diff mbox

[PATCHv2,3/3] clk: keystone: Add sci-clk driver support

Message ID 1472733635-22661-4-git-send-email-t-kristo@ti.com (mailing list archive)
State Changes Requested, archived
Delegated to: Stephen Boyd
Headers show

Commit Message

Tero Kristo Sept. 1, 2016, 12:40 p.m. UTC
In K2G, the clock handling is done through firmware executing on a
separate core. Linux kernel needs to communicate to the firmware
through TI system control interface to access any power management
related resources, including clocks.

The keystone sci-clk driver does this, by communicating to the
firmware through the TI SCI driver. The driver adds support for
registering clocks through DT, and basic required clock operations
like prepare/get_rate, etc.

Signed-off-by: Tero Kristo <t-kristo@ti.com>
---
 MAINTAINERS                    |   1 +
 drivers/clk/Kconfig            |   9 +
 drivers/clk/keystone/Makefile  |   1 +
 drivers/clk/keystone/sci-clk.c | 537 +++++++++++++++++++++++++++++++++++++++++
 4 files changed, 548 insertions(+)
 create mode 100644 drivers/clk/keystone/sci-clk.c

Comments

Stephen Boyd Sept. 2, 2016, 11:32 p.m. UTC | #1
On 09/01, Tero Kristo wrote:
> +
> +/**
> + * _sci_clk_get - Gets a handle for an SCI clock
> + * @provider: Handle to SCI clock provider
> + * @dev_id: device ID for the clock to register
> + * @clk_id: clock ID for the clock to register
> + * @parse_parents: indicator whether parents for this clock should be handled
> + *
> + * Gets a handle to an existing TI SCI hw clock, or builds a new clock
> + * entry and registers it with the common clock framework. Called from
> + * the common clock framework, when a corresponding of_clk_get call is
> + * executed, or recursively from itself when parsing parent clocks.
> + * Returns a pointer to the hw clock struct, or ERR_PTR value in failure.
> + */
> +static struct clk_hw *_sci_clk_get(struct sci_clk_provider *provider,
> +				   u16 dev_id, u8 clk_id, bool parse_parents)

I still don't get it. We should be registering hw pointers in
probe and handing them out in the xlate function. Not register
hws in the xlate function and then handing them out as well. I
haven't reviewed anything else in this patch.
Tero Kristo Sept. 5, 2016, 7:07 a.m. UTC | #2
On 03/09/16 02:32, Stephen Boyd wrote:
> On 09/01, Tero Kristo wrote:
>> +
>> +/**
>> + * _sci_clk_get - Gets a handle for an SCI clock
>> + * @provider: Handle to SCI clock provider
>> + * @dev_id: device ID for the clock to register
>> + * @clk_id: clock ID for the clock to register
>> + * @parse_parents: indicator whether parents for this clock should be handled
>> + *
>> + * Gets a handle to an existing TI SCI hw clock, or builds a new clock
>> + * entry and registers it with the common clock framework. Called from
>> + * the common clock framework, when a corresponding of_clk_get call is
>> + * executed, or recursively from itself when parsing parent clocks.
>> + * Returns a pointer to the hw clock struct, or ERR_PTR value in failure.
>> + */
>> +static struct clk_hw *_sci_clk_get(struct sci_clk_provider *provider,
>> +				   u16 dev_id, u8 clk_id, bool parse_parents)
>
> I still don't get it. We should be registering hw pointers in
> probe and handing them out in the xlate function. Not register
> hws in the xlate function and then handing them out as well. I
> haven't reviewed anything else in this patch.

Ok, let me try to explain the functionality.

Probe time hw pointer registration is only needed for special clocks 
that require extra flags, like adding the spread spectrum flag. There is 
ever going to be only handful of these.

Xlate checks out the sci_clk list, and picks up an existing hw clock if 
it is there. If not, it will create a new one. The reason this is done 
like this, the device IDs / clock IDs don't mean anything to the driver 
itself, the driver just passes these forward to the underlying SCI fwk. 
And, we don't have inherent knowledge of valid device / clock IDs 
either, the SCI fwk returns a failure if a specific clock ID is bad. The 
device / clock IDs are going to be different between different 
generations of SoC also, and in addition there can be holes in the ID 
ranges.

