kernel panic due to swap problem or broken SCSI driver or ???

kernel panic due to swap problem or broken SCSI driver or ???

Post by Dennis G. Allar » Fri, 22 Aug 1997 04:00:00



I'm running Red Hat 4.1 full install
kernel 2.0.27.
Intel DX4-100 CPU
64MB RAM
100MB swap partition
Adaptec AHA-2842VL Vesa Local Bus SCSI
No IDE controller
Quantum 3.2 Gig SCSI hard drive (nice drive, by the way, real quiet)
JAZ 1 Gig cartridge drive
Fujitsu 128MB Magneto Optical cartridge drive
Phillips 8X CD-ROM drive
3COM ethernet LAN connection to NT box via Samba, X, telnet, etc.

[Begin Caveat.
I originally wanted to configure to a Cyrix P166/PCI with an Adaptec
Ultra-wide SCSI card but just could not get RedHat install to go
through.
So, I stuck with my good ole 486DX4-100 - been running Linux for two
years, originally on Slackware (which gave me NO critical problems).
I decided to move to RedHat when the need for ELF binaries arose,
deciding I did not wish to hassle an in-place move to ELF and having
read that RedHat RPM was cool).
End Caveat.]

The system runs fine for most purposes.  But.  It hoses up at
various times which seem to be related to heavy use of memory
by either large Lisp processes or when I am copying very large files
to my JAZ drive (which I wish to use for daily backups).

I hate to be vague, but the problem or problems manifest themselves
in different ways.

Here are a couple of exemplars which appeared on one of my pseudo
terminals about the time the system hosed up or started to hose up...

I have LITTLE idea of what any of this means.  I'm just copying
it to here in the hopes that a guru will see this and have some
wisdom to impart...

Exemplar 1:

wait-queue is bad (eip=001149f6)
     q=018fa528
    *q=0154bfb0
Unable to handle kernel NULL pointer dereference at virtual address
C0000004
current->tss.cr3=0058f000,%cr3=0058f000
*pde=00102067
*pte=00000027
Ooops: 0000
CPU:     0
EIP:     0010:[<00115614>]
.... (further hex-like output omitted)

Exemplar 2:

aic7xxx: (reset_channel) Channel reset, requencer restarted
aic7xxx: (done_aborted_scbs) Aborting scb0, TCL=5/0/0
aic7xxx: (done_aborted_scbs) Aborting scb1, TCL=5/0/0
aic7xxx: (done_aborted_scbs) Aborting scb2, TCL=0/0/0
aic7xxx: (done_aborted_scbs) Aborting scb3, TCL=0/0/0
scsi0: Target 5, channelA, now synchronous at 10.0 MHz, offset 15
scsi: aborting command due to time out: pid 18602, scsi0, channel 0,
id5, lun 0 write(6) 06 8c 6c f4 00
aic7xxx: (abort) Aborting scb1, TCL 5/0/0
SCSI host 0 abort (pid 18602) timed out - resetting
SCSI bus is being reset for host 0 channel 0
aic7xxx: (reset) target/channel 5/0
aic7xxx: (abort_reset) scb state 0x1, Message_In phase, SCSISIGI=0xe6
....
scsi0: No active CB for reconnecting target 5, channel A - Issuing
ABORT...
Kernel panic: scsi 0 : Target 5, channel A, did not send an IDENTITY
message.  SAVED_TCL 0x50
In swapper task - not synching

Any clues as to how I should proceed to diagnose, things to try,
etc., would be much appreciated.

--
Dennis Allard
http://oceanpark.com

 
 
 

1. Sparcstation-2 PANICS due to 32-bit driver built with Forte 6.2 C compiler

Any help with the following problem will be greatly appreciated:

A 32 bit driver built with the Forte 6 Update 2 C compiler with the
default options for the target architecture instruction set
(i.e. -xarch=generic) seems to cause a system PANIC on a SparcStation-2.
However, it seems to work fine on an UltraSparc-5 running Solaris 2.x
(x >= 6) in 32-bit mode. Furthermore, the problem seems to disappear when
the driver is built with the same compiler with the `xarch' option set to
`v7'.

I was wondering if this is a known issue with the compiler, etal?

Additionally, the PANIC seems to happen when the driver tries to do an
integer multiplication or division operation, which results in an illegal
instruction trap.

Thanks,

-- Karthik.

2. sar -q .. possible wrong

3. Kernel 2.4.7 does not boot due a problem involving Adaptec AIC-7896 SCSI

4. Allow Anyone to Cancel a Request - urgent pls

5. Kernel panic due to NF_IP_LOCAL_OUT handler calling itself again

6. Help Redhat 5.1 Backpack CDROM

7. IWILL SCSI card - Kernel panic, I panic:-)

8. Debian say's goodbye!

9. change to drivers/scsi/scsi.c as of 1.3.9 causes panic

10. Kernel 2.0.33 + SCSI - Kernel panic

11. Kernel Panic with aic7xxx driver 6.2.1 in kernel 2.4.9

12. aic7xxx driver panics under heavy swap.

13. PROBLEM: Kernel Panic Promise driver