diff mbox series

[v4,16/16] PM / devfreq: Introduce driver for NVIDIA Tegra20

Message ID 20190501233815.32643-17-digetx@gmail.com (mailing list archive)
State Not Applicable, archived
Headers show
Series NVIDIA Tegra devfreq improvements and Tegra20/30 support | expand

Commit Message

Dmitry Osipenko May 1, 2019, 11:38 p.m. UTC
Add devfreq driver for NVIDIA Tegra20 SoC's. The driver periodically
reads out Memory Controller counters and adjusts memory frequency based
on the memory clients activity.

Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com>
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
---
 MAINTAINERS                       |   8 ++
 drivers/devfreq/Kconfig           |  10 ++
 drivers/devfreq/Makefile          |   1 +
 drivers/devfreq/tegra20-devfreq.c | 212 ++++++++++++++++++++++++++++++
 4 files changed, 231 insertions(+)
 create mode 100644 drivers/devfreq/tegra20-devfreq.c

Comments

Thierry Reding June 4, 2019, 11:25 a.m. UTC | #1
On Thu, May 02, 2019 at 02:38:15AM +0300, Dmitry Osipenko wrote:
> Add devfreq driver for NVIDIA Tegra20 SoC's. The driver periodically
> reads out Memory Controller counters and adjusts memory frequency based
> on the memory clients activity.
> 
> Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com>
> Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
> ---
>  MAINTAINERS                       |   8 ++
>  drivers/devfreq/Kconfig           |  10 ++
>  drivers/devfreq/Makefile          |   1 +
>  drivers/devfreq/tegra20-devfreq.c | 212 ++++++++++++++++++++++++++++++
>  4 files changed, 231 insertions(+)
>  create mode 100644 drivers/devfreq/tegra20-devfreq.c
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 98edc38bfd7b..e7e434f74038 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -10098,6 +10098,14 @@ F:	include/linux/memblock.h
>  F:	mm/memblock.c
>  F:	Documentation/core-api/boot-time-mm.rst
>  
> +MEMORY FREQUENCY SCALING DRIVER FOR NVIDIA TEGRA20
> +M:	Dmitry Osipenko <digetx@gmail.com>
> +L:	linux-pm@vger.kernel.org
> +L:	linux-tegra@vger.kernel.org
> +T:	git git://git.kernel.org/pub/scm/linux/kernel/git/mzx/devfreq.git
> +S:	Maintained
> +F:	drivers/devfreq/tegra20-devfreq.c
> +
>  MEMORY MANAGEMENT
>  L:	linux-mm@kvack.org
>  W:	http://www.linux-mm.org
> diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig
> index a6bba6e1e7d9..1530dbefa31f 100644
> --- a/drivers/devfreq/Kconfig
> +++ b/drivers/devfreq/Kconfig
> @@ -100,6 +100,16 @@ config ARM_TEGRA_DEVFREQ
>  	  It reads ACTMON counters of memory controllers and adjusts the
>  	  operating frequencies and voltages with OPP support.
>  
> +config ARM_TEGRA20_DEVFREQ
> +	tristate "NVIDIA Tegra20 DEVFREQ Driver"
> +	depends on (TEGRA_MC && TEGRA20_EMC) || COMPILE_TEST
> +	select DEVFREQ_GOV_SIMPLE_ONDEMAND
> +	select PM_OPP

Again, I'm not sure the COMPILE_TEST will work here unless you add a few
more dependencies.

