diff mbox series

[meta-arago,master,1/2] conf: machine: am62a: Fix overlays in KERNEL_DEVICETREE_PREFIX

Message ID 20260316221815.2502448-2-jm@ti.com
State Accepted
Delegated to: Ryan Eatmon
Headers show
Series Some overlay updates | expand

Commit Message

Judith Mendez March 16, 2026, 10:18 p.m. UTC
There have been changes in k3-am62x* overlays due to conflicts
in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.

Signed-off-by: Judith Mendez <jm@ti.com>
---
 meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

Comments

Andrew Davis March 16, 2026, 10:19 p.m. UTC | #1
On 3/16/26 5:18 PM, Judith Mendez wrote:
> There have been changes in k3-am62x* overlays due to conflicts
> in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.
> 
> Signed-off-by: Judith Mendez <jm@ti.com>
> ---
>   meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
>   1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf b/meta-ti-bsp/conf/machine/am62axx-evm.conf
> index 32ae2490..01955470 100644
> --- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
> +++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
> @@ -6,7 +6,13 @@ require conf/machine/include/am62axx.inc
>   
>   KERNEL_DEVICETREE_PREFIX = " \
>       ti/k3-am62a7 \
> -    ti/k3-am62x-sk \
> +    ti/k3-am62x-sk-csi2 \

"ti/k3-am62x-sk" is already a prefix of "ti/k3-am62x-sk-csi2", does this change do anything?

Andrew

> +    ti/k3-am62x-sk-ecap-capture \
> +    ti/k3-am62x-sk-eqep \
> +    ti/k3-am62x-sk-hdmi \
> +    ti/k3-am62x-sk-lpm \
> +    ti/k3-am62x-sk-mcspi-loopback \
> +    ti/k3-am62x-sk-pwm \
>       ti/k3-fpdlink \
>       ti/k3-v3link \
>   "
Ryan Eatmon March 17, 2026, 1:21 p.m. UTC | #2
On 3/16/2026 5:19 PM, Andrew Davis wrote:
> On 3/16/26 5:18 PM, Judith Mendez wrote:
>> There have been changes in k3-am62x* overlays due to conflicts
>> in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.
>>
>> Signed-off-by: Judith Mendez <jm@ti.com>
>> ---
>>   meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
>>   1 file changed, 7 insertions(+), 1 deletion(-)
>>
>> diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf 
>> b/meta-ti-bsp/conf/machine/am62axx-evm.conf
>> index 32ae2490..01955470 100644
>> --- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
>> +++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
>> @@ -6,7 +6,13 @@ require conf/machine/include/am62axx.inc
>>   KERNEL_DEVICETREE_PREFIX = " \
>>       ti/k3-am62a7 \
>> -    ti/k3-am62x-sk \
>> +    ti/k3-am62x-sk-csi2 \
> 
> "ti/k3-am62x-sk" is already a prefix of "ti/k3-am62x-sk-csi2", does this 
> change do anything?

Andrew,

If you are right, then this locks the machine down to not pick up future 
additions under the ti/k3-am62x-sk pattern in the future.  That's what 
it does...  which would not be "Good (TM)" if you are correct.


> Andrew
> 
>> +    ti/k3-am62x-sk-ecap-capture \
>> +    ti/k3-am62x-sk-eqep \
>> +    ti/k3-am62x-sk-hdmi \
>> +    ti/k3-am62x-sk-lpm \
>> +    ti/k3-am62x-sk-mcspi-loopback \
>> +    ti/k3-am62x-sk-pwm \
>>       ti/k3-fpdlink \
>>       ti/k3-v3link \
>>   "
>
Judith Mendez March 17, 2026, 2:10 p.m. UTC | #3
Hi Andrew, Ryan,

