Re: [RFC PATCH v9 12/19] virtio/vsock: add SEQPACKET receive logic

[Stefano Garzarella]

On Sat, May 08, 2021 at 07:35:40PM +0300, Arseny Krasnov wrote:
> This modifies current receive logic for SEQPACKET support:
> 1) Inserts 'RW' packet to socket's rx queue, but without merging with
>   buffer of last packet in queue.

This is not true anymore, right?

> 2) Performs check for packet and socket types on receive(if mismatch,
>   then reset connection).
>
> Signed-off-by: Arseny Krasnov <arseny.krasnov@xxxxxxxxxxxxx>
> Reviewed-by: Stefano Garzarella <sgarzare@xxxxxxxxxx>

Also this patch is changed :-)

> ---
> net/vmw_vsock/virtio_transport_common.c | 28 +++++++++++++++++++++++--
> 1 file changed, 26 insertions(+), 2 deletions(-)
>
> diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> index f649a21dd23b..7fea0a2192f7 100644
> --- a/net/vmw_vsock/virtio_transport_common.c
> +++ b/net/vmw_vsock/virtio_transport_common.c
> @@ -165,6 +165,14 @@ void virtio_transport_deliver_tap_pkt(struct virtio_vsock_pkt *pkt)
> }
> EXPORT_SYMBOL_GPL(virtio_transport_deliver_tap_pkt);
>
> +static u16 virtio_transport_get_type(struct sock *sk)
> +{
> +	if (sk->sk_type == SOCK_STREAM)
> +		return VIRTIO_VSOCK_TYPE_STREAM;
> +	else
> +		return VIRTIO_VSOCK_TYPE_SEQPACKET;
> +}
> +
> /* This function can only be used on connecting/connected sockets,
>  * since a socket assigned to a transport is required.
>  *
> @@ -980,11 +988,15 @@ virtio_transport_recv_enqueue(struct vsock_sock *vsk,
> 		/* If there is space in the last packet queued, we copy the
> 		 * new packet in its buffer.
> 		 */
> -		if (pkt->len <= last_pkt->buf_len - last_pkt->len) {
> +		if ((pkt->len <= last_pkt->buf_len - last_pkt->len) &&
> +		    !(le32_to_cpu(last_pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR)) {

Maybe we should update the comment above.

> 			memcpy(last_pkt->buf + last_pkt->len, pkt->buf,
> 			       pkt->len);
> 			last_pkt->len += pkt->len;
> 			free_pkt = true;
> +
> +			if (le32_to_cpu(pkt->hdr.flags) & VIRTIO_VSOCK_SEQ_EOR)
> +				last_pkt->hdr.flags |= cpu_to_le32(VIRTIO_VSOCK_SEQ_EOR);

What about doing the following in any case?

			last_pkt->hdr.flags |= pkt->hdr.flags;

> 			goto out;
> 		}
> 	}