Dear all, dear ekpulse,
if you do have a hardware problem, then it looks like I have it too
because my machine behaves exactly as you describe.
However, after the last thread and my post I did some more testing as
indicated by someone else on this list. I hooked up the NSLU2
directely to my laptop with a very short xover cable and... sure
enough, the problems remained. Same things happened (oopsing and
crashing while copying very large files).
Then I did a different test to understand whether it's a problem with
the network. I attempted to do several copies of a 1.2 Gbyte file from
the NSLU2 disk right into the same disk, no network involved and...
yes, oopsed and crashed again. I rebooted, fsck'ed the disk, tried to
remove whatever was left from the crashed copies and... crashed again.
Finally performing one remove at a time I succeded to remove the copies.
The oopses from copying/removing both start with:
Unable to handle kernel paging request at virtual address 00200000
pgd = c1830000
Internal error: Oops: f5 [#1]
Modules linked in: tcp_diag ixp425_eth ixp400 ext3 jbd ipv6 ext2 mbcache
and then, I noticed a (perhaps) strange thing. The 'free' command gave
me, right after each crash, what follows:
total used free shared buffers
Mem: 30680 29548 1132 0 284
Swap: 56216 0 56216
Total: 86896 29548 57348
That is: RAM almost to capacity but swap untouched. Could it possibly
be that the kernel crashes at the moment of swapping because of some
problem with the swapfile or something? I have properly formatted the
swap space with -v1, and the disk does not have bad blocks. But other
times (like when running fsck, I can see the swap being used) so this
could be completely wrong.
ekpulse, do you observe oops traces in your messages file
(/var/log/messages, look for 'Oops')? It would be really good if we
could pin this down. Any other hint?
P.S. I have also installed and modprobe'd the tcp_diag module, but I
don't have a clue on how to use it.
P.S. My machine does exactly the same boot sequence as the one you
posted. Does'nt seem to be a problem (but I might be wrong).
--- In email@example.com, "epkulse" <uffe.z@t...> wrote:
> I´m now trying to run Openslug. Have previously tried FW r24 and r63.
> Also tried Unslung BETA 5.5. All the time, though, I have encountered
> instability of all kinds. In the beginning mainly when handling disk
> accesses. But now, even when setting up Openslug it is unstable. Lots
> of different interesting situations. Anyway - I´m more and more
> convinced that something is incorrect in the HW. I have enclosed some
> errors that I encounter in the dmesg - perhaps someone can verify if
> these errors could indicate a faulty HW?
> Linux version 220.127.116.11 (openslug@o...) (gcc version 3.4.4) #1
> Mon Aug 8 08:52:36 PDT 2005
> CPU: XScale-IXP42x Family [690541f1] revision 1 (ARMv5TE)
> CPU0: D VIVT undefined 5 cache
> CPU0: I cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
> CPU0: D cache: 32768 bytes, associativity 32, 32 byte lines, 32 sets
> Machine: Linksys NSLU2
> Warning: bad configuration page, trying to continue
> Memory policy: ECC disabled, Data cache writeback
> On node 0 totalpages: 8192
> DMA zone: 8192 pages, LIFO batch:3
> Normal zone: 0 pages, LIFO batch:1
> HighMem zone: 0 pages, LIFO batch:1
> Built 1 zonelists
> Kernel command line: root=/dev/mtdblock4 rw rootfstype=jffs2
> mem=32M@0x00000000 init=/linuxrc reboot=s noirqdebug
> ixp400: module license 'unspecified' taints kernel.
> Module init.
> Initializing IXP425 NPE Ethernet driver software v. 1.1A
> ixp425_eth: CPU clock speed (approx) = 0 MHz
> [error] ixEthMiiPhyScan : unexpected Mii PHY ID 00008201
> ixp425_eth: npe0 is using the PHY at address 0
> ixp425_eth: npe1 is using the PHY at address 1
> NET: Registered protocol family 17
> ixp425_eth: ixEthMiiLinkStatus failed on PHY0.
> Can't determine
> the auto negotiated parameters. Using default values.