From: Heng Qi <hengqi@linux.alibaba.com>
To: Jason Wang <jasowang@redhat.com>
Cc: netdev@vger.kernel.org, bpf@vger.kernel.org,
"Michael S. Tsirkin" <mst@redhat.com>,
Paolo Abeni <pabeni@redhat.com>, Jakub Kicinski <kuba@kernel.org>,
John Fastabend <john.fastabend@gmail.com>,
"David S. Miller" <davem@davemloft.net>,
Daniel Borkmann <daniel@iogearbox.net>,
Alexei Starovoitov <ast@kernel.org>,
Eric Dumazet <edumazet@google.com>
Subject: Re: [RFC PATCH 5/9] virtio_net: build xdp_buff with multi buffers
Date: Thu, 8 Dec 2022 16:25:15 +0800 [thread overview]
Message-ID: <ebf67352-d072-f004-97d1-184ca03933bb@linux.alibaba.com> (raw)
In-Reply-To: <CACGkMEvBDmGdP7e1c-8s2OQFEYQ2LLbhnDF+qN+yPVwvkxPjCw@mail.gmail.com>
在 2022/12/6 下午2:14, Jason Wang 写道:
> On Tue, Nov 22, 2022 at 3:44 PM Heng Qi <hengqi@linux.alibaba.com> wrote:
>> Support xdp for multi buffer packets.
>>
>> Putting the first buffer as the linear part for xdp_buff,
>> and the rest of the buffers as non-linear fragments to struct
>> skb_shared_info in the tailroom belonging to xdp_buff.
>>
>> Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
>> Reviewed-by: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
>> ---
>> drivers/net/virtio_net.c | 74 ++++++++++++++++++++++++++++++++++++++++
>> 1 file changed, 74 insertions(+)
>>
>> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
>> index cd65f85d5075..20784b1d8236 100644
>> --- a/drivers/net/virtio_net.c
>> +++ b/drivers/net/virtio_net.c
>> @@ -911,6 +911,80 @@ static struct sk_buff *receive_big(struct net_device *dev,
>> return NULL;
>> }
>>
>> +static int virtnet_build_xdp_buff(struct net_device *dev,
>> + struct virtnet_info *vi,
>> + struct receive_queue *rq,
>> + struct xdp_buff *xdp,
>> + void *buf,
>> + unsigned int len,
>> + unsigned int frame_sz,
>> + u16 *num_buf,
>> + unsigned int *xdp_frags_truesize,
>> + struct virtnet_rq_stats *stats)
>> +{
>> + unsigned int tailroom = SKB_DATA_ALIGN(sizeof(struct skb_shared_info));
>> + struct virtio_net_hdr_mrg_rxbuf *hdr = buf;
>> + unsigned int truesize, headroom;
>> + struct skb_shared_info *shinfo;
>> + unsigned int xdp_frags_truesz = 0;
>> + unsigned int cur_frag_size;
>> + struct page *page;
>> + skb_frag_t *frag;
>> + int offset;
>> + void *ctx;
>> +
>> + xdp_init_buff(xdp, frame_sz, &rq->xdp_rxq);
>> + xdp_prepare_buff(xdp, buf - VIRTIO_XDP_HEADROOM,
>> + VIRTIO_XDP_HEADROOM + vi->hdr_len, len - vi->hdr_len, true);
>> + shinfo = xdp_get_shared_info_from_buff(xdp);
>> + shinfo->nr_frags = 0;
>> + shinfo->xdp_frags_size = 0;
>> +
>> + if ((*num_buf - 1) > MAX_SKB_FRAGS)
>> + return -EINVAL;
>> +
>> + while ((--*num_buf) >= 1) {
>> + buf = virtqueue_get_buf_ctx(rq->vq, &len, &ctx);
> So this works only for a mergeable buffer, I wonder if it's worth it
> to make it work for big mode as well. Or at least we can mention it as
> a TODO somewhere and rename this function (with mergeable suffix).
Yes, I'm leaning towards the latter, I'll rename it with a mergeable
suffix in the next version.
Thanks.
>
> Others look good.
