Radeonhd card displays whitescreen on login when nvidia pci-e card removed
I have a radeonhd card: mpower@dodtsair:~$ lspci | grep VGA 01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3200 Graphics 02:00.0 VGA compatible controller: nVidia Corporation G70 [GeForce 7800 GT] (rev a1) Works great, but I decided to rebuild a gaming machine and since I was happy with the ATI in linux and I was not using the nVidia card I decided I would use it for my gaming box until the new nvidia cards came out. So I shutdown the computer removed the card and to my dismay instead of a ubuntu login screen I get a whitescreen. I am not sure what happened I played with it a bit. The only error I could find anywhere was when running startx from the command prompt. It displayed a white screen and then continuely output something like Protocol not specified. Attached are two logs Xorg.0.log which is the log of the xserver that I am currently running in. You'll see that it has two video cards. The Xorg.0.log.old file is the xserver that failed to start. I diffed them using meld and despite the huge lack of EDID querying and resolution stuff at the end I see no problem. It seems like the xserver decided to stop half way through start up and display a white screen. I have no clue why. This is an onboard ATI card. I have heard something about needing to specify the BusID in the config due to some bug somewhere. I am going to fire this email off and then add BusID "PCI:1:5:0" if that does not work I'll try BusID "PCI:0:1:5" Hopefully that will work if not... I could use some help :) Mike Power # nvidia-xconfig: X configuration file generated by nvidia-xconfig # nvidia-xconfig: version 1.0 (buildmeister@builder58) Fri Aug 14 18:33:37 PDT 2009 # xorg.conf (X.Org X Window System server configuration file) # # This file was generated by dexconf, the Debian X Configuration tool, using # values from the debconf database. # # Edit this file with caution, and see the xorg.conf manual page. # (Type "man xorg.conf" at the shell prompt.) # # This file is automatically updated on xserver-xorg package upgrades *only* # if it has not been modified since the last upgrade of the xserver-xorg # package. # # Note that some configuration settings that could be done previously # in this file, now are automatically configured by the server and settings # here are ignored. # # If you have edited this file but would like it to be automatically updated # again, run the following command: # sudo dpkg-reconfigure -phigh xserver-xorg Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" 0 0 InputDevice "Keyboard0" "CoreKeyboard" InputDevice "Mouse0" "CorePointer" EndSection Section "InputDevice" # generated from default Identifier "Keyboard0" Driver "keyboard" EndSection Section "InputDevice" # generated from default Identifier "Mouse0" Driver "mouse" Option "Protocol" "auto" Option "Device" "/dev/psaux" Option "Emulate3Buttons" "no" Option "ZAxisMapping" "4 5" EndSection Section "Monitor" Identifier "Configured Monitor" EndSection Section "Device" Identifier "Configured Video Device" Driver "radeonhd" EndSection Section "Screen" Identifier "Default Screen" Device "Configured Video Device" Monitor "Configured Monitor" SubSection "Display" Virtual 3840 1200 EndSubSection EndSection
I have a radeonhd card: mpower@dodtsair:~$ lspci | grep VGA 01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3200 Graphics 02:00.0 VGA compatible controller: nVidia Corporation G70 [GeForce 7800 GT] (rev a1)
Works great, but I decided to rebuild a gaming machine and since I was happy with the ATI in linux and I was not using the nVidia card I decided I would use it for my gaming box until the new nvidia cards came out.
So I shutdown the computer removed the card and to my dismay instead of a ubuntu login screen I get a whitescreen. I am not sure what happened I played with it a bit.
The only error I could find anywhere was when running startx from the command prompt. It displayed a white screen and then continuely output something like Protocol not specified. Attached are two logs Xorg.0.log which is the log of the xserver that I am currently running in. You'll see that it has two video cards. The Xorg.0.log.old file is the xserver that failed to start. I diffed them using meld and despite the huge lack of EDID querying and resolution stuff at the end I see no problem. It seems like the xserver decided to stop half way through start up and display a white screen.
I have no clue why.
