Merge tag 'pm-6.5-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm

Pull more power management updates from Rafael Wysocki:
 "These add support for new hardware (ap807 and AM62A7), fix several
  issues in cpufreq drivers and in the operating performance points
  (OPP) framework, fix up intel_idle after recent changes and add
  documentation.

  Specifics:

   - Add missing __init annotation to one function in the intel_idle
     drvier (Rafael Wysocki)

   - Make intel_pstate use a correct scaling factor when mapping HWP
     performance levels to frequency values on hybrid-capable systems
     with disabled E-cores (Srinivas Pandruvada)

   - Fix Kconfig dependencies of the cpufreq-dt-platform driver (Viresh
     Kumar)

   - Add support to build cpufreq-dt-platdev as a module (Zhipeng Wang)

   - Don't allocate Sparc's cpufreq_driver dynamically (Viresh Kumar)

   - Add support for TI's AM62A7 platform (Vibhore Vardhan)

   - Add support for Armada's ap807 platform (Russell King (Oracle))

   - Add support for StarFive JH7110 SoC (Mason Huo)

   - Fix voltage selection for Mediatek Socs (Daniel Golle)

   - Fix error handling in Tegra's cpufreq driver (Christophe JAILLET)

   - Document Qualcomm's IPQ8074 in DT bindings (Robert Marko)

   - Don't warn for disabling a non-existing frequency for imx6q cpufreq
     driver (Christoph Niedermaier)

   - Use dev_err_probe() in Qualcomm's cpufreq driver (Andrew Halaney)

   - Simplify performance state related logic in the OPP core (Viresh
     Kumar)

   - Fix use-after-free and improve locking around lazy_opp_tables
     (Viresh Kumar, Stephan Gerhold)

   - Minor cleanups - using dev_err_probe() and rate-limiting debug
     messages (Andrew Halaney, Adrián Larumbe)"

* tag 'pm-6.5-rc1-2' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: (23 commits)
  cpufreq: intel_pstate: Fix scaling for hybrid-capable systems with disabled E-cores
  cpufreq: Make CONFIG_CPUFREQ_DT_PLATDEV depend on OF
  intel_idle: Add __init annotation to matchup_vm_state_with_baremetal()
  OPP: Properly propagate error along when failing to get icc_path
  OPP: Use dev_err_probe() when failing to get icc_path
  cpufreq: qcom-cpufreq-hw: Use dev_err_probe() when failing to get icc paths
  cpufreq: mediatek: correct voltages for MT7622 and MT7623
  cpufreq: armada-8k: add ap807 support
  OPP: Simplify the over-designed pstate <-> level dance
  OPP: pstate is only valid for genpd OPP tables
  OPP: don't drop performance constraint on OPP table removal
  OPP: Protect `lazy_opp_tables` list with `opp_table_lock`
  OPP: Staticize `lazy_opp_tables` in of.c
  cpufreq: dt-platdev: Support building as module
  opp: Fix use-after-free in lazy_opp_tables after probe deferral
  dt-bindings: cpufreq: qcom-cpufreq-nvmem: document IPQ8074
  cpufreq: dt-platdev: Blacklist ti,am62a7 SoC
  cpufreq: ti-cpufreq: Add support for AM62A7
  OPP: rate-limit debug messages when no change in OPP is required
  cpufreq: imx6q: don't warn for disabling a non-existing frequency
  ...

Documentation/devicetree/bindings/cpufreq/qcom-cpufreq-nvmem.yaml

@@ -28,6 +28,7 @@ select:
           - qcom,apq8064
           - qcom,apq8096
           - qcom,ipq8064
+          - qcom,ipq8074
           - qcom,msm8939
           - qcom,msm8960
           - qcom,msm8974

drivers/cpufreq/Kconfig

@@ -218,7 +218,8 @@ config CPUFREQ_DT
 	  If in doubt, say N.
 
 config CPUFREQ_DT_PLATDEV
-	bool
+	tristate "Generic DT based cpufreq platdev driver"
+	depends on OF
 	help
 	  This adds a generic DT based cpufreq platdev driver for frequency
 	  management.  This creates a 'cpufreq-dt' platform device, on the

drivers/cpufreq/armada-8k-cpufreq.c

@@ -21,6 +21,13 @@
 #include <linux/pm_opp.h>
 #include <linux/slab.h>
 
+static const struct of_device_id __maybe_unused armada_8k_cpufreq_of_match[] = {
+	{ .compatible = "marvell,ap806-cpu-clock" },
+	{ .compatible = "marvell,ap807-cpu-clock" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, armada_8k_cpufreq_of_match);
+
 /*
  * Setup the opps list with the divider for the max frequency, that
  * will be filled at runtime.
@@ -127,7 +134,8 @@ static int __init armada_8k_cpufreq_init(void)
 	struct device_node *node;
 	struct cpumask cpus;
 
