[v4,2/3] arm64: dts: ti: k3-j721e: Add ESM support

Message ID 20230503093310.85779-3-n-francis@ti.com
State New
Headers
Series Add support for ESM |

Commit Message

Neha Malcom Francis May 3, 2023, 9:33 a.m. UTC
  Add address entry mapping ESM on J721E.

Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
---
 arch/arm64/boot/dts/ti/k3-j721e.dtsi | 1 +
 1 file changed, 1 insertion(+)
  

Comments

Nishanth Menon May 3, 2023, 11:48 a.m. UTC | #1
On 15:03-20230503, Neha Malcom Francis wrote:
> Add address entry mapping ESM on J721E.
> 
> Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
> ---
>  arch/arm64/boot/dts/ti/k3-j721e.dtsi | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/boot/dts/ti/k3-j721e.dtsi b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> index b912143b6a11..52bcde601eb8 100644
> --- a/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> +++ b/arch/arm64/boot/dts/ti/k3-j721e.dtsi

Why is'nt esm node introduced here?

> @@ -131,6 +131,7 @@ cbass_main: bus@100000 {
>  		#size-cells = <2>;
>  		ranges = <0x00 0x00100000 0x00 0x00100000 0x00 0x00020000>, /* ctrl mmr */
>  			 <0x00 0x00600000 0x00 0x00600000 0x00 0x00031100>, /* GPIO */
> +			 <0x00 0x00700000 0x00 0x00700000 0x00 0x00001000>, /* ESM */
>  			 <0x00 0x00900000 0x00 0x00900000 0x00 0x00012000>, /* serdes */
>  			 <0x00 0x00a40000 0x00 0x00a40000 0x00 0x00000800>, /* timesync router */
>  			 <0x00 0x06000000 0x00 0x06000000 0x00 0x00400000>, /* USBSS0 */
> -- 
> 2.34.1
>
  
Neha Malcom Francis May 3, 2023, 11:55 a.m. UTC | #2
Hi Nishanth,

On 03/05/23 17:18, Nishanth Menon wrote:
> On 15:03-20230503, Neha Malcom Francis wrote:
>> Add address entry mapping ESM on J721E.
>>
>> Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
>> ---
>>   arch/arm64/boot/dts/ti/k3-j721e.dtsi | 1 +
>>   1 file changed, 1 insertion(+)
>>
>> diff --git a/arch/arm64/boot/dts/ti/k3-j721e.dtsi b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
>> index b912143b6a11..52bcde601eb8 100644
>> --- a/arch/arm64/boot/dts/ti/k3-j721e.dtsi
>> +++ b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> 
> Why is'nt esm node introduced here?
> 

As it stands esm node is already in 
arch/arm/dts/k3-j721e-r5-common-proc-board.dts in u-boot and since this 
patch is motivated to keep the u-boot and kernel dts in sync, I haven't 
added it here.

Although... I could add it here similar to j7200 and take action to move 
to the same in u-boot... maybe that is better route.

>> @@ -131,6 +131,7 @@ cbass_main: bus@100000 {
>>   		#size-cells = <2>;
>>   		ranges = <0x00 0x00100000 0x00 0x00100000 0x00 0x00020000>, /* ctrl mmr */
>>   			 <0x00 0x00600000 0x00 0x00600000 0x00 0x00031100>, /* GPIO */
>> +			 <0x00 0x00700000 0x00 0x00700000 0x00 0x00001000>, /* ESM */
>>   			 <0x00 0x00900000 0x00 0x00900000 0x00 0x00012000>, /* serdes */
>>   			 <0x00 0x00a40000 0x00 0x00a40000 0x00 0x00000800>, /* timesync router */
>>   			 <0x00 0x06000000 0x00 0x06000000 0x00 0x00400000>, /* USBSS0 */
>> -- 
>> 2.34.1
>>
>
  
Nishanth Menon May 3, 2023, 12:02 p.m. UTC | #3
On 17:25-20230503, Neha Malcom Francis wrote:
> Hi Nishanth,
> 
> On 03/05/23 17:18, Nishanth Menon wrote:
> > On 15:03-20230503, Neha Malcom Francis wrote:
> > > Add address entry mapping ESM on J721E.
> > > 
> > > Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
> > > ---
> > >   arch/arm64/boot/dts/ti/k3-j721e.dtsi | 1 +
> > >   1 file changed, 1 insertion(+)
> > > 
> > > diff --git a/arch/arm64/boot/dts/ti/k3-j721e.dtsi b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> > > index b912143b6a11..52bcde601eb8 100644
> > > --- a/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> > > +++ b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
> > 
> > Why is'nt esm node introduced here?
> > 
> 
> As it stands esm node is already in
> arch/arm/dts/k3-j721e-r5-common-proc-board.dts in u-boot and since this
> patch is motivated to keep the u-boot and kernel dts in sync, I haven't
> added it here.
> 
> Although... I could add it here similar to j7200 and take action to move to
> the same in u-boot... maybe that is better route.

