Submitting the V5 version of exynos5440 cpufreq driver. This patchset addresses
all the coding and design concerns raised especially by Viresh.
Changes in V5:
* Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
* Fixed all coding and comments issue raised by Viresh.
Changes in V4:
* Added dev_err logs instead of pr_err.
* Used the devm_ioremap_resource API.
* Implemented several coding guidelines and minor error comments from Sylwester,
Russell and Viresh.
Changes in V3:
* Converted the driver to probe based as suggested by Viresh. This is also
beneficial for multiplatform kernel.
* Other coding guidelines related changes.
* Moved the DT node outside cpu0 node as the driver is now a platform
driver.
Changes in V2:
* Added OPP library support to parse DT parameters.
* Removed a hack to handle interrupts in bootup.
* Implemented other review comments from Viresh and Inder.
All these patches are dependent on Thomas Abraham common clock patches.
(http://www.mail-archive.com/[email protected]/msg15860.html)
This whole patch series is based on 3.9-rc4.
Amit Daniel Kachhap (4):
cpufreq: exynos: Add cpufreq driver for exynos5440
cpufreq: exynos: Remove error return even if no soc is found
arm: exynos: Enable OPP library support for exynos5440
arm: dts: Add cpufreq controller node for Exynos5440 SoC
.../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
arch/arm/boot/dts/exynos5440.dtsi | 12 +
arch/arm/mach-exynos/Kconfig | 2 +
drivers/cpufreq/Kconfig.arm | 9 +
drivers/cpufreq/Makefile | 1 +
drivers/cpufreq/exynos-cpufreq.c | 2 +-
drivers/cpufreq/exynos5440-cpufreq.c | 474 ++++++++++++++++++++
7 files changed, 527 insertions(+), 1 deletions(-)
create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
This patch adds dvfs support for exynos5440 SOC. This soc has 4 cores and
they scale at same frequency. The nature of exynos5440 clock controller is
different from previous exynos controllers so not using the common exynos
cpufreq framework. The major difference being interrupt notification for
frequency change. Also, OPP library is used for device tree parsing to get
different parameters like frequency, voltage etc. Since the opp library sorts
the frequency table in ascending order so they are again re-arranged in
descending order. This will have one-to-one mapping with the clock controller
state management logic.
Signed-off-by: Amit Daniel Kachhap <[email protected]>
---
.../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
drivers/cpufreq/Kconfig.arm | 9 +
drivers/cpufreq/Makefile | 1 +
drivers/cpufreq/exynos5440-cpufreq.c | 474 ++++++++++++++++++++
4 files changed, 512 insertions(+), 0 deletions(-)
create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
diff --git a/Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt b/Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
new file mode 100644
index 0000000..caff1a5
--- /dev/null
+++ b/Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
@@ -0,0 +1,28 @@
+
+Exynos5440 cpufreq driver
+-------------------
+
+Exynos5440 SoC cpufreq driver for CPU frequency scaling.
+
+Required properties:
+- interrupts: Interrupt to know the completion of cpu frequency change.
+- operating-points: Table of frequencies and voltage CPU could be transitioned into,
+ in the decreasing order. Frequency should be in KHz units and voltage
+ should be in microvolts.
+
+Optional properties:
+- clock-latency: Clock monitor latency in microsecond.
+
+All the required listed above must be defined under node cpufreq.
+
+Example:
+--------
+ cpufreq@160000 {
+ compatible = "samsung,exynos5440-cpufreq";
+ reg = <0x160000 0x1000>;
+ interrupts = <0 57 0>;
+ operating-points = <
+ 1000000 975000
+ 800000 925000>;
+ clock-latency = <100000>;
+ };
diff --git a/drivers/cpufreq/Kconfig.arm b/drivers/cpufreq/Kconfig.arm
index 030ddf6..7ed9c4a 100644
--- a/drivers/cpufreq/Kconfig.arm
+++ b/drivers/cpufreq/Kconfig.arm
@@ -77,6 +77,15 @@ config ARM_EXYNOS5250_CPUFREQ
This adds the CPUFreq driver for Samsung EXYNOS5250
SoC.
+config ARM_EXYNOS5440_CPUFREQ
+ def_bool SOC_EXYNOS5440
+ depends on HAVE_CLK && PM_OPP && OF
+ help
+ This adds the CPUFreq driver for Samsung EXYNOS5440
+ SoC. The nature of exynos5440 clock controller is
+ different than previous exynos controllers so not using
+ the common exynos framework.
+
config ARM_KIRKWOOD_CPUFREQ
def_bool ARCH_KIRKWOOD && OF
help
diff --git a/drivers/cpufreq/Makefile b/drivers/cpufreq/Makefile
index 863fd18..c841438 100644
--- a/drivers/cpufreq/Makefile
+++ b/drivers/cpufreq/Makefile
@@ -52,6 +52,7 @@ obj-$(CONFIG_ARM_EXYNOS_CPUFREQ) += exynos-cpufreq.o
obj-$(CONFIG_ARM_EXYNOS4210_CPUFREQ) += exynos4210-cpufreq.o
obj-$(CONFIG_ARM_EXYNOS4X12_CPUFREQ) += exynos4x12-cpufreq.o
obj-$(CONFIG_ARM_EXYNOS5250_CPUFREQ) += exynos5250-cpufreq.o
+obj-$(CONFIG_ARM_EXYNOS5440_CPUFREQ) += exynos5440-cpufreq.o
obj-$(CONFIG_ARM_KIRKWOOD_CPUFREQ) += kirkwood-cpufreq.o
obj-$(CONFIG_ARM_OMAP2PLUS_CPUFREQ) += omap-cpufreq.o
obj-$(CONFIG_ARM_SPEAR_CPUFREQ) += spear-cpufreq.o
diff --git a/drivers/cpufreq/exynos5440-cpufreq.c b/drivers/cpufreq/exynos5440-cpufreq.c
new file mode 100644
index 0000000..ee67bd7
--- /dev/null
+++ b/drivers/cpufreq/exynos5440-cpufreq.c
@@ -0,0 +1,474 @@
+/*
+ * Copyright (c) 2013 Samsung Electronics Co., Ltd.
