Problem(s) installing OpenBSD/hppa 3.3 on a model 735/125
Peter Wright <[email protected]>
| Newsgroups | gmane.os.openbsd.hppa |
|---|---|
| Message-ID | <[email protected]> |
Hi all. I've got a HPPA/PA-RISC machine, model 735/125, that I've been
hoping I could turn into a replacement for my current (OpenBSD 3.0/i386)
gateway/firewall/mailserver machine.
I've currently got Debian/hppa installed on the machine and working
fine, but I thought OpenBSD/hppa looked like a much more interesting
proposition (I thought it'd be particularly cool to have the new W^X
feature).
So... I downloaded the 3.3 install tarballs and the lif33.fs. Set
up rbootd on a Linux machine (cartman) on my local LAN, as well as the
tftp server. Rebooted the hppa machine (using the serial console,
connected using minicom on cartman).
I ran into the following problem - a kernel panic where I think it's
having some issues with the SCSI bus (message: "panic: unable to find
reselecting device"). I include a dump of the boot sequence, copied
from minicom. I also include 'trace' and 'ps' output, as helpfully
requested by the debugging prompt. :-)
[ NB. the following is with the "official" 3.3 release lif33.fs
(1667072 bytes)]:
----------- minicom terminal output ----------------------------------
(c) Copyright. Hewlett-Packard Company. 1992.
All rights reserved.
PDC ROM rev. 2.7
IODC ROM rev. 1.1
128 MB of memory configured and tested.
Selecting a system to boot.
To stop selection process, press and hold the ESCAPE key.
Selection process stopped.
Searching for Potential Boot Devices.
To terminate search, press and hold the ESCAPE key.
Device Selection Device Path Device Type
----------------------------------------------------------------------------
P0 scsi.5.0 DEC DSP3210S
P1 lan.0004e2-0b86db.0.0 cartman
b) Boot from specified device
s) Search for bootable devices
a) Enter Boot Administration mode
x) Exit and continue boot sequence
?) Help
Select from menu: b
Enter boot selection, (h)elp, or e(x)it: p1
Trying lan.0004e2-0b86db.0.0
Boot path initialized.
Attempting to load IPL.
Hard booted.
>> OpenBSD/hppa BOOT 0.8
boot>
booting dk0a:/bsd: 1607412+2699264+352256=0x6a9158
SPID bits: 0x0, error = -2
pdc_coproc: c0, c0
[ bsd ELF symbol table not valid: symtab unaligned ]
[ bsd symbol table has bad string table address 0x7fb3dc4a ]
[ no symbol table formats found ]
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
Copyright (c) 1995-2003 OpenBSD. All rights reserved. http://www.OpenBSD.org
OpenBSD 3.3 (RAMDISK) #4: Thu Mar 27 14:22:32 EST 2003
[email protected]:/usr/src/sys/arch/hppa/compile/RAMDISK
HP9000/735/130 (Snake Cheetah) PA-RISC 1.1a
real mem = 134217728 (262144 reserved for PROM, 8351744 used by OpenBSD)
avail mem = 108937216
using 8421 buffers containing 13393920 bytes of memory
mainbus0 (root) [flex fff80000]
pdc0 at mainbus0
mem0 at mainbus0 offset ffbf000: size 128MB
cpu0 at mainbus0 offset ffbe000 irq 31: PCXL L1-A 125MHz, FPU PCXT (Rolex - CMOS-26B) rev 1
cpu0: 256K(32b/l) Icache, 256K(32b/l) wr-back coherent Dcache, 120 coherent TLB, 16 BTLB
mg0 at mainbus0 offset c000000 irq 17: HWPC000 rev 34, 33 MHz
eisa0 at mg0
isa at mg0 not configured
asp0 at mainbus0 offset 82f000 irq 28: Hardball rev 19, lan 1 scsi 7
gsc0 at asp0: wordleds
"Advanced audio (int.)" at gsc0 (type a, sv 7f) offset 0 not configured
"Core FW SCSI" at gsc0 (type a, sv 7c) offset 30000 not configured
lpt0 at gsc0 offset 24000 irq 7
com1 at gsc0 offset 22000 irq 6: ns16550a, 16 byte fifo
com0 at gsc0 offset 23000 irq 5: ns16550a, 16 byte fifo
com0: console
hil0 at gsc0 offset 21000 irq 1
ie0 at gsc0 offset 26000 irq 8: i82596DX v0.0, address 08:00:09:41:b6:44
osiop0 at gsc0 offset 25000 irq 9: NCR53C700 rev 0, 31MHz, SCSI ID 7
osiop0: resetting chip
scsibus0 at osiop0: 8 targets
osiop0: target 5 now using 8 bit 7 MHz 8 REQ/ACK offset xfers
sd0 at scsibus0 targ 5 lun 0: <DEC, DSP3210S, 441G> SCSI2 0/direct fixed
osiop0: target ID 07 reselect nexus_list 0xc024a3dc
panic: unable to find reselecting device
Stopped at 0x172e98: break 0,5
RUN AT LEAST 'trace' AND 'ps' AND INCLUDE OUTPUT WHEN REPORTING THIS PANIC!
