLP]Remove ExtraArgs from reductions.
2024-07-29 13:23:56 -04:00
gather-move-out-of-loop.ll
[SLP][NFC] Add option to bypass early profitability check. (#88594)
2024-04-15 09:09:39 -07:00
gather-node-same-as-vect-but-order.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
gather-node-same-reduced.ll
[SLP]Fix PR89438: check for all tree entries for the resized value.
2024-04-22 06:38:38 -07:00
gather-nodes-different-bb.ll
[SLP]Fix a crash for gather node with instructions from different bbs,
2024-03-21 08:03:06 -07:00
gather-with-cmp-user.ll
[SLP]Fix minbitwidth analysis for gather nodes with icmp users.
2024-08-26 11:40:44 -07:00
gathered-delayed-nodes-with-reused-user.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
gep_mismatch.ll
…
gep-nodes-with-non-gep-inst.ll
…
gep-nonconst-idx-transformed-to-const.ll
…
gep.ll
…
geps-non-pow-2.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
hadd-inseltpoison.ll
[SLP]Improve/fix extracts calculations for non-power-of-2 elements.
2024-05-24 09:33:36 -04:00
hadd.ll
[SLP]Improve/fix extracts calculations for non-power-of-2 elements.
2024-05-24 09:33:36 -04:00
hoist.ll
…
horizontal-list.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
horizontal-minmax.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
horizontal-smax.ll
…
horizontal-store-many-uses.ll
…
horizontal.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
hsub-inseltpoison.ll
…
hsub.ll
…
implicitfloat.ll
…
in-tree-user.ll
…
insert-after-bundle.ll
…
insertelements-with-reused-indices.ll
[SLP][NFC]Replace unreachable instructions by rets, NFC.
2024-10-14 07:00:56 -07:00
insertvalue.ll
…
inst_size_bug.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
intrinsic_with_scalar_param.ll
…
intrinsic.ll
…
jumbled-load-multiuse.ll
…
jumbled-load-shuffle-placement.ll
…
jumbled-load-used-in-phi.ll
…
jumbled-load.ll
…
landing_pad.ll
[SLP]Add subvector vectorization for non-load nodes
2024-09-25 10:23:41 -04:00
lit.local.cfg
…
load-bitcast-vec.ll
…
load-merge-inseltpoison.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
load-merge.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
load-partial-vector-shuffle.ll
…
long_chains.ll
…
long-full-reg-stores.ll
[SLP]Initial support for non-power-of-2 (but whole reg) vectorization for stores
2024-10-09 15:22:44 -04:00
lookahead.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
loopinvariant.ll
…
malformed_phis.ll
…
many_stores.ll
…
matched-shuffled-entries.ll
[SLP]Do not ignore undefs when trying to replace with "poisonous" shuffles
2024-10-24 07:47:23 -07:00
matching-gather-nodes-phi-users.ll
…
memory-runtime-checks.ll
…
metadata.ll
…
minbitwidth-cast.ll
…
minbitwidth-drop-wrapping-flags.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
minbitwidth-icmp-to-trunc.ll
[SLP]Improve minbitwidth analysis for operands of IToFP and ICmp instructions.
2024-04-03 14:18:45 -07:00
minbitwidth-node-but-not-operands.ll
[SLP]Do extra analysis int minbitwidth if some checks return false.
2024-03-20 05:48:55 -07:00
minbitwidth-transformed-operand.ll
[SLP]Improve minbitwidth analysis for trun'ed gather nodes.
2024-07-17 07:41:00 -07:00
minbw-multiused-from-gather.ll
[SLP]Look for vector user when estimating the cost
2024-09-27 04:14:28 -07:00
minbw-user-non-sizable.ll
[SLP]Fix PR107036: Check if the type of the user is sizable before requesting its size.
2024-09-03 05:29:59 -07:00
minimum-sizes.ll
…
mul64.ll
…
multi_block.ll
…
multi_user.ll
…
multi-extracts-bv-combined.ll
[SLP]Fix PR108421: Correctly deduce VF from the masks
2024-09-12 13:43:44 -07:00
multi-node-with-save-values.ll
…
multi-nodes-bv-vectorized.ll
[SLP]Fix PR106667: carefully look for operand nodes.
