[PATCH] s3c-fb: add support for runtime pm

This patch adds support for runtime pm using the functions.
 - pm_runtime_get_sync()
 - pm_runtime_put_sync()

pm_runtime_get_sync() and pm_runtime_put_sync() are called when
open or release function of framebufer driver is called to inform
the system if hardware is idle or not.

Signed-off-by: Jingoo Han <redacted>
---
 drivers/video/s3c-fb.c |  111 ++++++++++++++++++++++++++++++++++++++++++++++--
 1 files changed, 107 insertions(+), 4 deletions(-)

diff --git a/drivers/video/s3c-fb.c b/drivers/video/s3c-fb.c
index f9aca9d..83ce9a0 100644
--- a/drivers/video/s3c-fb.c
+++ b/drivers/video/s3c-fb.c

@@ -23,6 +23,7 @@
 #include <linux/io.h>
 #include <linux/uaccess.h>
 #include <linux/interrupt.h>
+#include <linux/pm_runtime.h>
 
 #include <mach/map.h>
 #include <plat/regs-fb-v4.h>

@@ -1013,8 +1014,30 @@ static int s3c_fb_ioctl(struct fb_info *info, unsigned int cmd,
 	return ret;
 }
 
+static int s3c_fb_open(struct fb_info *info, int user)
+{
+	struct s3c_fb_win *win = info->par;
+	struct s3c_fb *sfb = win->parent;
+
+	pm_runtime_get_sync(sfb->dev);
+
+	return 0;
+}
+
+static int s3c_fb_release(struct fb_info *info, int user)
+{
+	struct s3c_fb_win *win = info->par;
+	struct s3c_fb *sfb = win->parent;
+
+	pm_runtime_put_sync(sfb->dev);
+
+	return 0;
+}
+
 static struct fb_ops s3c_fb_ops = {
 	.owner		= THIS_MODULE,
+	.fb_open	= s3c_fb_open,
+	.fb_release	= s3c_fb_release,
 	.fb_check_var	= s3c_fb_check_var,
 	.fb_set_par	= s3c_fb_set_par,
 	.fb_blank	= s3c_fb_blank,

@@ -1322,6 +1345,8 @@ static int __devinit s3c_fb_probe(struct platform_device *pdev)
 
 	clk_enable(sfb->bus_clk);
 
+	pm_runtime_enable(sfb->dev);
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(dev, "failed to find registers\n");

@@ -1360,6 +1385,9 @@ static int __devinit s3c_fb_probe(struct platform_device *pdev)
 
 	dev_dbg(dev, "got resources (regs %p), probing windows\n", sfb->regs);
 
+	platform_set_drvdata(pdev, sfb);
+	pm_runtime_get_sync(sfb->dev);
+
 	/* setup gpio and output polarity controls */
 
 	pd->setup_gpio();

@@ -1400,6 +1428,7 @@ static int __devinit s3c_fb_probe(struct platform_device *pdev)
 	}
 
 	platform_set_drvdata(pdev, sfb);
+	pm_runtime_put_sync(sfb->dev);
 
 	return 0;
 

@@ -1434,6 +1463,8 @@ static int __devexit s3c_fb_remove(struct platform_device *pdev)
 	struct s3c_fb *sfb = platform_get_drvdata(pdev);
 	int win;
 
+	pm_runtime_get_sync(sfb->dev);
+
 	for (win = 0; win < S3C_FB_MAX_WIN; win++)
 		if (sfb->windows[win])
 			s3c_fb_release_win(sfb, sfb->windows[win]);

@@ -1450,12 +1481,74 @@ static int __devexit s3c_fb_remove(struct platform_device *pdev)
 
 	kfree(sfb);
 
+	pm_runtime_put_sync(sfb->dev);
+	pm_runtime_disable(sfb->dev);
+
 	return 0;
 }
 
