[llvm-commits] [llvm] r47711 - in /llvm/trunk: lib/Target/X86/X86ISelDAGToDAG.cpp test/CodeGen/X86/test-nofold.ll

Chris Lattner clattner at apple.com
Thu Feb 28 09:28:18 PST 2008


On Feb 27, 2008, at 11:46 PM, Evan Cheng wrote:

> Author: evancheng
> Date: Thu Feb 28 01:46:38 2008
> New Revision: 47711
>
> URL: http://llvm.org/viewvc/llvm-project?rev=47711&view=rev
> Log:
> Set to default: x86 no longer fold and into test if it has more than  
> one use.

Yay.  Should the AlwaysFoldAndInTest option be removed?

-Chris

>
>
> Modified:
>    llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp
>    llvm/trunk/test/CodeGen/X86/test-nofold.ll
>
> Modified: llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp?rev=47711&r1=47710&r2=47711&view=diff
>
> = 
> = 
> = 
> = 
> = 
> = 
> = 
> = 
> ======================================================================
> --- llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp (original)
> +++ llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp Thu Feb 28  
> 01:46:38 2008
> @@ -48,7 +48,7 @@
>   static cl::opt<bool>
>   AlwaysFoldAndInTest("always-fold-and-in-test",
>                 cl::desc("Always fold and operation in test"),
> -                cl::init(true), cl::Hidden);
> +                cl::init(false), cl::Hidden);
> }
>
> // 
> = 
> = 
> = 
> ----------------------------------------------------------------------= 
> ==//
>
> Modified: llvm/trunk/test/CodeGen/X86/test-nofold.ll
> URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/test-nofold.ll?rev=47711&r1=47710&r2=47711&view=diff
>
> = 
> = 
> = 
> = 
> = 
> = 
> = 
> = 
> ======================================================================
> --- llvm/trunk/test/CodeGen/X86/test-nofold.ll (original)
> +++ llvm/trunk/test/CodeGen/X86/test-nofold.ll Thu Feb 28 01:46:38  
> 2008
> @@ -1,6 +1,5 @@
> ; RUN: llvm-as < %s | llc -march=x86 -mcpu=yonah | grep {testl.* 
> %e.x.*%e.x}
> ; rdar://5752025
> -; XFAIL: *
>
> ; We don't want to fold the and into the test, because the and  
> clobbers its
> ; input forcing a copy.  We want:
>
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits




More information about the llvm-commits mailing list