2024-08-30 10:19:27 -07:00
multi-nodes-to-shuffle.ll
…
multi-tracked-reduced-value.ll
[SLP]Fix PR107037: correctly track origonal/modified after vectorizations reduced values
2024-09-04 04:54:32 -07:00
no_alternate_divrem.ll
…
no-scheduled-instructions.ll
…
non-load-reduced-as-part-of-bv.ll
[SLP]Do not check for clustered loads only
2024-10-24 08:16:42 -07:00
non-power-of-2-num-elems.ll
[SLP]Check if number of elements forms a full register
2024-09-27 12:54:56 -07:00
non-power-of-2-order-detection.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
non-scheduled-inst-reused-as-last-inst.ll
[SLP][NFC] Add option to bypass early profitability check. (#88594)
2024-04-15 09:09:39 -07:00
odd_store.ll
[SLP]Correctly decide if the non-power-of-2 number of stores can be vectorized.
2024-08-29 12:40:31 -07:00
opaque-ptr-2.ll
…
opaque-ptr.ll
[SLP]Keep externally used GEPs as GEPs, if possible instead of extractelement.
2024-04-16 14:54:06 -04:00
operandorder.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
opt.ll
…
ordering-bug.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
ordering.ll
…
partail.ll
[SLP][NFC]Replace unreachable instructions by rets, NFC.
2024-10-14 07:00:56 -07:00
phi3.ll
[IR] Remove support for icmp and fcmp constant expressions (#93038)
2024-06-04 08:31:03 +02:00
phi_landingpad.ll
…
phi_overalignedtype.ll
…
phi-node-deletion.ll
[SLP]Fix PR98133: Inserting PHI after debug-records!
2024-07-09 05:44:45 -07:00
phi-node-reshuffled-part.ll
[SLP]Return proper value for phi vectorized node
2024-09-13 11:30:29 -07:00
phi-nodes-as-operand-reorder.ll
[SLP]Fix PR105120: fix the order of phi nodes vectorization.
2024-08-21 12:22:01 -07:00
phi-removed-on-operand-vectorization.ll
[SLP]Fix a crash on attempt to revectorize vectorized phi.
2024-07-09 14:11:08 -07:00
phi-reordered-reshuffled.ll
[SLP]Fix PR89614: phis can be reordered, if reuses are not empty.
2024-04-22 08:40:19 -07:00
phi.ll
[SLP]Add subvector vectorization for non-load nodes
2024-09-25 10:23:41 -04:00
phis-different-order.ll
…
poison-element-shuffle.ll
…
powi-regression.ll
…
powi.ll
…
powof2div.ll
…
powof2mul.ll
…
pr16571.ll
…
pr16628.ll
…
pr16899.ll
…
pr18060.ll
…
pr19657.ll
…
pr23510.ll
…
pr27163.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
pr35497.ll
…
pr40522.ll
…
pr42022-inseltpoison.ll
…
pr42022.ll
…
pr44067-inseltpoison.ll
…
pr44067.ll
…
pr46983.ll
[SLP]Attempt to vectorize long stores, if short one failed.
2024-04-26 06:53:44 -07:00
pr47623.ll
[ConstantFolding] Canonicalize constexpr GEPs to i8 (#89872)
2024-05-20 11:47:30 +02:00
pr47629-inseltpoison.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
pr47629.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
pr47642.ll
…
pr48879-sroa.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
pr49081.ll
…
pr49933.ll
…
pr52275.ll
…
pr54465.ll
…
pr59867.ll
…
pr63668.ll
[X86] AMD Zen 5 Initial enablement (#107964)
2024-09-13 17:45:33 +01:00
pr98978.ll
[BasicAA] Fix handling of indirect assumption based results (#100130)
2024-07-25 12:25:19 +02:00
PR31847.ll
…
PR32086.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
PR34635.ll
…
PR35628_1.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
PR35628_2.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
PR35777.ll
…
PR35865-inseltpoison.ll
…
PR35865.ll
…
PR36280.ll
…
PR39774.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
PR40310.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
propagate_ir_flags.ll
…
reassociated-loads.ll
…
reduce-constants.ll
…
reduce-with-folded-to-consts.ll
…
reduced-val-extracted-and-externally-used.ll
[SLP]Fix PR108620: Need to check, if the reduced value was transformed
2024-09-13 15:43:06 -07:00
reduced-val-vectorized-in-transform.ll
[SLP]Transform nodes before building externally used values
2024-10-02 06:01:25 -07:00
reduced-value-replace-extractelement.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
reduced-value-vectorized-later.ll
[SLP]Update TrackedToOrig mappings after reduction vectorization
2024-09-23 10:52:03 -07:00
reduction2.ll
…
reduction_unrolled.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
reduction-bool-logic-op-inside.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
reduction-extracted-value.ll
…
reduction-logical.ll
[SLP]Add subvector vectorization for non-load nodes
2024-09-25 10:23:41 -04:00
reduction-same-vals.ll
…
reduction-transpose.ll
…
reduction-value-in-tree.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
reduction.ll
…
redux-feed-buildvector.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
redux-feed-insertelement.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
remark_alternate.ll
…
remark_extract_broadcast.ll
[SLP]Do not ignore undefs when trying to replace with "poisonous" shuffles
2024-10-24 07:47:23 -07:00
remark_gather-load-redux-cost.ll
[SLP]Improve masked loads vectorization, attempting gathered loads
2024-10-08 16:43:10 -04:00
remark_horcost.ll
…
remark_listcost.ll
…
remark_not_all_parts.ll
…
remark_unsupported.ll
…
remark-masked-loads-consecutive-loads-same-ptr.ll
[SLP]Better cost estimation for masked gather or "clustered" loads.
