On 2007-08-20 13:22 +1200, Charles Manning wrote:
> There  were a few problems in checkpointing that were fixed in rev 1.49 that 
> are likely causing the problem.

Were these such that if I make a new kernel with the new yaffs code it
will not-crash on seeing the poorly filesystem, or would it merely
prevent the FS getting into a poorly state in the first place?

I suppose we should try and arrange things so that the kernel never
crashes, no matter how shafted the FS.

I'll try an updated kernel first and see what happens. 

> Curious... why is the device being remounted readonly?

Which bit of the log shows that it is? - I failed to spot it.

I'm not expecting it to be RO.

Wookey
-- 
Principal hats:  Balloonz - Toby Churchill - Aleph One - Debian
http://wookware.org/