-	node = of_find_compatible_node(NULL, NULL, "marvell,ap806-cpu-clock");
+	node = of_find_matching_node_and_match(NULL, armada_8k_cpufreq_of_match,
+					       NULL);
 	if (!node || !of_device_is_available(node)) {
 		of_node_put(node);
 		return -ENODEV;
@@ -204,12 +212,6 @@ static void __exit armada_8k_cpufreq_exit(void)
 }
 module_exit(armada_8k_cpufreq_exit);
 
-static const struct of_device_id __maybe_unused armada_8k_cpufreq_of_match[] = {
-	{ .compatible = "marvell,ap806-cpu-clock" },
-	{ },
-};
-MODULE_DEVICE_TABLE(of, armada_8k_cpufreq_of_match);
-
 MODULE_AUTHOR("Gregory Clement <gregory.clement@bootlin.com>");
 MODULE_DESCRIPTION("Armada 8K cpufreq driver");
 MODULE_LICENSE("GPL");

drivers/cpufreq/cpufreq-dt-platdev.c

@@ -5,6 +5,7 @@
  */
 
 #include <linux/err.h>
+#include <linux/module.h>
 #include <linux/of.h>
 #include <linux/platform_device.h>
 
@@ -85,6 +86,8 @@ static const struct of_device_id allowlist[] __initconst = {
 	{ .compatible = "st-ericsson,u9500", },
 	{ .compatible = "st-ericsson,u9540", },
 
+	{ .compatible = "starfive,jh7110", },
+
 	{ .compatible = "ti,omap2", },
 	{ .compatible = "ti,omap4", },
 	{ .compatible = "ti,omap5", },
@@ -165,6 +168,7 @@ static const struct of_device_id blocklist[] __initconst = {
 	{ .compatible = "ti,dra7", },
 	{ .compatible = "ti,omap3", },
 	{ .compatible = "ti,am625", },
+	{ .compatible = "ti,am62a7", },
 
 	{ .compatible = "qcom,ipq8064", },
 	{ .compatible = "qcom,apq8064", },
@@ -214,3 +218,4 @@ static int __init cpufreq_dt_platdev_init(void)
 			       sizeof(struct cpufreq_dt_platform_data)));
 }
 core_initcall(cpufreq_dt_platdev_init);
+MODULE_LICENSE("GPL");

drivers/cpufreq/imx6q-cpufreq.c

@@ -209,6 +209,14 @@ static struct cpufreq_driver imx6q_cpufreq_driver = {
 	.suspend = cpufreq_generic_suspend,
 };
 
+static void imx6x_disable_freq_in_opp(struct device *dev, unsigned long freq)
+{
+	int ret = dev_pm_opp_disable(dev, freq);
+
+	if (ret < 0 && ret != -ENODEV)
+		dev_warn(dev, "failed to disable %ldMHz OPP\n", freq / 1000000);
+}
+
 #define OCOTP_CFG3			0x440
 #define OCOTP_CFG3_SPEED_SHIFT		16
 #define OCOTP_CFG3_SPEED_1P2GHZ		0x3
@@ -254,17 +262,15 @@ static int imx6q_opp_check_speed_grading(struct device *dev)
 	val &= 0x3;
 
 	if (val < OCOTP_CFG3_SPEED_996MHZ)
-		if (dev_pm_opp_disable(dev, 996000000))
-			dev_warn(dev, "failed to disable 996MHz OPP\n");
+		imx6x_disable_freq_in_opp(dev, 996000000);
 
 	if (of_machine_is_compatible("fsl,imx6q") ||
 	    of_machine_is_compatible("fsl,imx6qp")) {
 		if (val != OCOTP_CFG3_SPEED_852MHZ)
-			if (dev_pm_opp_disable(dev, 852000000))
-				dev_warn(dev, "failed to disable 852MHz OPP\n");
+			imx6x_disable_freq_in_opp(dev, 852000000);
+
 		if (val != OCOTP_CFG3_SPEED_1P2GHZ)
-			if (dev_pm_opp_disable(dev, 1200000000))
-				dev_warn(dev, "failed to disable 1.2GHz OPP\n");
+			imx6x_disable_freq_in_opp(dev, 1200000000);
 	}
 
 	return 0;
@@ -316,20 +322,16 @@ static int imx6ul_opp_check_speed_grading(struct device *dev)
 	val >>= OCOTP_CFG3_SPEED_SHIFT;
 	val &= 0x3;
 