-Tero
--
To unsubscribe from this list: send the line "unsubscribe linux-clk" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Stephen Boyd Sept. 16, 2016, 11:01 p.m. UTC | #3
On 09/05, Tero Kristo wrote:
> On 03/09/16 02:32, Stephen Boyd wrote:
> >I still don't get it. We should be registering hw pointers in
> >probe and handing them out in the xlate function. Not register
> >hws in the xlate function and then handing them out as well. I
> >haven't reviewed anything else in this patch.
> 
> Ok, let me try to explain the functionality.
> 
> Probe time hw pointer registration is only needed for special clocks
> that require extra flags, like adding the spread spectrum flag.
> There is ever going to be only handful of these.
> 
> Xlate checks out the sci_clk list, and picks up an existing hw clock
> if it is there. If not, it will create a new one. The reason this is
> done like this, the device IDs / clock IDs don't mean anything to
> the driver itself, the driver just passes these forward to the
> underlying SCI fwk. And, we don't have inherent knowledge of valid
> device / clock IDs either, the SCI fwk returns a failure if a
> specific clock ID is bad. The device / clock IDs are going to be
> different between different generations of SoC also, and in addition
> there can be holes in the ID ranges.

Ok. Thanks for the explanation.

I understand the desire to make this SoC agnostic and consumer/dt
driven. Constructor pattern and all. Unfortunately the OF clk
provider is a getter pattern; not a constructor pattern. I'm
seriously concerned about the locking here because we're in the
framework creating a clk and tying it into a consumer list which
could cause all sorts of problems if we want to reenter the
framework and register more clks. Recursion abounds and my head
explodes! The recursive clk prepare mutex is not something anyone
should be thrilled about keeping around.

So, just don't do this. Instead, register the clks during probe
that exist for the firmware. That list could be discoverable with
firmware calls, or known based on different compatible strings
from DT that have the soc name in it, or something else. Even
scanning the entire DT tree for

	clocks = <&my_clk_node x y>
	
would be better, but probably hugely inefficient and outright
buggy with DT overlays so don't do it. If the firmware can't tell
us what clks are there, just have a table based on compatible
string and be done.
Tero Kristo Sept. 23, 2016, 8:06 a.m. UTC | #4
On 17/09/16 02:01, Stephen Boyd wrote:
> On 09/05, Tero Kristo wrote:
>> On 03/09/16 02:32, Stephen Boyd wrote:
>>> I still don't get it. We should be registering hw pointers in
>>> probe and handing them out in the xlate function. Not register
>>> hws in the xlate function and then handing them out as well. I
>>> haven't reviewed anything else in this patch.
>>
>> Ok, let me try to explain the functionality.
>>
>> Probe time hw pointer registration is only needed for special clocks
>> that require extra flags, like adding the spread spectrum flag.
>> There is ever going to be only handful of these.
>>
>> Xlate checks out the sci_clk list, and picks up an existing hw clock
>> if it is there. If not, it will create a new one. The reason this is
>> done like this, the device IDs / clock IDs don't mean anything to
>> the driver itself, the driver just passes these forward to the
>> underlying SCI fwk. And, we don't have inherent knowledge of valid
>> device / clock IDs either, the SCI fwk returns a failure if a
>> specific clock ID is bad. The device / clock IDs are going to be
>> different between different generations of SoC also, and in addition
>> there can be holes in the ID ranges.
>
> Ok. Thanks for the explanation.
>
> I understand the desire to make this SoC agnostic and consumer/dt
> driven. Constructor pattern and all. Unfortunately the OF clk
> provider is a getter pattern; not a constructor pattern. I'm
> seriously concerned about the locking here because we're in the
> framework creating a clk and tying it into a consumer list which
> could cause all sorts of problems if we want to reenter the
> framework and register more clks. Recursion abounds and my head
> explodes! The recursive clk prepare mutex is not something anyone
> should be thrilled about keeping around.
>
> So, just don't do this. Instead, register the clks during probe
> that exist for the firmware. That list could be discoverable with
> firmware calls, or known based on different compatible strings
> from DT that have the soc name in it, or something else. Even
> scanning the entire DT tree for
>
> 	clocks = <&my_clk_node x y>
> 	
> would be better, but probably hugely inefficient and outright
> buggy with DT overlays so don't do it. If the firmware can't tell
> us what clks are there, just have a table based on compatible
> string and be done.

