Return-path: Received: from sabertooth01.qualcomm.com ([65.197.215.72]:11375 "EHLO sabertooth01.qualcomm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750731Ab3GLEX0 (ORCPT ); Fri, 12 Jul 2013 00:23:26 -0400 From: Kalle Valo To: Johannes Berg CC: Sarah Sharp , Xenia Ragiadakou , OPW Kernel Interns List , , , Steven Rostedt Subject: Re: Help adding trace events to xHCI References: <51DB0257.1010709@gmail.com> <20130711162002.GA5240@xanatos> <1373562533.8201.33.camel@jlt4.sipsolutions.net> Date: Fri, 12 Jul 2013 07:23:16 +0300 In-Reply-To: <1373562533.8201.33.camel@jlt4.sipsolutions.net> (Johannes Berg's message of "Thu, 11 Jul 2013 19:08:53 +0200") Message-ID: <8738rk5ruz.fsf@kamboji.qca.qualcomm.com> (sfid-20130712_062344_672719_B25E9205) MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: linux-wireless-owner@vger.kernel.org List-ID: Johannes Berg writes: >> (Mentors and wireless folks, we're struggling a bit with adding trace >> events to the xHCI USB host controller driver. I'm trying to look at >> the ath6kl driver trace events as an example. We could use some help >> and/or advice.) > > Those were in turn modelled on mac80211, cfg80211 and/or iwlwifi, I > think; might be worth also looking there. In general though, it's all > pretty similar. Correct, I used your code as example. > The message tracing was really more of an after-thought in iwlwifi (and > ath6kl as well I guess) because we already had a lot of debug messages > and capturing it all together can be useful. > > However, tracing all the debug messages is actually fairly expensive, I > think in part because of the string formatting and in part because of > the always-max allocations in the ringbuffer. The reason why added message tracing to ath6kl was that I wanted to combine wpasupplicant, cfg80211 and ath6kl debug messages all into file with synchronised timestamps. In this case I wasn't actually concerned about performance at all. -- Kalle Valo