Re: [PATCH net] page_pool: fix inconsistency for page_pool_ring_[un]lock()

[Jesper Dangaard Brouer]

On 23/05/2023 04.13, Yunsheng Lin wrote:
> Do you still working on optimizing the page_pool destroy
> process? If not, do you mind if I carry it on based on
> that?

I'm still working on improving the page_pool destroy process.
I prefer to do the implementation myself.

I've not submitted another version, because I'm currently using the
workqueue to detect/track a memory leak in mlx5.

The mlx5 driver combined with XDP redirect is leaking memory, and will
eventually lead to OOM.  The workqueue warning doesn't point to the
actual problem, but it makes is easier to notice that there is a problem.

--Jesper