Skip to contentSkip to conversation

dt-bindings: thermal: imx8mm-thermal: Document optional nvmem-cells

PATCH Applied Marek Vasut 3 participants 8 messages Dec 2, 2022 – Dec 31, 2022

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH v3 1/5 dt-bindings: thermal: imx8mm-thermal: Document optional nvmem-cells

The TMU TASR, TCALIVn, TRIM registers must be explicitly programmed with calibration values from OCOTP. Document optional phandle to OCOTP nvmem provider.

---
Cc: Adam Ford <[email protected]>
Cc: Alice Guo <[email protected]>
Cc: Amit Kucheria <[email protected]>
Cc: Daniel Lezcano <[email protected]>
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Li Jun <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Markus Niebel <[email protected]>
Cc: NXP Linux Team <[email protected]>
Cc: Peng Fan <[email protected]>
Cc: Pengutronix Kernel Team <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Richard Cochran <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Zhang Rui <[email protected]>
Cc: [email protected]
To: [email protected]
To: [email protected]
---
V2: Add AB from Krzysztof
V3: No change

---
 .../devicetree/bindings/thermal/imx8mm-thermal.yaml        | 7 +++++++
 1 file changed, 7 insertions(+)
Documentation/devicetree/bindings/thermal/imx8mm-thermal.yaml +7 -0
index 89c54e08ee61b..b90726229ac9c 100644
@@ -32,6 +32,13 @@ properties:
   clocks:
     maxItems: 1
 
+  nvmem-cells:
+    maxItems: 1
+    description: Phandle to the calibration data provided by ocotp
+
+  nvmem-cell-names:
+    const: calib
+
   "#thermal-sensor-cells":
     description: |
       Number of cells required to uniquely identify the thermal
Signature · 7 lines
2.35.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#1 lore raw

PATCH v3 2/5 arm64: dts: imx8m: Align SoC unique ID node unit address

Align the SoC unique ID DT node unit address with its reg property.

Fixes cbff23797fa1 ("arm64: dts: imx8m: add NVMEM provider and consumer to read soc unique ID")

---
Cc: Adam Ford <[email protected]>
Cc: Alice Guo <[email protected]>
Cc: Amit Kucheria <[email protected]>
Cc: Daniel Lezcano <[email protected]>
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Li Jun <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Markus Niebel <[email protected]>
Cc: NXP Linux Team <[email protected]>
Cc: Peng Fan <[email protected]>
Cc: Pengutronix Kernel Team <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Richard Cochran <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Zhang Rui <[email protected]>
Cc: [email protected]
To: [email protected]
To: [email protected]
---
V2: Add RB from Peng
V3: No change

---
 arch/arm64/boot/dts/freescale/imx8mm.dtsi | 2 +-
 arch/arm64/boot/dts/freescale/imx8mn.dtsi | 2 +-
 arch/arm64/boot/dts/freescale/imx8mp.dtsi | 2 +-
 arch/arm64/boot/dts/freescale/imx8mq.dtsi | 2 +-
 4 files changed, 4 insertions(+), 4 deletions(-)
arch/arm64/boot/dts/freescale/imx8mm.dtsi +1 -1
index 0c97aca8db6b6..423cb36cbcd53 100644
@@ -563,7 +563,7 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mm_uid: unique-id@410 {
+				imx8mm_uid: unique-id@4 {
 					reg = <0x4 0x8>;
 				};
 
arch/arm64/boot/dts/freescale/imx8mn.dtsi +1 -1
index 9b3a9e1384ae9..312e3abc35ea8 100644
@@ -564,7 +564,7 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mn_uid: unique-id@410 {
+				imx8mn_uid: unique-id@4 {
 					reg = <0x4 0x8>;
 				};
 
arch/arm64/boot/dts/freescale/imx8mp.dtsi +1 -1
index f18cf611f778e..c9459ed21b243 100644
@@ -426,7 +426,7 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mp_uid: unique-id@420 {
+				imx8mp_uid: unique-id@8 {
 					reg = <0x8 0x8>;
 				};
 
arch/arm64/boot/dts/freescale/imx8mq.dtsi +1 -1
index 5246b44a37d4a..2b6d3f4ff5d93 100644
@@ -593,7 +593,7 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mq_uid: soc-uid@410 {
+				imx8mq_uid: soc-uid@4 {
 					reg = <0x4 0x8>;
 				};
 