+ * http://www.samsung.com
+ *
+ * Amit Daniel Kachhap <[email protected]>
+ *
+ * EXYNOS5440 - CPU frequency scaling support
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+*/
+
+#include <linux/clk.h>
+#include <linux/cpu.h>
+#include <linux/cpufreq.h>
+#include <linux/err.h>
+#include <linux/interrupt.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/opp.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+/* Register definitions */
+#define XMU_DVFS_CTRL 0x0060
+#define XMU_PMU_P0_7 0x0064
+#define XMU_C0_3_PSTATE 0x0090
+#define XMU_P_LIMIT 0x00a0
+#define XMU_P_STATUS 0x00a4
+#define XMU_PMUEVTEN 0x00d0
+#define XMU_PMUIRQEN 0x00d4
+#define XMU_PMUIRQ 0x00d8
+
+/* PMU mask and shift definations */
+#define P_VALUE_MASK 0x7
+
+#define XMU_DVFS_CTRL_EN_SHIFT 0
+
+#define P0_7_CPUCLKDEV_SHIFT 21
+#define P0_7_CPUCLKDEV_MASK 0x7
+#define P0_7_ATBCLKDEV_SHIFT 18
+#define P0_7_ATBCLKDEV_MASK 0x7
+#define P0_7_CSCLKDEV_SHIFT 15
+#define P0_7_CSCLKDEV_MASK 0x7
+#define P0_7_CPUEMA_SHIFT 28
+#define P0_7_CPUEMA_MASK 0xf
+#define P0_7_L2EMA_SHIFT 24
+#define P0_7_L2EMA_MASK 0xf
+#define P0_7_VDD_SHIFT 8
+#define P0_7_VDD_MASK 0x7f
+#define P0_7_FREQ_SHIFT 0
+#define P0_7_FREQ_MASK 0xff
+
+#define C0_3_PSTATE_VALID_SHIFT 8
+#define C0_3_PSTATE_CURR_SHIFT 4
+#define C0_3_PSTATE_NEW_SHIFT 0
+
+#define PSTATE_CHANGED_EVTEN_SHIFT 0
+
+#define PSTATE_CHANGED_IRQEN_SHIFT 0
+
+#define PSTATE_CHANGED_SHIFT 0
+
+/* some constant values for clock divider calculation */
+#define CPU_DIV_FREQ_MAX 500
+#define CPU_DBG_FREQ_MAX 375
+#define CPU_ATB_FREQ_MAX 500
+
+#define PMIC_LOW_VOLT 0x30
+#define PMIC_HIGH_VOLT 0x28
+
+#define CPUEMA_HIGH 0x2
+#define CPUEMA_MID 0x4
+#define CPUEMA_LOW 0x7
+
+#define L2EMA_HIGH 0x1
+#define L2EMA_MID 0x3
+#define L2EMA_LOW 0x4
+
+#define DIV_TAB_MAX 2
+/* frequency unit is 20MHZ */
+#define FREQ_UNIT 20
+#define MAX_VOLTAGE 1550000 /* In microvolt */
+#define VOLTAGE_STEP 12500 /* In microvolt */
+
+#define CPUFREQ_NAME "exynos5440_dvfs"
+#define DEF_TRANS_LATENCY 100000
+
+enum cpufreq_level_index {
+ L0, L1, L2, L3, L4,
+ L5, L6, L7, L8, L9,
+};
+#define CPUFREQ_LEVEL_END (L7 + 1)
+
+struct exynos_dvfs_data {
+ void __iomem *base;
+ struct resource *mem;
+ int irq;
+ struct clk *cpu_clk;
+ unsigned int cur_frequency;
+ unsigned int latency;
+ struct cpufreq_frequency_table *freq_table;
+ unsigned int freq_count;
+ struct device *dev;
+ bool dvfs_enabled;
+ struct work_struct irq_work;
+};
+
+static struct exynos_dvfs_data *dvfs_info;
+static DEFINE_MUTEX(cpufreq_lock);
+static struct cpufreq_freqs freqs;
+
+static int init_div_table(void)
+{
+ struct cpufreq_frequency_table *freq_tbl = dvfs_info->freq_table;
+ unsigned int tmp, clk_div, ema_div, freq, volt_id;
+ int i = 0;
+ struct opp *opp;
+
+ for (i = 0; freq_tbl[i].frequency != CPUFREQ_TABLE_END; i++) {
+
+ opp = opp_find_freq_exact(dvfs_info->dev,
+ freq_tbl[i].frequency * 1000, true);
+ if (IS_ERR(opp)) {
+ dev_err(dvfs_info->dev,
+ "failed to find valid OPP for %u KHZ\n",
+ freq_tbl[i].frequency);
+ return PTR_ERR(opp);
+ }
+
+ freq = freq_tbl[i].frequency / 1000; /* In MHZ */
+ clk_div = ((freq / CPU_DIV_FREQ_MAX) & P0_7_CPUCLKDEV_MASK)
+ << P0_7_CPUCLKDEV_SHIFT;
+ clk_div |= ((freq / CPU_ATB_FREQ_MAX) & P0_7_ATBCLKDEV_MASK)
+ << P0_7_ATBCLKDEV_SHIFT;
+ clk_div |= ((freq / CPU_DBG_FREQ_MAX) & P0_7_CSCLKDEV_MASK)
+ << P0_7_CSCLKDEV_SHIFT;
+
+ /* Calculate EMA */
+ volt_id = opp_get_voltage(opp);
+ volt_id = (MAX_VOLTAGE - volt_id) / VOLTAGE_STEP;
+ if (volt_id < PMIC_HIGH_VOLT) {
+ ema_div = (CPUEMA_HIGH << P0_7_CPUEMA_SHIFT) |
+ (L2EMA_HIGH << P0_7_L2EMA_SHIFT);
+ } else if (volt_id > PMIC_LOW_VOLT) {
+ ema_div = (CPUEMA_LOW << P0_7_CPUEMA_SHIFT) |
+ (L2EMA_LOW << P0_7_L2EMA_SHIFT);
+ } else {
+ ema_div = (CPUEMA_MID << P0_7_CPUEMA_SHIFT) |
+ (L2EMA_MID << P0_7_L2EMA_SHIFT);
+ }
+
