From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [RFC] Generic DMA chaining APIs Date: Fri, 10 Aug 2007 00:33:03 -0700 Message-ID: <20070810073303.GF8658@atomide.com> References: <9C23CDD79DA20A479D4615857B2E2C4701120BC1@dlee13.ent.ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <9C23CDD79DA20A479D4615857B2E2C4701120BC1@dlee13.ent.ti.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-omap-open-source-bounces@linux.omap.com Errors-To: linux-omap-open-source-bounces@linux.omap.com To: "Syed Mohammed, Khasim" Cc: Linux OMAP List-Id: linux-omap@vger.kernel.org * Syed Mohammed, Khasim [070605 22:07]: > Hi all: > > I would like to port Generic DMA chaining API support from TI's 2430 DMA driver to our OMAP GIT DMA driver. > > - Basically these APIs give flexibility for handling DMA chaining for drivers like McBSP (for audio), McSPI (for LCD), UART, etc. > > - Chaining is very important for drivers like Audio, LCD etc. > > - Our current DMA driver on GIT does support APIs to enable /configure DMA logical channel chaining, but they have to be again used in client drivers to implement the queuing logic. > > - Since there are multiple drivers in need of these kind of chaining support, it's preferable to implement them in a generic way so that all other drivers can leverage the code instead of creating a redundant copy in every driver. > > A code snippet (from TI's DMA driver) is as shown below. Please do let me know if you find any issues with this kind of implementation/approach before I start (this week) this porting activity. > > Thanks in advance. > > APIs to be ported are: > ===================== > omap_request_dma_chain > omap_free_dma_chain > omap_dma_chain_a_transfer > omap_modify_dma_chain_params > omap_start_dma_chain_transfers > omap_stop_dma_chain_transfers > > omap_dma_chain_status > omap_get_dma_chain_index > omap_get_dma_chain_dst_pos > omap_get_dma_chain_src_pos The new functions look OK to me. Regards, Tony