[PATCH v2 0/7] qcom-cpufreq-hw: Add CPU clock provider support

[PATCH v2 0/7] qcom-cpufreq-hw: Add CPU clock provider support

[Manivannan Sadhasivam]

Hello,

This series adds clock provider support to the Qcom CPUFreq driver for
supplying the clocks to the CPU cores in Qcom SoCs.

The Qualcomm platforms making use of CPUFreq HW Engine (EPSS/OSM) supply
clocks to the CPU cores. But this is not represented clearly in devicetree.
There is no clock coming out of the CPUFreq HW node to the CPU. This created
an issue [1] with the OPP core when a recent enhancement series was submitted.
Eventhough the issue got fixed in the OPP framework in the meantime, that's
not a proper solution and this series aims to fix it properly.

There was also an attempt made by Viresh [2] to fix the issue by moving the
clocks supplied to the CPUFreq HW node to the CPU. But that was not accepted
since those clocks belong to the CPUFreq HW node only.

The proposal here is to add clock provider support to the Qcom CPUFreq HW
driver to supply clocks to the CPUs that comes out of the EPSS/OSM block.
This correctly reflects the hardware implementation.

The clock provider is a simple one that just provides the frequency of the
clocks supplied to each frequency domain in the SoC using .recalc_rate()
callback. The frequency supplied by the driver will be the actual frequency
that comes out of the EPSS/OSM block after the DCVS operation. This frequency
is not same as what the CPUFreq framework has set but it is the one that gets
supplied to the CPUs after throttling by LMh.

This series has been tested on SM8450 based dev board with the OPP hack removed
and hence there is a DTS change only for that platform. Once this series gets
accepted, rest of the platform DTS can also be modified and finally the hack on
the OPP core can be dropped.

Thanks,
Mani

[1] https://lore.kernel.org/lkml/YsxSkswzsqgMOc0l@hovoldconsulting.com/
[2] https://lore.kernel.org/lkml/20220801054255.GA12039@thinkpad/t/

Changes in v2:

* Moved the qcom_cpufreq_data allocation to probe
* Added single clock provider with multiple clks for each freq domain
* Moved soc_data to qcom_cpufreq struct
* Added Rob's review for binding

Manivannan Sadhasivam (7):
  dt-bindings: cpufreq: cpufreq-qcom-hw: Add cpufreq clock provider
  arm64: dts: qcom: sm8450: Supply clock from cpufreq node to CPUs
  cpufreq: qcom-hw: Remove un-necessary cpumask_empty() check
  cpufreq: qcom-hw: Allocate qcom_cpufreq_data during probe
  cpufreq: qcom-hw: Use cached dev pointer in probe()
  cpufreq: qcom-hw: Move soc_data to struct qcom_cpufreq
  cpufreq: qcom-hw: Add CPU clock provider support

 .../bindings/cpufreq/cpufreq-qcom-hw.yaml     |  12 ++
 arch/arm64/boot/dts/qcom/sm8450.dtsi          |   9 +
 drivers/cpufreq/qcom-cpufreq-hw.c             | 166 ++++++++++--------
 3 files changed, 118 insertions(+), 69 deletions(-)

-- 
2.25.1

[PATCH v2 1/7] dt-bindings: cpufreq: cpufreq-qcom-hw: Add cpufreq clock provider

[Manivannan Sadhasivam]

Qualcomm platforms making use of CPUFreq HW Engine (EPSS/OSM) supply clocks
to the CPU cores. Document the same in the binding to reflect the actual
implementation.

CPUFreq HW will become the clock provider and CPU cores will become the
clock consumers.

