From owner-freebsd-x11@FreeBSD.ORG Sat Jul 14 13:33:30 2012 Return-Path: Delivered-To: x11@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 2DBB11065670; Sat, 14 Jul 2012 13:33:30 +0000 (UTC) (envelope-from marius@alchemy.franken.de) Received: from alchemy.franken.de (alchemy.franken.de [194.94.249.214]) by mx1.freebsd.org (Postfix) with ESMTP id 819F68FC12; Sat, 14 Jul 2012 13:33:29 +0000 (UTC) Received: from alchemy.franken.de (localhost [127.0.0.1]) by alchemy.franken.de (8.14.4/8.14.4/ALCHEMY.FRANKEN.DE) with ESMTP id q6EDXDA5093957; Sat, 14 Jul 2012 15:33:13 +0200 (CEST) (envelope-from marius@alchemy.franken.de) Received: (from marius@localhost) by alchemy.franken.de (8.14.4/8.14.4/Submit) id q6EDXDUN093956; Sat, 14 Jul 2012 15:33:13 +0200 (CEST) (envelope-from marius) Date: Sat, 14 Jul 2012 15:33:13 +0200 From: Marius Strobl To: x11@freebsd.org Message-ID: <20120714133313.GA93858@alchemy.franken.de> References: <20120619104247.GA13630@mech-cluster241.men.bris.ac.uk> <20120630121634.GA94551@alchemy.franken.de> <20120709094958.GB52954@mech-cluster241.men.bris.ac.uk> <20120709130430.GN63893@alchemy.franken.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20120709130430.GN63893@alchemy.franken.de> User-Agent: Mutt/1.4.2.3i Cc: flo@freebsd.org, freebsd-sparc64@freebsd.org, freebsd-ppc@freebsd.org Subject: Re: graphics/libGL regression on sparc64 X-BeenThere: freebsd-x11@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: X11 on FreeBSD -- maintaining and support List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 14 Jul 2012 13:33:30 -0000 On Mon, Jul 09, 2012 at 03:04:30PM +0200, Marius Strobl wrote: > On Mon, Jul 09, 2012 at 10:49:59AM +0100, Anton Shterenlikht wrote: > > On Sat, Jun 30, 2012 at 02:16:34PM +0200, Marius Strobl wrote: > > > On Tue, Jun 19, 2012 at 11:42:47AM +0100, Anton Shterenlikht wrote: > > > > On sparc64 r235474, > > > > updating from libGL-7.4.4 to 7.6.1 I get: > > > > > > There are several problems preventing Xorg bits to build on sparc64 > > > (and powerpc) since the update to 7.5.2. First, make sure you have a > > > ports tree with graphics/libdrm/Makefile rev. 1.25. Then apply the > > > following patches: > > > http://people.freebsd.org/~bapt/fix-hal-on-sparc64.diff > > > http://people.freebsd.org/~marius/dri_libGL_libdrm.diff > > > > > > According to a quick test, the old server works fine with both the > > > mach64 and the sunffb driver on sparc64. When running `Xorg -configure` > > > you need to manually fix the resulting configuration file though > > > as any device on the PCI bus not being mach64 compatible is detected > > > as a radeon chip. > > > However, while the new server selected with WITH_NEW_XORG builds > > > just fine on sparc64 with these patches, it doesn't work there. For > > > mach64, there isn't any indication in the log why this doesn't work > > > besides "no screens found", although the configuration is correct, > > > libdrm is built without KMS support and the mach64 being detected. > > > For sunffb, it just segfaults. > > > > > > Marius > > > > Is there a PR on this? I'd like to track it. > > No, not currently; I'm waiting for bapt@ to complete a -exp-run for > the options fix and to commit that patch and on feedback from miwi@ > regarding the patch for fixing the dri, libGL and libdrm ports on > powerpc and sparc64. > Given that I haven't received any feedback on the above mentioned dri_libGL_libdrm.diff so far, I'm going to commit it on June 16th unless someone comes up with an objection and given that I can get an approval from a ports committer. %% - Since the update to the Xorg 7.5.2 bits, graphics/libdrm no longer installs the necessary headers for compiling the Intel DRI drivers on !x86 (see graphics/libdrm/Makefile rev. 1.25), so exclude them on powerpc and sparc64 as they are of no use there anyway. This now generally follows the behavior used on Linux of only building the DRI drivers that are of possible use on a given architecture. This includes no longer building the i810 DRI driver on amd64 as there is no amd64 machine in existence where it could be used. - Fix/add some endian conversion bits also preventing things from building on powerpc and sparc64. %% Marius