Kernel Panic: SCSI Adaptec 2920 Problem

Kernel Panic: SCSI Adaptec 2920 Problem

Post by [SS]Strob » Tue, 04 May 1999 04:00:00


I encountered a frustrating problem with my scsi card:

On RH5.2, kernel 2.0.36, the scsi works fine with embedded drivers for
the future domain chipset, but, sometimes, my whole system locked for 10
seconds or so when a sysmessage like:

module <fdomain> idle request on 0x7000 irq 11,

11 rows of hexcode,
(11 sec effective idle time)

module <fdomain> resume request on 0x7000 irq 11,

(then my harddisk is ronking like hell!)
After that i could just play futher with my computer. The fact is, all
apm and scsi-idle-programs aren't running at all !
So what i did as solution: i turned off the scsi with a tool
(scsi-start/stop) because my linux was installed on /dev/hda3, i didn't
use the scsi in rh5.2, only for storage usage...

Now, i installed RH6.0, kernel 2.2.5 with Increased SCSI support (as
they decline !)
My scsi works fine, i installed the whole stuff on /dev/sda6, incluis a
swap partition, it still works fine untill ...

module <fdomain> idle request on 0x7000 irq 11,

11 rows of hexcode,
(11 sec effective idle time)

module <fdomain> resume request on 0x7000 irq 11...

But then, no resume ! only a damn kernel panic: NO REENTRANT INTERRUPTS.

I recompiled the kernel with all the scsiverbosity on, but it's very
frustrating because it always happens on heavy diskusage, the only thing
it can safe is the on-off button, a spectactular e2fschk with LOTS of
errors on the disk...

Somebody similar problems ?
Anybody hints ?

I *really* can't scsi-stop the scsi now, it wouldn't do my linux
anygood,not ? ;-)





1. Help: Adaptec 2920, Quantum Fireball, 2.0.xx, Kernel panic


  I've been having some mean problems with my Quantum
Disk (A 3.2 GB Quantum Fireball) under Linux.  Basically,
whenever I start writing a lot of data to it, the kernel
panics with IO errors from the disk.  I've tried this with
2.0.27 and 2.0.30, and in both cases, kablooie.

  I booted my machine under WinBlows and installed
the SCSI drivers from Adaptec.  There, I filled the
Quantum with a copy of my WinBlows disk (and yes I did
paritition it into a 2 GB and a 1GB disk...).  So I know
that the hardware is probably okay.

  Actually, my linux problem occurs specifically under
these conditions:

  1. 1 Linux Swap Paritition and 1 Linux Native paritition
     on disk
  2. Installing Accel-X CDE 1.1 or Accel-X CDE 1.0.10.

  I talked to Accel-X tech support about this, they said
that the installation program is a script, so I doubt that
the problem is specific to Accel-X.  I think the panics
come around when the system starts to swap.  I could
easily be wrong..

  Here's my system configuration:

Cyrix 6x86-150MHz
  64 MB RAM
  512K Cache
  2GB IDE Disk (IBM)
  3.2GB SCSI Disk (Quantum) - Not functional!!
  Number Nine Motion 331 2MB DRAM
  Exabyte Eagle Tr-3 Tape Drive
  Motorola ModemSurfr 28.8 Modem (NOT in PnP mode)
  Sound Blaster 16 PnP
  Adaptec 2920 PCI SCSI interface
  3Com 3c905-Tx PCI 10BaseT Ethernet board (installed after problems
had first occurred)

  I'd greatly appreciate any and all help I can get.  Also, I don't
have much time to read through the newsgroups (maybe once every three
weeks) and my news server expires very quickly.. So I ask those kind

  Thanks in advance!

Lally Singh -------------------------------------------
Internet Technology Director | Contractor
Geoloc Corporation           | US Naval Research Lab    |
(703) 812-8500               | (202) 404-2540
--Linux: Where Reboots Are For Hardware Upgrades-------

2. Ethernet 100BaseT adaptor recommendations?

3. Adaptec 2920 with kernel panic

4. question on program

5. Adaptec 2920 with CD-ROM create kernel panic

6. Add rename_region().

7. 2.0.30 & Kernel Panics with the Adaptec 2920

8. 7500 install - cliffdiving moron

9. Traxdata and Adaptec 2920 Kernel Panic

10. Adaptec 2920 with kernel panic

11. Need Kernel which supports Adaptec 2920 PCI SCSI adapeter

12. Both Adaptec 2940 and 2920 cards (no 2920!)

13. Adaptec 2920 scsi card problems