* Tim Brown [19. Nov 2008]: > security_note(port:port, data:uname); > osversion = ssh_cmd(socket:sock, cmd:"uname -r"); > ... > > However, it appears that osversion etc never get populated.
Odd. Just to clarify: - Is osversion *never* populated or *not always*? - Do uname -r etc yield reasonable output on your target system? > If I reorder the script so that the Solaris checks are carried out > directly after the initial uname then it works. > > In essence, the later ssh_cmd do not appear to run correctly. Yes, I'm suspecting this too. I think it has something to do with the way ssh_func.inc handles connections in ssh_reuse_connection(). There were some changes in there which seem to cause problem with ssh_cmd acquiring the shared ssh socket, but I haven't yet been able to pinpoint the exact issue. Could you try your tests with an older version of ssh_func.inc, prior to rev 1226? I had the issue that gather-package-list.nasl would sometime correctly report the remote OS, but on the next run wouldn't report anything at all. This is probably SSH-related as well, did you observe something similar? > I did try your patch to libopenvas MIchael, but that didn't seem to > resolve it. Well, it didn't resolve it either for me. ;) But since it was obviously broken, I decided to fix it before we forget the issue. Regards, Michael -- Michael Wiegand | OpenPGP key: D7D049EC | http://www.intevation.de/ Intevation GmbH, Neuer Graben 17, 49074 Osnabrück | AG Osnabrück, HR B 18998 Geschäftsführer: Frank Koormann, Bernhard Reiter, Dr. Jan-Oliver Wagner _______________________________________________ Openvas-devel mailing list Openvas-devel@wald.intevation.org http://lists.wald.intevation.org/mailman/listinfo/openvas-devel
