SA8540P-ride is one of the Qualcomm platforms that does not have access
to UEFI runtime services and on which the RTC registers are read-only,
as described in:
https://lore.kernel.org/all/20230202155448.6715-1-johan+linaro@kernel.org/

Reserve four bytes in one of the PMIC registers to hold the RTC offset
the same way as it was done for sc8280xp-crd which has similar
limitations:
    commit e67b45582c5e ("arm64: dts: qcom: sc8280xp-crd: enable rtc")

One small difference on SA8540P-ride, the PMIC register bank SDAM6 is
not writable, so use SDAM7 instead.

Reviewed-by: Caleb Connolly <caleb.connolly@linaro.org>
Signed-off-by: Eric Chanudet <echanude@redhat.com>
---
Changes in v2:
* Default pmm8540a_sdam_7 status to "disabled" in sa8540p-pmics.dtsi.

 arch/arm64/boot/dts/qcom/sa8540p-pmics.dtsi | 11 ++++++++++-
 arch/arm64/boot/dts/qcom/sa8540p-ride.dts   | 15 +++++++++++++++
 2 files changed, 25 insertions(+), 1 deletion(-)

diff --git a/arch/arm64/boot/dts/qcom/sa8540p-pmics.dtsi b/arch/arm64/boot/dts/qcom/sa8540p-pmics.dtsi
index 1221be89b3de..85a077619ae4 100644
--- a/arch/arm64/boot/dts/qcom/sa8540p-pmics.dtsi
+++ b/arch/arm64/boot/dts/qcom/sa8540p-pmics.dtsi
@@ -14,7 +14,7 @@ pmm8540a: pmic@0 {
 		#address-cells = <1>;
 		#size-cells = <0>;
 
-		rtc@6000 {
+		pmm8540a_rtc: rtc@6000 {
 			compatible = "qcom,pm8941-rtc";
 			reg = <0x6000>, <0x6100>;
 			reg-names = "rtc", "alarm";
@@ -22,6 +22,15 @@ rtc@6000 {
 			wakeup-source;
 		};
 
+		pmm8540a_sdam_7: nvram@b610 {
+			compatible = "qcom,spmi-sdam";
+			reg = <0xb610>;
+			#address-cells = <1>;
+			#size-cells = <1>;
+			ranges = <0 0xb610 0xb0>;
+			status = "disabled";
+		};
+
 		pmm8540a_gpios: gpio@c000 {
 			compatible = "qcom,pm8150-gpio", "qcom,spmi-gpio";
 			reg = <0xc000>;
diff --git a/arch/arm64/boot/dts/qcom/sa8540p-ride.dts b/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
index 5a26974dcf8f..608dd71a3f1c 100644
--- a/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
+++ b/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
@@ -407,6 +407,21 @@ &pcie3a_phy {
 	status = "okay";
 };
 
+&pmm8540a_rtc {
+	nvmem-cells = <&rtc_offset>;
+	nvmem-cell-names = "offset";
+
+	status = "okay";
+};
+
+&pmm8540a_sdam_7 {
+	status = "okay";
+
+	rtc_offset: rtc-offset@ac {
+		reg = <0xac 0x4>;
+	};
+};
+
 &qup0 {
 	status = "okay";
 };
-- 
2.41.0

On Tue, Jul 18, 2023 at 10:46:10AM -0400, Eric Chanudet wrote:
> diff --git a/arch/arm64/boot/dts/qcom/sa8540p-ride.dts b/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
[..]
> +&pmm8540a_sdam_7 {
> +	status = "okay";
> +
> +	rtc_offset: rtc-offset@ac {
> +		reg = <0xac 0x4>;

I'm still trying to get confirmation that this is a good choice.

Regards,
Bjorn

On Fri, Jul 21, 2023 at 08:59:30PM -0700, Bjorn Andersson wrote:
> On Tue, Jul 18, 2023 at 10:46:10AM -0400, Eric Chanudet wrote:
> > diff --git a/arch/arm64/boot/dts/qcom/sa8540p-ride.dts b/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
> [..]
> > +&pmm8540a_sdam_7 {
> > +	status = "okay";
> > +
> > +	rtc_offset: rtc-offset@ac {
> > +		reg = <0xac 0x4>;
> 
> I'm still trying to get confirmation that this is a good choice.
> 

I'm recommended that you use 0xa0 from SDAM2, "preferably in the second
PMM8540", instead.

Can you please give this a try, Eric?

Regards,
Bjorn

On Tue, Aug 08, 2023 at 02:36:46PM -0700, Bjorn Andersson wrote:
> On Fri, Jul 21, 2023 at 08:59:30PM -0700, Bjorn Andersson wrote:
> > On Tue, Jul 18, 2023 at 10:46:10AM -0400, Eric Chanudet wrote:
> > > diff --git a/arch/arm64/boot/dts/qcom/sa8540p-ride.dts b/arch/arm64/boot/dts/qcom/sa8540p-ride.dts
> > [..]
> > > +&pmm8540a_sdam_7 {
> > > +	status = "okay";
> > > +
> > > +	rtc_offset: rtc-offset@ac {
> > > +		reg = <0xac 0x4>;
> > 
> > I'm still trying to get confirmation that this is a good choice.
> > 
> 
> I'm recommended that you use 0xa0 from SDAM2, "preferably in the second
> PMM8540", instead.
> 
> Can you please give this a try, Eric?

That worked, the offset was saved and read back upon rebooting from the
system. I did get a defer on some tests waiting for the second pmic,
which I presume is not surprising:
[    0.257064] spmi spmi-0: PMIC arbiter version v5 (0x50020000)
[    8.340386] platform c440000.spmi:pmic@0:rtc@6000: error -EPROBE_DEFER: wait for supplier /soc@0/spmi@c440000/pmic@4/nvram@b110/rtc-offset@a0
[    8.393201] platform c440000.spmi:pmic@0:rtc@6000: error -EPROBE_DEFER: wait for supplier /soc@0/spmi@c440000/pmic@4/nvram@b110/rtc-offset@a0
[    8.465407] rtc-pm8xxx c440000.spmi:pmic@0:rtc@6000: registered as rtc0
[    8.479612] rtc-pm8xxx c440000.spmi:pmic@0:rtc@6000: setting system clock to 2023-08-09T19:16:08 UTC (1691608568)

Here is the v3 with the changes:
    https://lore.kernel.org/linux-arm-msm/20230809203506.1833205-1-echanude@redhat.com

Thank you again for the feedback.

-- 
Eric Chanudet