[Olsr-users] Fwd: Question about hidden node behavior

[Markus Kittenberger]

On Thu, Jan 13, 2011 at 1:40 PM, Juliusz Chroboczek <
(spam-protected)> wrote:

> > So far i haven't seen a hw queue filled, they are emptied really fast,
> > not even on multi-bssid scenarios,
>
> Think a while about what that means.
>
>  What Nick is saying is that ath9k doesn't normally drop packets.

hmm normally *G

imho a quick peek at the code looks like it usually drops, as soon as its
queues are full,..
(and i asume it`s easy to fill them *G)


> Hence,
> it must apply backpressure somewhere.  If my understanding is correct,
> this happens in ath9k/xmit.c around line 1768.
>
yep here we have ath_tx_start() which has useful return codes,..

but this function is not directly called by the "kernel" ...

it is called by ath9k_tx() @main.c:1323
which is the in ieee80211_ops.tx registered entry point for sending packets
afaics,..

but there the return code of ath_tx_start is not passed back to the ath9k_tx
callee

<snip>
if (ath_tx_start(hw, skb, &txctl) != 0) {
ath_print(common, ATH_DBG_XMIT, "TX failed\n");
goto exit;
}

return 0;
exit:
dev_kfree_skb_any(skb);
return 0;
}
</snip>

imho the second return 0; looks wrong, as this way callees (probably
mac80211) will assume the packet was sent succesfully,..

Markus


>                                        Juliusz
>
> --
> Olsr-users mailing list
> (spam-protected)
> http://lists.olsr.org/mailman/listinfo/olsr-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.olsr.org/pipermail/olsr-users/attachments/20110113/94b0e9c5/attachment.html>