 #ifdef CONFIG_PM
-static int s3c_fb_suspend(struct platform_device *pdev, pm_message_t state)
+static int s3c_fb_suspend(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct s3c_fb *sfb = platform_get_drvdata(pdev);
+	struct s3c_fb_win *win;
+	int win_no;
+
+	for (win_no = S3C_FB_MAX_WIN - 1; win_no >= 0; win_no--) {
+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		/* use the blank function to push into power-down */
+		s3c_fb_blank(FB_BLANK_POWERDOWN, win->fbinfo);
+	}
+
+	clk_disable(sfb->bus_clk);
+	return 0;
+}
+
+static int s3c_fb_resume(struct device *dev)
+{
+	struct platform_device *pdev = to_platform_device(dev);
+	struct s3c_fb *sfb = platform_get_drvdata(pdev);
+	struct s3c_fb_platdata *pd = sfb->pdata;
+	struct s3c_fb_win *win;
+	int win_no;
+
+	clk_enable(sfb->bus_clk);
+
+	/* setup registers */
+	writel(pd->vidcon1, sfb->regs + VIDCON1);
+
+	/* zero all windows before we do anything */
+	for (win_no = 0; win_no < sfb->variant.nr_windows; win_no++)
+		s3c_fb_clear_win(sfb, win_no);
+
+	for (win_no = 0; win_no < sfb->variant.nr_windows - 1; win_no++) {
+		void __iomem *regs = sfb->regs + sfb->variant.keycon;
+
+		regs += (win_no * 8);
+		writel(0xffffff, regs + WKEYCON0);
+		writel(0xffffff, regs + WKEYCON1);
+	}
+
+	/* restore framebuffers */
+	for (win_no = 0; win_no < S3C_FB_MAX_WIN; win_no++) {
+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		dev_dbg(&pdev->dev, "resuming window %d\n", win_no);
+		s3c_fb_set_par(win->fbinfo);
+	}
+
+	return 0;
+}
+
+int s3c_fb_runtime_suspend(struct device *dev)
 {
+	struct platform_device *pdev = to_platform_device(dev);
 	struct s3c_fb *sfb = platform_get_drvdata(pdev);
 	struct s3c_fb_win *win;
 	int win_no;

@@ -1473,8 +1566,9 @@ static int s3c_fb_suspend(struct platform_device *pdev, pm_message_t state)
 	return 0;
 }
 
-static int s3c_fb_resume(struct platform_device *pdev)
+int s3c_fb_runtime_resume(struct device *dev)
 {
+	struct platform_device *pdev = to_platform_device(dev);
 	struct s3c_fb *sfb = platform_get_drvdata(pdev);
 	struct s3c_fb_platdata *pd = sfb->pdata;
 	struct s3c_fb_win *win;

@@ -1509,9 +1603,12 @@ static int s3c_fb_resume(struct platform_device *pdev)
 
 	return 0;
 }
+
 #else
 #define s3c_fb_suspend NULL
 #define s3c_fb_resume  NULL
+#define s3c_fb_runtime_suspend NULL
+#define s3c_fb_runtime_resume NULL
 #endif
 
 

@@ -1710,15 +1807,21 @@ static struct platform_device_id s3c_fb_driver_ids[] = {
 };
 MODULE_DEVICE_TABLE(platform, s3c_fb_driver_ids);
 
+static const struct dev_pm_ops s3cfb_pm_ops = {
+	.suspend	= s3c_fb_suspend,
+	.resume		= s3c_fb_resume,
+	.runtime_suspend	= s3c_fb_runtime_suspend,
+	.runtime_resume		= s3c_fb_runtime_resume,
+};
+
 static struct platform_driver s3c_fb_driver = {
 	.probe		= s3c_fb_probe,
 	.remove		= __devexit_p(s3c_fb_remove),
-	.suspend	= s3c_fb_suspend,
-	.resume		= s3c_fb_resume,
 	.id_table	= s3c_fb_driver_ids,
 	.driver		= {
 		.name	= "s3c-fb",
 		.owner	= THIS_MODULE,
+		.pm	= &s3cfb_pm_ops,
 	},
 };
 

-- 
1.6.2.5

On Fri, Dec 17, 2010 at 04:45:46PM +0900, Jingoo Han wrote:

This patch adds support for runtime pm using the functions.
 - pm_runtime_get_sync()
 - pm_runtime_put_sync()

pm_runtime_get_sync() and pm_runtime_put_sync() are called when
open or release function of framebufer driver is called to inform
the system if hardware is idle or not.

