Some of the shared library magic is very hackish, however the net-snmp
library building make-fu is hackish as well, so I don't feel bad about it.
This commit also (re-)enables the host module for much more information.
It also fixes a long standing bug where 'snmpwalk localhost public' would
hang.
this change (with a repocopy) and update various references.
this does not upgrade this port to the most recent version of net-snmp,
that will be coming soon.
options `start' and `stop' now (unless I have forgotten any). This allows
us to call the scripts from /etc/rc.shutdown with the correct option.
The (42 or so) ports that already DTRT before are unchanged.
previous commit message to bsd.port.mk, which said INSTALL_SHLIBS. Boo.)
Line up the rhs of variable assignments nicely. Remove a couple of extra
whitespaces while I'm here.
Suggested by: sobomax
who have a -CURRENT before:
Fri May 5 9:58:59 2000 UTC by phk
.. however, since no __FreeBSD_version bump was performed, I can't
handle this with the usual Makefile magic.
Submitted by: kuriyama-san
Reviewed by: Matt Ayres <mayres@chc-chimes.com> ("it works now")
o Look for perl in the right place (patch-ah)
o Add a typedef that ucd-snmp wants (patch-ag)
o Install headers files with the right permissions
Submitted by: <jack@germanium.xtalwind.net>
PR: ports/13539
----------------------------------------------------------------------
o Remove extraneous HAS_CONFIGURE introduced by nectar in rev1.19
o Use the PATCH_* framework and grab 012.patch from the authors.
Upgrade p5-SNMP to 1.8.1 and mark it broken as it is incompatible
with the ucd-snmp changes. If this is a huge problem, then a
repo copy of ucd-snmp can be made and we can have the hassle of
keeping two copies in our tree.
3.6.1 tickles a linker bug in the a.out ld:
``ld: internal error: RRS relocs exceed allocation 343''
* Remove the patches.aout directory since it was forgotten during
upgrade 3.5.3->3.6.1 by <kuriyama@FreeBSD.ORG>
* Remove most of the crud from the main Makefile that was there
for the a.out build
I'm going to add a tag for the 3.5.3 (1999/04/06 really) port in
case someone must build ucd-snmp for a 2.2.x FreeBSD
Runs fine on stable, on SMP current I still get race
conditions very easily, so that tkined says "no Response"
on snmp queries, and in the meantime it fills your /var
partition with the /var/log/snmp.log file:
1998-04-06 22:35:16 UCD-SNMP version 3.3.1
Opening port(s): 161
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
system.sysORLastChange.0 -- OID Doesn't exist
klookup(3, 0xefbfbf60, 96): klread: Bad address
TCP_Count_Connections - inpcb: Bad address
klookup(3, 0xefbfbf60, 96): klread: Bad address
TCP_Count_Connections - inpcb: Bad address
klookup(3, 0xefbfbf60, 96): klread: Bad address
TCP_Count_Connections - inpcb: Bad address
klookup(3, 0xefbfbf60, 96): klread: Bad address
TCP_Count_Connections - inpcb: Bad address
and so on.
This might be -current and/or SMP related.
John Hay says it runs ok on his -current.
I'll committ it anyay¸ because it runs on -STABLE
and that counts.
BTW: added an @unexec in PLIST ...
patches a bit (author incorporated a fair number of them, and made
even more cross-platfrm portability improvements). I'll deal with
any packaging changes tomorrow.
