Created attachment 488100 [details] Error displayed. Description of problem: I tried to install Fedora 15 alpha / nightly (2011/3/27) and I get a kerneloops as soon as I try to start the livecd with the default grub entry. Version-Release number of selected component (if applicable): Fedora 15 Alpha How reproducible: Always. Steps to Reproduce: 1. Boot livecd on usb stick 2. 3. Actual results: A kerneloops and modprobe failure. Expected results: The livecd boots fine. Additional info: Backtrace are attached as images.
Created attachment 488101 [details] error (2)
Created attachment 488102 [details] error (3)
Booting with nomodeset helps - the oops does not appear.
Upgrading to 2.6.39-0.r0.git11.0.fc16 does not help.
Blacklisting i915 and radeon in grub is leading to a blank screen.
It is actually not a kerneloops, but the first line starts with: BUG: : unable to handle kernel paging [...]
On a side note: The laptop is using such a hybrid setup with two cards: # lspci | grep VGA 00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07) 01:00.0 VGA compatible controller: ATI Technologies Inc M92 LP [Mobility Radeon HD 4300 Series]
Thanks for the bug report. We have reviewed the information you have provided above, and there is some additional information especially concerning your hardware we require that will be helpful in our diagnosis of this issue. If the computer is not completely frozen when installation fails, switch to the console (Ctrl+Alt+F2) and copy /tmp/X* and /var/log/anaconda.xlog to some other place -- USB stick, some other computer via network, somewhere on the Internet, and please attach it to the bug report as individual uncompressed file attachments using the bugzilla file attachment link above. If the computer is completely useless after installation fails, you can also install Fedora with nomodeset, reboot computer in the runlevel 3 (add 3 to the kernel command line), and then run startx as a normal user after that. If it fails, still /var/log/Xorg.0.log and the output of dmesg program from the failed attempt to start X would be useful. We will review this issue again once you've had a chance to attach this information. Thank you very much in advance.
Created attachment 488360 [details] xorg.log
Created attachment 488361 [details] lspci
Created attachment 488362 [details] messages
Created attachment 488363 [details] xorg.log
The - just a couple of seconds ago - submitted files were gathered from the system after a "nomodeset" boot.
I've also noticed that /sys/kernel/debug/vgaswitcheroo is missing. It was present on 2.6.35. I do not if this is a sidecase of the nomodeset param or of something else.
Could this be related to https://bugzilla.kernel.org/show_bug.cgi?id=30052, as another hybrid solutions fails to start rawhide (https://bugzilla.kernel.org/show_bug.cgi?id=30052#c10)
(In reply to comment #15) > Could this be related to https://bugzilla.kernel.org/show_bug.cgi?id=30052, as > another hybrid solutions fails to start rawhide > (https://bugzilla.kernel.org/show_bug.cgi?id=30052#c10) Does not seem to be plausible, as the stack trace does not mention switcheroo in any way. And it is normal that the debugfs file is missing, as switcheroo needs kms.
You rigth, it's likely hybrid issue, can you desactive the intel gpu in the bios and see if it works then ?
The bios/model (HP Envy 13) hasn't got the functionality to disable any of the graphic cards.
I just wanted to note that this was fine (aka no problems with both cards) until pre-2.6.38 fedora kernels. I also tried to use rd.blacklist=i915 to prevent the intel module fom loading, but this is to late, the module is loaded eralier.
So, booting 2.6.38-1.fc15 without the "rhgb" and "quite" bootoptions helps: The system boots up. But this does not help with 2.6.39 (the one I tried before, from rawhide)
Booting with 2.6.38.1 also works when omitting rhgb and quiet! It is also important to note, that this only works, if the laptop is cold booted. If I reboot with omitting rhgb and quiet, the error apperas again.
Correction: It seems to be a race-condition, as a cold boot is no reliable way to start the system. So sometimes it works, sometimes not ...
Are there ideas or suggestions?
Additionally switching the two cards using the vgaswitcheroo/switch file does not work. The kernel also throws some error. This is a annoying regression from 2.6.35. Anything that can be done to support fixing this issue?
Are there any news, any way to help to track down this problem? this pronblem is very annoying. sometimes it takes 15 reboots to get a bootable system.
The latest update to F15beta brought some changes - the laptop booted fine a couple of times. the updates were: ... Updated xorg-x11-drv-cirrus-1.3.2-7.fc15.x86_64 Update 1.3.2-8.fc15.x86_64 Updated xorg-x11-drv-synaptics-1.3.99.901-2.fc15.x86_64 Update 1.4.0.901-1.fc15.x86_64 Updated xorg-x11-drv-wacom-0.10.99-1.20110315.fc15.x86_64 Update 0.11.0-1.fc15.x86_64 Updated xorg-x11-server-Xephyr-1.10.0-7.fc15.x86_64 Update 1.10.1-8.fc15.x86_64 Updated xorg-x11-server-Xorg-1.10.0-7.fc15.x86_64 Update 1.10.1-8.fc15.x86_64 Updated xorg-x11-server-common-1.10.0-7.fc15.x86_64 Update 1.10.1-8.fc15.x86_64 Updated xorg-x11-server-devel-1.10.0-7.fc15.x86_64 Update 1.10.1-8.fc15.x86_64
Thank you for letting us know.
Sorry, it just happened again. Same error. Nothing special, just booted the machine.
Sometimes the order of the pci devices in the vgaswitcheroo/switch file changes. Maybe this is related to this problem, as I suppose that this indicates, that the devices are initialised in a different order ...
This might be fixed in the latest kernel available here http://kojipkgs.fedoraproject.org/packages/kernel/2.6.38.6/26.rc1.fc15/ Please test it and let us know.
Thanks for the notice! I've updated to 2.6.38.6-26.rc1 yesterday and this problem is not fixed. I am currently running: [fabiand@sober ~]$ uname -a Linux sober 2.6.38.6-26.rc1.fc15.x86_64 #1 SMP Mon May 9 20:45:15 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux Anything more I can do?
2.6.39-0.rc7.git6.1 also does not solve the problem :-/ Are there any ideas why this problem appears at all?
A couple of beers if someone can give hint's how to solve this.
I am also willing to pay someone to get this bug fixed.
There seems to be an improvement on F15 (all updates) aka 2.6.39-1.fc16.x86_64. Let's see how stable this will be.
Waiting on your report then.
The problem did not appear over the last six boots. Some packages: xorg-x11-drv-ati-6.14.1-2.20110525gitfe5c42f51.fc15.x86_64 kernel-2.6.39-0.rc7.git6.1.fc16.x86_64 kernel-2.6.39-1.fc16.x86_64 Linux sober 2.6.39-1.fc16.x86_64 #1 SMP Sat May 21 02:34:01 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux So thanks and closing this for now :)