cpuidle: Refactor and move out NVIDIA Tegra20 driver into drivers/cpuidle

The driver's code is refactored in a way that will make it easy to
support Tegra30/114/124 SoCs by this unified driver later on. The
current functionality is equal to the old Tegra20 driver, only the
code's structure changed a tad. This is also a proper platform driver
now.

Acked-by: Peter De Schrijver <pdeschrijver@nvidia.com>
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Thierry Reding <treding@nvidia.com>

3 files changed, 289 insertions, 0 deletions

diff --git a/drivers/cpuidle/Kconfig.arm b/drivers/cpuidle/Kconfig.arm
index 62272ecfa771..99a2d72ac02b 100644
--- a/drivers/cpuidle/Kconfig.arm
+++ b/drivers/cpuidle/Kconfig.arm
@@ -86,3 +86,11 @@ config ARM_MVEBU_V7_CPUIDLE
 	depends on (ARCH_MVEBU || COMPILE_TEST) && !ARM64
 	help
 	  Select this to enable cpuidle on Armada 370, 38x and XP processors.
+
+config ARM_TEGRA_CPUIDLE
+	bool "CPU Idle Driver for NVIDIA Tegra SoCs"
+	depends on ARCH_TEGRA && !ARM64
+	select ARCH_NEEDS_CPU_IDLE_COUPLED if SMP
+	select ARM_CPU_SUSPEND
+	help
+	  Select this to enable cpuidle for NVIDIA Tegra20/30/114/124 SoCs.
diff --git a/drivers/cpuidle/Makefile b/drivers/cpuidle/Makefile
index cc8c769d7fa9..55a464f6a78b 100644
--- a/drivers/cpuidle/Makefile
+++ b/drivers/cpuidle/Makefile
@@ -24,6 +24,7 @@ obj-$(CONFIG_ARM_CPUIDLE)		+= cpuidle-arm.o
 obj-$(CONFIG_ARM_PSCI_CPUIDLE)		+= cpuidle_psci.o
 cpuidle_psci-y				:= cpuidle-psci.o
 cpuidle_psci-$(CONFIG_PM_GENERIC_DOMAINS_OF) += cpuidle-psci-domain.o
+obj-$(CONFIG_ARM_TEGRA_CPUIDLE)		+= cpuidle-tegra.o
 
