Skip to contentSkip to conversation

Initio miles tests :-(

DISCUSSION [email protected] 2 participants 3 messages May 24, 1999

lore
Share & tools
lore
#0 lore raw

OK so far it's not really good...

Let's begin by what's working...

mke2fs /dev/disk-on-miles is ok
fsck too

mount -t hfs
mount -t ext2 as well

cp a 2M file (kernel) ok, not corrupted when cp back to mesh (gzip'ed and de-gzip'ed succesfully)

umount ok after cping this small file.

Now what doesn't work...

pdisk /dev/disk-on-miles destroy the partition map if re-written, even just renaming a partition

mke2fs -c gives a kernel panic ("kernel stack overflow")

cp /usr/src/linux/* /mnt/disk-on miles/ goes ok few... and then I got this kind of lines (several times):
scsi: aborting command due to timeout: pid 20175, scsi1, channel 0, id5, lun0, Write (6) 04 9C 45 80 00
SCSI bus is being reset for host 1 channel 0
the kernel "panics" soon after

I got his behaviour also when cping 10 megs of files (/boot/*) and issuing umount when done.

Finally hdparm -Tt gives a kernel panic, or smth just under

Each time the ext2 partition is completely screwed, I must reformat to continue tests, the hfs is screwed once over 2.
I think I will go on with a ramdisk !

So I guess the patch wasn't enough...

Any way of investigation someone ?

Signature · 7 lines
Guillaume Laurès - student @ ESIEA (www.esiea.fr)


[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]
#1 lore raw
Quoted 3 lines · On Mon, 24 May 1999 [email protected] wrote:

Each time the ext2 partition is completely screwed, I must reformat to continue tests, the hfs is screwed once over 2.
I think I will go on with a ramdisk !

Can you do all the stress testing if you don't pdisk the drive? Also can you go:
pdisk /dev/blank-inito-disk
... create partitions ...
stress testing
?

---
Tom Rini (TR1265)
http://gate.crashing.org/~trini/

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]] [[ not forced back to the list, so be sure to Cc linuxppc-dev if your ]] [[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting. ]]

#2 lore raw
Quoted 3 lines · Tom Rini

Can you do all the stress testing if you don't pdisk the drive? Also can

What I did: partitionned the disk under MacOs, two hfs, the second mke2fsck'ed in linux after insertion of the module. That's basically what I'm always doing, see below why.

Quoted 3 lines

you go:
pdisk /dev/blank-inito-disk
... create partitions ...

This is not possible since pdisk reports a incredible number of 512 bytes block for a 256M disk (the test drive), with a total free space of 126G!!!
Here pdisk fail at creating the partition map, and at modifying it sjightly.
I wish I could show you the output of pdisk, is it possible to redirect output on a file, still setting the keyborad as input ?

Thx for your help :-)

Signature · 7 lines
Guillaume Laurès - student @ ESIEA (www.esiea.fr)


[[ This message was sent via the linuxppc-dev mailing list.  Replies are ]]
[[ not  forced  back  to the list, so be sure to Cc linuxppc-dev if your ]]
[[ reply is of general interest. Please check http://lists.linuxppc.org/ ]]
[[ and http://www.linuxppc.org/ for useful information before posting.   ]]