Re: Suggested buildd log check
* Jonathan Nieder <jrnieder@gmail.com> [111126 01:15]:
> > If you have ideas for warnings/diagnostics visibile in buildd logs
> > that are worth having listed in a view like that, please let me
> > know.
>
> I'm not sure how your infrastructure copes with something like this,
> but I would find it useful to have gcc 4.6's machine-parsable warnings
> in such a list. They look like this:
>
> ../../src/gcc/graphite.c:59:29: warning: non-local variable 'cloog_pointers__' with anonymous type is questionable in C++ [-Wc++-compat]
>
> I guess a regex like "\[-W[-+=a-z0-9]*\]" would do, plus
> "\[-pedantic\]" and "\[-fpermissive\]"..
>
> I suppose these could all be lumped as one diagnostic type, except
> that when new gcc releases introduce new warnings (like the recent
> -Wunused-but-set-variable and -Wunused-but-set-parameter), they could
> get a different tag. What do you think?
While this could be extracted, I'm not sure it makes that much sense.
Few packages are totally warning free and some are quite harmless, to
it could distract to have too much information. (it could be info and
not error and warning, but it still might not have that much
information).
Also the logs are usually done with very different compiler versions.
Having packages in unstable with the last log two years old is not that
uncommon.
I definitely plan to add information pages for any dangerous warnings
(currently a global rescan also looking for "warning: assignment makes
pointer from integer without a cast" is running), and I'd love to get
hints for other interesting warnings like those to add to that list.
I can also gather more information, but would like to see some idea what
to actually do with that vast amount of information.
Bernhard R. Link
Reply to: