Date: Sun, 19 Jun 2011 00:36:06 +1000 (EST) From: Bruce Evans <brde@optusnet.com.au> To: Ben Laurie <benl@freebsd.org> Cc: svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r223262 - in head: cddl/contrib/opensolaris/lib/libdtrace/common contrib/binutils/bfd contrib/binutils/gas contrib/binutils/gas/config contrib/binutils/ld contrib/binutils/opcodes contr... Message-ID: <20110619000207.V7011@besplex.bde.org> In-Reply-To: <201106181356.p5IDuXhW044171@svn.freebsd.org> References: <201106181356.p5IDuXhW044171@svn.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, 18 Jun 2011, Ben Laurie wrote: > Author: benl > Date: Sat Jun 18 13:56:33 2011 > New Revision: 223262 > URL: http://svn.freebsd.org/changeset/base/223262 > > Log: > Fix clang warnings. > > Approved by: philip (mentor) Most of these seem to be bugs in clang, so source code can only be broken by avoiding the warnings there. In particular: - casting to void is a large style bug in gnu code. Warnings in vendor code shouldn't be fixed anyway. - it is an old gcc bug to warn about use of the extremely unsurprising precedence of AND operators over OR operators. This bug is not much of a problem since it is controlled by -Wparentheses. -Wparentheses is impiled by -Wall, which is enables at WARNS >= 2. But it is a new clang bug to make this warning unconditional (it can be turned off or configured off, but this is not supported in FreeBSD or documented in clang.1). So it is now fatal at WARNS >=1, since WARNS >=1 turns warnings into errors. Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110619000207.V7011>