The patch was submitted because we did see the issue. On Jan 4, 2011, at 5:43 PM, Ethan Du wrote: > Hi: > > I have not seen case (1) and (2) across >20 types of micro SD > cards, and >10 types of Sandisk/Toshiba eMMC chips, so I believe the > 2GB condition is ok. > Probing of the OCR bit to chose between byte and sector mode will > also work, which I have tried in a large scope. > > Regards, > -Ethan > > On Wed, Jan 5, 2011 at 12:59 AM, Sujit Reddy <sujitreddy243@xxxxxxxxx> wrote: >> Hi, >> >> Any comments on below mentioned query? >> >> Thanks >> Sujit >> >> On Thu, Dec 23, 2010 at 1:35 PM, Sujit Reddy <sujitreddy243@xxxxxxxxx> wrote: >>> Hi, >>> >>> According to eMMC spec v4.3, Section 6.1 says that greater than 2GB >>> density cards are sector addressable and less than 2GB are byte addressable. >>> But Section 7.3.3 says that OCR bit 30 needs to be used which access mode >>> the host must use for all its future transactions. >>> >>> In mainline kernel the support for less than 2GB cards is by checking >>> the density of card. >>> If it is greater than 2GB it is assumed to be supporting sector addressing. >>> >>> But actually we need to check the OCR value to determine the correct access mode >>> See patch (http://www.spinics.net/lists/linux-mmc/msg01466.html) by Philip. >>> >>> Some eMMC cards even though less than 2GB have SEC_COUNT value defined >>> in EXT_CSD register but does not support sector access mode. >>> >>> I have a couple of questions here: >>> 1) Are there any cards that support only byte access mode even though >>> the density is >>> greater than 2GB? >>> 2) If there are cards which have capacity less than 2GB and supports only sector >>> access mode then what's the solution? I am not sure If there are any such cards. >>> >>> Philip's patch handles both these scenarios effectively. >>> Can anyone mention the reason not taking Philip's patch into account, >>> which I feel is much >>> more fool proof than the implementation in mainline kernel >>> (http://www.mail-archive.com/linux-mmc@xxxxxxxxxxxxxxx/msg02307.html)? >>> >>> Thanks >>> Sujit >>> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-mmc" in >> the body of a message to majordomo@xxxxxxxxxxxxxxx >> More majordomo info at http://vger.kernel.org/majordomo-info.html >> -- To unsubscribe from this list: send the line "unsubscribe linux-mmc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
