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;
MIME-Version: 1.0
References: <20200827112922.48889-1-linmiaohe@huawei.com>
In-Reply-To: <20200827112922.48889-1-linmiaohe@huawei.com>
From:   Eric Dumazet <edumazet@google.com>
Date:   Thu, 27 Aug 2020 04:35:32 -0700
Message-ID: <CANn89iK3CKrXPj5fNYys26zd8P67jz8GZEF2WjLD6Xw05SimcA@mail.gmail.com>
Subject: Re: [PATCH] net: Set trailer iff skb1 is the last one
To:     Miaohe Lin <linmiaohe@huawei.com>
Cc:     David Miller <davem@davemloft.net>,
        Jakub Kicinski <kuba@kernel.org>,
        Pravin B Shelar <pshelar@ovn.org>,
        Florian Westphal <fw@strlen.de>, martin.varghese@nokia.com,
        Davide Caratti <dcaratti@redhat.com>,
        Steffen Klassert <steffen.klassert@secunet.com>,
        Paolo Abeni <pabeni@redhat.com>, shmulik@metanetworks.com,
        kyk.segfault@gmail.com, netdev <netdev@vger.kernel.org>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Aug 27, 2020 at 4:31 AM Miaohe Lin <linmiaohe@huawei.com> wrote:
>
> Set trailer iff skb1 is the skbuff where the tailbits space begins.
>
> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
> ---
>  net/core/skbuff.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/net/core/skbuff.c b/net/core/skbuff.c
> index 0b24aed04060..18ed56316e56 100644
> --- a/net/core/skbuff.c
> +++ b/net/core/skbuff.c
> @@ -4488,8 +4488,9 @@ int skb_cow_data(struct sk_buff *skb, int tailbits, struct sk_buff **trailer)
>                         skb1 = skb2;
>                 }
>                 elt++;
> -               *trailer = skb1;
>                 skb_p = &skb1->next;
> +               if (!*skb_p)
> +                       *trailer = skb1;
>

Why is adding a conditional test going to help ?

cpu will have hard time predicting this one, I doubt this kind of
change is a win.