The clock index for each CPU core is based on the frequency domain index.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Rob Herring <robh@kernel.org>
---
 .../devicetree/bindings/cpufreq/cpufreq-qcom-hw.yaml | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/Documentation/devicetree/bindings/cpufreq/cpufreq-qcom-hw.yaml b/Documentation/devicetree/bindings/cpufreq/cpufreq-qcom-hw.yaml
index cbba8979fe0e..2e0336163ffb 100644
--- a/Documentation/devicetree/bindings/cpufreq/cpufreq-qcom-hw.yaml
+++ b/Documentation/devicetree/bindings/cpufreq/cpufreq-qcom-hw.yaml
@@ -57,6 +57,9 @@ properties:
   '#freq-domain-cells':
     const: 1
 
+  '#clock-cells':
+    const: 1
+
 required:
   - compatible
   - reg
@@ -84,6 +87,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_0>;
         qcom,freq-domain = <&cpufreq_hw 0>;
+        clocks = <&cpufreq_hw 0>;
         L2_0: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -100,6 +104,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_100>;
         qcom,freq-domain = <&cpufreq_hw 0>;
+        clocks = <&cpufreq_hw 0>;
         L2_100: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -113,6 +118,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_200>;
         qcom,freq-domain = <&cpufreq_hw 0>;
+        clocks = <&cpufreq_hw 0>;
         L2_200: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -126,6 +132,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_300>;
         qcom,freq-domain = <&cpufreq_hw 0>;
+        clocks = <&cpufreq_hw 0>;
         L2_300: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -139,6 +146,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_400>;
         qcom,freq-domain = <&cpufreq_hw 1>;
+        clocks = <&cpufreq_hw 1>;
         L2_400: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -152,6 +160,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_500>;
         qcom,freq-domain = <&cpufreq_hw 1>;
+        clocks = <&cpufreq_hw 1>;
         L2_500: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -165,6 +174,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_600>;
         qcom,freq-domain = <&cpufreq_hw 1>;
+        clocks = <&cpufreq_hw 1>;
         L2_600: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -178,6 +188,7 @@ examples:
         enable-method = "psci";
         next-level-cache = <&L2_700>;
         qcom,freq-domain = <&cpufreq_hw 1>;
+        clocks = <&cpufreq_hw 1>;
         L2_700: l2-cache {
           compatible = "cache";
           next-level-cache = <&L3_0>;
@@ -198,6 +209,7 @@ examples:
         clock-names = "xo", "alternate";
 
         #freq-domain-cells = <1>;
+        #clock-cells = <1>;
       };
     };
 ...
-- 
2.25.1

[PATCH v2 2/7] arm64: dts: qcom: sm8450: Supply clock from cpufreq node to CPUs

[Manivannan Sadhasivam]

Qualcomm platforms making use of CPUFreq HW Engine (EPSS/OSM) supply clocks
to the CPU cores. But this relationship is not represented in DTS so far.

So let's make cpufreq node as the clock provider and CPU nodes as the
consumers. The clock index for each CPU node is based on the frequency
domain index.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 arch/arm64/boot/dts/qcom/sm8450.dtsi | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/arch/arm64/boot/dts/qcom/sm8450.dtsi b/arch/arm64/boot/dts/qcom/sm8450.dtsi
index 6c18cfca9a34..8f26cf9aad01 100644
--- a/arch/arm64/boot/dts/qcom/sm8450.dtsi
+++ b/arch/arm64/boot/dts/qcom/sm8450.dtsi
@@ -52,6 +52,7 @@ CPU0: cpu@0 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 0>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 0>;
 			L2_0: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -71,6 +72,7 @@ CPU1: cpu@100 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 0>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 0>;
 			L2_100: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -87,6 +89,7 @@ CPU2: cpu@200 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 0>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 0>;
 			L2_200: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -103,6 +106,7 @@ CPU3: cpu@300 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 0>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 0>;
 			L2_300: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -119,6 +123,7 @@ CPU4: cpu@400 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 1>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 1>;
 			L2_400: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -135,6 +140,7 @@ CPU5: cpu@500 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 1>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 1>;
 			L2_500: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -152,6 +158,7 @@ CPU6: cpu@600 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 1>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 1>;
 			L2_600: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -168,6 +175,7 @@ CPU7: cpu@700 {
 			power-domain-names = "psci";
 			qcom,freq-domain = <&cpufreq_hw 2>;
 			#cooling-cells = <2>;
+			clocks = <&cpufreq_hw 2>;
 			L2_700: l2-cache {
 			      compatible = "cache";
 			      next-level-cache = <&L3_0>;
@@ -3804,6 +3812,7 @@ cpufreq_hw: cpufreq@17d91000 {
 				     <GIC_SPI 19 IRQ_TYPE_LEVEL_HIGH>;
 			interrupt-names = "dcvsh-irq-0", "dcvsh-irq-1", "dcvsh-irq-2";
 			#freq-domain-cells = <1>;
+			#clock-cells = <1>;
 		};
 
