From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Ujfalusi Subject: Re: [PATCH 03/17] ASoC: Add a generic dmaengine_pcm driver Date: Thu, 18 Apr 2013 09:30:47 +0200 Message-ID: <516FA127.40606@ti.com> References: <1366046404-8759-1-git-send-email-lars@metafoo.de> <1366046404-8759-4-git-send-email-lars@metafoo.de> <516E6FD8.5040400@ti.com> <516E7DF1.6080905@metafoo.de> <516E9E5E.4090901@ti.com> <516EEDAD.40706@metafoo.de> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: Received: from comal.ext.ti.com (comal.ext.ti.com [198.47.26.152]) by alsa0.perex.cz (Postfix) with ESMTP id BCB4F265E15 for ; Thu, 18 Apr 2013 09:30:55 +0200 (CEST) In-Reply-To: <516EEDAD.40706@metafoo.de> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: alsa-devel-bounces@alsa-project.org Sender: alsa-devel-bounces@alsa-project.org To: Lars-Peter Clausen Cc: Ola Lilja , alsa-devel@alsa-project.org, Rajeev Kumar , Russell King , Ryan Mallon , Stephen Warren , Vinod Koul , Liam Girdwood , Hartley Sweeten , Mark Brown , Laxman Dewangan , Sebastien Guiriec , Markus Pargmann , Shawn Guo , Lee Jones , Jarkko Nikula List-Id: alsa-devel@alsa-project.org Hi Lars, On 04/17/2013 08:45 PM, Lars-Peter Clausen wrote: > On 04/17/2013 03:06 PM, Peter Ujfalusi wrote: >> On 04/17/2013 12:48 PM, Lars-Peter Clausen wrote: >>> This sounds like it needs more than just support for custom channel nam= es. >>> The generic driver currently expects that there is exactly one DMA chan= nel >>> for capture and one for playback (well and in case the DAI link is simp= lex >>> only either one of them). >>> >>> What are these 8 channels exactly used for? >> >> We have 7 FIFOs which can be configured either playback or capture. The = last >> channel is used for debug purposes only right now. >> Basically we can dynamically configure the 8 DMA channel for playback/ca= pture >> and the mixing of incoming streams done inside of the AESS. >> We have 7 DAIs and we have playback/capture on those (some only have pla= yback >> some only capture some duplex). Right now we have kind of static DMA bin= dings >> for them, I'm working on a dynamic model right now. > = > Hm, if you need to re-assign the DMA channels dynamically at runtime it > maybe better to stick with a custom PCM driver. Should I drop OMAP from t= his > series for now? I think this would the best for now. However I did tested the series and it worked with the code upstream. Later I can revisit this and probably have two PCM driver for OMAP, but I'm not sure about this right now, I will know more when I have rebased my AESS/ABE work on top of 3.10-rc1. Can you drop the OMAP patches for now? Thanks, P=E9ter