Signed-off-by: Jingoo Han <redacted>

Barring any objections, I'll apply it, thanks.

Hi Jingoo,

Here is a quick review.

Best regards,

Jingoo Han wrote:

This patch adds support for runtime pm using the functions.
 - pm_runtime_get_sync()
 - pm_runtime_put_sync()

pm_runtime_get_sync() and pm_runtime_put_sync() are called when open or
release function of framebufer driver is called to inform the system if

hardware is

f9aca9d..83ce9a0

<snip>

+static int s3c_fb_release(struct fb_info *info, int user) {
+	struct s3c_fb_win *win = info->par;
+	struct s3c_fb *sfb = win->parent;
+
+	pm_runtime_put_sync(sfb->dev);
+
+	return 0;
+}
+
 static struct fb_ops s3c_fb_ops = {
 	.owner		= THIS_MODULE,
+	.fb_open	= s3c_fb_open,
+	.fb_release	= s3c_fb_release,
 	.fb_check_var	= s3c_fb_check_var,
 	.fb_set_par	= s3c_fb_set_par,
 	.fb_blank	= s3c_fb_blank,

How about use pm_runtime_put (sfb->dev) instead
pm_runtime_put_sync(sfb->dev) ?
In my opinion no need to sync option in *put* case.

platform_device

*pdev)

 	clk_enable(sfb->bus_clk);

+	pm_runtime_enable(sfb->dev);
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(dev, "failed to find registers\n"); @@ -1360,6

+1385,9

How about move pm_runtime_enable upper pm_runtime_get_sync in s3c_fb_probe?

sfb->regs);

+	platform_set_drvdata(pdev, sfb);
+	pm_runtime_get_sync(sfb->dev);
+
 	/* setup gpio and output polarity controls */

 	pd->setup_gpio();

platform_set_drvdata(pdev, sfb); already exists line# 1402.

<snip>

sfb->variant.nr_windows is better than S3C_FB_MAX_WIN.

1450,12 +1481,74 @@ static int __devexit s3c_fb_remove(struct

platform_device

*pdev)

<snip>

+
+	for (win_no = S3C_FB_MAX_WIN - 1; win_no >= 0; win_no--) {

Ditto

+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		/* use the blank function to push into power-down */
+		s3c_fb_blank(FB_BLANK_POWERDOWN, win->fbinfo);
+	}
+
+	clk_disable(sfb->bus_clk);
+	return 0;
+}
+
+static int s3c_fb_resume(struct device *dev) {
+	struct platform_device *pdev = to_platform_device(dev);
+	struct s3c_fb *sfb = platform_get_drvdata(pdev);
+	struct s3c_fb_platdata *pd = sfb->pdata;
+	struct s3c_fb_win *win;
+	int win_no;
+
+	clk_enable(sfb->bus_clk);
+
+	/* setup registers */
+	writel(pd->vidcon1, sfb->regs + VIDCON1);
+
+	/* zero all windows before we do anything */
+	for (win_no = 0; win_no < sfb->variant.nr_windows; win_no++)
+		s3c_fb_clear_win(sfb, win_no);

Is it right "win_no < sfb->variant.nr_windows" not "nr_windows - 1" ?

+
+	for (win_no = 0; win_no < sfb->variant.nr_windows - 1; win_no++) {
+		void __iomem *regs = sfb->regs + sfb->variant.keycon;
+
+		regs += (win_no * 8);
+		writel(0xffffff, regs + WKEYCON0);
+		writel(0xffffff, regs + WKEYCON1);
+	}
+
+	/* restore framebuffers */
+	for (win_no = 0; win_no < S3C_FB_MAX_WIN; win_no++) {
+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		dev_dbg(&pdev->dev, "resuming window %d\n", win_no);
+		s3c_fb_set_par(win->fbinfo);
+	}
+
+	return 0;
+}
+

How about merge three for loop ?

In my opinion s3c_fb_blank(FB_BLANK_UNBLANK, win->fbinfo) should be called
in resume.
Because sfb->enabled was cleared in suspend function using
s3c_fb_blank(FB_BLANK_POWERDOWN, win->fbinfo).

