Mailing-List: contact cygwin-developers-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-developers-owner AT cygwin DOT com Delivered-To: mailing list cygwin-developers AT cygwin DOT com Date: Thu, 10 Jan 2002 17:00:45 -0500 From: Christopher Faylor To: cygwin-developers AT cygwin DOT com Subject: Re: Am I using dumper.exe correctly? Message-ID: <20020110220045.GA30612@redhat.com> Reply-To: cygwin-developers AT cygwin DOT com Mail-Followup-To: cygwin-developers AT cygwin DOT com References: <20020110214350 DOT 30413 DOT qmail AT lizard DOT curl DOT com> <20020110214755 DOT GA30469 AT redhat DOT com> <20020110215709 DOT 30544 DOT qmail AT lizard DOT curl DOT com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20020110215709.30544.qmail@lizard.curl.com> User-Agent: Mutt/1.3.23.1i On Thu, Jan 10, 2002 at 04:57:09PM -0500, Jonathan Kamens wrote: >> Date: Thu, 10 Jan 2002 16:47:55 -0500 >> From: Christopher Faylor >> >> Is there a .stackdump file being created? > >It's difficult to diagnose exactly what's happening, since we run our >automated builds in an off-screen desktop, i.e., we don't expect them >to pop up any dialogs, and if any dialogs *do* pop up, the build hangs >because there's a dialog off in never-never-land waiting for someone >to click "OK". We think that's what's happening here, because (a) the >builds are hanging with only Cygwin processes in the process list and >(b) we've seen dialogs of this sort occasionally when running manual >builds *not* in an off-screen desktop. > >I don't know why the dialog gets popped up instead of a stackdump >being generated. Ah. I've seen this from time to time. I'm never 100% sure why you see a dialog rather than a stack dump. Initially, I thought it was because a SEGV happened just prior to the initialization required for trapping an exception but I would swear that I've seen a dialog pop up when that wasn't the case, too. Anyway, I'll look at dcrt0.cc and see if the exception handling can be started a little earlier. cgf