+ tmp = (clk_div | ema_div | (volt_id << P0_7_VDD_SHIFT)
+ | ((freq / FREQ_UNIT) << P0_7_FREQ_SHIFT));
+
+ __raw_writel(tmp, dvfs_info->base + XMU_PMU_P0_7 + 4 * i);
+ }
+
+ return 0;
+}
+
+static void exynos_enable_dvfs(void)
+{
+ unsigned int tmp, i, cpu;
+ struct cpufreq_frequency_table *freq_table = dvfs_info->freq_table;
+ /* Disable DVFS */
+ __raw_writel(0, dvfs_info->base + XMU_DVFS_CTRL);
+
+ /* Enable PSTATE Change Event */
+ tmp = __raw_readl(dvfs_info->base + XMU_PMUEVTEN);
+ tmp |= (1 << PSTATE_CHANGED_EVTEN_SHIFT);
+ __raw_writel(tmp, dvfs_info->base + XMU_PMUEVTEN);
+
+ /* Enable PSTATE Change IRQ */
+ tmp = __raw_readl(dvfs_info->base + XMU_PMUIRQEN);
+ tmp |= (1 << PSTATE_CHANGED_IRQEN_SHIFT);
+ __raw_writel(tmp, dvfs_info->base + XMU_PMUIRQEN);
+
+ /* Set initial performance index */
+ for (i = 0; freq_table[i].frequency != CPUFREQ_TABLE_END; i++)
+ if (freq_table[i].frequency == dvfs_info->cur_frequency)
+ break;
+
+ if (freq_table[i].frequency == CPUFREQ_TABLE_END) {
+ dev_crit(dvfs_info->dev, "Boot up frequency not supported\n");
+ /* Assign the highest frequency */
+ i = 0;
+ dvfs_info->cur_frequency = freq_table[i].frequency;
+ }
+
+ dev_info(dvfs_info->dev, "Setting dvfs initial frequency = %uKHZ",
+ dvfs_info->cur_frequency);
+
+ for (cpu = 0; cpu < CONFIG_NR_CPUS; cpu++) {
+ tmp = __raw_readl(dvfs_info->base + XMU_C0_3_PSTATE + cpu * 4);
+ tmp &= ~(P_VALUE_MASK << C0_3_PSTATE_NEW_SHIFT);
+ tmp |= (i << C0_3_PSTATE_NEW_SHIFT);
+ __raw_writel(tmp, dvfs_info->base + XMU_C0_3_PSTATE + cpu * 4);
+ }
+
+ /* Enable DVFS */
+ __raw_writel(1 << XMU_DVFS_CTRL_EN_SHIFT,
+ dvfs_info->base + XMU_DVFS_CTRL);
+}
+
+static int exynos_verify_speed(struct cpufreq_policy *policy)
+{
+ return cpufreq_frequency_table_verify(policy,
+ dvfs_info->freq_table);
+}
+
+static unsigned int exynos_getspeed(unsigned int cpu)
+{
+ return dvfs_info->cur_frequency;
+}
+
+static int exynos_target(struct cpufreq_policy *policy,
+ unsigned int target_freq,
+ unsigned int relation)
+{
+ unsigned int index, tmp;
+ int ret = 0, i;
+ struct cpufreq_frequency_table *freq_table = dvfs_info->freq_table;
+
+ mutex_lock(&cpufreq_lock);
+ freqs.old = dvfs_info->cur_frequency;
+
+ ret = cpufreq_frequency_table_target(policy, freq_table,
+ target_freq, relation, &index);
+ if (ret)
+ goto out;
+
+ freqs.new = freq_table[index].frequency;
+ freqs.cpu = policy->cpu;
+
+ for_each_cpu(freqs.cpu, policy->cpus)
+ cpufreq_notify_transition(&freqs, CPUFREQ_PRECHANGE);
+
+ /* Set the target frequency in all C0_3_PSTATE register */
+ for_each_cpu(i, policy->cpus) {
+ tmp = __raw_readl(dvfs_info->base + XMU_C0_3_PSTATE + i * 4);
+ tmp &= ~(P_VALUE_MASK << C0_3_PSTATE_NEW_SHIFT);
+ tmp |= (index << C0_3_PSTATE_NEW_SHIFT);
+
+ __raw_writel(tmp, dvfs_info->base + XMU_C0_3_PSTATE + i * 4);
+ }
+out:
+ mutex_unlock(&cpufreq_lock);
+ return ret;
+}
+
+static void exynos_cpufreq_work(struct work_struct *work)
+{
+ unsigned int cur_pstate, index;
+ struct cpufreq_policy *policy = cpufreq_cpu_get(0); /* boot CPU */
+ struct cpufreq_frequency_table *freq_table = dvfs_info->freq_table;
+
+ /* Ensure we can access cpufreq structures */
+ if (unlikely(dvfs_info->dvfs_enabled == false))
+ goto skip_work;
+
+ mutex_lock(&cpufreq_lock);
+ freqs.old = dvfs_info->cur_frequency;
+
+ cur_pstate = __raw_readl(dvfs_info->base + XMU_P_STATUS);
+ if (cur_pstate >> C0_3_PSTATE_VALID_SHIFT & 0x1)
+ index = (cur_pstate >> C0_3_PSTATE_CURR_SHIFT) & P_VALUE_MASK;
+ else
+ index = (cur_pstate >> C0_3_PSTATE_NEW_SHIFT) & P_VALUE_MASK;
+
+ if (likely(index < dvfs_info->freq_count)) {
+ freqs.new = freq_table[index].frequency;
+ for_each_cpu(freqs.cpu, policy->cpus)
+ cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
+ dvfs_info->cur_frequency = freqs.new;
+ } else {
+ dev_crit(dvfs_info->dev, "New frequency out of range\n");
+ }
+
+ cpufreq_cpu_put(policy);