 ###############################################################################
 # MIPS drivers
diff --git a/drivers/cpuidle/cpuidle-tegra.c b/drivers/cpuidle/cpuidle-tegra.c
new file mode 100644
index 000000000000..5691bdcf11cb
--- /dev/null
+++ b/drivers/cpuidle/cpuidle-tegra.c
@@ -0,0 +1,280 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * CPU idle driver for Tegra CPUs
+ *
+ * Copyright (c) 2010-2013, NVIDIA Corporation.
+ * Copyright (c) 2011 Google, Inc.
+ * Author: Colin Cross <ccross@android.com>
+ *         Gary King <gking@nvidia.com>
+ *
+ * Rework for 3.3 by Peter De Schrijver <pdeschrijver@nvidia.com>
+ *
+ * Tegra20/124 driver unification by Dmitry Osipenko <digetx@gmail.com>
+ */
+
+#define pr_fmt(fmt)	"tegra-cpuidle: " fmt
+
+#include <linux/atomic.h>
+#include <linux/cpuidle.h>
+#include <linux/cpumask.h>
+#include <linux/cpu_pm.h>
+#include <linux/delay.h>
+#include <linux/errno.h>
+#include <linux/platform_device.h>
+#include <linux/types.h>
+
+#include <linux/clk/tegra.h>
+
+#include <soc/tegra/cpuidle.h>
+#include <soc/tegra/flowctrl.h>
+#include <soc/tegra/fuse.h>
+#include <soc/tegra/irq.h>
+#include <soc/tegra/pm.h>
+
+#include <asm/cpuidle.h>
+#include <asm/smp_plat.h>
+#include <asm/suspend.h>
+
+enum tegra_state {
+	TEGRA_C1,
+	TEGRA_CC6,
+	TEGRA_STATE_COUNT,
+};
+
+static atomic_t tegra_idle_barrier;
+static atomic_t tegra_abort_flag;
+
+static void tegra_cpuidle_report_cpus_state(void)
+{
+	unsigned long cpu, lcpu, csr;
+
+	for_each_cpu(lcpu, cpu_possible_mask) {
+		cpu = cpu_logical_map(lcpu);
+		csr = flowctrl_read_cpu_csr(cpu);
+
+		pr_err("cpu%lu: online=%d flowctrl_csr=0x%08lx\n",
+		       cpu, cpu_online(lcpu), csr);
+	}
+}
+
+static int tegra_cpuidle_wait_for_secondary_cpus_parking(void)
+{
+	unsigned int retries = 3;
+
+	while (retries--) {
+		unsigned int delay_us = 10;
+		unsigned int timeout_us = 500 * 1000 / delay_us;
+
+		/*
+		 * The primary CPU0 core shall wait for the secondaries
+		 * shutdown in order to power-off CPU's cluster safely.
+		 * The timeout value depends on the current CPU frequency,
+		 * it takes about 40-150us in average and over 1000us in
+		 * a worst case scenario.
+		 */
+		do {
+			if (tegra_cpu_rail_off_ready())
+				return 0;
+
+			udelay(delay_us);
+
+		} while (timeout_us--);
+
+		pr_err("secondary CPU taking too long to park\n");
+
+		tegra_cpuidle_report_cpus_state();
+	}
+
+	pr_err("timed out waiting secondaries to park\n");
+
+	return -ETIMEDOUT;
+}
+
+static void tegra_cpuidle_unpark_secondary_cpus(void)
+{
+	unsigned int cpu, lcpu;
+
+	for_each_cpu(lcpu, cpu_online_mask) {
+		cpu = cpu_logical_map(lcpu);
+
+		if (cpu > 0) {
+			tegra_enable_cpu_clock(cpu);
+			tegra_cpu_out_of_reset(cpu);
+			flowctrl_write_cpu_halt(cpu, 0);
+		}
+	}
+}
+
+static int tegra_cpuidle_cc6_enter(unsigned int cpu)
+{
+	int ret;
+
+	if (cpu > 0) {
+		ret = cpu_suspend(cpu, tegra_pm_park_secondary_cpu);
+	} else {
+		ret = tegra_cpuidle_wait_for_secondary_cpus_parking();
+		if (!ret)
+			ret = tegra_pm_enter_lp2();
+
+		tegra_cpuidle_unpark_secondary_cpus();
+	}
+
+	return ret;
+}
+
+static int tegra_cpuidle_coupled_barrier(struct cpuidle_device *dev)
+{
+	if (tegra_pending_sgi()) {
+		/*
+		 * CPU got local interrupt that will be lost after GIC's
+		 * shutdown because GIC driver doesn't save/restore the
+		 * pending SGI state across CPU cluster PM.  Abort and retry
+		 * next time.
+		 */
+		atomic_set(&tegra_abort_flag, 1);
+	}
+
+	cpuidle_coupled_parallel_barrier(dev, &tegra_idle_barrier);
+
+	if (atomic_read(&tegra_abort_flag)) {
+		cpuidle_coupled_parallel_barrier(dev, &tegra_idle_barrier);
+		atomic_set(&tegra_abort_flag, 0);
+		return -EINTR;
+	}
+
