[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <4dda1ca6-f684-4b6a-8972-3915921cbf6d@intel.com>
Date: Mon, 21 Oct 2024 16:20:24 +0200
From: Alexander Lobakin <aleksander.lobakin@...el.com>
To: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
CC: "David S. Miller" <davem@...emloft.net>, Eric Dumazet
<edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>, Paolo Abeni
<pabeni@...hat.com>, Toke Høiland-Jørgensen
<toke@...hat.com>, Alexei Starovoitov <ast@...nel.org>, Daniel Borkmann
<daniel@...earbox.net>, John Fastabend <john.fastabend@...il.com>, "Andrii
Nakryiko" <andrii@...nel.org>, Stanislav Fomichev <sdf@...ichev.me>, "Magnus
Karlsson" <magnus.karlsson@...el.com>,
<nex.sw.ncis.osdt.itp.upstreaming@...el.com>, <bpf@...r.kernel.org>,
<netdev@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH net-next v2 12/18] xdp: add generic
xdp_build_skb_from_buff()
From: Maciej Fijalkowski <maciej.fijalkowski@...el.com>
Date: Thu, 17 Oct 2024 14:34:33 +0200
> On Tue, Oct 15, 2024 at 04:53:44PM +0200, Alexander Lobakin wrote:
>> The code which builds an skb from an &xdp_buff keeps multiplying itself
>> around the drivers with almost no changes. Let's try to stop that by
>> adding a generic function.
[...]
>> + skb_record_rx_queue(skb, rxq->queue_index);
>> +
>> + if (unlikely(nr_frags)) {
>> + u32 ts;
>
> nit: spell out truesize? ts confuse my brain with timestamp TBH
Ack.
>
>> +
>> + ts = sinfo->xdp_frags_truesize ? : nr_frags * xdp->frame_sz;
>> + xdp_update_skb_shared_info(skb, nr_frags,
>> + sinfo->xdp_frags_size, ts,
>> + xdp_buff_is_frag_pfmemalloc(xdp));
>> + }
>> +
>> + skb->protocol = eth_type_trans(skb, rxq->dev);
>
> could we leave this out to be set by drivers? i see in ice for example
> netdev ptr is retrieved in different ways here.
As for ice, it's for representors only + unlikely() + we could override
it manually there later (set skb->dev + call eth_skb_pkt_type()).
OTOH this looks more consistent with xdp_build_skb_from_frame() + saves
some object code (you won't need to generate a call to eth_type_trans()
in each driver).
>
>> +
>> + return skb;
>> +}
>> +EXPORT_SYMBOL_GPL(xdp_build_skb_from_buff);
>> +
>> struct sk_buff *__xdp_build_skb_from_frame(struct xdp_frame *xdpf,
>> struct sk_buff *skb,
>> struct net_device *dev)
>> --
>> 2.46.2
Thanks,
Olek
Powered by blists - more mailing lists