Re: [PATCH net-next 5/9] page_pool: don't use driver-set flags field directly

From: Jakub Kicinski
Date: Thu Aug 03 2023 - 12:01:11 EST

On Thu, 3 Aug 2023 16:56:22 +0200 Alexander Lobakin wrote:
> > FWIW I'm tempted to do something like the patch below (an obvious move,
> > I suspect). I want to add another pointer (netdev) to the params and
> Just take napi->dev as I do in libie :)

:) The fields have extra semantics, like napi implies that recycling
is allowed, and netdev implies that there is only _one_ netdev eating
from the PP. There's also a way to get the pp <> netdev from the memory
model registration. But I feel like explicit field is cleanest.

Anyone, conversation for a later time :)

> > I don't want it to eat up bytes in the first cache line.
> > The patch is incomplete, we need to stash a one-bit indication in
> > the first cache line to know init_callback is not present without
> > having to look at @slow. I'll defer doing that cleanly until your
> > patches land.
> I would propose to include it in the series, but it has grown a bunch
> already and it's better to do that later separately :s

Yeah.. I'd be trying to split your series up a little to make progress
rather than add more things :( I was going to suggest that you post
just the first 3 patches for instance. Should be an easy merge.