+ mutex_unlock(&cpufreq_lock);
+skip_work:
+ enable_irq(dvfs_info->irq);
+}
+
+static irqreturn_t exynos_cpufreq_irq(int irq, void *id)
+{
+ unsigned int tmp;
+
+ tmp = __raw_readl(dvfs_info->base + XMU_PMUIRQ);
+ if (tmp >> PSTATE_CHANGED_SHIFT & 0x1) {
+ __raw_writel(tmp, dvfs_info->base + XMU_PMUIRQ);
+ disable_irq_nosync(irq);
+ schedule_work(&dvfs_info->irq_work);
+ }
+ return IRQ_HANDLED;
+}
+
+static void exynos_sort_descend_freq_table(void)
+{
+ struct cpufreq_frequency_table *freq_tbl = dvfs_info->freq_table;
+ int i = 0, index;
+ unsigned int tmp_freq;
+ /*
+ * Exynos5440 clock controller state logic expects the cpufreq table to
+ * be in descending order. But the OPP library constructs the table in
+ * ascending order. So to make the table descending we just need to
+ * swap the i element with the N - i element.
+ */
+ for (i = 0; i < dvfs_info->freq_count / 2; i++) {
+ index = dvfs_info->freq_count - i - 1;
+ tmp_freq = freq_tbl[i].frequency;
+ freq_tbl[i].frequency = freq_tbl[index].frequency;
+ freq_tbl[index].frequency = tmp_freq;
+ }
+}
+
+static int exynos_cpufreq_cpu_init(struct cpufreq_policy *policy)
+{
+ policy->cur = dvfs_info->cur_frequency;
+ cpufreq_frequency_table_get_attr(dvfs_info->freq_table, policy->cpu);
+
+ /* set the transition latency value */
+ policy->cpuinfo.transition_latency = dvfs_info->latency;
+
+ cpumask_setall(policy->cpus);
+
+ return cpufreq_frequency_table_cpuinfo(policy, dvfs_info->freq_table);
+}
+
+static struct cpufreq_driver exynos_driver = {
+ .flags = CPUFREQ_STICKY,
+ .verify = exynos_verify_speed,
+ .target = exynos_target,
+ .get = exynos_getspeed,
+ .init = exynos_cpufreq_cpu_init,
+ .name = CPUFREQ_NAME,
+};
+
+static const struct of_device_id exynos_cpufreq_match[] = {
+ {
+ .compatible = "samsung,exynos5440-cpufreq",
+ },
+ {},
+};
+MODULE_DEVICE_TABLE(of, exynos_cpufreq_match);
+
+static int exynos_cpufreq_probe(struct platform_device *pdev)
+{
+ int ret = -EINVAL;
+ struct device_node *np;
+ struct resource res;
+
+ np = pdev->dev.of_node;
+ if (!np)
+ return -ENODEV;
+
+ dvfs_info = devm_kzalloc(&pdev->dev, sizeof(*dvfs_info), GFP_KERNEL);
+ if (!dvfs_info) {
+ ret = -ENOMEM;
+ goto err_put_node;
+ }
+
+ dvfs_info->dev = &pdev->dev;
+
+ ret = of_address_to_resource(np, 0, &res);
+ if (ret)
+ goto err_put_node;
+
+ dvfs_info->base = devm_ioremap_resource(dvfs_info->dev, &res);
+ if (IS_ERR(dvfs_info->base)) {
+ ret = PTR_ERR(dvfs_info->base);
+ goto err_put_node;
+ }
+
+ dvfs_info->irq = irq_of_parse_and_map(np, 0);
+ if (!dvfs_info->irq) {
+ dev_err(dvfs_info->dev, "No cpufreq irq found\n");
+ ret = -ENODEV;
+ goto err_put_node;
+ }
+
+ ret = of_init_opp_table(dvfs_info->dev);
+ if (ret) {
+ dev_err(dvfs_info->dev, "failed to init OPP table: %d\n", ret);
+ goto err_put_node;
+ }
+
+ ret = opp_init_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
+ if (ret) {
+ dev_err(dvfs_info->dev,
+ "failed to init cpufreq table: %d\n", ret);
+ goto err_put_node;
+ }
+ dvfs_info->freq_count = opp_get_opp_count(dvfs_info->dev);
+ exynos_sort_descend_freq_table();
+
+ if (of_property_read_u32(np, "clock-latency", &dvfs_info->latency))
+ dvfs_info->latency = DEF_TRANS_LATENCY;
+
+ dvfs_info->cpu_clk = devm_clk_get(dvfs_info->dev, "armclk");
+ if (IS_ERR(dvfs_info->cpu_clk)) {
+ dev_err(dvfs_info->dev, "Failed to get cpu clock\n");
+ ret = PTR_ERR(dvfs_info->cpu_clk);
+ goto err_free_table;
+ }
+
+ dvfs_info->cur_frequency = clk_get_rate(dvfs_info->cpu_clk);
+ if (!dvfs_info->cur_frequency) {
+ dev_err(dvfs_info->dev, "Failed to get clock rate\n");
+ ret = -EINVAL;
+ goto err_free_table;
+ }
+ dvfs_info->cur_frequency /= 1000;
+
+ INIT_WORK(&dvfs_info->irq_work, exynos_cpufreq_work);
+ ret = devm_request_irq(dvfs_info->dev, dvfs_info->irq,
+ exynos_cpufreq_irq, IRQF_TRIGGER_NONE,
+ CPUFREQ_NAME, dvfs_info);
+ if (ret) {
