search

qca / open-ath9k-htc-firmware

The firmware for QCA AR7010/AR9271 802.11n USB NICs
Other
428 stars 182 forks source link

Only increment xretries statistic for unicast frames #97

Closed DC-fgervais closed 7 years ago

DC-fgervais commented 7 years ago

It looks like the ar5416 only has the AR_ExcessiveRetries flag to let the outside know it didn't receive an ACK.

So in case of broadcast/multicast it gets set even though it's not "excessive retry" per se. These frames just don't acknowledge/retry in a 802.11 transmission.

This patch makes sure broadcast/multicast transmissions won't increment the xretries statistic.

olerem commented 7 years ago

Looks ok. Compiled and working in STA mode.

olerem commented 7 years ago

@erikarn , so probably one more pull before new version ?

erikarn commented 7 years ago

I think we can roll these together. lemme go accept it.

-a

On 23 September 2016 at 03:11, Oleksij Rempel notifications@github.com wrote:

@erikarn https://github.com/erikarn , so probably one more pull before new version ?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/qca/open-ath9k-htc-firmware/pull/97#issuecomment-249153956, or mute the thread https://github.com/notifications/unsubscribe-auth/ABGl7YooVRUrV7nWjfeCuttN1n14-1g_ks5qs6XCgaJpZM4KDTeX .

olerem commented 7 years ago

For the archive, @DC-fgervais what tests are you using for this?

DC-fgervais commented 7 years ago

The easiest test for me is to ping the local broadcast address. For example, my network number is 172.16.8.0/22 so I ping as follow:

ping -c 1 -b 172.16.11.255

Without the patch, the Xretries counter would increment by 1.

I also have wireshark on a monitor interface on another machine to make sure that the over-the-air result is as expected.