Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin-developers AT sources DOT redhat DOT com Date: Fri, 30 Mar 2001 10:45:32 -0500 From: Christopher Faylor To: cygwin-developers AT cygwin DOT com Subject: Re: clearerr Message-ID: <20010330104532.D12718@redhat.com> Reply-To: cygwin-developers AT cygwin DOT com Mail-Followup-To: cygwin-developers AT cygwin DOT com References: <02ff01c0b916$66480970$0200a8c0 AT lifelesswks> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.11i In-Reply-To: <02ff01c0b916$66480970$0200a8c0@lifelesswks>; from robert.collins@itdomain.com.au on Fri, Mar 30, 2001 at 09:33:49PM +1000 On Fri, Mar 30, 2001 at 09:33:49PM +1000, Robert Collins wrote: >I'm confused - newlib's clearerr calls __sclearerr, but I can't find >that anywhere. cygwin.dindefines clearerr and _clearerr, but I can't >find them in the source.. > >In testing the fifo semantics, I've found that the expected behaviour >for a fifo on freeBSD is that clearerr() should re-open the fifo, >allowing further writes to take take place.. > >i.e. > >a reader can loop >clearerr() >read to eof > >and the pipe can have multiple separate writers write to the pipe and >exit. Without clearerr or a rewind(), the pipe gets marked eof and stays >eof. > >most OS's seem to simply use rewind() to the same effect, so I'm >implementing that. The question I have is: will clearerr() actually call >any cygwin code? Why are you asking a question like this which is so easily verified by a simple test case? gdb is your friend. cgf