Simon Pilgrim d8cd8d56ea

[SLP] getSpillCost - fully populate IntrinsicCostAttributes to improve cost analysis. (#124129) ...

We were only constructing the IntrinsicCostAttributes with the arg type info, and not the args themselves, preventing more detailed cost analysis (constant / uniform args etc.)

Just pass the whole IntrinsicInst to the constructor and let it resolve everything it can.

Noticed while having yet another attempt at #63980

2025-01-23 16:57:13 +00:00

..

32-bit.ll

…

64-bit-vector.ll

…

accelerate-vector-functions-inseltpoison.ll

[Analysis] atan2: isTriviallyVectorizable; add to massv and accelerate veclibs (#113637)

2024-11-08 16:07:38 -08:00

accelerate-vector-functions.ll

[Analysis] atan2: isTriviallyVectorizable; add to massv and accelerate veclibs (#113637)

2024-11-08 16:07:38 -08:00

alternate-vectorization-split-node.ll

[SLP][NFC]Add a test with potential alternate node, marked for minbitwidth size

2025-01-22 06:48:34 -08:00

buildvector-reduce.ll

[SLP] Remove -slp-optimize-identity-hor-reduction-ops option (#106238)

2024-08-27 13:21:57 -07:00

buildvector-vectorize.ll

…

commute.ll

…

div.ll

[SLP][NFC] Update test for PR #118055 (#122696)

2025-01-22 03:28:34 -08:00

ext-trunc.ll

[SLP]Improve minbitwidth analysis.

2024-03-19 08:19:45 -07:00

external-non-inst-use.ll

[SLP]Represent externally used values as original scalars, if profitable.

2024-08-12 10:15:02 -04:00

external-use-icmp.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

extractelements-to-shuffle.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

extracts-from-scalarizable-vector.ll

…

fmulladd.ll

…

fpsat.ll

…

fshl.ll

…

gather-buildvector-with-minbitwidth-user.ll

[SLP]Support minbitwidth analisys for buildvector nodes.

2024-04-29 09:57:37 -04:00

gather-cost.ll

[ConstantFolding] Infer getelementptr nuw flag (#119214)

2024-12-09 16:44:05 +01:00

gather-load-128.ll

…

gather-load-min-required-vf-2.ll

…

gather-reduce.ll

[InstCombine] Move gep of phi fold into separate function

2024-12-05 15:20:56 +01:00

gather-root.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

gather-with-minbith-user.ll

[SLP]Fix PR91467: Look through scalar cast, when trying to cast to another type.

2024-05-09 04:19:43 -07:00

getelementptr2.ll

[InstCombine] Infer nusw + nneg -> nuw for getelementptr (#111144)

2024-12-05 14:36:40 +01:00

getelementptr.ll

[InstCombine] Infer nusw + nneg -> nuw for getelementptr (#111144)

2024-12-05 14:36:40 +01:00

horizontal.ll

[SLP]Do extra analysis int minbitwidth if some checks return false.

2024-03-20 05:48:55 -07:00

insertelement-inseltpoison.ll

…

insertelement.ll

…

InstructionsState-is-invalid-0.ll

[SLP] Replace MainOp and AltOp in TreeEntry with InstructionsState. (#122443)

2025-01-18 10:23:20 +08:00

invalid_type.ll

…

landing_pad.ll

[SLP]Do not vectorize code in EH and non-returning blocks

2024-10-31 13:50:02 -04:00

lit.local.cfg

…

load-store-q.ll

…

loadi8.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

loadorder.ll

[SLP] getSpillCost - fully populate IntrinsicCostAttributes to improve cost analysis. (#124129)

2025-01-23 16:57:13 +00:00

matmul.ll

…

memory-runtime-checks-in-loops.ll

…

memory-runtime-checks.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

minbitwidth-cast-node-length.ll

[SLP]Use number of scalars as a vector length for minbw cast

2024-09-26 13:06:19 -07:00

minimum-sizes.ll

[llvm] Remove br i1 undef from some regression tests [NFC] (#118419)

2024-12-03 20:54:36 +00:00

mismatched-intrinsics.ll

…

mul.ll

…

multiple_reduction.ll

[SLP]Improve/fix subvectors in gather/buildvector nodes handling

2024-08-23 06:45:22 -07:00

nontemporal.ll

…

phi-use-order-scalable.ll

…

PR38339.ll

…

reduce-add-i64.ll

[SLP]Improve minbitwidth analysis.

2024-03-19 08:19:45 -07:00

reduce-fadd.ll

[SLP] getSpillCost - fully populate IntrinsicCostAttributes to improve cost analysis. (#124129)

2025-01-23 16:57:13 +00:00

remarks.ll

…

reorder-fmuladd-crash.ll

[llvm] Remove br i1 undef from some regression tests [NFC] (#118419)

2024-12-03 20:54:36 +00:00

reused-scalar-repeated-in-node.ll

[SLP]Add cost estimation for gather node reshuffling

2024-12-24 15:35:29 -05:00

scalable-external-user.ll

[SLP] Protect against scalable vector users.

2024-04-05 11:30:14 +01:00

scalable-vector.ll

…

scalarization-overhead.ll

[SLP]Add cost estimation for gather node reshuffling

2024-12-24 15:35:29 -05:00

sdiv-pow2.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

shuffle-vectors-mask-size.ll

[SLP]Improve/fix subvectors in gather/buildvector nodes handling

2024-08-23 06:45:22 -07:00

slp-abs.ll

…

slp-and-reduction.ll

…

slp-fma-loss.ll

[SLP]Represent externally used values as original scalars, if profitable.

2024-08-12 10:15:02 -04:00

slp-frem.ll

[AArch64][LV][SLP] Vectorizers use call cost for vectorized frem (#82488)

2024-03-14 17:20:29 +00:00

slp-or-reduction.ll

…

slp-xor-reduction.ll

…

spillcost-di.ll

[SLP]Remove operands upon marking instruction for deletion.

2024-07-08 07:56:48 -07:00

spillcost-order.ll

…

splat-loads.ll

…

store-ptr.ll

…

transpose-inseltpoison.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

transpose.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

trunc-insertion.ll

[llvm] Remove br i1 undef from some regression tests [NFC] (#118419)

2024-12-03 20:54:36 +00:00

tsc-s116.ll

[SLP]Initial non-power-of-2 support (but still whole register) for reductions

2024-10-21 12:25:39 -07:00

tsc-s352.ll

…

unsigned-after-sext-node.ll

[SLP]Fix PR91309: Do not consider SExt as always producing signed result.

2024-05-07 08:57:52 -07:00

uselistorder.ll

Fix potential crash in SLPVectorizer caused by missing check (#95937)

2024-07-02 08:15:51 -04:00

user-node-not-in-bitwidths.ll

[SLP]Fix PR91467: Look through scalar cast, when trying to cast to another type.

2024-05-09 04:19:43 -07:00

vec3-base.ll

Revert "[SLP] Make getSameOpcode support different instructions if they have same semantics. (#112181)"

2024-12-12 20:38:31 -08:00

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-reorder-reshuffle.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

vec15-base.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

vector-getelementptr.ll

…

vectorizable-selects-min-max.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00

vectorizable-selects-uniform-cmps.ll

[SLP]Do not include subvectors for fully matched buildvectors

2025-01-13 07:24:16 -08:00

vectorize-free-extracts-inserts.ll

[SLP][NFC]Replace undef with constant zero in tests, NFC

2025-01-17 09:48:03 -08:00

widen.ll

[LLVM][IR] Use splat syntax when printing Constant[Data]Vector. (#112548)

2024-11-06 11:53:33 +00:00