Re: [nmap-svn] r33574 - nmap-exp/d33tah/nmap-nsock-scan

[Fyodor <fyodor () nmap org>]

On Mon, Aug 18, 2014 at 10:15 AM, <commit-mailer () nmap org> wrote:

> Author: d33tah
> Date: Mon Aug 18 17:15:01 2014
> New Revision: 33574
>
> Log:
> I have no idea why this comment was there.
>
> Modified:
>    nmap-exp/d33tah/nmap-nsock-scan/nsock_scan.cc
>
> Modified: nmap-exp/d33tah/nmap-nsock-scan/nsock_scan.cc
>
> ==============================================================================
> --- nmap-exp/d33tah/nmap-nsock-scan/nsock_scan.cc       (original)
> +++ nmap-exp/d33tah/nmap-nsock-scan/nsock_scan.cc       Mon Aug 18
> 17:15:01 2014
> @@ -718,8 +718,6 @@
>  void NsockTCPScanEngine::errnoToPortState(int connect_errno, int
> *port_state,
>                                            int *reason_id) {
>    switch (connect_errno) {
> -    /* This can happen on localhost, successful/failing connection
> -       immediately in non-blocking mode. */

   case ECONNREFUSED:
>      *port_state = PORT_CLOSED;
>      *reason_id = ER_CONREFUSED;

Just FYI, it's because some people are surprised that a nonblocking
connect() call can return connection success or refused immediately, since
we normally only get that information later (e.g. after doing a select() on
the socket).  After all, you're telling the system that you aren't willing
to wait.  But if you are doing the connect against localhost, the connect()
call can actually succeed (or fail) right there on the spot.  So we need to
handle that case here.  Maybe the comment doesn't make this clear enough
though.

Cheers,
Fyodor
_______________________________________________
Sent through the dev mailing list
http://nmap.org/mailman/listinfo/dev
Archived at http://seclists.org/nmap-dev/