DO NOT EVEN BOTHER REPORTING THIS WITHOUT INCLUDING THAT INFORMATION!
ddb> trace
(null)(192964,f0823800,5,6e7800) at 0x172e98
(null)(18c680,f0823800,5,6e7800) at 0x82f44
(null)(0,a,6e8800,c023a000) at 0x82f28
(null)(5,1603000,b,a) at 0x49f54
(null)(3000,0,0,c0243a30) at 0x46038
(null)(c0243ac8,6f23c8,1a6000,1a6000) at 0x45d28
(null)(6f23c8,686918,156c0c,686900) at 0x154b84
(null)(1a64bc,c0243a00,6f2648,2) at 0x154dc8
(null)(191470,c0243a14,c0245614,2) at 0x154980
(null)(c023a074,685c78,156540,685c60) at 0x15c860
(null)(c023a074,685c78,156540,685c60) at 0x7d0f0
(null)(1a527c,8,c023a074,6e7800) at 0x15749c
(null)(191470,c0245614,c023a014,2) at 0x156a40
(null)(1,68f320,c023a5bc,8) at 0x156724
(null)(18ba7c,1f,7,6e7800) at 0x7d0f0
(null)(1b5128,9,5,6e7800) at 0x7cdac
(null)(191470,c023a014,c023a814,2) at 0x45928
(null)(191414,f0823800,5,6e7800) at 0x16ec8c
(null)(0,0,0,0) at 0x7d0f0
(null)(1b57d4,f0823800,5,6e7800) at 0x7cdac
(null)(191470,c023a814,c0244e94,2) at 0x172a7c
(null)(191470,c0245f14,c0246814,4024f0) at 0x16db68
(null)(1b5428,1b53d4,13,1) at 0x7d0f0
(null)(191470,c0244e94,c0245fd4,2) at 0x7cdac
(null)(0,a,6e8800,a) at 0x16aa18
(null)(0,0,0,0) at 0x7d0f0
(null)(1b50ac,fff80000,6f1308,693a10) at 0x7cdac
(null)(191464,c0245fd4,6f1408,2) at 0x172a7c
(null)(6d0e40,711000,70f800,a5a59) at 0x168a60
(null)(70f800,6d0d50,50,6d0d50) at 0x7d0f0
(null)(18f36c,f0823800,5,6e7800) at 0x7ce88
(null)(0,0,0,0) at 0x1716b4
(null)(0,0,0,0) at 0x67a34
-- trap #0
0x40154:
ddb> ps
PID PPID PGRP UID S FLAGS WAIT COMMAND
* 0 -1 0 0 2 0x80204 swapper
ddb>
----------- minicom terminal output ----------------------------------
I tried that a few times, just to make sure that it was at least a
consistent and repeatable bug.... it certainly was. *wry grin*
After a bit of thought, I decided it might be worth checking the
"snapshot" releases for OpenBSD/hppa and at least have a go with a
different lifimage.
Hence, with the June 10 "snapshot" lif33.fs (1640448 bytes):
----------- minicom terminal output ----------------------------------
(c) Copyright. Hewlett-Packard Company. 1992.
All rights reserved.
PDC ROM rev. 2.7
IODC ROM rev. 1.1
128 MB of memory configured and tested.
Selecting a system to boot.
To stop selection process, press and hold the ESCAPE key.
Selection process stopped.
Searching for Potential Boot Devices.
To terminate search, press and hold the ESCAPE key.