 		gem_noc: interconnect@19100000 {
-- 
2.25.1

[PATCH v2 3/7] cpufreq: qcom-hw: Remove un-necessary cpumask_empty() check

[Manivannan Sadhasivam]

CPUFreq core will always set the "policy->cpus" bitmask with the bitfield
of the CPU that goes first per domain/policy. So there is no way the
"policy->cpus" bitmask will be empty during qcom_cpufreq_hw_cpu_init().

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/cpufreq/qcom-cpufreq-hw.c | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
index d5ef3c66c762..a5b3b8d0e164 100644
--- a/drivers/cpufreq/qcom-cpufreq-hw.c
+++ b/drivers/cpufreq/qcom-cpufreq-hw.c
@@ -552,11 +552,6 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 		data->per_core_dcvs = true;
 
 	qcom_get_related_cpus(index, policy->cpus);
-	if (cpumask_empty(policy->cpus)) {
-		dev_err(dev, "Domain-%d failed to get related CPUs\n", index);
-		ret = -ENOENT;
-		goto error;
-	}
 
 	policy->driver_data = data;
 	policy->dvfs_possible_from_any_cpu = true;
-- 
2.25.1

[PATCH v2 4/7] cpufreq: qcom-hw: Allocate qcom_cpufreq_data during probe

[Manivannan Sadhasivam]

qcom_cpufreq_data is allocated based on the number of frequency domains
defined in DT which is static and won't change during runtime. There is
no real reason to allocate it during the CPU init() callback and deallocate
it during exit(). Hence, move the allocation to probe() and use the
allocated memory during init().

This also allows us to use devm_platform_get_and_ioremap_resource() helper
for acquiring the freq-domain resources from DT.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/cpufreq/qcom-cpufreq-hw.c | 86 +++++++++++++------------------
 1 file changed, 37 insertions(+), 49 deletions(-)

diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
index a5b3b8d0e164..1842e9facaa1 100644
--- a/drivers/cpufreq/qcom-cpufreq-hw.c
+++ b/drivers/cpufreq/qcom-cpufreq-hw.c
@@ -58,6 +58,10 @@ struct qcom_cpufreq_data {
 	bool per_core_dcvs;
 };
 
+static struct {
+	struct qcom_cpufreq_data *data;
+} qcom_cpufreq;
+
 static unsigned long cpu_hw_rate, xo_rate;
 static bool icc_scaling_enabled;
 
@@ -489,8 +493,6 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 	struct of_phandle_args args;
 	struct device_node *cpu_np;
 	struct device *cpu_dev;
-	struct resource *res;
-	void __iomem *base;
 	struct qcom_cpufreq_data *data;
 	int ret, index;
 
@@ -512,43 +514,16 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 		return ret;
 
 	index = args.args[0];
-
-	res = platform_get_resource(pdev, IORESOURCE_MEM, index);
-	if (!res) {
-		dev_err(dev, "failed to get mem resource %d\n", index);
-		return -ENODEV;
-	}
-
-	if (!request_mem_region(res->start, resource_size(res), res->name)) {
-		dev_err(dev, "failed to request resource %pR\n", res);
-		return -EBUSY;
-	}
-
-	base = ioremap(res->start, resource_size(res));
-	if (!base) {
-		dev_err(dev, "failed to map resource %pR\n", res);
-		ret = -ENOMEM;
-		goto release_region;
-	}
-
-	data = kzalloc(sizeof(*data), GFP_KERNEL);
-	if (!data) {
-		ret = -ENOMEM;
-		goto unmap_base;
-	}
-
 	data->soc_data = of_device_get_match_data(&pdev->dev);
-	data->base = base;
-	data->res = res;
+	data = &qcom_cpufreq.data[index];
 
