From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bernhard Urban Date: Sat, 17 Nov 2012 20:12:44 +0100 Subject: [ath9k-devel] [PATCH] ath9k_htc: update RSSI values only when the device is associated In-Reply-To: <509E1B47.3050107@openwrt.org> References: <1352480966-17462-1-git-send-email-lewurm@gmail.com> <509E1B47.3050107@openwrt.org> Message-ID: <20121117191244.GA18782@phenom> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ath9k-devel@lists.ath9k.org Thanks for the reviews! On Sat, Nov 10, 2012 at 9:42 AM, Sujith wrote: > Looks good, but a cleaner fix would be to update the RSSI only for > beacons that are destined for the station. You mean like in recv.c? There's a `is_mybeacon' field, which is computed in `ath_rx_tasklet(...)' and used in `ath9k_process_rssi(...)'. Adding this check to the if-guard changes nothing at the result, so I think it's fine if you consider it as cleaner. See attached patch. Btw, could you please elaborate what `is_mybeacon' represents? On Sat, Nov 10, 2012 at 10:15 AM, Felix Fietkau wrote: > Please check how RSSI is handled in ath9k, and use that as reference. > The per-packet RSSI should not be set from some internal average, it > should contain the real RSSI value of the packet. This change disables > RSSI reporting for anything but station mode, which is not a good idea. Replacing if (priv->num_sta_assoc_vif != 0) with if (priv->ah->opmode == NL80211_IFTYPE_STATION) works out well (which makes sense to me). However, adding `&& priv->ah->opmode == NL80211_IFTYPE_ADHOC' (like in ath9k/recv.c) produces bad results like before. I don't understand the effect, as my device is always in "managed mode" according to `iw wlan0 info'. I'm right by assuming "managed mode" is the same as NL80211_IFTYPE_STATION? Also, I don't know what you mean with "some internal average". Can you please elaborate that? Thanks for your help, Bernhard