[cfe-commits] r86523 - in /cfe/trunk: include/clang/Analysis/PathSensitive/SVals.h lib/Analysis/FixedAddressChecker.cpp lib/Analysis/GRExprEngineInternalChecks.cpp lib/Analysis/GRExprEngineInternalChecks.h lib/Analysis/SVals.cpp test/Analysis/ptr

[Ted Kremenek]

On Nov 9, 2009, at 4:33 PM, Zhongxing Xu wrote:

> 2009/11/10 Ted Kremenek <kremenek at apple.com>:
>> Hi Zhongxing,
>>
>> Very nice  I vaguely recall we use to do this check, albeit not as  
>> directly.
>
> Perhaps in EvalLocation().
>
>>  How do we want to handle cases like PR 5272
>> (http://llvm.org/bugs/show_bug.cgi?id=5272)?
>
> Leave it as a false positive.
>
>> Using hardcoded addresses is
>> fine in certain circumstances.
>
> Definitely. So I see this kind of check as some kind of guide to the
> programmer, not as some absolute judger of the code. False positives
> are allowed. There should not be too much.

Makes sense.  Eventually we should provide a mechanism to silence  
false positives of this nature.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20091109/46274439/attachment.html>