On Oct 1, 2013, at 22:16 , Daniel Marjamäki <[email protected]> wrote:
> > Hello! > >> This should be a warning, not a static analyser check, shouldn't it? > > Is anybody against this? > > Personally I feel very confident about this check regarding signal/noise. > However putting it in the static analyser to start with felt like a safe > choice. I'm fine with this. I didn't realize at first it was just for character literals; with that restriction this does seem like false positives would be unlikely. Thanks, Richard. (Sorry to force another iteration, Anders. The existing warning is in SemaExpr.cpp: diagnoseStringPlusInt.) Daniel, delaying += and char plus string seem fine to me, though they might just be handled if you hook into diagnoseStringPlusInt. And thanks for the reminder about things like s8, u8, etc. Jordan _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
