[llvm] r313701 - Revert "[GVNSink] Remove dependency on SmallPtrSet iteration order."

[Benjamin Kramer via llvm-commits]

This now fails again with reverse iteration order. Why do we have test
coverage for iteration order but not for the behavior you depend on
here?

On Wed, Sep 20, 2017 at 2:07 AM, Daniel Berlin via llvm-commits
<llvm-commits at lists.llvm.org> wrote:
> Author: dannyb
> Date: Tue Sep 19 17:07:25 2017
> New Revision: 313701
>
> URL: http://llvm.org/viewvc/llvm-project?rev=313701&view=rev
> Log:
> Revert "[GVNSink] Remove dependency on SmallPtrSet iteration order."
>
> This reverts commit r312156, because now the op and block arrays are not in the same order :(.
>
> Modified:
>     llvm/trunk/lib/Transforms/Scalar/GVNSink.cpp
>
> Modified: llvm/trunk/lib/Transforms/Scalar/GVNSink.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Scalar/GVNSink.cpp?rev=313701&r1=313700&r2=313701&view=diff
> ==============================================================================
> --- llvm/trunk/lib/Transforms/Scalar/GVNSink.cpp (original)
> +++ llvm/trunk/lib/Transforms/Scalar/GVNSink.cpp Tue Sep 19 17:07:25 2017
> @@ -229,14 +229,12 @@ public:
>    ModelledPHI(const VArray &V, const BArray &B) {
>      std::copy(V.begin(), V.end(), std::back_inserter(Values));
>      std::copy(B.begin(), B.end(), std::back_inserter(Blocks));
> -    std::sort(Blocks.begin(), Blocks.end());
>    }
>
>    /// Create a PHI from [I[OpNum] for I in Insts].
>    template <typename BArray>
>    ModelledPHI(ArrayRef<Instruction *> Insts, unsigned OpNum, const BArray &B) {
>      std::copy(B.begin(), B.end(), std::back_inserter(Blocks));
> -    std::sort(Blocks.begin(), Blocks.end());
>      for (auto *I : Insts)
>        Values.push_back(I->getOperand(OpNum));
>    }
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits