Red Hat Bugzilla – Bug 20199
nmapfe dumps core when scanning a Solaris 7 box
Last modified: 2008-05-01 11:37:59 EDT
nmapfe dumps core when scanning a Solaris 7 box.
nmap produces correct results.
RPMS: nmap-2.53-5, namp-frontend-2.53-5
launch nmapfe, scan a Solaris 7 box as nmap -sS -O xxx.xxx.xxx.xxx
launch nmapfe, scan the same box as nmap -sS -R xxx.xxx.xxx.xxx
scan again as nmap -sS -O xxx.xxx.xxx.xxx
Is this happening when just using nmap or just with nmapfe? I'm just trying to
make sure so that the maintainers of nmap/nmapfe are aware of the problem.
Wait. I misread your report. Sorry. Do later versions of nmapfe work correctly
on a Solaris 7 machine? The reason I'm asking is beacuse I don't have one I
could test on.
I could not download 2.54BETA7 from www.insecure.org to test it out. The site
seems unreachable right now.
I found an alternate download. I'm putting together packages of it now. I'll let
you know where to get them once they are finished.
I have uploaded the new package for you to test. You can get them from:
and the SRPM from:
Let me know whether it works or not. Oh, sorry for the delay, had to go to a
Ok, it seems to work after some tweeking.
Details: Scanning the same Solaris box as nmap -sS -O ... (default)
Binary rpms do not go any further stating that:
OS scan requested but I cannot find nmap-os-fingerprints file. It should be in
/usr/share/nmap, ~/.nmap/ or .
After I built the rpms from the SRPM, doing a rpm -bb, and installed it
afterwards, it worked ok (no core dump).
It seems that it created a /usr/share/nmap/ directory with various files, not
owned by any package. This directory though does not belong to any rpm, and was
probably added during the rpm -bb phase.
Possible solution: include it in the spec file, so that /usr/share/nmap/* is
part of the nmap rpm.
Actually it was an error in the spec file that I didn't catch. I forgot to
include $RPM_BUILD_ROOT in the install path for $(nmapdatadir). I built another
package, located in the same place:
and the SRPM:
Please try this one out. It should work now.
Ok, that seemed to do the trick.
OK. I have it going through the QA process now. I'm closing this bug with the
resolution of "errata".
Just so that you know, an errata was released today on this.