Re: [PATCH v10 01/19] zram: sleepable entry locking

[Sergey Senozhatsky]

On (26/07/07 12:02), Sebastian Andrzej Siewior wrote:
[..]
> If you have zram0 and zram1 and both have different lock chains which
> could result in a deadlock from theoretical point of view but not in
> reality because the zram0-swap never clash with zram1-ext4 then you
> should have a dynamic key. That means add it struct zram and register
> with lockdep_register_key() and assign with the lock as you did.

Sounds good.  Do you have time to send a patch or do you prefer me to
cook one (giving you the credit)?