I think we're good to go for tomorrow.
Alan Buxey ran into an issue with an infinite loop when a home server was marked down. I've just tracked it down and committed a fix. I've also run the server through valgrind, and corrected one or two other memory leaks on exit. They're not critical, but it's useful to fix. Alan DeKok.
Alan DeKok wrote:
Alan Buxey ran into an issue with an infinite loop when a home server was marked down. I've just tracked it down and committed a fix.
I've also run the server through valgrind, and corrected one or two other memory leaks on exit. They're not critical, but it's useful to fix.
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/devel.html
I've been running recent CVS head for around a week on the production servers and found no problems. CPU usage idling around 6% with 5.16MB private memory usage. No problems with EAP authentication, and no strange zombie thread issues. All looks very shiny. One thing, in the man pages for unlang it might be worth including a small example for using attribute manipulation operators like -=, as it isn't explicitly clear. Thanks, Arran -- Arran Cudbard-Bell (A.Cudbard-Bell@sussex.ac.uk) Authentication, Authorisation and Accounting Officer Infrastructure Services | ENG1 E1-1-08 University Of Sussex, Brighton EXT:01273 873900 | INT: 3900
Hi, the fix for hints file is not processing packets when they contain no User-Name is still not in CVS (see bug #335). Some time ago I was promised it gets into -pre2, but that appears to have slipped though... Stefan -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
Oh. Assuming that update request { ... } in unlang works, the issue is obsolete. Sorry for the noise. I guess I need a break. Stefan -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
On Wed, Jan 09, 2008 at 05:07:58PM +0100, Alan DeKok wrote:
Alan Buxey ran into an issue with an infinite loop when a home server was marked down. I've just tracked it down and committed a fix.
I've also run the server through valgrind, and corrected one or two other memory leaks on exit. They're not critical, but it's useful to fix.
Did you get a chance to look at the fix for bug #458/#459 (rlm_ldap/compare_check_items)? I've updated the patch for the current CVS HEAD, and attached it to bug 459. Thanks, -- Chris Mikkelson | Problems are posed by fools like me; cmikk@qwest.net | But only Heuristics can search a tree.
Chris Mikkelson wrote:
Did you get a chance to look at the fix for bug #458/#459 (rlm_ldap/compare_check_items)? I've updated the patch for the current CVS HEAD, and attached it to bug 459.
Thanks. Given the relatively few comments about the feature, I think almost no one uses it. I took a look at the patches a while ago, but there was a fair amount of code change, and the LDAP module is already huge and complicated. I think it can wait for 2.0.1. Alan DeKok.
Alan DeKok wrote:
Alan Buxey ran into an issue with an infinite loop when a home server was marked down. I've just tracked it down and committed a fix.
I've also run the server through valgrind, and corrected one or two other memory leaks on exit. They're not critical, but it's useful to fix.
I've rebuild a package from a 5-minutes-ago CVS checkout and it builds, installs and runs out of the box on Debian. Although I've done only very basic tests (sorry) I don't see any reason to wait longer. From my side everything looks ok. -- Nicolas Baradakis
Alan DeKok wrote:
Alan Buxey ran into an issue with an infinite loop when a home server was marked down. I've just tracked it down and committed a fix.
I've also run the server through valgrind, and corrected one or two other memory leaks on exit. They're not critical, but it's useful to fix.
