diff mbox series

ARM: dts: stm32: Reinstate card detect behavior on DHSOM

Message ID 20201019213752.888493-1-marex@denx.de
State Superseded
Delegated to: Patrice Chotard
Headers show
Series ARM: dts: stm32: Reinstate card detect behavior on DHSOM | expand

Commit Message

Marek Vasut Oct. 19, 2020, 9:37 p.m. UTC
The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
unsupported, reinstate the old cd-gpios behavior until this handling
is fully implemented. This permits the DHSOM to boot from SD again,
without this patch the card detect fails.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Cc: Patrice Chotard <patrice.chotard@st.com>
---
 arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
 1 file changed, 3 insertions(+)

Comments

Patrick DELAUNAY Oct. 22, 2020, 5:17 p.m. UTC | #1
Hi Marek,

> From: Marek Vasut <marex@denx.de>
> Sent: lundi 19 octobre 2020 23:38
> 
> The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
> unsupported, reinstate the old cd-gpios behavior until this handling is fully
> implemented. This permits the DHSOM to boot from SD again, without this patch
> the card detect fails.

It is strange if it is not working... I will check why the configuration is not managed
correctly in stm32 gpio/pincontrol driver.

> Signed-off-by: Marek Vasut <marex@denx.de>
> Cc: Patrick Delaunay <patrick.delaunay@st.com>
> Cc: Patrice Chotard <patrice.chotard@st.com>
> ---
>  arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> index 92345b7ba3..73bb5f1c6d 100644
> --- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> +++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> @@ -265,6 +265,9 @@
> 
>  &sdmmc1 {
>  	u-boot,dm-spl;
> +	broken-cd;
> +	/delete-property/ cd-gpios;
> +	/delete-property/ disable-wp;
>  };
> 
>  &sdmmc1_b4_pins_a {
> --
> 2.28.0

Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>

But I will be try to fix the issue before to accept this patch.

I will merge it only if I can't found the root cause for v2020.01-rc2....

Thanks

Patrick
Marek Vasut Oct. 22, 2020, 6:18 p.m. UTC | #2
On 10/22/20 7:17 PM, Patrick DELAUNAY wrote:
> Hi Marek,

Hi,

>> From: Marek Vasut <marex@denx.de>
>> Sent: lundi 19 octobre 2020 23:38
>>
>> The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
>> unsupported, reinstate the old cd-gpios behavior until this handling is fully
>> implemented. This permits the DHSOM to boot from SD again, without this patch
>> the card detect fails.
> 
> It is strange if it is not working... I will check why the configuration is not managed
> correctly in stm32 gpio/pincontrol driver.
> 
>> Signed-off-by: Marek Vasut <marex@denx.de>
>> Cc: Patrick Delaunay <patrick.delaunay@st.com>
>> Cc: Patrice Chotard <patrice.chotard@st.com>
>> ---
>>  arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>> b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>> index 92345b7ba3..73bb5f1c6d 100644
>> --- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>> +++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>> @@ -265,6 +265,9 @@
>>
>>  &sdmmc1 {
>>  	u-boot,dm-spl;
>> +	broken-cd;
>> +	/delete-property/ cd-gpios;
>> +	/delete-property/ disable-wp;
>>  };
>>
>>  &sdmmc1_b4_pins_a {
>> --
>> 2.28.0
> 
> Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
> 
> But I will be try to fix the issue before to accept this patch.
> 
> I will merge it only if I can't found the root cause for v2020.01-rc2....

Do you not see the same behavior on EV/DK ?

btw it is bugged in SPL.
Patrick DELAUNAY Oct. 23, 2020, 2:39 p.m. UTC | #3
Hi,

> From: Marek Vasut <marex@denx.de>
> Sent: jeudi 22 octobre 2020 20:19
> To: Patrick DELAUNAY <patrick.delaunay@st.com>; u-boot@lists.denx.de
> Cc: Patrice CHOTARD <patrice.chotard@st.com>
> Subject: Re: [PATCH] ARM: dts: stm32: Reinstate card detect behavior on
> DHSOM
> Importance: High
> 
> On 10/22/20 7:17 PM, Patrick DELAUNAY wrote:
> > Hi Marek,
> 
> Hi,
> 
> >> From: Marek Vasut <marex@denx.de>
> >> Sent: lundi 19 octobre 2020 23:38
> >>
> >> The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
> >> unsupported, reinstate the old cd-gpios behavior until this handling
> >> is fully implemented. This permits the DHSOM to boot from SD again,
> >> without this patch the card detect fails.
> >
> > It is strange if it is not working... I will check why the
> > configuration is not managed correctly in stm32 gpio/pincontrol driver.
> >
> >> Signed-off-by: Marek Vasut <marex@denx.de>
> >> Cc: Patrick Delaunay <patrick.delaunay@st.com>
> >> Cc: Patrice Chotard <patrice.chotard@st.com>
> >> ---
> >>  arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
> >>  1 file changed, 3 insertions(+)
> >>
> >> diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >> b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >> index 92345b7ba3..73bb5f1c6d 100644
> >> --- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >> +++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >> @@ -265,6 +265,9 @@
> >>
> >>  &sdmmc1 {
> >>  	u-boot,dm-spl;
> >> +	broken-cd;
> >> +	/delete-property/ cd-gpios;
> >> +	/delete-property/ disable-wp;
> >>  };
> >>
> >>  &sdmmc1_b4_pins_a {
> >> --
> >> 2.28.0
> >
> > Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
> >
> > But I will be try to fix the issue before to accept this patch.
> >
> > I will merge it only if I can't found the root cause for v2020.01-rc2....
> 
> Do you not see the same behavior on EV/DK ?

