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: Wed, 13 Jun 2001 11:25:16 -0400 From: Christopher Faylor To: cygwin-developers AT cygwin DOT com Subject: Re: fileutils-4.0-3 Message-ID: <20010613112516.D7001@redhat.com> Reply-To: cygwin-developers AT cygwin DOT com Mail-Followup-To: cygwin-developers AT cygwin DOT com References: <20010613105845 DOT D1144 AT cygbert DOT vinschen DOT de> <12812308989 DOT 20010613152159 AT logos-m DOT ru> <20010613153123 DOT K1144 AT cygbert DOT vinschen DOT de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.11i In-Reply-To: <20010613153123.K1144@cygbert.vinschen.de>; from vinschen@redhat.com on Wed, Jun 13, 2001 at 03:31:23PM +0200 On Wed, Jun 13, 2001 at 03:31:23PM +0200, Corinna Vinschen wrote: >On Wed, Jun 13, 2001 at 03:21:59PM +0400, egor duda wrote: >> Hi! >> >> Wednesday, 13 June, 2001 Corinna Vinschen cygwin AT cygwin DOT com wrote: >> >> CV> Typically it's the other way around. If a Win32 application opens >> CV> a file using the `CreateFile' call, it has exclusive access to the >> CV> file while it's opened. If the application want's to share the >> CV> file with other apps, it can do that by giving additional flags >> CV> to `CreateFile' (FILE_SHARE_READ, FILE_SHARE_WRITE). Cygwin's >> CV> open(2) call uses these flags by default. >> >> well, CreateFile() accepts 0 as second argument, which is what we >> need-- just query information no matter if anyone opened file in >> DENYALL mode. i've just tested it on nt4.0 -- it works fine. >> >> the only question is whether we should add new parameter to >> fhandler::open(), say 'int cygwin_flags', or define new flag in >> fcntl.h? for me, the first one looks preferable. > >Wow. I just read the MSDN entry of CreateFile and I must admit >that I always slipped over that sentence without reading it. >It seems obvious now. If that really works (as you state), >it would be the ultimate solution for `fstat'. > >I think you're right using some internal flag. It's not needed >to create a new fcntl flag. I don't see how fstat can do this, though. Do you mean stat? cgf