[PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Francesco Dolcini]

From: Francesco Dolcini <francesco.dolcini@toradex.com>

MMC clock pins must be configured as OUTPUT-only per TRM [0] with
receiver disabled (RXACTIVE bit = 0). Change MMC CLK pins from
PIN_INPUT to PIN_OUTPUT to match documentation.

Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]
Fixes: 316b80246b16 ("arm64: dts: ti: add verdin am62")
Cc: stable@vger.kernel.org
Signed-off-by: Francesco Dolcini <francesco.dolcini@toradex.com>
---
 arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
index e97b2b047d10..895b1b7dcac7 100644
--- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
+++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
@@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
 	pinctrl_sdhci0: main-mmc0-default-pins {
 		pinctrl-single,pins = <
 			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
-			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
+			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */
 			AM62X_IOPAD(0x214, PIN_INPUT,        0) /* (AA2) MMC0_DAT0 */
 			AM62X_IOPAD(0x210, PIN_INPUT_PULLUP, 0) /* (AA1) MMC0_DAT1 */
 			AM62X_IOPAD(0x20c, PIN_INPUT_PULLUP, 0) /* (AA3) MMC0_DAT2 */
@@ -589,7 +589,7 @@ AM62X_IOPAD(0x1f8, PIN_INPUT_PULLUP, 0) /* (AC2) MMC0_DAT7 */
 	pinctrl_sdhci1: main-mmc1-default-pins {
 		pinctrl-single,pins = <
 			AM62X_IOPAD(0x23c, PIN_INPUT,        0) /* (A21) MMC1_CMD  */ /* SODIMM 74 */
-			AM62X_IOPAD(0x234, PIN_INPUT,        0) /* (B22) MMC1_CLK  */ /* SODIMM 78 */
+			AM62X_IOPAD(0x234, PIN_OUTPUT,       0) /* (B22) MMC1_CLK  */ /* SODIMM 78 */
 			AM62X_IOPAD(0x230, PIN_INPUT,        0) /* (A22) MMC1_DAT0 */ /* SODIMM 80 */
 			AM62X_IOPAD(0x22c, PIN_INPUT,        0) /* (B21) MMC1_DAT1 */ /* SODIMM 82 */
 			AM62X_IOPAD(0x228, PIN_INPUT,        0) /* (C21) MMC1_DAT2 */ /* SODIMM 70 */
@@ -601,7 +601,7 @@ AM62X_IOPAD(0x224, PIN_INPUT,        0) /* (D22) MMC1_DAT3 */ /* SODIMM 72 */
 	pinctrl_sdhci2: main-mmc2-default-pins {
 		pinctrl-single,pins = <
 			AM62X_IOPAD(0x120, PIN_INPUT, 0) /* (C24) MMC2_CMD   */ /* WiFi_SDIO_CMD   */
-			AM62X_IOPAD(0x118, PIN_INPUT, 0) /* (D25) MMC2_CLK   */ /* WiFi_SDIO_CLK   */
+			AM62X_IOPAD(0x118, PIN_OUTPUT,0) /* (D25) MMC2_CLK   */ /* WiFi_SDIO_CLK   */
 			AM62X_IOPAD(0x114, PIN_INPUT, 0) /* (B24) MMC2_DAT0  */ /* WiFi_SDIO_DATA0 */
 			AM62X_IOPAD(0x110, PIN_INPUT, 0) /* (C25) MMC2_DAT1  */ /* WiFi_SDIO_DATA1 */
 			AM62X_IOPAD(0x10c, PIN_INPUT, 0) /* (E23) MMC2_DAT2  */ /* WiFi_SDIO_DATA2 */
-- 
2.47.3

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver (RXACTIVE=0), which breaks the internal clock loopback required by the MMC controller for read operations.
--

--- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT ---
commit b7e315f9edd8f485234748a2f8e267539b5b0f30
Author: Francesco Dolcini <francesco.dolcini@toradex.com>

arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

This commit changes the MMC clock pins for the K3 AM62x Verdin module
from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, effectively
disabling the receiver (RXACTIVE = 0).

Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]

> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> index e97b2b047d104..895b1b7dcac71 100644
> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
>  	pinctrl_sdhci0: main-mmc0-default-pins {
>  		pinctrl-single,pins = <
>  			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
> -			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
> +			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */

[Severity: High]
Does configuring this pin as PIN_OUTPUT disable the receiver (RXACTIVE=0) and
break the internal clock loopback?

TI MMC controllers generally rely on this internal clock loopback path to
synchronize incoming data for read operations. While the TRM documents the
logical direction of the signal to the external card, disabling the input
receiver might lead to eMMC read failures and timeouts.

>  			AM62X_IOPAD(0x214, PIN_INPUT,        0) /* (AA2) MMC0_DAT0 */
>  			AM62X_IOPAD(0x210, PIN_INPUT_PULLUP, 0) /* (AA1) MMC0_DAT1 */

[ ... ]

> @@ -589,7 +589,7 @@ AM62X_IOPAD(0x1f8, PIN_INPUT_PULLUP, 0) /* (AC2) MMC0_DAT7 */
>  	pinctrl_sdhci1: main-mmc1-default-pins {
>  		pinctrl-single,pins = <
>  			AM62X_IOPAD(0x23c, PIN_INPUT,        0) /* (A21) MMC1_CMD  */ /* SODIMM 74 */
> -			AM62X_IOPAD(0x234, PIN_INPUT,        0) /* (B22) MMC1_CLK  */ /* SODIMM 78 */
> +			AM62X_IOPAD(0x234, PIN_OUTPUT,       0) /* (B22) MMC1_CLK  */ /* SODIMM 78 */

[Severity: High]
Could this change similarly break the SD card read loopback?

>  			AM62X_IOPAD(0x230, PIN_INPUT,        0) /* (A22) MMC1_DAT0 */ /* SODIMM 80 */
>  			AM62X_IOPAD(0x22c, PIN_INPUT,        0) /* (B21) MMC1_DAT1 */ /* SODIMM 82 */

[ ... ]

> @@ -601,7 +601,7 @@ AM62X_IOPAD(0x224, PIN_INPUT,        0) /* (D22) MMC1_DAT3 */ /* SODIMM 72 */
>  	pinctrl_sdhci2: main-mmc2-default-pins {
>  		pinctrl-single,pins = <
>  			AM62X_IOPAD(0x120, PIN_INPUT, 0) /* (C24) MMC2_CMD   */ /* WiFi_SDIO_CMD   */
> -			AM62X_IOPAD(0x118, PIN_INPUT, 0) /* (D25) MMC2_CLK   */ /* WiFi_SDIO_CLK   */
> +			AM62X_IOPAD(0x118, PIN_OUTPUT,0) /* (D25) MMC2_CLK   */ /* WiFi_SDIO_CLK   */

[Severity: High]
Will setting MMC2_CLK to PIN_OUTPUT clear the RXACTIVE bit and prevent the
WiFi SDIO from properly reading data?

>  			AM62X_IOPAD(0x114, PIN_INPUT, 0) /* (B24) MMC2_DAT0  */ /* WiFi_SDIO_DATA0 */
>  			AM62X_IOPAD(0x110, PIN_INPUT, 0) /* (C25) MMC2_DAT1  */ /* WiFi_SDIO_DATA1 */

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260813111238.56892-1-francesco@dolcini.it?part=1

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Francesco Dolcini]