2024-08-30 14:27:51 -04:00
remark-partial-loads-vectorize.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
reorder_diamond_match.ll
[SLP]Do not propagate nuw/nsw flags for alt nodes, affected by
2024-03-26 10:24:09 -07:00
reorder_phi.ll
…
reorder_repeated_ops.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
reorder_with_external_users.ll
Revert "[SLP]Initial non-power-of-2 support (but still whole register) for reductions"
2024-10-21 13:37:44 +01:00
reorder_with_reordered_users.ll
[SLP] Regenerate test checks to reduce NFC diff in #100904
2024-08-08 17:08:53 +01:00
reorder-phi-operand.ll
…
reorder-possible-strided-node.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
reorder-reused-masked-gather2.ll
[SLP]Better cost estimation for masked gather or "clustered" loads.
2024-08-30 14:27:51 -04:00
reorder-reused-masked-gather.ll
…
reorder-vf-to-resize.ll
[SLP]Keep the original order in the reductions.
2024-07-09 12:26:42 -04:00
replaced-external-in-reduction.ll
[SLP]Represent externally used values as original scalars, if profitable.
2024-08-12 10:15:02 -04:00
resched.ll
[SLP][NFC]Replace more unreachable terminators by rets, NFC
2024-10-14 07:50:07 -07:00
reuse-extracts-in-wider-vect.ll
…
reused-extractelements.ll
[SLP]Do not vectorize small (<=2) buildvector/buildvalue sequences with MaxVF==true.
2024-07-16 12:45:58 -04:00
reused-extracts.ll
…
reused-mask-with-poison-index.ll
[SLP]Fix a crash on accessing element with index -1 for reused mask with PoisonMaskElem
2024-10-03 08:24:05 -07:00
reused-reductions-with-minbitwidth.ll
…
reused-scalars-in-buildvector.ll
…
reused-undefs.ll
…
revec-fix-99411.ll
Move the test to the correct folder. A test specified for a target
2024-07-18 10:06:19 -07:00
revectorized_rdx_crash.ll
…
reverse_extract_elements.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
rgb_phi.ll
…
root-gather-reused-scalar.ll
[SLP]Ignore root gather node, when searching for reuses
2024-10-21 09:16:16 -07:00
saxpy.ll
…
scatter-vectorize-reorder-non-empty.ll
…
scatter-vectorize-reorder.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
scatter-vectorize-reused-pointer.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
schedule_budget_debug_info.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
schedule_budget.ll
…
schedule-bundle.ll
…
scheduling.ll
…
select-reduction-op.ll
[SLP]Fix PR98838: do no replace condition of select-based logical op by poison.
2024-07-15 07:27:54 -07:00
sext-inseltpoison.ll
[SLP]Fix PR87011: Missing sign extension of demoted type before zero extension
2024-04-01 06:07:18 -07:00
sext.ll
[SLP]Fix PR87011: Missing sign extension of demoted type before zero extension
2024-04-01 06:07:18 -07:00
shift-ashr.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
shift-lshr.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
shift-shl.ll
[SLP]Remove operands upon marking instruction for deletion.
