I am having some strange problems with a new Linux install I did on an
old Sparc 5.
Here's a quick system spec:
Sparc 5, 64M , 2 1G HDD, 1 CD-ROM, Red Hat 5.1 (2.0.33 kernel)
Did install from CD-ROM and had some trouble:
Repartitioned both disks as follows:
Disk /dev/sda (Sun disk label): 14 heads, 72 sectors, 2036 cylinders
Units = cylinders of 1008 * 512 bytes
Device Flag Begin Start End Blocks Id System
/dev/sda1 0 0 203 102312 83 Linux native (/)
/dev/sda2 u 203 203 333 65520 82 Linux swap
/dev/sda3 0 0 2036 1026144 5 Whole disk
/dev/sda4 333 333 2036 858312 83 Linux native
(/usr)
Disk /dev/sdb (Sun disk label): 8 heads, 112 sectors, 2313 cylinders
Units = cylinders of 896 * 512 bytes
Device Flag Begin Start End Blocks Id System
/dev/sdb3 0 0 2313 1036224 5 Whole disk
/dev/sdb4 0 0 2313 1036224 83 Linux native
(/home)
Chose format everything & verify format, mount points and packages,
etc...
It formated the first disk (sda) no p roblem (AFAIK) (contains /,
/usr and swap)
It seemed to format the 2nd (sdb) disk ok, but then it hung.
(contains 1 partition [sdb4] for /home)
Eventually, I got a "unexpected signal 10" which caused everything to
abort.
Tried again, but this time w/o formating.
When it's supposed to start installing packages, it got the same
"signal 10" error.
3rd try: Just use sda (i.e. don't mount /sdb)
lo and behold! Installation proceeded w/o problems. Everything
boots. I'm cookin'!
--- extract of boot log ---
Feb 11 01:02:40 huckle kernel: klogd 1.3-3, log source = /proc/kmsg
started.
Feb 11 01:02:41 huckle kernel: Loaded 4879 symbols from
/boot/System.map.
Feb 11 01:02:41 huckle kernel: Symbols match kernel version 2.0.33.
Feb 11 01:02:41 huckle kernel: Error seeking in /dev/kmem
Feb 11 01:02:41 huckle kernel: Error adding kernel module table entry.
Feb 11 01:02:41 huckle kernel: PROMLIB: Sun Boot Prom Version 3 Revision
2
Feb 11 01:02:41 huckle kernel: ARCH: SUN4M
Feb 11 01:02:41 huckle kernel: TYPE: SPARCstation 5
Feb 11 01:02:41 huckle kernel: Ethernet address: 8:0:20:74:2d:28
Feb 11 01:02:41 huckle kernel: Found 1 CPU prom device tree node(s).
Feb 11 01:02:41 huckle kernel: IOMMU: impl 0 vers 4 page table at
f0280000 of si
ze 262144 bytes
Feb 11 01:02:41 huckle kernel: sbus0: Clock 21.1250 MHz
Feb 11 01:02:41 huckle kernel: dma0: Revision 2
Feb 11 01:02:41 huckle kernel: dma1: Revision 2
Feb 11 01:02:41 huckle kernel: Memory: 62372k available (1076k kernel
code, 1776
k data) [f0000000,f4000000]
Feb 11 01:02:41 huckle kernel: Linux version 2.0.33
([EMAIL PROTECTED]) (gcc version egcs-2.90.27 980315 (egcs-1.0.2
release)) #1 Tue Jun 2 07:29:25 EDT 1998
Feb 11 01:02:41 huckle kernel: Starting kswapd v 1.4.2.2
Feb 11 01:02:41 huckle kernel: Sparc Zilog8530 serial driver version
1.18.2.15
Feb 11 01:02:41 huckle kernel: Floppy drive(s): fd0 is 1.44M
Feb 11 01:02:41 huckle kernel: FDC 0 is a post-1991 82077
Feb 11 01:02:41 huckle kernel: md driver 0.35 MAX_MD_DEV=4, MAX_REAL=8
Feb 11 01:02:41 huckle kernel: linear personality registered
Feb 11 01:02:41 huckle kernel: raid0 personality registered
Feb 11 01:22:00 huckle kernel: esp0: IRQ 36 SCSI ID 7 Clock 40 MHz
CCF=8 Time-Out 167 NCR53C9XF(espfa
st) detected
Feb 11 01:22:00 huckle kernel: esp0: FAST chip is fas100a (family=0,
version=2)
Feb 11 01:22:00 huckle kernel: ESP: Total of 1 ESP hosts found, 1
actually in use.