<snip>

s3c_fb_driver_ids[] > {  };  MODULE_DEVICE_TABLE(platform, s3c_fb_driver_ids);

+static const struct dev_pm_ops s3cfb_pm_ops = {
+	.suspend	= s3c_fb_suspend,
+	.resume		= s3c_fb_resume,
+	.runtime_suspend	= s3c_fb_runtime_suspend,
+	.runtime_resume		= s3c_fb_runtime_resume,
+};

s3c_fb_suspend and s3c_fb_resume are exactly same with
s3c_fb_runtime_suspend s3c_fb_runtime_resume.
Why don't you register the same function ?

<snip>

--
1.6.2.5

--
To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in

the body

of a message to majordomo@vger.kernel.org More majordomo info at
http://vger.kernel.org/majordomo-info.html

Hi Jingoo,

How about adding pm_runtime_* in s3c_fb_blank ?
In FB_BLANK_POWERDOWN state no need to enable clock or power-domain.

Best regards,

-----Original Message-----
From: linux-fbdev-owner@vger.kernel.org [mailto:linux-fbdev-
owner@vger.kernel.org] On Behalf Of Jonghun Han
Sent: Tuesday, December 21, 2010 5:30 PM
To: 'Jingoo Han'; linux-fbdev@vger.kernel.org; 'Paul Mundt';

linux-samsung-

soc@vger.kernel.org
Cc: ben-linux@fluff.org; 'Ilho Lee'
Subject: RE: [PATCH] s3c-fb: add support for runtime pm

Hi Jingoo,

Here is a quick review.

Best regards,

Jingoo Han wrote:

quoted

This patch adds support for runtime pm using the functions.
 - pm_runtime_get_sync()
 - pm_runtime_put_sync()

pm_runtime_get_sync() and pm_runtime_put_sync() are called when open
or release function of framebufer driver is called to inform the
system if

hardware is

quoted

idle or not.

Signed-off-by: Jingoo Han <redacted>
---
 drivers/video/s3c-fb.c |  111
++++++++++++++++++++++++++++++++++++++++++++++--
 1 files changed, 107 insertions(+), 4 deletions(-)

diff --git a/drivers/video/s3c-fb.c b/drivers/video/s3c-fb.c index

f9aca9d..83ce9a0

quoted

100644

--- a/drivers/video/s3c-fb.c
+++ b/drivers/video/s3c-fb.c

<snip>

quoted

+static int s3c_fb_release(struct fb_info *info, int user) {
+	struct s3c_fb_win *win = info->par;
+	struct s3c_fb *sfb = win->parent;
+
+	pm_runtime_put_sync(sfb->dev);
+
+	return 0;
+}
+
 static struct fb_ops s3c_fb_ops = {
 	.owner		= THIS_MODULE,
+	.fb_open	= s3c_fb_open,
+	.fb_release	= s3c_fb_release,
 	.fb_check_var	= s3c_fb_check_var,
 	.fb_set_par	= s3c_fb_set_par,
 	.fb_blank	= s3c_fb_blank,

How about use pm_runtime_put (sfb->dev) instead
pm_runtime_put_sync(sfb->dev) ?
In my opinion no need to sync option in *put* case.

quoted

@@ -1322,6 +1345,8 @@ static int __devinit s3c_fb_probe(struct

platform_device

quoted

*pdev)

 	clk_enable(sfb->bus_clk);

+	pm_runtime_enable(sfb->dev);
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (!res) {
 		dev_err(dev, "failed to find registers\n"); @@ -1360,6

+1385,9

How about move pm_runtime_enable upper pm_runtime_get_sync in

s3c_fb_probe?

quoted

@@ static int __devinit s3c_fb_probe(struct platform_device *pdev)

 	dev_dbg(dev, "got resources (regs %p), probing windows\n",

sfb->regs);

quoted

+	platform_set_drvdata(pdev, sfb);
+	pm_runtime_get_sync(sfb->dev);
+
 	/* setup gpio and output polarity controls */

 	pd->setup_gpio();

platform_set_drvdata(pdev, sfb); already exists line# 1402.

<snip>

quoted

@@ -1434,6 +1463,8 @@ static int __devexit s3c_fb_remove(struct

platform_device *pdev)
 	struct s3c_fb *sfb = platform_get_drvdata(pdev);
 	int win;

+	pm_runtime_get_sync(sfb->dev);
+
 	for (win = 0; win < S3C_FB_MAX_WIN; win++)
 		if (sfb->windows[win])
 			s3c_fb_release_win(sfb, sfb->windows[win]); @@ -

sfb->variant.nr_windows is better than S3C_FB_MAX_WIN.

quoted

1450,12 +1481,74 @@ static int __devexit s3c_fb_remove(struct

platform_device

quoted

*pdev)

<snip>

quoted

+
+	for (win_no = S3C_FB_MAX_WIN - 1; win_no >= 0; win_no--) {

Ditto

quoted

+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		/* use the blank function to push into power-down */
+		s3c_fb_blank(FB_BLANK_POWERDOWN, win->fbinfo);
+	}
+
+	clk_disable(sfb->bus_clk);
+	return 0;
+}
+
+static int s3c_fb_resume(struct device *dev) {
+	struct platform_device *pdev = to_platform_device(dev);
+	struct s3c_fb *sfb = platform_get_drvdata(pdev);
+	struct s3c_fb_platdata *pd = sfb->pdata;
+	struct s3c_fb_win *win;
+	int win_no;
+
+	clk_enable(sfb->bus_clk);
+
+	/* setup registers */
+	writel(pd->vidcon1, sfb->regs + VIDCON1);
+
+	/* zero all windows before we do anything */
+	for (win_no = 0; win_no < sfb->variant.nr_windows; win_no++)
+		s3c_fb_clear_win(sfb, win_no);

Is it right "win_no < sfb->variant.nr_windows" not "nr_windows - 1" ?

Sorry my mistake.

quoted

+
+	for (win_no = 0; win_no < sfb->variant.nr_windows - 1; win_no++) {
+		void __iomem *regs = sfb->regs + sfb->variant.keycon;
+
+		regs += (win_no * 8);
+		writel(0xffffff, regs + WKEYCON0);
+		writel(0xffffff, regs + WKEYCON1);
+	}
+
+	/* restore framebuffers */
+	for (win_no = 0; win_no < S3C_FB_MAX_WIN; win_no++) {
+		win = sfb->windows[win_no];
+		if (!win)
+			continue;
+
+		dev_dbg(&pdev->dev, "resuming window %d\n", win_no);
+		s3c_fb_set_par(win->fbinfo);
+	}
+
+	return 0;
+}
+

How about merge three for loop ?

So how about merge two for loops ?

In my opinion s3c_fb_blank(FB_BLANK_UNBLANK, win->fbinfo) should be called
in resume.
Because sfb->enabled was cleared in suspend function using
s3c_fb_blank(FB_BLANK_POWERDOWN, win->fbinfo).

<snip>

quoted

@@ -1710,15 +1807,21 @@ static struct platform_device_id

s3c_fb_driver_ids[] > > {  };  MODULE_DEVICE_TABLE(platform, s3c_fb_driver_ids);

quoted

+static const struct dev_pm_ops s3cfb_pm_ops = {
+	.suspend	= s3c_fb_suspend,
+	.resume		= s3c_fb_resume,
+	.runtime_suspend	= s3c_fb_runtime_suspend,
+	.runtime_resume		= s3c_fb_runtime_resume,
+};

s3c_fb_suspend and s3c_fb_resume are exactly same with
s3c_fb_runtime_suspend s3c_fb_runtime_resume.
Why don't you register the same function ?

<snip>

quoted

--
1.6.2.5

--
To unsubscribe from this list: send the line "unsubscribe linux-fbdev"
in

the body

quoted

of a message to majordomo@vger.kernel.org More majordomo info at
http://vger.kernel.org/majordomo-info.html

--
To unsubscribe from this list: send the line "unsubscribe linux-fbdev" in

the body

of a message to majordomo@vger.kernel.org More majordomo info at
http://vger.kernel.org/majordomo-info.html