From mboxrd@z Thu Jan 1 00:00:00 1970 From: Markus Trippelsdorf Subject: Re: [PATCH] drm/ttm: fix delayed ttm_bo_cleanup_refs_and_unlock delayed handling Date: Thu, 17 Jan 2013 19:57:18 +0100 Message-ID: <20130117185718.GA242@x4> References: <1358263615.29800.32.camel@thor.local> <20130115163202.GA220@x4> <20130115170338.GA500@x4> <20130116231002.GA220@x4> <20130117084634.GA227@x4> <20130117161015.GA227@x4> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: Received: from mail.ud10.udmedia.de (ud10.udmedia.de [194.117.254.50]) by gabe.freedesktop.org (Postfix) with ESMTP id 4B6CEE5CB7 for ; Thu, 17 Jan 2013 10:57:20 -0800 (PST) Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dri-devel-bounces+sf-dri-devel=m.gmane.org@lists.freedesktop.org Errors-To: dri-devel-bounces+sf-dri-devel=m.gmane.org@lists.freedesktop.org To: Jerome Glisse Cc: Dave Airlie , Michel =?iso-8859-1?Q?D=E4nzer?= , Jerome Glisse , "dri-devel@lists.freedesktop.org" List-Id: dri-devel@lists.freedesktop.org On 2013.01.17 at 13:28 -0500, Jerome Glisse wrote: > On Thu, Jan 17, 2013 at 11:10 AM, Markus Trippelsdorf > wrote: > > On 2013.01.17 at 10:44 -0500, Jerome Glisse wrote: > >> On Thu, Jan 17, 2013 at 3:46 AM, Markus Trippelsdorf > >> wrote: > >> > On 2013.01.16 at 19:18 -0500, Jerome Glisse wrote: > >> >> On Wed, Jan 16, 2013 at 6:10 PM, Markus Trippelsdorf > >> >> wrote: > >> >> > On 2013.01.16 at 17:36 -0500, Alex Deucher wrote: > >> >> >> On Tue, Jan 15, 2013 at 12:03 PM, Markus Trippelsdorf > >> >> >> wrote: > >> >> >> > On 2013.01.15 at 17:32 +0100, Markus Trippelsdorf wrote: > >> >> >> >> On 2013.01.15 at 16:26 +0100, Michel D=E4nzer wrote: > >> >> >> >> > On Die, 201301-15 at 16:23 +0100, Markus Trippelsdorf wrote: > >> >> >> >> > > On 2013.01.15 at 15:43 +0100, Michel D=E4nzer wrote: > >> >> >> >> > > > On Sam, 2013-01-05 at 11:41 +0100, Markus Trippelsdorf = wrote: > >> >> >> >> > > > > On 2012.12.20 at 14:58 +0100, Markus Trippelsdorf wro= te: > >> >> >> >> > > > > > > >> >> >> >> > > > > > And just in case it got lost in the noise yesterday: > >> >> >> >> > > > > > The image corruption is caused by Dave's commit: > >> >> >> >> > > > > > > >> >> >> >> > > > > > commit dd54fee7d440c4a9756cce2c24a50c15e4c17ccb > >> >> >> >> > > > > > Author: Dave Airlie > >> >> >> >> > > > > > Date: Fri Dec 14 21:04:46 2012 +1000 > >> >> >> >> > > > > > > >> >> >> >> > > > > > radeon: fix regression with eviction since evic= t caching changes > >> >> >> >> > > > > > > >> >> >> >> > > > > > Reverting it 'fixes' the issue. > >> >> >> >> > > > > > >> >> >> >> > > > > Ping. > >> >> >> >> > > > > The issue still happens with todays Linus git tree. > >> >> >> >> > > > > >> >> >> >> > > > Does the corruption also occur with > >> >> >> >> > > > dd54fee7d440c4a9756cce2c24a50c15e4c17ccb applied manual= ly on top of > >> >> >> >> > > > 0d0b3e7443bed6b49cb90fe7ddc4b5578a83a88d? > >> >> >> >> > > > >> >> >> >> > > No. > >> >> >> >> > > >> >> >> >> > So, can you bisect which change between those two actually = introduced > >> >> >> >> > the corruption? > >> >> >> > > >> >> >> > The real cause of the image corruption is: > >> >> >> > > >> >> >> > d025e9e2b890db679f1246037bf65bd4be512627 is the first bad comm= it > >> >> >> > commit d025e9e2b890db679f1246037bf65bd4be512627 > >> >> >> > Author: Jerome Glisse > >> >> >> > Date: Thu Nov 29 10:35:41 2012 -0500 > >> >> >> > > >> >> >> > drm/radeon: do not move bo to different placement at each = cs > >> >> >> > > >> >> >> > The bo creation placement is where the bo will be. Instead= of trying > >> >> >> > to move bo at each command stream let this work to another= worker > >> >> >> > thread that will use more advance heuristic. > >> >> >> > > >> >> >> > agd5f: remove leftover unused variable > >> >> >> > > >> >> >> > Signed-off-by: Jerome Glisse > >> >> >> > Reviewed-by: Alex Deucher > >> >> >> > > >> >> >> > Reverting d025e9e2b890d on top of Linus' tree fixes the issue. > >> >> >> > >> >> >> Can you try this patch from Jerome: > >> >> >> https://bugzilla.kernel.org/attachment.cgi?id=3D91421 > >> >> > > >> >> > It fixes the corruption, but it degrades performance so much that= it > >> >> > takes several seconds to switch virtual desktops under xmonad. And > >> >> > sometimes the website used for the scroll test is stuck for sever= al > >> >> > seconds and unscrollable during that time. > >> >> > > >> >> > -- > >> >> > Markus > >> >> > >> >> What about this patch instead : > >> >> http://people.freedesktop.org/~glisse/0001-drm-radeon-exclude-syste= m-placement-when-validating-.patch > >> > > >> > This one doesn't work: > >> > >> Same address updated patch > >> > >> http://people.freedesktop.org/~glisse/0001-drm-radeon-exclude-system-p= lacement-when-validating-.patch > > > > It still doesn't work unfortunately. Can you please just revert > > d025e9e2b89 for now? Maybe it's better to wait for the next kernel > > release for another solution. > > > > Jan 17 17:05:34 x4 kernel: radeon 0000:01:05.0: GPU lockup CP stall for= more than 10000msec > > Jan 17 17:05:34 x4 kernel: radeon 0000:01:05.0: GPU lockup (waiting for= 0x000000000000022b last fence id 0x0000000000000224) > > Jan 17 17:05:34 x4 kernel: [drm:radeon_cs_ioctl] *ERROR* Failed to pars= e relocation -12! > > Jan 17 17:05:34 x4 kernel: [drm:radeon_gem_object_create] *ERROR* Faile= d to allocate GEM object (7000064, 6, 4096, -12) > > Jan 17 17:05:34 x4 kernel: [drm:radeon_gem_object_create] *ERROR* Faile= d to allocate GEM object (7000064, 6, 4096, -12) > > Jan 17 17:05:34 x4 kernel: [drm:radeon_gem_object_create] *ERROR* Faile= d to allocate GEM object (7000064, 6, 4096, -12) > > Jan 17 17:05:34 x4 kernel: radeon 0000:01:05.0: couldn't schedule ib > > Jan 17 17:05:34 x4 kernel: [drm:radeon_cs_ib_chunk] *ERROR* Failed to s= chedule IB ! > > Jan 17 17:05:34 x4 kernel: [drm:radeon_gem_object_create] *ERROR* Faile= d to allocate GEM object (7098368, 6, 4096, -12) > > Jan 17 17:05:34 x4 kernel: [drm:radeon_gem_object_create] *ERROR* Faile= d to allocate GEM object (7278592, 2, 4096, -12) > > > > -- > > Markus > = > For 3.9 sake can you try if > http://people.freedesktop.org/~glisse/0001-drm-radeon-keep-original-user-= requested-placement-ar.patch > = > on top of revert d025e9e2b89 works Yes, this combination works just fine. -- = Markus