SDHCI support

[Simon Arlott]

I've finished SDHCI PIO and DMA support but it could do with lots of
testing to see if some of the quirks from the other driver show up when
using other cards.

DMA: https://github.com/lp0/linux/commit/ffd17318c29c1bdfcbd9b92e379823b87dcf5deb
SDHCI PIO: https://github.com/lp0/linux/commit/bdd858088d01a82a31f4620311710bd8917ddc73
SDHCI DMA https://github.com/lp0/linux/commit/af5adccd8ebef6d5dc3a6a99da47a3fd93a530ba
HEAD: https://github.com/lp0/linux/commit/015e1324d5459e05f86d39d243606644573f03e4

All of the mmc_test results pass except 15 and 17 which are short writes
that return an unexpected CRC error.

Test 18 recovers ok in PIO mode but I have to reset the DMA channel to
get data transfers to work properly again otherwise tests 23+ fail.

You'll need to have the bcm2708 regulator enabled or it will fail to
register the sdhci host.

You'll need to set the /axi/dma/broadcom,channels parameter to 0x3c (or
0x783c but the extra 4 channels can't support slave DMA).

If you get a repeatable bug, turn on CONFIG_MMC_DEBUG, DMADEVICES_DEBUG
and DMADEVICES_VDEBUG.

-- 
Simon Arlott

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: dma_test: BCM2708 SDHCI PIO.txt
URL: <http://lists.infradead.org/pipermail/linux-rpi-kernel/attachments/20120615/baef88c9/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: dma_test: BCM2708 SDHCI DMA.txt
URL: <http://lists.infradead.org/pipermail/linux-rpi-kernel/attachments/20120615/baef88c9/attachment-0001.txt>