-	if (of_machine_is_compatible("fsl,imx6ul")) {
+	if (of_machine_is_compatible("fsl,imx6ul"))
 		if (val != OCOTP_CFG3_6UL_SPEED_696MHZ)
-			if (dev_pm_opp_disable(dev, 696000000))
-				dev_warn(dev, "failed to disable 696MHz OPP\n");
-	}
+			imx6x_disable_freq_in_opp(dev, 696000000);
 
 	if (of_machine_is_compatible("fsl,imx6ull")) {
 		if (val != OCOTP_CFG3_6ULL_SPEED_792MHZ)
-			if (dev_pm_opp_disable(dev, 792000000))
-				dev_warn(dev, "failed to disable 792MHz OPP\n");
+			imx6x_disable_freq_in_opp(dev, 792000000);
 
 		if (val != OCOTP_CFG3_6ULL_SPEED_900MHZ)
-			if (dev_pm_opp_disable(dev, 900000000))
-				dev_warn(dev, "failed to disable 900MHz OPP\n");
+			imx6x_disable_freq_in_opp(dev, 900000000);
 	}
 
 	return ret;

drivers/cpufreq/intel_pstate.c

@@ -302,6 +302,13 @@ static bool hwp_forced __read_mostly;
 
 static struct cpufreq_driver *intel_pstate_driver __read_mostly;
 
+#define HYBRID_SCALING_FACTOR	78741
+
+static inline int core_get_scaling(void)
+{
+	return 100000;
+}
+
 #ifdef CONFIG_ACPI
 static bool acpi_ppc;
 #endif
@@ -400,6 +407,26 @@ static int intel_pstate_get_cppc_guaranteed(int cpu)
 
 	return cppc_perf.nominal_perf;
 }
+
+static int intel_pstate_cppc_get_scaling(int cpu)
+{
+	struct cppc_perf_caps cppc_perf;
+	int ret;
+
+	ret = cppc_get_perf_caps(cpu, &cppc_perf);
+
+	/*
+	 * If the nominal frequency and the nominal performance are not
+	 * zero and the ratio between them is not 100, return the hybrid
+	 * scaling factor.
+	 */
+	if (!ret && cppc_perf.nominal_perf && cppc_perf.nominal_freq &&
+	    cppc_perf.nominal_perf * 100 != cppc_perf.nominal_freq)
+		return HYBRID_SCALING_FACTOR;
+
+	return core_get_scaling();
+}
+
 #else /* CONFIG_ACPI_CPPC_LIB */
 static inline void intel_pstate_set_itmt_prio(int cpu)
 {
@@ -492,6 +519,11 @@ static inline int intel_pstate_get_cppc_guaranteed(int cpu)
 {
 	return -ENOTSUPP;
 }
+
+static int intel_pstate_cppc_get_scaling(int cpu)
+{
+	return core_get_scaling();
+}
 #endif /* CONFIG_ACPI_CPPC_LIB */
 
 /**
@@ -1897,11 +1929,6 @@ static int core_get_turbo_pstate(int cpu)
 	return ret;
 }
 
-static inline int core_get_scaling(void)
-{
-	return 100000;
-}
-
 static u64 core_get_val(struct cpudata *cpudata, int pstate)
 {
 	u64 val;
@@ -1938,16 +1965,28 @@ static void hybrid_get_type(void *data)
 	*cpu_type = get_this_hybrid_cpu_type();
 }
 
-static int hybrid_get_cpu_scaling(int cpu)
+static int hwp_get_cpu_scaling(int cpu)
 {
 	u8 cpu_type = 0;
 
 	smp_call_function_single(cpu, hybrid_get_type, &cpu_type, 1);
 	/* P-cores have a smaller perf level-to-freqency scaling factor. */
 	if (cpu_type == 0x40)
-		return 78741;
+		return HYBRID_SCALING_FACTOR;
 
-	return core_get_scaling();
+	/* Use default core scaling for E-cores */
+	if (cpu_type == 0x20)
+		return core_get_scaling();
+
+	/*
+	 * If reached here, this system is either non-hybrid (like Tiger
+	 * Lake) or hybrid-capable (like Alder Lake or Raptor Lake) with
+	 * no E cores (in which case CPUID for hybrid support is 0).
+	 *
+	 * The CPPC nominal_frequency field is 0 for non-hybrid systems,
+	 * so the default core scaling will be used for them.
+	 */
+	return intel_pstate_cppc_get_scaling(cpu);
 }
 
 static void intel_pstate_set_pstate(struct cpudata *cpu, int pstate)
@@ -3395,8 +3434,7 @@ static int __init intel_pstate_init(void)
 			if (!default_driver)
 				default_driver = &intel_pstate;
 
