From mboxrd@z Thu Jan 1 00:00:00 1970 From: Simon Horman Subject: Re: [PATCH v4 0/4] UHS-I SDR-104 support for sh_mobile_sdhi Date: Thu, 11 Aug 2016 10:43:02 +0200 Message-ID: <20160811084301.GF30475@verge.net.au> References: <1469592803-13842-1-git-send-email-horms+renesas@verge.net.au> <20160810131224.GB1741@katana> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20160810131224.GB1741@katana> Sender: linux-renesas-soc-owner@vger.kernel.org To: Wolfram Sang Cc: Wolfram Sang , Ulf Hansson , Magnus Damm , linux-mmc@vger.kernel.org, linux-renesas-soc@vger.kernel.org List-Id: linux-mmc@vger.kernel.org On Wed, Aug 10, 2016 at 03:12:24PM +0200, Wolfram Sang wrote: > Hi Simon, > > > In this patchset I have attempted to address all review received > > for the v3 patch-set. I also believe that I have resolved a problem > > where tuning would timeout under some circumstances: this seems to > > have been due to several bugs introduced between v1 and v3. > > So, you didn't see timeouts anymore? Yes, that is correct. I have not seen them at all with this patchset. I will extend my testing as I integrate these changes on other SoCs and boards. > > * Currently I am seeing speeds of up to 48MB/s with these patches and > > 30MB/s without using a SanDisk Extreme Pro 8Gb microSDHC UHS-1 card. > > > > * I am also seeing 45MB/s with these patches and 34MB/s without using > > a SanDisk Ultra 64Gb microSDXC UHS-1 card. > > I got similar numbers for my SanDisk and Samsung cards. While the SDR104 > speed is not that what I hoped for, it is still an improvement over > SDR50. We are currently investigating if DMA causes a bottleneck > somewhere perhaps. Except for a minor comment, I'd think these patches > are good to go in and we could improve incrementally from here: > > Tested-by: Wolfram Sang > > Maybe Ulf can have another high-level view on those. I guess he is more > experienced with the tuning stuff? Yes, of course such input would be welcome. I'd like to work towards getting these changes merged.