Hi Mr.Jeon One question...if we used predefined transfer, didn't send stop-command? then i think that didn't need to enter this condition..how about this? In __dw_mci_start_reqeust() function(at your patch) if (mrq->stop) host->stop_cmdr = dw_mci_prepare_command(slot->mmc, mrq->stop); If i misunderstood something, plz let me know...Thanks. Tested-by: Jaehoon Chung <jh80.chung@xxxxxxxxxxx> On 09/28/2011 05:23 PM, Seungwon Jeon wrote: > Andrei Warkentin wrote: >> 2011/9/26 Seungwon Jeon <tgih.jun@xxxxxxxxxxx>: >>> Andrei Warkentin wrote: >>>> Hi Seungwon, >>>> >>>> ----- Original Message ----- >>>>> From: "Seungwon Jeon" <tgih.jun@xxxxxxxxxxx> >>>>> To: linux-mmc@xxxxxxxxxxxxxxx >>>>> Cc: "Chris Ball" <cjb@xxxxxxxxxx>, linux-samsung-soc@xxxxxxxxxxxxxxx, >>>> "kgene kim" <kgene.kim@xxxxxxxxxxx>, "dh han" >>>>> <dh.han@xxxxxxxxxxx>, "Seungwon Jeon" <tgih.jun@xxxxxxxxxxx> >>>>> Sent: Monday, September 26, 2011 7:46:59 AM >>>>> Subject: [PATCH] mmc: dw_mmc: Support predefined multiple block >>>> transfers. >>>>> >>>>> This patch adds the support for predefined multiple block read/write. >>>>> >>>>> Signed-off-by: Seungwon Jeon <tgih.jun@xxxxxxxxxxx> >>>> >>>> Without knowing much about dw_mmc host, your logic otherwise looks ok, >>>> given what >>>> I've previously done for SDHCI as far as CMD23/Auto-CMD23 enhancement. >>>> Just curious, what eMMC cards did you test this on, and what >> improvement >>>> did you see? >>> >>> Thank you for review. >>> As you done, predefined transfer is required for reliable writes and >> eMMC4.5 feature. >>> Sadly, I didn't gain an improvement in my case. >>> (I don't know whether I can clarify the tested eMMC card, just one >> sample.) >>> >> >> So far I've seen some Sandisk cards have a noticeable real-life >> improvement (30%) over >> open-ended transfers. >> >> You might wish to try out https://github.com/andreiw/superalign to be >> certain. > > I had already tested this patch with IOZONE, but I found no difference. > Maybe a result depends on eMMC device. > As your recommend, I applied above benchmark tool. > Predefined transfer seems like a little better, but there is no difference. > > Thanks. > Seungwon Jeon. > >> >> A >> -- >> 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 > -- 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
