shylie/llvm-project
1
0
Fork 0
Code Issues Pull Requests Actions 6 Packages Projects Releases Wiki Activity
llvm-project/llvm/test/Transforms/SLPVectorizer/X86
History
Simon Pilgrim 6a75c36ea9 [SLP] Add support for commutative icmp/fcmp predicates 
For the cases where the icmp/fcmp predicate is commutative, use reorderInputsAccordingToOpcode to collect and commute the operands.

This requires a helper to recognise commutativity in both general Instruction and CmpInstr types - the CmpInst::isCommutative doesn't overload the Instruction::isCommutative method for reasons I'm not clear on (maybe because its based on predicate not opcode?!?).

Differential Revision: https://reviews.llvm.org/D59992

llvm-svn: 357266
2019-03-29 15:28:25 +00:00
..
addsub.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
aggregate.ll
…
align.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
alternate-calls.ll
[SLPVectorizer][X86] Begin adding alternate tests for call operators
2018-07-02 17:23:45 +00:00
alternate-cast.ll
[SLPVectorizer] Add initial alternate opcode support for cast instructions. (REAPPLIED-2)
2018-07-13 11:09:52 +00:00
alternate-fp.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
alternate-int.ll
[SLPVectorizer] Update file missed in rL356913
2019-03-25 16:14:21 +00:00
arith-add-saddo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-add-ssat.ll
[TTI] Add generic SADDSAT/SSUBSAT costs
2019-01-27 13:51:59 +00:00
arith-add-uaddo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-add-usat.ll
[TTI] Add generic UADDSAT/USUBSAT costs
2019-01-24 12:27:10 +00:00
arith-add.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
arith-fix.ll
[Vectorizer] Add vectorization support for fixed smul/umul intrinsics
2019-02-25 15:42:02 +00:00
arith-fp.ll
…
arith-mul-smulo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-mul-umulo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-mul.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
arith-sub-ssat.ll
[TTI] Add generic SADDSAT/SSUBSAT costs
2019-01-27 13:51:59 +00:00
arith-sub-ssubo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-sub-usat.ll
[TTI] Add generic UADDSAT/USUBSAT costs
2019-01-24 12:27:10 +00:00
arith-sub-usubo.ll
[SLPVectorizer][X86] Add add/sub/mul overflow tests
2019-02-20 12:04:54 +00:00
arith-sub.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
atomics.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
bad_types.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
barriercall.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
bitreverse.ll
…
blending-shuffle.ll
[SLP] Fix invalid triple in X86 tests
2019-03-05 17:56:35 +00:00
bswap.ll
…
call.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
cast.ll
…
cmp_commute.ll
[SLP] Add support for commutative icmp/fcmp predicates
2019-03-29 15:28:25 +00:00
cmp_sel.ll
[SLPVectorizer][X86] Regenerate some tests. NFCI
2018-04-04 13:53:51 +00:00
commutativity.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
compare-reduce.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
consecutive-access.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
continue_vectorizing.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_7zip.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
crash_binaryop.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_bullet3.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
crash_bullet.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
crash_cmpop.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
crash_dequeue.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_flop7.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_gep.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_lencod.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_mandeltext.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
crash_netbsd_decompress.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_scheduling.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
crash_sim4b1.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
crash_smallpt.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
crash_vectorizeTree.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
cross_block_slp.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
cse.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
ctlz.ll
…
ctpop.ll
…
cttz.ll
…
cycle_dup.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
debug_info.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
diamond.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
external_user_jumbled_load.ll
[SLP] Fix PR36481: vectorize reassociated instructions.
2018-04-03 17:14:47 +00:00
external_user.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
extract_in_tree_user.ll
…
extract-shuffle.ll
[SLP] Fix invalid triple in X86 tests
2019-03-05 17:56:35 +00:00
extract.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
extractcost.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
extractelement.ll
…
fabs-cost-softfp.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
fabs.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
fcopysign.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
flag.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
fma.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
fptosi.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
fptoui.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
fround.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
funclet.ll
…
gep_mismatch.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
gep.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
hadd.ll
[X86][AVX] Reduce v4f64/v4i64 shuffle costs (PR37882)
2018-06-21 11:37:13 +00:00
hoist.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
horizontal-list.ll
[SLP]PR39774: Update references of the replaced external instructions.
2018-11-30 15:14:20 +00:00
