Erik de Castro Lopo <mle+s...@mega-nerd.com> writes:
> Daniel Pittman wrote:
>
>> You do mean, specifically, bugs that have the appropriate sparse tests
>> and/or type annotations in place, right?
>
> Possibly not. Remeber I played with Sparse for all of 10 minutes.
I wasn't sure how much time you spent investigating it. :)
>> Can you render them down to a trivial example that demonstrates the
>> failure?
>
> Well a trivial example:
>
> int pos, array [100] ;
> read (fd, &pos, sizeof (pos));
> array [pos] = 0 ;
>
> I know analysis tools like coverity can find stuff like this.
*nod* I don't think that is one of the problems that sparse tests
for[1], in large part because GCC already has a warning for
use-without-initialization and sparse tends to be "things GCC can't
test".
In the bigger picture, your original question about static analysis
tools for C missed one critical detail: what do you actually want it to
statically analyse?
After all, sloccount is technically a static analysis tool for C, but
the only thing it outputs are line counts and estimates for time and
dollar-cost to reproduce the software — so, probably not what you are
after. :)
Regards,
Daniel
Footnotes:
[1] ...though I confess to not looking at it in the last few months, so
the most bleeding edge version may have changed this.
_______________________________________________
coders mailing list
coders@slug.org.au
http://lists.slug.org.au/listinfo/coders
