cppcheck

Thanks for the quick answer. The hint on the files size makes sense, I will try it out and try to measure... By the way: looking at the code it seems that cppcheck always gives back either EXIT_SUCCESS or EXIT_FAILURE as return code, is that correct? ... [More] And if it returns EXIT_FAILURE, does it always include exactly one statement "cppcheck: error: ... " in its output and always as the last statement? And my last question: is it okay, if I use your cppcheck logo in my project of a plugin? Great many thanks and regards frithjof [Less]

I recommend that you don't put headers in the --file-list. Use #include to include them from your source files - then they will be analysed properly. The order should have no impact on performance as far as I see. I don't know if it makes any real ... [More] difference but if you use multithreaded checking it could be best if the smallest files are checked last (to avoid that cores are idle in the end while a single core finish checking the large file). [Less]

Hello Daniel, I just have a question regarding an feature you added to cppcheck some time ago on request of one of the developers of the code::blocks ide. The feature is the possibility to give a file with a line wise list of the files to be checked ... [More] by cppcheck using --file-list=<file> parameter. My question is if the sorting of the files in the list does have an effect on the execution time of cppcheck? Like if it will execute faster if in the list first come all header files and then all the source files... or any other sorting. Thanks in advance [Less]

sure

I read in the wiki about registering an account - May I send you my htpasswd string via e-mail as well?

We had to move trac: https://sourceforge.net/p/forge/community-docs/Hosted%20Apps%20Retirement/

Today I tried to file an issue, but I found that my Sourceforge login no longer works on the Cppcheck Trac instance. When and why has this changed?

the destructor only needs to be virtual if a derived class with a non-empty destructor is deleted using a base class pointer You are absolutely right. But issues that are (kind of) "prepared" by this argument are at least leaks, sometimes ... [More] catastrophic errors. And they may be hard to detect. And I came to the conclusion, that another argument works better: if there is already a VTBL, a virtual destructor is relatively cheap. The presence of virtual functions (especially pure ones) indicate that there will be inheritance... ...and sometimes also derived destructor calls. Regardless whether inconclusive or not: many thanks for this new check! :) [Less]

I especially like the new(?) virtualDestructor check. Why is it inconclusive? The reason is that the destructor only needs to be virtual if a derived class with a non-empty destructor is deleted using a base class pointer. This warning is inconclusive because the check does not try to determine if that happens.

Minor issues between this version and the previous one. In this version, I had to manually download msvcp120.dll and msvcr120.dll since they were not available on my non-development machine. In addition, the GUI version still will not run and gives ... [More] me the following error: cppcheckgui This application failed to start because it could not find or load the Qt platform plugin "windows". Reinstalling the application may fix this problem. OK Thanks, Bassam [Less]