# Re: \[EXT\] Re: \[PATCH v1 net\] page\_pool: Cap queue size to 32k.

- From: Jesper Dangaard Brouer \<hawk@kernel.org\>
- Date: 2023-08-14 08:46:40 
- List: netdev
- Page: https://ratatoskr.run/netdev/2023/08/2302179
- Thread: https://ratatoskr.run/netdev/2023/08/2301940/t.md

```text


On 14/08/2023 10.05, Ratheesh Kannoth wrote:
>> From: Johannes Berg <johannes@sipsolutions.net>
>> Subject: [EXT] Re: [PATCH v1 net] page_pool: Cap queue size to 32k.
>>> Please find discussion at
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__lore.kernel.org_l
>>>
>> I'm not the one who's going to apply this, but honestly, I don't think that will
>> work as a commit message for such a change ...
>> Sure, link to it by all means, but also summarize it and make sense of it for
>> the commit message?
>
> Why do you think it will not work ?. There is a long discussion about pros and cons of different approaches by
> Page pool maintainers in the discussion link.  However I  summarize it here, as commit message, it will
> Lead to some more questions by reviewers.
>

I agree with Johannes, this commit message is too thin.

It makes sense to give a summary of the discussion, because it show us
(page_pool maintainers) what you concluded for the discussion.

Further more, you also send another patch:
 - "[PATCH net-next] page_pool: Set page pool size"
 - https://lore.kernel.org/all/20230809021920.913324-1-rkannoth@marvell.com/

That patch solves the issue for your driver marvell/octeontx2 and I like
than change.

Why did you conclude that PP core should also change?

--Jesper
(p.s. Cc/To list have gotten excessive with 89 recipients)



```