+ dev_err(dvfs_info->dev, "Failed to register IRQ\n");
+ goto err_free_table;
+ }
+
+ ret = init_div_table();
+ if (ret) {
+ dev_err(dvfs_info->dev, "Failed to initialise div table\n");
+ goto err_free_table;
+ }
+
+ exynos_enable_dvfs();
+ ret = cpufreq_register_driver(&exynos_driver);
+ if (ret) {
+ dev_err(dvfs_info->dev,
+ "%s: failed to register cpufreq driver\n", __func__);
+ goto err_free_table;
+ }
+
+ of_node_put(np);
+ dvfs_info->dvfs_enabled = true;
+ return 0;
+
+err_free_table:
+ opp_free_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
+err_put_node:
+ of_node_put(np);
+ dev_err(dvfs_info->dev, "%s: failed initialization\n", __func__);
+ return ret;
+}
+
+static int exynos_cpufreq_remove(struct platform_device *pdev)
+{
+ cpufreq_unregister_driver(&exynos_driver);
+ opp_free_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
+ return 0;
+}
+
+static struct platform_driver exynos_cpufreq_platdrv = {
+ .driver = {
+ .name = "exynos5440-cpufreq",
+ .owner = THIS_MODULE,
+ .of_match_table = exynos_cpufreq_match,
+ },
+ .probe = exynos_cpufreq_probe,
+ .remove = exynos_cpufreq_remove,
+};
+module_platform_driver(exynos_cpufreq_platdrv);
+
+MODULE_AUTHOR("Amit Daniel Kachhap <[email protected]>");
+MODULE_DESCRIPTION("Exynos5440 cpufreq driver");
+MODULE_LICENSE("GPL");
--
1.7.1
The OPP library support is needed for exynos5440 cpu frequency
dynamic scaling driver.
Signed-off-by: Amit Daniel Kachhap <[email protected]>
---
arch/arm/mach-exynos/Kconfig | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/arch/arm/mach-exynos/Kconfig b/arch/arm/mach-exynos/Kconfig
index 70f94c8..d5dde07 100644
--- a/arch/arm/mach-exynos/Kconfig
+++ b/arch/arm/mach-exynos/Kconfig
@@ -72,10 +72,12 @@ config SOC_EXYNOS5440
bool "SAMSUNG EXYNOS5440"
default y
depends on ARCH_EXYNOS5
+ select ARCH_HAS_OPP
select ARM_ARCH_TIMER
select AUTO_ZRELADDR
select PINCTRL
select PINCTRL_EXYNOS5440
+ select PM_OPP
help
Enable EXYNOS5440 SoC support
--
1.7.1
Add cpufreq controller device node for Exynos5440 SoC for passing
parameters like controller base address, interrupt and cpufreq
table. This node is added outside cpu0 as this driver is now a platform
driver and a new device structure is needed.
Signed-off-by: Amit Daniel Kachhap <[email protected]>
---
arch/arm/boot/dts/exynos5440.dtsi | 12 ++++++++++++
1 files changed, 12 insertions(+), 0 deletions(-)
diff --git a/arch/arm/boot/dts/exynos5440.dtsi b/arch/arm/boot/dts/exynos5440.dtsi
index 9a99755..94074ea 100644
--- a/arch/arm/boot/dts/exynos5440.dtsi
+++ b/arch/arm/boot/dts/exynos5440.dtsi
@@ -63,6 +63,18 @@
};
+ cpufreq@160000 {
+ compatible = "samsung,exynos5440-cpufreq";
+ reg = <0x160000 0x1000>;
+ interrupts = <0 57 0>;
+ operating-points = <
+ /* KHz uV */
+ 1200000 1025000
+ 1000000 975000
+ 800000 925000
+ >;
+ };
+
serial@B0000 {
compatible = "samsung,exynos4210-uart";
reg = <0xB0000 0x1000>;
--
1.7.1
This patch helps to have single binary for exynos5440 and previous
exynos soc's. This change is needed for adding exynos5440 cpufreq driver
which does not uses exynos-cpufreq common file and adds it own driver.
Signed-off-by: Amit Daniel Kachhap <[email protected]>
---
drivers/cpufreq/exynos-cpufreq.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/cpufreq/exynos-cpufreq.c b/drivers/cpufreq/exynos-cpufreq.c
index 78057a3..ee75997 100644
--- a/drivers/cpufreq/exynos-cpufreq.c
+++ b/drivers/cpufreq/exynos-cpufreq.c
@@ -297,7 +297,7 @@ static int __init exynos_cpufreq_init(void)
else if (soc_is_exynos5250())
ret = exynos5250_cpufreq_init(exynos_info);
else
- pr_err("%s: CPU type not found\n", __func__);
+ return 0;
if (ret)
goto err_vdd_arm;
--
1.7.1
On 28 March 2013 13:35, Amit Daniel Kachhap <[email protected]> wrote:
> Submitting the V5 version of exynos5440 cpufreq driver. This patchset addresses
> all the coding and design concerns raised especially by Viresh.
>
> Changes in V5:
> * Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
> * Fixed all coding and comments issue raised by Viresh.
People normally forget what they commented on earlier and that's why
this change log phenomenon was introduced. Some more details would
be better here for next time.
On 28 March 2013 13:35, Amit Daniel Kachhap <[email protected]> wrote:
> This patch adds dvfs support for exynos5440 SOC. This soc has 4 cores and
> they scale at same frequency. The nature of exynos5440 clock controller is
> different from previous exynos controllers so not using the common exynos
> cpufreq framework. The major difference being interrupt notification for
> frequency change. Also, OPP library is used for device tree parsing to get
> different parameters like frequency, voltage etc. Since the opp library sorts
> the frequency table in ascending order so they are again re-arranged in
> descending order. This will have one-to-one mapping with the clock controller
> state management logic.
>
> Signed-off-by: Amit Daniel Kachhap <[email protected]>
Mostly okay now, just minor comments:
> diff --git a/drivers/cpufreq/exynos5440-cpufreq.c b/drivers/cpufreq/exynos5440-cpufreq.c
> +static void exynos_cpufreq_work(struct work_struct *work)
> +{
> + if (likely(index < dvfs_info->freq_count)) {
> + freqs.new = freq_table[index].frequency;
> + for_each_cpu(freqs.cpu, policy->cpus)
> + cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
> + dvfs_info->cur_frequency = freqs.new;
> + } else {
> + dev_crit(dvfs_info->dev, "New frequency out of range\n");
> + }
I believe there is something wrong here. For failure cases too
we need to issue cpufreq_notify_transition(&freqs, CPUFREQ_POSTCHANGE);
with the old frequency.
> +static int exynos_cpufreq_probe(struct platform_device *pdev)
> +{
> + ret = cpufreq_register_driver(&exynos_driver);
> + if (ret) {
> + dev_err(dvfs_info->dev,
> + "%s: failed to register cpufreq driver\n", __func__);
> + goto err_free_table;
> + }
> +
> + of_node_put(np);
Don't we need to put node everytime?
> + dvfs_info->dvfs_enabled = true;
> + return 0;
> +
> +err_free_table:
> + opp_free_cpufreq_table(dvfs_info->dev, &dvfs_info->freq_table);
> +err_put_node:
> + of_node_put(np);
> + dev_err(dvfs_info->dev, "%s: failed initialization\n", __func__);
> + return ret;
> +}
--
viresh
On 28 March 2013 13:35, Amit Daniel Kachhap <[email protected]> wrote:
> +++ b/drivers/cpufreq/exynos5440-cpufreq.c
> +static int exynos_cpufreq_cpu_init(struct cpufreq_policy *policy)
> +{
> + policy->cur = dvfs_info->cur_frequency;
> + cpufreq_frequency_table_get_attr(dvfs_info->freq_table, policy->cpu);
> +
> + /* set the transition latency value */
> + policy->cpuinfo.transition_latency = dvfs_info->latency;
> +
> + cpumask_setall(policy->cpus);
> +
> + return cpufreq_frequency_table_cpuinfo(policy, dvfs_info->freq_table);
I missed this earlier and saw this mistake in yet another driver.
You need to call cpufreq_frequency_table_get_attr() only when
cpufreq_frequency_table_cpuinfo() has passed.
> +}
On Thursday, March 28, 2013 01:35:18 PM Amit Daniel Kachhap wrote:
> Submitting the V5 version of exynos5440 cpufreq driver. This patchset addresses
> all the coding and design concerns raised especially by Viresh.
>
> Changes in V5:
> * Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
> * Fixed all coding and comments issue raised by Viresh.
>
> Changes in V4:
> * Added dev_err logs instead of pr_err.
> * Used the devm_ioremap_resource API.
> * Implemented several coding guidelines and minor error comments from Sylwester,
> Russell and Viresh.
>
> Changes in V3:
> * Converted the driver to probe based as suggested by Viresh. This is also
> beneficial for multiplatform kernel.
> * Other coding guidelines related changes.
> * Moved the DT node outside cpu0 node as the driver is now a platform
> driver.
>
> Changes in V2:
> * Added OPP library support to parse DT parameters.
> * Removed a hack to handle interrupts in bootup.
> * Implemented other review comments from Viresh and Inder.
>
> All these patches are dependent on Thomas Abraham common clock patches.
> (http://www.mail-archive.com/[email protected]/msg15860.html)
> This whole patch series is based on 3.9-rc4.
>
> Amit Daniel Kachhap (4):
> cpufreq: exynos: Add cpufreq driver for exynos5440
> cpufreq: exynos: Remove error return even if no soc is found
> arm: exynos: Enable OPP library support for exynos5440
> arm: dts: Add cpufreq controller node for Exynos5440 SoC
>
> .../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
> arch/arm/boot/dts/exynos5440.dtsi | 12 +
> arch/arm/mach-exynos/Kconfig | 2 +
> drivers/cpufreq/Kconfig.arm | 9 +
> drivers/cpufreq/Makefile | 1 +
> drivers/cpufreq/exynos-cpufreq.c | 2 +-
> drivers/cpufreq/exynos5440-cpufreq.c | 474 ++++++++++++++++++++
> 7 files changed, 527 insertions(+), 1 deletions(-)
> create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
> create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
Is this for me or the Samsung tree?
If it's for me, I need ACKs from the Samsung tree maintainers. Also, please
address the Viresh's comments.
Thanks,
Rafael
--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.