Thierry

> +	help
> +	  This adds the DEVFREQ driver for the Tegra20 family of SoCs.
> +	  It reads Memory Controller counters and adjusts the operating
> +	  frequencies and voltages with OPP support.
> +
>  config ARM_RK3399_DMC_DEVFREQ
>  	tristate "ARM RK3399 DMC DEVFREQ Driver"
>  	depends on ARCH_ROCKCHIP
> diff --git a/drivers/devfreq/Makefile b/drivers/devfreq/Makefile
> index 47e5aeeebfd1..338ae8440db6 100644
> --- a/drivers/devfreq/Makefile
> +++ b/drivers/devfreq/Makefile
> @@ -11,6 +11,7 @@ obj-$(CONFIG_DEVFREQ_GOV_PASSIVE)	+= governor_passive.o
>  obj-$(CONFIG_ARM_EXYNOS_BUS_DEVFREQ)	+= exynos-bus.o
>  obj-$(CONFIG_ARM_RK3399_DMC_DEVFREQ)	+= rk3399_dmc.o
>  obj-$(CONFIG_ARM_TEGRA_DEVFREQ)		+= tegra30-devfreq.o
> +obj-$(CONFIG_ARM_TEGRA20_DEVFREQ)	+= tegra20-devfreq.o
>  
>  # DEVFREQ Event Drivers
>  obj-$(CONFIG_PM_DEVFREQ_EVENT)		+= event/
> diff --git a/drivers/devfreq/tegra20-devfreq.c b/drivers/devfreq/tegra20-devfreq.c
> new file mode 100644
> index 000000000000..ff82bac9ee4e
> --- /dev/null
> +++ b/drivers/devfreq/tegra20-devfreq.c
> @@ -0,0 +1,212 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * NVIDIA Tegra20 devfreq driver
> + *
> + * Copyright (C) 2019 GRATE-DRIVER project
> + */
> +
> +#include <linux/clk.h>
> +#include <linux/devfreq.h>
> +#include <linux/io.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/of_device.h>
> +#include <linux/platform_device.h>
> +#include <linux/pm_opp.h>
> +#include <linux/slab.h>
> +
> +#include <soc/tegra/mc.h>
> +
> +#include "governor.h"
> +
> +#define MC_STAT_CONTROL				0x90
> +#define MC_STAT_EMC_CLOCK_LIMIT			0xa0
> +#define MC_STAT_EMC_CLOCKS			0xa4
> +#define MC_STAT_EMC_CONTROL			0xa8
> +#define MC_STAT_EMC_COUNT			0xb8
> +
> +#define EMC_GATHER_CLEAR			(1 << 8)
> +#define EMC_GATHER_ENABLE			(3 << 8)
> +
> +struct tegra_devfreq {
> +	struct devfreq *devfreq;
> +	struct clk *emc_clock;
> +	void __iomem *regs;
> +};
> +
> +static int tegra_devfreq_target(struct device *dev, unsigned long *freq,
> +				u32 flags)
> +{
> +	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
> +	struct devfreq *devfreq = tegra->devfreq;
> +	struct dev_pm_opp *opp;
> +	unsigned long rate;
> +	int err;
> +
> +	opp = devfreq_recommended_opp(dev, freq, flags);
> +	if (IS_ERR(opp))
> +		return PTR_ERR(opp);
> +
> +	rate = dev_pm_opp_get_freq(opp);
> +	dev_pm_opp_put(opp);
> +
> +	err = clk_set_min_rate(tegra->emc_clock, rate);
> +	if (err)
> +		return err;
> +
> +	err = clk_set_rate(tegra->emc_clock, 0);
> +	if (err)
> +		goto restore_min_rate;
> +
> +	return 0;
> +
> +restore_min_rate:
> +	clk_set_min_rate(tegra->emc_clock, devfreq->previous_freq);
> +
> +	return err;
> +}
> +
> +static int tegra_devfreq_get_dev_status(struct device *dev,
> +					struct devfreq_dev_status *stat)
> +{
> +	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
> +
> +	/*
> +	 * EMC_COUNT returns number of memory events, that number is lower
> +	 * than the number of clocks. Conversion ratio of 1/8 results in a
> +	 * bit higher bandwidth than actually needed, it is good enough for
> +	 * the time being because drivers don't support requesting minimum
> +	 * needed memory bandwidth yet.
> +	 *
> +	 * TODO: adjust the ratio value once relevant drivers will support
> +	 * memory bandwidth management.
> +	 */
> +	stat->busy_time = readl_relaxed(tegra->regs + MC_STAT_EMC_COUNT);
> +	stat->total_time = readl_relaxed(tegra->regs + MC_STAT_EMC_CLOCKS) / 8;
> +	stat->current_frequency = clk_get_rate(tegra->emc_clock);
> +
> +	writel_relaxed(EMC_GATHER_CLEAR, tegra->regs + MC_STAT_CONTROL);
> +	writel_relaxed(EMC_GATHER_ENABLE, tegra->regs + MC_STAT_CONTROL);
> +
> +	return 0;
> +}
> +
> +static struct devfreq_dev_profile tegra_devfreq_profile = {
> +	.polling_ms	= 500,
> +	.target		= tegra_devfreq_target,
> +	.get_dev_status	= tegra_devfreq_get_dev_status,
> +};
> +
> +static struct tegra_mc *tegra_get_memory_controller(void)
> +{
> +	struct platform_device *pdev;
> +	struct device_node *np;
> +	struct tegra_mc *mc;
> +
> +	np = of_find_compatible_node(NULL, NULL, "nvidia,tegra20-mc-gart");
> +	if (!np)
> +		return ERR_PTR(-ENOENT);
> +
> +	pdev = of_find_device_by_node(np);
> +	of_node_put(np);
> +	if (!pdev)
> +		return ERR_PTR(-ENODEV);
> +
> +	mc = platform_get_drvdata(pdev);
> +	if (!mc)
> +		return ERR_PTR(-EPROBE_DEFER);
> +
> +	return mc;
> +}
> +
> +static int tegra_devfreq_probe(struct platform_device *pdev)
> +{
> +	struct tegra_devfreq *tegra;
> +	struct tegra_mc *mc;
> +	unsigned long max_rate;