No but I think we have pull-up on board.

But after investigation the internall pull-up is not configurated in stm32 pinctrol
(cheked with pinmux command) even it is requested in device-tree of EV1/DK2.

It is clearly a bug (I isolate the issue) and I am working on a patch 
(it  should be sent to u-boot mailing liste next week).

> btw it is bugged in SPL.

Ah, what is the issue.

In the stm32 driver or in the framework ?

I will cross-check it also on EV1/DK2.

Patrick
Marek Vasut Oct. 23, 2020, 3:04 p.m. UTC | #4
On 10/23/20 4:39 PM, Patrick DELAUNAY wrote:

Hi,

[...]

>>>> From: Marek Vasut <marex@denx.de>
>>>> Sent: lundi 19 octobre 2020 23:38
>>>>
>>>> The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
>>>> unsupported, reinstate the old cd-gpios behavior until this handling
>>>> is fully implemented. This permits the DHSOM to boot from SD again,
>>>> without this patch the card detect fails.
>>>
>>> It is strange if it is not working... I will check why the
>>> configuration is not managed correctly in stm32 gpio/pincontrol driver.
>>>
>>>> Signed-off-by: Marek Vasut <marex@denx.de>
>>>> Cc: Patrick Delaunay <patrick.delaunay@st.com>
>>>> Cc: Patrice Chotard <patrice.chotard@st.com>
>>>> ---
>>>>  arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
>>>>  1 file changed, 3 insertions(+)
>>>>
>>>> diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>>>> b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>>>> index 92345b7ba3..73bb5f1c6d 100644
>>>> --- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>>>> +++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
>>>> @@ -265,6 +265,9 @@
>>>>
>>>>  &sdmmc1 {
>>>>  	u-boot,dm-spl;
>>>> +	broken-cd;
>>>> +	/delete-property/ cd-gpios;
>>>> +	/delete-property/ disable-wp;
>>>>  };
>>>>
>>>>  &sdmmc1_b4_pins_a {
>>>> --
>>>> 2.28.0
>>>
>>> Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
>>>
>>> But I will be try to fix the issue before to accept this patch.
>>>
>>> I will merge it only if I can't found the root cause for v2020.01-rc2....
>>
>> Do you not see the same behavior on EV/DK ?
> 
> No but I think we have pull-up on board.
> 
> But after investigation the internall pull-up is not configurated in stm32 pinctrol
> (cheked with pinmux command) even it is requested in device-tree of EV1/DK2.
> 
> It is clearly a bug (I isolate the issue) and I am working on a patch 
> (it  should be sent to u-boot mailing liste next week).
> 
>> btw it is bugged in SPL.
> 
> Ah, what is the issue.
> 
> In the stm32 driver or in the framework ?
> 
> I will cross-check it also on EV1/DK2.

The card is not detected in SPL again, same fail mode as before.
Now that I think about it, note to self, I should check whether the CD
GPIO controller node is u-boot,dm-spl .
Patrick DELAUNAY Oct. 28, 2020, 10:20 a.m. UTC | #5
Hi Marek,

