[PATCH] D72689: [SLP] Vectorize partially vectorized instructions

[Anton Afanasyev via Phabricator via llvm-commits]

anton-afanasyev created this revision.
anton-afanasyev added reviewers: RKSimon, ABataev, spatel, dtemirbulatov.
Herald added subscribers: llvm-commits, hiraditya.
Herald added a project: LLVM.

This is just a draft with a lot of dirty hacks to show proof-of-concept
for the vectorization of partially vectorized instructions.

The main motivation comes from here: http://llvm.org/pr42022.
We need to merge `store <2 x float>, store <2 x float>` to `store <4 x float>`.
Such cases happen, for instance, when SLP-optimizing already SLP-optimized inlined
code.

The solution is to permit `insertelement`'s as vectorization tree node, but this should
be done carefully: `insertelement`'s couldn't be scheduled (they have inner deps).

Early feedback is appreciated since this unfinished work could concern another revisions
(like D44067 <https://reviews.llvm.org/D44067> or PR35732).

This is follow-up of D70587 <https://reviews.llvm.org/D70587> and D70924 <https://reviews.llvm.org/D70924>


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D72689

Files:
  llvm/lib/Transforms/Vectorize/SLPVectorizer.cpp

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D72689.237897.patch
Type: text/x-patch
Size: 11487 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20200114/3d018676/attachment-0001.bin>