JFFS2 eats memory

[David Woodhouse]

On Mon, 2004-07-19 at 17:15 +0200, Øyvind Harboe wrote:
> At this point I'm just trying to find a fix that is correct. 

Well yes, but this _is_ a performance optimisation you're doing :)

> > Can't you adjust gc_node instead?
> 
> I don't know.
> 
> What should I adjust it to?

gc_node is the next node to be garbage-collected. We don't need to
garbage-collect nodes which are already obsolete. So if you're freeing
the object which is currently pointed to by jeb->gc_node, you can just
make gc_node point to the next_phys node which you're _not_ freeing.

-- 
dwmw2