Re: [RFA] remote debugging patches

[Michal Ludvig <mludvig@suse.cz>]

Andreas Schwab wrote:
> Maybe it's just me, but I hate arbitrary limits.  There is no need for
> sprintf at this point anyway.  I'd suggest to define another variable that
> either points to name or to the result of inet_ntoa and use that in the
> fprintf call.  No need for namebuf either.

OK, the 100-chars limit is removed and the patch is getting smaller and 
smaller :-)) Do you like it more this way, Andreas?
Index: gdbserver/remote-utils.c
===================================================================
RCS file: /cvs/src/src/gdb/gdbserver/remote-utils.c,v
retrieving revision 1.8
diff -c -3 -p -B -b -r1.8 remote-utils.c
*** remote-utils.c      2002/02/14 06:21:22     1.8
--- remote-utils.c      2002/03/13 19:05:33
*************** remote_open (char *name)
*** 99,105 ****
         }
   #endif

!
       }
     else
       {
--- 99,105 ----
         }
   #endif

!       fprintf (stderr, "Remote debugging using %s\n", name);
       }
     else
       {
*************** remote_open (char *name)
*** 154,159 ****
--- 154,163 ----

         signal (SIGPIPE, SIG_IGN);      /* If we don't do this, then 
gdbserver simply
                                            exits when the remote side 
dies.  */
+
+       /* Convert IP address to string.  */
+       fprintf (stderr, "Remote debugging from host %s\n",
+          inet_ntoa (sockaddr.sin_addr));
       }

   #if defined(F_SETFL) && defined (FASYNC)
*************** remote_open (char *name)
*** 164,170 ****
   #endif
   #endif
     disable_async_io ();
-   fprintf (stderr, "Remote debugging using %s\n", name);
   }

   void
--- 168,173 ----