--- layout: fc_discuss_archives title: Message 8 from Frama-C-discuss on September 2011 ---
Le 14/09/2011 16:18, haihao shen a ?crit : > Thanks Pascal and Julien. > Some of research work showed that macro may hidden some bugs, so > sometimes, we still need to do static analysis based on the purely > source code I think. Macros may hidden bugs to developers but, as far I know, inlining macros (as preprocessing does) does not remove bugs: if you use a correct preprocessor and next a correct static analysis tool on the resulting file, you do find bugs introduced by misused macros. The major drawback of this approach is that the exact origin of the bug may be more difficult to identify for the end-user. That is one reason why you suggestion would be useful. But, if the Frama-C team chooses to implement a new C preprocessor from scratch, it would take a huge amount of time used to reinvent the wheel instead of writing new static analysers which is our primary goal. Also it would introduce a new large and complex piece of code in Frama-C which could contain critical bugs impacting each end-user. That is two reasons why the Frama-C team chooses to not implement such a preprocessor. Fortunatly, Frama-C is an open source and extensible platform: anyone is free to implement itself such a preprocessor and plug in to Frama-C. The Frama-C team could even provide some helps through this mailing list. Maybe someone reading this message would like to contribute this way :-). -- Julien