On 12/07/2016 06:04 PM, Masahiro Yamada wrote: > Hi Tom > > 2016-12-06 4:10 GMT+09:00 Tom Rini <tr...@konsulko.com>: >> On Sun, Dec 04, 2016 at 10:23:13PM +0900, Masahiro Yamada wrote: >> >>> While I moved the options, I also renamed them so that they are all >>> prefixed with MMC_SDHCI_. >> [snip] >>> diff --git a/drivers/mmc/Kconfig b/drivers/mmc/Kconfig >>> index 8e9fa2d..4785d71 100644 >>> --- a/drivers/mmc/Kconfig >>> +++ b/drivers/mmc/Kconfig >>> @@ -132,6 +132,68 @@ config MMC_SDHCI >>> >>> If unsure, say N. >>> >>> +config MMC_SDHCI_SDMA >>> + bool "Support SDHCI SDMA" >>> + depends on MMC_SDHCI >>> + help >>> + This enables support for the SDMA (Single Operation DMA) defined >>> + in the SD Host Controller Standard Specification Version 1.00 . >> >> I think this should be a default y. > > No. > > After my conversion, > > masahiro@pug:~/workspace/u-boot/configs$ git grep CONFIG_MMC_SDHCI=y | wc > 54 54 2306 > masahiro@pug:~/workspace/u-boot/configs$ git grep CONFIG_MMC_SDHCI_SDMA=y | wc > 21 21 995 > > > So, SDMA is disabled for more than half of boards with SDHCI driver.
Right, MMC_SDHCI_SDMA doesn't need to enable by default. > > > I know one more reason that we should not enable option by default. > > SDMA is so old that it only supports 32 bit physical address. > > U-Boot uses the tail of memory area as work RAM. > > If SDMA is enabled on 64bit boards, MMC does not work > because the controller cannot read EXT-CSD data. > > Given increasing ARMv8 boards, > we want to use ADMA(2), but it is not supported in U-Boot. As i know, u-boot host controller didn't fully support for ARM64. It needs to take a ToDo. Best Regards, Jaehoon Chung > > Anyway, I believe SDMA does not deserve default. > _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
