Johannes Berg wrote:
> On Fri, 2007-06-29 at 13:51 +0200, Patrick McHardy wrote:
>
>
>>Do multicast groups have to have a seperate name? Or would it suffice
>>to have them associated with the genl family and be able to find out
>>the starting group number? In that case something like
>>
>>struct genl_mc_groups {
>> struct genl_family *family or char *family_name or similar;
>> unsigned int group_off;
>> unsigned int group_num;
>> unsigned long groups[];
>>};
>>
>>seems to make more sense since you only need a single struct
>>per family.
>
>
> Hm. For me that'd work too but Jamal wanted dynamically allocated groups
> if I understood him correctly. I'm not too concerned with that case, I'd
> think most people know the groups up-front. On the other hand, I can see
> something like a group per netdev or whatever other instance too.
Maybe use a mix. Use the bitmap, but allow families to register
multiple of them. In the common case it would only be a single
one, but it would be possible to register groups dynamically.
>>Why would you care about holes? If you really want to use sparse
>>bitmaps that would complicate the code a lot.
>
>
> No, not sparse bitmaps, but the bitmap could have a hole when a family
> goes away, and we could reuse that group number later. If we have it in
> a bitmap we know without checking all group IDs.
Right.
-
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