Feb 11 01:22:00 huckle kernel: QPTI: Total of 0 PTI Qlogic/ISP hosts
found, 0 actually in use.
Feb 11 01:02:41 huckle kernel: scsi0 : Sparc ESP100A-FAST
Feb 11 01:02:41 huckle kernel: scsi : 1 host.
Feb 11 01:02:41 huckle kernel: Vendor: IBM Model:
DPES31080SUN1.05 Rev: S80B
Feb 11 01:02:41 huckle kernel: Type:
Direct-Access ANSI SCSI revision: 02
Feb 11 01:02:41 huckle kernel: Detected scsi disk sda at scsi0, channel
0, id 1, lun 0
Feb 11 01:02:41 huckle kernel: Vendor: CONNER Model: CFP1060E
1.05GB Rev: 2035
Feb 11 01:02:41 huckle kernel: Type:
Direct-Access ANSI SCSI revision: 02
Feb 11 01:02:41 huckle kernel: Detected scsi disk sdb at scsi0, channel
0, id 3, lun 0
Feb 11 01:02:41 huckle kernel: Vendor: TOSHIBA Model:
XM-4101TASUNSLCD Rev: 1084
Feb 11 01:02:41 huckle kernel: Type:
CD-ROM ANSI SCSI revision: 02
Feb 11 01:02:41 huckle kernel: Detected scsi CD-ROM sr0 at scsi0,
channel 0, id 6, lun 0
Feb 11 01:02:41 huckle kernel: scsi : detected 1 SCSI cdrom 2 SCSI disks
total.
Feb 11 01:02:41 huckle kernel: esp0: target 1 [period 100ns offset 15
10.00MHz FAST SCSI-II]
Feb 11 01:02:41 huckle kernel: SCSI device sda: hdwr sector= 512 bytes.
Sectors= 2118144 [1034 MB] [1.
0 GB]
Feb 11 01:02:41 huckle kernel: esp0: target 3 [period 100ns offset 15
10.00MHz FAST SCSI-II]
Feb 11 01:02:41 huckle kernel: SCSI device sdb: hdwr sector= 512 bytes.
Sectors= 2074880 [1013 MB] [1.
0 GB]
Feb 11 01:02:41 huckle kernel: sunlance.c:v1.9 21/Aug/96 Miguel de Icaza
([EMAIL PROTECTED])
Feb 11 01:02:41 huckle kernel: eth0: LANCE 08:00:20:74:2d:28
Feb 11 01:02:41 huckle kernel: eth0: using auto-carrier-detection.
Feb 11 01:02:41 huckle kernel: Partition check:
Feb 11 01:02:41 huckle kernel: sda: sda1 sda2 sda3 sda4 sda5 sda6 sda7
sda8
Feb 11 01:02:41 huckle kernel: sdb: sdb1 sdb2 sdb3 sdb4 sdb5 sdb6 sdb7
sdb8
Feb 11 01:02:41 huckle kernel: VFS: Mounted root (ext2 filesystem)
readonly.
Feb 11 01:02:41 huckle kernel: Freeing unused kernel memory: 40k code,
4k data
Feb 11 01:02:41 huckle kernel: Adding Swap: 65516k swap-space (priority
-1)
...
--- end extract ---
This is easy...I'll just mount sdb4 now..
woa....look at all those messages....I'm cooked!
--- extract begin ---
Feb 11 01:16:49 huckle kernel: esp0: Target will not disconnect
Feb 11 01:16:49 huckle kernel: esp0: resetting bus...
