[Olsr-dev] [PATCH v1 09/23] kernel_route: olsr_os_inetgw_tunnel_route can now take the table

[Ferry Huberts]

That extra patch is now also pushed onto the multigw branch...

On 13/12/12 12:59, Henning Rogge wrote:
> On 12/13/2012 12:33 PM, Ferry Huberts wrote:
>>
>>
>> On 13/12/12 12:30, Henning Rogge wrote:
>>> On 12/13/2012 11:26 AM, Ferry Huberts wrote:
>>>>
>>>>
>>>> On 13/12/12 11:23, Henning Rogge wrote:
>>>>> On 12/13/2012 11:15 AM, Henning Rogge wrote:
>>>>>> My other question would be if "table" should be an integer larger
>>>>>> than 8
>>>>>> bit. Because I think there can be more than 255 routing tables.
>>>>>
>>>>> Okay, seems to be a kernel version issue... lets keep the 255 maximum
>>>>> for the moment (but my comment about the pointer is still valid I
>>>>> think).
>>>>
>>>> I use a pointer to be able to easily signal that it is present or
>>>> not (I
>>>> want to be able to have 0 as a valid table value)
>>>
>>> Maybe we can use an "int" as a parameter and use -1 as the "just take
>>> the default" parameter? Less chances for things to go wrong.
>>>
>>
>>
>> I'd rather not. Then I'd have to ensure the int conforms to - <= int <=
>> 255.
>>
>> If you don't like the current code then I propose to always specify the
>> table explicitly.
>
> This could work too, the tunnel mechanism is only used by "SmartGW" at
> the moment anyways... a new user would need its okay default anyways.
>
> Henning Rogge
>

-- 
Ferry Huberts