Re: [lwip-devel] Removing the restriction on number of sockets

[Joan Lledó]

Hello,

I've wrote a patch[1] that implements this new macro.

Since you haven't decided whether use a pool for sockets, that isn't
included in the patch and the old array is still there. I'll be glad
to remove the old code and create a new pool if you want me to.

Please note that I wrote and tested my patch for the version 2.0.1
which I'm using, this patch is an adapted version and may contain
errors.

-----------
[1] https://savannah.nongnu.org/patch/index.php?9431

2017-08-08 15:38 GMT+02:00 Joel Cunningham <address@hidden>:
>
> On Aug 8, 2017, at 8:16 AM, Dirk Ziegelmeier <address@hidden> wrote:
>
> Yes, for many sockets a linked list may become slow, but usually every
> socket implies a netconn and a PCB -> PCBs are iterated for every incoming
> packet, too.
>
>
> The PCB iteration is a good point to contrast.  I would expect input packets
> to be incoming at a higher frequency then socket API calls
>
> If we add the same “caching” feature that our PCB lists have, where the most
> recently used are moved to head of the list, that may negate some of the
> performance loss
>
> Joel
>
> _______________________________________________
> lwip-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/lwip-devel
>