+	return 0;
+}
+
+static int tegra_cpuidle_state_enter(struct cpuidle_device *dev,
+				     int index, unsigned int cpu)
+{
+	int ret;
+
+	/*
+	 * CC6 state is the "CPU cluster power-off" state.  In order to
+	 * enter this state, at first the secondary CPU cores need to be
+	 * parked into offline mode, then the last CPU should clean out
+	 * remaining dirty cache lines into DRAM and trigger Flow Controller
+	 * logic that turns off the cluster's power domain (which includes
+	 * CPU cores, GIC and L2 cache).
+	 */
+	if (index == TEGRA_CC6) {
+		ret = tegra_cpuidle_coupled_barrier(dev);
+		if (ret)
+			return ret;
+	}
+
+	local_fiq_disable();
+	tegra_pm_set_cpu_in_lp2();
+	cpu_pm_enter();
+
+	switch (index) {
+	case TEGRA_CC6:
+		ret = tegra_cpuidle_cc6_enter(cpu);
+		break;
+
+	default:
+		ret = -EINVAL;
+		break;
+	}
+
+	cpu_pm_exit();
+	tegra_pm_clear_cpu_in_lp2();
+	local_fiq_enable();
+
+	return ret;
+}
+
+static int tegra_cpuidle_enter(struct cpuidle_device *dev,
+			       struct cpuidle_driver *drv,
+			       int index)
+{
+	unsigned int cpu = cpu_logical_map(dev->cpu);
+	int err;
+
+	err = tegra_cpuidle_state_enter(dev, index, cpu);
+	if (err && err != -EINTR)
+		pr_err_once("cpu%u failed to enter idle state %d err: %d\n",
+			    cpu, index, err);
+
+	return err ? -1 : index;
+}
+
+/*
+ * The previous versions of Tegra CPUIDLE driver used a different "legacy"
+ * terminology for naming of the idling states, while this driver uses the
+ * new terminology.
+ *
+ * Mapping of the old terms into the new ones:
+ *
+ * Old | New
+ * ---------
+ * LP3 | C1	(CPU core clock gating)
+ * LP2 | C7	(CPU core power gating)
+ * LP2 | CC6	(CPU cluster power gating)
+ *
+ * Note that that the older CPUIDLE driver versions didn't explicitly
+ * differentiate the LP2 states because these states either used the same
+ * code path or because CC6 wasn't supported.
+ */
+static struct cpuidle_driver tegra_idle_driver = {
+	.name = "tegra_idle",
+	.states = {
+		[TEGRA_C1] = ARM_CPUIDLE_WFI_STATE_PWR(600),
+		[TEGRA_CC6] = {
+			.enter			= tegra_cpuidle_enter,
+			.exit_latency		= 5000,
+			.target_residency	= 10000,
+			.power_usage		= 0,
+			.flags			= CPUIDLE_FLAG_TIMER_STOP |
+						  CPUIDLE_FLAG_COUPLED,
+			.name			= "CC6",
+			.desc			= "CPU cluster powered off",
+		},
+	},
+	.state_count = TEGRA_STATE_COUNT,
+	.safe_state_index = TEGRA_C1,
+};
+
+static inline void tegra_cpuidle_disable_state(enum tegra_state state)
+{
+	cpuidle_driver_state_disabled(&tegra_idle_driver, state, true);
+}
+
+/*
+ * Tegra20 HW appears to have a bug such that PCIe device interrupts, whether
+ * they are legacy IRQs or MSI, are lost when CC6 is enabled.  To work around
+ * this, simply disable CC6 if the PCI driver and DT node are both enabled.
+ */
+void tegra_cpuidle_pcie_irqs_in_use(void)
+{
+	struct cpuidle_state *state_cc6 = &tegra_idle_driver.states[TEGRA_CC6];
+
+	if ((state_cc6->flags & CPUIDLE_FLAG_UNUSABLE) ||
+	    tegra_get_chip_id() != TEGRA20)
+		return;
+
+	pr_info("disabling CC6 state, since PCIe IRQs are in use\n");
+	tegra_cpuidle_disable_state(TEGRA_CC6);
+}
+
+static int tegra_cpuidle_probe(struct platform_device *pdev)
+{
+	/*
+	 * Required suspend-resume functionality, which is provided by the
+	 * Tegra-arch core and PMC driver, is unavailable if PM-sleep option
+	 * is disabled.
+	 */
+	if (!IS_ENABLED(CONFIG_PM_SLEEP))
+		tegra_cpuidle_disable_state(TEGRA_CC6);
+
+	return cpuidle_register(&tegra_idle_driver, cpu_possible_mask);
+}
+
+static struct platform_driver tegra_cpuidle_driver = {
+	.probe = tegra_cpuidle_probe,
+	.driver = {
+		.name = "tegra-cpuidle",
+	},
+};
+builtin_platform_driver(tegra_cpuidle_driver);