The core problem we have is that part of the hardware description is in
u-boot and a large part is in kernel. That is wrong. eventually, we will
drop u-boot.dtsi and r5-xyz.dts files.. This cleanup journey is towards
that action.

we want kernel to be the canonical description of the hardware. which
means u-boot or zephyr or something else will eventually just copy the
kernel dts and it has information enough about the hardware to be able
to boot.

each of the dependent ecosystems (u-boot, zephyr etc..) have a little
bit of their own binding extensions (such as binman for u-boot), which
is maintained in their own repos. But, the SoC hardware description,
entirely should be available in kernel device tree.

So, please ensure esm is completely described as well.
  
Neha Malcom Francis May 3, 2023, 12:16 p.m. UTC | #4
Hi Nishanth

On 03/05/23 17:32, Nishanth Menon wrote:
> On 17:25-20230503, Neha Malcom Francis wrote:
>> Hi Nishanth,
>>
>> On 03/05/23 17:18, Nishanth Menon wrote:
>>> On 15:03-20230503, Neha Malcom Francis wrote:
>>>> Add address entry mapping ESM on J721E.
>>>>
>>>> Signed-off-by: Neha Malcom Francis <n-francis@ti.com>
>>>> ---
>>>>    arch/arm64/boot/dts/ti/k3-j721e.dtsi | 1 +
>>>>    1 file changed, 1 insertion(+)
>>>>
>>>> diff --git a/arch/arm64/boot/dts/ti/k3-j721e.dtsi b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
>>>> index b912143b6a11..52bcde601eb8 100644
>>>> --- a/arch/arm64/boot/dts/ti/k3-j721e.dtsi
>>>> +++ b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
>>>
>>> Why is'nt esm node introduced here?
>>>
>>
>> As it stands esm node is already in
>> arch/arm/dts/k3-j721e-r5-common-proc-board.dts in u-boot and since this
>> patch is motivated to keep the u-boot and kernel dts in sync, I haven't
>> added it here.
>>
>> Although... I could add it here similar to j7200 and take action to move to
>> the same in u-boot... maybe that is better route.
> 
> The core problem we have is that part of the hardware description is in
> u-boot and a large part is in kernel. That is wrong. eventually, we will
> drop u-boot.dtsi and r5-xyz.dts files.. This cleanup journey is towards
> that action.
> 
> we want kernel to be the canonical description of the hardware. which
> means u-boot or zephyr or something else will eventually just copy the
> kernel dts and it has information enough about the hardware to be able
> to boot.
> 
> each of the dependent ecosystems (u-boot, zephyr etc..) have a little
> bit of their own binding extensions (such as binman for u-boot), which
> is maintained in their own repos. But, the SoC hardware description,
> entirely should be available in kernel device tree.
> 
> So, please ensure esm is completely described as well.
> 

Right got it, will add the node in v5, thanks!
  

Patch

diff --git a/arch/arm64/boot/dts/ti/k3-j721e.dtsi b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
index b912143b6a11..52bcde601eb8 100644
--- a/arch/arm64/boot/dts/ti/k3-j721e.dtsi
+++ b/arch/arm64/boot/dts/ti/k3-j721e.dtsi
@@ -131,6 +131,7 @@  cbass_main: bus@100000 {
 		#size-cells = <2>;
 		ranges = <0x00 0x00100000 0x00 0x00100000 0x00 0x00020000>, /* ctrl mmr */
 			 <0x00 0x00600000 0x00 0x00600000 0x00 0x00031100>, /* GPIO */
+			 <0x00 0x00700000 0x00 0x00700000 0x00 0x00001000>, /* ESM */
 			 <0x00 0x00900000 0x00 0x00900000 0x00 0x00012000>, /* serdes */
 			 <0x00 0x00a40000 0x00 0x00a40000 0x00 0x00000800>, /* timesync router */
 			 <0x00 0x06000000 0x00 0x06000000 0x00 0x00400000>, /* USBSS0 */