Date: Thu, 22 Aug 2013 21:49:18 +1000 (EST) From: Bruce Evans <brde@optusnet.com.au> To: =?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?= <des@des.no> Cc: svn-src-stable@FreeBSD.org, svn-src-all@FreeBSD.org, src-committers@FreeBSD.org, svn-src-stable-9@FreeBSD.org, "David E. O'Brien" <obrien@FreeBSD.org> Subject: Re: svn commit: r254626 - stable/9/usr.bin/yacc Message-ID: <20130822213005.H1824@besplex.bde.org> In-Reply-To: <86y57ugn00.fsf@nine.des.no> References: <201308212257.r7LMvUmY053608@svn.freebsd.org> <8638q2i1oo.fsf@nine.des.no> <86y57ugn00.fsf@nine.des.no>
next in thread | previous in thread | raw e-mail | index | archive | help
This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --0-597222542-1377172158=:1824 Content-Type: TEXT/PLAIN; charset=X-UNKNOWN; format=flowed Content-Transfer-Encoding: QUOTED-PRINTABLE On Thu, 22 Aug 2013, [utf-8] Dag-Erling Sm=C3=B8rgrav wrote: > Dag-Erling Sm=C3=B8rgrav <des@des.no> writes: >> This breaks the build in usr.bin/expr because expr.y already provides a >> prototype, and gcc does not accept multiple prototypes, even if they're >> identical. That is arguably a bug in gcc, ISTR that it took several rounds of changes to fix the same bug in -current in expr and many other places. > Oh, it's actually our own fault for including -Wredundant-decls in > CFLAGS. Is there a good reason to keep it? > > -Wredundant-decls > Warn if anything is declared more than once in the same scope, = even > in cases where multiple declaration is valid and changes nothin= g. How else would you detect bogus declarations of library functions? Almost all -W flags are for detecting bugs that are mostly style bugs like this one. This one only becomes more than a style bug if: - the declaration becomes wrong - the header with the correct declaration becomes not included. Another bug in the commit is that it auto-generates code with a style bug (redundant extern). yacc in -current seems to auto-generate a function prototype for just one function (the same one), and this has the style bug too. This may be needed for K&R1 version compilers from 1978, but not for the STDC case. Bruce --0-597222542-1377172158=:1824--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20130822213005.H1824>