Created attachment 23546 [details] Xorg.1.log from booting 8.10 (-ati 6.9.x) - failed Forwarding this bug from a Ubuntu reporter: https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/292421 [Problem] Black screen on X boot. Problem is a regression that first started in Intrepid with the 6.9.x driver, and is still seen on 6.10.0 and 6.11.0. At times, attempting an fglrx boot (which fails), and then booting -ati will make it work (log attached), but just for one boot. Included will be logs and register dumps from this case, to compare with the plain vanilla boot. Also attempted clean boots with no fglrx bits installed, but same behavior was seen. [lspci] 00:00.0 Host bridge: Intel Corporation 82975X Memory Controller Hub (rev c0) 05:00.0 VGA compatible controller: ATI Technologies Inc RV530LE [Radeon X1600/X1650 PRO] [Original Report] After updating to 8.10 the Xserver wasn't unable to display anything (putting my monitor into sleep mode). The Xorg logs don't show anything strange. I have noted the following differences when it not works: Output 68 disable success Blank CRTC 0 success Disable CRTC 0 success Blank CRTC 1 success Disable CRTC 1 success (II) RADEON(0): RADEONRestoreMemMapRegisters() : (II) RADEON(0): MC_FB_LOCATION : 0xdfffd000 0xdfffd000 (II) RADEON(0): MC_AGP_LOCATION : 0x00000000 (II) RADEON(0): avivo_restore ! Enable CRTC 1 success Unblank CRTC 1 success Enable CRTC 0 success Unblank CRTC 0 success May it be that the Enable and Unblanl don't work properly on the driver thus making that nothing can be seen on the screen?
Created attachment 23547 [details] [review] Xorg.0.log - functioning, 8.10 This is the log after using fglrx then restarting the xserver (as t also fails) and then using the radeon module.
Created attachment 23548 [details] Xorg.log.nodri Log from attempting a boot with DRI turned off - this didn't seem to make a difference.
Created attachment 23549 [details] Xorg.0.log - working case, nodri And for comparison, here is the same configuration as previously, but after doing the fglrx-crash trick. Here's the diff between the two files: 15c15 < (==) Log file: "/var/log/Xorg.0.log", Time: Sun Nov 9 15:41:12 2008 --- > (==) Log file: "/var/log/Xorg.0.log", Time: Sat Nov 8 17:44:54 2008 505c505 < (II) RADEON(0): MC_FB_LOCATION : 0xdfffd000 0xcfffc000 --- > (II) RADEON(0): MC_FB_LOCATION : 0xdfffd000 0xdfffd000 634,661c634,646 < (II) RADEON(0): I2C device "VGA-0:ddc2" registered at address 0xA0. < (II) RADEON(0): I2C device "VGA-0:ddc2" removed. < (II) RADEON(0): Output: VGA-0, Detected Monitor Type: 0 < Dac detection success < in RADEONProbeOutputModes < (II) RADEON(0): Adding Screen mode: 1280x1024 < (II) RADEON(0): Adding Screen mode: 1024x768 < (II) RADEON(0): Adding Screen mode: 800x600 < (II) RADEON(0): Adding Screen mode: 640x480 < (II) RADEON(0): Total number of valid Screen mode(s) added: 4 < (II) RADEON(0): I2C device "DVI-0:ddc2" registered at address 0xA0. < (II) RADEON(0): I2C device "DVI-0:ddc2" removed. < (II) RADEON(0): Output: DVI-0, Detected Monitor Type: 0 < Dac detection success < (II) RADEON(0): I2C device "VGA-0:ddc2" registered at address 0xA0. < (II) RADEON(0): I2C device "VGA-0:ddc2" removed. < (II) RADEON(0): Output: VGA-0, Detected Monitor Type: 0 < Dac detection success < in RADEONProbeOutputModes < (II) RADEON(0): Adding Screen mode: 1280x1024 < (II) RADEON(0): Adding Screen mode: 1024x768 < (II) RADEON(0): Adding Screen mode: 800x600 < (II) RADEON(0): Adding Screen mode: 640x480 < (II) RADEON(0): Total number of valid Screen mode(s) added: 4 < (II) RADEON(0): I2C device "DVI-0:ddc2" registered at address 0xA0. < (II) RADEON(0): I2C device "DVI-0:ddc2" removed. < (II) RADEON(0): Output: DVI-0, Detected Monitor Type: 0 < Dac detection success --- > Output 68 disable success > Blank CRTC 0 success > Disable CRTC 0 success > Blank CRTC 1 success > Disable CRTC 1 success > (II) RADEON(0): RADEONRestoreMemMapRegisters() : > (II) RADEON(0): MC_FB_LOCATION : 0xdfffd000 0xdfffd000 > (II) RADEON(0): MC_AGP_LOCATION : 0x00000000 > (II) RADEON(0): avivo_restore ! > Enable CRTC 1 success > Unblank CRTC 1 success > Enable CRTC 0 success > Unblank CRTC 0 success
Created attachment 23550 [details] register dumps regdump_vesa.txt (using vesa which is what I'm using now and seems to work fine) regdump_bad.txt (using radeon after vesa, fails) regdump_good.txt (using fglrx, killing the Xserver then using radeon, seems to work fine)
Created attachment 23551 [details] Xorg.0.log - jaunty Updated log after confirming issue is still present on Jaunty.