+Judith

On Thu, Aug 13, 2026 at 11:24:58AM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> - [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver (RXACTIVE=0), which breaks the internal clock loopback required by the MMC controller for read operations.
> --
> 
> --- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT ---
> commit b7e315f9edd8f485234748a2f8e267539b5b0f30
> Author: Francesco Dolcini <francesco.dolcini@toradex.com>
> 
> arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT
> 
> This commit changes the MMC clock pins for the K3 AM62x Verdin module
> from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, effectively
> disabling the receiver (RXACTIVE = 0).
> 
> Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]
> 
> > diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > index e97b2b047d104..895b1b7dcac71 100644
> > --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
> >  	pinctrl_sdhci0: main-mmc0-default-pins {
> >  		pinctrl-single,pins = <
> >  			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
> > -			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
> > +			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */
> 
> [Severity: High]
> Does configuring this pin as PIN_OUTPUT disable the receiver (RXACTIVE=0) and
> break the internal clock loopback?
> 
> TI MMC controllers generally rely on this internal clock loopback path to
> synchronize incoming data for read operations. While the TRM documents the
> logical direction of the signal to the external card, disabling the input
> receiver might lead to eMMC read failures and timeouts.

Judith: is sashiko right on this?

Thanks,
Francesco

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Mendez, Judith]

Hi Francesco,

On 8/13/2026 8:22 AM, Francesco Dolcini wrote:
> +Judith
> 
> On Thu, Aug 13, 2026 at 11:24:58AM +0000, sashiko-bot@kernel.org wrote:
>> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
>> - [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver (RXACTIVE=0), which breaks the internal clock loopback required by the MMC controller for read operations.
>> --
>>
>> --- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT ---
>> commit b7e315f9edd8f485234748a2f8e267539b5b0f30
>> Author: Francesco Dolcini <francesco.dolcini@toradex.com>
>>
>> arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT
>>
>> This commit changes the MMC clock pins for the K3 AM62x Verdin module
>> from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, effectively
>> disabling the receiver (RXACTIVE = 0).
>>
>> Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]
>>
>>> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> index e97b2b047d104..895b1b7dcac71 100644
>>> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>> @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
>>>   	pinctrl_sdhci0: main-mmc0-default-pins {
>>>   		pinctrl-single,pins = <
>>>   			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
>>> -			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
>>> +			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */
>>
>> [Severity: High]
>> Does configuring this pin as PIN_OUTPUT disable the receiver (RXACTIVE=0) and
>> break the internal clock loopback?
>>
>> TI MMC controllers generally rely on this internal clock loopback path to
>> synchronize incoming data for read operations. While the TRM documents the
>> logical direction of the signal to the external card, disabling the input
>> receiver might lead to eMMC read failures and timeouts.
> 
> Judith: is sashiko right on this?
Changing to PIN_OUTPUT should not break anything for you on AM62x based
board. By default the input buffer for CLK pin is not used, CLKLB input
buffer is used instead. CLKLB input is already configured for you by
default, you should not be modifying CLKLB for MMC0 nor MMC1.

This logic does not apply for MMC2.

In any case, you should test once. If by any chance something does
break, you modified the default in emmcsd_io_clklb_sel CTRL MMR
register.

~ Judith

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Francesco Dolcini]

On Thu, Aug 13, 2026 at 10:45:35AM -0500, Mendez, Judith wrote:
> Hi Francesco,
> 
> On 8/13/2026 8:22 AM, Francesco Dolcini wrote:
> > +Judith
> > 
> > On Thu, Aug 13, 2026 at 11:24:58AM +0000, sashiko-bot@kernel.org wrote:
> > > Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
> > > - [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver (RXACTIVE=0), which breaks the internal clock loopback required by the MMC controller for read operations.
> > > --
> > > 
> > > --- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT ---
> > > commit b7e315f9edd8f485234748a2f8e267539b5b0f30
> > > Author: Francesco Dolcini <francesco.dolcini@toradex.com>
> > > 
> > > arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT
> > > 
> > > This commit changes the MMC clock pins for the K3 AM62x Verdin module
> > > from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, effectively
> > > disabling the receiver (RXACTIVE = 0).
> > > 
> > > Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]
> > > 
> > > > diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > > index e97b2b047d104..895b1b7dcac71 100644
> > > > --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > > +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > > @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
> > > >   	pinctrl_sdhci0: main-mmc0-default-pins {
> > > >   		pinctrl-single,pins = <
> > > >   			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
> > > > -			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
> > > > +			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */
> > > 
> > > [Severity: High]
> > > Does configuring this pin as PIN_OUTPUT disable the receiver (RXACTIVE=0) and
> > > break the internal clock loopback?
> > > 
> > > TI MMC controllers generally rely on this internal clock loopback path to
> > > synchronize incoming data for read operations. While the TRM documents the
> > > logical direction of the signal to the external card, disabling the input
> > > receiver might lead to eMMC read failures and timeouts.
> > 
> > Judith: is sashiko right on this?
> Changing to PIN_OUTPUT should not break anything for you on AM62x based
> board. By default the input buffer for CLK pin is not used, CLKLB input
> buffer is used instead. CLKLB input is already configured for you by
> default, you should not be modifying CLKLB for MMC0 nor MMC1.
> 
> This logic does not apply for MMC2.

