On 4/25/2018 5:27 PM, Adrien Mazarguil wrote:
> On Wed, Apr 25, 2018 at 09:38:16PM +0800, Qi Zhang wrote:
>> After add RSS hash offload check, default rss_hf will fail on
>> devices that not support all bits, the patch take rss_hf as
>> a suggest value and only set bits that device supported base on
>> rte_eth_dev_get_info, also rss_hf will only be updated when new
>> rss offload is successfully updated on all ports by
>> "port config all rss [!default]" command.
>>
>> Fixes: 586ac442be96 ("ethdev: add supported hash function check")
>> Fixes: 8c1f4aff92a6 ("app/testpmd: new parameter for port config all RSS
>> command")
>> Signed-off-by: Qi Zhang <qi.z.zh...@intel.com>
>
> Although this approach prevents updating rss_hf if at least one port doesn't
> support RSS configuration (i.e. it can't be used to update defaults for
> subsequent flow rules on mlx4),
Yes it does, and perhaps we need a value per port, instead of single value for
all.
> I confirm it does fix the startup issue,
> therefore:
>
> Tested-by: Adrien Mazarguil <adrien.mazarg...@6wind.com>
>
