[cfe-commits] r50176 - in /cfe/trunk: include/clang/Analysis/PathSensitive/GRExprEngine.h include/clang/Analysis/PathSensitive/GRTransferFuncs.h lib/Analysis/CFRefCount.cpp lib/Analysis/GRExprEngine.cpp lib/Analysis/GRSimpleVals.cpp lib/Analysis/GRSimpleVals.h

Ted Kremenek kremenek at apple.com
Wed Apr 23 23:04:49 PDT 2008

On Apr 23, 2008, at 5:55 PM, Chris Lattner wrote:

> On Apr 23, 2008, at 1:12 PM, Ted Kremenek wrote:
>> Author: kremenek
>> Date: Wed Apr 23 15:12:28 2008
>> New Revision: 50176
>> URL: http://llvm.org/viewvc/llvm-project?rev=50176&view=rev
>> Log:
>> Fixed: <rdar://problem/5881148>
>> Problem:
>> In the recently refactored VisitDeref (which processes  
>> dereferences), we
>> were incorrectly skipping the node just generated for the  
>> subexpression
>> of the dereference.  This was a horrible regression.
> No regression test? :)

I wiil add a test for this.  This bug manifested by producing a false  
path, essentially by bifurcating the ExplodedGraph and creating two  
paths where there should have been one... except that on one path some  
state would not have been recorded.  Consequently, the bug was  
detected when some checker hit a weird false positive.  I can add  
regression tests for these kind of bugs as they come along, but  
because they manifest in weird ways their hard to detect without just  
writing a lot of tests.

More information about the cfe-commits mailing list