Re: [patch net-next 3/8] mlxsw: spectrum_dpipe: Add IPv6 host table initial support

[Jiri Pirko <jiri@resnulli.us>]

Wed, Aug 30, 2017 at 07:36:13PM CEST, dsahern@gmail.com wrote:
>On 8/30/17 6:03 AM, Jiri Pirko wrote:
>> @@ -328,9 +329,21 @@ static int mlxsw_sp_dpipe_table_host_matches_dump(struct sk_buff *skb, int type)
>>  	if (err)
>>  		return err;
>>  
>> -	match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT;
>> -	match.header = &devlink_dpipe_header_ipv4;
>> -	match.field_id = DEVLINK_DPIPE_FIELD_IPV4_DST_IP;
>> +	switch (type) {
>> +	case AF_INET:
>> +		match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT;
>> +		match.header = &devlink_dpipe_header_ipv4;
>> +		match.field_id = DEVLINK_DPIPE_FIELD_IPV4_DST_IP;
>> +		break;
>> +	case AF_INET6:
>> +		match.type = DEVLINK_DPIPE_MATCH_TYPE_FIELD_EXACT;
>> +		match.header = &devlink_dpipe_header_ipv6;
>> +		match.field_id = DEVLINK_DPIPE_FIELD_IPV6_DST_IP;
>> +		break;
>> +	default:
>> +		WARN_ON(1);
>
>Why a warn for dump request of an unsupported family?

It's a handling of default case that should not happen unless there is a
bug in kernel.


>
>> +		return -EINVAL;
>> +	}
>>  
>>  	return devlink_dpipe_match_put(skb, &match);
>>  }
>> @@ -342,7 +355,7 @@ mlxsw_sp_dpipe_table_host4_matches_dump(void *priv, struct sk_buff *skb)
>