[PATCH] D124218: [LoadStoreVectorizer] Consider if operation is faster than before

Fri Nov 18 14:31:24 PST 2022

rampitec added a comment.

In D124218#3934383 <https://reviews.llvm.org/D124218#3934383>, @rampitec wrote:

> In D124218#3932136 <https://reviews.llvm.org/D124218#3932136>, @arsenm wrote:
>
>> Needs testcase
>
> This is NFCI. Even the next in stack (D124219 <https://reviews.llvm.org/D124219>) does not really change the codegen, just makes model more sane.

This is really NFCI, since no targets use it and return a relative speed yet. However, subsequent D124219 <https://reviews.llvm.org/D124219> could have some tests for sub-dword LDS operations with alignment smaller than 4. I will add these to the D124219 <https://reviews.llvm.org/D124219>.
The other patch which should depend on this (incomplete, as this one hung for a long time) is the scratch handling, it is always faster to use a dword access no matter what.

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D124218/new/

https://reviews.llvm.org/D124218