On Mon, Feb 15, 2010 at 10:16:55PM -0500, Christopher Faylor wrote: > On Tue, Feb 16, 2010 at 02:20:55AM +0100, Enrico Forestieri wrote: > >On Sun, Feb 14, 2010 at 08:54:27PM -0500, Christopher Faylor wrote: > > > >> I just checked in YA in my series of attempts to get this working right. > >> It will behave marginally better now but there are still problems if you > >> attempt to write to a fifo before anything is reading it and then try to > >> do a non-blocking read. > > > >Thank you for your efforts. It indeed works better with the 2010-02-15 > >snapshot. > > > >In order to help you pinpointing another glitch, I attach here a slightly > >slightly modified version of the original test case and a shell script. > >Launching the program in a terminal and then the shell script in another > >one, it seems that the program hangs on the write() call. > >Apparently, the problem is the O_RDWR flag, as replacing it with O_WRONLY, > >everything works. The test case works with O_RDWR both in cygwin 1.5 and > >linux. > > Yes, non-blocking writes are not working. They never really were, in fact.
Please, note that the write fails when using O_RDWR instead of O_WRONLY. In the test case, O_NONBLOCK is not used. -- Enrico -- 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