Libsmi May 2005

libsmi@ibr.cs.tu-bs.de

3 participants
5 discussions

Failing build from subversion repository
by Keith Dart 13 Dec '05

13 Dec '05

I obtained a subversion copy of libsmi, and try to build it (on two machines), I get the following: gcc -DHAVE_CONFIG_H -I. -I. -I.. -g -O2 -MT parser-smi.lo -MD -MP -MF .deps/parser-smi.Tpo -c parser-smi.c -o parser-smi.o >/dev/null 2>&1 make[2]: *** No rule to make target `parser-smi.tab.h', needed by `scanner-smi.c'. Stop. make[2]: Leaving directory `/u01/home/kdart/src/libsmi/trunk/lib' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/u01/home/kdart/src/libsmi/trunk' make: *** [all] Error 2 But I have flex and bison: 403 !$ bison --version bison (GNU Bison) 1.875 404 $ flex --version flex version 2.5.4 It looks like the Makefile does not include a rule for this, after all. -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Keith Dart <kdart(a)kdart.com> public key: ID: F3D288E4 =====================================================================

2 1

Detailed change log
by Arkadiusz Gabriel 06 May '05

06 May '05

Hi, All! I have looked through the libsmi website and the svn repository, but I have found only rough change log. Is there a detailed change log for smilib? Especially for bug fixes? Regards Arek ps. Sorry if it's not the right place to ask, but seems to be the only one...

2 1

smidump/mosy problem
by Allen McIntosh 04 May '05

04 May '05

Happens with latest source. Input: FOO DEFINITIONS ::= BEGIN IMPORTS MODULE-IDENTITY FROM SNMPv2-SMI; fooMIB MODULE-IDENTITY LAST-UPDATED "200006270000Z" ORGANIZATION "Carrots Inc." CONTACT-INFO "Bugs Bunny" DESCRIPTION "Testing" REVISION "200006270000Z" DESCRIPTION "1 2 3" ::= {iso(1) member-body(2) us(840) 802dot3(10006) snmpmibs(300) 43 } fooObjects OBJECT IDENTIFIER ::= { fooMIB 1 } Output (smidump --format=mosy FOO.my): fooMIB snmpmibs.43 %n0 fooMIB module-identity fooObjects fooMIB.1 %n0 fooObjects object-id The ancestry of snmpmibs is clearly stated in the input, but does not appear in the output. I have access to a mosy descendent (still called mosy) that outputs fooMIB 1.2.840.802.10006.300.43 (Rest of output is the same) I don't think that smidump should necessarily produce identical output, but it ought to be possible to re-create the object ID.

2 1

Unsigned64
by Allen McIntosh 03 May '05

03 May '05

"the SPPI construct/keyword `Unsigned64' may not be used in a MIB" Why? When did SMI disallow it? I agree that it might not be wise, but making this a fatal error in smidump --format=mosy CISCO-CLASS-BASED-QOS-MIB.my seems a little extreme.

2 1

Libsmi 0.4.3 memory fault (Linux/gcc)
by Allen McIntosh 03 May '05

03 May '05

The libsmi parser memory faults when parsing mibs/ietf/DOT12-RPTR-MIB. Repeat by running smilint mibs/ietf/DOT12-RPTR-MIB after installing libsmi. This happened to me under RedHat 9 (gcc 3.2.2/x86) and RedHat 7.3 (gcc 2.96/x86). The parser usually runs off the rails thus: $ gdb /usr/local/bin/smilint [stuff omitted] (gdb) run mibs/ietf/DOT12-RPTR-MIB Starting program: /usr/local/bin/smilint mibs/ietf/DOT12-RPTR-MIB mibs/ietf/DOT12-RPTR-MIB:36: revision for last update is missing mibs/ietf/DOT12-RPTR-MIB:151: warning: use Integer32 instead of INTEGER in SMIv2 Program received signal SIGSEGV, Segmentation fault. 0x40037b15 in checkObjects (parserPtr=0xbffff5f0, modulePtr=0x80559a0) at parser-smi.y:220 220 if (objectPtr->nodePtr->parentPtr && (gdb) where #0 0x40037b15 in checkObjects (parserPtr=0xbffff5f0, modulePtr=0x80559a0) at parser-smi.y:220 #1 0x4003a26c in smiparse (parserPtr=0xbffff5f0) at parser-smi.y:1641 #2 0x40031021 in loadModule ( modulename=0xbffff841 "mibs/ietf/DOT12-RPTR-MIB", parserPtr=0x0) at data.c:3792 #3 0x40035110 in smiLoadModule (module=0xbffff841 "mibs/ietf/DOT12-RPTR-MIB") at smi.c:416 #4 0x080492f2 in main (argc=2, argv=0xbffff6f4) at smilint.c:314 #5 0x42017589 in __libc_start_main () from /lib/i686/libc.so.6 (gdb) print objectPtr->nodePtr->parentPtr $1 = (struct Node *) 0x62696d2f Running with ElectricFence and EF_PROTECT_BELOW on turns up a problem, if that helps. I can provide details if that would help.

2 2

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

Libsmi May 2005