Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[FUKAUMI Naoki]

Hi Chukun,

Could you check Product Brief P.4?

  https://dl.radxa.com/cm3i/e25/radxa_e25_product_brief.pdf

There is "USB3.0 OTG Switch".

Best regards,

--
FUKAUMI Naoki
Radxa Computer (Shenzhen) Co., Ltd.

On 12/15/24 17:38, Chukun Pan wrote:
> The Radxa E25 board has only one USB3.0 Type A connector,
> changing it to host mode is more appropriate.
> 
> Cc: FUKAUMI Naoki <naoki@radxa.com>
> Signed-off-by: Chukun Pan <amadeus@jmu.edu.cn>
> ---
>   arch/arm64/boot/dts/rockchip/rk3568-radxa-e25.dts | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/boot/dts/rockchip/rk3568-radxa-e25.dts b/arch/arm64/boot/dts/rockchip/rk3568-radxa-e25.dts
> index 98cfa3abb809..90a3cb54fd47 100644
> --- a/arch/arm64/boot/dts/rockchip/rk3568-radxa-e25.dts
> +++ b/arch/arm64/boot/dts/rockchip/rk3568-radxa-e25.dts
> @@ -218,6 +218,7 @@ &usb_host0_ohci {
>   };
>   
>   &usb_host0_xhci {
> +	dr_mode = "host";
>   	status = "okay";
>   };
>   

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[Chukun Pan]

Hi,
> Could you check Product Brief P.4?
>
>  https://dl.radxa.com/cm3i/e25/radxa_e25_product_brief.pdf
>
> There is "USB3.0 OTG Switch".
>
> Best regards,

Oh, sorry, it was covered by the shell so I didn't notice it.
BTW do you see any issues in the sd-uhs-sdr104 mode?

Thanks,
Chukun

-- 
2.25.1

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[FUKAUMI Naoki]

Hi Chukun,

On 12/22/24 13:30, Chukun Pan wrote:
> Hi,
>> Could you check Product Brief P.4?
>>
>>   https://dl.radxa.com/cm3i/e25/radxa_e25_product_brief.pdf
>>
>> There is "USB3.0 OTG Switch".
>>
>> Best regards,
> 
> Oh, sorry, it was covered by the shell so I didn't notice it.
> BTW do you see any issues in the sd-uhs-sdr104 mode?

I've been using the uSD card for a long time as a bootloader only, so I 
don't know what the problem is.

Do you have any information on how to reproduce the problem?

Best regards,

--
FUKAUMI Naoki
Radxa Computer (Shenzhen) Co., Ltd.

> Thanks,
> Chukun
> 

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[Chukun Pan]

Hi,
> I've been using the uSD card for a long time as a bootloader only,
> so I don't know what the problem is.
>
> Do you have any information on how to reproduce the problem?
>
> Best regards,

Not stable on openwrt/lede, sometimes falling to sd-highspeed.
Maybe it would be better to switch to sd-uhs-sdr50 mode.

Thanks,
Chukun

-- 
2.25.1

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[Jonas Karlman]

On 2024-12-23 04:00, Chukun Pan wrote:
> Hi,
>> I've been using the uSD card for a long time as a bootloader only,
>> so I don't know what the problem is.
>>
>> Do you have any information on how to reproduce the problem?
>>
>> Best regards,
> 
> Not stable on openwrt/lede, sometimes falling to sd-highspeed.
> Maybe it would be better to switch to sd-uhs-sdr50 mode.

I would advise against using any sd-uhs- mode for sd-card on Rockchip
RK356x boards because of a timing issue during boot between mmc,
regulator and IO-domain drivers.

Any board that support changing between 1v8 or 3v3 io-voltage for the
sd-card is possible affected.

There is a race condition during boot and when IO-domain driver has not
been probed before UHS tuning is started the tuning is done using a 3v3
IO-domain configuration instead of 1v8.

This can typically be seen by comparing UHS tuning during boot and when
same sd-card is removed and re-inserted after boot.

Ideally this race condition should be solved instead of limiting use
of UHS modes. However, use of sd-uhs-sdr50 is typically more forgiving
and has been used on other boards.

Regards,
Jonas

> 
> Thanks,
> Chukun
> 

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[Chukun Pan]

Hi,
> I would advise against using any sd-uhs- mode for sd-card on Rockchip
> RK356x boards because of a timing issue during boot between mmc,
> regulator and IO-domain drivers.

This problem seems to occur only on rk356x?

> Ideally this race condition should be solved instead of limiting use
> of UHS modes. However, use of sd-uhs-sdr50 is typically more forgiving
> and has been used on other boards.

It seems that sd-uhs-sdr50 does not need 1.8V IO voltage.
I tested the speed on a rk3528 board with sdio wifi:
(The IO voltage of sdio on this board is 3.3V)

highspeed: 170Mbps DL
sdr50:     290Mbps DL
sdr104:    290Mbps DL

So sdr50 should be safe.

Thanks,
Chukun

-- 
2.25.1

Re: [PATCH 1/1] arm64: dts: rockchip: default to host mode for USB on Radxa E25

[Chukun Pan]

Hi,
>> @@ -31,5 +31,14 @@ ipq6018_s2: s2 {
>>  			regulator-max-microvolt = <1062500>;
>>  			regulator-always-on;
>>  		};
>> +
>> +		ipq6018_l2: l2 {
>
> s/ipq6018/mp5496

Does ipq6018_s2 also need to be changed?

-- 
2.25.1