This is an onboard ATI card. I have heard something about needing to specify the BusID in the config due to some bug somewhere. I am going to fire this email off and then add
BusID "PCI:1:5:0" if that does not work I'll try BusID "PCI:0:1:5"
Hopefully that will work if not... I could use some help :)
Mike Power
BusID "PCI:1:5:0" seems to be the correct bus id, however it reproduces the same problem. BusId "PCI:0:5:0" and "PCI:0:1:5" results in X erroring out with no devices found. The resulting OS is corrupted and I was not able to get the text prompt to reconfigure it. The text display was corrupted beyond recollection. I was able to ssh into the box and reconfigure it. Even after that a simple restart of gdm did not resurrect the display. I had to reboot the box to fix the display Changing the priority of the bios from internal graphics card to pcie card did not help the driver use the onboard chipset. At this point in time I am able to use the onboard chipset by simply removing the Driver "radeonhd" line. Obviously not what I desired, as that means I am probably using the ATI closed source driver. But given the options between white screen and text display I'll just have to take it. The fact that radeonhd fails and the ati driver works suggests to me that there is a driver problem or perhaps a conflict with that driver. I have taken this problem as far as I can, if anyone has anything else to suggest I'll try it out otherwise I'll have to stick with the closed source driver. Thanks All. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Sat, Mar 06, 2010 at 06:00:51PM -0800, Mike Power wrote:
I have a radeonhd card: mpower@dodtsair:~$ lspci | grep VGA 01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3200 Graphics 02:00.0 VGA compatible controller: nVidia Corporation G70 [GeForce 7800 GT] (rev a1)
Works great, but I decided to rebuild a gaming machine and since I was happy with the ATI in linux and I was not using the nVidia card I decided I would use it for my gaming box until the new nvidia cards came out.
So I shutdown the computer removed the card and to my dismay instead of a ubuntu login screen I get a whitescreen. I am not sure what happened I played with it a bit.
The only error I could find anywhere was when running startx from the command prompt. It displayed a white screen and then continuely output something like Protocol not specified. Attached are two logs Xorg.0.log which is the log of the xserver that I am currently running in. You'll see that it has two video cards. The Xorg.0.log.old file is the xserver that failed to start. I diffed them using meld and despite the huge lack of EDID querying and resolution stuff at the end I see no problem. It seems like the xserver decided to stop half way through start up and display a white screen.
I have no clue why.
This is an onboard ATI card. I have heard something about needing to specify the BusID in the config due to some bug somewhere. I am going to fire this email off and then add
BusID "PCI:1:5:0" if that does not work I'll try BusID "PCI:0:1:5"
Hopefully that will work if not... I could use some help :)
Mike Power
BusID "PCI:1:5:0" seems to be the correct bus id, however it reproduces the same problem.
BusId "PCI:0:5:0" and "PCI:0:1:5" results in X erroring out with no devices found. The resulting OS is corrupted and I was not able to get the text prompt to reconfigure it. The text display was corrupted beyond recollection. I was able to ssh into the box and reconfigure it. Even after that a simple restart of gdm did not resurrect the display. I had to reboot the box to fix the display
Changing the priority of the bios from internal graphics card to pcie card did not help the driver use the onboard chipset.
At this point in time I am able to use the onboard chipset by simply removing the Driver "radeonhd" line. Obviously not what I desired, as that means I am probably using the ATI closed source driver. But given the options between white screen and text display I'll just have to take it.
The fact that radeonhd fails and the ati driver works suggests to me that there is a driver problem or perhaps a conflict with that driver.
I have taken this problem as far as I can, if anyone has anything else to suggest I'll try it out otherwise I'll have to stick with the closed source driver.
Thanks All.
White screen usually means that the MC is not set up ok. Luc Verhaegen. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Sat, Mar 06, 2010 at 06:00:51PM -0800, Mike Power wrote:
I have a radeonhd card: mpower@dodtsair:~$ lspci | grep VGA 01:05.0 VGA compatible controller: ATI Technologies Inc Radeon HD 3200 Graphics 02:00.0 VGA compatible controller: nVidia Corporation G70 [GeForce 7800 GT] (rev a1)
Works great, but I decided to rebuild a gaming machine and since I was happy with the ATI in linux and I was not using the nVidia card I decided I would use it for my gaming box until the new nvidia cards came out.
So I shutdown the computer removed the card and to my dismay instead of a ubuntu login screen I get a whitescreen. I am not sure what happened I played with it a bit.
The only error I could find anywhere was when running startx from the command prompt. It displayed a white screen and then continuely output something like Protocol not specified. Attached are two logs Xorg.0.log which is the log of the xserver that I am currently running in. You'll see that it has two video cards. The Xorg.0.log.old file is the xserver that failed to start. I diffed them using meld and despite the huge lack of EDID querying and resolution stuff at the end I see no problem. It seems like the xserver decided to stop half way through start up and display a white screen.
I have no clue why.
