DFI Lanparty NF4 Ultra-D AMD64

[Søren Schmidt]

John Baldwin wrote:
> On Wednesday 16 November 2005 08:53 pm, Christopher Thomas wrote:
> 
>>Hello,
>>
>>first time poster, long time user.
>>
>>System:
>>DFI Lanparty NF4 ultra-D
>>AMD64 3500+
>>1GB GeIL DDR400 Dual Channel
>>Maxtor 12GB 5400RM
>>Lite-ON CDRW
>>
>>Problem:
>>Can install FBSD 6.0-Release/amd64 and i386 just fine, but upon first
>>reboot I get.
>>
>>Fatal Trap 18: integer divide fault while in kernel mode
>>instruction pointer  = 0x8:0xffffffff8022b99a
>>stack pointer         = 0x10:0xfffffff809769f0
>>frame pointer         = 0x10:0xfffff003b8af800
>>code segment        = base 0x0, limit 0xfffff, type 0x1b
>>                             = DBL 0, pres 1, long 1, def32 0, gran 1
>>processor flags      = interrupt enabled, resume, IOPL = 0
>>current process     = 0 (swapper)
>>trap number          = 18
>>panic: integer divide fault
>>Uptime: 1s
>>Cannot dump. No dump device defined.
>>(hand typed on my WinXP box (same mobo))
> 
> 
> It does seem related to ataraid.  I just ran gdb on the kernel.debug on the 
> 6.0 install CD and that instruction pointer comes out to:
> 
> (gdb) l *0xffffffff8022b99a
> 0xffffffff8022b99a is in ata_raid_promise_read_meta 
> (/usr/src/sys/dev/ata/ata-raid.c:2497).
> 
> Which corresponds to the line:
> 
> 
>     if (ata_raid_rw(parent, PR_LBA(parent),
> 		    meta, sizeof(struct promise_raid_conf), ATA_R_READ)) {
> 
> It looks like it probably died in the PR_LBA macro.  Perhaps the ata-raid code 
> isn't checking for a 'heads' value of 0 and it needs to?

Hmm heads and sectors cannot be null or at least then something else has 
gone wrong before this. Is the disk found and reported correctly (boot 
verbose) or does the geometry etc look strange ?


-Søren