> +	unsigned long rate;
> +	int err;
> +
> +	mc = tegra_get_memory_controller();
> +	if (IS_ERR(mc)) {
> +		err = PTR_ERR(mc);
> +		dev_err(&pdev->dev, "failed to get memory controller: %d\n",
> +			err);
> +		return err;
> +	}
> +
> +	tegra = devm_kzalloc(&pdev->dev, sizeof(*tegra), GFP_KERNEL);
> +	if (!tegra)
> +		return -ENOMEM;
> +
> +	/* EMC is a system-critical clock that is always enabled */
> +	tegra->emc_clock = devm_clk_get(&pdev->dev, "emc");
> +	if (IS_ERR(tegra->emc_clock)) {
> +		err = PTR_ERR(tegra->emc_clock);
> +		dev_err(&pdev->dev, "failed to get emc clock: %d\n", err);
> +		return err;
> +	}
> +
> +	tegra->regs = mc->regs;
> +
> +	max_rate = clk_round_rate(tegra->emc_clock, ULONG_MAX);
> +
> +	for (rate = 0; rate <= max_rate; rate++) {
> +		rate = clk_round_rate(tegra->emc_clock, rate);
> +
> +		err = dev_pm_opp_add(&pdev->dev, rate, 0);
> +		if (err) {
> +			dev_err(&pdev->dev, "failed to add opp: %d\n", err);
> +			goto remove_opps;
> +		}
> +	}
> +
> +	/*
> +	 * Reset statistic gathers state, select global bandwidth for the
> +	 * statistics collection mode and set clocks counter saturation
> +	 * limit to maximum.
> +	 */
> +	writel_relaxed(0x00000000, tegra->regs + MC_STAT_CONTROL);
> +	writel_relaxed(0x00000000, tegra->regs + MC_STAT_EMC_CONTROL);
> +	writel_relaxed(0xffffffff, tegra->regs + MC_STAT_EMC_CLOCK_LIMIT);
> +
> +	platform_set_drvdata(pdev, tegra);
> +
> +	tegra->devfreq = devfreq_add_device(&pdev->dev, &tegra_devfreq_profile,
> +					    DEVFREQ_GOV_SIMPLE_ONDEMAND, NULL);
> +	if (IS_ERR(tegra->devfreq)) {
> +		err = PTR_ERR(tegra->devfreq);
> +		goto remove_opps;
> +	}
> +
> +	return 0;
> +
> +remove_opps:
> +	dev_pm_opp_remove_all_dynamic(&pdev->dev);
> +
> +	return err;
> +}
> +
> +static int tegra_devfreq_remove(struct platform_device *pdev)
> +{
> +	struct tegra_devfreq *tegra = platform_get_drvdata(pdev);
> +
> +	devfreq_remove_device(tegra->devfreq);
> +	dev_pm_opp_remove_all_dynamic(&pdev->dev);
> +
> +	return 0;
> +}
> +
> +static struct platform_driver tegra_devfreq_driver = {
> +	.probe		= tegra_devfreq_probe,
> +	.remove		= tegra_devfreq_remove,
> +	.driver		= {
> +		.name	= "tegra20-devfreq",
> +	},
> +};
> +module_platform_driver(tegra_devfreq_driver);
> +
> +MODULE_ALIAS("platform:tegra20-devfreq");
> +MODULE_AUTHOR("Dmitry Osipenko <digetx@gmail.com>");
> +MODULE_DESCRIPTION("NVIDIA Tegra20 devfreq driver");
> +MODULE_LICENSE("GPL v2");
> -- 
> 2.21.0
>
Dmitry Osipenko June 4, 2019, 1:57 p.m. UTC | #2
04.06.2019 14:25, Thierry Reding пишет:
> On Thu, May 02, 2019 at 02:38:15AM +0300, Dmitry Osipenko wrote:
>> Add devfreq driver for NVIDIA Tegra20 SoC's. The driver periodically
>> reads out Memory Controller counters and adjusts memory frequency based
>> on the memory clients activity.
>>
>> Reviewed-by: Chanwoo Choi <cw00.choi@samsung.com>
>> Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
>> ---
>>  MAINTAINERS                       |   8 ++
>>  drivers/devfreq/Kconfig           |  10 ++
>>  drivers/devfreq/Makefile          |   1 +
>>  drivers/devfreq/tegra20-devfreq.c | 212 ++++++++++++++++++++++++++++++
>>  4 files changed, 231 insertions(+)
>>  create mode 100644 drivers/devfreq/tegra20-devfreq.c
>>
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index 98edc38bfd7b..e7e434f74038 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -10098,6 +10098,14 @@ F:	include/linux/memblock.h
>>  F:	mm/memblock.c
>>  F:	Documentation/core-api/boot-time-mm.rst
>>  
>> +MEMORY FREQUENCY SCALING DRIVER FOR NVIDIA TEGRA20
>> +M:	Dmitry Osipenko <digetx@gmail.com>
>> +L:	linux-pm@vger.kernel.org
>> +L:	linux-tegra@vger.kernel.org
>> +T:	git git://git.kernel.org/pub/scm/linux/kernel/git/mzx/devfreq.git
>> +S:	Maintained
>> +F:	drivers/devfreq/tegra20-devfreq.c
>> +
>>  MEMORY MANAGEMENT
>>  L:	linux-mm@kvack.org
>>  W:	http://www.linux-mm.org
>> diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig
>> index a6bba6e1e7d9..1530dbefa31f 100644
>> --- a/drivers/devfreq/Kconfig
>> +++ b/drivers/devfreq/Kconfig
>> @@ -100,6 +100,16 @@ config ARM_TEGRA_DEVFREQ
>>  	  It reads ACTMON counters of memory controllers and adjusts the
>>  	  operating frequencies and voltages with OPP support.
>>  
>> +config ARM_TEGRA20_DEVFREQ
>> +	tristate "NVIDIA Tegra20 DEVFREQ Driver"
>> +	depends on (TEGRA_MC && TEGRA20_EMC) || COMPILE_TEST
>> +	select DEVFREQ_GOV_SIMPLE_ONDEMAND
>> +	select PM_OPP
> 
> Again, I'm not sure the COMPILE_TEST will work here unless you add a few
> more dependencies.

