[PATCH] D39407: [(new) Pass Manager] instantiate SimplifyCFG with the same options as the old PM

[Brooks Moses via llvm-commits]

Further updating this: Taking the compiler from rL318299 with the new
llvm/Transforms/Scalar/SimplifyCFG.h and setting needCanonicalLoops to
false fixes the issue; taking the
old llvm/Transforms/Scalar/SimplifyCFG.h and setting needCanonicalLoops to
true reintroduces the issue.  The values of other flags seem to be
unrelated to this regression.

With profiling we're particularly seeing slowdowns in
the DecompressAllTags<SnappyArrayWriter> function.


On Fri, Dec 15, 2017 at 2:09 PM, Chandler Carruth via llvm-commits <
llvm-commits at lists.llvm.org> wrote:

> FYI, this benchmark actually exists in some form open source:
> https://github.com/google/snappy
>
> Still, we'll need to investigate to know why this is slowing things down
> (so much).
>
> On Fri, Dec 15, 2017 at 2:02 PM Brooks Moses via Phabricator via
> llvm-commits <llvm-commits at lists.llvm.org> wrote:
>
>> brooksmoses added a comment.
>>
>> We're seeing about 10% regressions in an internal benchmark as a result
>> of this change.  Still digging further, but wanted to send a heads-up.
>>
>>
>> Repository:
>>   rL LLVM
>>
>> https://reviews.llvm.org/D39407
>>
>>
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at lists.llvm.org
>> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
>>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20171215/7d03b44b/attachment.html>