compatibility between systems and use cases. [Henri Doreau]
There were known issues with the following cases:
* epoll and regular files
* kqueue and regular files
* WSAPoll and stdin
The --nsock-engine=<ENGINE> can still be used to bypass this
default choice.
See: http://seclists.org/nmap-dev/2012/q4/173
This is needed from the Visual C++ GUI build, not only from make, and
the GUI doesn't have a way to generate this file.
It may be possible to have a Visual Basic script to the automatic
updating of the version number, according to this article.
http://support.microsoft.com/kb/237870
All the information passed as arguments is present in an iod, and we
only call this with members of one iod. Change it to accept just an iod
as an argument.
This code was not always properly cleaning up automatically generated
sockets; for example in the case of "Connection refused". It would
delete sockets even if you gave one manually with --source; it's not
clear that's the right behavior.
Added tests for UNIX domain sockets to the ncat_test.pl
script. Added tests for:
1. listening on and connecting to UNIX STREAM socket.
2. listening on and connecting to (client with random
socket) UNIX DGRAM socket.
3. connecting to UNIX DGRAM socket with given source
socket (with "-s").
Implementation of UNIX-domain sockets functionality
for Nsock. Also some minor necessary changes to
existing Nsock functions, to work properly with
UNIX-domain sockets.
o.af is AF_UNSPEC at this point in the common case that neither -4 nor
-6 was given. The code was falling through to the "else" IPv6 case. I
think we were getting luck with this because the port number happens to
be in the same place in sockaddr_in and sockaddr_in6.
Since r29743, ncat closes connection after reading EOF. This patch first
introduces a new test to verify it, and also updates two existing tests
to prevent them from closing STDIN too early.
Make new versioned nmap.rc.in and nsis/Nmap.nsi.in, and use those to
generate unversioned nmap.rc and nsis/Nmap.nsi. The changes made to
those files were constantly causing problems with updating.
There was one case where we previously didn't fatal, in nse_dnet.cc.
Move the fatal calls out of nmap_raw_socket and into the calling scope,
with the exception of the one in nse_dnet.cc.
The problem was reported by Rob Nicholls.
http://seclists.org/nmap-dev/2012/q4/186
