X-Spam-Check-By: sourceware.org Date: Sat, 10 Dec 2005 18:25:58 -0500 From: Christopher Faylor To: cygwin AT cygwin DOT com Subject: Re: d_ino deprecated in latest snapshot Message-ID: <20051210232558.GA24474@trixie.casa.cgf.cx> Reply-To: cygwin AT cygwin DOT com References: <20051205212651 DOT GA12440 AT trixie DOT casa DOT cgf DOT cx> <439B601C DOT 4070207 AT byu DOT net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <439B601C.4070207@byu.net> User-Agent: Mutt/1.5.11 Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm Precedence: bulk List-Unsubscribe: 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 On Sat, Dec 10, 2005 at 04:09:16PM -0700, Eric Blake wrote: >According to Christopher Faylor on 12/5/2005 2:26 PM: >>I've deprecated dirent's d_ino field in the latest snapshot. I've >>taken this step because in the majority of cases this field was >>different from the inode field returned by stat() and this presents >>problems for programs which correctly think that they should be the >>same. >> >>This field will continue to be filled out (in)correctly for older >>programs but the field has now been renamed to __deprecated_d_ino so it >>will not be detectable by configure. The __deprecated_d_ino field is >>now zero in any applications built with the snapshot or newer versions >>of libcygwin.a. > >Actually, sys/dirent.h only has __deprecated_d_ino when >__INSIDE_CYGWIN__, but still used d_ino in __CYGWIN_USE_BIG_TYPES__. >Therefore a compile of findutils-4.2.27 against the headers of snapshot >20051205 claimed: > >checking for d_ino member in directory struct... yes > >> >>What this means is that, with luck, packages like bash and find which >>try to use this field in some meaningful way will switch to using the >>reliable st_ino field instead. > >Not quite. I'll wait for the next snapshot before putting >findutils-4.2.27-1 online as a test version, so that you can get >sys/dirent.h fixed first. I'll make the change but have you completely given up on the concept of "a patch" now? It seems like you could have tested this pretty easily by just making changes to the header and then rectified the behavior by submitting a patch. You didn't even need to check out cygwin source code for this one. cgf -- 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/