Alan DeKok. - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/devel.html
ssing-prototypes -Wmissing-declarations -Wnested-externs -W -Wredundant-decls -Wundef -I/usr/local/src/freeradius-cvscurrent/src -I/usr/local/src/freeradius-cvscurrent/libltdl -c rlm_acctlog.c -o rlm_acctlog.o >/dev/null 2>&1 /usr/local/src/freeradius-cvscurrent/libtool --mode=link gcc -release 2.0.0-beta \ -module -export-dynamic -o rlm_acctlog.la \ -rpath /usr/local/freeradius-cvs100108/lib rlm_acctlog.lo rlm_acctlog.c /usr/local/src/freeradius-cvscurrent/src/lib/libfreeradius-radius.la -framework DirectoryService -lresolv -lpthread gcc ${wl}-flat_namespace ${wl}-undefined ${wl}suppress -o .libs/rlm_acctlog-2.0.0-beta.so -bundle .libs/rlm_acctlog.o /usr/local/src/freeradius-cvscurrent/src/lib/.libs/libfreeradius-radius.dylib -lresolv -lpthread -framework DirectoryService (cd .libs && rm -f rlm_acctlog.so && ln -s rlm_acctlog-2.0.0-beta.so rlm_acctlog.so) ar cru .libs/rlm_acctlog.a rlm_acctlog.o ranlib .libs/rlm_acctlog.a creating rlm_acctlog.la (cd .libs && rm -f rlm_acctlog.la && ln -s ../rlm_acctlog.la rlm_acctlog.la) usage: cp [-R [-H | -L | -P]] [-f | -i | -n] [-pv] src target cp [-R [-H | -L | -P]] [-f | -i | -n] [-pv] src1 ... srcN directory make[6]: *** [build-module] Error 64 make[5]: *** [common] Error 2 make[4]: *** [all] Error 2 make[3]: *** [common] Error 2 make[2]: *** [all] Error 2 make[1]: *** [common] Error 2 make: *** [all] Error 2 MACs use BSD Style cp command, if that makes a difference. Also I see you've cleaned up the code to get rid of build time warnings . But theres still issues in misc.c /usr/local/src/freeradius-cvscurrent/libtool --mode=compile gcc -g -O2 -D_REENTRANT -D_POSIX_PTHREAD_SEMANTICS -DDARWIN -Wall -D_GNU_SOURCE -g -Wshadow -Wpointer-arith -Wcast-qual -Wcast-align -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wnested-externs -W -Wredundant-decls -Wundef -D_LIBRADIUS -I/usr/local/src/freeradius-cvscurrent/src -c misc.c gcc -g -O2 -D_REENTRANT -D_POSIX_PTHREAD_SEMANTICS -DDARWIN -Wall -D_GNU_SOURCE -g -Wshadow -Wpointer-arith -Wcast-qual -Wcast-align -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wnested-externs -W -Wredundant-decls -Wundef -D_LIBRADIUS -I/usr/local/src/freeradius-cvscurrent/src -c misc.c -fno-common -DPIC -o .libs/misc.o misc.c: In function 'rad_lockfd': misc.c:60: warning: unused parameter 'lock_len' misc.c: In function 'rad_lockfd_nonblock': misc.c:83: warning: unused parameter 'lock_len' misc.c: In function 'rad_unlockfd': misc.c:106: warning: unused parameter 'lock_len'
I was about to report the same rlm_acct error, I'm on Suse 10.3. -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
Stefan Winter wrote:
I was about to report the same rlm_acct error, I'm on Suse 10.3.
Do you have space characters in the path of your source tree? Maybe the following patch could help. Index: src/modules/rules.mak =================================================================== RCS file: /source/radiusd/src/modules/rules.mak,v retrieving revision 1.67 diff -u -r1.67 rules.mak --- src/modules/rules.mak 25 Nov 2007 14:20:02 -0000 1.67 +++ src/modules/rules.mak 10 Jan 2008 10:01:09 -0000 @@ -108,7 +108,7 @@ build-module: $(TARGET).la $(RLM_UTILS) @[ "x$(RLM_SUBDIRS)" = "x" ] || $(MAKE) $(MFLAGS) WHAT_TO_MAKE=all common - @cp $< $(top_builddir)/src/modules/lib + @cp $< "$(top_builddir)/src/modules/lib" $(TARGET).la: $(LT_OBJS) $(LIBTOOL) --mode=link $(CC) -release $(RADIUSD_VERSION) \ -- Nicolas Baradakis
On Thu 10 Jan 2008, Stefan Winter wrote:
I was about to report the same rlm_acct error, I'm on Suse 10.3.
I dont see that error in last nights snapshot on any SUSE version... -- Peter Nixon http://peternixon.net/
Arran Cudbard-Bell wrote: ...
usage: cp [-R [-H | -L | -P]] [-f | -i | -n] [-pv] src target cp [-R [-H | -L | -P]] [-f | -i | -n] [-pv] src1 ... srcN directory
That change *just* went in, about 10 minutes before you sent the email.
MACs use BSD Style cp command, if that makes a difference.
Nope. The issue is that the target directory doesn't exist. I've added a line creating it. It works now on the Macs I have access to.
Also I see you've cleaned up the code to get rid of build time warnings . But theres still issues in misc.c
Fixed, thanks. Alan DeKok.
Nope. The issue is that the target directory doesn't exist. I've added a line creating it. It works now on the Macs I have access to.
Works for me now, too. Stefan -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
participants (6)
-
Alan DeKok -
Arran Cudbard-Bell -
Chris Mikkelson -
Nicolas Baradakis -
Peter Nixon -
Stefan Winter