[PATCH] D46283: [AArch64] Set vectorizer-maximize-bandwidth as default true

[Renato Golin via Phabricator via llvm-commits]

rengolin requested changes to this revision.
rengolin added a comment.
This revision now requires changes to proceed.

In https://reviews.llvm.org/D46283#1083378, @zatrazz wrote:

> [1] [llvm] r305960 - Enable vectorizer-maximize-bandwidth by default. (Dehao)
>  [llvm] r305990 - Revert "Enable vectorizer-maximize-bandwidth by default." (Diana Picus)
>  [llvm] r306336 - Enable vectorizer-maximize-bandwidth by default. (Dehao)
>  [llvm] r306344 - revert r306336 for breaking ppc test. (Dehao)
>  [llvm] r306473 - re-commit r306336: Enable vectorizer-maximize-bandwidth by default. (Dehao)
>  [llvm] r306792 - Revert "r306473 - re-commit r306336: Enable vectorizer-maximize-bandwidth by default." (Daniel)
>  [llvm] r306933 - Enable vectorizer-maximize-bandwidth by default.
>  [llvm] r306934 - revert r306336 for breaking ppc test.
>  [llvm] r306935 - re-commit r306336: Enable vectorizer-maximize-bandwidth by default.
>  [llvm] r306936 - Revert "r306473 - re-commit r306336: Enable vectorizer-maximize-bandwidth by default."


This is a horrible history, even worse if it broke things, not just performance...

Worse still is the number of targets (Arm, PPC) and it was re-committed a number of times without modification.

This trend is worrying and we should be *really* worried about the overall implementation before assuming that, because it didn't break SPEC2k6 on A72, it's a good patch.

We need a full analysis on what happened, why it's breaking so much, why the noise is so great and how can we reduce it while fixing the actual bugs.

Re-committing the patch as is simply won't work.

cheers,
--renato


Repository:
  rL LLVM

https://reviews.llvm.org/D46283