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 content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Subject: RE: found possible suspect for characters out of order bug X-MimeOLE: Produced By Microsoft Exchange V6.0.4417.0 Date: Tue, 11 Sep 2001 12:43:58 +1000 Message-ID: X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: found possible suspect for characters out of order bug Thread-Index: AcE6YmTr/q/IqeewSwSjtEa3ghbkwAAB6PRg From: "Robert Collins" To: Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id WAA18054 > -----Original Message----- > From: Christopher Faylor [mailto:cgf AT redhat DOT com] > Sent: Tuesday, September 11, 2001 11:51 AM > To: cygwin-developers AT cygwin DOT com > Subject: Re: found possible suspect for characters out of order bug > > > On Tue, Sep 11, 2001 at 10:39:57AM +1000, Robert Collins wrote: > >Remember the bug that keeps getting odd reports? Well I've got it > >happening regularly to me at home. I couldn't discuss it > last night - no > >internet connection :[. > > > >What I think is happening is that the fhander_termios > edit_line method > >is getting called as a signal. If any out of order queueing > occurs with > >signals, that would cause the percieved symptoms. > > I don't know what you mean by "is getting called as a signal". Do you > mean it is being called in a signal handler? > > If so, you'll have to expand on a scenario during which that > could happen. > I can't imagine one. Uhmm, this is a new area of cygwin for me. And the code window with my cursor at the right place is not available right now... (no net @home still). I just followed my nose to get here :}. I just went to webcvs and nosed around to find what I saw at home... in int fhandler_console::read (void *pv, size_t buflen) there is code that look like it handles being called as a signal handler. I _think_ that this is what I was looking at. The behaviour from a user point of view is that strings of a few characters (I've seen up three and IIRC 4) in length get muddled. I'm not sure if they get reversed or just muddled :[. What I'm trying to figure out is whether the problem is in windows or cygwin. Reentrancy or concurrency seems an easy explanation, but so would windows not keeping order in queues of some sort.. Rob