Signature · 1 line
2.35.1
#2 lore raw

Re: [PATCH v3 2/5] arm64: dts: imx8m: Align SoC unique ID node unit address

Quoted 5 lines · Marek Vasut

Align the SoC unique ID DT node unit address with its reg property.

Fixes cbff23797fa1 ("arm64: dts: imx8m: add NVMEM provider and consumer to read soc unique ID")

Applied 2/5, 3/5 and 4/5, thanks!

#3 lore raw

PATCH v3 3/5 arm64: dts: imx8m: Document the fuse address calculation

In reply to Marek Vasut

The mapping from OCOTP reg DT property to Fusemap Descriptions Table in the datasheet is often unclear. Add a comment to make it easier to find out how it works. No functional change.

---
Cc: Adam Ford <[email protected]>
Cc: Alice Guo <[email protected]>
Cc: Amit Kucheria <[email protected]>
Cc: Daniel Lezcano <[email protected]>
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Li Jun <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Markus Niebel <[email protected]>
Cc: NXP Linux Team <[email protected]>
Cc: Peng Fan <[email protected]>
Cc: Pengutronix Kernel Team <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Richard Cochran <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Zhang Rui <[email protected]>
Cc: [email protected]
To: [email protected]
To: [email protected]
---
V2: Add RB from Peng
V3: No change

---
 arch/arm64/boot/dts/freescale/imx8mm.dtsi | 19 ++++++++++++++++---
 arch/arm64/boot/dts/freescale/imx8mn.dtsi | 19 ++++++++++++++++---
 arch/arm64/boot/dts/freescale/imx8mp.dtsi | 21 +++++++++++++++++----
 arch/arm64/boot/dts/freescale/imx8mq.dtsi | 19 ++++++++++++++++---
 4 files changed, 65 insertions(+), 13 deletions(-)
arch/arm64/boot/dts/freescale/imx8mm.dtsi +16 -3
index 423cb36cbcd53..513c2de0caa15 100644
@@ -563,15 +563,28 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mm_uid: unique-id@4 {
+				/*
+				 * The register address below maps to the MX8M
+				 * Fusemap Description Table entries this way.
+				 * Assuming
+				 *   reg = <ADDR SIZE>;
+				 * then
+				 *   Fuse Address = (ADDR * 4) + 0x400
+				 * Note that if SIZE is greater than 4, then
+				 * each subsequent fuse is located at offset
+				 * +0x10 in Fusemap Description Table (e.g.
+				 * reg = <0x4 0x8> describes fuses 0x410 and
+				 * 0x420).
+				 */
+				imx8mm_uid: unique-id@4 { /* 0x410-0x420 */
 					reg = <0x4 0x8>;
 				};
 
-				cpu_speed_grade: speed-grade@10 {
+				cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
-				fec_mac_address: mac-address@90 {
+				fec_mac_address: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
 			};
arch/arm64/boot/dts/freescale/imx8mn.dtsi +16 -3
index 312e3abc35ea8..068f599cdf757 100644
@@ -564,15 +564,28 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mn_uid: unique-id@4 {
+				/*
+				 * The register address below maps to the MX8M
+				 * Fusemap Description Table entries this way.
+				 * Assuming
+				 *   reg = <ADDR SIZE>;
+				 * then
+				 *   Fuse Address = (ADDR * 4) + 0x400
+				 * Note that if SIZE is greater than 4, then
+				 * each subsequent fuse is located at offset
+				 * +0x10 in Fusemap Description Table (e.g.
+				 * reg = <0x4 0x8> describes fuses 0x410 and
+				 * 0x420).
+				 */
+				imx8mn_uid: unique-id@4 { /* 0x410-0x420 */
 					reg = <0x4 0x8>;
 				};
 
-				cpu_speed_grade: speed-grade@10 {
+				cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
-				fec_mac_address: mac-address@90 {
+				fec_mac_address: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
 			};
arch/arm64/boot/dts/freescale/imx8mp.dtsi +17 -4
index c9459ed21b243..ddcd5e23ba47d 100644
@@ -426,19 +426,32 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mp_uid: unique-id@8 {
+				/*
+				 * The register address below maps to the MX8M
+				 * Fusemap Description Table entries this way.
+				 * Assuming
+				 *   reg = <ADDR SIZE>;
+				 * then
+				 *   Fuse Address = (ADDR * 4) + 0x400
+				 * Note that if SIZE is greater than 4, then
+				 * each subsequent fuse is located at offset
+				 * +0x10 in Fusemap Description Table (e.g.
+				 * reg = <0x8 0x8> describes fuses 0x420 and
+				 * 0x430).
+				 */
+				imx8mp_uid: unique-id@8 { /* 0x420-0x430 */
 					reg = <0x8 0x8>;
 				};
 
-				cpu_speed_grade: speed-grade@10 {
+				cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
-				eth_mac1: mac-address@90 {
+				eth_mac1: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
 
-				eth_mac2: mac-address@96 {
+				eth_mac2: mac-address@96 { /* 0x658 */
 					reg = <0x96 6>;
 				};
 			};
arch/arm64/boot/dts/freescale/imx8mq.dtsi +16 -3
index 2b6d3f4ff5d93..8a2ec90b493d9 100644
@@ -593,15 +593,28 @@ ocotp: efuse@30350000 {
 				#address-cells = <1>;
 				#size-cells = <1>;
 
-				imx8mq_uid: soc-uid@4 {
+				/*
+				 * The register address below maps to the MX8M
+				 * Fusemap Description Table entries this way.
+				 * Assuming
+				 *   reg = <ADDR SIZE>;
+				 * then
+				 *   Fuse Address = (ADDR * 4) + 0x400
+				 * Note that if SIZE is greater than 4, then
+				 * each subsequent fuse is located at offset
+				 * +0x10 in Fusemap Description Table (e.g.
+				 * reg = <0x4 0x8> describes fuses 0x410 and
+				 * 0x420).
+				 */
+				imx8mq_uid: soc-uid@4 { /* 0x410-0x420 */
 					reg = <0x4 0x8>;
 				};
 
-				cpu_speed_grade: speed-grade@10 {
+				cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
-				fec_mac_address: mac-address@90 {
+				fec_mac_address: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
 			};
Signature · 7 lines
2.35.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#4 lore raw

PATCH v3 4/5 arm64: dts: imx8m: Add TMU phandle to calibration data in OCOTP

In reply to Marek Vasut

The TMU TASR, TCALIVn, TRIM registers must be explicitly programmed with calibration values in OCOTP. Add the OCOTP calibration values phandle so the TMU driver can perform this programming.

The MX8MM/MX8MN TMUv1 uses only one OCOTP cell, while MX8MP TMUv2 uses 4.

---
Cc: Adam Ford <[email protected]>
Cc: Alice Guo <[email protected]>
Cc: Amit Kucheria <[email protected]>
Cc: Daniel Lezcano <[email protected]>
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Li Jun <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Markus Niebel <[email protected]>
Cc: NXP Linux Team <[email protected]>
Cc: Peng Fan <[email protected]>
Cc: Pengutronix Kernel Team <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Richard Cochran <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Zhang Rui <[email protected]>
Cc: [email protected]
To: [email protected]
To: [email protected]
---
V2: Add RB from Peng
V3: No change