Ok, I will add a static mapping of valid clock IDs within the driver and 
init all of these during probe. I have a working solution for this 
already, but will post it out only once 4.9-rc1 is out.

-Tero
--
To unsubscribe from this list: send the line "unsubscribe linux-clk" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/MAINTAINERS b/MAINTAINERS
index bcb7b9e..2ed4d38 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11575,6 +11575,7 @@  F:	drivers/firmware/ti_sci*
 F:	include/linux/soc/ti/ti_sci_protocol.h
 F:	Documentation/devicetree/bindings/clock/ti,sci-clk.txt
 F:	include/dt-bindings/clock/k2g.h
+F:	drivers/clk/keystone/sci-clk.c
 
 THANKO'S RAREMONO AM/FM/SW RADIO RECEIVER USB DRIVER
 M:	Hans Verkuil <hverkuil@xs4all.nl>
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index e2d9bd7..dbe792a7 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -179,6 +179,15 @@  config COMMON_CLK_NXP
 	---help---
 	  Support for clock providers on NXP platforms.
 
+config TI_SCI_CLK
+	tristate "TI System Control Interface clock drivers"
+	depends on (TI_SCI_PROTOCOL && COMMON_CLK_KEYSTONE) || COMPILE_TEST
+	default TI_SCI_PROTOCOL
+	---help---
+	  This adds the clock driver support over TI System Control Interface.
+	  If you wish to use clock resources from the PMMC firmware, say Y.
+	  Otherwise, say N.
+
 config COMMON_CLK_PALMAS
 	tristate "Clock driver for TI Palmas devices"
 	depends on MFD_PALMAS
diff --git a/drivers/clk/keystone/Makefile b/drivers/clk/keystone/Makefile
index 0477cf6..0e7993d 100644
--- a/drivers/clk/keystone/Makefile
+++ b/drivers/clk/keystone/Makefile
@@ -1 +1,2 @@ 
 obj-y			+= pll.o gate.o
+obj-$(CONFIG_TI_SCI_CLK)	+= sci-clk.o
diff --git a/drivers/clk/keystone/sci-clk.c b/drivers/clk/keystone/sci-clk.c
new file mode 100644
index 0000000..852fc49
--- /dev/null
+++ b/drivers/clk/keystone/sci-clk.c
@@ -0,0 +1,537 @@ 
+/*
+ * SCI Clock driver for keystone based devices
+ *
+ * Copyright (C) 2015-2016 Texas Instruments Incorporated - http://www.ti.com/
+ *	Tero Kristo <t-kristo@ti.com>
+ *
+ * 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.
+ *
+ * This program is distributed "as is" WITHOUT ANY WARRANTY of any
+ * kind, whether express or implied; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+#include <linux/clk-provider.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/soc/ti/ti_sci_protocol.h>
+
+#define SCI_CLK_SSC_ENABLE		BIT(0)
+#define SCI_CLK_ALLOW_FREQ_CHANGE	BIT(1)
+#define SCI_CLK_INPUT_TERMINATION	BIT(2)
+
+/**
+ * struct sci_clk_provider - TI SCI clock provider representation
+ * @sci:    Handle to the System Control Interface protocol handler
+ * @ops:    Pointer to the SCI ops to be used by the clocks
+ * @dev:    Device pointer for the clock provider
+ * @clocks: List of all registered clocks
+ * @lock:   Mutex for locking access to the @clocks list
+ */
+struct sci_clk_provider {
+	const struct ti_sci_handle *sci;
+	const struct ti_sci_clk_ops *ops;
+	struct device *dev;
+	struct list_head clocks;
+	struct mutex lock; /* Protects access to the @clocks list */
+};
+
+/**
+ * struct sci_clk - TI SCI clock representation
+ * @hw:		 Hardware clock cookie for common clock framework
+ * @dev_id:	 Device index
+ * @clk_id:	 Clock index
+ * @node:	 Clocks list link
+ * @provider:	 Master clock provider
+ * @flags:	 Flags for the clock
