[U-Boot] [PATCH 0/5] mmc: omap_hsmmc: Add support for ADMA

[Jean-Jacques Hiblot]

On 17/07/2017 23:39, Adam Ford wrote:
> On Mon, Jul 17, 2017 at 3:29 PM, Jean-Jacques Hiblot <jjhiblot at ti.com> wrote:
>>
>> On 14/07/2017 15:18, Adam Ford wrote:
>>> On Thu, Jul 13, 2017 at 2:22 AM, Jean-Jacques Hiblot <jjhiblot at ti.com>
>>> wrote:
>>>>
>>>> On 12/07/2017 19:30, Tom Rini wrote:
>>>>> On Tue, Jul 11, 2017 at 06:20:09PM +0200, Jean-Jacques Hiblot wrote:
>>>>>
>>>>>> This series enables the ADMA present in some OMAP SOCs.
>>>>>> On a DRA7 the performances when reading from the eMMC go from 20MB/s
>>>>>> to 40MB/s.
>>>>>> Also while were at it, fix some incorrect bit operations
>>>>>>
>>>>>> This is the first series of 3 which wille enable HS200 and UHS on the
>>>>>> omap5
>>>>>> platforms (dra7 and am57).
>>>>>> This series applies on u-boot-mmc/hs200-ufs-testing although it could
>>>>>> be
>>>>>> modified to apply on u-boot/master.
>>>>> What hardware have you runtime tested this on?  If you haven't done
>>>>> omap3, I can do some run-time testing there.  Thanks!
>>>> I haven't been able to test on omap3.
>>>> I did BBB, am437x SK, AM57X evm and DRA7 evm.
>>>>
>>> I had to fix some other SPL issues first and once I was able to do
>>> that, I applied this 5-patch series and tested on DM3730 booting with
>>> SPL using Device Tree in U-boot.  (I might still be the only OMAP3
>>> board which enables CONFIG_OF_CONTROL=y).
>>>
>>> The only compile concern I found was:
>>>
>>>    CC      drivers/usb/gadget/f_fastboot.o
>>> drivers/mmc/omap_hsmmc.c: In function ‘omap_hsmmc_send_cmd’:
>>> drivers/mmc/omap_hsmmc.c:486:14: warning: unused variable ‘mmc’
>>> [-Wunused-variable]
>>>     struct mmc *mmc = upriv->mmc;
>>>                 ^~~
>>>     LD      board/logicpd/omap3som/built-in.o
>>>     CC      lib/smbios.o
>>>
>>>
>>> However, my board booted correctly via SD card.
>> Thanks for testing. I'll fix the warning in the next version
>>
> no problem.
> For what it's worth, someone from TI asked me not to use the #ifdef to
> set the base address offsets.  I originally did your way, and it was
> rejected.  You might want to make sure that everyone at TI is on
> board.
I understand how it feels. But it's a bit more than just changing the 
base address. There are actually new registers defined in this area, 
that do not exist on omap3.
JJ
>
>
> adam
>
>> JJ
>>
>>>
>>>
>>> adam
>>>
>>>> JJ
>>>>>