Created attachment 23552 [details] register dumps with 6.11.0 - Failure case
Created attachment 23553 [details] register dumps with 6.11.0 - Working case
Unfortunately, these dumps are from radeontool (rather than avivotool) so they aren't much use on this chip. for r1xx-r4xx use radeontool, for everything else, use avivotool. Any chance you could update the dumps?
I'm the original reporter. Thanks Bryce for all :) The following are the results when executing on Intrepid with the following X.org version: X.Org X Server 1.5.2 Release Date: 10 October 2008 X Protocol Version 11, Revision 0 Build Operating System: Linux 2.6.24-19-server i686 Ubuntu Current Operating System: Linux KDK 2.6.27-13-generic #1 SMP Thu Feb 26 07:26:43 UTC 2009 i686 Build Date: 09 March 2009 10:48:54AM xorg-server 2:1.5.2-2ubuntu3.1 (buildd@rothera.buildd) Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Module Loader present According to aptitude the radeon module version is 6.9.0 I can try it on jaunty or on another live distro if needed. Those are the avivotool reports :) avivo_bad.txt Report when using the radeon driver, screen gets blank. avivo_good.txt Report when using the vesa driver all is ok.
Created attachment 23837 [details] Results of sudo ./avivotool regmatch '*' with the latest git version
Created attachment 23838 [details] Results of sudo ./avivotool regmatch '*' with the latest git version when using radeon driver
(In reply to comment #9) > > According to aptitude the radeon module version is 6.9.0 Is this still happening with 6.11.0 or newer?
(In reply to comment #12) > (In reply to comment #9) > > > > According to aptitude the radeon module version is 6.9.0 > > Is this still happening with 6.11.0 or newer? I could say yes as it happened too with jaunty, though those reports aren't from that version. I can boot into jaunty later if necessary and make new reports-
Just a side comment. Using xorg-server-1.5.3 and xf86-video-ati-6.12.1-r1 from gentoo seems to solve the issue. But I don't know if it will work with xorg-server-1.6. I'll try and report later.
Created attachment 25239 [details] xorg.log on gentoo when it began failing again It's odd it began failing again (on gentoo) without touching anything. I got the following (strange) errors at dmesg: mtrr: no MTRR for d0000000,10000000 found mtrr: no MTRR for d0000000,1000000 found Also I go the following xorg.log mtrrs are enabled on kernel.
Does xf86-video-ati git master work for you?
(In reply to comment #16) > Does xf86-video-ati git master work for you? > I have the same issue (Asus EAX1650XT video card, Ubuntu 9.04). I tried installing the latest version of the driver (following the instructions at http://www.x.org/wiki/radeon). The result is still the same. When I start X, I get a blank screen, sometimes with some garbage near the top. Sometimes, I also have garbage where the mouse pointer is supposed to be. gdm plays its welcome sound, so the system is not completely locked up, but they keyboard doesn't seem to do anything anymore.
I am experiencing similar symptoms that are described in this bug. I have recently installed Fedora 11 on a machine with an X1650 XT video card. When I try to start X with the radeon driver, I get a blank screen with only a bit of garbage at the top. The keyboard and mouse are unresponsive, and I have to power off the machine. I have tried disabling DRI, which results in an equally useless display, but I get a mouse pointer, and I can kill the X server and return to a console. I've also tried disabling kernel mode setting, and that didn't help. Driver version = 6.12.2-14 X Server version = 1.6.1.901-1 kernel version = 2.6.29.4-167.fc11.i686.PAE Relevant lspci output: 01:00.0 VGA compatible controller: ATI Technologies Inc Radeon X1650 XT (Primary) (PCIE) (rev 9a) 01:00.1 Display controller: ATI Technologies Inc Radeon X1650 XT (Secondary) (PCIE) (rev 9a) Please let me know if I can provide any additional information.
After updating the driver and enabling ATOMTvOut the screens began working again. Now the question is, why isn't it enabled when the screen is analogical and the chip requires it?
(In reply to comment #19) > After updating the driver and enabling ATOMTvOut the screens began working > again. > > Now the question is, why isn't it enabled when the screen is analogical and the > chip requires it? > What are you asking? The ATOMTvOut option is only required for using tv-out. Other outputs (DVI, VGA, etc.) should work just fine without the ATOMTvOut option, if they don't then that's a bug. Or was this whole bug about tv-out not working? It's not clear from the description.
(In reply to comment #20) > What are you asking? The ATOMTvOut option is only required for using tv-out. > Other outputs (DVI, VGA, etc.) should work just fine without the ATOMTvOut > option, if they don't then that's a bug. Then it's a bug, the screen not working was a VGA monitor attached to the VGA port of the card. > Or was this whole bug about tv-out > not working? It's not clear from the description. Well I have a TV attached to the S-Video output but I don't use it. The problem was the VGA monitor.
Is this still an issue with KMS or a newer version of the driver?
I think this was fixed at some point of time but I forgot to close this bug. Even if it wasn't fixed I don't have access to the hardware anyway.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.