Bug 18132

Summary: BTX dumps trying to boot w/ dedicated SCSI disks present
Product: Base System Reporter: Doug White <dwhite>
Component: i386Assignee: freebsd-bugs (Nobody) <bugs>
Status: Closed FIXED    
Severity: Affects Only Me    
Priority: Normal    
Version: 3.2-STABLE   
Hardware: Any   
OS: Any   

Description Doug White freebsd_committer freebsd_triage 2000-04-21 20:00:01 UTC 
When booting the above system with four IBM SCSI disks, one with a proper DOS
partition and the other three in dedicated mode with one single BSD partition,
the boot register dumps in what appears to be boot1 -- there are no boot messages
and no spinner.

The register dump is (hand-copied and leading zeros removed):
int=0 err=0 efl=00030246 eip=00001d29
eax=0 ebx=0 ecx=0 edx=0
esi=9e3d edi=1c09 ebp=039c esp=390
cs=c800 ds=40 es=9e3d fs=0 gs=0 ss=9e3d
cs: eip=f7f133d28a4ef6f7-f13dff037603b8ff
ss: esp=00003f0000000000-00000200220a00c8
System halted
int=d err=0 efl=10096 eip=90e3
eax=ffff ebf=394 ecx=20 edx=0
esi=9e3d edi=1c09 ebp=39c esp=\746e690a
cs=18 ds=20 fs=20 gs=20 ss=303d
cs: eip=eae890000031c08e-d08ed8bb0870e80f
ss: esp=b66ddbb66ddb-b66ddbb66ddb6ddb
System halted

Fix: 

Refromat dedicated mode disks to have proper slice tables.
Comment 1 John Baldwin freebsd_committer freebsd_triage 2000-05-09 12:29:47 UTC 
On 09-May-00 Doug White wrote:
> If you're going to decode boot panics, could you look at i386/18132
> too? :)

Yeah:

00000000  F7F1              div cx
00000002  33D2              xor dx,dx
00000004  8A4EF6            mov cl,[bp-0xa]
00000007  F7F1              div cx
00000009  3DFF03            cmp ax,0x3ff
0000000C  7603              jna 0x11
0000000E  B8FFFF            mov ax,0xffff

Since %cs=c800, we are in some extra hardware's ROM.  The actual fault
since %cx=0000.  There may be a return address of f300:0000 on the stack,
would be in the ROM BIOS.  It may be that your SCSI controller doesn't
like disks without a normal slice table.

-- 

John Baldwin <jhb@FreeBSD.org> -- http://www.FreeBSD.org/~jhb/
PGP Key: http://www.cslab.vt.edu/~jobaldwi/pgpkey.asc
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/
Comment 2 dwhite 2000-05-09 20:35:50 UTC 
On Tue, 9 May 2000, John Baldwin wrote:

> Since %cs=c800, we are in some extra hardware's ROM.  The actual fault
> since %cx=0000.  There may be a return address of f300:0000 on the stack,
> would be in the ROM BIOS.  It may be that your SCSI controller doesn't
> like disks without a normal slice table.

Chaulk one up to lame Adaptec software then.

Doug White                    |  FreeBSD: The Power to Serve
dwhite@resnet.uoregon.edu     |  www.FreeBSD.org
Comment 3 Mike Barcroft freebsd_committer freebsd_triage 2001-07-22 02:44:47 UTC 
State Changed
From-To: open->closed


This is apparently a hardware problem.