Red Hat Bugzilla – Bug 129923
[AGP] /sbin/loader panics when installing from non-onboard video
Last modified: 2015-01-04 17:08:46 EST
Description of problem: Shuttle SB52G2 (http://eu.shuttle.com/sb52g2.htm) p4/2600 512mb pc2100 ddr geforce fx5200 in pci slot 2 Version-Release number of selected component (if applicable): fedora core 3 test 1 How reproducible: always Steps to Reproduce: 1.install PCI video card (eg geforce fx5200) 2.setup bios to boot off PCI card instead of onboard intel agp video 3.watch installer blow up Actual results: /sbin/loader panics and the install halts Expected results: it should not panic :P Additional info: it appears /sbin/loader is trying to setup onboard video anyway, even though it's been disabled in bios? if i boot off onboard video it's ok. but the onboard video sucks, which is why i installed a geforce fx5200 into one of the pci slots.
any info besides "panics" ??? we have very little to go on here...
looks basically exactly the same as #124007
did this get fixed for fc3 final ?
nope. looks like a bug in intel-agp.ko it tries to setup agp even if it's disabled in bios, and panics the kernel. intel-agp.ko should just bail out instead. you can duplicate it on i82845G chipset (probably others too), just disable agp in bios and try loading intel-agp.ko for nice fireworks and a panicked kernel.
booting with agp=off should get you past the agp initialisation. Can you paste the exact agp messages that show up in the 'crashes' case ?
agp=off does nothing, still crashes. the oops backtrace when it crashes (copied by hand, since the kernel is hardlocked) show_stack show_registers die do_page_fault error_code apic_timer_interrupt agp_backend_initialize [agpgart] agp_add_bridge pci_device_probe_static __pci_device_probe pci_device_probe bus_match driver_attach bus_add_driver driver_register pci_register_driver init_module [intel_agp] sys_init_module sysenter_past_esp
repeated attempts over several months to contact the kernel developers responsible for this module have been fruitless. maybe someone at redhat will have better luck.
An update has been released for Fedora Core 3 (kernel-2.6.12-1.1372_FC3) which may contain a fix for your problem. Please update to this new kernel, and report whether or not it fixes your problem. If you have updated to Fedora Core 4 since this bug was opened, and the problem still occurs with the latest updates for that release, please change the version field of this bug to 'fc4'. Thank you.