+ */
+struct sci_clk {
+	struct clk_hw hw;
+	u16 dev_id;
+	u8 clk_id;
+	struct list_head node;
+	struct sci_clk_provider *provider;
+	u8 flags;
+};
+
+#define to_sci_clk(_hw) container_of(_hw, struct sci_clk, hw)
+
+/**
+ * sci_clk_prepare - Prepare (enable) a TI SCI clock
+ * @hw: clock to prepare
+ *
+ * Prepares a clock to be actively used. Returns the SCI protocol status.
+ */
+static int sci_clk_prepare(struct clk_hw *hw)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	bool enable_ssc = clk->flags & SCI_CLK_SSC_ENABLE;
+	bool allow_freq_change = clk->flags & SCI_CLK_ALLOW_FREQ_CHANGE;
+	bool input_termination = clk->flags & SCI_CLK_INPUT_TERMINATION;
+
+	return clk->provider->ops->get_clock(clk->provider->sci, clk->dev_id,
+					     clk->clk_id, enable_ssc,
+					     allow_freq_change,
+					     input_termination);
+}
+
+/**
+ * sci_clk_unprepare - Un-prepares (disables) a TI SCI clock
+ * @hw: clock to unprepare
+ *
+ * Un-prepares a clock from active state.
+ */
+static void sci_clk_unprepare(struct clk_hw *hw)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	int ret;
+
+	ret = clk->provider->ops->put_clock(clk->provider->sci, clk->dev_id,
+					    clk->clk_id);
+	if (ret)
+		dev_err(clk->provider->dev,
+			"unprepare failed for dev=%d, clk=%d, ret=%d\n",
+			clk->dev_id, clk->clk_id, ret);
+}
+
+/**
+ * sci_clk_is_prepared - Check if a TI SCI clock is prepared or not
+ * @hw: clock to check status for
+ *
+ * Checks if a clock is prepared (enabled) in hardware. Returns non-zero
+ * value if clock is enabled, zero otherwise.
+ */
+static int sci_clk_is_prepared(struct clk_hw *hw)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	bool req_state, current_state;
+	int ret;
+
+	ret = clk->provider->ops->is_on(clk->provider->sci, clk->dev_id,
+					clk->clk_id, &req_state,
+					&current_state);
+	if (ret) {
+		dev_err(clk->provider->dev,
+			"is_prepared failed for dev=%d, clk=%d, ret=%d\n",
+			clk->dev_id, clk->clk_id, ret);
+		return 0;
+	}
+
+	return req_state;
+}
+
+/**
+ * sci_clk_recalc_rate - Get clock rate for a TI SCI clock
+ * @hw: clock to get rate for
+ * @parent_rate: parent rate provided by common clock framework, not used
+ *
+ * Gets the current clock rate of a TI SCI clock. Returns the current
+ * clock rate, or zero in failure.
+ */
+static unsigned long sci_clk_recalc_rate(struct clk_hw *hw,
+					 unsigned long parent_rate)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	u64 freq;
+	int ret;
+
+	ret = clk->provider->ops->get_freq(clk->provider->sci, clk->dev_id,
+					   clk->clk_id, &freq);
+	if (ret) {
+		dev_err(clk->provider->dev,
+			"recalc-rate failed for dev=%d, clk=%d, ret=%d\n",
+			clk->dev_id, clk->clk_id, ret);
+		return 0;
+	}
+
+	return (u32)freq;
+}
+
+/**
+ * sci_clk_determine_rate - Determines a clock rate a clock can be set to
+ * @hw: clock to change rate for
+ * @req: requested rate configuration for the clock
+ *
+ * Determines a suitable clock rate and parent for a TI SCI clock.
+ * The parent handling is un-used, as generally the parent clock rates
+ * are not known by the kernel; instead these are internally handled
+ * by the firmware. Returns 0 on success, negative error value on failure.
+ */
+static int sci_clk_determine_rate(struct clk_hw *hw,
+				  struct clk_rate_request *req)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	int ret;
