On Tue, Aug 24, 2004 at 06:48:55PM +0200, Harti Brandt wrote: > On Tue, 24 Aug 2004, Ruslan Ermilov wrote: > > RE>On Tue, Aug 24, 2004 at 05:11:32PM +0200, Harti Brandt wrote: > RE>> On Tue, 24 Aug 2004, Tomas Verbaitis wrote: > RE>> > RE>> TV>2) Next failure was this: > RE>> TV>===> snmp_atm > RE>> TV>cat > RE>> TV>/usr/src/lib/libbsnmp/modules/snmp_atm/../../../../contrib/ngatm/snmp_atm/atm_tree.def > RE>> TV>/usr/src/lib/libbsnmp/modules/snmp_atm/atm_freebsd.def | gensnmptree -e > RE>> TV>begemotAtm > atm_oid.h > RE>> TV>line 110: junk after closing ')' > RE>> > RE>> The problem seems to be that it is running an old gensnmptree from > RE>> /usr/bin although it should build and run a new one. Can you tell me what > RE>> is the __FreeBSD_version in /usr/include/osreldate.h on the system you are > RE>> trying to upgrade? > RE>> > RE>Harti, nothing to worry about, you can safely close PR. This is what > RE>happens when people do "make includes" without understanding that it > RE>will ruin their build environment. ``make buildworld OSRELDATE=0'' > RE>usually helps these pour souls. ;) > > Ok. I was now half an hour thinking about the versions. It was my fault > that I did not bump the FreeBSD_version then, so the actual commit of the > gensnmptree is between two version bumps. I choose the later one for the > check in Makefile.inc1 and think this should be safe. > It should be okay, and I thought we already fixed this in Makefile.inc1 by using the correct __FreeBSD_version check. Ask for the ``ls -l /usr/include/osreldate.h /usr/bin/gensnmptree'' output when diagnosing the problem -- like I said, chances are good that people fooled their build environment to think it's more fresh than is actually. Cheers, -- Ruslan Ermilov ru_at_FreeBSD.org FreeBSD committer
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:08 UTC