On 5/28/24 13:39, Eelco Chaudron wrote:
> This patch fixes an uninitialized gso_type case in
> netdev_linux_prepend_vnet_hdr() by returning an error.
> 
> Fixes: 3337e6d91c5b ("userspace: Enable L4 checksum offloading by default.")
> Signed-off-by: Eelco Chaudron <echau...@redhat.com>
> ---
>  lib/netdev-linux.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/lib/netdev-linux.c b/lib/netdev-linux.c
> index eb0c5c624..7cffc0e13 100644
> --- a/lib/netdev-linux.c
> +++ b/lib/netdev-linux.c
> @@ -7167,6 +7167,11 @@ netdev_linux_prepend_vnet_hdr(struct dp_packet *b, int 
> mtu)
>              vnet->gso_type = VIRTIO_NET_HDR_GSO_TCPV4;
>          } else if (dp_packet_hwol_tx_ipv6(b)) {
>              vnet->gso_type = VIRTIO_NET_HDR_GSO_TCPV6;
> +        } else {
> +            VLOG_ERR_RL(&rl, "Unknown gso_type for TSO hw offload packet. "
> +                        "Flags: %"PRIu64,
> +                        (uint64_t)*dp_packet_ol_flags_ptr(b));

I'm not sure if this should be an error or warning, up to you.  But I'd suggest
removing the 'hw offload' part from the message since 'TSO' contains 'offload'
and it's not necessarily hardware here.  Also, flags are better printed in hex,
i.e. %#"PRIx64.  And there is a missing space between the cast and a variable.

Best regards, Ilya Maximets.
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to