---
 arch/arm64/boot/dts/freescale/imx8mm.dtsi | 6 ++++++
 arch/arm64/boot/dts/freescale/imx8mn.dtsi | 6 ++++++
 arch/arm64/boot/dts/freescale/imx8mp.dtsi | 6 ++++++
 3 files changed, 18 insertions(+)
arch/arm64/boot/dts/freescale/imx8mm.dtsi +6 -0
index 513c2de0caa15..0cd7fff47c44d 100644
@@ -496,6 +496,8 @@ tmu: tmu@30260000 {
 				compatible = "fsl,imx8mm-tmu";
 				reg = <0x30260000 0x10000>;
 				clocks = <&clk IMX8MM_CLK_TMU_ROOT>;
+				nvmem-cells = <&tmu_calib>;
+				nvmem-cell-names = "calib";
 				#thermal-sensor-cells = <0>;
 			};
 
@@ -584,6 +586,10 @@ cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
+				tmu_calib: calib@3c { /* 0x4f0 */
+					reg = <0x3c 4>;
+				};
+
 				fec_mac_address: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
arch/arm64/boot/dts/freescale/imx8mn.dtsi +6 -0
index 068f599cdf757..5eef9b274edde 100644
@@ -498,6 +498,8 @@ tmu: tmu@30260000 {
 				compatible = "fsl,imx8mn-tmu", "fsl,imx8mm-tmu";
 				reg = <0x30260000 0x10000>;
 				clocks = <&clk IMX8MN_CLK_TMU_ROOT>;
+				nvmem-cells = <&tmu_calib>;
+				nvmem-cell-names = "calib";
 				#thermal-sensor-cells = <0>;
 			};
 
@@ -585,6 +587,10 @@ cpu_speed_grade: speed-grade@10 { /* 0x440 */
 					reg = <0x10 4>;
 				};
 
+				tmu_calib: calib@3c { /* 0x4f0 */
+					reg = <0x3c 4>;
+				};
+
 				fec_mac_address: mac-address@90 { /* 0x640 */
 					reg = <0x90 6>;
 				};
arch/arm64/boot/dts/freescale/imx8mp.dtsi +6 -0
index ddcd5e23ba47d..0173e394ad4d8 100644
@@ -380,6 +380,8 @@ tmu: tmu@30260000 {
 				compatible = "fsl,imx8mp-tmu";
 				reg = <0x30260000 0x10000>;
 				clocks = <&clk IMX8MP_CLK_TSENSOR_ROOT>;
+				nvmem-cells = <&tmu_calib>;
+				nvmem-cell-names = "calib";
 				#thermal-sensor-cells = <1>;
 			};
 
@@ -454,6 +456,10 @@ eth_mac1: mac-address@90 { /* 0x640 */
 				eth_mac2: mac-address@96 { /* 0x658 */
 					reg = <0x96 6>;
 				};
+
+				tmu_calib: calib@264 { /* 0xd90-0xdc0 */
+					reg = <0x264 0x10>;
+				};
 			};
 
 			anatop: clock-controller@30360000 {
Signature · 7 lines
2.35.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
#5 lore raw

PATCH v3 5/5 thermal/drivers/imx: Add support for loading calibration data from OCOTP

In reply to Marek Vasut

The TMU TASR, TCALIVn, TRIM registers must be explicitly programmed with calibration values in OCOTP. Add support for reading the OCOTP calibration data and programming those into the TMU hardware.

The MX8MM/MX8MN TMUv1 uses only one OCOTP cell, while MX8MP TMUv2 uses 4, the programming differs in each case.

Based on U-Boot commits:
70487ff386c ("imx8mm: Load fuse for TMU TCALIV and TASR") ebb9aab318b ("imx: load calibration parameters from fuse for i.MX8MP")

---
Cc: Adam Ford <[email protected]>
Cc: Alice Guo <[email protected]>
Cc: Amit Kucheria <[email protected]>
Cc: Daniel Lezcano <[email protected]>
Cc: Fabio Estevam <[email protected]>
Cc: Krzysztof Kozlowski <[email protected]>
Cc: Li Jun <[email protected]>
Cc: Lucas Stach <[email protected]>
Cc: Markus Niebel <[email protected]>
Cc: NXP Linux Team <[email protected]>
Cc: Peng Fan <[email protected]>
Cc: Pengutronix Kernel Team <[email protected]>
Cc: Rafael J. Wysocki <[email protected]>
Cc: Richard Cochran <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Zhang Rui <[email protected]>
Cc: [email protected]
To: [email protected]
To: [email protected]
---
V2: - Add RB from Peng
- Include slab.h to get kfree() definition
V3: - Replace %ld with %zu

