Return-path: Received: from mail-yw0-f51.google.com ([209.85.213.51]:56699 "EHLO mail-yw0-f51.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756717Ab2FYNyk (ORCPT ); Mon, 25 Jun 2012 09:54:40 -0400 Received: by yhnn12 with SMTP id n12so3326350yhn.10 for ; Mon, 25 Jun 2012 06:54:39 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: <87395j4j3y.fsf@purkki.adurom.net> References: <1340610505-7713-1-git-send-email-eliad@wizery.com> <87395j4j3y.fsf@purkki.adurom.net> Date: Mon, 25 Jun 2012 16:54:39 +0300 Message-ID: (sfid-20120625_155443_992173_83726872) Subject: Re: [PATCH] mac80211: don't require associated->beacon_ies for ps From: Eliad Peller To: Kalle Valo Cc: Johannes Berg , linux-wireless@vger.kernel.org Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Mon, Jun 25, 2012 at 4:27 PM, Kalle Valo wrote: >> beacon_ies is needed only in order to extract the dtim >> period. However, even if it's missing we can still enter >> ps with dtim=1 (which also happens if the TIM ie is invalid). >> >> Most drivers don't use conf.max_sleep_period/ps_dtim_period >> anyway, and this check prevents them from entering ps if >> they don't have beacon (but only probe response), even though >> the beacon is not needed at all. > > Does this increase the chances of accidentally using dtim 1 even though > AP has dtim > 1? I'm just worried that it's difficult to detect cases > when we are forcing dtim to 1 and nobody might not notice it. How often > will this happen? > doesn't dtim=1 is still better than not entering ps at all? i think the only bad effect of using dtim=1 (instead of greater value) is wrt power saving. but entering psm with dtim=1 is still better than not entering psm at all. > Should we add a warning message for that case? Or few years ago we > talked about waiting for a beacon before enabling ps mode, should we > reconsider that? (Or maybe it's implemented already?) the current behavior is not entering ps if there is no beacon. however, as i've just explained, i think we can drop this requirement. i don't think we should add a warning, as this might be a legitimate behavior (as probe response might be enough in order to connect). Eliad.