Alexey Bataev 56140a8258 [SLP]Fix PR104422: Wrong value truncation ...

The minbitwidth restrictions can be skipped only for immediate reduced
values, for other nodes still need to check if external users allow
bitwidth reduction.

Fixes https://github.com/llvm/llvm-project/issues/104422

2024-08-15 08:00:08 -07:00

..

AArch64

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

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

AMDGPU

TTI: Check legalization cost of min/max ISD nodes (#100514)

2024-08-08 17:06:11 +04:00

ARM

…

NVPTX

…

PowerPC

…

RISCV

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

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

SystemZ

[SLP]Add index of the node to the short name output.

2024-08-08 08:57:14 -07:00

VE

…

WebAssembly

[InstCombine] Canonicalize constant GEPs to i8 source element type (#68882)

2024-01-24 15:25:29 +01:00

X86

[SLP]Fix PR104422: Wrong value truncation

2024-08-15 08:00:08 -07:00

XCore

…

alt-cmp-vectorize.ll

[SLP]Improve minbitwidth analysis.

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

alternate-non-profitable.ll

[SLP]Buildvector for alternate instructions with non-profitable gather operands.

2024-04-10 14:33:56 -04:00

const-bool-logical-or-reduction.ll

…

ext-after-phi-node.ll

[SLP]Fix PR90780: insert cast instruction for PHI nodes after all phi nodes.

2024-05-02 06:30:14 -07:00

insertelement-across-zero.ll

[SLP]Fix PR89187: fixx assertion check.

2024-04-18 04:22:25 -07:00

int_sideeffect.ll

…

is_fpclass.ll

…

orig-btiwidth-les-projected.ll

Revert "[SLP]Improve minbitwidth analysis for trun'ed gather nodes."

2024-07-17 05:51:54 -07:00

phi-not-ordered.ll

…

rdar128092379.ll

[llvm][SLPVectorizer] Fix a bad cast assertion (#97621)

2024-07-03 16:25:32 -07:00

reschedule.ll

…

revec-shufflevector.ll

[SLP][REVEC] Make SLP vectorize shufflevector. (#102489)

2024-08-13 11:19:29 +08:00

revec.ll

[SLP][REVEC] Make CastInst support vector instructions. (#103216)

2024-08-13 23:52:32 +08:00

slp-hr-with-reuse.ll

…

slp-max-phi-size.ll

…

slp-max-reg-size.ll

…

slp-non-pow-2-insertelement.ll

…

slp-umax-rdx-matcher-crash.ll

[SLP]Keep the original order in the reductions.

2024-07-09 12:26:42 -04:00

slp-variable-insertelement.ll

…

vectorizable-functions-inseltpoison.ll

…

vectorizable-functions.ll

…