[dpdk-dev] [PATCH] doc: add mlx5 limitation of RSS hash result

[dpdk-dev] [PATCH] doc: add mlx5 limitation of RSS hash result

[Yongseok Koh]

Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
---
 doc/guides/nics/mlx5.rst | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/doc/guides/nics/mlx5.rst b/doc/guides/nics/mlx5.rst
index 52e1213cf5..cbcc233d03 100644
--- a/doc/guides/nics/mlx5.rst
+++ b/doc/guides/nics/mlx5.rst
@@ -136,6 +136,12 @@ Limitations
   enabled (``rxq_cqe_comp_en``) at the same time, RSS hash result is not fully
   supported. Some Rx packets may not have PKT_RX_RSS_HASH.
 
+- In general, PKT_RX_RSS_HASH shall be set in a received mbuf if RSS is enabled
+  and the packet is subject to hash cacalculation. However, in case of
+  vectorized Rx (``rx_vec_en``), PKT_RX_RSS_HASH will always be set as long as
+  RSS is enabled. Zero hash result (mbuf->hash.rss) means the packet isn't
+  subject to the configured RSS hash type and the flag should be disregarded.
+
 Statistics
 ----------
 
-- 
2.11.0

Re: [dpdk-dev] [PATCH] doc: add mlx5 limitation of RSS hash result

[Yongseok Koh]

> On Aug 14, 2018, at 9:57 AM, Yongseok Koh <yskoh@mellanox.com> wrote:
> 
> Signed-off-by: Yongseok Koh <yskoh@mellanox.com>
> ---
> doc/guides/nics/mlx5.rst | 6 ++++++
> 1 file changed, 6 insertions(+)
> 
> diff --git a/doc/guides/nics/mlx5.rst b/doc/guides/nics/mlx5.rst
> index 52e1213cf5..cbcc233d03 100644
> --- a/doc/guides/nics/mlx5.rst
> +++ b/doc/guides/nics/mlx5.rst
> @@ -136,6 +136,12 @@ Limitations
>   enabled (``rxq_cqe_comp_en``) at the same time, RSS hash result is not fully
>   supported. Some Rx packets may not have PKT_RX_RSS_HASH.
> 
> +- In general, PKT_RX_RSS_HASH shall be set in a received mbuf if RSS is enabled
> +  and the packet is subject to hash cacalculation. However, in case of
> +  vectorized Rx (``rx_vec_en``), PKT_RX_RSS_HASH will always be set as long as
> +  RSS is enabled. Zero hash result (mbuf->hash.rss) means the packet isn't
> +  subject to the configured RSS hash type and the flag should be disregarded.
> +
> Statistics
> ----------

I have to take it back.
We concluded there's a bug in the implementation and decided to fix this issue.

Thanks,
Yongseok