Device Selection Device Path Device Type
----------------------------------------------------------------------------
P0 scsi.5.0 DEC DSP3210S
P1 lan.0004e2-0b86db.0.0 cartman
b) Boot from specified device
s) Search for bootable devices
a) Enter Boot Administration mode
x) Exit and continue boot sequence
?) Help
Select from menu: b
Enter boot selection, (h)elp, or e(x)it: p1
Trying lan.0004e2-0b86db.0.0
Boot path initialized.
Attempting to load IPL.
Hard booted.
>> OpenBSD/hppa BOOT 0.8
booting dk0a:/bsd: 1579300+2650112+364544=0x660158
SPID bits: 0x0, error = -2
[ bsd ELF symbol table not valid: symtab unaligned ]
[ bsd symbol table has bad string table address 0x7fb34c4a ]
[ no symbol table formats found ]
Copyright (c) 1982, 1986, 1989, 1991, 1993
The Regents of the University of California. All rights reserved.
Copyright (c) 1995-2003 OpenBSD. All rights reserved. http://www.OpenBSD.org
OpenBSD 3.3-current (RAMDISK) #4: Tue Jun 10 19:29:20 EDT 2003
[email protected]:/usr/src/sys/arch/hppa/compile/RAMDISK
HP 9000/735/130 (Snake Cheetah) PA-RISC 1.1a
real mem = 134217728 (524288 reserved for PROM, 8560640 used by OpenBSD)
avail mem = 108990464
using 8421 buffers containing 13393920 bytes of memory
mainbus0 (root) [flex fff80000]
pdc0 at mainbus0
mem0 at mainbus0 offset ffbf000: size 128MB
cpu0 at mainbus0 offset ffbe000 irq 31: PCXL L1-A 125MHz, FPU PCXT (Rolex - CMOS-26B) rev 1
cpu0: 256K(32b/l) Icache, 256K(32b/l) wr-back coherent Dcache, 120 coherent TLB, 16 BTLB
mg0 at mainbus0 offset c000000 irq 17: HWPC000 rev 34, 33 MHz
eisa0 at mg0
isa at mg0 not configured
asp0 at mainbus0 offset 82f000 irq 28: Hardball rev 19, lan 1 scsi 7
gsc0 at asp0: wordleds
"Advanced audio (int.)" at gsc0 (type a, sv 7f) offset 0 not configured
"Core FW SCSI" at gsc0 (type a, sv 7c) offset 30000 not configured
lpt0 at gsc0 offset 24000 irq 7
com1 at gsc0 offset 22000 irq 6: ns16550a, 16 byte fifo
com0 at gsc0 offset 23000 irq 5: ns16550a, 16 byte fifo
com0: console
hil0 at gsc0 offset 21000 irq 1
ie0 at gsc0 offset 26000 irq 8: i82596DX v0.0, address 08:00:09:41:b6:44
osiop0 at gsc0 offset 25000 irq 9: NCR53C710 rev 0, 31MHz, SCSI ID 7
scsibus0 at osiop0: 8 targets
osiop0: osiop_select while connected?
----------- minicom terminal output ----------------------------------
...and then the boot process just "hung" there, with no further
output. I gave up on it after about five minutes.
I also just had a go booting with the lif33.fs snapshot dated as
follows:
OpenBSD 3.3-current (RAMDISK) #5: Tue Jun 24 15:34:19 EDT 2003
...but got exactly the same result with that as I did with the June 10
lifimage (ie. a "hang" on the line "osiop0: osiop_select while connected?").
I really just wanted to know if anyone on the list could suggest
another approach to getting OpenBSD 3.3/hppa installed - even if it's
just "wait until the next snapshot lif33.fs release and try that" :).
Alternatively, could the problem might be caused by something unusual
with my hardware setup? Debian/hppa and NetBSD/hppa boot/install
lifimages seem to work fine, so I'd think that's _probably_ not the
case, but still...
I'm not in any major hurry, so if OpenBSD/hppa-current is under active
development and/or bugfixing at the moment, I guess I'll just keep
trying the newer lifimages as they appear.
However, any other suggestions would be very welcome.
Thanks,
Pete.
--
http://akira.apana.org.au/~pete/
preventive maintenance, n. A superstitious ritual in which an engineer
is allowed to break something under controlled conditions in the
forlorn hope that this will appease some unknown gods.
-- The Devil's IT Dictionary