On Wed, May 12, 2021 at 11:50 AM Ilias Apalodimas
[off-list ref] wrote:

[...]

quoted

quoted

Since we added an extra argument on __skb_frag_unref() to handle
recycling, update the current users of the function with that.

This part could be done with a preliminary patch, only adding this
extra boolean, this would keep the 'complex' patch smaller.

Sure

[...]

quoted

quoted

 #include <linux/uaccess.h>
 #include <trace/events/skb.h>

@@ -645,6 +648,11 @@ static void skb_free_head(struct sk_buff *skb)
 {
    unsigned char *head = skb->head;

+#if IS_BUILTIN(CONFIG_PAGE_POOL)

Why IS_BUILTIN() ?

No reason, we'll replace it with an ifdef

quoted

PAGE_POOL is either y or n

IS_ENABLED() would look better, since we use IS_BUILTIN() for the cases where a module might be used.

Or simply #ifdef CONFIG_PAGE_POOL

quoted

+   if (skb->pp_recycle && page_pool_return_skb_page(head))

This probably should be attempted only in the (skb->head_frag) case ?

I think the extra check makes sense.

What do you mean here ?

quoted

Also this patch misses pskb_expand_head()

I am not sure I am following. Misses what? pskb_expand_head() will either
call skb_release_data() or skb_free_head(), which would either recycle or
unmap the buffer for us (depending on the page refcnt)

 pskb_expand_head() allocates a new skb->head, from slab.

We should clear skb->pp_recycle for consistency of the skb->head_frag
clearing we perform there.

But then, I now realize you use skb->pp_recycle bit for both skb->head
and fragments,
and rely on this PP_SIGNATURE thing (I note that patch 1 changelog
does not describe why a random page will _not_ have this signature by
bad luck)

Please document/describe which struct page fields are aliased with
page->signature ?

Thanks !