 	/* HW should be in enabled state to proceed */
-	if (!(readl_relaxed(base + data->soc_data->reg_enable) & 0x1)) {
+	if (!(readl_relaxed(data->base + data->soc_data->reg_enable) & 0x1)) {
 		dev_err(dev, "Domain-%d cpufreq hardware not enabled\n", index);
-		ret = -ENODEV;
-		goto error;
+		return -ENODEV;
 	}
 
-	if (readl_relaxed(base + data->soc_data->reg_dcvs_ctrl) & 0x1)
+	if (readl_relaxed(data->base + data->soc_data->reg_dcvs_ctrl) & 0x1)
 		data->per_core_dcvs = true;
 
 	qcom_get_related_cpus(index, policy->cpus);
@@ -559,14 +534,13 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 	ret = qcom_cpufreq_hw_read_lut(cpu_dev, policy);
 	if (ret) {
 		dev_err(dev, "Domain-%d failed to read LUT\n", index);
-		goto error;
+		return ret;
 	}
 
 	ret = dev_pm_opp_get_opp_count(cpu_dev);
 	if (ret <= 0) {
 		dev_err(cpu_dev, "Failed to add OPPs\n");
-		ret = -ENODEV;
-		goto error;
+		return -ENODEV;
 	}
 
 	if (policy_has_boost_freq(policy)) {
@@ -575,18 +549,7 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 			dev_warn(cpu_dev, "failed to enable boost: %d\n", ret);
 	}
 
-	ret = qcom_cpufreq_hw_lmh_init(policy, index);
-	if (ret)
-		goto error;
-
-	return 0;
-error:
-	kfree(data);
-unmap_base:
-	iounmap(base);
-release_region:
-	release_mem_region(res->start, resource_size(res));
-	return ret;
+	return qcom_cpufreq_hw_lmh_init(policy, index);
 }
 
 static int qcom_cpufreq_hw_cpu_exit(struct cpufreq_policy *policy)
@@ -643,7 +606,7 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 {
 	struct device *cpu_dev;
 	struct clk *clk;
-	int ret;
+	int ret, i, num_domains;
 
 	clk = clk_get(&pdev->dev, "xo");
 	if (IS_ERR(clk))
@@ -670,6 +633,31 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 	if (ret)
 		return ret;
 
+	/* Allocate qcom_cpufreq_data based on the available frequency domains in DT */
+	num_domains = of_property_count_elems_of_size(pdev->dev.of_node, "reg", sizeof(u32) * 4);
+	if (num_domains <= 0)
+		return num_domains;
+
+	qcom_cpufreq.data = devm_kzalloc(&pdev->dev, sizeof(struct qcom_cpufreq_data) * num_domains,
+					 GFP_KERNEL);
+	if (!qcom_cpufreq.data)
+		return -ENOMEM;
+
+	for (i = 0; i < num_domains; i++) {
+		struct qcom_cpufreq_data *data = &qcom_cpufreq.data[i];
+		struct resource *res;
+		void __iomem *base;
+
+		base = devm_platform_get_and_ioremap_resource(pdev, i, &res);
+		if (IS_ERR(base)) {
+			dev_err(&pdev->dev, "Failed to map resource %pR\n", res);
+			return PTR_ERR(base);
+		}
+
+		data->base = base;
+		data->res = res;
+	}
+
 	ret = cpufreq_register_driver(&cpufreq_qcom_hw_driver);
 	if (ret)
 		dev_err(&pdev->dev, "CPUFreq HW driver failed to register\n");
-- 
2.25.1

[PATCH v2 5/7] cpufreq: qcom-hw: Use cached dev pointer in probe()

[Manivannan Sadhasivam]

