[PATCH v3 2/3] mmc: dw_mmc-rockchip: Add v2 tuning support

[Shawn Lin]

Hi Heiko,

在 2024/8/15 21:17, Heiko Stübner 写道:
> Am Donnerstag, 15. August 2024, 02:55:37 CEST schrieb Shawn Lin:
>> Hi Detlev
>>
>> 在 2024/8/15 6:34, Detlev Casanova 写道:
>>> From: Shawn Lin <shawn.lin at rock-chips.com>
>>>
>>> v2 tuning will inherit pre-stage loader's phase settings for the first
>>> time, and do re-tune if necessary.
>>> Re-tune will still try the rough degrees, for instance, 90, 180, 270,
>>> 360 but continue to do the fine tuning if sample window isn't good
>>> enough.
>>>
>>> Signed-off-by: Shawn Lin <shawn.lin at rock-chips.com>
>>> Signed-off-by: Detlev Casanova <detlev.casanova at collabora.com>
> 
>>> @@ -277,6 +322,10 @@ static int dw_mci_rk3288_parse_dt(struct dw_mci *host)
>>>    					&priv->default_sample_phase))
>>>    		priv->default_sample_phase = 0;
>>>    
>>> +	priv->use_v2_tuning =
>>> +		of_device_is_compatible(host->dev->of_node,
>>> +					"rockchip,rk3576-dw-mshc");
>>> +
>>
>> v2 is a kind of software decision instead of hardware dependency.
>> So in theory, any SoC can claim to use it via DT.
> 
> which actually makes it unsuitable for dt. >

Understood.

> Devicetree describes hardware-properties and should _not_ be used for
> software configuration.
> 
>  From the comment above, I assume the rk3576 does not need that feature
> and can just work with the regular tuning?

Yep, your are right.

> 
> So there are two routes for the immediate future:
> (1) rk3576 _needs_ that feature, then going with the compatible is fine
> 
> (2) rk3576 does not need absolutely need that feature, then I'd expect
> the basic rk3576 to first come without, as I'd expect a lot more explanation
> on why it is actually needed, and which cases it does improve.
> The commit message does not really explain that much about why this
> is a great/needed feature and which areas it does improve.
> 

Vote for the 2nd. rk3576 just need
[PATCH v3 3/3] mmc: dw_mmc-rockchip: Add internal phase support


> 
> Heiko
> 
>