2024-07-08 07:56:48 -07:00
shrink_after_reorder.ll
…
shuffle-multiple-nodes.ll
…
shuffled-gathers-diff-size.ll
…
sign-extend-inseltpoison.ll
…
sign-extend.ll
…
simple-loop.ll
…
simplebb.ll
…
sin-sqrt.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
sitofp-inseltpoison.ll
…
sitofp-minbitwidth-node.ll
[SLP]Consider unsigned nodes, feeding into sitofp, being converted using
2024-04-11 10:39:54 -07:00
sitofp.ll
…
slp-fma-loss.ll
Prefer use of 0.0 over -0.0 for fadd reductions w/nsz (in IR) (#106770)
2024-09-03 09:16:37 -07:00
slp-schedule-use-order.ll
…
slp-throttle.ll
[SLP][NFC]Replace unreachable instructions by rets, NFC.
2024-10-14 07:00:56 -07:00
small-graph-diff-block-instructions.ll
[SLP]Small buidlvector only graph should contains scalars from same block
2024-10-23 10:46:38 -07:00
splat-score-adjustment.ll
[SLP]Better analysis of the repeated instructions during operands reordering
2024-09-24 14:03:10 -07:00
split-load8_2_unord_geps.ll
…
split-load8_2-unord.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
sqrt.ll
…
stackrestore-dependence.ll
…
stacksave-dependence.ll
[SLP]Keep externally used GEPs as GEPs, if possible instead of extractelement.
2024-04-16 14:54:06 -04:00
store-abs-minbitwidth.ll
[SLP]Consider (f)sub, being operand of llvm.(f)abs/icmp eq/ne 0, commutative.
2024-04-11 14:56:39 -04:00
store-insertelement-minbitwidth.ll
…
store-jumbled.ll
…
stores_constant_float.ll
…
stores_mix_sizes.ll
[SLP] The order of store chains needs to consider the size of the values. (#101810)
2024-08-07 11:01:53 +08:00
stores_vectorize.ll
…
subvector-minbitwidth-unsigned-value.ll
[SLP]Do correct signedness analysis for clustered nodes
2024-10-02 12:56:49 -07:00
supernode.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
tiny-tree.ll
[SLP]Improve/fix subvectors in gather/buildvector nodes handling
2024-08-23 06:45:22 -07:00
trunc-node-reused.ll
[SLP]Fix PR108700: correctly identify id of the operand node
2024-09-16 09:44:47 -07:00
trunc-store-value-ty-not-power-of-2.ll
[SLP] Make sure MinVF is a power-of-2 by using PowerOf2Ceil.
2024-04-16 13:29:35 +01:00
uitofp.ll
…
undef_vect.ll
…
unreachable.ll
…
used-reduced-op.ll
…
value-bug-inseltpoison.ll
…
value-bug.ll
…
vec3-base.ll
[SLP] Initial vectorization of non-power-of-2 ops. (#77790)
2024-04-13 08:14:40 +01:00
vec3-calls.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
vec3-gather-some-loads.ll
[SLP] Initial vectorization of non-power-of-2 ops. (#77790)
2024-04-13 08:14:40 +01:00
vec3-reorder-reshuffle.ll
[SLP] Enable reordering for non-power-of-two vectors (#106638)
2024-09-05 07:52:27 -07:00
vec_list_bias_external_insert_shuffled.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
vec_list_bias-inseltpoison.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
vec_list_bias.ll
[SLP]Vectorize gathered loads
2024-09-21 15:41:06 -07:00
vect_copyable_in_binops.ll
[SLP] Initial vectorization of non-power-of-2 ops. (#77790)
2024-04-13 08:14:40 +01:00
vect-gather-same-nodes.ll
[SLP]Initial non-power-of-2 support (but still whole register) for reductions
2024-10-21 12:25:39 -07:00
vector_gep.ll
Reland "[X86] Remove knl/knm specific ISAs supports (#92883)" (#93136)
2024-05-24 13:46:34 +08:00
vector.ll
…
vectorize-cmps.ll
…
vectorize-pair-path.ll
…
vectorize-reorder-reuse.ll
…
vectorize-widest-phis.ll
…
zext-inseltpoison.ll
[SLP]Fix PR87011: Missing sign extension of demoted type before zero extension
2024-04-01 06:07:18 -07:00
zext.ll
[SLP]Fix PR87011: Missing sign extension of demoted type before zero extension
2024-04-01 06:07:18 -07:00