-			if (boot_cpu_has(X86_FEATURE_HYBRID_CPU))
-				pstate_funcs.get_cpu_scaling = hybrid_get_cpu_scaling;
+			pstate_funcs.get_cpu_scaling = hwp_get_cpu_scaling;
 
 			goto hwp_cpu_matched;
 		}

drivers/cpufreq/mediatek-cpufreq.c

@@ -696,9 +696,16 @@ static const struct mtk_cpufreq_platform_data mt2701_platform_data = {
 static const struct mtk_cpufreq_platform_data mt7622_platform_data = {
 	.min_volt_shift = 100000,
 	.max_volt_shift = 200000,
-	.proc_max_volt = 1360000,
+	.proc_max_volt = 1350000,
 	.sram_min_volt = 0,
-	.sram_max_volt = 1360000,
+	.sram_max_volt = 1350000,
+	.ccifreq_supported = false,
+};
+
+static const struct mtk_cpufreq_platform_data mt7623_platform_data = {
+	.min_volt_shift = 100000,
+	.max_volt_shift = 200000,
+	.proc_max_volt = 1300000,
 	.ccifreq_supported = false,
 };
 
@@ -734,7 +741,7 @@ static const struct of_device_id mtk_cpufreq_machines[] __initconst = {
 	{ .compatible = "mediatek,mt2701", .data = &mt2701_platform_data },
 	{ .compatible = "mediatek,mt2712", .data = &mt2701_platform_data },
 	{ .compatible = "mediatek,mt7622", .data = &mt7622_platform_data },
-	{ .compatible = "mediatek,mt7623", .data = &mt7622_platform_data },
+	{ .compatible = "mediatek,mt7623", .data = &mt7623_platform_data },
 	{ .compatible = "mediatek,mt8167", .data = &mt8516_platform_data },
 	{ .compatible = "mediatek,mt817x", .data = &mt2701_platform_data },
 	{ .compatible = "mediatek,mt8173", .data = &mt2701_platform_data },

drivers/cpufreq/qcom-cpufreq-hw.c

@@ -661,7 +661,7 @@ static int qcom_cpufreq_hw_driver_probe(struct platform_device *pdev)
 
 	ret = dev_pm_opp_of_find_icc_paths(cpu_dev, NULL);
 	if (ret)
-		return ret;
+		return dev_err_probe(dev, ret, "Failed to find icc paths\n");
 
 	for (num_domains = 0; num_domains < MAX_FREQ_DOMAINS; num_domains++)
 		if (!platform_get_resource(pdev, IORESOURCE_MEM, num_domains))

drivers/cpufreq/sparc-us2e-cpufreq.c

@@ -20,8 +20,6 @@
 #include <asm/asi.h>
 #include <asm/timer.h>
 
-static struct cpufreq_driver *cpufreq_us2e_driver;
-
 struct us2e_freq_percpu_info {
 	struct cpufreq_frequency_table table[6];
 };
@@ -300,12 +298,19 @@ static int __init us2e_freq_cpu_init(struct cpufreq_policy *policy)
 
 static int us2e_freq_cpu_exit(struct cpufreq_policy *policy)
 {
-	if (cpufreq_us2e_driver)
-		us2e_freq_target(policy, 0);
-
+	us2e_freq_target(policy, 0);
 	return 0;
 }
 
+static struct cpufreq_driver cpufreq_us2e_driver = {
+	.name = "UltraSPARC-IIe",
+	.init = us2e_freq_cpu_init,
+	.verify = cpufreq_generic_frequency_table_verify,
+	.target_index = us2e_freq_target,
+	.get = us2e_freq_get,
+	.exit = us2e_freq_cpu_exit,
+};
+
 static int __init us2e_freq_init(void)
 {
 	unsigned long manuf, impl, ver;
@@ -319,39 +324,15 @@ static int __init us2e_freq_init(void)
 	impl  = ((ver >> 32) & 0xffff);
 
