[llvm-commits] [llvm] r147683 - /llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp

[Cameron Zwarich]

Is there a test case where this does something wrong?

Cameron

On Jan 6, 2012, at 1:23 PM, Andrew Trick wrote:

> Author: atrick
> Date: Fri Jan  6 15:23:43 2012
> New Revision: 147683
> 
> URL: http://llvm.org/viewvc/llvm-project?rev=147683&view=rev
> Log:
> SCEVExpander: hoistStep should check strict dominance.
> 
> Modified:
>    llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp
> 
> Modified: llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp?rev=147683&r1=147682&r2=147683&view=diff
> ==============================================================================
> --- llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp (original)
> +++ llvm/trunk/lib/Analysis/ScalarEvolutionExpander.cpp Fri Jan  6 15:23:43 2012
> @@ -1564,7 +1564,7 @@
>   for (User::op_iterator OI = IncV->op_begin(), OE = IncV->op_end();
>        OI != OE; ++OI) {
>     Instruction *OInst = dyn_cast<Instruction>(OI);
> -    if (OInst && !DT->dominates(OInst, InsertPos))
> +    if (OInst && (OInst == InsertPos || !DT->dominates(OInst, InsertPos)))
>       return false;
>   }
>   IncV->moveBefore(InsertPos);
> 
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits