Re: mvsata problem

[Christoph Egger <Christoph_Egger%gmx.de@localhost>]

On 15.10.10 04:33, Mark Davies wrote:
> On Thu, 14 Oct 2010, KIYOHARA Takashi wrote:
>> Can you try rebuilded kernel with MVSATA_DEBUG and ATADEBUG again?
> 
> That produced:
> 
>    [...]
> wd0: 931 GB, 1938021 cyl, 16 head, 63 sec, 512 bytes/sect x 1953525168 
> sectors
> mvsata0:3: mvsata_exec_command: drive=0, bcount=512, r_command=0xec, 
> r_head=0x0, r_cyl=0x0, r_sector=0x0, r_count=0x0, r_features=0x0
> mvsata0:3: mvsata_cmd_start: drive=0
> mvsata0:0: mvsata_intr: cause=0x00000800
> mvsata0:3: mvsata_cmd_intr: drive=0
> mvsata0:3: mvsata_cmd_done: drive=0, flags=0x49
> mvsata0:3: mvsata_bio: drive=0, blkno=0, bcount=512
> mvsata0:3: mvsata_bio_start: drive=0
> mvsata0:0: mvsata_intr: cause=0x00000800
> fatal page fault in supervisor mode
<snippet from first mail>
trap type 6 code 0 eip c07871a1 cs 8 eflags 10246 cr2
Stopped in pid 0.2(system) at netbsd:wdcintr+0xc1: call *0x30(%ebx)
</snippet>
>    [...]

After the crash you see a "ddb>" prompt, right?
When type "show register" there I suppose, you will see
a line like

ebx           0

If this is the case, then the problem is a dereferencing a
NULL-pointer.

Christoph