There are multiple instances of dev pointer used in the probe() function.
Instead of referencing pdev->dev all the time, let's use a cached dev
pointer to simplify the code.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/cpufreq/qcom-cpufreq-hw.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
index 1842e9facaa1..bc991ef10c05 100644
--- a/drivers/cpufreq/qcom-cpufreq-hw.c
+++ b/drivers/cpufreq/qcom-cpufreq-hw.c
@@ -604,18 +604,19 @@ static struct cpufreq_driver cpufreq_qcom_hw_driver = {
 
 static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 {
+	struct device *dev = &pdev->dev;
 	struct device *cpu_dev;
 	struct clk *clk;
 	int ret, i, num_domains;
 
-	clk = clk_get(&pdev->dev, "xo");
+	clk = clk_get(dev, "xo");
 	if (IS_ERR(clk))
 		return PTR_ERR(clk);
 
 	xo_rate = clk_get_rate(clk);
 	clk_put(clk);
 
-	clk = clk_get(&pdev->dev, "alternate");
+	clk = clk_get(dev, "alternate");
 	if (IS_ERR(clk))
 		return PTR_ERR(clk);
 
@@ -634,11 +635,11 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 		return ret;
 
 	/* Allocate qcom_cpufreq_data based on the available frequency domains in DT */
-	num_domains = of_property_count_elems_of_size(pdev->dev.of_node, "reg", sizeof(u32) * 4);
+	num_domains = of_property_count_elems_of_size(dev->of_node, "reg", sizeof(u32) * 4);
 	if (num_domains <= 0)
 		return num_domains;
 
-	qcom_cpufreq.data = devm_kzalloc(&pdev->dev, sizeof(struct qcom_cpufreq_data) * num_domains,
+	qcom_cpufreq.data = devm_kzalloc(dev, sizeof(struct qcom_cpufreq_data) * num_domains,
 					 GFP_KERNEL);
 	if (!qcom_cpufreq.data)
 		return -ENOMEM;
@@ -650,7 +651,7 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 
 		base = devm_platform_get_and_ioremap_resource(pdev, i, &res);
 		if (IS_ERR(base)) {
-			dev_err(&pdev->dev, "Failed to map resource %pR\n", res);
+			dev_err(dev, "Failed to map resource %pR\n", res);
 			return PTR_ERR(base);
 		}
 
@@ -660,9 +661,9 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 
 	ret = cpufreq_register_driver(&cpufreq_qcom_hw_driver);
 	if (ret)
-		dev_err(&pdev->dev, "CPUFreq HW driver failed to register\n");
+		dev_err(dev, "CPUFreq HW driver failed to register\n");
 	else
-		dev_dbg(&pdev->dev, "QCOM CPUFreq HW driver initialized\n");
+		dev_dbg(dev, "QCOM CPUFreq HW driver initialized\n");
 
 	return ret;
 }
-- 
2.25.1

[PATCH v2 6/7] cpufreq: qcom-hw: Move soc_data to struct qcom_cpufreq

[Manivannan Sadhasivam]

soc_data is a static info of the driver and thus no need to cache it inside
the qcom_cpufreq_data struct which is allocated per frequency domain. So,
move it inside qcom_cpufreq struct.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/cpufreq/qcom-cpufreq-hw.c | 27 ++++++++++++++-------------
 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
index bc991ef10c05..76f840636828 100644
--- a/drivers/cpufreq/qcom-cpufreq-hw.c
+++ b/drivers/cpufreq/qcom-cpufreq-hw.c
@@ -42,7 +42,6 @@ struct qcom_cpufreq_soc_data {
 struct qcom_cpufreq_data {
 	void __iomem *base;
 	struct resource *res;
-	const struct qcom_cpufreq_soc_data *soc_data;
 
 	/*
 	 * Mutex to synchronize between de-init sequence and re-starting LMh
@@ -60,6 +59,7 @@ struct qcom_cpufreq_data {
 
 static struct {
 	struct qcom_cpufreq_data *data;
+	const struct qcom_cpufreq_soc_data *soc_data;
 } qcom_cpufreq;
 
 static unsigned long cpu_hw_rate, xo_rate;
@@ -110,7 +110,7 @@ static int qcom_cpufreq_hw_target_index(struct cpufreq_policy *policy,
 					unsigned int index)
 {
 	struct qcom_cpufreq_data *data = policy->driver_data;
-	const struct qcom_cpufreq_soc_data *soc_data = data->soc_data;
+	const struct qcom_cpufreq_soc_data *soc_data = qcom_cpufreq.soc_data;
 	unsigned long freq = policy->freq_table[index].frequency;
 	unsigned int i;
 