On 3/17/26 8:21 AM, Ryan Eatmon wrote:
> 
> 
> On 3/16/2026 5:19 PM, Andrew Davis wrote:
>> On 3/16/26 5:18 PM, Judith Mendez wrote:
>>> There have been changes in k3-am62x* overlays due to conflicts
>>> in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.
>>>
>>> Signed-off-by: Judith Mendez <jm@ti.com>
>>> ---
>>>   meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
>>>   1 file changed, 7 insertions(+), 1 deletion(-)
>>>
>>> diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf b/meta-ti-bsp/ 
>>> conf/machine/am62axx-evm.conf
>>> index 32ae2490..01955470 100644
>>> --- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>> +++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>> @@ -6,7 +6,13 @@ require conf/machine/include/am62axx.inc
>>>   KERNEL_DEVICETREE_PREFIX = " \
>>>       ti/k3-am62a7 \
>>> -    ti/k3-am62x-sk \
>>> +    ti/k3-am62x-sk-csi2 \
>>
>> "ti/k3-am62x-sk" is already a prefix of "ti/k3-am62x-sk-csi2", does 
>> this change do anything?

I am removing ti/k3-am62x-sk first andrew.
> 
> Andrew,
> 
> If you are right, then this locks the machine down to not pick up future 
> additions under the ti/k3-am62x-sk pattern in the future.  That's what 
> it does...  which would not be "Good (TM)" if you are correct.

Some ti/k3-am62x-* overlays no longer apply to am62a platform. We can
keep ti/k3-am62x-sk but it will cause confusion since at least the
following overlays no longer apply to am62a:

k3-am62x-sk-dmtimer-pwm.dtso
k3-am62x-sk-mcan.dtso

Is there some kind of blacklist for overlays not to be included for a
platform?

~ Judith



>>
>>> +    ti/k3-am62x-sk-ecap-capture \
>>> +    ti/k3-am62x-sk-eqep \
>>> +    ti/k3-am62x-sk-hdmi \
>>> +    ti/k3-am62x-sk-lpm \
>>> +    ti/k3-am62x-sk-mcspi-loopback \
>>> +    ti/k3-am62x-sk-pwm \
>>>       ti/k3-fpdlink \
>>>       ti/k3-v3link \
>>>   "
>>
>
Ryan Eatmon March 17, 2026, 3:26 p.m. UTC | #4
On 3/17/2026 9:10 AM, Judith Mendez wrote:
> Hi Andrew, Ryan,
> 
> On 3/17/26 8:21 AM, Ryan Eatmon wrote:
>>
>>
>> On 3/16/2026 5:19 PM, Andrew Davis wrote:
>>> On 3/16/26 5:18 PM, Judith Mendez wrote:
>>>> There have been changes in k3-am62x* overlays due to conflicts
>>>> in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.
>>>>
>>>> Signed-off-by: Judith Mendez <jm@ti.com>
>>>> ---
>>>>   meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
>>>>   1 file changed, 7 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf 
>>>> b/meta-ti-bsp/ conf/machine/am62axx-evm.conf
>>>> index 32ae2490..01955470 100644
>>>> --- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>>> +++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>>> @@ -6,7 +6,13 @@ require conf/machine/include/am62axx.inc
>>>>   KERNEL_DEVICETREE_PREFIX = " \
>>>>       ti/k3-am62a7 \
>>>> -    ti/k3-am62x-sk \
>>>> +    ti/k3-am62x-sk-csi2 \
>>>
>>> "ti/k3-am62x-sk" is already a prefix of "ti/k3-am62x-sk-csi2", does 
>>> this change do anything?
> 
> I am removing ti/k3-am62x-sk first andrew.
>>
>> Andrew,
>>
>> If you are right, then this locks the machine down to not pick up 
>> future additions under the ti/k3-am62x-sk pattern in the future.  
>> That's what it does...  which would not be "Good (TM)" if you are 
>> correct.
> 
> Some ti/k3-am62x-* overlays no longer apply to am62a platform. We can
> keep ti/k3-am62x-sk but it will cause confusion since at least the
> following overlays no longer apply to am62a:
> 
> k3-am62x-sk-dmtimer-pwm.dtso
> k3-am62x-sk-mcan.dtso
> 
> Is there some kind of blacklist for overlays not to be included for a
> platform?

There is not.  So this might be the best way.


