This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Code coverage tool


David Flynn <dav@chess.plus.com> writes:

>  Are there any plans to add a code coverage tool to gcc?
>
>  Essentially it would be like profiling but at greater detail.
>
>  {} nests (including functions), each () on ifs, whiles, etc, both sides of 
> the : in () ? statment 1 : statement 2  expressions, each case in a switch, 
> etc would have a counter. Basically there is a counter for every possible 
> path of C code. When the code goes through there, the counter is incremented 
> (may have to be a 8 byte counter for certain apps). Then it should be 
> possible to dump the counts and for a separate tool (like gprof) to be able 
> to line up the counts with the source itself. That way you can determine if 
> all paths through your code have been executed for testing, or for statistics 
> if you test based upon likelyhood of code being used.
>
>  It should be also possible to combine two counter files, for multiple tests 
> when the program exits.
>
>  In doing so it makes it easy to prove that all code has been executed and 
> means there being fewer bugs where programmers have forgotten to test paths 
> of code (particularly in complicated if logic statements).

What do you want to do that cannot be done with gcov and 
gcc -fprofile-arcs -ftest-coverage as documented in the GCC manual (I
just checked gcc 2.95)?

Andreas
-- 
 Andreas Jaeger
  SuSE Labs aj@suse.de
   private aj@arthur.inka.de
    http://www.suse.de/~aj


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]