---
 drivers/thermal/imx8mm_thermal.c | 164 +++++++++++++++++++++++++++++++
 1 file changed, 164 insertions(+)
drivers/thermal/imx8mm_thermal.c +164 -0
index e2c2673025a7a..a8b04dc0023de 100644
@@ -10,9 +10,11 @@
 #include <linux/err.h>
 #include <linux/io.h>
 #include <linux/module.h>
+#include <linux/nvmem-consumer.h>
 #include <linux/of.h>
 #include <linux/of_device.h>
 #include <linux/platform_device.h>
+#include <linux/slab.h>
 #include <linux/thermal.h>
 
 #include "thermal_core.h"
@@ -20,6 +22,22 @@
 #define TER			0x0	/* TMU enable */
 #define TPS			0x4
 #define TRITSR			0x20	/* TMU immediate temp */
+/* TMU calibration data registers */
+#define TASR			0x28
+#define TASR_BUF_SLOPE_MASK	GENMASK(19, 16)
+#define TASR_BUF_VREF_MASK	GENMASK(4, 0)	/* TMU_V1 */
+#define TASR_BUF_VERF_SEL_MASK	GENMASK(1, 0)	/* TMU_V2 */
+#define TCALIV(n)		(0x30 + ((n) * 4))
+#define TCALIV_EN		BIT(31)
+#define TCALIV_HR_MASK		GENMASK(23, 16)	/* TMU_V1 */
+#define TCALIV_RT_MASK		GENMASK(7, 0)	/* TMU_V1 */
+#define TCALIV_SNSR105C_MASK	GENMASK(27, 16)	/* TMU_V2 */
+#define TCALIV_SNSR25C_MASK	GENMASK(11, 0)	/* TMU_V2 */
+#define TRIM			0x3c
+#define TRIM_BJT_CUR_MASK	GENMASK(23, 20)
+#define TRIM_BGR_MASK		GENMASK(31, 28)
+#define TRIM_VLSB_MASK		GENMASK(15, 12)
+#define TRIM_EN_CH		BIT(7)
 
 #define TER_ADC_PD		BIT(30)
 #define TER_EN			BIT(31)
@@ -32,6 +50,25 @@
 #define SIGN_BIT		BIT(7)
 #define TEMP_VAL_MASK		GENMASK(6, 0)
 
+/* TMU OCOTP calibration data bitfields */
+#define ANA0_EN			BIT(25)
+#define ANA0_BUF_VREF_MASK	GENMASK(24, 20)
+#define ANA0_BUF_SLOPE_MASK	GENMASK(19, 16)
+#define ANA0_HR_MASK		GENMASK(15, 8)
+#define ANA0_RT_MASK		GENMASK(7, 0)
+#define TRIM2_VLSB_MASK		GENMASK(23, 20)
+#define TRIM2_BGR_MASK		GENMASK(19, 16)
+#define TRIM2_BJT_CUR_MASK	GENMASK(15, 12)
+#define TRIM2_BUF_SLOP_SEL_MASK	GENMASK(11, 8)
+#define TRIM2_BUF_VERF_SEL_MASK	GENMASK(7, 6)
+#define TRIM3_TCA25_0_LSB_MASK	GENMASK(31, 28)
+#define TRIM3_TCA40_0_MASK	GENMASK(27, 16)
+#define TRIM4_TCA40_1_MASK	GENMASK(31, 20)
+#define TRIM4_TCA105_0_MASK	GENMASK(19, 8)
+#define TRIM4_TCA25_0_MSB_MASK	GENMASK(7, 0)
+#define TRIM5_TCA105_1_MASK	GENMASK(23, 12)
+#define TRIM5_TCA25_1_MASK	GENMASK(11, 0)
+
 #define VER1_TEMP_LOW_LIMIT	10000
 #define VER2_TEMP_LOW_LIMIT	-40000
 #define VER2_TEMP_HIGH_LIMIT	125000
@@ -128,6 +165,129 @@ static void imx8mm_tmu_probe_sel_all(struct imx8mm_tmu *tmu)
 	writel_relaxed(val, tmu->base + TPS);
 }
 