Feb 11 01:16:49 huckle kernel: esp0: SCSI bus reset interrupt
Feb 11 01:16:49 huckle kernel: esp0: SCSI bus reset interrupt
Feb 11 01:17:09 huckle kernel: scsi : aborting command due to timeout :
pid 2008, scsi0, channel 0, id
3, lun 0 Read (6) 0b 80 0e 02 00
Feb 11 01:17:09 huckle kernel: esp0: Aborting command
Feb 11 01:17:09 huckle kernel: esp0: dumping state
Feb 11 01:17:09 huckle kernel: esp0: dma -- cond_reg<a6400310>
addr<f3fa589a>
Feb 11 01:17:09 huckle kernel: esp0: SW [sreg<11> sstep<04> ireg<18>]
Feb 11 01:17:09 huckle kernel: esp0: HW reread [sreg<01> sstep<cc>
ireg<00>]
Feb 11 01:17:09 huckle kernel: esp0: current command [tgt<01> lun<00>
pphase<CLUELESS> cphase<DATAIN>]
Feb 11 01:17:09 huckle kernel: esp0: disconnected
Feb 11 01:17:09 huckle kernel: SCSI disk error : host 0 channel 0 id 3
lun 0 return code = 50000
Feb 11 01:17:09 huckle kernel: scsidisk I/O error: dev 08:14, sector
753678, absolute sector 753678
Feb 11 01:17:09 huckle kernel: scsi : aborting command due to timeout :
pid 2007, scsi0, channel 0, id
1, lun 0 Request Sense 00 00 00 10 00
Feb 11 01:17:09 huckle kernel: esp0: Aborting command
Feb 11 01:17:09 huckle kernel: esp0: dumping state
Feb 11 01:17:09 huckle kernel: esp0: dma -- cond_reg<a6400310>
addr<f3fa589a>
Feb 11 01:17:09 huckle kernel: esp0: SW [sreg<11> sstep<04> ireg<18>]
Feb 11 01:17:09 huckle kernel: esp0: HW reread [sreg<01> sstep<cc>
ireg<00>]
Feb 11 01:17:09 huckle kernel: esp0: current command [tgt<01> lun<00>
pphase<CLUELESS> cphase<DATAIN>]
Feb 11 01:17:09 huckle kernel: esp0: disconnected
Feb 11 01:17:09 huckle kernel: EXT2-fs error (device 08:14):
read_inode_bitmap: Cannot read inode bitm
ap - block_group = 46, inode_bitmap = 376839
Feb 11 01:17:10 huckle kernel: SCSI host 0 abort (pid 2007) timed out -
resetting
Feb 11 01:17:11 huckle kernel: SCSI bus is being reset for host 0
channel 0.
Feb 11 01:17:11 huckle kernel: esp0: Resetting scsi bus
Feb 11 01:17:12 huckle kernel: esp0: Gross error sreg=40
Feb 11 01:17:12 huckle kernel: esp0: SCSI bus reset interrupt
Feb 11 01:17:14 huckle kernel: esp0: Target will not disconnect
Feb 11 01:17:14 huckle kernel: esp0: resetting bus...
Feb 11 01:17:14 huckle kernel: esp0: SCSI bus reset interrupt
Feb 11 01:17:14 huckle kernel: esp0: SCSI bus reset interrupt
Feb 11 01:17:14 huckle kernel: esp0: Warning, live target 1 not
responding to selection.
Feb 11 01:17:15 huckle kernel: EXT2-fs error (device 08:14):
ext2_check_inodes_bitmap: Wrong free inod
es count in super block, stored = 259069, counted = 257029
Interestingly, after all the flury of messages passes, /home seems
to be mounted...
(unmounting causes just as many complaints from the system!)
Let's try to put sdb4 in the fstab and see if it will work better.
Big mistake...same bunch of messages with 20-30 SCSI bus reset
interrupt sprinkled just for fun...
Luckily, it eventually subsides and the boot process completes.
Anybody have any suggestions as to what I should do ?
(at some point, there is i/o errors...should I try to reformat
sdb...and how do I do that again?)
Is the weird partition 3 (that Whole disk thing) significant if I'm
not going to run solaris ?
Is that what's causing problems on sdb ?
Can anyone help ?
Thanks,
Benoit
--
| Benoit Lavigne, Newbridge Networks Corporation |
| Bin 10T3, 600 March Road, Kanata, Ontario, Canada, K2K 2E6 |
| Voice: (613)780-3535 ext. 3185 |
| e-mail:[EMAIL PROTECTED] |
