[PATCH] [InstCombine] Combine adjacent i8 loads.

[Chandler Carruth]

On Thu, May 1, 2014 at 11:38 PM, Andrew Trick <atrick at apple.com> wrote:

> On May 1, 2014, at 11:30 PM, Michael Spencer <bigcheesegs at gmail.com>
> wrote:
>
> Conceptually this is doing SLP vectorization, but it doesn't fit with our
> SLP algortihm which first finds a vectorizable use-def tree. Sticking it in
> GVN is another option, but again I'm concerned about it running before SLP.
> Maybe it can run in the SLP pass after the main algorithm.
>
>
> In this case wouldn't it be just as easy to make it a separate pass that
> runs after SLP?
>
>
> Yes, I thought we might be avoiding pass proliferation though. I’m neutral
> on that. The main things I’m concerned with is that it doesn’t interfere
> with SLP and is generally profitable.
>

I think we don't need to worry about pass proliferation. The current costs
that incurs are one of the things I'm keeping an eye on w/ the new manager.

I'm curious why not to use the def-use tree approach with the SLP
vectorizer so that we can also handle bitwise operations as Raul mentioned?
That seems like a nice advantage of using an integrated approach.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20140501/257dc092/attachment.html>