[Refdb-devel] [ refdb-Bugs-1599713 ] checkref: *** glibc detected *** double free or corruption

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 22:49
Message generated for change (Comment added) made by davidnebauer
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-21 03:12

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Comment added) made by mhoenicka
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open

Resolution: Works For Me

Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)

Date: 2006-11-20 22:54

Message:
Logged In: YES
user_id=85809
Originator: NO

I've checked my setup on FreeBSD using valgrind to detect allocated memory
problems. There were no errors reported, except a number of memory leaks
related to the checkref command. I've plugged these leaks, although they
most likely cannot not cause the problem that you reported. In any case
you may want to check out svn revision 238 and try again. If this does not
fix the problem, please send me your test data (both the datasets in your
database and in the file that you want to check) off-list. It could just
so happen that my data miss a bordercase that triggers the problem.

----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-20 18:42

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Comment added) made by dreusser
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open
Resolution: Works For Me
Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: Dominik (dreusser)

Date: 2006-11-22 14:47

Message:
Logged In: YES
user_id=1572139
Originator: YES

When reporting the bug, I've been using the refdb 0.9.8-pre8 (as shown in
refdb-bug.txt). I upgraded today to svn revision 238 (dpkg -i
refdb-server_0.0-svn-20061121_i386.deb
refdb-0.0-svn_20061121.orig.tar.gz). Is it possible, that refdba:viewstat
still reports 0.9.8-pre8 after upgrading? If so, then the error persists
with the latest svn.
With respect to the debian system: I'm also using an up-to-date
debian/testing box except for 9 packages which are held back (apache2
apache2-mpm-prefork apache2-utils gnupg libapache-mod-php4
libapache2-mod-perl2 libapache2-mod-php4 php4-common php4-mysql).

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-20 22:54

Message:
Logged In: YES
user_id=85809
Originator: NO

I've checked my setup on FreeBSD using valgrind to detect allocated memory
problems. There were no errors reported, except a number of memory leaks
related to the checkref command. I've plugged these leaks, although they
most likely cannot not cause the problem that you reported. In any case
you may want to check out svn revision 238 and try again. If this does not
fix the problem, please send me your test data (both the datasets in your
database and in the file that you want to check) off-list. It could just
so happen that my data miss a bordercase that triggers the problem.

----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-20 18:42

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Comment added) made by mhoenicka
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open
Resolution: Works For Me
Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)

Date: 2006-11-22 14:57

Message:
Logged In: YES
user_id=85809
Originator: NO

It is correct that the current svn version still claims to be pre8. I
usually bump up the version numbers only just before a new (pre)release. I
think I should add the svn revision number to the viewstat output as well.
As for the bug itself, please provide your database (as a ris or risx
dump) and the ris file that you want to check. As valgrind does not detect
any incorrect access to allocated memory with my data, I'll have to try
with your data in order to reproduce the problem.

----------------------------------------------------------------------

Comment By: Dominik (dreusser)
Date: 2006-11-22 14:47

Message:
Logged In: YES
user_id=1572139
Originator: YES

When reporting the bug, I've been using the refdb 0.9.8-pre8 (as shown in
refdb-bug.txt). I upgraded today to svn revision 238 (dpkg -i
refdb-server_0.0-svn-20061121_i386.deb
refdb-0.0-svn_20061121.orig.tar.gz). Is it possible, that refdba:viewstat
still reports 0.9.8-pre8 after upgrading? If so, then the error persists
with the latest svn.
With respect to the debian system: I'm also using an up-to-date
debian/testing box except for 9 packages which are held back (apache2
apache2-mpm-prefork apache2-utils gnupg libapache-mod-php4
libapache2-mod-perl2 libapache2-mod-php4 php4-common php4-mysql).

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-20 22:54

Message:
Logged In: YES
user_id=85809
Originator: NO

I've checked my setup on FreeBSD using valgrind to detect allocated memory
problems. There were no errors reported, except a number of memory leaks
related to the checkref command. I've plugged these leaks, although they
most likely cannot not cause the problem that you reported. In any case
you may want to check out svn revision 238 and try again. If this does not
fix the problem, please send me your test data (both the datasets in your
database and in the file that you want to check) off-list. It could just
so happen that my data miss a bordercase that triggers the problem.

----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-20 18:42

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Comment added) made by mhoenicka
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None
Status: Open

Resolution: Fixed

Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)

Date: 2006-11-22 22:17

Message:
Logged In: YES
user_id=85809
Originator: NO

Turns out the problem was an insufficiently sized buffer in dbfncs.c. The
problem does not depend on particular input data, but it occurs only if
you use MySQL as the database engine. I've fixed this bug in subversion
revision 241. A one-line patch is also attached to this bug report.

Please let me know whether revision 241 runs ok on Debian with MySQL.

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-22 14:57

Message:
Logged In: YES
user_id=85809
Originator: NO

It is correct that the current svn version still claims to be pre8. I
usually bump up the version numbers only just before a new (pre)release. I
think I should add the svn revision number to the viewstat output as well.
As for the bug itself, please provide your database (as a ris or risx
dump) and the ris file that you want to check. As valgrind does not detect
any incorrect access to allocated memory with my data, I'll have to try
with your data in order to reproduce the problem.

