Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cygwin-owner AT cygwin DOT com Mail-Followup-To: cygwin AT cygwin DOT com Delivered-To: mailing list cygwin AT cygwin DOT com Message-Id: <3.0.5.32.20031007203808.008239f0@incoming.verizon.net> X-Sender: vze1u1tg AT incoming DOT verizon DOT net Date: Tue, 07 Oct 2003 20:38:08 -0400 To: cygwin AT cygwin DOT com From: "Pierre A. Humblet" Subject: Re: cygwin-1.5.4-1 breaks fetchmail on Win9x (Pierre can you comment?) In-Reply-To: <20031007153217.GA20840@redhat.com> References: <20031007134914 DOT GA19358 AT redhat DOT com> <3F7F77B0 DOT 853C8E40 AT alphalink DOT com DOT au> <20031007021300 DOT GA1596 AT tishler DOT net> <20031007025837 DOT GD10639 AT redhat DOT com> <20031007131904 DOT GA2128 AT tishler DOT net> <20031007134914 DOT GA19358 AT redhat DOT com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" At 11:32 AM 10/7/2003 -0400, Christopher Faylor wrote: >On Tue, Oct 07, 2003 at 09:49:14AM -0400, Christopher Faylor wrote: >>On Tue, Oct 07, 2003 at 09:19:05AM -0400, Jason Tishler wrote: >>>On Mon, Oct 06, 2003 at 10:58:37PM -0400, Christopher Faylor wrote: >>>>On Mon, Oct 06, 2003 at 10:13:00PM -0400, Jason Tishler wrote: >>>>>BTW, there seemed to be some gyration regarding this section of >>>>>unlink() during that time period: >>>> >>>>...which might be illuminated by reading the archives, I suspect... >>> >>>I tried searching the archives via Google and Cygwin's mailing list >>>search engine, but came up empty. Would you be willing to enlighten >>>me? >> >>Actually, I am trying not to have to do the search myself. I recall >>that there was a discussion about this with Pierre which caused the >>change to take place. It might have been in the cygwin-developers >>list. Hmm, I remember investigating what's happening when a Win9X mounts a file system on NT and there are hard links.. This is what I see on Win98/Me: - DELETE_ON_CLOSE works if the file is not yet opened. - If it is opened for writing, CreateFile (DELETE_ON_CLOSE) fails and the file is eventually put on the delete queue, at least if it is local. Why not if it's remote? - However if the file is opened for reading, then CreateFile (DELETE_ON_CLOSE) succeeds, CloseHandle returns 1, but the file is not deleted. That's the case that Mark Ord examined. It's an MS bug, the documentation states that CreateFile should fail. So there is indeed a current problem. Until the next Cygwin release fetchmail could possibly patch things up by opening the file for writing. Pierre -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/