Ext3 users - Dec 2001 - crashing ext3 is easy and reproducible

This is my crash script for 2.4.17-rc1 and rc2 

It works exactly as expected first time. 

Second time your system is toast with an exception error. 
I think it kills the daemon. 

I've tried it with varying ram disk sizes, up to 256Mb and 
in UP and SMP mode. 

The only thing I've seen was putting two of them into a script 
with the intention of having a one stop crasher didn't work,-- at least
once. 

I've never had it fail to crash when run manually. 
I wonder if a time delay is part of it??? 

The crash occurs in the rm -rf statement and produces enough verbiage 
as to scroll the top of the oops off the top of the screen. 

Regards 

john 

#!/bin/sh
#hisdad's script of death

mkfs /dev/ram0
tune2fs -j /dev/ram0
mount -t ext3 /dev/ram0 /mnt/ramdisk

rm -rf /mnt/ramdisk/lost+found                <==== oops on 2nd time
umount /mnt/ramdisk

John Huttley wrote:
> 
> This is my crash script for 2.4.17-rc1 and rc2
> 
> It works exactly as expected first time.
> 
> Second time your system is toast with an exception error.
> I think it kills the daemon.
> 
> I've tried it with varying ram disk sizes, up to 256Mb and
> in UP and SMP mode.
> 
> The only thing I've seen was putting two of them into a script
> with the intention of having a one stop crasher didn't work,-- at least
> once.
> 
> I've never had it fail to crash when run manually.
> I wonder if a time delay is part of it???
> 
> The crash occurs in the rm -rf statement and produces enough verbiage
> as to scroll the top of the oops off the top of the screen.
> 
whoops.   I recall back in April deciding that the ramdisk
driver was incompatible with ext3, but I forget why I
decided that.  Plus the rd driver has changed a lot since
then.

I'll take a look into this, see what can be done.  Am I
correct in assuming that it works OK for you on a real disk?

-