This is an onboard ATI card. I have heard something about needing to specify the BusID in the config due to some bug somewhere. I am going to fire this email off and then add
BusID "PCI:1:5:0" if that does not work I'll try BusID "PCI:0:1:5"
Hopefully that will work if not... I could use some help :)
Mike Power
BusID "PCI:1:5:0" seems to be the correct bus id, however it reproduces the same problem.
BusId "PCI:0:5:0" and "PCI:0:1:5" results in X erroring out with no devices found. The resulting OS is corrupted and I was not able to get the text prompt to reconfigure it. The text display was corrupted beyond recollection. I was able to ssh into the box and reconfigure it. Even after that a simple restart of gdm did not resurrect the display. I had to reboot the box to fix the display
Changing the priority of the bios from internal graphics card to pcie card did not help the driver use the onboard chipset.
At this point in time I am able to use the onboard chipset by simply removing the Driver "radeonhd" line. Obviously not what I desired, as that means I am probably using the ATI closed source driver. But given the options between white screen and text display I'll just have to take it.
The fact that radeonhd fails and the ati driver works suggests to me that there is a driver problem or perhaps a conflict with that driver.
I have taken this problem as far as I can, if anyone has anything else to suggest I'll try it out otherwise I'll have to stick with the closed source driver.
Thanks All.
White screen usually means that the MC is not set up ok.
Luc Verhaegen.
So what is an MC? Mike Power -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Sat, Mar 06, 2010 at 07:20:46PM -0800, Mike Power wrote:
So what is an MC?
Mike Power
Memory controller; src/rhd_mc.c Luc Verhaegen -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Mar 06, 10 18:00:51 -0800, Mike Power wrote:
At this point in time I am able to use the onboard chipset by simply removing the Driver "radeonhd" line. Obviously not what I desired, as that means I am probably using the ATI closed source driver. But given the options between white screen and text display I'll just have to take it.
Which means that you have the fglrx driver installed while testing
radeonhd. This is something *completely* unsupported, because the fglrx
kernel module messes around with the hardware beyond repair.
Matthias
--
Matthias Hopf
On Mar 06, 10 18:00:51 -0800, Mike Power wrote:
At this point in time I am able to use the onboard chipset by simply removing the Driver "radeonhd" line. Obviously not what I desired, as that means I am probably using the ATI closed source driver. But given the options between white screen and text display I'll just have to take it.
Which means that you have the fglrx driver installed while testing radeonhd. This is something *completely* unsupported, because the fglrx kernel module messes around with the hardware beyond repair.
Matthias
Okay so I created a bug in Ubuntu to make radeonhd conflict with fglrx so they both can not be installed at the same time. However I think I made a mistake. I do not think fglrx is on the system. mpower@dodtsair:~/Desktop$ apt-cache show xorg-driver-fglx W: Unable to locate package xorg-driver-fglx E: No packages found mpower@dodtsair:~/Desktop$ locate fglrx /usr/share/app-install/desktop/fglrx-driver.desktop /usr/share/apport/package-hooks/source_fglrx-installer.py /usr/share/doc/fglrx-modaliases /usr/share/doc/fglrx-modaliases/changelog.Debian.gz /usr/share/doc/fglrx-modaliases/copyright /usr/share/jockey/handlers/fglrx.py /usr/share/jockey/handlers/fglrx.pyc /usr/share/jockey/modaliases/fglrx-modules.alias.override /var/cache/jockey/fglrx.oldconf /var/lib/dpkg/info/fglrx-modaliases.list /var/lib/dpkg/info/fglrx-modaliases.md5sums mpower@dodtsair:~/Desktop$ Based on the Xorg log (II) LoadModule: "radeon" (II) Loading /usr/lib/xorg/modules/drivers//radeon_drv.so (II) Module radeon: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0 (II) RADEON: Driver for ATI Radeon chipsets: And yet even as I read the log I see: (II) LoadModule: "ati" (II) Loading /usr/lib/xorg/modules/drivers//ati_drv.so (II) Module ati: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0 Is that bad, is that the fglrx driver, or is it dependent on the fglrx driver? Mike Power -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Thu, Mar 11, 2010 at 09:17:11AM -0800, Mike Power wrote:
Okay so I created a bug in Ubuntu to make radeonhd conflict with fglrx so they both can not be installed at the same time.
