From: dahms AT ifk20 DOT mach DOT uni-karlsruhe DOT de Subject: RE: "read" routine! 12 Dec 1997 19:44:39 -0800 Message-ID: <009BEAF8.A91D2440.30151.cygnus.gnu-win32@ifk20.mach.uni-karlsruhe.de> To: hasmik AT maker DOT com Cc: gnu-win32 AT cygnus DOT com, dahms AT ifk20 DOT mach DOT uni-karlsruhe DOT de Hi Hasmik, you wrote: : My application starts a Tcl shell and if I issue a Ctrl-C from : the debugger it exits the program. I have debugged this to the point : that I think the "read" routine is not behaving like the "read" on the unix : system. The "read" on NT system returns a 0 (zero) and the Tcl application : treats that as the end-of-file and exits the program. Shouldn't the "read" : routine return some sort of error (EINTR)?? I guess it would, but first the default signal handler terminates it. You should ignore SIGINT or install a handler, maybe which writes "ouch!" to fd 2 (stderr). After ^c, (my) programs by default die under Unix, too, and so does tclsh. And I'm pretty shure tclsh retries after EINTR. If (debugger == gdb), then under Unix, it doesn't pass SIGINT to the debuggee by default. Check output of "info signals" on cygwin32 and Unix... Bye, Heribert (dahms AT ifk20 DOT mach DOT uni-karlsruhe DOT de) - For help on using this list (especially unsubscribing), send a message to "gnu-win32-request AT cygnus DOT com" with one line of text: "help".