xfs
[Top] [All Lists]

Re: Possible problem: 2.4.17cvs oops.

To: David Chambers <davidc@xxxxxxxxxxxxx>
Subject: Re: Possible problem: 2.4.17cvs oops.
From: Steve Lord <lord@xxxxxxx>
Date: 14 Jan 2002 17:39:16 -0600
Cc: linux-xfs@xxxxxxxxxxx
In-reply-to: <20020111214715.4e67c0f9.davidc@ccmi.salk.edu>
References: <20020111214715.4e67c0f9.davidc@ccmi.salk.edu>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Fri, 2002-01-11 at 23:47, David Chambers wrote:
> Hi!
> 
> I do not know whether any of the following is of any use, and no core file
> was generated, but here goes:
> 
> System is a Tyan motherboard, dual PIII, VIA chipset.  Running 1 x Seagate
> disk on hda, 8 x Maxtor Diamondmax 80Gb as RAID 5 controlled by a 3ware
> 7850 card. NIC is 3com 3C905B.  1Gb RAM.  All partitions are XFS.
> 
> Red Hat 7.2, Kernel 2.4.17 from the XFS 2.4.17-cvs tree on Jan. 9th.
> 3c905 is a module, 3w-xxxx is also a module.
> 
> During testing, copying about 60GB from an NFS mounted filesystem to the
> 3ware array I get the messages appended below.
> 
> I am a total newbie at kernel diagnostics.  I do not know what to do
> to proceed with this!  Is this report of any use?? What can/should I do
> to be more helpful?
> 
> - David
> 
> In syslog: (the "memory shortage" occurs whether 3c905 is compiled
> into the kernel or is a module.  I'm a little suspicious of it!)

If an XFS memory allocation failed, then it is possible it was the root
cause of the oops here. XFS comes from an environment where a memory
allocation does not fail as it can on Linux.

Not sure I have a fix for you right now, but it does tell me this type
of thing can still happen.

Steve


-- 

Steve Lord                                      voice: +1-651-683-3511
Principal Engineer, Filesystem Software         email: lord@xxxxxxx


<Prev in Thread] Current Thread [Next in Thread>