[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Today's bug...

dwmw2@xxxxxxx.org said:
>  Somehow, jffs_add_node() is being called with node->fm == NULL.
> I threw in code to detect that and return -EIO, and strangely, the
> filesystem still appears to work.

I'm fairly concerned by the fact that my symptomatic fix actually appears 
to work. As far as I'm concerned, it just shouldn't.

Are there circumstances under which it's permitted for a jffs_node->fm to 
be NULL when that node is split by jffs_delete_data(), and if so, what is 
the expected behaviour of jffs_add_node() when it's asked to add the second 
half of the original node into the list?