From mboxrd@z Thu Jan 1 00:00:00 1970 From: robert.jarzmik@free.fr (Robert Jarzmik) Date: Thu, 05 May 2016 21:14:40 +0200 Subject: [RFC PATCH 2/7] ALSA: ac97: add an ac97 bus In-Reply-To: <20160504162200.GX6292@sirena.org.uk> (Mark Brown's message of "Wed, 4 May 2016 17:22:00 +0100") References: <1462050939-27940-1-git-send-email-robert.jarzmik@free.fr> <1462050939-27940-3-git-send-email-robert.jarzmik@free.fr> <20160503162938.GT6292@sirena.org.uk> <87twiforl3.fsf@belgarion.home> <20160504162200.GX6292@sirena.org.uk> Message-ID: <87lh3opba7.fsf@belgarion.home> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Mark Brown writes: > On Tue, May 03, 2016 at 09:43:20PM +0200, Robert Jarzmik wrote: > >> You probably mean the BITCLK clock. > >> What is a bit pesky about this clock is that it can either be mastered by >> digital controller and the codec is a slave, or the other way around. > > That's a bit surprising - I've never encountered a system that > impelemnts this, it may be permitted by the spec but it's always the > CODEC. The master clock from the CODEC is often provided by the SoC but > I've not seen systems where anything other than the CODEC drives the > actual AC'97 bus. Ok, so let's assume a one direction. I'll have a look how I could add this clock to the probe() exactly as amba_probe() does. > It's not really adding anything though, it's just clumsy wording on > their part - it's not like we need to distinguish this from analogue > or any other type of AC'97 controllers. > >> Now if you prefer "ac97_controller" or something like that, that's as you wish, >> the name does not matter that much to me ;) > > Yes. Okay, let me add this to my todo list for v2. I already have suspend/resume/pm_suspend/pm_resume. Cheers. -- Robert