kicad-developers team mailing list archive

[jp charras <jp.charras@xxxxxxxxxx>]

Le 05/04/2016 13:23, Mário Luzeiro a écrit :
> Hi all,
> 
> Thanks for the ccpcheck suggestion JP, that was something I was looking for some time...
> (I found already a bug with that!)
> 
> "There are still quite a few "variable not used" warnings but those
> are in code contributed by Mario so I have left them alone for now."
> 
> I am running this also in my branch and found that issues.
> I am not sure why but they are false positives, I guess it is the way the structures are constructed, example:
> 
> GLM_ALIGNED_STRUCT(CLASS_ALIGNMENT) RAY
> 
> is uses a define GLM_ALIGNED_STRUCT from GLM library.
> probably cppcheck is not able to resolve this define in a proper way.
> 
> Any suggestions?
> 
> Mario


Cppcheck, like all static analyzers, is useful to find some errors and mistakes.

However, it is not usually possible (and perhaps not good) to fix all "errors" find by a static
analyzer.
There are always false positive or issues detected by the detected which cannot really happen due to
the logic of the code.

Just, be sure remaining errors or warnings are acceptable (false positive),
and actual errors and warnings are fixed.

> 
> ________________________________________
> From: Kicad-developers [kicad-developers-bounces+mrluzeiro=ua.pt@xxxxxxxxxxxxxxxxxxx] on behalf of Cirilo Bernardo [cirilo.bernardo@xxxxxxxxx]
> Sent: 05 April 2016 11:45
> To: jp charras
> Cc: KiCad Developers
> Subject: Re: [Kicad-developers] Merging 3d_initial_merge branch.
> 
> Hi Jean-Pierre, Wayne,
> 
>  I've run cppcheck and fixed a number of issues:
> 1. ensured "delete [] X" where "delete X" was accidentally used
> 2. ensured variable initialization
> 3. fixed a number of errors and warnings reported by cppcheck
> (most of those warnings were in fact errors)
> 
> There are still quite a few "variable not used" warnings but those
> are in code contributed by Mario so I have left them alone for now.
> 
> cppcheck reports a number of "(error, inconclusive)" but I have
> checked to make sure that those are in fact not errors and that
> the program has actually performed appropriate checks.
> 
> All changes have been committed to r6558.  I hope all goes
> smoothly and we can start to get more feedback on this new
> 3D plugin system.
> 
> - Cirilo
> 
> 
> On Tue, Apr 5, 2016 at 6:02 PM, jp charras <jp.charras@xxxxxxxxxx<mailto:jp.charras@xxxxxxxxxx>> wrote:
> Hi, Cirilo.
> 
> We (Wayne and me) are willing to merge your 3d_initial_merge branch.
> 
> It looks good for now, and adds very good features.
> 
> Could you run cppcheck on you code (the 3d-viewer and others new folders)?
> (see http://cppcheck.sourceforge.net)
> cppcheck is fast, very easy to install and use (just run cppcheck on a given folder).
> 
> It is no as powerful as Coverity, but it is very helpful especially for a first basic analysis
> before Coverity is run.
> 
> Especially, not initialized members and 3 errors (I am thinking related to "delete" versus
> "delete[]") must be fixed (but some others could be fixed).
> not initialized members create a lot of warning with Coverity, and are potential issues, so I am
> thinking initializing them is good, even when it is not really needed.
> 
> Also there are some not used variables detected by the compiler.
> 
> Thanks for your work.
> 
> --
> Jean-Pierre CHARRAS


-- 
Jean-Pierre CHARRAS