 	if (manuf == 0x17 && impl == 0x13) {
-		struct cpufreq_driver *driver;
-
-		ret = -ENOMEM;
-		driver = kzalloc(sizeof(*driver), GFP_KERNEL);
-		if (!driver)
-			goto err_out;
-
-		us2e_freq_table = kzalloc((NR_CPUS * sizeof(*us2e_freq_table)),
-			GFP_KERNEL);
+		us2e_freq_table = kzalloc(NR_CPUS * sizeof(*us2e_freq_table),
+					  GFP_KERNEL);
 		if (!us2e_freq_table)
-			goto err_out;
-
-		driver->init = us2e_freq_cpu_init;
-		driver->verify = cpufreq_generic_frequency_table_verify;
-		driver->target_index = us2e_freq_target;
-		driver->get = us2e_freq_get;
-		driver->exit = us2e_freq_cpu_exit;
-		strcpy(driver->name, "UltraSPARC-IIe");
+			return -ENOMEM;
 
-		cpufreq_us2e_driver = driver;
-		ret = cpufreq_register_driver(driver);
+		ret = cpufreq_register_driver(&cpufreq_us2e_driver);
 		if (ret)
-			goto err_out;
+			kfree(us2e_freq_table);
 
-		return 0;
-
-err_out:
-		if (driver) {
-			kfree(driver);
-			cpufreq_us2e_driver = NULL;
-		}
-		kfree(us2e_freq_table);
-		us2e_freq_table = NULL;
 		return ret;
 	}
 
@@ -360,13 +341,8 @@ static int __init us2e_freq_init(void)
 
 static void __exit us2e_freq_exit(void)
 {
-	if (cpufreq_us2e_driver) {
-		cpufreq_unregister_driver(cpufreq_us2e_driver);
-		kfree(cpufreq_us2e_driver);
-		cpufreq_us2e_driver = NULL;
-		kfree(us2e_freq_table);
-		us2e_freq_table = NULL;
-	}
+	cpufreq_unregister_driver(&cpufreq_us2e_driver);
+	kfree(us2e_freq_table);
 }
 
 MODULE_AUTHOR("David S. Miller <davem@redhat.com>");

drivers/cpufreq/sparc-us3-cpufreq.c

@@ -19,8 +19,6 @@
 #include <asm/head.h>
 #include <asm/timer.h>
 
-static struct cpufreq_driver *cpufreq_us3_driver;
-
 struct us3_freq_percpu_info {
 	struct cpufreq_frequency_table table[4];
 };
@@ -144,12 +142,19 @@ static int __init us3_freq_cpu_init(struct cpufreq_policy *policy)
 
 static int us3_freq_cpu_exit(struct cpufreq_policy *policy)
 {
-	if (cpufreq_us3_driver)
-		us3_freq_target(policy, 0);
-
+	us3_freq_target(policy, 0);
 	return 0;
 }
 
+static struct cpufreq_driver cpufreq_us3_driver = {
+	.name = "UltraSPARC-III",
+	.init = us3_freq_cpu_init,
+	.verify = cpufreq_generic_frequency_table_verify,
+	.target_index = us3_freq_target,
+	.get = us3_freq_get,
+	.exit = us3_freq_cpu_exit,
+};
+
 static int __init us3_freq_init(void)
 {
 	unsigned long manuf, impl, ver;
@@ -167,39 +172,15 @@ static int __init us3_freq_init(void)
 	     impl == CHEETAH_PLUS_IMPL ||
 	     impl == JAGUAR_IMPL ||
 	     impl == PANTHER_IMPL)) {
-		struct cpufreq_driver *driver;
-
-		ret = -ENOMEM;
-		driver = kzalloc(sizeof(*driver), GFP_KERNEL);
-		if (!driver)
-			goto err_out;
-
-		us3_freq_table = kzalloc((NR_CPUS * sizeof(*us3_freq_table)),
-			GFP_KERNEL);
+		us3_freq_table = kzalloc(NR_CPUS * sizeof(*us3_freq_table),
+					 GFP_KERNEL);
 		if (!us3_freq_table)
-			goto err_out;
-
-		driver->init = us3_freq_cpu_init;
-		driver->verify = cpufreq_generic_frequency_table_verify;
-		driver->target_index = us3_freq_target;
-		driver->get = us3_freq_get;
-		driver->exit = us3_freq_cpu_exit;
-		strcpy(driver->name, "UltraSPARC-III");
+			return -ENOMEM;
 
-		cpufreq_us3_driver = driver;
-		ret = cpufreq_register_driver(driver);
+		ret = cpufreq_register_driver(&cpufreq_us3_driver);
 		if (ret)
-			goto err_out;
+			kfree(us3_freq_table);
 