horizontal-minmax.ll
[CostModel][X86] Add explicit vector select costs
2019-01-20 13:55:01 +00:00
horizontal.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
hsub.ll
[X86][AVX] Reduce v4f64/v4i64 shuffle costs (PR37882)
2018-06-21 11:37:13 +00:00
implicitfloat.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
in-tree-user.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
insert-after-bundle.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
insert-element-build-vector.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
insertvalue.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
intrinsic.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
jumbled-load-multiuse.ll
[SLPVectorizer][X86] Regenerate some tests. NFCI
2018-04-04 13:53:51 +00:00
jumbled-load-shuffle-placement.ll
[SLPVectorizer][X86] Regenerate some tests. NFCI
2018-04-04 13:53:51 +00:00
jumbled-load-used-in-phi.ll
[SLP] Fix PR36481: vectorize reassociated instructions.
2018-04-03 17:14:47 +00:00
jumbled-load.ll
[SLP] Fix PR36481: vectorize reassociated instructions.
2018-04-03 17:14:47 +00:00
limit.ll
…
lit.local.cfg
…
load-merge.ll
…
long_chains.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
loopinvariant.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
metadata.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
minimum-sizes.ll
…
multi_block.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
multi_user.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
odd_store.ll
[SLPVectorizer][X86] Regenerate some tests. NFCI
2018-04-04 13:53:51 +00:00
operandorder.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
opt.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
ordering.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
partail.ll
…
phi3.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
phi_landingpad.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
phi_overalignedtype.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
phi.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
powof2div.ll
[SLPVectorizer] Recognise non uniform power of 2 constants
2018-06-26 16:20:16 +00:00
pr16571.ll
…
pr16628.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
pr16899.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
pr18060.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
pr19657.ll
…
pr23510.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
pr27163.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
pr31599.ll
…
pr35497.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
PR32086.ll
…
PR34635.ll
…
PR35628_1.ll
[SLP]PR39774: Update references of the replaced external instructions.
2018-11-30 15:14:20 +00:00
PR35628_2.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
PR35777.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
PR35865.ll
…
PR36280.ll
…
PR39774.ll
[SLPVectorizer] Merge reorderAltShuffleOperands into reorderInputsAccordingToOpcode
2019-03-25 20:05:27 +00:00
PR40310.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
propagate_ir_flags.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
reassociated-loads.ll
[SLP] Fix PR36481: vectorize reassociated instructions.
2018-04-03 17:14:47 +00:00
reduction2.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
reduction_loads.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
reduction_unrolled.ll
[CostModel][X86] Fix overcounting arithmetic cost in illegal types in getArithmeticReductionCost/getMinMaxReductionCost
2018-12-07 18:20:56 +00:00
reduction.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
remark_horcost.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
remark_listcost.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
remark_not_all_parts.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
remark_unsupported.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
reorder_phi.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
reorder_repeated_ops.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
resched.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
return.ll
…
reverse_extract_elements.ll
…
rgb_phi.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
saxpy.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
schedule_budget.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
schedule-bundle.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
scheduling.ll
…
sext.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
shift-ashr.ll
[X86] Split skylake-avx512 run lines in SLP vectorizer tests to cover -mprefer=vector-width=256 and -mprefer-vector-width=512.
2018-11-30 22:53:21 +00:00
shift-lshr.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
shift-shl.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
sign-extend.ll
…
simple-loop.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
simplebb.ll
…
sitofp.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
slp-throttle.ll
[SLPVectorizer] Add test related to SLP Throttling support, NFCI.
2019-03-22 14:50:53 +00:00
sqrt.ll
…
store-jumbled.ll
[SLP] Fix PR36481: vectorize reassociated instructions.
2018-04-03 17:14:47 +00:00
stores_vectorize.ll
…
tiny-tree.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
uitofp.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00
undef_vect.ll
…
unreachable.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
value-bug.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
vect_copyable_in_binops.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
vector_gep.ll
[SLP]Update test checks for the SPL vectorizer, NFC.
2019-01-11 20:21:14 +00:00
vector.ll
…
vectorize-reorder-reuse.ll
[SLPVectorizer] reorderInputsAccordingToOpcode - remove non-Instruction canonicalization
2019-03-25 15:53:55 +00:00
zext.ll
Temporarily Revert "[X86][SLP] Enable SLP vectorization for 128-bit horizontal X86 instructions (add, sub)"
2019-02-20 04:42:07 +00:00