What do you mean? that MMC2_CLK should stay as input? 

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Mendez, Judith]

On 8/13/2026 10:47 AM, Francesco Dolcini wrote:
> On Thu, Aug 13, 2026 at 10:45:35AM -0500, Mendez, Judith wrote:
>> Hi Francesco,
>>
>> On 8/13/2026 8:22 AM, Francesco Dolcini wrote:
>>> +Judith
>>>
>>> On Thu, Aug 13, 2026 at 11:24:58AM +0000, sashiko-bot@kernel.org wrote:
>>>> Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
>>>> - [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver (RXACTIVE=0), which breaks the internal clock loopback required by the MMC controller for read operations.
>>>> --
>>>>
>>>> --- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT ---
>>>> commit b7e315f9edd8f485234748a2f8e267539b5b0f30
>>>> Author: Francesco Dolcini <francesco.dolcini@toradex.com>
>>>>
>>>> arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT
>>>>
>>>> This commit changes the MMC clock pins for the K3 AM62x Verdin module
>>>> from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, effectively
>>>> disabling the receiver (RXACTIVE = 0).
>>>>
>>>> Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. MMCSDi I/O Signals [0]
>>>>
>>>>> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>> index e97b2b047d104..895b1b7dcac71 100644
>>>>> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>> @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
>>>>>    	pinctrl_sdhci0: main-mmc0-default-pins {
>>>>>    		pinctrl-single,pins = <
>>>>>    			AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) MMC0_CMD  */
>>>>> -			AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) MMC0_CLK  */
>>>>> +			AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) MMC0_CLK  */
>>>>
>>>> [Severity: High]
>>>> Does configuring this pin as PIN_OUTPUT disable the receiver (RXACTIVE=0) and
>>>> break the internal clock loopback?
>>>>
>>>> TI MMC controllers generally rely on this internal clock loopback path to
>>>> synchronize incoming data for read operations. While the TRM documents the
>>>> logical direction of the signal to the external card, disabling the input
>>>> receiver might lead to eMMC read failures and timeouts.
>>>
>>> Judith: is sashiko right on this?
>> Changing to PIN_OUTPUT should not break anything for you on AM62x based
>> board. By default the input buffer for CLK pin is not used, CLKLB input
>> buffer is used instead. CLKLB input is already configured for you by
>> default, you should not be modifying CLKLB for MMC0 nor MMC1.
>>
>> This logic does not apply for MMC2.
> 
> What do you mean? that MMC2_CLK should stay as input?

Well, there is a reason MMC2 CLKLB pin is defined in pinmux nodes on TI
boards (: By default, MMC2 CLKLB is disabled after reset unlike MMC0 and
MMC1. If we do not enable it in software -> PIN_INPUT, then the MMCSD 
loopback clock input would not work and your MMC2 would not work.

