carl9170 - single packet injection causes 3 packets to be sent.

carl9170 - single packet injection causes 3 packets to be sent.

[marco]

Hello,

I noticed a problem with packet injection on the carl9170.  When I inject a
packet, I see 3 packets go out.  I see them with a secondary sniffer on a
different PC.  One interesting thing is, if the packet has a broadcast DA mac,
only one packet is injected (as I would expect).  I poked around in the
carl9170 driver a little, and it *appears* that it is only sending the single
packet across the usb stick, so I'm guessing the issue is with the
firmware/hardware (?).  The only other thing I can think of is perhaps a
radiotap header misconfiguration.  Is this a known problem or am I the only one
seeing such a thing?  I've seen it across a few driver/firmware revisions.

I'm running: 
[    8.310000] usb 1-1: driver   API: 1.9.4 2011-08-15 [1-1] 
[    8.310000] usb 1-1: firmware API: 1.9.4 2011-06-30

I'm currently running compat-wireless-2011-11-22 on a 2.6.37 kernel.

Thanks for any insight you can provide,

Marco Fonseca

Re: carl9170 - single packet injection causes 3 packets to be sent.

[Daniel Smith]

On 6/26/2012 6:26 PM, marco wrote:
> Hello,
>
> I noticed a problem with packet injection on the carl9170.  When I inject a
> packet, I see 3 packets go out.  I see them with a secondary sniffer on a
> different PC.  One interesting thing is, if the packet has a broadcast DA mac,
> only one packet is injected (as I would expect).  I poked around in the
> carl9170 driver a little, and it *appears* that it is only sending the single
> packet across the usb stick, so I'm guessing the issue is with the
> firmware/hardware (?).  The only other thing I can think of is perhaps a
> radiotap header misconfiguration.  Is this a known problem or am I the only one
> seeing such a thing?  I've seen it across a few driver/firmware revisions.
>
> I'm running:
> [    8.310000] usb 1-1: driver   API: 1.9.4 2011-08-15 [1-1]
> [    8.310000] usb 1-1: firmware API: 1.9.4 2011-06-30
>
> I'm currently running compat-wireless-2011-11-22 on a 2.6.37 kernel.
>
> Thanks for any insight you can provide,
>
> Marco Fonseca
> --
> To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
  Hi Marco,

I have dealt with a similar issue with the ath9k chipset. On those chips 
there is a retries parameter in the TX descriptor. By default the ath9k 
driver would set this to three as well. I forget where all I had to make 
changes (probably some in mac80211 as well as ath9k) so that I could 
pass the retries field in the radiotap header and have it honored by 
ath9k. Hopefully this help point you in the right direction.

V/r,
Daniel