@@ -138,7 +138,7 @@ static unsigned int qcom_cpufreq_hw_get(unsigned int cpu)
 		return 0;
 
 	data = policy->driver_data;
-	soc_data = data->soc_data;
+	soc_data = qcom_cpufreq.soc_data;
 
 	index = readl_relaxed(data->base + soc_data->reg_perf_state);
 	index = min(index, LUT_MAX_ENTRIES - 1);
@@ -150,7 +150,7 @@ static unsigned int qcom_cpufreq_hw_fast_switch(struct cpufreq_policy *policy,
 						unsigned int target_freq)
 {
 	struct qcom_cpufreq_data *data = policy->driver_data;
-	const struct qcom_cpufreq_soc_data *soc_data = data->soc_data;
+	const struct qcom_cpufreq_soc_data *soc_data = qcom_cpufreq.soc_data;
 	unsigned int index;
 	unsigned int i;
 
@@ -174,7 +174,7 @@ static int qcom_cpufreq_hw_read_lut(struct device *cpu_dev,
 	unsigned long rate;
 	int ret;
 	struct qcom_cpufreq_data *drv_data = policy->driver_data;
-	const struct qcom_cpufreq_soc_data *soc_data = drv_data->soc_data;
+	const struct qcom_cpufreq_soc_data *soc_data = qcom_cpufreq.soc_data;
 
 	table = kcalloc(LUT_MAX_ENTRIES + 1, sizeof(*table), GFP_KERNEL);
 	if (!table)
@@ -291,10 +291,10 @@ static unsigned long qcom_lmh_get_throttle_freq(struct qcom_cpufreq_data *data)
 {
 	unsigned int lval;
 
-	if (data->soc_data->reg_current_vote)
-		lval = readl_relaxed(data->base + data->soc_data->reg_current_vote) & 0x3ff;
+	if (qcom_cpufreq.soc_data->reg_current_vote)
+		lval = readl_relaxed(data->base + qcom_cpufreq.soc_data->reg_current_vote) & 0x3ff;
 	else
-		lval = readl_relaxed(data->base + data->soc_data->reg_domain_state) & 0xff;
+		lval = readl_relaxed(data->base + qcom_cpufreq.soc_data->reg_domain_state) & 0xff;
 
 	return lval * xo_rate;
 }
@@ -366,9 +366,9 @@ static irqreturn_t qcom_lmh_dcvs_handle_irq(int irq, void *data)
 	disable_irq_nosync(c_data->throttle_irq);
 	schedule_delayed_work(&c_data->throttle_work, 0);
 
-	if (c_data->soc_data->reg_intr_clr)
+	if (qcom_cpufreq.soc_data->reg_intr_clr)
 		writel_relaxed(GT_IRQ_STATUS,
-			       c_data->base + c_data->soc_data->reg_intr_clr);
+			       c_data->base + qcom_cpufreq.soc_data->reg_intr_clr);
 
 	return IRQ_HANDLED;
 }
@@ -514,16 +514,15 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
 		return ret;
 
 	index = args.args[0];
-	data->soc_data = of_device_get_match_data(&pdev->dev);
 	data = &qcom_cpufreq.data[index];
 
 	/* HW should be in enabled state to proceed */
-	if (!(readl_relaxed(data->base + data->soc_data->reg_enable) & 0x1)) {
+	if (!(readl_relaxed(data->base + qcom_cpufreq.soc_data->reg_enable) & 0x1)) {
 		dev_err(dev, "Domain-%d cpufreq hardware not enabled\n", index);
 		return -ENODEV;
 	}
 
