Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT sources DOT redhat DOT com Delivered-To: mailing list cygwin AT sources DOT redhat DOT com Date: Sun, 21 Oct 2001 00:39:13 -0400 From: Christopher Faylor To: cygwin AT cygwin DOT com Cc: bgat AT billgatliff DOT com Subject: Re: [1.3.3] breaks serial i/o? Message-ID: <20011021003913.A29884@redhat.com> Reply-To: cygwin AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com, bgat AT billgatliff DOT com References: <20011018161003 DOT A3059 AT saturn DOT billgatliff DOT com> <20011018222406 DOT C11830 AT redhat DOT com> <20011019085618 DOT A5013 AT saturn DOT billgatliff DOT com> <20011019114712 DOT A23101 AT visi DOT com> <028e01c1594f$47cf7030$0200a8c0 AT lifelesswks> <20011020150809 DOT A3610 AT saturn DOT billgatliff DOT com> <20011020190254 DOT A27597 AT redhat DOT com> <20011020203703 DOT A4217 AT saturn DOT billgatliff DOT com> <20011020223934 DOT A28028 AT redhat DOT com> <20011021003044 DOT A29855 AT redhat DOT com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20011021003044.A29855@redhat.com> User-Agent: Mutt/1.3.21i On Sun, Oct 21, 2001 at 12:30:44AM -0400, Christopher Faylor wrote: >On Sat, Oct 20, 2001 at 10:39:34PM -0400, Christopher Faylor wrote: >>However, before we reach this drastic step, I'd *really* suggest at >>least trying the 5.1 branch of gdb. Your errors sound very similar >>to others that were reported with ARM. It is very possible that >>they were fixed subsequent to 5.0. > >I have just spelunked through the cvs logs in the rdi-share directory >and it looks like most relevant changes to gdb for arm/rdi did make >it into gdb 5.0 so this probably is not going to hold out much promise. > >>From looking at gdb's code it seems like select is returning a TRUE >for something to select on the serial line but when it is time to >read something, apparently nothing is available. > >I guess I really will have to make a cable and connect my linux and >windows systems to test this... One suggestion for the gdb code -- add an '| O_BINARY' to all of the opens of the serial line in unixcomm.c. I can see a pathological condition in cygwin where a VMIN setting will not be properly honored if there is a readahead condition due to a '\n' -> '\r\n' conversion. This could cause the code to block but even if it wasn't blocking, I assume that you probably don't want to do newline conversions. This probably isn't going to fix the problem since the default is for cygwin to not do this kind of conversion on devices but it's best to be safe anyway. cgf -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/