Kernel panic 8.0-STABLE

Hello everyone!

Today i decided to go on 8.0-STABLE from 8.0-RELEASE and got this problem.

I did:

1. Grabbed source and run

Code:
#make buildworld

2. Copied GENERIC kernel, added
Code:
device sound
device snd_hda

changed

Code:
ident GENERIC

to

Code:
ident MYKERNELNAME

built and installed new kernel, rebooted and got panic

Code:
atapci0: <Marvell 88SX6141 UDMA133 controller> port 0x9c00-0x9c07, 0x9880-0x9883, 0x9800-0x983f, 0x9480 mem 0xfa7ffc00-0xfa7fffff irq 17 at device 0.0 on pci2
atapci1: <AHCI SATA controller> on atapci0
atapci1: unable to map interrupt
device_attach: atapci1 attach returned 6
atapci1: <AHCI SATA controller> on atapci0
panic: resource_list_alloc: resource entry is busy
cpuid = 0
Uptime: 1s
Automatic reboot in 15 second ...

My motherboard is ASUS P5WD2-E Premium. My BIOS Sata Mode set to ahci, and to Marvell Sata controller i connected 1 Sata HDD.

I hope i'll get some help. Thanks in advance.
 
forgot to say: first to time i got this was from single user mode, cause i didn't installed 'new world' yet.

booting from old kernel is fine..
 
I also tried to manually load ahci using old kernel, but system corresponded with version mistmatch (as i expected).

Also, atamarvell module is loaded too (using old kernel) though there is no definition in kernel conf...
 
mav@ said:
Then ahci(4) probably wasn't loaded. You should at least see `ahci0:` instead of `atapci1:` there.
Oops you are right.

Code:
atapci0: <Marvell 88SX6141 UDMA133 controller> port 0x9c00-0x9c07, 0x9880-0x9883, 0x9800-0x983f, 0x9480 mem 0xfa7ffc00-0xfa7fffff irq 17 at device 0.0 on pci2
ahci0: <Marvell 88SX6141 AHCI SATA controller> on atapci0
ahci0: unable to map interrupt
device_attach: atapci1 attach returned 6
ahci0: <Marvell 88XS6141 AHCI SATA controller> on atapci0
panic: resource_list_alloc: resource entry is busy
cpuid = 0
Uptime: 1s
Automatic reboot in 15 second ...
 
Do you have any loader tunables set?
Can you show related part of `pciconf -lvcb` output?
Try to enable verbose boot messages.
 
Hi,

for anyone interested:

i submitted a bug report for the problem with verbose boot output and backtrace. ( serial connection is needed since dumpdev is not available during kernel boot ).

http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/145064

Would be great if this could be resolved as this mainboard with onboard 8x sata ports (Marvell+Intel) makes for a nice cheap whitebox freebsd zfs fileserver.

thanks.
 
Back
Top