~ Judith

Re: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT

[Mendez, Judith]

Francesco,

On 8/13/2026 10:58 AM, Mendez, Judith wrote:
> 
> 
> On 8/13/2026 10:47 AM, Francesco Dolcini wrote:
>> On Thu, Aug 13, 2026 at 10:45:35AM -0500, Mendez, Judith wrote:
>>> Hi Francesco,
>>>
>>> On 8/13/2026 8:22 AM, Francesco Dolcini wrote:
>>>> +Judith
>>>>
>>>> On Thu, Aug 13, 2026 at 11:24:58AM +0000, sashiko-bot@kernel.org wrote:
>>>>> Thank you for your contribution! Sashiko AI review found 1 
>>>>> potential issue(s) to consider:
>>>>> - [High] Setting MMC clock pins to PIN_OUTPUT disables the receiver 
>>>>> (RXACTIVE=0), which breaks the internal clock loopback required by 
>>>>> the MMC controller for read operations.
>>>>> -- 
>>>>>
>>>>> --- Patch [1]: [PATCH v1] arm64: dts: ti: k3-am62-verdin: Set MMC 
>>>>> clock pins as OUTPUT ---
>>>>> commit b7e315f9edd8f485234748a2f8e267539b5b0f30
>>>>> Author: Francesco Dolcini <francesco.dolcini@toradex.com>
>>>>>
>>>>> arm64: dts: ti: k3-am62-verdin: Set MMC clock pins as OUTPUT
>>>>>
>>>>> This commit changes the MMC clock pins for the K3 AM62x Verdin module
>>>>> from PIN_INPUT to PIN_OUTPUT to match the TRM documentation, 
>>>>> effectively
>>>>> disabling the receiver (RXACTIVE = 0).
>>>>>
>>>>> Link: https://www.ti.com/lit/pdf/spruiv7 Rev. C: Table 12-246. 
>>>>> MMCSDi I/O Signals [0]
>>>>>
>>>>>> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/ 
>>>>>> arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>>> index e97b2b047d104..895b1b7dcac71 100644
>>>>>> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>>> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
>>>>>> @@ -573,7 +573,7 @@ AM62X_IOPAD(0x15c, PIN_INPUT, 0)  /* (AB22) 
>>>>>> MDIO0_MDIO */ /* ETH_1_MDIO, SODIMM
>>>>>>        pinctrl_sdhci0: main-mmc0-default-pins {
>>>>>>            pinctrl-single,pins = <
>>>>>>                AM62X_IOPAD(0x220, PIN_INPUT,        0) /*  (Y3) 
>>>>>> MMC0_CMD  */
>>>>>> -            AM62X_IOPAD(0x218, PIN_INPUT,        0) /* (AB1) 
>>>>>> MMC0_CLK  */
>>>>>> +            AM62X_IOPAD(0x218, PIN_OUTPUT,       0) /* (AB1) 
>>>>>> MMC0_CLK  */
>>>>>
>>>>> [Severity: High]
>>>>> Does configuring this pin as PIN_OUTPUT disable the receiver 
>>>>> (RXACTIVE=0) and
>>>>> break the internal clock loopback?
>>>>>
>>>>> TI MMC controllers generally rely on this internal clock loopback 
>>>>> path to
>>>>> synchronize incoming data for read operations. While the TRM 
>>>>> documents the
>>>>> logical direction of the signal to the external card, disabling the 
>>>>> input
>>>>> receiver might lead to eMMC read failures and timeouts.
>>>>
>>>> Judith: is sashiko right on this?
>>> Changing to PIN_OUTPUT should not break anything for you on AM62x based
>>> board. By default the input buffer for CLK pin is not used, CLKLB input
>>> buffer is used instead. CLKLB input is already configured for you by
>>> default, you should not be modifying CLKLB for MMC0 nor MMC1.
>>>
>>> This logic does not apply for MMC2.
>>
>> What do you mean? that MMC2_CLK should stay as input?
> 
> Well, there is a reason MMC2 CLKLB pin is defined in pinmux nodes on TI
> boards (: By default, MMC2 CLKLB is disabled after reset unlike MMC0 and

Meant to say MMC2 CLKLB input buffer is disabled after reset by default,
as in, RX_ACTIVE=0.

> MMC1. If we do not enable it in software -> PIN_INPUT, then the MMCSD 
> loopback clock input would not work and your MMC2 would not work.
> 
> ~ Judith
>