mbox series

[v2,0/3] Add DTS for SDM845 SoC and MTP

Message ID 20180131161941.29865-1-rnayak@codeaurora.org
Headers show
Series Add DTS for SDM845 SoC and MTP | expand

Message

Rajendra Nayak Jan. 31, 2018, 4:19 p.m. UTC
Now that we have serial[1], pinctrl[2] and clock[3]
drivers for SDM845 SoC out on the lists, heres the
basic dts files needed to boot a SDM845 based MTP
board to a ramfs based serial console shell.

[1] https://patchwork.ozlabs.org/cover/860251/
[2] https://patchwork.kernel.org/patch/10157143/
[3] https://lkml.org/lkml/2018/1/22/78

changes in v2:
* dropped cpu-map
* dropped GIC_CPU_MASK_SIMPLE()
* Added new cpu compatible for kryo385
* added ITS node, marked as disabled
* other cosmetic fixes 

Rajendra Nayak (3):
  Documentation: dt-bindings: arm: Document kryo385 cpu
  arm64: dts: sdm845: Add minimal dts/dtsi files for sdm845 SoC and MTP
  arm64: dts: sdm845: Add serial console support

 Documentation/devicetree/bindings/arm/cpus.txt |   1 +
 arch/arm64/boot/dts/qcom/Makefile              |   1 +
 arch/arm64/boot/dts/qcom/sdm845-mtp.dts        |  55 +++++
 arch/arm64/boot/dts/qcom/sdm845.dtsi           | 298 +++++++++++++++++++++++++
 4 files changed, 355 insertions(+)
 create mode 100644 arch/arm64/boot/dts/qcom/sdm845-mtp.dts
 create mode 100644 arch/arm64/boot/dts/qcom/sdm845.dtsi

Comments

Doug Anderson Feb. 6, 2018, 7:35 p.m. UTC | #1
Hi,

On Wed, Jan 31, 2018 at 8:19 AM, Rajendra Nayak <rnayak@codeaurora.org> wrote:
> Add the qup uart node and geni se instance needed to
> support the serial console on the MTP.
>
> Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
> ---
>  arch/arm64/boot/dts/qcom/sdm845-mtp.dts | 42 +++++++++++++++++++++++++++++++++
>  arch/arm64/boot/dts/qcom/sdm845.dtsi    | 21 +++++++++++++++++
>  2 files changed, 63 insertions(+)
>
> diff --git a/arch/arm64/boot/dts/qcom/sdm845-mtp.dts b/arch/arm64/boot/dts/qcom/sdm845-mtp.dts
> index 617c7bb25fb1..42fbf2ab9a2d 100644
> --- a/arch/arm64/boot/dts/qcom/sdm845-mtp.dts
> +++ b/arch/arm64/boot/dts/qcom/sdm845-mtp.dts
> @@ -10,4 +10,46 @@
>  / {
>         model = "Qualcomm Technologies, Inc. SDM845 MTP";
>         compatible = "qcom,sdm845-mtp";
> +
> +       aliases {
> +               serial0 = &qup_uart2;
> +       };
> +
> +       chosen {
> +               stdout-path = "serial0";
> +       };
> +
> +       soc {

I don't know if there's an official position, but in general I'm seen
people use the actual "soc" alias here.  AKA at the top level of this
dts, just do:

&soc {
  ...
};

Normally doing stuff like that is useful so you don't need to
replicate the whole hierarchy.  In this case that's not a huge
savings, but it can be nice to be consistent.  In the very least it
saves you a level of indentation.


> +               serial@a84000 {
> +                       status = "okay";
> +               };

Similarly here you can use the alias from the sdm845.dtsi file to
avoid replicating the hierarchy.  AKA at the top level do:

&qup_uart2 {
  status = "okay";
};

In this case it saves you 2 levels of indentation.

> +
> +               pinctrl@3400000 {
> +                       qup_uart2_default: qup_uart2_default {

I'm not sure how persnickety I should be, but according to
<https://elinux.org/Device_Tree_Linux>:

  node names use dash "-" instead of underscore "_"

...but, of course, labels can't use dashes (and the same page says to
use underscore for labels).  This is why, in rk3288 for instance, you
see:

i2c2_xfer: i2c2-xfer {
  rockchip,pins = <6 9 RK_FUNC_1 &pcfg_pull_none>,
  <6 10 RK_FUNC_1 &pcfg_pull_none>;
};

AKA the label and the node name are the same but the label uses "_"
and the node names use "-".


