[PATCH v3] net: use bulk free in kfree_skb_list

[PATCH v3] net: use bulk free in kfree_skb_list

[Felix Fietkau]

Since we're freeing multiple skbs, we might as well use bulk free to save a
few cycles. Use the same conditions for bulk free as in napi_consume_skb.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
v3: reorder checks to prevent skb double unref
v2: call kmem_cache_free_bulk once the skb array is full instead of
    falling back to kfree_skb
 net/core/skbuff.c | 40 ++++++++++++++++++++++++++++++++++++----
 1 file changed, 36 insertions(+), 4 deletions(-)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 2415d9cb9b89..ca0308485669 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -666,12 +666,44 @@ EXPORT_SYMBOL(kfree_skb);
 
 void kfree_skb_list(struct sk_buff *segs)
 {
-	while (segs) {
-		struct sk_buff *next = segs->next;
+	struct sk_buff *next = segs;
+	void *skbs[16];
+	int n_skbs = 0;
 
-		kfree_skb(segs);
-		segs = next;
+	while ((segs = next) != NULL) {
+		next = segs->next;
+
+		if (segs->fclone != SKB_FCLONE_UNAVAILABLE) {
+			kfree_skb(segs);
+			continue;
+		}
+
+		if (!skb_unref(segs))
+			continue;
+
+		trace_kfree_skb(segs, __builtin_return_address(0));
+
+		/* drop skb->head and call any destructors for packet */
+		skb_release_all(segs);
+
+#ifdef CONFIG_SLUB
+		/* SLUB writes into objects when freeing */
+		prefetchw(segs);
+#endif
+
+		skbs[n_skbs++] = segs;
+
+		if (n_skbs < ARRAY_SIZE(skbs))
+			continue;
+
+		kmem_cache_free_bulk(skbuff_head_cache, n_skbs, skbs);
+		n_skbs = 0;
 	}
+
+	if (!n_skbs)
+		return;
+
+	kmem_cache_free_bulk(skbuff_head_cache, n_skbs, skbs);
 }
 EXPORT_SYMBOL(kfree_skb_list);
 
-- 
2.17.0

Re: [PATCH v3] net: use bulk free in kfree_skb_list

[Eric Dumazet]

On 03/25/2019 02:14 AM, Felix Fietkau wrote:
> Since we're freeing multiple skbs, we might as well use bulk free to save a
> few cycles. Use the same conditions for bulk free as in napi_consume_skb.
> 
> Signed-off-by: Felix Fietkau <nbd@nbd.name>
> ---
> v3: reorder checks to prevent skb double unref
> v2: call kmem_cache_free_bulk once the skb array is full instead of
>     falling back to kfree_skb
>  net/core/skbuff.c | 40 ++++++++++++++++++++++++++++++++++++----
>  1 file changed, 36 insertions(+), 4 deletions(-)
> 
> diff --git a/net/core/skbuff.c b/net/core/skbuff.c
> index 2415d9cb9b89..ca0308485669 100644
> --- a/net/core/skbuff.c
> +++ b/net/core/skbuff.c
> @@ -666,12 +666,44 @@ EXPORT_SYMBOL(kfree_skb);
>  
>  void kfree_skb_list(struct sk_buff *segs)
>  {
> -	while (segs) {
> -		struct sk_buff *next = segs->next;
> +	struct sk_buff *next = segs;
> +	void *skbs[16];
> +	int n_skbs = 0;
>  
> -		kfree_skb(segs);
> -		segs = next;
> +	while ((segs = next) != NULL) {
> +		next = segs->next;
> +
> +		if (segs->fclone != SKB_FCLONE_UNAVAILABLE) {
> +			kfree_skb(segs);
> +			continue;
> +		}
> +
> +		if (!skb_unref(segs))
> +			continue;
> +
> +		trace_kfree_skb(segs, __builtin_return_address(0));
> +
> +		/* drop skb->head and call any destructors for packet */
> +		skb_release_all(segs);
> +


> +#ifdef CONFIG_SLUB
> +		/* SLUB writes into objects when freeing */
> +		prefetchw(segs);
> +#endif

This is done too late : 
You should probably either remove this prefetchw()
or do it before reading segs->next at the beginning of the loop.

> +
> +		skbs[n_skbs++] = segs;
> +
> +		if (n_skbs < ARRAY_SIZE(skbs))
> +			continue;
> +
> +		kmem_cache_free_bulk(skbuff_head_cache, n_skbs, skbs);
> +		n_skbs = 0;
>  	}
> +
> +	if (!n_skbs)
> +		return;
> +
> +	kmem_cache_free_bulk(skbuff_head_cache, n_skbs, skbs);
>  }
>  EXPORT_SYMBOL(kfree_skb_list);
>  
> 

Re: [PATCH v3] net: use bulk free in kfree_skb_list

[Jesper Dangaard Brouer]

On Mon, 25 Mar 2019 02:27:14 -0700
Eric Dumazet <eric.dumazet@gmail.com> wrote:

> > +#ifdef CONFIG_SLUB
> > +		/* SLUB writes into objects when freeing */
> > +		prefetchw(segs);
> > +#endif  
> 
> This is done too late : 
> You should probably either remove this prefetchw()
> or do it before reading segs->next at the beginning of the loop.

Agree. Not sure the prefetchw optimization makes sense here. IHMO just
drop it in this patch.

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer