I seem to recall an actual functionality bug (maybe in regex?) In libc++ due to a missing return somewhere earlier this year...
Might be worth prioritizing some warnings and cleaning them up more aggressively. On Aug 23, 2014 12:51 AM, "Chandler Carruth" <[email protected]> wrote: > > On Sat, Aug 23, 2014 at 12:39 AM, Eric Fiselier <[email protected]> wrote: > >> > We have a Clang warning for this - I would like to know why that didn't >> catch it? Are we not building this with -Wreturn-type? >> >> Currently we don't compile our tests with any warning flags. I'm not sure >> if it was just too noisy but looking at other defects >> Coverity found its clear we need to enable at least some of them. There >> are a bunch of assignments that should be comparisons as well. >> > > I would suggest compiling with most warnings (-Wall -Wextra -pedantic > maybe?) and using narrow and scoped pragmas to disable the warnings that we > intentionally trigger to test warned-on behavior that the library is > required to support. > > It'll likely be a ton of work though. > > _______________________________________________ > cfe-commits mailing list > [email protected] > http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits > >
_______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