-	if (readl_relaxed(data->base + data->soc_data->reg_dcvs_ctrl) & 0x1)
+	if (readl_relaxed(data->base + qcom_cpufreq.soc_data->reg_dcvs_ctrl) & 0x1)
 		data->per_core_dcvs = true;
 
 	qcom_get_related_cpus(index, policy->cpus);
@@ -644,6 +643,8 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 	if (!qcom_cpufreq.data)
 		return -ENOMEM;
 
+	qcom_cpufreq.soc_data = of_device_get_match_data(dev);
+
 	for (i = 0; i < num_domains; i++) {
 		struct qcom_cpufreq_data *data = &qcom_cpufreq.data[i];
 		struct resource *res;
-- 
2.25.1

[PATCH v2 7/7] cpufreq: qcom-hw: Add CPU clock provider support

[Manivannan Sadhasivam]

Qcom CPUFreq hardware (EPSS/OSM) controls clock and voltage to the CPU
cores. But this relationship is not represented with the clk framework
so far.

So, let's make the qcom-cpufreq-hw driver a clock provider. This makes the
clock producer/consumer relationship cleaner and is also useful for CPU
related frameworks like OPP to know the frequency at which the CPUs are
running.

The clock frequency provided by the driver is for each frequency domain.
We cannot get the frequency of each CPU core because, not all platforms
support per-core DCVS feature.

Also the frequency supplied by the driver is the actual frequency that
comes out of the EPSS/OSM block after the DCVS operation. This frequency is
not same as what the CPUFreq framework has set but it is the one that gets
supplied to the CPUs after throttling by LMh.

Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
---
 drivers/cpufreq/qcom-cpufreq-hw.c | 43 +++++++++++++++++++++++++++++++
 1 file changed, 43 insertions(+)

diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
index 76f840636828..66677db3e267 100644
--- a/drivers/cpufreq/qcom-cpufreq-hw.c
+++ b/drivers/cpufreq/qcom-cpufreq-hw.c
@@ -4,6 +4,7 @@
  */
 
 #include <linux/bitfield.h>
+#include <linux/clk-provider.h>
 #include <linux/cpufreq.h>
 #include <linux/init.h>
 #include <linux/interconnect.h>
@@ -53,6 +54,7 @@ struct qcom_cpufreq_data {
 	bool cancel_throttle;
 	struct delayed_work throttle_work;
 	struct cpufreq_policy *policy;
+	struct clk_hw cpu_clk;
 
 	bool per_core_dcvs;
 };
@@ -601,8 +603,20 @@ static struct cpufreq_driver cpufreq_qcom_hw_driver = {
 	.ready		= qcom_cpufreq_ready,
 };
 
+static unsigned long qcom_cpufreq_hw_recalc_rate(struct clk_hw *hw, unsigned long parent_rate)
+{
+	struct qcom_cpufreq_data *data = container_of(hw, struct qcom_cpufreq_data, cpu_clk);
+
+	return qcom_lmh_get_throttle_freq(data) / HZ_PER_KHZ;
+}
+
+static const struct clk_ops qcom_cpufreq_hw_clk_ops = {
+	.recalc_rate = qcom_cpufreq_hw_recalc_rate,
+};
+
 static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 {
+	struct clk_hw_onecell_data *clk_data;
 	struct device *dev = &pdev->dev;
 	struct device *cpu_dev;
 	struct clk *clk;
@@ -645,8 +659,16 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 
 	qcom_cpufreq.soc_data = of_device_get_match_data(dev);
 
+	clk_data = devm_kzalloc(dev, struct_size(clk_data, hws, num_domains), GFP_KERNEL);
+	if (!clk_data)
+		return -ENOMEM;
+
+	clk_data->num = num_domains;
+
 	for (i = 0; i < num_domains; i++) {
 		struct qcom_cpufreq_data *data = &qcom_cpufreq.data[i];
+		static struct clk_init_data init = {};
+		const char *clk_name;
 		struct resource *res;
 		void __iomem *base;
 
