[llvm-commits] [patch] Tautology?

Chris Lattner clattner at apple.com
Wed Mar 17 23:20:15 PDT 2010


On Mar 17, 2010, at 11:16 PM, Evan Cheng wrote:

> 
> On Mar 17, 2010, at 11:00 PM, Chris Lattner wrote:
> 
>> Yeah, that is weird.  Evan, you added this in r40033, do you happen to know what you meant here?
> 
> No. I am blanking. Looks like copy-n-paste bug. I think all the != should be ==?

No idea!  Please make whatever fix you think is right, optionally just removing the code as dead.

-Chris

> 
> Evan
> 
>> 
>> -Chris
>> 
>> On Mar 17, 2010, at 7:49 PM, Zhongxing Xu wrote:
>> 
>>> Index: utils/TableGen/Record.cpp
>>> ===================================================================
>>> --- utils/TableGen/Record.cpp    (版本 98703)
>>> +++ utils/TableGen/Record.cpp    (工作副本)
>>> @@ -644,20 +644,6 @@
>>>    DagInit *LHSs = dynamic_cast<DagInit*>(LHS);
>>>    DagInit *RHSs = dynamic_cast<DagInit*>(RHS);
>>>    if (LHSs && RHSs) {
>>> -      DefInit *LOp = dynamic_cast<DefInit*>(LHSs->getOperator());
>>> -      DefInit *ROp = dynamic_cast<DefInit*>(RHSs->getOperator());
>>> -      if (LOp->getDef() != ROp->getDef()) {
>>> -        bool LIsOps =
>>> -          LOp->getDef()->getName() == "outs" ||
>>> -          LOp->getDef()->getName() != "ins" ||
>>> -          LOp->getDef()->getName() != "defs";
>>> -        bool RIsOps =
>>> -          ROp->getDef()->getName() == "outs" ||
>>> -          ROp->getDef()->getName() != "ins" ||
>>> -          ROp->getDef()->getName() != "defs";
>>> -        if (!LIsOps || !RIsOps)
>>> -          throw "Concated Dag operators do not match!";
>>> -      }
>>>      std::vector<Init*> Args;
>>>      std::vector<std::string> ArgNames;
>>>      for (unsigned i = 0, e = LHSs->getNumArgs(); i != e; ++i) {
>>> 
>>> 
>>> LOp->getDef()->getName() != "ins" || LOp->getDef()->getName() != "defs";
>>> will always be true.
>>> _______________________________________________
>>> 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