Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Mon, 11 Jan 2021 18:29:00 +0000
To:     "David S. Miller" <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>
From:   Alexander Lobakin <alobakin@pm.me>
Cc:     Eric Dumazet <edumazet@google.com>,
        Edward Cree <ecree@solarflare.com>,
        Jonathan Lemon <jonathan.lemon@gmail.com>,
        Willem de Bruijn <willemb@google.com>,
        Miaohe Lin <linmiaohe@huawei.com>,
        Alexander Lobakin <alobakin@pm.me>,
        Steffen Klassert <steffen.klassert@secunet.com>,
        Guillaume Nault <gnault@redhat.com>,
        Yadu Kishore <kyk.segfault@gmail.com>,
        Al Viro <viro@zeniv.linux.org.uk>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org
Reply-To: Alexander Lobakin <alobakin@pm.me>
Subject: [PATCH net-next 3/5] skbuff: reuse skbuff_heads from flush_skb_cache if available
Message-ID: <20210111182801.12609-3-alobakin@pm.me>
In-Reply-To: <20210111182801.12609-1-alobakin@pm.me>
References: <20210111182655.12159-1-alobakin@pm.me> <20210111182801.12609-1-alobakin@pm.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Instead of unconditional allocating a new skbuff_head and
unconditional flushing of flush_skb_cache, reuse the ones queued
up for flushing if there are any.
skbuff_heads stored in flush_skb_cache are already unreferenced
from any pages or extensions and almost ready for use. We perform
zeroing in __napi_alloc_skb() anyway, regardless of where did our
skbuff_head came from.

Signed-off-by: Alexander Lobakin <alobakin@pm.me>
---
 net/core/skbuff.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/core/skbuff.c b/net/core/skbuff.c
index 3c904c29efbb..0e8c597ff6ce 100644
--- a/net/core/skbuff.c
+++ b/net/core/skbuff.c
@@ -487,6 +487,9 @@ EXPORT_SYMBOL(__netdev_alloc_skb);
=20
 static struct sk_buff *__napi_decache_skb(struct napi_alloc_cache *nc)
 {
+=09if (nc->flush_skb_count)
+=09=09return nc->flush_skb_cache[--nc->flush_skb_count];
+
 =09return kmem_cache_alloc(skbuff_head_cache, GFP_ATOMIC);
 }
=20
--=20
2.30.0