Created attachment 22796 [details]
Output of lspci,lsmod,modinfo on my PC.
When I added a second video card to my Debian Linux PC, the boot process fails with kernel panic or to an infinite loop with kernel error messages. The problem disappears when I remove one kernel module:
Hence, I believe that this module is incompatible with the module for the new video card.
Please find additional details below. As I am not a kernel programmer, I have no experience in debugging such problems. But I will happily cooperate to track down this bug if I receive instructions how to do this.
* Please find the output of the following commands attached:
* modinfo intel-agp
* modinfo radeon
* modinfo drm
* modinfo agpgart
* modinfo video
* modinfo output
* I guess the interesting lspci lines are these:
00:02.0 Display controller: Intel Corporation 82845G/GL[Brookdale-G]/GE Chipset Integrated Graphics Device (rev 03)
03:00.1 Display controller: ATI Technologies Inc RV280 [Radeon 9200 PRO] (Secondary) (rev 01)
* I guess the conflicting drivers are: intel-agp and radeon
* My PC has an on-board video card (I guess: 82845G/GL)
* When I upgraded my monitor, I found that it doesn't work with that card.
* I added a secondary card (I guess: ATI Radeon 9200) that works with the new monitor.
* After adding the card, boot fails with kernel panic/error messages.
* After removing the intel-agp.ko module, boot and monitor work fine with the secondary video card. I don't use the onbord video card.
* I use standard Debian Linux (stable or testing releases). The problem appears in many recent linux-image packages.
could you attach dmesg when boot failure?
As we don't have ATI card to produce this one, disable intel-agp through module blacklist is workaround for you.
ping for response...bug reporter
(In reply to comment #2)
> ping for response...bug reporter
I need your help.
When I reproduce the bug, the boot-up process gets stuck with a kernel panic (sometimes it just enters an infinite loop). As a result, the computer is not usable and I have to switch it off.
When I then reboot it, the dmesg of the failed boot is not anymore present in memory. I hope that the error messages are also written to some logfile --- but I can't find it.
Please advise how I can extract the error messages of a previous failed boot in a subsequent boot that doesn't show the same problem. Maybe I need to instruct the kernel to write it's messages to a file before it boots?
At this stage, I could only copy the log messages manually from the screen. But most likely, the interesting part is scrolled off the screen anyhow.
(In reply to comment #3)
> (In reply to comment #2)
> > ping for response...bug reporter
> I need your help.
> When I reproduce the bug, the boot-up process gets stuck with a kernel panic
> (sometimes it just enters an infinite loop). As a result, the computer is not
> usable and I have to switch it off.
> When I then reboot it, the dmesg of the failed boot is not anymore present in
> memory. I hope that the error messages are also written to some logfile --- but
> I can't find it.
> Please advise how I can extract the error messages of a previous failed boot in
> a subsequent boot that doesn't show the same problem.
Does this still exist? or we're time out here.