From: Alexander H Duyck <redacted>
Date: Wed, 31 May 2023 08:21:03 -0700

On Tue, 2023-05-30 at 17:00 +0200, Alexander Lobakin wrote:

quoted

Currently, touching <net/page_pool.h> triggers a rebuild of more than
a half of the kernel. That's because it's included in <linux/skbuff.h>.

In 6a5bcd84e886 ("page_pool: Allow drivers to hint on SKB recycling"),
Matteo included it to be able to call a couple functions defined there.
Then, in 57f05bc2ab24 ("page_pool: keep pp info as long as page pool
owns the page") one of the calls was removed, so only one left.
It's call to page_pool_return_skb_page() in napi_frag_unref(). The
function is external and doesn't have any dependencies. Having include
of very niche page_pool.h only for that looks like an overkill.
Instead, move the declaration of that function to skbuff.h itself, with
a small comment that it's a special guest and should not be touched.
Now, after a few include fixes in the drivers, touching page_pool.h
only triggers rebuilding of the drivers using it and a couple core
networking files.

Signed-off-by: Alexander Lobakin <aleksander.lobakin@intel.com>
---
 drivers/net/ethernet/engleder/tsnep_main.c               | 1 +
 drivers/net/ethernet/freescale/fec_main.c                | 1 +
 drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c | 1 +
 drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c     | 1 +
 drivers/net/ethernet/mellanox/mlx5/core/en/params.c      | 1 +
 drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c         | 1 +
 drivers/net/wireless/mediatek/mt76/mt76.h                | 1 +
 include/linux/skbuff.h                                   | 4 +++-
 include/net/page_pool.h                                  | 2 --
 9 files changed, 10 insertions(+), 3 deletions(-)

<...>

quoted

diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 5951904413ab..6d5eee932b95 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h

@@ -32,7 +32,6 @@
 #include <linux/if_packet.h>
 #include <linux/llist.h>
 #include <net/flow.h>
-#include <net/page_pool.h>
 #if IS_ENABLED(CONFIG_NF_CONNTRACK)
 #include <linux/netfilter/nf_conntrack_common.h>
 #endif

@@ -3422,6 +3421,9 @@ static inline void skb_frag_ref(struct sk_buff *skb, int f)
 	__skb_frag_ref(&skb_shinfo(skb)->frags[f]);
 }
 
+/* Internal from net/core/page_pool.c, do not use in drivers directly */
+bool page_pool_return_skb_page(struct page *page, bool napi_safe);
+
 static inline void
 napi_frag_unref(skb_frag_t *frag, bool recycle, bool napi_safe)
 {

diff --git a/include/net/page_pool.h b/include/net/page_pool.h
index 126f9e294389..2a9ce2aa6eb2 100644
--- a/include/net/page_pool.h
+++ b/include/net/page_pool.h

@@ -240,8 +240,6 @@ inline enum dma_data_direction page_pool_get_dma_dir(struct page_pool *pool)
 	return pool->p.dma_dir;
 }
 
-bool page_pool_return_skb_page(struct page *page, bool napi_safe);
-
 struct page_pool *page_pool_create(const struct page_pool_params *params);
 
 struct xdp_mem_info;

So the code as-is works, so I am providing my "Reviewed-by".
Reviewed-by: Alexander Duyck <alexanderduyck@fb.com>

Consider the rest of this a suggestion or a "nice to have".

I wonder if we shouldn't also look at restructuring the function and
just moving it to net/core/skbuff.c somewhere next to skb_pp_recycle.

I suspect we could look at pulling parts of it out as well. The
pp_magic check should always be succeeding unless we have pages getting
routed the wrong way somewhere. So maybe we should look at pulling it
out and moving it to another part of the path such as
__page_pool_put_page() and making it a bit more visible to catch those
cases.

I've just noticed that this function is exported with no modular users ._.
Anyway, I feel like it's a good way to go. The entire function, apart
from the magic check, can be moved and made static. And the magic can be
moved one level up, right...
v4 will happen either way I guess, so maybe I'll replace this patch with
that kinda change.

Thanks,
Olek