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: Tue, 13 Nov 2001 13:13:20 -0500 From: Christopher Faylor To: cygwin AT cygwin DOT com Subject: Re: AW: cygwin vfork Message-ID: <20011113181320.GA959@redhat.com> Reply-To: cygwin AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com References: <000101c16c39$19f18980$651c440a AT BRAMSCHE> <3BF13F87 DOT 2070600 AT ece DOT gatech DOT edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <3BF13F87.2070600@ece.gatech.edu> User-Agent: Mutt/1.3.23.1i On Tue, Nov 13, 2001 at 10:43:03AM -0500, Charles Wilson wrote: >> The VM comment is referring to the large footprint of XEmacs which means >> that doing a fork requires copying an awful lot of data (and hence takes a >> long time), most OS's do copy-on-write for vfork so the overhead is never >> incurred. > >And of course, cgf chimed in on this thread, but I can't find his message >in my mail archive, and (as mentioned elsewhere) the cygwin ml archive is >missing his message as well, so I can't quote it here for you. With vfork, there is very little copying going on. Only cygwin's heap is copied. Of course, if you don't use vfork then cygwin's fork implementation *is* pretty slow. 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/