Mail Archives: cygwin/2010/11/17/23:57:54
On Nov 17 11:20, Mark Rivers wrote:
> Folks,
>
> I believe I have discovered a problem with blocking sockets on Cygwin.
> The problem is that very rarely a call to send() will never return.
>
> I am having problem when using the Channel Access TCP/IP protocol in the
> EPICS control system (http://www.aps.anl.gov/epics/).
>
> I have isolated the problem to a call to send() on the server which
> never returns. This problem happens with a large number of well-tested
> clients, running on both Linux and Windows. I can build the same EPICS
> server without Cygwin, using the MS VC++ Windows compiler and libraries
> and it does not fail. When the call to send() hangs it is sending a
> large packet (just over 8KB). The failure rate is very low, perhaps
> once in 10 million calls to send().
Wow, that's not very often. You missed all the stuff from
http://cygwin.com/problems.html, so we don't know what Cygwin release
you're talking about. If it's Cygwin 1.5, you're out of luck since we
don't support that anymore. Otherwise...
> I realize that my application is difficult to reproduce for the Cygwin
> developers. I am willing to try to write a simple socket server and
> client to try to reproduce the problem in a simple test setup.
...that would be most helpful.
> But
> before I do that I would like to first know if there is a known problem
> of this sort that has already been reported, and perhaps fixed.
No, not really. I wouldn't claim that Cygwin's socket implementation is
bug free, but this kind of very occassional hang is new to me.
Corinna
--
Corinna Vinschen Please, send mails regarding Cygwin to
Cygwin Project Co-Leader cygwin AT cygwin DOT com
Red Hat
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
- Raw text -