Re: Breaking old compilation paths
-On [20020504 20:15], Jeroen Ruigrok/asmodai ([EMAIL PROTECTED]) wrote: It is only here that warnings are treated as errors. I am thinking something in the upgrade path is not 100% thought out, trying to see what exactly, but my focus has been way more on STABLE than CURRENT, so some more exposed experience gladly received. Of course, the entire build complains due to WARNS being raised above 0 and NO_WERROR has not been defined. The only reference I can find in UPDATING is: 20020225: Warnings are now errors in the kernel. Unless you are a developer, you should add -DNO_WERROR to your make line. I definately do not call libiberty kernel related. Nor do we, by default, advocate NO_WERROR. So in effect, we broke moving from older versions to newer versions of CURRENT. -- Jeroen Ruigrok van der Werven / asmodai / Kita no Mono [EMAIL PROTECTED], finger [EMAIL PROTECTED] http://www.softweyr.com/asmodai/ | http://www.tendra.org/ I am, I was, and I will be... To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: Breaking old compilation paths
On Sun, May 05, 2002 at 02:42:22PM +0200, Jeroen Ruigrok/asmodai wrote: I definately do not call libiberty kernel related. Nor do we, by default, advocate NO_WERROR. So in effect, we broke moving from older versions to newer versions of CURRENT. Can you track down what the actual cause of the breakage is? I can't immediately see why compiling on an older version should cause extra warnings, because the build should be using the up-to-date system headers and code. Kris msg38061/pgp0.pgp Description: PGP signature
Re: Breaking old compilation paths
-On [20020506 00:30], Kris Kennaway ([EMAIL PROTECTED]) wrote: On Sun, May 05, 2002 at 02:42:22PM +0200, Jeroen Ruigrok/asmodai wrote: I definately do not call libiberty kernel related. Nor do we, by default, advocate NO_WERROR. So in effect, we broke moving from older versions to newer versions of CURRENT. Can you track down what the actual cause of the breakage is? I can't immediately see why compiling on an older version should cause extra warnings, because the build should be using the up-to-date system headers and code. The libiberty Makefile sets WARNS higher than 0, which effectively [through share/mk] also sets -Werror since NO_WERROR is not defined. I would need to see if it is something libc related or something which causes the function failure. -- Jeroen Ruigrok van der Werven / asmodai / Kita no Mono [EMAIL PROTECTED], finger [EMAIL PROTECTED] http://www.softweyr.com/asmodai/ | http://www.tendra.org/ Where does the world end if the enternal optimist loses faith..? To Unsubscribe: send mail to [EMAIL PROTECTED] with unsubscribe freebsd-current in the body of the message
Re: Breaking old compilation paths
On Mon, May 06, 2002 at 07:31:30AM +0200, Jeroen Ruigrok/asmodai wrote: -On [20020506 00:30], Kris Kennaway ([EMAIL PROTECTED]) wrote: On Sun, May 05, 2002 at 02:42:22PM +0200, Jeroen Ruigrok/asmodai wrote: I definately do not call libiberty kernel related. Nor do we, by default, advocate NO_WERROR. So in effect, we broke moving from older versions to newer versions of CURRENT. Can you track down what the actual cause of the breakage is? I can't immediately see why compiling on an older version should cause extra warnings, because the build should be using the up-to-date system headers and code. The libiberty Makefile sets WARNS higher than 0, which effectively [through share/mk] also sets -Werror since NO_WERROR is not defined. Yes, but why is this only a problem for upgrade builds from older systems? That's what doesn't sound right to me. Kris msg38064/pgp0.pgp Description: PGP signature