[Make-wifi-fast] [PATCH v2] mac80211: Move crypto IV generation to after TXQ dequeue.

[Dave Taht]

On Wed, Aug 17, 2016 at 9:49 PM, Johannes Berg
<johannes at sipsolutions.net> wrote:
> Hi,
>
> You need to work on coding style, a lot of your indentation is
> completely messed up.
>
>> +     switch (sdata->vif.type) {
>> +     case NL80211_IFTYPE_STATION:
>> +             if (sdata->u.mgd.use_4addr) {
>> +                     pn_offs = 30;
>> +                     break;
>> +             }
>> +             pn_offs = 24;
>> +             break;
>> +     case NL80211_IFTYPE_AP_VLAN:
>> +             if (sdata->wdev.use_4addr) {
>> +                     pn_offs = 30;
>> +                     break;
>> +             }
>> +             /* fall through */
>> +     case NL80211_IFTYPE_ADHOC:
>> +     case NL80211_IFTYPE_AP:
>> +             pn_offs = 24;
>> +             break;
>> +     default:
>> +             return;
>> +     }
>> +
>> +     if (sta->sta.wme) {
>> +             pn_offs += 2;
>> +     }
>
> I think you just reinvented ieee80211_hdrlen(). No?
>
>> -     if (fast_tx->pn_offs) {
>> -             u64 pn;
>> -             u8 *crypto_hdr = skb->data + fast_tx->pn_offs;
>
> No need to undo the pn_offs optimisation for the !txq case, you can
> pass it in to the new function that will fill it.
>
> However, you're still doing it wrong - now you haven't fixed anything
> for TKIP, which won't hit the fastpath.

well, we're getting there. the results of both patch attempts were
really nice, and brought encrypted performance with fq back into line
with unencrypted. Still running crypted tests as I write...

So fixing TKIP would be next, forcing the AP to use that? What other
scenarios do we have to worry about? WDS?


> johannes
> _______________________________________________
> Make-wifi-fast mailing list
> Make-wifi-fast at lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/make-wifi-fast



-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
http://blog.cerowrt.org