> ~ Judith
> 
> 
> 
>>>
>>>> +    ti/k3-am62x-sk-ecap-capture \
>>>> +    ti/k3-am62x-sk-eqep \
>>>> +    ti/k3-am62x-sk-hdmi \
>>>> +    ti/k3-am62x-sk-lpm \
>>>> +    ti/k3-am62x-sk-mcspi-loopback \
>>>> +    ti/k3-am62x-sk-pwm \
>>>>       ti/k3-fpdlink \
>>>>       ti/k3-v3link \
>>>>   "
>>>
>>
>
Andrew Davis March 17, 2026, 3:29 p.m. UTC | #5
On 3/17/26 9:10 AM, Judith Mendez wrote:
> Hi Andrew, Ryan,
> 
> On 3/17/26 8:21 AM, Ryan Eatmon wrote:
>>
>>
>> On 3/16/2026 5:19 PM, Andrew Davis wrote:
>>> On 3/16/26 5:18 PM, Judith Mendez wrote:
>>>> There have been changes in k3-am62x* overlays due to conflicts
>>>> in DT, so update KERNEL_DEVICETREE_PREFIX for AM62A SK appropriately.
>>>>
>>>> Signed-off-by: Judith Mendez <jm@ti.com>
>>>> ---
>>>>   meta-ti-bsp/conf/machine/am62axx-evm.conf | 8 +++++++-
>>>>   1 file changed, 7 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf b/meta-ti-bsp/ conf/machine/am62axx-evm.conf
>>>> index 32ae2490..01955470 100644
>>>> --- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>>> +++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
>>>> @@ -6,7 +6,13 @@ require conf/machine/include/am62axx.inc
>>>>   KERNEL_DEVICETREE_PREFIX = " \
>>>>       ti/k3-am62a7 \
>>>> -    ti/k3-am62x-sk \
>>>> +    ti/k3-am62x-sk-csi2 \
>>>
>>> "ti/k3-am62x-sk" is already a prefix of "ti/k3-am62x-sk-csi2", does this change do anything?
> 
> I am removing ti/k3-am62x-sk first andrew.
>>
>> Andrew,
>>
>> If you are right, then this locks the machine down to not pick up future additions under the ti/k3-am62x-sk pattern in the future.  That's what it does...  which would not be "Good (TM)" if you are correct.
> 
> Some ti/k3-am62x-* overlays no longer apply to am62a platform. We can
> keep ti/k3-am62x-sk but it will cause confusion since at least the
> following overlays no longer apply to am62a:
> 
> k3-am62x-sk-dmtimer-pwm.dtso
> k3-am62x-sk-mcan.dtso
> 
> Is there some kind of blacklist for overlays not to be included for a
> platform?
> 

There is no issue with adding extra DTB and DTBO files. This is always
the case as one machine config often supports multiple similar boards.
It is up to the user/bootloader to pick up and apply the correct DTB
and overlays for the given board.

Andrew

> ~ Judith
> 
> 
> 
>>>
>>>> +    ti/k3-am62x-sk-ecap-capture \
>>>> +    ti/k3-am62x-sk-eqep \
>>>> +    ti/k3-am62x-sk-hdmi \
>>>> +    ti/k3-am62x-sk-lpm \
>>>> +    ti/k3-am62x-sk-mcspi-loopback \
>>>> +    ti/k3-am62x-sk-pwm \
>>>>       ti/k3-fpdlink \
>>>>       ti/k3-v3link \
>>>>   "
>>>
>>
>
diff mbox series

Patch

diff --git a/meta-ti-bsp/conf/machine/am62axx-evm.conf b/meta-ti-bsp/conf/machine/am62axx-evm.conf
index 32ae2490..01955470 100644
--- a/meta-ti-bsp/conf/machine/am62axx-evm.conf
+++ b/meta-ti-bsp/conf/machine/am62axx-evm.conf
@@ -6,7 +6,13 @@  require conf/machine/include/am62axx.inc
 
 KERNEL_DEVICETREE_PREFIX = " \
     ti/k3-am62a7 \
-    ti/k3-am62x-sk \
+    ti/k3-am62x-sk-csi2 \
+    ti/k3-am62x-sk-ecap-capture \
+    ti/k3-am62x-sk-eqep \
+    ti/k3-am62x-sk-hdmi \
+    ti/k3-am62x-sk-lpm \
+    ti/k3-am62x-sk-mcspi-loopback \
+    ti/k3-am62x-sk-pwm \
     ti/k3-fpdlink \
     ti/k3-v3link \
 "