Christopher Reder wrote:
In trying to track down potential ram issues, I was messing around with
different devices. I have a ram chip, 32meg on the bus of a at91rm9200.
Is there a way to view the ram memory? When I try cat /dev/mem, I get a
bad virtual address and segmentation fault.
man mem suggests this is the way to examine the system memory and it
actually works on my system (hexdump -C -n 512 /dev/mem).
When I try a cat /dev/ram0 ,
IMHO, this not what you think it is - interface to the system ram. This
device entry is used by the RAM disk driver that uses part of the system
memory as a block device. But, maybe it could be helpful to diagnose
your memory problems with the use of RAM disk to claim the system memory
(then seen as a block device) and then filling it with known values.
the cursor will blink like it is working
and I can see that the kernel hasn’t crashed. Eventually, it fails as
well, but on a null virtual address. I have the output below and have
cut it short so as to not include too much. Can someone tell me if this
is normal and if it isn’t, what could be the cause of it (setup issue
with kernel etc). I am running 2.6.20, with the AT91 patches.
Not normal.
# cat /dev/ram0
<1>Unable to handle kernel NULL pointer dereference at virtual address
00000000
[ 459.880000] pgd = c1f5c000
[ 459.880000] [00000000] *pgd=21f44031, *pte=00000000, *ppte=00000000
[ 459.880000] Internal error: Oops: 17 [#2]
This line suggests that something is very wrong. It caused kernel to
oops. See linux/Documentation/oops-tracing.txt and do the google search
on how to decode the oops message to something more readable so that
people can comment with use of ksymoops utility.
I don't know if it is usable for the WHOLE system memory, but I've used
a utility called devmem2.c, for inspection I/O register on my ARM
device. It might be of use - see google for the source file.
HTH,
Hinko
--
ČETRTA POT, d.o.o., Kranj
Planina 3
4000 Kranj
Slovenia, Europe
Tel. +386 (0) 4 280 66 03
E-mail: hinko.kocevar@xxxxxxxxxxxx
Http: www.cetrtapot.si
--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx
Please read the FAQ at http://kernelnewbies.org/FAQ
