Alexey Bataev e1ea86e849 [SLP]Do not try to use interleaved loads, if reordering is required ...

If the interleaved loads require reordering, better to avoid generate
load + shuffle sequence, which in this case cannot be recognized as
interleaved load. Also, it fixes the issue with the incorrect codegen.

Fixes #138923

2025-05-12 14:12:51 -07:00

..

check-node-without-vector-user.ll

…

combined-loads-stored.ll

…

complex-loads.ll

[SLP]Improve reordering of the alternate nodes

2025-04-24 14:33:10 -04:00

ctpop.ll

…

floating-point.ll

…

fminimumnum.ll

[RISCV] Support vectorizing FMINIMUMNUM and FMAXIMUMNUM (#135727)

2025-04-27 19:10:02 +08:00

fround.ll

…

gather-node-with-no-users.ll

…

gep.ll

…

getpointerschaincost.ll

…

horizontal-list.ll

[SLP]Improved reduction cost/codegen

2025-02-14 11:03:33 -08:00

init-ext-node-not-truncable.ll

…

interleave-greater-than-slice.ll

…

lit.local.cfg

…

load-binop-store.ll

…

load-store.ll

…

loads-ordering.ll

…

long-gep-chains.ll

…

long-mask-split.ll

…

math-function.ll

[SLP]Fix/improve getSpillCost analysis

2025-03-04 15:47:23 -05:00

minbw-with-and-and-scalar-trunc.ll

…

mixed-extracts-types.ll

…

partial-vec-invalid-cost.ll

…

phi-const.ll

…

reduced-value-repeated-and-vectorized.ll

…

reduction-extension-after-bitwidth.ll

…

reductions.ll

[RISCV][TTI] Use processShuffleMask for shuffle legalization estimate (#136191)

2025-04-22 10:50:20 -07:00

remark-zext-incoming-for-neg-icmp.ll

…

remarks_cmp_sel_min_max.ll

[SLP]Represent SLP graph as a tree

2025-02-21 07:15:02 -05:00

remarks-insert-into-small-vector.ll

…

reordered-buildvector-scalars.ll

[SLP]Fix a crash when checking a scalar in a reordered buildvector node

2025-02-21 14:59:43 -08:00

reordered-interleaved-loads.ll

[SLP]Do not try to use interleaved loads, if reordering is required

2025-05-12 14:12:51 -07:00

repeated-address-store.ll

…

revec-getGatherCost.ll

…

revec.ll

[SLP] Sort PHIs by ExtractElements when relevant (#131229)

2025-03-17 14:19:46 -07:00

reversed-strided-node-with-external-ptr.ll

[SLP] Make getSameOpcode support interchangeable instructions. (#135797)

2025-04-16 00:08:59 +08:00

rvv-min-vector-size.ll

…

scalable-type-to-vect.ll

…

scatter-vectorize-reversed.ll

…

segmented-loads-simple.ll

[SLP]Prefer segmented/deinterleaved loads to strided and fix codegen

2025-04-22 12:11:01 -04:00

segmented-loads.ll

…

segmented-stores.ll

…

select-profitability.ll

…

shuffled-gather-casted.ll

…

small-phi-tree.ll

[SLP]Represent SLP graph as a tree

2025-02-21 07:15:02 -05:00

small-tree-not-schedulable-bv-node.ll

…

smax-reduction-unsigned-missing-sign.ll

[SLP]Fix type promotion for smax reduction with unsigned reduced operands

2025-04-16 10:14:29 -07:00

smax-unsigned-operand.ll

…

smin-reduction-unsigned-missing-sign.ll

[SLP]Mark smin reduction as signed compare

2025-04-15 07:24:17 -07:00

smin-signed-zextended.ll

…

spillcost.ll

[SLP]Fix/improve getSpillCost analysis

2025-03-04 15:47:23 -05:00

strided-loads-vectorized.ll

…

strided-loads-with-external-indices.ll

…

strided-loads-with-external-use-ptr.ll

…

strided-loads.ll

…

strided-stores-vectorized.ll

…

strided-unsupported-type.ll

…

struct-gep.ll

…

trunc-bv-multi-uses.ll

…

trunc-to-large-than-bw.ll

…

unordered-loads-operands.ll

[SLP][NFC]Add a test with ordering of the operands of unordered loads

2025-05-11 08:09:51 -07:00

unsigned-icmp-signed-op.ll

…

unsigned-node-trunc-with-signed-users.ll

…

vec3-base.ll

[SLP] Make getSameOpcode support interchangeable instructions. (#135797)

2025-04-16 00:08:59 +08:00

vec15-base.ll

…