From mboxrd@z Thu Jan  1 00:00:00 1970
From: vinod.koul@intel.com (Vinod Koul)
Date: Tue, 14 May 2013 11:09:43 +0530
Subject: [PATCH 0/2 V4] DMA: AT91: Fix get residue function in at_hdmac.c
In-Reply-To: <1368154798-9279-1-git-send-email-elen.song@atmel.com>
References: <1368154798-9279-1-git-send-email-elen.song@atmel.com>
Message-ID: <20130514053943.GD27639@intel.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Fri, May 10, 2013 at 10:59:58AM +0800, Elen Song wrote:
> The residue actually means the bytes remain in transfer identified by current cookie.
> The get residue api did not return the correct values in at_hdmac.c.
> This patch serial focus on fix the issue.
> 
> This patch serials is based on DMA: AT91: Get residual bytes in dma buffer v2.
> It is split into 3 aimed patch, the first one is to get transmit buswidth,
> the second one is to get resiude in cookie, it use the buswidth to get residual bytes,
> the third part is to handle the status that there are still some bytes remain in dma fifo.
Applied, Thanks

--
~Vinod
> 
> Compare to v2, this patch serials improve:
> 1) consider buswidth to get correct residue.
> 2) handle dma fifo empty status.
> 
> Compare to v3, this patch serials improve:
> 1) To state clearly, replace buswidth with transfer width.
> 2) Do not pause dma when get residue.
> 3) Let issue pending handler remain work if data remain in dma fifo, polling empty flag is too time consuming.
> 
> Elen Song (2):
>   DMA: AT91: Get transfer width
>   DMA: AT91: Get residual bytes in dma buffer
> 
>  drivers/dma/at_hdmac.c      |  151 ++++++++++++++++++++++++++++++++++++++-----
>  drivers/dma/at_hdmac_regs.h |    5 ++
>  2 files changed, 139 insertions(+), 17 deletions(-)
> 
> -- 
> 1.7.9.5
>