[cfe-commits] r139437 - in /cfe/trunk: include/clang/AST/Stmt.h lib/AST/Stmt.cpp lib/Analysis/CFG.cpp test/SemaCXX/return-noreturn.cpp

[Ted Kremenek]

Ah, nevermind.  I misread the patch.

On Sep 9, 2011, at 9:43 PM, Ted Kremenek wrote:

> 
> On Sep 9, 2011, at 5:02 PM, Chandler Carruth wrote:
> 
>> Modified: cfe/trunk/lib/Analysis/CFG.cpp
>> URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Analysis/CFG.cpp?rev=139437&r1=139436&r2=139437&view=diff
>> ==============================================================================
>> --- cfe/trunk/lib/Analysis/CFG.cpp (original)
>> +++ cfe/trunk/lib/Analysis/CFG.cpp Fri Sep  9 19:02:34 2011
>> @@ -723,9 +723,7 @@
>>   if (CompoundStmt *CS = dyn_cast<CompoundStmt>(S)) {
>>     for (CompoundStmt::body_iterator BI = CS->body_begin(), BE = CS->body_end()
>>         ; BI != BE; ++BI) {
>> -      Stmt *SI = *BI;
>> -      if (LabelStmt *LS = dyn_cast<LabelStmt>(SI))
>> -        SI = LS->getSubStmt();
>> +      Stmt *SI = (*BI)->stripLabelLikeStatements();
>>       if (DeclStmt *DS = dyn_cast<DeclStmt>(SI))
>>         Scope = addLocalScopeForDeclStmt(DS, Scope);
>>     }
>> @@ -734,9 +73
> 
> Chandler,
> 
> I don't think this is right.  The CFG builder explicitly processes CaseStmts and DefaultStmts.  It shouldn't be skipping over them here.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20110909/2bee479e/attachment.html>