> The default policy for offload-specific fields is that
> they are undefined unless the corresponding offloads are
> requested in mbuf ol_flags. This is also the case for outer
> L2 and L3 length fields which must not be assumed to contain
> zeros for non-tunnel packets. The patch clarifies this behaviour
> in the comments. PMDs which mistakenly assume these fields
> to be zero for non-tunnel packets are expected to comply
> with the clarified comment and have dedicated fixes.
>
> Signed-off-by: Ivan Malov <ivan.ma...@oktetlabs.ru>
> ---
> lib/librte_mbuf/rte_mbuf.h | 13 ++++++++++++-
> 1 file changed, 12 insertions(+), 1 deletion(-)
>
> diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h
> index 0d9fef0..26a0b14 100644
> --- a/lib/librte_mbuf/rte_mbuf.h
> +++ b/lib/librte_mbuf/rte_mbuf.h
> @@ -702,7 +702,18 @@ struct rte_mbuf {
> uint64_t tso_segsz:RTE_MBUF_TSO_SEGSZ_BITS;
> /**< TCP TSO segment size */
>
> - /* fields for TX offloading of tunnels */
> + /*
> + * Fields for Tx offloading of tunnels.
> + * These are undefined for packets which don't request
> + * any tunnel offloads (outer IP or UDP checksum,
> + * tunnel TSO).
> + *
> + * PMDs are advised not to use these fields
> + * unconditionally when calculating offsets.
I'd use something stronger then 'are advised not to'.
Might be something like: 'should not use'.
BTW, from the previous discussion, right now we do
have some misbehaving PMDs that need to be fixed, correct?
Apart from that:
Acked-by: Konstantin Ananyev <konstantin.anan...@intel.com>
> + *
> + * Applications are expected to set appropriate tunnel
> + * offload flags when they fill in these fields.
> + */
> uint64_t outer_l3_len:RTE_MBUF_OUTL3_LEN_BITS;
> /**< Outer L3 (IP) Hdr Length. */
> uint64_t outer_l2_len:RTE_MBUF_OUTL2_LEN_BITS;
> --
> 1.8.3.1