+static int imx8mm_tmu_probe_set_calib_v1(struct platform_device *pdev,
+					 struct imx8mm_tmu *tmu)
+{
+	struct device *dev = &pdev->dev;
+	u32 ana0;
+	int ret;
+
+	ret = nvmem_cell_read_u32(&pdev->dev, "calib", &ana0);
+	if (ret) {
+		dev_warn(dev, "Failed to read OCOTP nvmem cell (%d).\n", ret);
+		return ret;
+	}
+
+	writel(FIELD_PREP(TASR_BUF_VREF_MASK,
+			  FIELD_GET(ANA0_BUF_VREF_MASK, ana0)) |
+	       FIELD_PREP(TASR_BUF_SLOPE_MASK,
+			  FIELD_GET(ANA0_BUF_SLOPE_MASK, ana0)),
+	       tmu->base + TASR);
+
+	writel(FIELD_PREP(TCALIV_RT_MASK, FIELD_GET(ANA0_RT_MASK, ana0)) |
+	       FIELD_PREP(TCALIV_HR_MASK, FIELD_GET(ANA0_HR_MASK, ana0)) |
+	       ((ana0 & ANA0_EN) ? TCALIV_EN : 0),
+	       tmu->base + TCALIV(0));
+
+	return 0;
+}
+
+static int imx8mm_tmu_probe_set_calib_v2(struct platform_device *pdev,
+					 struct imx8mm_tmu *tmu)
+{
+	struct device *dev = &pdev->dev;
+	struct nvmem_cell *cell;
+	u32 trim[4] = { 0 };
+	size_t len;
+	void *buf;
+
+	cell = nvmem_cell_get(dev, "calib");
+	if (IS_ERR(cell))
+		return PTR_ERR(cell);
+
+	buf = nvmem_cell_read(cell, &len);
+	nvmem_cell_put(cell);
+
+	if (IS_ERR(buf))
+		return PTR_ERR(buf);
+
+	memcpy(trim, buf, min(len, sizeof(trim)));
+	kfree(buf);
+
+	if (len != 16) {
+		dev_err(dev,
+			"OCOTP nvmem cell length is %zu, must be 16.\n", len);
+		return -EINVAL;
+	}
+
+	/* Blank sample hardware */
+	if (!trim[0] && !trim[1] && !trim[2] && !trim[3]) {
+		/* Use a default 25C binary codes */
+		writel(FIELD_PREP(TCALIV_SNSR25C_MASK, 0x63c),
+		       tmu->base + TCALIV(0));
+		writel(FIELD_PREP(TCALIV_SNSR25C_MASK, 0x63c),
+		       tmu->base + TCALIV(1));
+		return 0;
+	}
+
+	writel(FIELD_PREP(TASR_BUF_VERF_SEL_MASK,
+			  FIELD_GET(TRIM2_BUF_VERF_SEL_MASK, trim[0])) |
+	       FIELD_PREP(TASR_BUF_SLOPE_MASK,
+			  FIELD_GET(TRIM2_BUF_SLOP_SEL_MASK, trim[0])),
+	       tmu->base + TASR);
+
+	writel(FIELD_PREP(TRIM_BJT_CUR_MASK,
+			  FIELD_GET(TRIM2_BJT_CUR_MASK, trim[0])) |
+	       FIELD_PREP(TRIM_BGR_MASK, FIELD_GET(TRIM2_BGR_MASK, trim[0])) |
+	       FIELD_PREP(TRIM_VLSB_MASK, FIELD_GET(TRIM2_VLSB_MASK, trim[0])) |
+	       TRIM_EN_CH,
+	       tmu->base + TRIM);
+
+	writel(FIELD_PREP(TCALIV_SNSR25C_MASK,
+			  FIELD_GET(TRIM3_TCA25_0_LSB_MASK, trim[1]) |
+			  (FIELD_GET(TRIM4_TCA25_0_MSB_MASK, trim[2]) << 4)) |
+	       FIELD_PREP(TCALIV_SNSR105C_MASK,
+			  FIELD_GET(TRIM4_TCA105_0_MASK, trim[2])),
+	       tmu->base + TCALIV(0));
+
+	writel(FIELD_PREP(TCALIV_SNSR25C_MASK,
+			  FIELD_GET(TRIM5_TCA25_1_MASK, trim[3])) |
+	       FIELD_PREP(TCALIV_SNSR105C_MASK,
+			  FIELD_GET(TRIM5_TCA105_1_MASK, trim[3])),
+	       tmu->base + TCALIV(1));
+
+	writel(FIELD_PREP(TCALIV_SNSR25C_MASK,
+			  FIELD_GET(TRIM3_TCA40_0_MASK, trim[1])) |
+	       FIELD_PREP(TCALIV_SNSR105C_MASK,
+			  FIELD_GET(TRIM4_TCA40_1_MASK, trim[2])),
+	       tmu->base + TCALIV(2));
+
+	return 0;
+}
+
+static int imx8mm_tmu_probe_set_calib(struct platform_device *pdev,
+				      struct imx8mm_tmu *tmu)
+{
+	struct device *dev = &pdev->dev;
+
+	/*
+	 * Lack of calibration data OCOTP reference is not considered
+	 * fatal to retain compatibility with old DTs. It is however
+	 * strongly recommended to update such old DTs to get correct
+	 * temperature compensation values for each SoC.
+	 */
+	if (!of_find_property(pdev->dev.of_node, "nvmem-cells", NULL)) {
+		dev_warn(dev,
+			 "No OCOTP nvmem reference found, SoC-specific calibration not loaded. Please update your DT.\n");
+		return 0;
+	}
+
+	if (tmu->socdata->version == TMU_VER1)
+		return imx8mm_tmu_probe_set_calib_v1(pdev, tmu);
+
+	return imx8mm_tmu_probe_set_calib_v2(pdev, tmu);
+}
+
 static int imx8mm_tmu_probe(struct platform_device *pdev)
 {
 	const struct thermal_soc_data *data;
@@ -180,6 +340,10 @@ static int imx8mm_tmu_probe(struct platform_device *pdev)
 
