Date: Mon, 7 Sep 1998 13:19:50 +1000 From: David Dawes <dawes@rf900.physics.usyd.edu.au> To: current@FreeBSD.ORG Subject: Re: cvs commit: src/sys/kern subr_bus.c uipc_socket2.c Message-ID: <19980907131950.D14380@rf900.physics.usyd.edu.au> In-Reply-To: <199809070103.SAA13578@bubba.whistle.com>; from Archie Cobbs on Sun, Sep 06, 1998 at 06:03:30PM -0700 References: <199809061801.LAA00223@austin.polstra.com> <199809070103.SAA13578@bubba.whistle.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Sep 06, 1998 at 06:03:30PM -0700, Archie Cobbs wrote: >John Polstra writes: >> > BTW: How far are we from using -Wall -Werror for the whole source >> > tree, including kernel ? >> >> We should never add -Werror to the standard flags. If somebody >> compiles without "-O", the compiler generates spurious warnings. >> These kill the build if -Werror is present. > >That's arguable. Anyone who's capable of removing the -O from some >included makefile is capable of removing -Werror from some included >makefile. > >In practice, however, -Werror may not be workable... it would be an >interesting experiment anyway. > >> Of course, we should still strive for the goal of "-O -Wall -Werror" >> working. > >Not to mention -Wstrict-prototypes, -Wmissing-prototypes, >-Wmissing-declarations, -Wnested-externs, ... what else?? In the current XFree86 devel source, we use those plus -Wpointer-arith and -Wredundant-decls. I consider the former important. The latter one generates a lot of noise from the system headers on many OSs (although FreeBSD is fairly clean in that regard). We don't use -Werror, and we'll probably relax the warnings before this development code is released. David To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980907131950.D14380>