On Thu, Nov 26, 2009 at 5:21 AM, Viktor Malyarchuk <mal...@gmail.com> wrote: > Hi Jerome, > > yes, you are right. Reverting f30f37def4c72e5df8d949022b806e5a4e99bbcb for > 2.6.32rc8 did not fix the problem. > > During bisecting here > > "bad: [f30f37def4c72e5df8d949022b806e5a4e99bbcb] drm/radeon/kms: Update memory > bandwidth requirements in mode_set_base hook." > > for some reason git move form 2.6.32-rc6 to 2.6.31-rc9, so I can not test last > point. > While making another "git bisect good/bad" I actually see final message: > > git bisect good > f30f37def4c72e5df8d949022b806e5a4e99bbcb is the first bad commit > commit f30f37def4c72e5df8d949022b806e5a4e99bbcb > Author: Michel Dänzer <daen...@vmware.com> > Date: Thu Oct 8 10:44:09 2009 +0200 > > drm/radeon/kms: Update memory bandwidth requirements in mode_set_base > hook. > > The hook may change the number of bytes per pixel being scanned out, which > affects the CRTC memory bandwidth requirements. E.g. booting in 8bpp and > then > running X in 32bpp would result in the bandwidth requirements being > underestimated for the latter and consequently in CRTC FIFO underruns > causing > visible artifacts with 3D intensive workloads. > > ATOM changes only compile-tested. > > Signed-off-by: Michel Dänzer <daen...@vmware.com> > Signed-off-by: Dave Airlie <airl...@redhat.com> > > :040000 040000 b4468fc813956f4084eb05f8fe95eeb52d3a6736 > 57f05271bd0d2dcb49c083c9b946b91025ad85f9 M drivers > > The same as before: f30f37def4c72e5df8d949022b806e5a4e99bbcb > > git bisect bad > e33497aaaef5ec9998cf7b007b13d74bb41be74e is the first bad commit > commit e33497aaaef5ec9998cf7b007b13d74bb41be74e > Author: Dave Airlie <airl...@redhat.com> > Date: Mon Oct 12 13:32:38 2009 +1000 > > drm/radeon/kms: don't refuse LUT because no framebuffer. > > The fb binding might be happening in a subsequent crtc setup call. > > gets rid of some lut issues. > > Signed-off-by: Dave Airlie <airl...@redhat.com> > > :040000 040000 44d217fc51520c4c6fa2218072dd8a33571cb0c4 > b4468fc813956f4084eb05f8fe95eeb52d3a6736 M drivers > > Another commit: e33497aaaef5ec9998cf7b007b13d74bb41be74e > > Unfortunately, reverting e33497aaaef5ec9998cf7b007b13d74bb41be74e also do not > help. Problem still there. > > Reverting both f30f37def4c72e5df8d949022b806e5a4e99bbcb and > e33497aaaef5ec9998cf7b007b13d74bb41be74e do not work either. > > I tried to boot with CRT attached and it behave differently from LVDS. > At grub screen: LVDS is active, CRT is not. At 130s part: LVDS is black but > CRT became active immediately and show distorted grub image during that time. >
The distorted image is expected if you don't have fbcon loaded. This happens when the vga engines are disables, but the crtcs are still active. Do you still have long delay with fbcon loaded or compiled in? Alex > Please, see attachments to get better idea. May be this will help. > > Best, > Viktor > > > On Wednesday 25 November 2009 03:26:46 pm Jerome Glisse wrote: >> On Wed, Nov 25, 2009 at 02:49:16AM -0600, Viktor Malyarchuk wrote: >> > Hi Alex, >> > >> > if I understand it right, this is the culprit: >> > f30f37def4c72e5df8d949022b806e5a4e99bbcb >> >> This seems unlikely, i don't see how this patch can trigger such >> slowdown/freeze. Can you confirm that this patch is really what's >> causing it by using 2.6.32rc8 and doing: >> >> git revert f30f37def4c72e5df8d949022b806e5a4e99bbcb >> >> Thanks, >> >> Jerome Glisse >> >> > git bisect log >> > git bisect start >> > # good: [b419148e567728f6af0c3b01965c1cc141e3e13a] Linux 2.6.32-rc6 >> > git bisect good b419148e567728f6af0c3b01965c1cc141e3e13a >> > # bad: [156171c71a0dc4bce12b4408bb1591f8fe32dc1a] Linux 2.6.32-rc7 >> > git bisect bad 156171c71a0dc4bce12b4408bb1591f8fe32dc1a >> > # good: [a544de2ee0bc321bf4a08fa20b331878f2405a77] Merge branch >> > 'for_linus' of >> > git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6 git >> > bisect good a544de2ee0bc321bf4a08fa20b331878f2405a77 >> > # bad: [6b46362c0ea472b174c336786fd406c504326ad4] Merge branch >> > 'drm-linus' of >> > git://git.kernel.org/pub/scm/linux/kernel/git/airlied/drm-2.6 git bisect >> > bad 6b46362c0ea472b174c336786fd406c504326ad4 >> > # bad: [ee59f2b462b0a7d2eb57ae59340d016589fa8e42] drm/radeon/kms/r600: >> > fix rs880 support v2 git bisect bad >> > ee59f2b462b0a7d2eb57ae59340d016589fa8e42 >> > # bad: [c850cb782626fda78e5e9e5baf18a5bd806a225c] drm/kms: fix kms/fbdev >> > colormap support properly. git bisect bad >> > c850cb782626fda78e5e9e5baf18a5bd806a225c >> > # bad: [39deb2d67515086f08a672e7574716ca0d3883a5] drm/radeon/kms: fix use >> > of index before proper initialisation. git bisect bad >> > 39deb2d67515086f08a672e7574716ca0d3883a5 >> > # bad: [5f9a0eb5af16906cd31fc3e02746aa3d61fd3aa9] drm/radeon/kms/atom: >> > reorder crtc dpms git bisect bad 5f9a0eb5af16906cd31fc3e02746aa3d61fd3aa9 >> > # bad: [f30f37def4c72e5df8d949022b806e5a4e99bbcb] drm/radeon/kms: Update >> > memory bandwidth requirements in mode_set_base hook. git bisect bad >> > f30f37def4c72e5df8d949022b806e5a4e99bbcb >> > >> > Best, >> > Viktor >> > ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july -- _______________________________________________ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel
