netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Yuya Kusakabe <yuya.kusakabe@gmail.com>
To: Jason Wang <jasowang@redhat.com>, "Michael S. Tsirkin" <mst@redhat.com>
Cc: andriin@fb.com, ast@kernel.org, bpf@vger.kernel.org,
	daniel@iogearbox.net, davem@davemloft.net, hawk@kernel.org,
	john.fastabend@gmail.com, kafai@fb.com, kuba@kernel.org,
	linux-kernel@vger.kernel.org, netdev@vger.kernel.org,
	songliubraving@fb.com, yhs@fb.com
Subject: Re: [PATCH bpf-next v5] virtio_net: add XDP meta data support
Date: Tue, 25 Feb 2020 09:52:40 +0900	[thread overview]
Message-ID: <f0a0c2bf-ac93-5704-13e7-ada1d1aa55c7@gmail.com> (raw)
In-Reply-To: <7272077b-4cf9-b81b-22b5-22a2b0aceeb6@redhat.com>

On 2/24/20 1:05 PM, Jason Wang wrote:
> 
> On 2020/2/23 下午4:14, Michael S. Tsirkin wrote:
>> On Fri, Feb 21, 2020 at 05:36:08PM +0900, Yuya Kusakabe wrote:
>>> On 2/21/20 1:23 PM, Jason Wang wrote:
>>>> On 2020/2/20 下午4:55, Yuya Kusakabe wrote:
>>>>> Implement support for transferring XDP meta data into skb for
>>>>> virtio_net driver; before calling into the program, xdp.data_meta points
>>>>> to xdp.data, where on program return with pass verdict, we call
>>>>> into skb_metadata_set().
>>>>>
>>>>> Tested with the script at
>>>>> https://github.com/higebu/virtio_net-xdp-metadata-test.
>>>>>
>>>>> Fixes: de8f3a83b0a0 ("bpf: add meta pointer for direct access")
>>>> I'm not sure this is correct since virtio-net claims to not support metadata by calling xdp_set_data_meta_invalid()?
>>> virtio_net doesn't support by calling xdp_set_data_meta_invalid() for now.
>>>
>>> https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git/tree/drivers/net/virtio_net.c?id=e42da4c62abb547d9c9138e0e7fcd1f36057b5e8#n686
>>> https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next.git/tree/drivers/net/virtio_net.c?id=e42da4c62abb547d9c9138e0e7fcd1f36057b5e8#n842
>>>
>>> And xdp_set_data_meta_invalid() are added by de8f3a83b0a0.
>>>
>>> $ git blame ./drivers/net/virtio_net.c | grep xdp_set_data_meta_invalid
>>> de8f3a83b0a0f (Daniel Borkmann           2017-09-25 02:25:51 +0200  686)                xdp_set_data_meta_invalid(&xdp);
>>> de8f3a83b0a0f (Daniel Borkmann           2017-09-25 02:25:51 +0200  842)                xdp_set_data_meta_invalid(&xdp);
>>>
>>> So I added `Fixes: de8f3a83b0a0 ("bpf: add meta pointer for direct access")` to the comment.
>>>
>>>>> Signed-off-by: Yuya Kusakabe<yuya.kusakabe@gmail.com>
>>>>> ---
>>>>> v5:
>>>>>    - page_to_skb(): copy vnet header if hdr_valid without checking metasize.
>>>>>    - receive_small(): do not copy vnet header if xdp_prog is availavle.
>>>>>    - __virtnet_xdp_xmit_one(): remove the xdp_set_data_meta_invalid().
>>>>>    - improve comments.
>>>>> v4:
>>>>>    - improve commit message
>>>>> v3:
>>>>>    - fix preserve the vnet header in receive_small().
>>>>> v2:
>>>>>    - keep copy untouched in page_to_skb().
>>>>>    - preserve the vnet header in receive_small().
>>>>>    - fix indentation.
>>>>> ---
>>>>>    drivers/net/virtio_net.c | 54 ++++++++++++++++++++++++----------------
>>>>>    1 file changed, 33 insertions(+), 21 deletions(-)
>>>>>
>>>>> diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
>>>>> index 2fe7a3188282..4ea0ae60c000 100644
>>>>> --- a/drivers/net/virtio_net.c
>>>>> +++ b/drivers/net/virtio_net.c
>>>>> @@ -371,7 +371,7 @@ static struct sk_buff *page_to_skb(struct virtnet_info *vi,
>>>>>                       struct receive_queue *rq,
>>>>>                       struct page *page, unsigned int offset,
>>>>>                       unsigned int len, unsigned int truesize,
>>>>> -                   bool hdr_valid)
>>>>> +                   bool hdr_valid, unsigned int metasize)
>>>>>    {
>>>>>        struct sk_buff *skb;
>>>>>        struct virtio_net_hdr_mrg_rxbuf *hdr;
>>>>> @@ -393,6 +393,7 @@ static struct sk_buff *page_to_skb(struct virtnet_info *vi,
>>>>>        else
>>>>>            hdr_padded_len = sizeof(struct padded_vnet_hdr);
>>>>>    +    /* hdr_valid means no XDP, so we can copy the vnet header */
>>>>>        if (hdr_valid)
>>>>>            memcpy(hdr, p, hdr_len);
>>>>>    @@ -405,6 +406,11 @@ static struct sk_buff *page_to_skb(struct virtnet_info *vi,
>>>>>            copy = skb_tailroom(skb);
>>>>>        skb_put_data(skb, p, copy);
>>>>>    +    if (metasize) {
>>>>> +        __skb_pull(skb, metasize);
>>>>> +        skb_metadata_set(skb, metasize);
>>>>> +    }
>>>>> +
>>>>>        len -= copy;
>>>>>        offset += copy;
>>>>>    @@ -450,10 +456,6 @@ static int __virtnet_xdp_xmit_one(struct virtnet_info *vi,
>>>>>        struct virtio_net_hdr_mrg_rxbuf *hdr;
>>>>>        int err;
>>>>>    -    /* virtqueue want to use data area in-front of packet */
>>>>> -    if (unlikely(xdpf->metasize > 0))
>>>>> -        return -EOPNOTSUPP;
>>>>> -
>>>>>        if (unlikely(xdpf->headroom < vi->hdr_len))
>>>>>            return -EOVERFLOW;
>>>>>    @@ -644,6 +646,7 @@ static struct sk_buff *receive_small(struct net_device *dev,
>>>>>        unsigned int delta = 0;
>>>>>        struct page *xdp_page;
>>>>>        int err;
>>>>> +    unsigned int metasize = 0;
>>>>>          len -= vi->hdr_len;
>>>>>        stats->bytes += len;
>>>>> @@ -683,8 +686,8 @@ static struct sk_buff *receive_small(struct net_device *dev,
>>>>>              xdp.data_hard_start = buf + VIRTNET_RX_PAD + vi->hdr_len;
>>>>>            xdp.data = xdp.data_hard_start + xdp_headroom;
>>>>> -        xdp_set_data_meta_invalid(&xdp);
>>>>>            xdp.data_end = xdp.data + len;
>>>>> +        xdp.data_meta = xdp.data;
>>>>>            xdp.rxq = &rq->xdp_rxq;
>>>>>            orig_data = xdp.data;
>>>>>            act = bpf_prog_run_xdp(xdp_prog, &xdp);
>>>>> @@ -695,6 +698,7 @@ static struct sk_buff *receive_small(struct net_device *dev,
>>>>>                /* Recalculate length in case bpf program changed it */
>>>>>                delta = orig_data - xdp.data;
>>>>>                len = xdp.data_end - xdp.data;
>>>>> +            metasize = xdp.data - xdp.data_meta;
>>>>>                break;
>>>>>            case XDP_TX:
>>>>>                stats->xdp_tx++;
>>>>> @@ -735,11 +739,14 @@ static struct sk_buff *receive_small(struct net_device *dev,
>>>>>        }
>>>>>        skb_reserve(skb, headroom - delta);
>>>>>        skb_put(skb, len);
>>>>> -    if (!delta) {
>>>>> +    if (!xdp_prog) {
>>>>>            buf += header_offset;
>>>>>            memcpy(skb_vnet_hdr(skb), buf, vi->hdr_len);
>>>>>        } /* keep zeroed vnet hdr since packet was changed by bpf */
>>>> I prefer to make this an independent patch and cc stable.
>>>>
>>>> Other looks good.
>>>>
>>>> Thanks
>>> I see. So I need to revert to delta from xdp_prog?
>>>
>>> Thank you.
>> So maybe send a 2 patch series: 1/2 is this chunk with the appropriate
>> description. Actually for netdev David prefers that people do not
>> cc stable directly, just include Fixes tag and mention in the
>> commit log it's also needed for stable. Patch 2/2 is the rest
>> handling metadata.
> 
> 
> +1
> 
> Thanks
> 
> 

