[PATCH] D26726: [SimplifyCFG] Fix for non-determinism in codegen

[Grang, Mandeep Singh via llvm-commits]

@dblaike

Yes, this is uncovered by the "reverse order iteration" and I put in a 
reference to that in the summary section of the patch.

The test case is Transforms/SimplifyCFG/bug-25299.ll which fails due to 
reverse iteration.

I will elaborate this in the patch comments.

Thanks,

Mandeep


On 11/21/2016 10:14 AM, David Blaikie wrote:
> it'd be good to include a minimized test case (& I assume this is 
> reproduced by the experimental "reverse iteration order" switch - so a 
> reference to that could be handy for those trying to evaluate whether 
> this is the best way to address the issue, etc) to help demonstrate 
> the problem/fix.
>
> Such things don't necessarily need to be committed (though if we're 
> thinking of having a "reverse iteration mode" build - that seems worth 
> keeping/committing the test cases for demonstration purposes & help 
> avoid regressions, etc)
>
> On Mon, Nov 21, 2016 at 10:07 AM Mandeep Singh Grang via llvm-commits 
> <llvm-commits at lists.llvm.org <mailto:llvm-commits at lists.llvm.org>> wrote:
>
>     mgrang added a comment.
>
>     Ping.
>
>
>     Repository:
>       rL LLVM
>
>     https://reviews.llvm.org/D26726
>
>
>
>     _______________________________________________
>     llvm-commits mailing list
>     llvm-commits at lists.llvm.org <mailto: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/20161121/fa7b8ab0/attachment.html>