Hello.

Not so long time ago we faced with some issues around IBSS mode in our lab.

Distro          : OpenWRT 18.06 with vanilla ath10k or ath10k-ct
System       : Qualcomm Atheros QCA956X ver 1 rev 0
Radio        : QCA988x
Firmware    : 10.1-ct beta-21 (full-htt-mgt-community)
Machine      : Ubiquiti UniFi-AC-LITE

We have 5 nodes mesh network for research purposes.
Sometimes throughput is good (300-350 mbps), but it can go down to
10-20 kbps or even bytes per second.
This issue looks like MTU related, but we can’t be sure and asking for your help

We get about 150-200 mbps when MTU is decreased to 1400 or lower.
MTU 1500 is also good, but sometimes we see a lot of dropped RX packets (1-5% )
MTU 1528 is the last good variant in our case with a little bit more
dropped packets (5-10%)
MTU 1529-1564 and more shows 0 kbps iperf throughput and we see that
all rx packets are dropped.

From time to time it looks like firmware changes it state and some
packets can be received correctly, but we don’t know why.

While  we did some debugging with ftrace kernel framework and found
out that driver drops frame in htt_rx.c:ath10k_htt_rx_h_unchain
function. As we think, that’s because  of aggregation/chaining in
wifi-native mode.
Also we found a comment:
  /* FIXME: Current unchaining logic can only handle simple case of raw
   * msdu chaining. If decapping is other than raw the chaining may be
   * more complex and this isn't handled by the current code. Don't even
   * try re-constructing such frames - it'll be pretty much garbage.
   */
Both vanilla ath10k and ath10k-ct are related to this issue.

Could you please help us resolve this problem by answering to
following questions:
1) What is actually means chaining?
2) Are there a valid frames on skbuff queue, but current driver code
can’t parse it? So is there any chance to correctly extract every
chained frame?
3) Is there a way to disable frames aggregation/chaining in firmware?

May be Ben Greear from Candelatech can tell us more about things
related to ath10k-ct  firmware.

-- 
Best Regards,
Lukonin Kirill

_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k

Reply via email to