Return-path: Received: from wolverine02.qualcomm.com ([199.106.114.251]:36634 "EHLO wolverine02.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752619AbaCXKrw (ORCPT ); Mon, 24 Mar 2014 06:47:52 -0400 From: Kalle Valo To: Janusz Dziedzic CC: "ath10k@lists.infradead.org" , Subject: Re: [PATCH 1/2] ath10k: introduce rx_status htt template References: <1395394437-4048-1-git-send-email-janusz.dziedzic@tieto.com> <87ior37vx1.fsf@kamboji.qca.qualcomm.com> Date: Mon, 24 Mar 2014 12:47:46 +0200 In-Reply-To: (Janusz Dziedzic's message of "Mon, 24 Mar 2014 11:16:54 +0100") Message-ID: <87a9cf7u0d.fsf@kamboji.qca.qualcomm.com> (sfid-20140324_114756_064022_FD8C2A10) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-wireless-owner@vger.kernel.org List-ID: Janusz Dziedzic writes: > On 24 March 2014 11:06, Kalle Valo wrote: >> Janusz Dziedzic writes: >> >>> Introduce rx_status htt template instead of >>> stack version. >>> >>> Signed-off-by: Janusz Dziedzic >> >> Why? What's the benefit from doing this? >> > This is the must in case we would like to report rssi + rates correctly. > I describe this in 2/2 - maybe squash them? Nah, better to keep them separated. But in patch 1 please document why you are doing that, so that I (and others cherry picking ath10k patches) don't need to guess. For example, you can say "in preparation for foo" or something like that. > Currently during heavy traffic we could have: >> 10 packets with START_VALID - correct RSSI + rates >> 10 packets with END_VALID >> 10 packets with START_VALID - correct RSSI + rates >> 10 packets with END_VALID - > ... > > Because of that we have to store rssi + rates for case when END_VALID only. > In other case when using monitor we will see: > - 10 packets with correct rssi/rates > - 10 packets with rssi=-95/rate=6Mbps Ok, I understand now. Please document that in patch 2. And in the future, please double check that your commit logs answer the question "why?". That will save everyone's time :) -- Kalle Valo