I have the same answer as I made in the comment to "Enable COMPILE_TEST
for the driver" patch. I think there is no real need to be overreactive
here as well, ACK?
diff mbox series

Patch

diff --git a/MAINTAINERS b/MAINTAINERS
index 98edc38bfd7b..e7e434f74038 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -10098,6 +10098,14 @@  F:	include/linux/memblock.h
 F:	mm/memblock.c
 F:	Documentation/core-api/boot-time-mm.rst
 
+MEMORY FREQUENCY SCALING DRIVER FOR NVIDIA TEGRA20
+M:	Dmitry Osipenko <digetx@gmail.com>
+L:	linux-pm@vger.kernel.org
+L:	linux-tegra@vger.kernel.org
+T:	git git://git.kernel.org/pub/scm/linux/kernel/git/mzx/devfreq.git
+S:	Maintained
+F:	drivers/devfreq/tegra20-devfreq.c
+
 MEMORY MANAGEMENT
 L:	linux-mm@kvack.org
 W:	http://www.linux-mm.org
diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig
index a6bba6e1e7d9..1530dbefa31f 100644
--- a/drivers/devfreq/Kconfig
+++ b/drivers/devfreq/Kconfig
@@ -100,6 +100,16 @@  config ARM_TEGRA_DEVFREQ
 	  It reads ACTMON counters of memory controllers and adjusts the
 	  operating frequencies and voltages with OPP support.
 
