From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Ujfalusi Subject: Re: [PATCH 5/6] dmaengine: omap-dma: Remove mapping between virtual channels and requests Date: Tue, 24 Feb 2015 19:01:41 +0200 Message-ID: <54ECAE75.3070501@ti.com> References: <1424787683-19151-1-git-send-email-peter.ujfalusi@ti.com> <1424787683-19151-6-git-send-email-peter.ujfalusi@ti.com> <20150224142833.GI8670@n2100.arm.linux.org.uk> Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-Reply-To: <20150224142833.GI8670@n2100.arm.linux.org.uk> Sender: linux-kernel-owner@vger.kernel.org To: Russell King - ARM Linux Cc: vinod.koul@intel.com, Tony Lindgren , grant.likely@linaro.org, dmaengine@vger.kernel.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org, robh+dt@kernel.org, nm@ti.com List-Id: linux-omap@vger.kernel.org On 02/24/2015 04:28 PM, Russell King - ARM Linux wrote: > On Tue, Feb 24, 2015 at 04:21:22PM +0200, Peter Ujfalusi wrote: >> Do not direct map the virtual channels to sDMA request number. When = the >> sDMA is behind of a crossbar this direct mapping can cause situation= s when >> certain channel can not be requested since the crossbar request numb= er >> will no longer match with the sDMA request line. >> The direct mapping for virtual channels with HW request lines will m= ake it >> harder to implement MEM_TO_MEM mode for the driver. >=20 > I assume when you talk about MEM_TO_MEM, you're referring to a DMA_ME= MCPY > driver. >=20 > mem2mem should not be handled by the slave driver. This should be a > separate DMA engine driver structure which does not have DMA_SLAVE se= t. >=20 > See how amba-pl08x handles this. Thanks for the pointer. I'm just planning to add the DMA_MEMCPY support= for omap-dma. With that in place we can convert the remaining legacy API us= ers to use dmaengine (as I recall all of them are using DMA for memcopy) --=20 P=E9ter From mboxrd@z Thu Jan 1 00:00:00 1970 From: peter.ujfalusi@ti.com (Peter Ujfalusi) Date: Tue, 24 Feb 2015 19:01:41 +0200 Subject: [PATCH 5/6] dmaengine: omap-dma: Remove mapping between virtual channels and requests In-Reply-To: <20150224142833.GI8670@n2100.arm.linux.org.uk> References: <1424787683-19151-1-git-send-email-peter.ujfalusi@ti.com> <1424787683-19151-6-git-send-email-peter.ujfalusi@ti.com> <20150224142833.GI8670@n2100.arm.linux.org.uk> Message-ID: <54ECAE75.3070501@ti.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 02/24/2015 04:28 PM, Russell King - ARM Linux wrote: > On Tue, Feb 24, 2015 at 04:21:22PM +0200, Peter Ujfalusi wrote: >> Do not direct map the virtual channels to sDMA request number. When the >> sDMA is behind of a crossbar this direct mapping can cause situations when >> certain channel can not be requested since the crossbar request number >> will no longer match with the sDMA request line. >> The direct mapping for virtual channels with HW request lines will make it >> harder to implement MEM_TO_MEM mode for the driver. > > I assume when you talk about MEM_TO_MEM, you're referring to a DMA_MEMCPY > driver. > > mem2mem should not be handled by the slave driver. This should be a > separate DMA engine driver structure which does not have DMA_SLAVE set. > > See how amba-pl08x handles this. Thanks for the pointer. I'm just planning to add the DMA_MEMCPY support for omap-dma. With that in place we can convert the remaining legacy API users to use dmaengine (as I recall all of them are using DMA for memcopy) -- P?ter From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753059AbbBXRCY (ORCPT ); Tue, 24 Feb 2015 12:02:24 -0500 Received: from bear.ext.ti.com ([192.94.94.41]:33014 "EHLO bear.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751682AbbBXRCW (ORCPT ); Tue, 24 Feb 2015 12:02:22 -0500 Message-ID: <54ECAE75.3070501@ti.com> Date: Tue, 24 Feb 2015 19:01:41 +0200 From: Peter Ujfalusi User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.4.0 MIME-Version: 1.0 To: Russell King - ARM Linux CC: , Tony Lindgren , , , , , , , , Subject: Re: [PATCH 5/6] dmaengine: omap-dma: Remove mapping between virtual channels and requests References: <1424787683-19151-1-git-send-email-peter.ujfalusi@ti.com> <1424787683-19151-6-git-send-email-peter.ujfalusi@ti.com> <20150224142833.GI8670@n2100.arm.linux.org.uk> In-Reply-To: <20150224142833.GI8670@n2100.arm.linux.org.uk> Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 02/24/2015 04:28 PM, Russell King - ARM Linux wrote: > On Tue, Feb 24, 2015 at 04:21:22PM +0200, Peter Ujfalusi wrote: >> Do not direct map the virtual channels to sDMA request number. When the >> sDMA is behind of a crossbar this direct mapping can cause situations when >> certain channel can not be requested since the crossbar request number >> will no longer match with the sDMA request line. >> The direct mapping for virtual channels with HW request lines will make it >> harder to implement MEM_TO_MEM mode for the driver. > > I assume when you talk about MEM_TO_MEM, you're referring to a DMA_MEMCPY > driver. > > mem2mem should not be handled by the slave driver. This should be a > separate DMA engine driver structure which does not have DMA_SLAVE set. > > See how amba-pl08x handles this. Thanks for the pointer. I'm just planning to add the DMA_MEMCPY support for omap-dma. With that in place we can convert the remaining legacy API users to use dmaengine (as I recall all of them are using DMA for memcopy) -- Péter