Re: [PATCH v3 4/9] virtio-net: build xdp_buff with multi buffers

Jason Wang <jasowang@xxxxxxxxxx> · Fri, 13 Jan 2023 10:58:47 +0800

在 2023/1/3 14:40, Heng Qi 写道:
Support xdp for multi buffer packets in mergeable mode.

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.

Let 'truesize' return to its literal meaning, that is, when
xdp is set, it includes the length of headroom and tailroom.

Signed-off-by: Heng Qi <hengqi@xxxxxxxxxxxxxxxxx>
Reviewed-by: Xuan Zhuo <xuanzhuo@xxxxxxxxxxxxxxxxx>


Acked-by: Jason Wang <jasowang@xxxxxxxxxx>

Thanks


---
  drivers/net/virtio_net.c | 108 ++++++++++++++++++++++++++++++++++++---
  1 file changed, 100 insertions(+), 8 deletions(-)

diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index 6fc5302ca5ff..699e376b8f8b 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -931,6 +931,91 @@ static struct sk_buff *receive_big(struct net_device *dev,
  	return NULL;
  }
  
+/* TODO: build xdp in big mode */
+static int virtnet_build_xdp_buff_mrg(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)
+{
+	struct virtio_net_hdr_mrg_rxbuf *hdr = buf;
+	unsigned int headroom, tailroom, room;
+	unsigned int truesize, cur_frag_size;
+	struct skb_shared_info *shinfo;
+	unsigned int xdp_frags_truesz = 0;
+	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);
+
+	if (*num_buf > 1) {
+		/* If we want to build multi-buffer xdp, we need
+		 * to specify that the flags of xdp_buff have the
+		 * XDP_FLAGS_HAS_FRAG bit.
+		 */
+		if (!xdp_buff_has_frags(xdp))
+			xdp_buff_set_frags_flag(xdp);
+
+		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);
+		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;
+		}
+
+		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);
+		tailroom = headroom ? sizeof(struct skb_shared_info) : 0;
+		room = SKB_DATA_ALIGN(headroom + tailroom);
+
+		cur_frag_size = truesize;
+		xdp_frags_truesz += cur_frag_size;
+		if (unlikely(len > truesize - room || cur_frag_size > PAGE_SIZE)) {
+			put_page(page);
+			pr_debug("%s: rx error: len %u exceeds truesize %lu\n",
+				 dev->name, len, (unsigned long)(truesize - room));
+			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,
@@ -949,15 +1034,17 @@ static struct sk_buff *receive_mergeable(struct net_device *dev,
  	unsigned int truesize = mergeable_ctx_to_truesize(ctx);
  	unsigned int headroom = mergeable_ctx_to_headroom(ctx);
  	unsigned int metasize = 0;
+	unsigned int tailroom = headroom ? sizeof(struct skb_shared_info) : 0;
+	unsigned int room = SKB_DATA_ALIGN(headroom + tailroom);
  	unsigned int frame_sz;
  	int err;
  
  	head_skb = NULL;
  	stats->bytes += len - vi->hdr_len;
  
-	if (unlikely(len > truesize)) {
+	if (unlikely(len > truesize - room)) {
  		pr_debug("%s: rx error: len %u exceeds truesize %lu\n",
-			 dev->name, len, (unsigned long)ctx);
+			 dev->name, len, (unsigned long)(truesize - room));
  		dev->stats.rx_length_errors++;
  		goto err_skb;
  	}
@@ -983,10 +1070,12 @@ static struct sk_buff *receive_mergeable(struct net_device *dev,
  		if (unlikely(hdr->hdr.gso_type))
  			goto err_xdp;
  
-		/* Buffers with headroom use PAGE_SIZE as alloc size,
-		 * see add_recvbuf_mergeable() + get_mergeable_buf_len()
+		/* Now XDP core assumes frag size is PAGE_SIZE, but buffers
+		 * with headroom may add hole in truesize, which
+		 * make their length exceed PAGE_SIZE. So we disabled the
+		 * hole mechanism for xdp. See add_recvbuf_mergeable().
  		 */
-		frame_sz = headroom ? PAGE_SIZE : truesize;
+		frame_sz = truesize;
  
  		/* This happens when rx buffer size is underestimated
  		 * or headroom is not enough because of the buffer
@@ -1139,9 +1228,12 @@ static struct sk_buff *receive_mergeable(struct net_device *dev,
  		page = virt_to_head_page(buf);
  
  		truesize = mergeable_ctx_to_truesize(ctx);
-		if (unlikely(len > truesize)) {
+		headroom = mergeable_ctx_to_headroom(ctx);
+		tailroom = headroom ? sizeof(struct skb_shared_info) : 0;
+		room = SKB_DATA_ALIGN(headroom + tailroom);
+		if (unlikely(len > truesize - room)) {
  			pr_debug("%s: rx error: len %u exceeds truesize %lu\n",
-				 dev->name, len, (unsigned long)ctx);
+				 dev->name, len, (unsigned long)(truesize - room));
  			dev->stats.rx_length_errors++;
  			goto err_skb;
  		}
@@ -1428,7 +1520,7 @@ static int add_recvbuf_mergeable(struct virtnet_info *vi,
  	}
  
  	sg_init_one(rq->sg, buf, len);
-	ctx = mergeable_len_to_ctx(len, headroom);
+	ctx = mergeable_len_to_ctx(len + room, headroom);
  	err = virtqueue_add_inbuf_ctx(rq->vq, rq->sg, 1, buf, ctx, gfp);
  	if (err < 0)
  		put_page(virt_to_head_page(buf));







