Hi, On Fri, Sep 01, 2000 at 03:35:33PM +0300, Matti Aarnio wrote: > > To repeat: (my theory) > > - Have a block device on which you can induce errors whenever > and what kind ever you want > - Make EXT2 on it > - Mount it, write something small on it, sync. > - Make the device fault on all read and write accesses > - append more to same files, sync I did this back in 2.0.2x time, and I think I caught all of the places where ext2 risked bombing on IO failures. It's quite possible that there are new faults in the 2.4 code, so I'll make a note to try to do a similar audit. Cheers, Stephen - To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to [EMAIL PROTECTED]