However I think I made a mistake. I do not think fglrx is on the system.
mpower@dodtsair:~/Desktop$ apt-cache show xorg-driver-fglx W: Unable to locate package xorg-driver-fglx E: No packages found mpower@dodtsair:~/Desktop$ locate fglrx /usr/share/app-install/desktop/fglrx-driver.desktop /usr/share/apport/package-hooks/source_fglrx-installer.py /usr/share/doc/fglrx-modaliases /usr/share/doc/fglrx-modaliases/changelog.Debian.gz /usr/share/doc/fglrx-modaliases/copyright /usr/share/jockey/handlers/fglrx.py /usr/share/jockey/handlers/fglrx.pyc /usr/share/jockey/modaliases/fglrx-modules.alias.override /var/cache/jockey/fglrx.oldconf /var/lib/dpkg/info/fglrx-modaliases.list /var/lib/dpkg/info/fglrx-modaliases.md5sums mpower@dodtsair:~/Desktop$
Based on the Xorg log
(II) LoadModule: "radeon" (II) Loading /usr/lib/xorg/modules/drivers//radeon_drv.so (II) Module radeon: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0 (II) RADEON: Driver for ATI Radeon chipsets:
And yet even as I read the log I see:
(II) LoadModule: "ati" (II) Loading /usr/lib/xorg/modules/drivers//ati_drv.so (II) Module ati: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0
Is that bad, is that the fglrx driver, or is it dependent on the fglrx driver?
Mike Power
This isn't even the radeonhd driver. And if you are trying to run the radeonhd driver next to it, there are symbols in radeon still which have the exact same name as some symbols in radeonhd (and they start with RHD), so that both drivers cannot be used at the same time. Luc Verhaegen. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Mar 11, 10 18:21:19 +0100, Luc Verhaegen wrote:
This isn't even the radeonhd driver. And if you are trying to run the radeonhd driver next to it, there are symbols in radeon still which have the exact same name as some symbols in radeonhd (and they start with RHD), so that both drivers cannot be used at the same time.
They can, at least in openSUSE. We have an additional patch in the
server that prefers module local symbols over external symbols.
When no xorg.conf is available, all known drivers are loaded for
probing, thus we noticed this clash (I think it's over 300 symbols that
clash, some with different ABI, some with subtle different semantics).
We had the patch discussed upstream, but as upstream already builds
drivers with non-exported symbols, people didn't care enough to apply
the patch (though it would be non-invasive). It is linux-centric, though
(it does work on OSes that support the same special dlopen flags).
Matthias
--
Matthias Hopf
On Mar 11, 10 09:17:11 -0800, Mike Power wrote:
And yet even as I read the log I see:
(II) LoadModule: "ati" (II) Loading /usr/lib/xorg/modules/drivers//ati_drv.so (II) Module ati: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0
Is that bad, is that the fglrx driver, or is it dependent on the fglrx driver?
No, that is the *radeon* driver (radeonhd's twin brother - or sister if
you insist). Choose the one that's working best for you.
Matthias
--
Matthias Hopf
On Mar 11, 10 09:17:11 -0800, Mike Power wrote:
And yet even as I read the log I see:
(II) LoadModule: "ati" (II) Loading /usr/lib/xorg/modules/drivers//ati_drv.so (II) Module ati: vendor="X.Org Foundation" compiled for 1.6.4, module version = 6.12.99 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 5.0
Is that bad, is that the fglrx driver, or is it dependent on the fglrx driver?
No, that is the *radeon* driver (radeonhd's twin brother - or sister if you insist). Choose the one that's working best for you.
Matthias
Well I get more mileage with the ati_drv.so as I am writing an email using it. So I'll just be good with this. However two things. 1) If you guys wanted to triage this hardware configuration I am still game 2) I get radeonhd, but what is radeon_drv.so versus ati_drv.so versus fglrx? I assume the fglrx is the closed source driver. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Mar 11, 10 22:38:45 -0800, Mike Power wrote:
1) If you guys wanted to triage this hardware configuration I am still game
Would like to, but ENOTIME.
2) I get radeonhd, but what is radeon_drv.so versus ati_drv.so versus fglrx? I assume the fglrx is the closed source driver.
Yes. radeon and radeonhd are two different open source drivers. In your
case both support your hardware. In principle.
Matthias
--
Matthias Hopf
On Fri, Mar 12, 2010 at 7:38 AM, Mike Power
2) I get radeonhd, but what is radeon_drv.so versus ati_drv.so versus fglrx? I assume the fglrx is the closed source driver.
radeon_drv is the "real" radeon driver; ati_drv is just a stub that loads the correct driver depending on the card (i.e. radeon_drv for radeon cards, r128_drv for rage and mach64_drv for older mach cards). Luca -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
participants (4)
-
Luc Verhaegen -
Luca Tettamanti -
Matthias Hopf -
Mike Power