-		return 0;
-
-err_out:
-		if (driver) {
-			kfree(driver);
-			cpufreq_us3_driver = NULL;
-		}
-		kfree(us3_freq_table);
-		us3_freq_table = NULL;
 		return ret;
 	}
 
@@ -208,13 +189,8 @@ static int __init us3_freq_init(void)
 
 static void __exit us3_freq_exit(void)
 {
-	if (cpufreq_us3_driver) {
-		cpufreq_unregister_driver(cpufreq_us3_driver);
-		kfree(cpufreq_us3_driver);
-		cpufreq_us3_driver = NULL;
-		kfree(us3_freq_table);
-		us3_freq_table = NULL;
-	}
+	cpufreq_unregister_driver(&cpufreq_us3_driver);
+	kfree(us3_freq_table);
 }
 
 MODULE_AUTHOR("David S. Miller <davem@redhat.com>");

drivers/cpufreq/tegra194-cpufreq.c

@@ -686,8 +686,10 @@ static int tegra194_cpufreq_probe(struct platform_device *pdev)
 
 	/* Check for optional OPPv2 and interconnect paths on CPU0 to enable ICC scaling */
 	cpu_dev = get_cpu_device(0);
-	if (!cpu_dev)
-		return -EPROBE_DEFER;
+	if (!cpu_dev) {
+		err = -EPROBE_DEFER;
+		goto err_free_res;
+	}
 
 	if (dev_pm_opp_of_get_opp_desc_node(cpu_dev)) {
 		err = dev_pm_opp_of_find_icc_paths(cpu_dev, NULL);

drivers/cpufreq/ti-cpufreq.c

@@ -337,6 +337,7 @@ static const struct of_device_id ti_cpufreq_of_match[] = {
 	{ .compatible = "ti,omap34xx", .data = &omap34xx_soc_data, },
 	{ .compatible = "ti,omap36xx", .data = &omap36xx_soc_data, },
 	{ .compatible = "ti,am625", .data = &am625_soc_data, },
+	{ .compatible = "ti,am62a7", .data = &am625_soc_data, },
 	/* legacy */
 	{ .compatible = "ti,omap3430", .data = &omap34xx_soc_data, },
 	{ .compatible = "ti,omap3630", .data = &omap36xx_soc_data, },

drivers/idle/intel_idle.c

@@ -2147,7 +2147,7 @@ static void __init intel_idle_cpuidle_devices_uninit(void)
  * All our short idle states are dominated by vmexit/vmenter latencies,
  * not the underlying hardware latencies so we keep our values for these.
  */
-static void matchup_vm_state_with_baremetal(void)
+static void __init matchup_vm_state_with_baremetal(void)
 {
 	int cstate;
 

drivers/opp/core.c

@@ -29,9 +29,6 @@
  */
 LIST_HEAD(opp_tables);
 
-/* OPP tables with uninitialized required OPPs */
-LIST_HEAD(lazy_opp_tables);
-
 /* Lock to allow exclusive modification to the device and opp lists */
 DEFINE_MUTEX(opp_table_lock);
 /* Flag indicating that opp_tables list is being updated at the moment */
@@ -230,17 +227,25 @@ EXPORT_SYMBOL_GPL(dev_pm_opp_get_level);
 unsigned int dev_pm_opp_get_required_pstate(struct dev_pm_opp *opp,
 					    unsigned int index)
 {
+	struct opp_table *opp_table = opp->opp_table;
+
 	if (IS_ERR_OR_NULL(opp) || !opp->available ||
-	    index >= opp->opp_table->required_opp_count) {
+	    index >= opp_table->required_opp_count) {
 		pr_err("%s: Invalid parameters\n", __func__);
 		return 0;
 	}
 
 	/* required-opps not fully initialized yet */
-	if (lazy_linking_pending(opp->opp_table))
+	if (lazy_linking_pending(opp_table))
 		return 0;
 
-	return opp->required_opps[index]->pstate;
+	/* The required OPP table must belong to a genpd */
+	if (unlikely(!opp_table->required_opp_tables[index]->is_genpd)) {
+		pr_err("%s: Performance state is only valid for genpds.\n", __func__);
+		return 0;
+	}
+
+	return opp->required_opps[index]->level;
 }
 EXPORT_SYMBOL_GPL(dev_pm_opp_get_required_pstate);
 
@@ -938,7 +943,7 @@ static int _set_opp_bw(const struct opp_table *opp_table,
 static int _set_performance_state(struct device *dev, struct device *pd_dev,
 				  struct dev_pm_opp *opp, int i)
 {
-	unsigned int pstate = likely(opp) ? opp->required_opps[i]->pstate : 0;
+	unsigned int pstate = likely(opp) ? opp->required_opps[i]->level: 0;
 	int ret;
 
 	if (!pd_dev)
@@ -1091,7 +1096,7 @@ static int _set_opp(struct device *dev, struct opp_table *opp_table,
 