+	u64 new_rate;
+
+	ret = clk->provider->ops->get_best_match_freq(clk->provider->sci,
+						      clk->dev_id,
+						      clk->clk_id,
+						      req->min_rate,
+						      req->rate,
+						      req->max_rate,
+						      &new_rate);
+	if (ret) {
+		dev_err(clk->provider->dev,
+			"determine-rate failed for dev=%d, clk=%d, ret=%d\n",
+			clk->dev_id, clk->clk_id, ret);
+		return ret;
+	}
+
+	req->rate = new_rate;
+
+	return 0;
+}
+
+/**
+ * sci_clk_set_rate - Set rate for a TI SCI clock
+ * @hw: clock to change rate for
+ * @rate: target rate for the clock
+ * @parent_rate: rate of the clock parent, not used for TI SCI clocks
+ *
+ * Sets a clock frequency for a TI SCI clock. Returns the TI SCI
+ * protocol status.
+ */
+static int sci_clk_set_rate(struct clk_hw *hw, unsigned long rate,
+			    unsigned long parent_rate)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+
+	return clk->provider->ops->set_freq(clk->provider->sci, clk->dev_id,
+					    clk->clk_id, rate, rate, rate);
+}
+
+/**
+ * sci_clk_get_parent - Get the current parent of a TI SCI clock
+ * @hw: clock to get parent for
+ *
+ * Returns the index of the currently selected parent for a TI SCI clock.
+ */
+static u8 sci_clk_get_parent(struct clk_hw *hw)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+	u8 parent_id;
+	int ret;
+
+	ret = clk->provider->ops->get_parent(clk->provider->sci, clk->dev_id,
+					     clk->clk_id, &parent_id);
+	if (ret) {
+		dev_err(clk->provider->dev,
+			"get-parent failed for dev=%d, clk=%d, ret=%d\n",
+			clk->dev_id, clk->clk_id, ret);
+		return 0;
+	}
+
+	return parent_id - clk->clk_id - 1;
+}
+
+/**
+ * sci_clk_set_parent - Set the parent of a TI SCI clock
+ * @hw: clock to set parent for
+ * @index: new parent index for the clock
+ *
+ * Sets the parent of a TI SCI clock. Return TI SCI protocol status.
+ */
+static int sci_clk_set_parent(struct clk_hw *hw, u8 index)
+{
+	struct sci_clk *clk = to_sci_clk(hw);
+
+	return clk->provider->ops->set_parent(clk->provider->sci, clk->dev_id,
+					      clk->clk_id,
+					      index + 1 + clk->clk_id);
+}
+
+static const struct clk_ops sci_clk_ops = {
+	.prepare = sci_clk_prepare,
+	.unprepare = sci_clk_unprepare,
+	.is_prepared = sci_clk_is_prepared,
+	.recalc_rate = sci_clk_recalc_rate,
+	.determine_rate = sci_clk_determine_rate,
+	.set_rate = sci_clk_set_rate,
+	.get_parent = sci_clk_get_parent,
+	.set_parent = sci_clk_set_parent,
+};
+
+/**
+ * _sci_clk_get - Gets a handle for an SCI clock
+ * @provider: Handle to SCI clock provider
+ * @dev_id: device ID for the clock to register
+ * @clk_id: clock ID for the clock to register
+ * @parse_parents: indicator whether parents for this clock should be handled
+ *
+ * Gets a handle to an existing TI SCI hw clock, or builds a new clock
+ * entry and registers it with the common clock framework. Called from
+ * the common clock framework, when a corresponding of_clk_get call is
+ * executed, or recursively from itself when parsing parent clocks.
+ * Returns a pointer to the hw clock struct, or ERR_PTR value in failure.
+ */
+static struct clk_hw *_sci_clk_get(struct sci_clk_provider *provider,
+				   u16 dev_id, u8 clk_id, bool parse_parents)
+{
+	struct clk_init_data init = { NULL };
+	struct sci_clk *sci_clk = NULL;
+	char *name = NULL;
+	char **parent_names = NULL;
