Re: [PATCH] clk: exynos5250: Fix divider values for sclk_mmc{0,1,2,3}

[Tushar Behera <tushar.behera@linaro.org>]

On 04/22/2013 11:10 PM, Olof Johansson wrote:
> Hi,
> 
> On Tue, Apr 16, 2013 at 12:35 PM, Doug Anderson <dianders@chromium.org> wrote:
>> Hi,
>>
>> On Mon, Apr 8, 2013 at 12:22 AM, Kukjin Kim <kgene.kim@samsung.com> wrote:
>>> Mike Turquette wrote:
>>>>
>>>> Quoting Tushar Behera (2013-04-02 01:20:40)
>>>>> In legacy setup, sclk_mmc{0,1,2,3} used PRE_RATIO bit-field (8-bit wide)
>>>>> instead of RATIO bit-field (4-bit wide) for dividing clock rate.
>>>>>
>>>>> With current common clock setup, we are using RATIO bit-field which
>>>>> is creating FIFO read errors while accessing eMMC. Changing over to
>>>>> use PRE_RATIO bit-field fixes this issue.
>>>>>
>>>>> dwmmc_exynos 12200000.dwmmc0: data FIFO error (status=00008020)
>>>>> mmcblk0: error -5 transferring data, sector 1, nr 7, cmd response 0x900,
>>>> card status 0x0
>>>>> end_request: I/O error, dev mmcblk0, sector 1
>>>>>
>>>>> Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
>>>>> CC: Thomas Abraham <thomas.abraham@linaro.org>
>>>>
>>>> I guess this will be applied through the samsung tree, so:
>>>>
>>>> Acked-by: Mike Turquette <mturquette@linaro.org>
>>>>
>>> Thanks, applied.
>>
>> I haven't yet had time to dig / track down why, but this patch totally
>> messes up access to the eMMC on the ARM Chromebook (exynos5250-snow).
>> I suddenly start getting FIFO errors like you show above.  When I
>> revert your change then I'm all happy.
>>
>> Perhaps I need a device tree setting change as well?  I always forget
>> how the "samsung,dw-mshc-ciu-div" / "samsung,dw-mshc-sdr-timing"
>> properties work...
>>
>> For the short term I'm going to revert locally since I've got a few
>> other things to do over the next few days.  If nobody else gets around
>> to it then I'll try to find time to dig further.
> 
> Unless I hear differently within 24 hours, I am going to revert this
> in arm-soc (since that is where it is merged right now).
> 

I will have a look at it today.

> It is obviously causing regressions on existing platforms. I am _NOT_
> happy to see dead silence about this for 6 days. Tushar??
> 

Apologies.

> 
> 
> -Olof
> 


-- 
Tushar Behera