Re: [RFC PATCH v1 1/4] virtio/vsock: rework MSG_PEEK for SOCK_STREAM

[Stefano Garzarella]

On Sun, Jun 18, 2023 at 09:24:48AM +0300, Arseniy Krasnov wrote:
> This reworks current implementation of MSG_PEEK logic:
> 1) Replaces 'skb_queue_walk_safe()' with 'skb_queue_walk()'. There is
>   no need in the first one, as there are no removes of skb in loop.
> 2) Removes nested while loop - MSG_PEEK logic could be implemented
>   without it: just iterate over skbs without removing it and copy
>   data from each until destination buffer is not full.
>
> Signed-off-by: Arseniy Krasnov <AVKrasnov@xxxxxxxxxxxxxx>
> ---
> net/vmw_vsock/virtio_transport_common.c | 41 ++++++++++++-------------
> 1 file changed, 19 insertions(+), 22 deletions(-)

Great clean up!

LGTM, but @Bobby can you also take a look?

Thanks,
Stefano

> diff --git a/net/vmw_vsock/virtio_transport_common.c b/net/vmw_vsock/virtio_transport_common.c
> index b769fc258931..2ee40574c339 100644
> --- a/net/vmw_vsock/virtio_transport_common.c
> +++ b/net/vmw_vsock/virtio_transport_common.c
> @@ -348,37 +348,34 @@ virtio_transport_stream_do_peek(struct vsock_sock *vsk,
> 				size_t len)
> {
> 	struct virtio_vsock_sock *vvs = vsk->trans;
> -	size_t bytes, total = 0, off;
> -	struct sk_buff *skb, *tmp;
> -	int err = -EFAULT;
> +	struct sk_buff *skb;
> +	size_t total = 0;
> +	int err;
>
> 	spin_lock_bh(&vvs->rx_lock);
>
> -	skb_queue_walk_safe(&vvs->rx_queue, skb,  tmp) {
> -		off = 0;
> +	skb_queue_walk(&vvs->rx_queue, skb) {
> +		size_t bytes;
>
> -		if (total == len)
> -			break;
> +		bytes = len - total;
> +		if (bytes > skb->len)
> +			bytes = skb->len;
>
> -		while (total < len && off < skb->len) {
> -			bytes = len - total;
> -			if (bytes > skb->len - off)
> -				bytes = skb->len - off;
> +		spin_unlock_bh(&vvs->rx_lock);
>
> -			/* sk_lock is held by caller so no one else can dequeue.
> -			 * Unlock rx_lock since memcpy_to_msg() may sleep.
> -			 */
> -			spin_unlock_bh(&vvs->rx_lock);
> +		/* sk_lock is held by caller so no one else can dequeue.
> +		 * Unlock rx_lock since memcpy_to_msg() may sleep.
> +		 */
> +		err = memcpy_to_msg(msg, skb->data, bytes);
> +		if (err)
> +			goto out;
>
> -			err = memcpy_to_msg(msg, skb->data + off, bytes);
> -			if (err)
> -				goto out;
> +		total += bytes;
>
> -			spin_lock_bh(&vvs->rx_lock);
> +		spin_lock_bh(&vvs->rx_lock);
>
> -			total += bytes;
> -			off += bytes;
> -		}
> +		if (total == len)
> +			break;
> 	}
>
> 	spin_unlock_bh(&vvs->rx_lock);
> --
> 2.25.1