+	int i;
+	int ret;
+
+	list_for_each_entry(sci_clk, &provider->clocks, node)
+		if (sci_clk->dev_id == dev_id && sci_clk->clk_id == clk_id)
+			return &sci_clk->hw;
+
+	sci_clk = devm_kzalloc(provider->dev, sizeof(*sci_clk), GFP_KERNEL);
+	if (!sci_clk) {
+		ret = -ENOMEM;
+		goto err;
+	}
+
+	sci_clk->dev_id = dev_id;
+	sci_clk->clk_id = clk_id;
+	sci_clk->provider = provider;
+
+	if (parse_parents) {
+		ret = provider->ops->get_num_parents(provider->sci, dev_id,
+						     clk_id,
+						     &init.num_parents);
+		if (ret)
+			goto err;
+	}
+
+	name = kasprintf(GFP_KERNEL, "%s:%d:%d", dev_name(provider->dev),
+			 sci_clk->dev_id, sci_clk->clk_id);
+
+	init.name = name;
+
+	if (init.num_parents < 2)
+		init.num_parents = 0;
+
+	if (init.num_parents) {
+		parent_names = devm_kcalloc(provider->dev, init.num_parents,
+					    sizeof(char *), GFP_KERNEL);
+
+		if (!parent_names) {
+			ret = -ENOMEM;
+			goto err;
+		}
+
+		for (i = 0; i < init.num_parents; i++) {
+			char *parent_name;
+
+			parent_name = devm_kzalloc(provider->dev, 20,
+						   GFP_KERNEL);
+			if (!parent_name) {
+				ret = -ENOMEM;
+				goto err;
+			}
+			snprintf(parent_name, 20, "%s:%d:%d",
+				 dev_name(provider->dev), sci_clk->dev_id,
+				 sci_clk->clk_id + 1 + i);
+			parent_names[i] = parent_name;
+
+			_sci_clk_get(provider, dev_id, clk_id + 1 + i, false);
+		}
+		init.parent_names = (const char * const *)parent_names;
+	}
+
+	init.ops = &sci_clk_ops;
+	sci_clk->hw.init = &init;
+
+	ret = devm_clk_hw_register(provider->dev, &sci_clk->hw);
+	if (ret) {
+		dev_err(provider->dev, "failed clk register with %d\n", ret);
+		goto err;
+	} else {
+		list_add(&sci_clk->node, &provider->clocks);
+	}
+
+	kfree(name);
+
+	return &sci_clk->hw;
+
+err:
+	if (parent_names) {
+		for (i = 0; i < init.num_parents; i++)
+			devm_kfree(provider->dev, parent_names[i]);
+
+		devm_kfree(provider->dev, parent_names);
+	}
+
+	devm_kfree(provider->dev, sci_clk);
+
+	kfree(name);
+
+	return ERR_PTR(ret);
+}
+
+/**
+ * sci_clk_get - Xlate function for getting clock handles
+ * @clkspec: device tree clock specifier
+ * @data: pointer to the clock provider
+ *
+ * Xlate function for retrieving clock TI SCI hw clock handles based on
+ * device tree clock specifier. Called from the common clock framework,
+ * when a corresponding of_clk_get call is executed. Returns a pointer
+ * to the TI SCI hw clock struct, or ERR_PTR value in failure.
+ */
+static struct clk_hw *sci_clk_get(struct of_phandle_args *clkspec, void *data)
+{
+	struct sci_clk_provider *provider = data;
+	struct clk_hw *clk;
+	u16 dev_id;
+	u8 clk_id;
+
+	if (clkspec->args_count != 2)
+		return ERR_PTR(-EINVAL);
+
+	mutex_lock(&provider->lock);
+
+	dev_id = clkspec->args[0];
+	clk_id = clkspec->args[1];
+
+	clk = _sci_clk_get(provider, dev_id, clk_id, true);
+
+	mutex_unlock(&provider->lock);
+
+	return clk;
+}
+
+static const struct of_device_id ti_sci_clk_of_match[] = {
+	{ .compatible = "ti,sci-clk" },
+	{ /* Sentinel */ },
+};
+MODULE_DEVICE_TABLE(of, ti_sci_clk_of_match);
+
+/**
+ * ti_sci_clk_parse_flags - Helper function to parse clock flag arrays
+ * @provider: clock provider handle
+ * @list_name: property name containing the clocks list
+ * @flag: flag to apply to the list of clocks
+ *
+ * Parses a DT list based on the provided data, and applies the flag value