----------------------------------------------------------------------

Comment By: Dominik (dreusser)
Date: 2006-11-22 14:47

Message:
Logged In: YES
user_id=1572139
Originator: YES

When reporting the bug, I've been using the refdb 0.9.8-pre8 (as shown in
refdb-bug.txt). I upgraded today to svn revision 238 (dpkg -i
refdb-server_0.0-svn-20061121_i386.deb
refdb-0.0-svn_20061121.orig.tar.gz). Is it possible, that refdba:viewstat
still reports 0.9.8-pre8 after upgrading? If so, then the error persists
with the latest svn.
With respect to the debian system: I'm also using an up-to-date
debian/testing box except for 9 packages which are held back (apache2
apache2-mpm-prefork apache2-utils gnupg libapache-mod-php4
libapache2-mod-perl2 libapache2-mod-php4 php4-common php4-mysql).

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-20 22:54

Message:
Logged In: YES
user_id=85809
Originator: NO

I've checked my setup on FreeBSD using valgrind to detect allocated memory
problems. There were no errors reported, except a number of memory leaks
related to the checkref command. I've plugged these leaks, although they
most likely cannot not cause the problem that you reported. In any case
you may want to check out svn revision 238 and try again. If this does not
fix the problem, please send me your test data (both the datasets in your
database and in the file that you want to check) off-list. It could just
so happen that my data miss a bordercase that triggers the problem.

----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-20 18:42

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

[SourceForge.net]

Bugs item #1599713, was opened at 2006-11-20 14:19
Message generated for change (Comment added) made by dreusser
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: refdbd
Group: None

Status: Closed

Resolution: Fixed
Priority: 5
Private: No
Submitted By: Dominik (dreusser)
Assigned to: Markus Hoenicka (mhoenicka)
Summary: checkref: *** glibc detected *** double free or corruption

Initial Comment:
On my debian box I get the following error while running checkref (I'm not sure whether this really is a problem in refdbd or some debian-glibc bug):

client:
refdbc: checkref test.ris
could not write to refdbd. Stop

server:
CREATE TEMPORARY TABLE t_temp_xlink (xlink_id BIGINT NOT NULL AUTO_INCREMENT,link_id BIGINT NOT NULL,xref_id BIGINT NOT NULL,xlink_type ENUM("URL","PDF","FULLTEXT","RELATED", "IMAGE"),xlink_source ENUM("REFERENCE","NOTE"),PRIMARY KEY (xlink_id),KEY (link_id),KEY (xref_id))
BEGIN
LOCK TABLES t_temp_refdb WRITE, t_temp_author WRITE, t_temp_keyword WRITE, t_temp_periodical WRITE, t_temp_user WRITE, t_temp_xauthor WRITE, t_temp_xkeyword WRITE, t_temp_xuser WRITE, t_temp_note WRITE, t_temp_xnote WRITE, t_temp_link WRITE, t_temp_xlink WRITE, refdb.t_journal_words WRITE
*** glibc detected *** double free or corruption (!prev): 0x080ff5a0 ***
child exited with code 0
server waiting n_max_fd=4



----------------------------------------------------------------------

Comment By: Dominik (dreusser)

Date: 2006-11-24 13:35

Message:
Logged In: YES
user_id=1572139
Originator: YES

works with revision 241.
Thanks

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-22 22:17

Message:
Logged In: YES
user_id=85809
Originator: NO

Turns out the problem was an insufficiently sized buffer in dbfncs.c. The
problem does not depend on particular input data, but it occurs only if
you use MySQL as the database engine. I've fixed this bug in subversion
revision 241. A one-line patch is also attached to this bug report.

Please let me know whether revision 241 runs ok on Debian with MySQL.

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-22 14:57

Message:
Logged In: YES
user_id=85809
Originator: NO

It is correct that the current svn version still claims to be pre8. I
usually bump up the version numbers only just before a new (pre)release. I
think I should add the svn revision number to the viewstat output as well.
As for the bug itself, please provide your database (as a ris or risx
dump) and the ris file that you want to check. As valgrind does not detect
any incorrect access to allocated memory with my data, I'll have to try
with your data in order to reproduce the problem.

----------------------------------------------------------------------

Comment By: Dominik (dreusser)
Date: 2006-11-22 14:47

Message:
Logged In: YES
user_id=1572139
Originator: YES

When reporting the bug, I've been using the refdb 0.9.8-pre8 (as shown in
refdb-bug.txt). I upgraded today to svn revision 238 (dpkg -i
refdb-server_0.0-svn-20061121_i386.deb
refdb-0.0-svn_20061121.orig.tar.gz). Is it possible, that refdba:viewstat
still reports 0.9.8-pre8 after upgrading? If so, then the error persists
with the latest svn.
With respect to the debian system: I'm also using an up-to-date
debian/testing box except for 9 packages which are held back (apache2
apache2-mpm-prefork apache2-utils gnupg libapache-mod-php4
libapache2-mod-perl2 libapache2-mod-php4 php4-common php4-mysql).

----------------------------------------------------------------------

Comment By: Markus Hoenicka (mhoenicka)
Date: 2006-11-20 22:54

Message:
Logged In: YES
user_id=85809
Originator: NO

I've checked my setup on FreeBSD using valgrind to detect allocated memory
problems. There were no errors reported, except a number of memory leaks
related to the checkref command. I've plugged these leaks, although they
most likely cannot not cause the problem that you reported. In any case
you may want to check out svn revision 238 and try again. If this does not
fix the problem, please send me your test data (both the datasets in your
database and in the file that you want to check) off-list. It could just
so happen that my data miss a bordercase that triggers the problem.

----------------------------------------------------------------------

Comment By: David Nebauer (davidnebauer)
Date: 2006-11-20 18:42

Message:
Logged In: YES
user_id=1080150
Originator: NO

Using refdb svn revision 224 on my up-to-date debian/testing box I find
the checkref command works perfectly and does not generate the error you
describe. Perhaps you could state the version of refdb you used and the
flavour of debian you are running it on.

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=385991&aid=1599713&group_id=26091