Re: 32 core net-next stack/netfilter "scaling"

[Patrick McHardy <kaber@trash.net>]

Eric Dumazet wrote:
> Patrick McHardy a écrit :
>   
>> Thats an interesting test-case, but one lock per conntrack just for
>> TCP tracking seems like overkill. We're trying to keep the conntrack
>> stuctures as small as possible, so I'd prefer an array of spinlocks
>> or something like that.
>>     
>
> Yes, this is wise. Current sizeof(struct nf_conn) is 220 (0xdc) on 32 bits,
> probably rounded to 0xE0 by SLAB/SLUB. I will provide a new patch using
> an array of say 512 spinlocks. (512 spinlocks use 2048 bytes if non
> debuging spinlocks, that spread to 32 x 64bytes cache lines)
>   

Sounds good, but it should be limited to NR_CPUS I guess.
> However I wonder if for very large number of cpus we should at least ask conntrack 
> to use hardware aligned "struct nf_conn" to avoid false sharing
>   

I'm not sure that is really a problem in practice, you usually have quite a
few inactive conntrack entries and false sharing would only happen when two
consequitive entries are active.