Return-path: Received: from mail.atheros.com ([12.36.123.2]:14292 "EHLO mail.atheros.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751387AbZAaIkw (ORCPT ); Sat, 31 Jan 2009 03:40:52 -0500 Received: from mail.atheros.com ([10.10.20.108]) by sidewinder.atheros.com for ; Sat, 31 Jan 2009 00:40:52 -0800 Date: Sat, 31 Jan 2009 14:10:45 +0530 From: Vasanthakumar Thiagarajan To: Bob Copeland CC: Vasanth Thiagarajan , "linville@tuxdriver.com" , "linux-wireless@vger.kernel.org" , Luis Rodriguez , Jouni Malinen , "ath9k-devel@lists.ath9k.org" Subject: Re: [PATCH] ath9k: Fix LED blink pattern Message-ID: <20090131084044.GA5999@vasanth-laptop> (sfid-20090131_094103_238430_68A0B3E4) References: <1233231739-4402-1-git-send-email-vasanth@atheros.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In-Reply-To: Sender: linux-wireless-owner@vger.kernel.org List-ID: On Thu, Jan 29, 2009 at 08:47:04PM +0530, Bob Copeland wrote: > > + struct delayed_work ath_led_blink_work; > > + int led_on_duration; > > + int led_off_duration; > > + int led_on_cnt; > > + int led_off_cnt; > > This is something that has come up from time to time from users for > ath5k as well. Half of the drivers now have their own timers for > turning leds on or off. > > Can we just add a decent blink algorithm to mac80211? Right now for > rx it just toggles every other packet. For tx, it's slightly smarter > but not overly so. It seems like this is one of those things everyone > wants to do and we could easily make it a standard service. ath9k maps all four led functionality (radio, assoc, rx and tx) into a single LED, I think ath5k is also doing in the similar way. But other drivers can map their led functionality differently in their h/w where the pattern will differ. mac80211 is already signalling appropriate leds, imho it is up to the driver to define the pattern based on the number of leds available/configured. > > -- > Bob Copeland %% www.bobcopeland.com