 	platform_set_drvdata(pdev, tmu);
 
+	ret = imx8mm_tmu_probe_set_calib(pdev, tmu);
+	if (ret)
+		goto disable_clk;
+
 	/* enable all the probes for V2 TMU */
 	if (tmu->socdata->version == TMU_VER2)
 		imx8mm_tmu_probe_sel_all(tmu);
Signature · 1 line
2.35.1
#6 lore raw

Re: [PATCH v3 5/5] thermal/drivers/imx: Add support for loading calibration data from OCOTP

Quoted 7 lines · Marek Vasut

The TMU TASR, TCALIVn, TRIM registers must be explicitly programmed with calibration values in OCOTP. Add support for reading the OCOTP calibration data and programming those into the TMU hardware.

The MX8MM/MX8MN TMUv1 uses only one OCOTP cell, while MX8MP TMUv2 uses 4, the programming differs in each case.

Quoted 6 lines

Based on U-Boot commits:
70487ff386c ("imx8mm: Load fuse for TMU TCALIV and TASR") ebb9aab318b ("imx: load calibration parameters from fuse for i.MX8MP")

Applied, thanks

Signature · 5 lines
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog
#7 lore raw
In reply to Marek Vasut
Quoted 7 lines · Marek Vasut

The TMU TASR, TCALIVn, TRIM registers must be explicitly programmed with calibration values from OCOTP. Document optional phandle to OCOTP nvmem provider.

Applied, thanks

Signature · 5 lines
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro:  <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog