[llvm-dev] [RFC] New pass: LoopExitValues

[Jake VanAdrighem via llvm-dev]

Do you have some specific performance measurements?

Jake

On Mon, Aug 31, 2015 at 10:16 AM, Steve King via llvm-dev <
llvm-dev at lists.llvm.org> wrote:

> Hello LLVM,
> This is a proposal for a new pass that improves performance and code
> size in some nested loop situations.  The pass is target independent.
> From the description in the file header:
>
> This optimization finds loop exit values reevaluated after the loop
> execution and replaces them by the corresponding exit values if they
> are available. Such sequences can arise after the
> SimplifyIndVals+LoopStrengthReduce passes. This pass should be run
> after LoopStrengthReduce.
>
> A former colleague created this pass back in LLVM 2.9 and we've been
> using it ever since.  I've done some light refactoring and
> modernization.
>
> This pass broke 4 existing tests that were sensitive to generated
> code.  I've corrected all these, but please give them special
> scrutiny.
>
> The patch is available here: http://reviews.llvm.org/D12494
>
> Please advise.
>
> Regards,
> -steve
> _______________________________________________
> LLVM Developers mailing list
> llvm-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20150831/535428eb/attachment.html>