 	/* Return early if nothing to do */
 	if (!forced && old_opp == opp && opp_table->enabled) {
-		dev_dbg(dev, "%s: OPPs are same, nothing to do\n", __func__);
+		dev_dbg_ratelimited(dev, "%s: OPPs are same, nothing to do\n", __func__);
 		return 0;
 	}
 
@@ -1358,7 +1363,10 @@ static struct opp_table *_allocate_opp_table(struct device *dev, int index)
 	return opp_table;
 
 remove_opp_dev:
+	_of_clear_opp_table(opp_table);
 	_remove_opp_dev(opp_dev, opp_table);
+	mutex_destroy(&opp_table->genpd_virt_dev_lock);
+	mutex_destroy(&opp_table->lock);
 err:
 	kfree(opp_table);
 	return ERR_PTR(ret);
@@ -1522,16 +1530,8 @@ static void _opp_table_kref_release(struct kref *kref)
 
 	WARN_ON(!list_empty(&opp_table->opp_list));
 
-	list_for_each_entry_safe(opp_dev, temp, &opp_table->dev_list, node) {
-		/*
-		 * The OPP table is getting removed, drop the performance state
-		 * constraints.
-		 */
-		if (opp_table->genpd_performance_state)
-			dev_pm_genpd_set_performance_state((struct device *)(opp_dev->dev), 0);
-
+	list_for_each_entry_safe(opp_dev, temp, &opp_table->dev_list, node)
 		_remove_opp_dev(opp_dev, opp_table);
-	}
 
 	mutex_destroy(&opp_table->genpd_virt_dev_lock);
 	mutex_destroy(&opp_table->lock);
@@ -2704,6 +2704,12 @@ int dev_pm_opp_xlate_performance_state(struct opp_table *src_table,
 	if (!src_table || !src_table->required_opp_count)
 		return pstate;
 
+	/* Both OPP tables must belong to genpds */
+	if (unlikely(!src_table->is_genpd || !dst_table->is_genpd)) {
+		pr_err("%s: Performance state is only valid for genpds.\n", __func__);
+		return -EINVAL;
+	}
+
 	/* required-opps not fully initialized yet */
 	if (lazy_linking_pending(src_table))
 		return -EBUSY;
@@ -2722,8 +2728,8 @@ int dev_pm_opp_xlate_performance_state(struct opp_table *src_table,
 	mutex_lock(&src_table->lock);
 
 	list_for_each_entry(opp, &src_table->opp_list, node) {
-		if (opp->pstate == pstate) {
-			dest_pstate = opp->required_opps[i]->pstate;
+		if (opp->level == pstate) {
+			dest_pstate = opp->required_opps[i]->level;
 			goto unlock;
 		}
 	}

drivers/opp/debugfs.c

@@ -152,7 +152,6 @@ void opp_debug_create_one(struct dev_pm_opp *opp, struct opp_table *opp_table)
 	debugfs_create_bool("dynamic", S_IRUGO, d, &opp->dynamic);
 	debugfs_create_bool("turbo", S_IRUGO, d, &opp->turbo);
 	debugfs_create_bool("suspend", S_IRUGO, d, &opp->suspend);
-	debugfs_create_u32("performance_state", S_IRUGO, d, &opp->pstate);
 	debugfs_create_u32("level", S_IRUGO, d, &opp->level);
 	debugfs_create_ulong("clock_latency_ns", S_IRUGO, d,
 			     &opp->clock_latency_ns);

drivers/opp/of.c

@@ -21,6 +21,9 @@
 
 #include "opp.h"
 
+/* OPP tables with uninitialized required OPPs, protected by opp_table_lock */
+static LIST_HEAD(lazy_opp_tables);
+
 /*
  * Returns opp descriptor node for a device node, caller must
  * do of_node_put().
@@ -145,7 +148,10 @@ static void _opp_table_free_required_tables(struct opp_table *opp_table)
 
 	opp_table->required_opp_count = 0;
 	opp_table->required_opp_tables = NULL;
+
+	mutex_lock(&opp_table_lock);
 	list_del(&opp_table->lazy);
+	mutex_unlock(&opp_table_lock);
 }
 
 /*
@@ -194,8 +200,15 @@ static void _opp_table_alloc_required_tables(struct opp_table *opp_table,
 	}
 
