linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
@ 2001-06-30  5:41 pt
  2001-06-30  7:20 ` Alan Cox
  0 siblings, 1 reply; 6+ messages in thread
From: pt @ 2001-06-30  5:41 UTC (permalink / raw)
  To: linux-kernel


I have configured a RAID5 volume, partitioned it and created
ext2 on it. I can mount it and everything seems to be ok, but
trying to copy a large (+-500MB) file from one directory on
the filesystem on the RAID to another results in something like:

i2o/iop0: No handler for event (0x00000400)
i2o/iop0 requires user configuration
Driver "I2O Block OSM" did not release device!
i2ob_del_device called, but not in dev table!
Driver "I2O Block OSM" did not release device!

and the copying process is frozen. After that I can read
some data from the filesystem (tried to ls some directories)
but cannot umount it and cannot kill the copying process.
The event code in the message's first line sometimes has different
values (0x00000020 for example), the first line happens to
be repeated a few times as well.
The problem is fully reproducible with my system.

The system is RedHat 7.1, the kernel is compiled with SMP,
I2O block OSM, I2O PCI support and without I2O SCSI and I2O net OSMs.
SCSI support is also off. I enabled the enhanced RTC because a
somewhat dated document from Intel concerning support for this
controller with RH62 says it has to be on.
GCC and binutils versions are: egcs-2.91.66 (stock RH71 "kgcc"),
binutils-2.10.91.0.2 (stock RH71).

I have the controller working in an Intel L440GX+ board with
two PIII 700MHz processors. The board specs are: 82440GX chipset,
Cirrus Logic CLGD5480 VGA, Adaptec AIC-7896 SCSI (unused), Intel 82559
EtherPro100 (unused). The SRCU31 controller works in an ordinary PCI
slot (32bit/33mhz), the only other PCI card in the system (apart
from the on-board devices) is an Intel EtherPro100.
There are four IBM DDYS-T18350 drives on the controller, three of
them are in the raid. The controller has 32MB RAM as cache,
the system has 512MB of ECC RAM.

greetings,

Przemek Tomala

PS: if you need additional info or you can help with this,
    please CC me, as I am not subscribed to the list


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
  2001-06-30  5:41 Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18 pt
@ 2001-06-30  7:20 ` Alan Cox
  0 siblings, 0 replies; 6+ messages in thread
From: Alan Cox @ 2001-06-30  7:20 UTC (permalink / raw)
  To: pt; +Cc: linux-kernel

> the filesystem on the RAID to another results in something like:
> 
> i2o/iop0: No handler for event (0x00000400)
> i2o/iop0 requires user configuration

That bit is fine. The device is asking you to comfigure it and it also sent
a DDM availability chnage for some reason

> Driver "I2O Block OSM" did not release device!
> i2ob_del_device called, but not in dev table!
> Driver "I2O Block OSM" did not release device!

If you can repeat it in 2.4.5 let me know. I fixed a very similar problem
the supertrak 100 showed up. ALso set up the i2o cgi tools and see why
the device wants to talk to you

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
  2001-07-04 14:11 pt
@ 2001-07-04 14:16 ` Alan Cox
  0 siblings, 0 replies; 6+ messages in thread
From: Alan Cox @ 2001-07-04 14:16 UTC (permalink / raw)
  To: pt; +Cc: linux-kernel

> Function: Media Unlock
> Error Code: Access Violation
> User Info: 00000000
> Disk Serial No is not set, just no data in the table field
> Timestamps are somewhat strange - three erroror logs one after
> another then about 3-4 minutes delay and again three messages

What that means in the intel log I dont know. Media unlock is a command we
issue when you unmount an array. It allows volume removal if the volume can
be removed or errors otherwise (an error we happily ignore in that case)



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
@ 2001-07-04 14:11 pt
  2001-07-04 14:16 ` Alan Cox
  0 siblings, 1 reply; 6+ messages in thread
From: pt @ 2001-07-04 14:11 UTC (permalink / raw)
  To: linux-kernel


> modprobe i2o_config

> may be needed

Yes of course, why didn't I think about it...

Anyway, the cgi tool doesn't seem to identify any problems with
the controller before the freeze. Then it freezes together with
the copying process. And after reboot I see errors in the
controller's log - they are all exactly same:

Function: Media Unlock
Error Code: Access Violation
User Info: 00000000
Disk Serial No is not set, just no data in the table field
Timestamps are somewhat strange - three erroror logs one after
another then about 3-4 minutes delay and again three messages

The RAID volume sometimes enters a state after reboot where all
three disks are working and the cgi says the volume is
initializing.



Przemek Tomala



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
  2001-07-02  8:11 pt
@ 2001-07-02 10:18 ` Alan Cox
  0 siblings, 0 replies; 6+ messages in thread
From: Alan Cox @ 2001-07-02 10:18 UTC (permalink / raw)
  To: pt; +Cc: linux-kernel

> > ALso set up the i2o cgi tools and see why
> > the device wants to talk to you
> 
> Tried to setup the Intel tools just as I did it before and I get
> only an "Error: could not open I2O system" in the browser under the
> RH-supplied kernel. I will keep trying to resolve this problem.

modprobe i2o_config 

may be needed


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18
@ 2001-07-02  8:11 pt
  2001-07-02 10:18 ` Alan Cox
  0 siblings, 1 reply; 6+ messages in thread
From: pt @ 2001-07-02  8:11 UTC (permalink / raw)
  To: linux-kernel


> If you can repeat it in 2.4.5 let me know.

Yes, I can reproduce it in 2.4.5 with exactly the same behaviour and
messages. I made another experiment by installing RH7.1 directly
on the raid partition (it was not possible to install with my mobo
before because of a problem in the RH installer) and I couldn't
reproduce the problem in 2.4.2 supplied by RedHat.

> ALso set up the i2o cgi tools and see why
> the device wants to talk to you

Tried to setup the Intel tools just as I did it before and I get
only an "Error: could not open I2O system" in the browser under the
RH-supplied kernel. I will keep trying to resolve this problem.

greetings

Przemek Tomala


^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2001-07-04 14:17 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-06-30  5:41 Intel SRCU3-1 RAID (I2O) and 2.4.5-ac18 pt
2001-06-30  7:20 ` Alan Cox
2001-07-02  8:11 pt
2001-07-02 10:18 ` Alan Cox
2001-07-04 14:11 pt
2001-07-04 14:16 ` Alan Cox

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).