> From: Marek Vasut <marex@denx.de>
> Sent: vendredi 23 octobre 2020 17:04
> 
> On 10/23/20 4:39 PM, Patrick DELAUNAY wrote:
> 
> Hi,
> 
> [...]
> 
> >>>> From: Marek Vasut <marex@denx.de>
> >>>> Sent: lundi 19 octobre 2020 23:38
> >>>>
> >>>> The cd-gpios with (GPIO_ACTIVE_LOW | GPIO_PULL_UP) gpio is thus far
> >>>> unsupported, reinstate the old cd-gpios behavior until this
> >>>> handling is fully implemented. This permits the DHSOM to boot from
> >>>> SD again, without this patch the card detect fails.
> >>>
> >>> It is strange if it is not working... I will check why the
> >>> configuration is not managed correctly in stm32 gpio/pincontrol driver.
> >>>
> >>>> Signed-off-by: Marek Vasut <marex@denx.de>
> >>>> Cc: Patrick Delaunay <patrick.delaunay@st.com>
> >>>> Cc: Patrice Chotard <patrice.chotard@st.com>
> >>>> ---
> >>>>  arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi | 3 +++
> >>>>  1 file changed, 3 insertions(+)
> >>>>
> >>>> diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >>>> b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >>>> index 92345b7ba3..73bb5f1c6d 100644
> >>>> --- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >>>> +++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
> >>>> @@ -265,6 +265,9 @@
> >>>>
> >>>>  &sdmmc1 {
> >>>>  	u-boot,dm-spl;
> >>>> +	broken-cd;
> >>>> +	/delete-property/ cd-gpios;
> >>>> +	/delete-property/ disable-wp;
> >>>>  };
> >>>>
> >>>>  &sdmmc1_b4_pins_a {
> >>>> --
> >>>> 2.28.0
> >>>
> >>> Reviewed-by: Patrick Delaunay <patrick.delaunay@st.com>
> >>>
> >>> But I will be try to fix the issue before to accept this patch.
> >>>
> >>> I will merge it only if I can't found the root cause for v2020.01-rc2....
> >>
> >> Do you not see the same behavior on EV/DK ?
> >
> > No but I think we have pull-up on board.
> >
> > But after investigation the internall pull-up is not configurated in
> > stm32 pinctrol (cheked with pinmux command) even it is requested in device-
> tree of EV1/DK2.
> >
> > It is clearly a bug (I isolate the issue) and I am working on a patch
> > (it  should be sent to u-boot mailing liste next week).
> >
> >> btw it is bugged in SPL.
> >
> > Ah, what is the issue.
> >
> > In the stm32 driver or in the framework ?
> >
> > I will cross-check it also on EV1/DK2.
> 
> The card is not detected in SPL again, same fail mode as before.
> Now that I think about it, note to self, I should check whether the CD GPIO
> controller node is u-boot,dm-spl

I found a big issue in stm32 pincontrol: the bias configuration was only managed for ouput pin.

With this serie [1], I checked the pull-up configuration (including in SPL by adding a  debug trace).

With [2], the bias configuration is now correct in DK2 / EV1 boards (even if the pull-up configuration is not mandatory,
because I don't see cart detection issue on ST board).

I expect this patch correct the DHSOM issue.

[1]: http://patchwork.ozlabs.org/project/uboot/list/?series=210501

[2]: patch [2/2] gpio: stm32: correct the bias management
      http://patchwork.ozlabs.org/project/uboot/patch/20201028094908.11031-2-patrick.delaunay@st.com/

Regards,

Patrick
Marek Vasut Oct. 28, 2020, 8:49 p.m. UTC | #6
On 10/28/20 11:20 AM, Patrick DELAUNAY wrote:
> Hi Marek,

Hello Patrick,

[...]

>>> But after investigation the internall pull-up is not configurated in
>>> stm32 pinctrol (cheked with pinmux command) even it is requested in device-
>> tree of EV1/DK2.
>>>
>>> It is clearly a bug (I isolate the issue) and I am working on a patch
>>> (it  should be sent to u-boot mailing liste next week).
>>>
>>>> btw it is bugged in SPL.
>>>
>>> Ah, what is the issue.
>>>
>>> In the stm32 driver or in the framework ?
>>>
>>> I will cross-check it also on EV1/DK2.
>>
>> The card is not detected in SPL again, same fail mode as before.
>> Now that I think about it, note to self, I should check whether the CD GPIO
>> controller node is u-boot,dm-spl
> 
> I found a big issue in stm32 pincontrol: the bias configuration was only managed for ouput pin.
> 
> With this serie [1], I checked the pull-up configuration (including in SPL by adding a  debug trace).
> 
> With [2], the bias configuration is now correct in DK2 / EV1 boards (even if the pull-up configuration is not mandatory,
> because I don't see cart detection issue on ST board).
> 
> I expect this patch correct the DHSOM issue.

It does not, but the following does:
[PATCH] ARM: dts: stm32: Fix uSD card-detect GPIO on DHCOM

So it's a bug on both ends, good thing we found them.

Thanks!
diff mbox series

Patch

diff --git a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
index 92345b7ba3..73bb5f1c6d 100644
--- a/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
+++ b/arch/arm/dts/stm32mp15xx-dhcom-u-boot.dtsi
@@ -265,6 +265,9 @@ 
 
 &sdmmc1 {
 	u-boot,dm-spl;
+	broken-cd;
+	/delete-property/ cd-gpios;
+	/delete-property/ disable-wp;
 };
 
 &sdmmc1_b4_pins_a {