Thank you for the detailed explanation. I will make a 2 patch series.

  reply	other threads:[~2020-02-25  0:52 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-06-27  8:06 [PATCH bpf-next] virtio_net: add XDP meta data support Yuya Kusakabe
2019-07-01  9:30 ` Jason Wang
2019-07-02  1:00   ` Yuya Kusakabe
2019-07-02  3:15     ` [PATCH bpf-next v2] " Yuya Kusakabe
2019-07-02  3:59       ` Jason Wang
2019-07-02  5:15         ` Yuya Kusakabe
2019-07-02  8:16           ` [PATCH bpf-next v3] " Yuya Kusakabe
2019-07-02  8:33             ` Jason Wang
2019-07-02 14:11               ` Yuya Kusakabe
2019-07-08 22:38                 ` Daniel Borkmann
2019-07-09  3:04                   ` Jason Wang
2019-07-09 20:03                     ` Daniel Borkmann
2019-07-10  2:30                       ` Jason Wang
2020-02-03 13:52                         ` Yuya Kusakabe
2020-02-04  3:31                           ` Jason Wang
2020-02-04  7:16                             ` [PATCH bpf-next v4] " Yuya Kusakabe
2020-02-05  4:10                               ` Jason Wang
2020-02-05  9:18                                 ` Yuya Kusakabe
2020-02-06  3:20                                   ` Jason Wang
2020-02-20  8:55                                     ` [PATCH bpf-next v5] " Yuya Kusakabe
2020-02-21  4:23                                       ` Jason Wang
2020-02-21  8:36                                         ` Yuya Kusakabe
2020-02-21 11:01                                           ` Michael S. Tsirkin
2020-02-23  8:14                                           ` Michael S. Tsirkin
2020-02-24  4:05                                             ` Jason Wang
2020-02-25  0:52                                               ` Yuya Kusakabe [this message]
2020-02-05  5:33                               ` [PATCH bpf-next v4] " Michael S. Tsirkin
2020-02-05  9:19                                 ` Yuya Kusakabe

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=f0a0c2bf-ac93-5704-13e7-ada1d1aa55c7@gmail.com \
    --to=yuya.kusakabe@gmail.com \
    --cc=andriin@fb.com \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=hawk@kernel.org \
    --cc=jasowang@redhat.com \
    --cc=john.fastabend@gmail.com \
    --cc=kafai@fb.com \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mst@redhat.com \
    --cc=netdev@vger.kernel.org \
    --cc=songliubraving@fb.com \
    --cc=yhs@fb.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;
as well as URLs for NNTP newsgroup(s).