+ * for each of the clocks identified by the list. Return 0 for success,
+ * negative error value for failure.
+ */
+static int ti_sci_clk_parse_flags(struct sci_clk_provider *provider,
+				  const char *list_name, u8 flag)
+{
+	int num_clks;
+	int i;
+	struct device_node *np = provider->dev->of_node;
+	struct device *dev = provider->dev;
+
+	num_clks = of_count_phandle_with_args(np, list_name, "#clock-cells");
+
+	for (i = 0; i < num_clks; i++) {
+		struct of_phandle_args clkspec;
+		struct sci_clk *sci_clk;
+		struct clk_hw *clk;
+		int ret;
+
+		ret = of_parse_phandle_with_args(np, list_name, "#clock-cells",
+						 i, &clkspec);
+		if (ret) {
+			dev_err(dev, "Failed to parse %s[%d] = %d\n", list_name,
+				i, ret);
+			return ret;
+		}
+		clk = sci_clk_get(&clkspec, provider);
+		if (IS_ERR(clk)) {
+			dev_err(dev, "clk_get failed for %s[%d] = %ld\n",
+				list_name, i, PTR_ERR(clk));
+			return PTR_ERR(clk);
+		}
+		sci_clk = to_sci_clk(clk);
+
+		sci_clk->flags |= flag;
+	}
+
+	return 0;
+}
+
+/**
+ * ti_sci_clk_probe - Probe function for the TI SCI clock driver
+ * @pdev: platform device pointer to be probed
+ *
+ * Probes the TI SCI clock device. Allocates a new clock provider
+ * and registers this to the common clock framework. Also applies
+ * any required flags to the identified clocks via clock lists
+ * supplied from DT. Returns 0 for success, negative error value
+ * for failure.
+ */
+static int ti_sci_clk_probe(struct platform_device *pdev)
+{
+	struct device *dev = &pdev->dev;
+	struct device_node *np = dev->of_node;
+	struct sci_clk_provider *provider;
+	const struct ti_sci_handle *handle;
+	int ret;
+
+	if (!np) {
+		dev_err(dev, "OF data missing\n");
+		return -EINVAL;
+	}
+
+	handle = devm_ti_sci_get_handle(dev);
+	if (IS_ERR(handle))
+		return PTR_ERR(handle);
+
+	provider = devm_kzalloc(dev, sizeof(*provider), GFP_KERNEL);
+	if (!provider)
+		return -ENOMEM;
+
+	INIT_LIST_HEAD(&provider->clocks);
+	mutex_init(&provider->lock);
+
+	provider->sci = handle;
+	provider->ops = &handle->ops.clk_ops;
+	provider->dev = dev;
+
+	ret = of_clk_add_hw_provider(np, sci_clk_get, provider);
+	if (ret)
+		return ret;
+
+	ti_sci_clk_parse_flags(provider, "ti,ssc-clocks", SCI_CLK_SSC_ENABLE);
+	ti_sci_clk_parse_flags(provider, "ti,allow-freq-change-clocks",
+			       SCI_CLK_ALLOW_FREQ_CHANGE);
+	ti_sci_clk_parse_flags(provider, "ti,input-term-clocks",
+			       SCI_CLK_INPUT_TERMINATION);
+
+	return 0;
+}
+
+/**
+ * ti_sci_clk_remove - Remove TI SCI clock device
+ * @pdev: platform device pointer for the device to be removed
+ *
+ * Removes the TI SCI device. Unregisters the clock provider registered
+ * via common clock framework. Any memory allocated for the device will
+ * be free'd silently via the devm framework. Returns 0 always.
+ */
+static int ti_sci_clk_remove(struct platform_device *pdev)
+{
+	of_clk_del_provider(pdev->dev.of_node);
+
+	return 0;
+}
+
+static struct platform_driver ti_sci_clk_driver = {
+	.probe = ti_sci_clk_probe,
+	.remove = ti_sci_clk_remove,
+	.driver = {
+		.name = "ti-sci-clk",
+		.of_match_table = of_match_ptr(ti_sci_clk_of_match),
+	},
+};
+module_platform_driver(ti_sci_clk_driver);
+
+MODULE_LICENSE("GPL v2");
+MODULE_DESCRIPTION("TI System Control Interface(SCI) Clock driver");
+MODULE_AUTHOR("Tero Kristo");
+MODULE_ALIAS("platform:ti-sci-clk");