> +                               pinmux {
> +                                       function = "qup9";
> +                                       pins = "gpio4", "gpio5";
> +                               };
> +
> +                               pinconf {
> +                                       pins = "gpio4", "gpio5";
> +                                       drive-strength = <2>;
> +                                       bias-disable;
> +                               };
> +                       };
> +
> +                       qup_uart2_sleep: qup_uart2_sleep {
> +                               pinmux {
> +                                       function = "gpio";
> +                                       pins = "gpio4", "gpio5";
> +                               };
> +
> +                               pinconf {
> +                                       pins = "gpio4", "gpio5";
> +                                       drive-strength = <2>;
> +                                       bias-disable;
> +                               };
> +                       };
> +               };

As per my (admittedly tardy) response to v1, these shouldn't be in the
board file because every board that happens to need uart2 will need to
duplicate these definitions.


> +       };
>  };
> diff --git a/arch/arm64/boot/dts/qcom/sdm845.dtsi b/arch/arm64/boot/dts/qcom/sdm845.dtsi
> index 02520f19e4ca..c4ce70840acf 100644
> --- a/arch/arm64/boot/dts/qcom/sdm845.dtsi
> +++ b/arch/arm64/boot/dts/qcom/sdm845.dtsi
> @@ -4,6 +4,7 @@
>   */
>
>  #include <dt-bindings/interrupt-controller/arm-gic.h>
> +#include <dt-bindings/clock/qcom,gcc-sdm845.h>
>
>  / {
>         model = "Qualcomm Technologies, Inc. SDM845";
> @@ -273,5 +274,25 @@
>                         cell-index = <0>;
>                 };
>
> +               qup_1: qcom,geni_se@ac0000 {
> +                       compatible = "qcom,geni-se-qup";
> +                       reg = <0xac0000 0x6000>;

I think you may have mentioned this in another context, but this
doesn't match the current bindings.  Some clocks should be here.
...and it looks like the uart should be a subnode.


> +               };
> +
> +               qup_uart2: serial@a84000 {
> +                       compatible = "qcom,geni-console", "qcom,geni-uart";
> +                       reg = <0xa84000 0x4000>;
> +                       reg-names = "se_phys";
> +                       clock-names = "se-clk", "m-ahb", "s-ahb";
> +                       clocks = <&gcc GCC_QUPV3_WRAP1_S1_CLK>,
> +                                <&gcc GCC_QUPV3_WRAP_1_M_AHB_CLK>,
> +                                <&gcc GCC_QUPV3_WRAP_1_S_AHB_CLK>;
> +                       pinctrl-names = "default", "sleep";
> +                       pinctrl-0 = <&qup_uart2_default>;
> +                       pinctrl-1 = <&qup_uart2_sleep>;

In general the "SoC" dtsi file shouldn't be relying on definitions
that are made in the board "dts" file.

...if everyone disagrees with me and we really decide to duplicate all
the pinctrl in all board files, we'll need to move the
"pinctrl-names", "pinctrl-0", and "pinctrl-1" to the board files too.
;)


-Doug
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Doug Anderson Feb. 6, 2018, 9:55 p.m. UTC | #2
Hi,

On Wed, Jan 31, 2018 at 8:19 AM, Rajendra Nayak <rnayak@codeaurora.org> wrote:
> diff --git a/arch/arm64/boot/dts/qcom/sdm845.dtsi b/arch/arm64/boot/dts/qcom/sdm845.dtsi
> new file mode 100644
> index 000000000000..02520f19e4ca
> --- /dev/null
> +++ b/arch/arm64/boot/dts/qcom/sdm845.dtsi
> @@ -0,0 +1,277 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (c) 2018, The Linux Foundation. All rights reserved.
> + */
> +
> +#include <dt-bindings/interrupt-controller/arm-gic.h>
> +
> +/ {
> +       model = "Qualcomm Technologies, Inc. SDM845";

I'm fairly certain that "model" doesn't belong in the SoC .dtsi file.
Only in the board .dts file.


> +       clocks {
> +               xo_board: xo_board {

Just to make it explicit: see my comments in patch 3/3 in this series
about using "_" in node names.  I believe this should be:

  xo_board: xo-board {


> +               spmi_bus: qcom,spmi@c440000 {

Drop the qcom in the node name.  AKA, I believe this should be:

spmi_bus: spmi@c440000 {

Specifically the node name is supposed to be a generic component name
then with an address.  I see that Rob Herring said the same thing when
he reviewed v1 of this patch just now (it seems like people are still
commenting there, so make sure you collect the latest feedback from
there when re-spinning).


-Doug
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Rajendra Nayak Feb. 7, 2018, 4:28 a.m. UTC | #3
[]..

