[dpdk-dev] [patch v3] doc: announce API change in ethdev offload flags
Ananyev, Konstantin
konstantin.ananyev at intel.com
Thu Aug 8 11:23:21 CEST 2019
Hi guys,
>
> From: Pavan Nikhilesh <pbhagavatula at marvell.com>
>
> Add new offload flags ``DEV_RX_OFFLOAD_PTYPE``, ``DEV_RX_OFFLOAD_RSS``
> and ``DEV_RX_OFFLOAD_FLOW_MARK``.
>
> Signed-off-by: Pavan Nikhilesh <pbhagavatula at marvell.com>
> Acked-by: Andrew Rybchenko <arybchenko at solarflare.com>
> Acked-by: Jerin Jacob <jerinj at marvell.com>
> ---
> v3 Changes:
> - DEV_RX_OFFLOAD_RSS -> DEV_RX_OFFLOAD_RSS_HASH (anndrew).
>
> v2 Changes:
> - Reword for clarity.
>
> doc/guides/rel_notes/deprecation.rst | 13 +++++++++++++
> 1 file changed, 13 insertions(+)
>
> diff --git a/doc/guides/rel_notes/deprecation.rst b/doc/guides/rel_notes/deprecation.rst
> index 37b8592b6..056c5709f 100644
> --- a/doc/guides/rel_notes/deprecation.rst
> +++ b/doc/guides/rel_notes/deprecation.rst
> @@ -78,3 +78,16 @@ Deprecation Notices
> to set new power environment if power environment was already initialized.
> In this case the function will return -1 unless the environment is unset first
> (using ``rte_power_unset_env``). Other function usage scenarios will not change.
> +
> +* ethdev: New offload flags ``DEV_RX_OFFLOAD_PTYPE``, ``DEV_RX_OFFLOAD_RSS_HASH``
> + and ``DEV_RX_OFFLOAD_FLOW_MARK`` will be added in 19.11.
One question about DEV_RX_OFFLOAD_PTYPE:
Does it mean that new ol_flags value (PKT_RX_PTYPE) will be introduced to
indicate that mbuf.packet_type value is set?
Or PMD will have to set mbuf.packet_type to zero,
when DEV_RX_OFFLOAD_PTYPE was not enabled by user?
If so, what is the advantage?
Again in that case, would it be more plausible to introduce something like:
rte_eth_dev_set_supported_ptypes(uint16_t port_id, uint32_t ptype_mask);
instead of DEV_RX_OFFLOAD_PTYPE?
Konstantin
> + This will allow application to enable or disable PMDs from updating
> + ``rte_mbuf`` fields ``rte_mbuf::packet_type``, ``rte_mbuf::hash::rss`` and
> + ``rte_mbuf::hash::fdir`` respectively.
> + This scheme will allow PMDs to avoid writes to ``rte_mbuf`` fields on Rx and
> + thereby improve Rx performance if application wishes do so.
> + In 19.11 PMDs will still update the fields even when the offloads are not
> + enabled.
> + The exact semantics of the flags will be worked out later either by making
> + them negative offloads to avoid application change or positive offload to
> + align with existing offload flag semantics.
> --
> 2.17.1
More information about the dev
mailing list