On Thu, Mar 28, 2013 at 6:12 PM, Viresh Kumar <[email protected]> wrote:
>
> On 28 March 2013 13:35, Amit Daniel Kachhap <[email protected]> wrote:
> > +++ b/drivers/cpufreq/exynos5440-cpufreq.c
> > +static int exynos_cpufreq_cpu_init(struct cpufreq_policy *policy)
> > +{
> > + policy->cur = dvfs_info->cur_frequency;
> > + cpufreq_frequency_table_get_attr(dvfs_info->freq_table, policy->cpu);
> > +
> > + /* set the transition latency value */
> > + policy->cpuinfo.transition_latency = dvfs_info->latency;
> > +
> > + cpumask_setall(policy->cpus);
> > +
> > + return cpufreq_frequency_table_cpuinfo(policy, dvfs_info->freq_table);
>
> I missed this earlier and saw this mistake in yet another driver.
> You need to call cpufreq_frequency_table_get_attr() only when
> cpufreq_frequency_table_cpuinfo() has passed.
ok. Submitted the V6 version with this change.
>
>
> > +}
> --
> To unsubscribe from this list: send the line "unsubscribe linux-samsung-soc" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Fri, Apr 5, 2013 at 6:08 PM, Rafael J. Wysocki <[email protected]> wrote:
> On Thursday, March 28, 2013 01:35:18 PM Amit Daniel Kachhap wrote:
>> Submitting the V5 version of exynos5440 cpufreq driver. This patchset addresses
>> all the coding and design concerns raised especially by Viresh.
>>
>> Changes in V5:
>> * Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
>> * Fixed all coding and comments issue raised by Viresh.
>>
>> Changes in V4:
>> * Added dev_err logs instead of pr_err.
>> * Used the devm_ioremap_resource API.
>> * Implemented several coding guidelines and minor error comments from Sylwester,
>> Russell and Viresh.
>>
>> Changes in V3:
>> * Converted the driver to probe based as suggested by Viresh. This is also
>> beneficial for multiplatform kernel.
>> * Other coding guidelines related changes.
>> * Moved the DT node outside cpu0 node as the driver is now a platform
>> driver.
>>
>> Changes in V2:
>> * Added OPP library support to parse DT parameters.
>> * Removed a hack to handle interrupts in bootup.
>> * Implemented other review comments from Viresh and Inder.
>>
>> All these patches are dependent on Thomas Abraham common clock patches.
>> (http://www.mail-archive.com/[email protected]/msg15860.html)
>> This whole patch series is based on 3.9-rc4.
>>
>> Amit Daniel Kachhap (4):
>> cpufreq: exynos: Add cpufreq driver for exynos5440
>> cpufreq: exynos: Remove error return even if no soc is found
>> arm: exynos: Enable OPP library support for exynos5440
>> arm: dts: Add cpufreq controller node for Exynos5440 SoC
>>
>> .../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
>> arch/arm/boot/dts/exynos5440.dtsi | 12 +
>> arch/arm/mach-exynos/Kconfig | 2 +
>> drivers/cpufreq/Kconfig.arm | 9 +
>> drivers/cpufreq/Makefile | 1 +
>> drivers/cpufreq/exynos-cpufreq.c | 2 +-
>> drivers/cpufreq/exynos5440-cpufreq.c | 474 ++++++++++++++++++++
>> 7 files changed, 527 insertions(+), 1 deletions(-)
>> create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-exynos5440.txt
>> create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
>
> Is this for me or the Samsung tree?
>
> If it's for me, I need ACKs from the Samsung tree maintainers. Also, please
> address the Viresh's comments.
>
> Thanks,
> Rafael
>
Hi Rafael,
I submitted the V6 version with all comments from Viresh addressed.
Hi Kukjin Kim,
Any suggestion on how this patch should be merged. In my opinion it
should go via Rafael's tree as this patch uses some fixes present in
Rafael's tree like modified cpufreq_notify_transition API.
Thanks,
Amit Daniel
>
> --
> I speak only for myself.
> Rafael J. Wysocki, Intel Open Source Technology Center.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
amit daniel kachhap wrote:
>
> On Fri, Apr 5, 2013 at 6:08 PM, Rafael J. Wysocki <[email protected]> wrote:
> > On Thursday, March 28, 2013 01:35:18 PM Amit Daniel Kachhap wrote:
> >> Submitting the V5 version of exynos5440 cpufreq driver. This patchset
> addresses
> >> all the coding and design concerns raised especially by Viresh.
> >>
> >> Changes in V5:
> >> * Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
> >> * Fixed all coding and comments issue raised by Viresh.
> >>
> >> Changes in V4:
> >> * Added dev_err logs instead of pr_err.
> >> * Used the devm_ioremap_resource API.
> >> * Implemented several coding guidelines and minor error comments from
> Sylwester,
> >> Russell and Viresh.
> >>
> >> Changes in V3:
> >> * Converted the driver to probe based as suggested by Viresh. This is
> also
> >> beneficial for multiplatform kernel.
> >> * Other coding guidelines related changes.
> >> * Moved the DT node outside cpu0 node as the driver is now a platform
> >> driver.
> >>
> >> Changes in V2:
> >> * Added OPP library support to parse DT parameters.
> >> * Removed a hack to handle interrupts in bootup.
> >> * Implemented other review comments from Viresh and Inder.
> >>
> >> All these patches are dependent on Thomas Abraham common clock patches.
> >> (http://www.mail-archive.com/linux-samsung-
> [email protected]/msg15860.html)
> >> This whole patch series is based on 3.9-rc4.