>> @@ -10,4 +10,46 @@
>>  / {
>>         model = "Qualcomm Technologies, Inc. SDM845 MTP";
>>         compatible = "qcom,sdm845-mtp";
>> +
>> +       aliases {
>> +               serial0 = &qup_uart2;
>> +       };
>> +
>> +       chosen {
>> +               stdout-path = "serial0";
>> +       };
>> +
>> +       soc {
> 
> I don't know if there's an official position, but in general I'm seen
> people use the actual "soc" alias here.  AKA at the top level of this
> dts, just do:
> 
> &soc {
>   ...
> };
> 
> Normally doing stuff like that is useful so you don't need to
> replicate the whole hierarchy.  In this case that's not a huge
> savings, but it can be nice to be consistent.  In the very least it
> saves you a level of indentation.
> 
> 
>> +               serial@a84000 {
>> +                       status = "okay";
>> +               };
> 
> Similarly here you can use the alias from the sdm845.dtsi file to
> avoid replicating the hierarchy.  AKA at the top level do:
> 
> &qup_uart2 {
>   status = "okay";
> };
> 
> In this case it saves you 2 levels of indentation.

Right. Andy/Bjorn, are there any preferences here?
I see we don't do this for the other board files, and I not sure
theres a specific reasoning for how its currently done and if we
need to stick to it.

> 
>> +
>> +               pinctrl@3400000 {
>> +                       qup_uart2_default: qup_uart2_default {
> 
> I'm not sure how persnickety I should be, but according to
> <https://elinux.org/Device_Tree_Linux>:
> 
>   node names use dash "-" instead of underscore "_"
> 
> ...but, of course, labels can't use dashes (and the same page says to
> use underscore for labels).  This is why, in rk3288 for instance, you
> see:
> 
> i2c2_xfer: i2c2-xfer {
>   rockchip,pins = <6 9 RK_FUNC_1 &pcfg_pull_none>,
>   <6 10 RK_FUNC_1 &pcfg_pull_none>;
> };
> 
> AKA the label and the node name are the same but the label uses "_"
> and the node names use "-".

Sure, I'll fix these up.

[]
>> diff --git a/arch/arm64/boot/dts/qcom/sdm845.dtsi b/arch/arm64/boot/dts/qcom/sdm845.dtsi
>> index 02520f19e4ca..c4ce70840acf 100644
>> --- a/arch/arm64/boot/dts/qcom/sdm845.dtsi
>> +++ b/arch/arm64/boot/dts/qcom/sdm845.dtsi
>> @@ -4,6 +4,7 @@
>>   */
>>
>>  #include <dt-bindings/interrupt-controller/arm-gic.h>
>> +#include <dt-bindings/clock/qcom,gcc-sdm845.h>
>>
>>  / {
>>         model = "Qualcomm Technologies, Inc. SDM845";
>> @@ -273,5 +274,25 @@
>>                         cell-index = <0>;
>>                 };
>>
>> +               qup_1: qcom,geni_se@ac0000 {
>> +                       compatible = "qcom,geni-se-qup";
>> +                       reg = <0xac0000 0x6000>;
> 
> I think you may have mentioned this in another context, but this
> doesn't match the current bindings.  Some clocks should be here.
> ...and it looks like the uart should be a subnode.

right, these were tested with the v1 set for serial. I will update them.

regards
Rajendra
Rajendra Nayak Feb. 7, 2018, 4:49 a.m. UTC | #4
On 02/07/2018 03:25 AM, Doug Anderson wrote:
> Hi,
> 
> On Wed, Jan 31, 2018 at 8:19 AM, Rajendra Nayak <rnayak@codeaurora.org> wrote:
>> diff --git a/arch/arm64/boot/dts/qcom/sdm845.dtsi b/arch/arm64/boot/dts/qcom/sdm845.dtsi
>> new file mode 100644
>> index 000000000000..02520f19e4ca
>> --- /dev/null
>> +++ b/arch/arm64/boot/dts/qcom/sdm845.dtsi
>> @@ -0,0 +1,277 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * Copyright (c) 2018, The Linux Foundation. All rights reserved.
>> + */
>> +
>> +#include <dt-bindings/interrupt-controller/arm-gic.h>
>> +
>> +/ {
>> +       model = "Qualcomm Technologies, Inc. SDM845";
> 
> I'm fairly certain that "model" doesn't belong in the SoC .dtsi file.
> Only in the board .dts file.
> 
> 
>> +       clocks {
>> +               xo_board: xo_board {
> 
> Just to make it explicit: see my comments in patch 3/3 in this series
> about using "_" in node names.  I believe this should be:
> 
>   xo_board: xo-board {
> 
> 
>> +               spmi_bus: qcom,spmi@c440000 {
> 
> Drop the qcom in the node name.  AKA, I believe this should be:
> 
> spmi_bus: spmi@c440000 {
> 
> Specifically the node name is supposed to be a generic component name
> then with an address.  I see that Rob Herring said the same thing when
> he reviewed v1 of this patch just now (it seems like people are still
> commenting there, so make sure you collect the latest feedback from
> there when re-spinning).

yes, I'll make sure I fix up based on Robs' review of the v1.