@@ -658,6 +680,27 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 
 		data->base = base;
 		data->res = res;
+
+		/* Register CPU clock for each frequency domain */
+		clk_name = devm_kasprintf(dev, GFP_KERNEL, "qcom_cpufreq%d", i);
+		init.name = clk_name;
+		init.flags = CLK_GET_RATE_NOCACHE;
+		init.ops = &qcom_cpufreq_hw_clk_ops;
+		data->cpu_clk.init = &init;
+
+		ret = devm_clk_hw_register(dev, &data->cpu_clk);
+		if (ret < 0) {
+			dev_err(dev, "Failed to register Qcom CPUFreq clock\n");
+			return ret;
+		}
+
+		clk_data->hws[i] = &data->cpu_clk;
+	}
+
+	ret = devm_of_clk_add_hw_provider(dev, of_clk_hw_onecell_get, clk_data);
+	if (ret < 0) {
+		dev_err(dev, "Failed to add Qcom CPUFreq clock provider\n");
+		return ret;
 	}
 
 	ret = cpufreq_register_driver(&cpufreq_qcom_hw_driver);
-- 
2.25.1

Re: [PATCH v2 7/7] cpufreq: qcom-hw: Add CPU clock provider support

[Viresh Kumar]

On 25-10-22, 13:02, Manivannan Sadhasivam wrote:
>  	for (i = 0; i < num_domains; i++) {
>  		struct qcom_cpufreq_data *data = &qcom_cpufreq.data[i];
> +		static struct clk_init_data init = {};

Why not make this part of a driver level structure instead of static ?
And what will {} do ? Initialize to 0 ? static would have already done
that.

-- 
viresh

Re: [PATCH v2 3/7] cpufreq: qcom-hw: Remove un-necessary cpumask_empty() check

[Viresh Kumar]

On 25-10-22, 13:02, Manivannan Sadhasivam wrote:
> CPUFreq core will always set the "policy->cpus" bitmask with the bitfield
> of the CPU that goes first per domain/policy. So there is no way the
> "policy->cpus" bitmask will be empty during qcom_cpufreq_hw_cpu_init().
> 
> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
> ---
>  drivers/cpufreq/qcom-cpufreq-hw.c | 5 -----
>  1 file changed, 5 deletions(-)
> 
> diff --git a/drivers/cpufreq/qcom-cpufreq-hw.c b/drivers/cpufreq/qcom-cpufreq-hw.c
> index d5ef3c66c762..a5b3b8d0e164 100644
> --- a/drivers/cpufreq/qcom-cpufreq-hw.c
> +++ b/drivers/cpufreq/qcom-cpufreq-hw.c
> @@ -552,11 +552,6 @@ static int qcom_cpufreq_hw_cpu_init(struct cpufreq_policy *policy)
>  		data->per_core_dcvs = true;
>  
>  	qcom_get_related_cpus(index, policy->cpus);
> -	if (cpumask_empty(policy->cpus)) {
> -		dev_err(dev, "Domain-%d failed to get related CPUs\n", index);
> -		ret = -ENOENT;
> -		goto error;
> -	}
>  
>  	policy->driver_data = data;
>  	policy->dvfs_possible_from_any_cpu = true;

Applied. Thanks.

I tried applying 4-6 as well, but git am failed. You can send such
cleanups separately, so they don't need to wait for others to reviews.

-- 
viresh

Re: [PATCH v2 7/7] cpufreq: qcom-hw: Add CPU clock provider support

[Manivannan Sadhasivam]

On Wed, Nov 02, 2022 at 09:58:49AM +0530, Viresh Kumar wrote:
> On 25-10-22, 13:02, Manivannan Sadhasivam wrote:
> >  	for (i = 0; i < num_domains; i++) {
> >  		struct qcom_cpufreq_data *data = &qcom_cpufreq.data[i];
> > +		static struct clk_init_data init = {};
> 
> Why not make this part of a driver level structure instead of static ?
> And what will {} do ? Initialize to 0 ? static would have already done
> that.
> 

Good catch! "static" is not needed here.

Thanks,
Mani

> -- 
> viresh

-- 
மணிவண்ணன் சதாசிவம்