+config ARM_TEGRA20_DEVFREQ
+	tristate "NVIDIA Tegra20 DEVFREQ Driver"
+	depends on (TEGRA_MC && TEGRA20_EMC) || COMPILE_TEST
+	select DEVFREQ_GOV_SIMPLE_ONDEMAND
+	select PM_OPP
+	help
+	  This adds the DEVFREQ driver for the Tegra20 family of SoCs.
+	  It reads Memory Controller counters and adjusts the operating
+	  frequencies and voltages with OPP support.
+
 config ARM_RK3399_DMC_DEVFREQ
 	tristate "ARM RK3399 DMC DEVFREQ Driver"
 	depends on ARCH_ROCKCHIP
diff --git a/drivers/devfreq/Makefile b/drivers/devfreq/Makefile
index 47e5aeeebfd1..338ae8440db6 100644
--- a/drivers/devfreq/Makefile
+++ b/drivers/devfreq/Makefile
@@ -11,6 +11,7 @@  obj-$(CONFIG_DEVFREQ_GOV_PASSIVE)	+= governor_passive.o
 obj-$(CONFIG_ARM_EXYNOS_BUS_DEVFREQ)	+= exynos-bus.o
 obj-$(CONFIG_ARM_RK3399_DMC_DEVFREQ)	+= rk3399_dmc.o
 obj-$(CONFIG_ARM_TEGRA_DEVFREQ)		+= tegra30-devfreq.o
+obj-$(CONFIG_ARM_TEGRA20_DEVFREQ)	+= tegra20-devfreq.o
 
 # DEVFREQ Event Drivers
 obj-$(CONFIG_PM_DEVFREQ_EVENT)		+= event/
diff --git a/drivers/devfreq/tegra20-devfreq.c b/drivers/devfreq/tegra20-devfreq.c
new file mode 100644
index 000000000000..ff82bac9ee4e
--- /dev/null
+++ b/drivers/devfreq/tegra20-devfreq.c
@@ -0,0 +1,212 @@ 
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * NVIDIA Tegra20 devfreq driver
+ *
+ * Copyright (C) 2019 GRATE-DRIVER project
+ */
+
+#include <linux/clk.h>
+#include <linux/devfreq.h>
+#include <linux/io.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/pm_opp.h>
+#include <linux/slab.h>
+
+#include <soc/tegra/mc.h>
+
+#include "governor.h"
+
+#define MC_STAT_CONTROL				0x90
+#define MC_STAT_EMC_CLOCK_LIMIT			0xa0
+#define MC_STAT_EMC_CLOCKS			0xa4
+#define MC_STAT_EMC_CONTROL			0xa8
+#define MC_STAT_EMC_COUNT			0xb8
+
+#define EMC_GATHER_CLEAR			(1 << 8)
+#define EMC_GATHER_ENABLE			(3 << 8)
+
+struct tegra_devfreq {
+	struct devfreq *devfreq;
+	struct clk *emc_clock;
+	void __iomem *regs;
+};
+
+static int tegra_devfreq_target(struct device *dev, unsigned long *freq,
+				u32 flags)
+{
+	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
+	struct devfreq *devfreq = tegra->devfreq;
+	struct dev_pm_opp *opp;
+	unsigned long rate;
+	int err;
+
+	opp = devfreq_recommended_opp(dev, freq, flags);
+	if (IS_ERR(opp))
+		return PTR_ERR(opp);
+
+	rate = dev_pm_opp_get_freq(opp);
+	dev_pm_opp_put(opp);
+
+	err = clk_set_min_rate(tegra->emc_clock, rate);
+	if (err)
+		return err;
+
+	err = clk_set_rate(tegra->emc_clock, 0);
+	if (err)
+		goto restore_min_rate;
+
+	return 0;
+
+restore_min_rate:
+	clk_set_min_rate(tegra->emc_clock, devfreq->previous_freq);
+
+	return err;
+}
+
+static int tegra_devfreq_get_dev_status(struct device *dev,
+					struct devfreq_dev_status *stat)
+{
+	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
+
+	/*
+	 * EMC_COUNT returns number of memory events, that number is lower
+	 * than the number of clocks. Conversion ratio of 1/8 results in a
+	 * bit higher bandwidth than actually needed, it is good enough for
+	 * the time being because drivers don't support requesting minimum
+	 * needed memory bandwidth yet.
+	 *
+	 * TODO: adjust the ratio value once relevant drivers will support
+	 * memory bandwidth management.
+	 */
+	stat->busy_time = readl_relaxed(tegra->regs + MC_STAT_EMC_COUNT);
+	stat->total_time = readl_relaxed(tegra->regs + MC_STAT_EMC_CLOCKS) / 8;
+	stat->current_frequency = clk_get_rate(tegra->emc_clock);
+
+	writel_relaxed(EMC_GATHER_CLEAR, tegra->regs + MC_STAT_CONTROL);
+	writel_relaxed(EMC_GATHER_ENABLE, tegra->regs + MC_STAT_CONTROL);
+
+	return 0;
+}
+
+static struct devfreq_dev_profile tegra_devfreq_profile = {
+	.polling_ms	= 500,
+	.target		= tegra_devfreq_target,
+	.get_dev_status	= tegra_devfreq_get_dev_status,
+};
+
+static struct tegra_mc *tegra_get_memory_controller(void)
+{
+	struct platform_device *pdev;
+	struct device_node *np;
+	struct tegra_mc *mc;
+
+	np = of_find_compatible_node(NULL, NULL, "nvidia,tegra20-mc-gart");
+	if (!np)
+		return ERR_PTR(-ENOENT);
+
+	pdev = of_find_device_by_node(np);
+	of_node_put(np);
+	if (!pdev)
+		return ERR_PTR(-ENODEV);
+
+	mc = platform_get_drvdata(pdev);
+	if (!mc)
+		return ERR_PTR(-EPROBE_DEFER);
+
+	return mc;
+}
+
+static int tegra_devfreq_probe(struct platform_device *pdev)
+{
+	struct tegra_devfreq *tegra;
+	struct tegra_mc *mc;
+	unsigned long max_rate;
+	unsigned long rate;
+	int err;
+
+	mc = tegra_get_memory_controller();
+	if (IS_ERR(mc)) {
+		err = PTR_ERR(mc);
+		dev_err(&pdev->dev, "failed to get memory controller: %d\n",
+			err);
+		return err;
+	}
+
+	tegra = devm_kzalloc(&pdev->dev, sizeof(*tegra), GFP_KERNEL);
+	if (!tegra)
+		return -ENOMEM;
+
+	/* EMC is a system-critical clock that is always enabled */
+	tegra->emc_clock = devm_clk_get(&pdev->dev, "emc");
+	if (IS_ERR(tegra->emc_clock)) {
+		err = PTR_ERR(tegra->emc_clock);
+		dev_err(&pdev->dev, "failed to get emc clock: %d\n", err);
+		return err;
+	}
+
+	tegra->regs = mc->regs;
+
+	max_rate = clk_round_rate(tegra->emc_clock, ULONG_MAX);
+
+	for (rate = 0; rate <= max_rate; rate++) {
+		rate = clk_round_rate(tegra->emc_clock, rate);
+
+		err = dev_pm_opp_add(&pdev->dev, rate, 0);
+		if (err) {
+			dev_err(&pdev->dev, "failed to add opp: %d\n", err);
+			goto remove_opps;
+		}
+	}
+
+	/*
+	 * Reset statistic gathers state, select global bandwidth for the
+	 * statistics collection mode and set clocks counter saturation
+	 * limit to maximum.
+	 */
+	writel_relaxed(0x00000000, tegra->regs + MC_STAT_CONTROL);
+	writel_relaxed(0x00000000, tegra->regs + MC_STAT_EMC_CONTROL);
+	writel_relaxed(0xffffffff, tegra->regs + MC_STAT_EMC_CLOCK_LIMIT);
+
+	platform_set_drvdata(pdev, tegra);
+
+	tegra->devfreq = devfreq_add_device(&pdev->dev, &tegra_devfreq_profile,
+					    DEVFREQ_GOV_SIMPLE_ONDEMAND, NULL);
+	if (IS_ERR(tegra->devfreq)) {
+		err = PTR_ERR(tegra->devfreq);
+		goto remove_opps;
+	}
+
+	return 0;
+
+remove_opps:
+	dev_pm_opp_remove_all_dynamic(&pdev->dev);
+
+	return err;
+}
+
+static int tegra_devfreq_remove(struct platform_device *pdev)
+{
+	struct tegra_devfreq *tegra = platform_get_drvdata(pdev);
+
+	devfreq_remove_device(tegra->devfreq);
+	dev_pm_opp_remove_all_dynamic(&pdev->dev);
+
+	return 0;
+}
+
+static struct platform_driver tegra_devfreq_driver = {
+	.probe		= tegra_devfreq_probe,
+	.remove		= tegra_devfreq_remove,
+	.driver		= {
+		.name	= "tegra20-devfreq",
+	},
+};
+module_platform_driver(tegra_devfreq_driver);
+
+MODULE_ALIAS("platform:tegra20-devfreq");
+MODULE_AUTHOR("Dmitry Osipenko <digetx@gmail.com>");
+MODULE_DESCRIPTION("NVIDIA Tegra20 devfreq driver");
+MODULE_LICENSE("GPL v2");