A malformed message could be returned as a result of a truncated network connection to xymond, or other sorts of interesting cases.
p = strchr(nextline, '\n'); if (p) *p = '\0';
strcpy(l, nextline);
if (p) nextline = p+1; else nextline = NULL;There does seem to be something a little odd there.
I suppose the below patch would at least keep us from writing past the end on 'l', but in a proper situation we shouldn't need something like it.
-jc
On Fri, October 2, 2015 4:58 am, Axel Beckert wrote:
Hi,
On Fri, Oct 02, 2015 at 01:32:51PM +0200, Axel Beckert wrote:
Today it finally crashed again after I deployed the debug packages. (It's seldom that I'm so happy to see a core dump. ;-)
Here's the [...] oldest one: [...] It's indeed in the WML generation code, should be this line here: https://sources.debian.net/src/xymon/4.3.21-1/xymongen/wmlgen.c/#L150
Looking at what status changed in the five minutes before the crash I found two status reports which had each line lengths of 6576 and 6897 characters, but both far shorter than the MAX_LINE_LEN of 16384 characters
Another thing I could imagine are lines with no trailing newline at the end. But then again I have no idea where they could come from nor how I should look for them.
Kind regards, Axel Beckert-- Axel Beckert <beckert at phys.ethz.ch> support: +41 44 633 26 68 IT Services Group, HPT H 6 voice: +41 44 633 41 89 Departement of Physics, ETH Zurich CH-8093 Zurich, Switzerland http://nic.phys.ethz.ch/
Xymon mailing list Xymon at xymon.com http://lists.xymon.com/mailman/listinfo/xymon