Re: CVS commit: pkgsrc/lang/mercury

[David Holland <dholland-pkgchanges%netbsd.org@localhost>]

On Wed, Mar 28, 2012 at 12:04:10AM +0400, Aleksej Saushev wrote:
 > David Holland <dholland-pkgchanges%netbsd.org@localhost> writes:
 > 
 > > On Mon, Jan 30, 2012 at 05:40:02AM +0000, Aleksej Saushev wrote:
 > >  > Modified Files:
 > >  >         pkgsrc/lang/mercury: Makefile PLIST distinfo
 > >  >         pkgsrc/lang/mercury/patches: patch-ab patch-ap patch-at
 > >  > Removed Files:
 > >  >         pkgsrc/lang/mercury/patches: patch-ar patch-as
 > >  > 
 > >  > Log Message:
 > >  > Update to Mercury 11.07
 > >
 > > /home/pkgbulk/c/obj/lang/mercury/work/mercury-compiler-11.07/boehm_gc/../scripts/mgnuc
 > >  --grade asm_fast.gc --no-ansi --no-check --no-mercury-stdlib-dir 
 > > -I./include -I./libatomic_ops-install/include -DNO_EXECUTE_PERMISSION 
 > > -DLARGE_CONFIG  -fpic -DMR_PIC   -DNO_DEBUGGING 
 > > -I/home/pkgbulk/b/pkg/include -I/usr/include  -c -o os_dep.o os_dep.c
 > > os_dep.c:503:9: error: redefinition of 'GC_data_start'
 > > os_dep.c:454:9: note: previous definition of 'GC_data_start' was here
 > > gmake[3]: *** [os_dep.o] Error 1
 > >
 > > Note that this is an improvement on the prior build behavior, which
 > > was SIGSEGV. :-/
 > 
 > No, it isn't. It just doesn't reach the stage where it broke previously.

Oh. Blah.

...anyhow, would it be better to make it use pkgsrc boehm-gc? (And is
it using boehm-gc and threads?)

-- 
David A. Holland
dholland%netbsd.org@localhost