On Sat, 10 May 2003, Poul-Henning Kamp wrote: > In message <200305101515.h4AFFU8c039592_at_grimreaper.grondar.org>, Mark Murray wr > ites: > >"Poul-Henning Kamp" writes: > >> >> http://phk.freebsd.dk/misc/_.ia64.buildworld > >> > > >> >AHA! > >> > > >> >Please try this patch (you may need to do it manually because of c&p > >> >fubaring): [Context with patch lost] > >> Started, report to follow... > > > >Thanks! :-) > > > >I'm very confident about this. If it works for all archs, I'd like to get > >it into 5.1. > > Well, it didn't. It failed for both ia64 and i386 so far :-( The patch seems to fix one aspect of the problem. It restores creation of ${WORLDTMP}/usr/include/openssl/ in the loop that creates ${WORLDTMP}/usr/include/. This cannot be a complete fix, since there are many other subdirs of ${WORLDTMP}/usr/include/ that need to be created before they are used. The main problem seems to be that the _includes target needs to build at least its second subdir (src/include) before the others, but it does the opposite of this -- it uses par-includes to run "make includes" in the subdirs concurrently. Untested quick fix: %%% Index: Makefile.inc1 =================================================================== RCS file: /home/ncvs/src/Makefile.inc1,v retrieving revision 1.356 diff -u -0 -r1.356 Makefile.inc1 --- Makefile.inc1 5 May 2003 12:54:26 -0000 1.356 +++ Makefile.inc1 11 May 2003 07:21:11 -0000 _at__at_ -330 +404,1 _at__at_ - cd ${.CURDIR}; ${WMAKE} SHARED=symlinks par-includes + cd ${.CURDIR}; ${WMAKE} SHARED=symlinks includes %%% BruceReceived on Sat May 10 2003 - 22:40:35 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:07 UTC