> >>
> >> Amit Daniel Kachhap (4):
> >> cpufreq: exynos: Add cpufreq driver for exynos5440
> >> cpufreq: exynos: Remove error return even if no soc is found
> >> arm: exynos: Enable OPP library support for exynos5440
> >> arm: dts: Add cpufreq controller node for Exynos5440 SoC
> >>
> >> .../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
> >> arch/arm/boot/dts/exynos5440.dtsi | 12 +
> >> arch/arm/mach-exynos/Kconfig | 2 +
> >> drivers/cpufreq/Kconfig.arm | 9 +
> >> drivers/cpufreq/Makefile | 1 +
> >> drivers/cpufreq/exynos-cpufreq.c | 2 +-
> >> drivers/cpufreq/exynos5440-cpufreq.c | 474
> ++++++++++++++++++++
> >> 7 files changed, 527 insertions(+), 1 deletions(-)
> >> create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-
> exynos5440.txt
> >> create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
> >
> > Is this for me or the Samsung tree?
> >
> > If it's for me, I need ACKs from the Samsung tree maintainers. Also,
> please
> > address the Viresh's comments.
> >
> > Thanks,
> > Rafael
> >
> Hi Rafael,
>
> I submitted the V6 version with all comments from Viresh addressed.
>
> Hi Kukjin Kim,
>
> Any suggestion on how this patch should be merged. In my opinion it
> should go via Rafael's tree as this patch uses some fixes present in
> Rafael's tree like modified cpufreq_notify_transition API.
>
Yes, I agree. The v6 series including resending 1/4 patch looks good to me.
But the patch V6 4/4 would be sent to upstream via samsung tree to avoid
useless merge conflicts and I think, it's possible.
So Rafael, please take 1/4 ~ 3/4 patches in your tree with my ack.
Acked-by: Kukjin Kim <[email protected]>
If any problems, please let me know.
Thanks.
- Kukjin
On Monday, April 08, 2013 09:02:44 PM Kukjin Kim wrote:
> amit daniel kachhap wrote:
> >
> > On Fri, Apr 5, 2013 at 6:08 PM, Rafael J. Wysocki <[email protected]> wrote:
> > > On Thursday, March 28, 2013 01:35:18 PM Amit Daniel Kachhap wrote:
> > >> Submitting the V5 version of exynos5440 cpufreq driver. This patchset
> > addresses
> > >> all the coding and design concerns raised especially by Viresh.
> > >>
> > >> Changes in V5:
> > >> * Removed the unnecessary DT look up entry from mach-exynos5-dt.c.
> > >> * Fixed all coding and comments issue raised by Viresh.
> > >>
> > >> Changes in V4:
> > >> * Added dev_err logs instead of pr_err.
> > >> * Used the devm_ioremap_resource API.
> > >> * Implemented several coding guidelines and minor error comments from
> > Sylwester,
> > >> Russell and Viresh.
> > >>
> > >> Changes in V3:
> > >> * Converted the driver to probe based as suggested by Viresh. This is
> > also
> > >> beneficial for multiplatform kernel.
> > >> * Other coding guidelines related changes.
> > >> * Moved the DT node outside cpu0 node as the driver is now a platform
> > >> driver.
> > >>
> > >> Changes in V2:
> > >> * Added OPP library support to parse DT parameters.
> > >> * Removed a hack to handle interrupts in bootup.
> > >> * Implemented other review comments from Viresh and Inder.
> > >>
> > >> All these patches are dependent on Thomas Abraham common clock patches.
> > >> (http://www.mail-archive.com/linux-samsung-
> > [email protected]/msg15860.html)
> > >> This whole patch series is based on 3.9-rc4.
> > >>
> > >> Amit Daniel Kachhap (4):
> > >> cpufreq: exynos: Add cpufreq driver for exynos5440
> > >> cpufreq: exynos: Remove error return even if no soc is found
> > >> arm: exynos: Enable OPP library support for exynos5440
> > >> arm: dts: Add cpufreq controller node for Exynos5440 SoC
> > >>
> > >> .../bindings/cpufreq/cpufreq-exynos5440.txt | 28 ++
> > >> arch/arm/boot/dts/exynos5440.dtsi | 12 +
> > >> arch/arm/mach-exynos/Kconfig | 2 +
> > >> drivers/cpufreq/Kconfig.arm | 9 +
> > >> drivers/cpufreq/Makefile | 1 +
> > >> drivers/cpufreq/exynos-cpufreq.c | 2 +-
> > >> drivers/cpufreq/exynos5440-cpufreq.c | 474
> > ++++++++++++++++++++
> > >> 7 files changed, 527 insertions(+), 1 deletions(-)
> > >> create mode 100644 Documentation/devicetree/bindings/cpufreq/cpufreq-
> > exynos5440.txt
> > >> create mode 100644 drivers/cpufreq/exynos5440-cpufreq.c
> > >
> > > Is this for me or the Samsung tree?
> > >
> > > If it's for me, I need ACKs from the Samsung tree maintainers. Also,
> > please
> > > address the Viresh's comments.
> > >
> > > Thanks,
> > > Rafael
> > >
> > Hi Rafael,
> >
> > I submitted the V6 version with all comments from Viresh addressed.
> >
> > Hi Kukjin Kim,
> >
> > Any suggestion on how this patch should be merged. In my opinion it
> > should go via Rafael's tree as this patch uses some fixes present in
> > Rafael's tree like modified cpufreq_notify_transition API.
> >
> Yes, I agree. The v6 series including resending 1/4 patch looks good to me.
> But the patch V6 4/4 would be sent to upstream via samsung tree to avoid
> useless merge conflicts and I think, it's possible.
>
> So Rafael, please take 1/4 ~ 3/4 patches in your tree with my ack.
>
> Acked-by: Kukjin Kim <[email protected]>
>
> If any problems, please let me know.
I will, thank you!
Rafael
--
I speak only for myself.
Rafael J. Wysocki, Intel Open Source Technology Center.