[422] in arla-drinkers
Arla 0.19 on Linux 2.1.131 core dump
daemon@ATHENA.MIT.EDU (Brandon S. Allbery)
Tue Dec 22 10:49:54 1998
From owner-arla-drinkers@stacken.kth.se Tue Dec 22 15:49:53 1998
Return-Path: <owner-arla-drinkers@stacken.kth.se>
Delivered-To: arla-drinkers-mtg@bloom-picayune.mit.edu
Received: (qmail 22012 invoked from network); 22 Dec 1998 15:49:53 -0000
Received: from unknown (HELO sundance.stacken.kth.se) (130.237.234.41)
by bloom-picayune.mit.edu with SMTP; 22 Dec 1998 15:49:53 -0000
Received: (from majordom@localhost)
by sundance.stacken.kth.se (8.8.8/8.8.8) id QAA11527
for arla-drinkers-list; Tue, 22 Dec 1998 16:41:43 +0100 (MET)
Received: (from root@localhost)
by sundance.stacken.kth.se (8.8.8/8.8.8) id QAA11522
for arla-drinkers@stacken.kth.se; Tue, 22 Dec 1998 16:41:24 +0100 (MET)
Received: (from root@localhost)
by sundance.stacken.kth.se (8.8.8/8.8.8) id QAA11473
for arla-drinkers@stacken.kth.se; Tue, 22 Dec 1998 16:38:45 +0100 (MET)
Message-Id: <199812221538.QAA11473@sundance.stacken.kth.se>
X-Mailer: exmh version 2.0.2 2/24/98
To: arla-drinkers@stacken.kth.se
Subject: Arla 0.19 on Linux 2.1.131 core dump
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Date: Tue, 22 Dec 1998 10:29:13 -0500
From: "Brandon S. Allbery" <allbery@hilfy.ece.cmu.edu>
Sender: owner-arla-drinkers@stacken.kth.se
Precedence: bulk
It's abort()ing in find_first_fs(), looking for the volume ID of
root.sup.readonly in the volume structure for root.cell. This makes me
think that it either inappropriately followed the mountpoint for root.sup
(/afs/ece.cmu.edu/support) or failed to follow it to select the correct
volume's cache structure.
This also happened with the 0.17.1 release; it's only just now that I've
had an opportunity to try to track down the arlad cores we've been having
on 2.1 kernels.
BTW, here's a minor diff to xfs/linux/getcwd.c so things don't break if you
reboot to a 2.0 kernel. The folks who are experimenting with 2.1 around
here jump around a lot, and it's moderately annoying to see the system
break if you forget to zap the LD_PRELOAD.
--- /var/tmp/arla-0.19/xfs/linux/getcwd.c Tue Dec 22 10:22:35 1998
+++ getcwd.c Mon Dec 21 09:32:42 1998
@@ -34,6 +34,9 @@
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
+ *
+ * 1998-12-21 allbery@ece.cmu.edu
+ * If we fail with ENOSYS, fall back on the old getcwd.
*/
#ifdef HAVE_CONFIG_H
@@ -58,6 +61,9 @@
retval = sys_getcwd(buf, size);
if (retval >= 0)
return buf;
- errno = -retval;
- return NULL;
+ if (retval != -ENOSYS) {
+ errno = -retval;
+ return NULL;
+ }
+ return __getcwd(buf, size);
}
--
brandon s. allbery [os/2][linux][solaris][japh] allbery@kf8nh.apk.net
system administrator [WAY too many hats] allbery@ece.cmu.edu
electrical and computer engineering KF8NH
carnegie mellon university ["God, root, what is difference?" -Pitr]
