Re: [PATCH net-next,v3 09/12] flow_dissector: add basic ethtool_rx_flow_spec to flow_rule structure translator

[Jiri Pirko <jiri@resnulli.us>]

Thu, Nov 22, 2018 at 05:59:27AM CET, f.fainelli@gmail.com wrote:
>
>
>On 11/20/2018 6:51 PM, Pablo Neira Ayuso wrote:
>> This patch adds a function to translate the ethtool_rx_flow_spec
>> structure to the flow_rule representation.
>> 
>> This allows us to reuse code from the driver side given that both flower
>> and ethtool_rx_flow interfaces use the same representation.
>> 
>> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
>> ---
>> v3: Suggested by Jiri Pirko:
>>         - Add struct ethtool_rx_flow_rule, keep placeholder to private
>>           dissector information.
>>     Reported by Manish Chopra:
>> 	- Fix incorrect dissector user_keys flags.
>> 
>>  include/linux/ethtool.h |  10 +++
>>  net/core/ethtool.c      | 189 ++++++++++++++++++++++++++++++++++++++++++++++++
>>  2 files changed, 199 insertions(+)
>> 
>> diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h
>> index afd9596ce636..99849e0858b2 100644
>> --- a/include/linux/ethtool.h
>> +++ b/include/linux/ethtool.h
>> @@ -400,4 +400,14 @@ struct ethtool_ops {
>>  	void	(*get_ethtool_phy_stats)(struct net_device *,
>>  					 struct ethtool_stats *, u64 *);
>>  };
>> +
>> +struct ethtool_rx_flow_rule {
>> +	struct flow_rule	*rule;
>> +	unsigned long		priv[0];
>
>This forces you to cast to/from that member, any reason this is just not
>a void *priv here?

The reason is single alloc call for rule and its priv. Quite usual along
the code.


>-- 
>Florian