Re: 3Ware 9500 and disklabel

From: Philip Murray <pmurray_at_nevada.net.nz>
Date: Thu, 22 Jul 2004 22:05:39 +1200
Harald Schmalzbauer wrote:

>Am Donnerstag, 22. Juli 2004 11:01 schrieb Philip Murray:
>  
>
>>Hi,
>>
>>I'm trying to test out -CURRENT (22/7/2004) on a dual Opteron (Tyan
>>board) with a 3Ware 9500-8 raid controller and I'm having problems with
>>disklabel.
>>
>>At the moment I have a RAID0 array totalling 600GB. I fdisk'd them to
>>create a single partition using the entire volume, fdisk (from
>>sysinstall) complained about the geometry, I wasn't sure what the
>>correct geometry was, so I forged ahead anyway.
>>
>>Then I went to try and disklabel it. It seemed to work from sysinstall
>>until I went to write the changes out and it complained "ERROR: Unable
>>to write data to disk da0!'. So I tried disklabel on the command line, as
>>so
>>    
>>
>
>AFAIK geom only prevents you from direct access to disks where your kernel 
>booted from. In your case it is ad4 and I think da1 shouldn't be affected but 
>perhaps you want to set the sysctl kern.geom.debugflags to 16.
>
>  
>
Trying to use disklabel from sysinstall with those geom flags resulted 
in a panic, I got the following trace from the debugger:

panic: spoiling cp->ace = 1
cpuid = 1;
KDB: enter: panic
[thread 100081]
Stopped at      kdb_enter+0x2f: nop
db> trace
kdb_enter() at kdb_enter+0x2f
panic() at panic+0x210
g_spoil() at g_spoil+0x7a
g_access() at g_access+0x1fb
g_dev_open() at g_dev_open+0x11b
spec_open() at spec_open+0x330
spec_vnoperate() at spec_vnoperate+0x14
vn_open_cred() at vn_open_cred+0x364
kern_open() at kern_open+0x109
syscall() at syscall+0x40e
Xfast_syscall() at Xfast_syscall+0xa8
--- syscall (5, FreeBSD ELF64, open), rip = 0x50ebc0, rsp = 
0x7fffffffd838, rbp = 0x7dd780 ---
db>

I don't know anything more useful I can do than that, so if there is 
something tell me. However, I have managed to somehow lockup the serial 
console so I can't do anymore until tomorrow.

Cheers

Philip
Received on Thu Jul 22 2004 - 08:05:53 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:02 UTC