On Thu, 2012-03-29 at 19:34 -0400, Josh Cartwright wrote:
> The locking policy is such that the erase_complete_block spinlock is
> nested within the alloc_sem mutex.  This fixes a case in which the
> acquisition order was erroneously reversed.  This issue was caught by
> the following lockdep splat:

I've added the follwing:

This was introduce in '81cfc9f jffs2: Fix serious write stall due to
erase'.
    
Cc: stable@kernel.org [2.6.37+]

and pushed to l2-mtd.git, thanks! CC-in Joakim to let him know that his
patch has a bug and that he might want this fix in his system(s) as
well.

-- 
Best Regards,
Artem Bityutskiy