From mboxrd@z Thu Jan 1 00:00:00 1970 From: gtolon at inti.gob.ar Date: Wed, 21 Mar 2012 10:11:49 -0300 Subject: [ath9k-devel] Question about Atheros AR9220 In-Reply-To: References: <000401ccf8ae$afafbe30$0f0f3a90$@tamu.edu> <4F544938.8060003@qca.qualcomm.com> <4F54FAAB.8010007@tamu.edu> <4F676E35.5050606@inti.gob.ar> Message-ID: <4F69D395.8090004@inti.gob.ar> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: ath9k-devel@lists.ath9k.org Thank you for the answer Adrian, i'll try with that. By the way, i've found in the code a register called RFCNT, i guess it's the RXCNT you mentioned. Gabriel El 20/03/2012 12:30 a.m., Adrian Chadd escribi?: > Hi, > > It's likely a mixture of both. > > * The channel survey gives you an idea of what percentage of time the > NIC thinks the air is busy (RCCNT). RXCNT is the time spent _RX_ing > frames, even if they (a) fail FCS/CRC, or (b) aren't destined for it. > It's still decoding them. So, you can calculate a rough estimation of > the time spent doing "other" by taking RCCNT and subtracting RXCNT. > * You should see FCS/CRC errors pop up in RX frames. There's bits in > the RX descriptor that get set. There should be CRC error statistics > kept by the driver somewhere accessible by debugfs. > > > Adrian >