>
> Thanks
>
>> + if (unlikely(!buf)) {
>> + pr_debug("%s: rx error: %d buffers out of %d missing\n",
>> + dev->name, *num_buf,
>> + virtio16_to_cpu(vi->vdev, hdr->num_buffers));
>> + dev->stats.rx_length_errors++;
>> + return -EINVAL;
>> + }
>> +
>> + if (!xdp_buff_has_frags(xdp))
>> + xdp_buff_set_frags_flag(xdp);
>> +
>> + stats->bytes += len;
>> + page = virt_to_head_page(buf);
>> + offset = buf - page_address(page);
>> + truesize = mergeable_ctx_to_truesize(ctx);
>> + headroom = mergeable_ctx_to_headroom(ctx);
>> +
>> + cur_frag_size = truesize + (headroom ? (headroom + tailroom) : 0);
>> + xdp_frags_truesz += cur_frag_size;
>> + if (unlikely(len > truesize || cur_frag_size > PAGE_SIZE)) {
>> + pr_debug("%s: rx error: len %u exceeds truesize %lu\n",
>> + dev->name, len, (unsigned long)ctx);
>> + dev->stats.rx_length_errors++;
>> + return -EINVAL;
>> + }
>> +
>> + frag = &shinfo->frags[shinfo->nr_frags++];
>> + __skb_frag_set_page(frag, page);
>> + skb_frag_off_set(frag, offset);
>> + skb_frag_size_set(frag, len);
>> + if (page_is_pfmemalloc(page))
>> + xdp_buff_set_frag_pfmemalloc(xdp);
>> +
>> + shinfo->xdp_frags_size += len;
>> + }
>> +
>> + *xdp_frags_truesize = xdp_frags_truesz;
>> + return 0;
>> +}
>> +
>> static struct sk_buff *receive_mergeable(struct net_device *dev,
>> struct virtnet_info *vi,
>> struct receive_queue *rq,
>> --
>> 2.19.1.6.gb485710b
>>
next prev parent reply other threads:[~2022-12-08 8:25 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-22 7:43 [RFC PATCH 0/9] virtio_net: support multi buffer xdp Heng Qi
2022-11-22 7:43 ` [RFC PATCH 1/9] virtio_net: disable the hole mechanism for xdp Heng Qi
2022-12-06 5:20 ` Jason Wang
2022-12-08 8:20 ` Heng Qi
2022-11-22 7:43 ` [RFC PATCH 2/9] virtio_net: set up xdp for multi buffer packets Heng Qi
2022-12-06 5:29 ` Jason Wang
2022-12-08 8:21 ` Heng Qi
2022-11-22 7:43 ` [RFC PATCH 3/9] virtio_net: update bytes calculation for xdp_frame Heng Qi
2022-12-06 5:31 ` Jason Wang
2022-12-08 8:35 ` Heng Qi
2022-11-22 7:43 ` [RFC PATCH 4/9] virtio_net: remove xdp related info from page_to_skb() Heng Qi
2022-12-06 5:36 ` Jason Wang
2022-12-08 8:23 ` Heng Qi
2022-11-22 7:43 ` [RFC PATCH 5/9] virtio_net: build xdp_buff with multi buffers Heng Qi
2022-12-06 6:14 ` Jason Wang
2022-12-08 8:25 ` Heng Qi [this message]
2022-11-22 7:43 ` [RFC PATCH 6/9] virtio_net: construct multi-buffer xdp in mergeable Heng Qi
2022-12-06 6:33 ` Jason Wang
2022-12-08 8:30 ` Heng Qi
2022-12-13 7:08 ` Jason Wang
2022-12-14 8:37 ` Heng Qi
2022-12-16 3:46 ` Jason Wang
2022-12-16 9:42 ` Heng Qi
2022-11-22 7:43 ` [RFC PATCH 7/9] virtio_net: build skb from multi-buffer xdp Heng Qi
2022-11-22 7:43 ` [RFC PATCH 8/9] virtio_net: transmit the " Heng Qi
2022-11-22 7:43 ` [RFC PATCH 9/9] virtio_net: support " Heng Qi
2022-12-06 6:42 ` Jason Wang
2022-12-08 8:31 ` Heng Qi
2022-12-02 4:50 ` [RFC PATCH 0/9] virtio_net: support multi buffer xdp Heng Qi
2022-12-02 5:41 ` Jason Wang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=ebf67352-d072-f004-97d1-184ca03933bb@linux.alibaba.com \
--to=hengqi@linux.alibaba.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=jasowang@redhat.com \
--cc=john.fastabend@gmail.com \
--cc=kuba@kernel.org \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox