[PATCH v2 0/3] i2c: designware: Add support for AMD I2C controller

[PATCH v2 0/3] i2c: designware: Add support for AMD I2C controller

[Mika Westerberg]

This is second iteration of the patch series. First version is here:

 http://patchwork.ozlabs.org/patch/390694/
 http://patchwork.ozlabs.org/patch/390695/

In order to get source clock to the core driver we need some way of
creating it in the first place. For Intel LPSS it is done in
drivers/acpi/acpi_lpss.c but that is not applicable for AMD I2C host
controller which is too different from private parts to the LPSS one.

We solve this by creating the clock in ACPI parts of the platform driver
when we detect the AMD I2C host controller. Doing this requires small
refactoring to be done to the probe() which is what patches [1-2/3] are
doing.

Changes to the previous version:
 * Rebased on top of i2c/for-next branch
 * New patch defaulting ACPI probe to fast mode
 * Instead of direct dependency to COMMON_CLK we depend on (ACPI &&
   COMMON_CLK) || ACPI. (suggested by Wolfram)

Carl, can you test that this still works on your machine?

Carl Peng (1):
  i2c: designware: Add support for AMD I2C controller

Mika Westerberg (2):
  i2c: designware: Default to fast mode in case of ACPI
  i2c: designware: Rework probe() to get clock a bit later

 drivers/i2c/busses/Kconfig                  |  1 +
 drivers/i2c/busses/i2c-designware-platdrv.c | 71 +++++++++++++++++++----------
 2 files changed, 48 insertions(+), 24 deletions(-)

-- 
2.1.0

[PATCH v2 3/3] i2c: designware: Add support for AMD I2C controller

[Mika Westerberg]

From: Carl Peng <carlpeng008@gmail.com>

Add support for AMD version of the DW I2C host controller. The device is
enumerated from ACPI namespace with ACPI ID AMD0010. Because the core
driver needs an input source clock, and this is not an Intel LPSS device
where clocks are provided through drivers/acpi/acpi_lpss.c, we register the
clock ourselves if the clock rate is given in ->driver_data

Signed-off-by: Carl Peng <carlpeng008@gmail.com>
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
---
 drivers/i2c/busses/Kconfig                  |  1 +
 drivers/i2c/busses/i2c-designware-platdrv.c | 27 +++++++++++++++++++++++++++
 2 files changed, 28 insertions(+)

diff --git a/drivers/i2c/busses/Kconfig b/drivers/i2c/busses/Kconfig
index 2ac87fa3058d..0715a239c87f 100644
--- a/drivers/i2c/busses/Kconfig
+++ b/drivers/i2c/busses/Kconfig
@@ -423,6 +423,7 @@ config I2C_DESIGNWARE_CORE
 config I2C_DESIGNWARE_PLATFORM
 	tristate "Synopsys DesignWare Platform"
 	select I2C_DESIGNWARE_CORE
+	depends on (ACPI && COMMON_CLK) || !ACPI
 	help
 	  If you say yes to this option, support will be included for the
 	  Synopsys DesignWare I2C adapter. Only master mode is supported.
diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index e59c63ae4d41..a7431150acf7 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -30,6 +30,7 @@
 #include <linux/delay.h>
 #include <linux/i2c.h>
 #include <linux/clk.h>
+#include <linux/clk-provider.h>
 #include <linux/errno.h>
 #include <linux/sched.h>
 #include <linux/err.h>
@@ -80,6 +81,7 @@ static void dw_i2c_acpi_params(struct platform_device *pdev, char method[],
 static int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
+	const struct acpi_device_id *id;
 
 	dev->adapter.nr = -1;
 	dev->tx_fifo_depth = 32;
@@ -93,9 +95,29 @@ static int dw_i2c_acpi_configure(struct platform_device *pdev)
 	dw_i2c_acpi_params(pdev, "FMCN", &dev->fs_hcnt, &dev->fs_lcnt,
 			   &dev->sda_hold_time);
 
+	/*
+	 * Provide a way for Designware I2C host controllers that are not
+	 * based on Intel LPSS to specify their input clock frequency via
+	 * id->driver_data.
+	 */
+	id = acpi_match_device(pdev->dev.driver->acpi_match_table, &pdev->dev);
+	if (id && id->driver_data)
+		clk_register_fixed_rate(&pdev->dev, dev_name(&pdev->dev), NULL,
+					CLK_IS_ROOT, id->driver_data);
+
 	return 0;
 }
 
+static void dw_i2c_acpi_unconfigure(struct platform_device *pdev)
+{
+	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
+	const struct acpi_device_id *id;
+
+	id = acpi_match_device(pdev->dev.driver->acpi_match_table, &pdev->dev);
+	if (id && id->driver_data)
+		clk_unregister(dev->clk);
+}
+
 static const struct acpi_device_id dw_i2c_acpi_match[] = {
 	{ "INT33C2", 0 },
 	{ "INT33C3", 0 },
@@ -103,6 +125,7 @@ static const struct acpi_device_id dw_i2c_acpi_match[] = {
 	{ "INT3433", 0 },
 	{ "80860F41", 0 },
 	{ "808622C1", 0 },
+	{ "AMD0010", 133 * 1000 * 1000 },
 	{ }
 };
 MODULE_DEVICE_TABLE(acpi, dw_i2c_acpi_match);
@@ -111,6 +134,7 @@ static inline int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	return -ENODEV;
 }
+static inline void dw_i2c_acpi_unconfigure(struct platform_device *pdev) { }
 #endif
 
 static int dw_i2c_probe(struct platform_device *pdev)
@@ -258,6 +282,9 @@ static int dw_i2c_remove(struct platform_device *pdev)
 	pm_runtime_put(&pdev->dev);
 	pm_runtime_disable(&pdev->dev);
 
+	if (ACPI_COMPANION(&pdev->dev))
+		dw_i2c_acpi_unconfigure(pdev);
+
 	return 0;
 }
 
-- 
2.1.0

[PATCH v2 1/3] i2c: designware: Default to fast mode in case of ACPI

[Mika Westerberg]

There is no way in ACPI to tell in which speed the host controller is
supposed to run, so we default to fast mode (400KHz). Since this has been
the default all the time there should be no functional changes with this
change.

This is the first step required to refactor the driver probe so that we can
supply source clock from ACPI part of the driver to the core.

Signed-off-by: Mika Westerberg <mika.westerberg-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
---
 drivers/i2c/busses/i2c-designware-platdrv.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index 4d6a6b94e2fa..353adc4dc583 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -80,7 +80,6 @@ static void dw_i2c_acpi_params(struct platform_device *pdev, char method[],
 static int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
-	bool fs_mode = dev->master_cfg & DW_IC_CON_SPEED_FAST;
 
 	if (!ACPI_HANDLE(&pdev->dev))
 		return -ENODEV;
@@ -93,10 +92,9 @@ static int dw_i2c_acpi_configure(struct platform_device *pdev)
 	 * Try to get SDA hold time and *CNT values from an ACPI method if
 	 * it exists for both supported speed modes.
 	 */
-	dw_i2c_acpi_params(pdev, "SSCN", &dev->ss_hcnt, &dev->ss_lcnt,
-			   fs_mode ? NULL : &dev->sda_hold_time);
+	dw_i2c_acpi_params(pdev, "SSCN", &dev->ss_hcnt, &dev->ss_lcnt, NULL);
 	dw_i2c_acpi_params(pdev, "FMCN", &dev->fs_hcnt, &dev->fs_lcnt,
-			   fs_mode ? &dev->sda_hold_time : NULL);
+			   &dev->sda_hold_time);
 
 	return 0;
 }
-- 
2.1.0

[PATCH v2 2/3] i2c: designware: Rework probe() to get clock a bit later

[Mika Westerberg]

In order to be able to create missing clock for AMD (and in future possibly
others) we move getting clock for the device a bit later. Also make ACPI/DT
configuration in the same place depending on from where the device was
enumerated from.

Signed-off-by: Mika Westerberg <mika.westerberg-VuQAYsv1563Yd54FQh9/CA@public.gmane.org>
---
 drivers/i2c/busses/i2c-designware-platdrv.c | 38 ++++++++++++++---------------
 1 file changed, 18 insertions(+), 20 deletions(-)

diff --git a/drivers/i2c/busses/i2c-designware-platdrv.c b/drivers/i2c/busses/i2c-designware-platdrv.c
index 353adc4dc583..e59c63ae4d41 100644
--- a/drivers/i2c/busses/i2c-designware-platdrv.c
+++ b/drivers/i2c/busses/i2c-designware-platdrv.c
@@ -81,9 +81,6 @@ static int dw_i2c_acpi_configure(struct platform_device *pdev)
 {
 	struct dw_i2c_dev *dev = platform_get_drvdata(pdev);
 
-	if (!ACPI_HANDLE(&pdev->dev))
-		return -ENODEV;
-
 	dev->adapter.nr = -1;
 	dev->tx_fifo_depth = 32;
 	dev->rx_fifo_depth = 32;
@@ -123,7 +120,7 @@ static int dw_i2c_probe(struct platform_device *pdev)
 	struct resource *mem;
 	struct dw_i2c_platform_data *pdata;
 	int irq, r;
-	u32 clk_freq;
+	u32 clk_freq, ht = 0;
 
 	irq = platform_get_irq(pdev, 0);
 	if (irq < 0) {
@@ -146,24 +143,14 @@ static int dw_i2c_probe(struct platform_device *pdev)
 	dev->irq = irq;
 	platform_set_drvdata(pdev, dev);
 
-	dev->clk = devm_clk_get(&pdev->dev, NULL);
-	dev->get_clk_rate_khz = i2c_dw_get_clk_rate_khz;
-
-	if (IS_ERR(dev->clk))
-		return PTR_ERR(dev->clk);
-	clk_prepare_enable(dev->clk);
-
 	/* fast mode by default because of legacy reasons */
 	clk_freq = 400000;
 
-	if (pdev->dev.of_node) {
-		u32 ht = 0;
-		u32 ic_clk = dev->get_clk_rate_khz(dev);
-
+	if (ACPI_COMPANION(&pdev->dev)) {
+		dw_i2c_acpi_configure(pdev);
+	} else if (pdev->dev.of_node) {
 		of_property_read_u32(pdev->dev.of_node,
 					"i2c-sda-hold-time-ns", &ht);
-		dev->sda_hold_time = div_u64((u64)ic_clk * ht + 500000,
-					     1000000);
 
 		of_property_read_u32(pdev->dev.of_node,
 				     "i2c-sda-falling-time-ns",
@@ -202,9 +189,20 @@ static int dw_i2c_probe(struct platform_device *pdev)
 		dev->master_cfg =  DW_IC_CON_MASTER | DW_IC_CON_SLAVE_DISABLE |
 			DW_IC_CON_RESTART_EN | DW_IC_CON_SPEED_FAST;
 
-	/* Try first if we can configure the device from ACPI */
-	r = dw_i2c_acpi_configure(pdev);
-	if (r) {
+	dev->clk = devm_clk_get(&pdev->dev, NULL);
+	dev->get_clk_rate_khz = i2c_dw_get_clk_rate_khz;
+	if (IS_ERR(dev->clk))
+		return PTR_ERR(dev->clk);
+	clk_prepare_enable(dev->clk);
+
+	if (!dev->sda_hold_time && ht) {
+		u32 ic_clk = dev->get_clk_rate_khz(dev);
+
+		dev->sda_hold_time = div_u64((u64)ic_clk * ht + 500000,
+					     1000000);
+	}
+
+	if (!dev->tx_fifo_depth) {
 		u32 param1 = i2c_dw_read_comp_param(dev);
 
 		dev->tx_fifo_depth = ((param1 >> 16) & 0xff) + 1;
-- 
2.1.0

Re: [PATCH v2 0/3] i2c: designware: Add support for AMD I2C controller

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1154 bytes --]

On Tue, Sep 30, 2014 at 01:04:52PM +0300, Mika Westerberg wrote:
> This is second iteration of the patch series. First version is here:
> 
>  http://patchwork.ozlabs.org/patch/390694/
>  http://patchwork.ozlabs.org/patch/390695/
> 
> In order to get source clock to the core driver we need some way of
> creating it in the first place. For Intel LPSS it is done in
> drivers/acpi/acpi_lpss.c but that is not applicable for AMD I2C host
> controller which is too different from private parts to the LPSS one.
> 
> We solve this by creating the clock in ACPI parts of the platform driver
> when we detect the AMD I2C host controller. Doing this requires small
> refactoring to be done to the probe() which is what patches [1-2/3] are
> doing.
> 
> Changes to the previous version:
>  * Rebased on top of i2c/for-next branch
>  * New patch defaulting ACPI probe to fast mode
>  * Instead of direct dependency to COMMON_CLK we depend on (ACPI &&
>    COMMON_CLK) || ACPI. (suggested by Wolfram)
> 
> Carl, can you test that this still works on your machine?

Yes, Carl that would be helpful.

Patches look good to me.

Thanks!


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

Re: [PATCH v2 0/3] i2c: designware: Add support for AMD I2C controller

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 1398 bytes --]

On Tue, Sep 30, 2014 at 01:04:52PM +0300, Mika Westerberg wrote:
> This is second iteration of the patch series. First version is here:
> 
>  http://patchwork.ozlabs.org/patch/390694/
>  http://patchwork.ozlabs.org/patch/390695/
> 
> In order to get source clock to the core driver we need some way of
> creating it in the first place. For Intel LPSS it is done in
> drivers/acpi/acpi_lpss.c but that is not applicable for AMD I2C host
> controller which is too different from private parts to the LPSS one.
> 
> We solve this by creating the clock in ACPI parts of the platform driver
> when we detect the AMD I2C host controller. Doing this requires small
> refactoring to be done to the probe() which is what patches [1-2/3] are
> doing.
> 
> Changes to the previous version:
>  * Rebased on top of i2c/for-next branch
>  * New patch defaulting ACPI probe to fast mode
>  * Instead of direct dependency to COMMON_CLK we depend on (ACPI &&
>    COMMON_CLK) || ACPI. (suggested by Wolfram)
> 
> Carl, can you test that this still works on your machine?
> 
> Carl Peng (1):
>   i2c: designware: Add support for AMD I2C controller
> 
> Mika Westerberg (2):
>   i2c: designware: Default to fast mode in case of ACPI
>   i2c: designware: Rework probe() to get clock a bit later
> 
>  drivers/i2c/busses/Kconfig                  |  1 +

Applied to for-next, thanks!


[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]