 	/* Let's do the linking later on */
-	if (lazy)
+	if (lazy) {
+		/*
+		 * The OPP table is not held while allocating the table, take it
+		 * now to avoid corruption to the lazy_opp_tables list.
+		 */
+		mutex_lock(&opp_table_lock);
 		list_add(&opp_table->lazy, &lazy_opp_tables);
+		mutex_unlock(&opp_table_lock);
+	}
 	else
 		_update_set_required_opps(opp_table);
 
@@ -500,11 +513,7 @@ int dev_pm_opp_of_find_icc_paths(struct device *dev,
 	for (i = 0; i < num_paths; i++) {
 		paths[i] = of_icc_get_by_index(dev, i);
 		if (IS_ERR(paths[i])) {
-			ret = PTR_ERR(paths[i]);
-			if (ret != -EPROBE_DEFER) {
-				dev_err(dev, "%s: Unable to get path%d: %d\n",
-					__func__, i, ret);
-			}
+			ret = dev_err_probe(dev, PTR_ERR(paths[i]), "%s: Unable to get path%d\n", __func__, i);
 			goto err;
 		}
 	}
@@ -932,9 +941,6 @@ static struct dev_pm_opp *_opp_add_static_v2(struct opp_table *opp_table,
 	if (ret)
 		goto free_required_opps;
 
-	if (opp_table->is_genpd)
-		new_opp->pstate = pm_genpd_opp_to_performance_state(dev, new_opp);
-
 	ret = _opp_add(dev, new_opp, opp_table);
 	if (ret) {
 		/* Don't return error for duplicate OPPs */
@@ -1021,14 +1027,6 @@ static int _of_add_opp_table_v2(struct device *dev, struct opp_table *opp_table)
 		goto remove_static_opp;
 	}
 
-	list_for_each_entry(opp, &opp_table->opp_list, node) {
-		/* Any non-zero performance state would enable the feature */
-		if (opp->pstate) {
-			opp_table->genpd_performance_state = true;
-			break;
-		}
-	}
-
 	lazy_link_required_opp_table(opp_table);
 
 	return 0;
@@ -1387,9 +1385,15 @@ int of_get_required_opp_performance_state(struct device_node *np, int index)
 		goto put_required_np;
 	}
 
+	/* The OPP tables must belong to a genpd */
+	if (unlikely(!opp_table->is_genpd)) {
+		pr_err("%s: Performance state is only valid for genpds.\n", __func__);
+		goto put_required_np;
+	}
+
 	opp = _find_opp_of_np(opp_table, required_np);
 	if (opp) {
-		pstate = opp->pstate;
+		pstate = opp->level;
 		dev_pm_opp_put(opp);
 	}
 

drivers/opp/opp.h

@@ -26,7 +26,7 @@ struct regulator;
 /* Lock to allow exclusive modification to the device and opp lists */
 extern struct mutex opp_table_lock;
 
-extern struct list_head opp_tables, lazy_opp_tables;
+extern struct list_head opp_tables;
 
 /* OPP Config flags */
 #define OPP_CONFIG_CLK			BIT(0)
@@ -78,7 +78,6 @@ struct opp_config_data {
  * @turbo:	true if turbo (boost) OPP
  * @suspend:	true if suspend OPP
  * @removed:	flag indicating that OPP's reference is dropped by OPP core.
- * @pstate: Device's power domain's performance state.
  * @rates:	Frequencies in hertz
  * @level:	Performance level
  * @supplies:	Power supplies voltage/current values
@@ -101,7 +100,6 @@ struct dev_pm_opp {
 	bool turbo;
 	bool suspend;
 	bool removed;
-	unsigned int pstate;
 	unsigned long *rates;
 	unsigned int level;
 
@@ -182,7 +180,6 @@ enum opp_table_access {
  * @paths: Interconnect path handles
  * @path_count: Number of interconnect paths
  * @enabled: Set to true if the device's resources are enabled/configured.
- * @genpd_performance_state: Device's power domain support performance state.
  * @is_genpd: Marks if the OPP table belongs to a genpd.
  * @set_required_opps: Helper responsible to set required OPPs.
  * @dentry:	debugfs dentry pointer of the real device directory (not links).
@@ -233,7 +230,6 @@ struct opp_table {
 	struct icc_path **paths;
 	unsigned int path_count;
 	bool enabled;
-	bool genpd_performance_state;
 	bool is_genpd;
 	int (*set_required_opps)(struct device *dev,
 		struct opp_table *opp_table, struct dev_pm_opp *opp, bool scaling_down);