Skip to contentSkip to conversation

Input: pxa27x_keypad keep clock on as wakeup source

PATCH Superseded Chao Xie 3 participants 6 messages Mar 28, 2012 – Mar 29, 2012

Subsystems (4)
Arm (32-bit) Arm sub-architectures HID Input drivers

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Chao Xie

author
#0 lore raw

PATCH 1/4 Input: pxa27x_keypad keep clock on as wakeup source

When the keypad is used as wake up source, the clock can not be disabled. Or it can not detect key pressing.
If the keypad is used as wake up source, when resume back, do not enable the clock and configure it again because the register content is retained.

Signed-off-by Chao Xie <[email protected]>
---
 drivers/input/keyboard/pxa27x_keypad.c |   27 ++++++++++++++++++---------
 1 files changed, 18 insertions(+), 9 deletions(-)
drivers/input/keyboard/pxa27x_keypad.c +18 -9
index 29fe1b2..1c5752f 100644
@@ -419,10 +419,14 @@ static int pxa27x_keypad_suspend(struct device *dev)
 	struct platform_device *pdev = to_platform_device(dev);
 	struct pxa27x_keypad *keypad = platform_get_drvdata(pdev);
 
-	clk_disable(keypad->clk);
-
+	/*
+	 * If the keypad is used a wake up source, clock can not be disabled.
+	 * Or it can not detect the key pressing.
+	 */
 	if (device_may_wakeup(&pdev->dev))
 		enable_irq_wake(keypad->irq);
+	else
+		clk_disable(keypad->clk);
 
 	return 0;
 }

@@ -433,19 +437,24 @@ static int pxa27x_keypad_resume(struct device *dev)
 	struct pxa27x_keypad *keypad = platform_get_drvdata(pdev);
 	struct input_dev *input_dev = keypad->input_dev;
 
+	/*
+	 * If the keypad is used as wake up source, the clock is not turned
+	 * off. So do not need configure it again.
+	 */
 	if (device_may_wakeup(&pdev->dev))
 		disable_irq_wake(keypad->irq);
+	else {
+		mutex_lock(&input_dev->mutex);
 
-	mutex_lock(&input_dev->mutex);
+		if (input_dev->users) {
+			/* Enable unit clock */
+			clk_enable(keypad->clk);
+			pxa27x_keypad_config(keypad);
+		}
 
-	if (input_dev->users) {
-		/* Enable unit clock */
-		clk_enable(keypad->clk);
-		pxa27x_keypad_config(keypad);
+		mutex_unlock(&input_dev->mutex);
 	}
 
-	mutex_unlock(&input_dev->mutex);
-
 	return 0;
 }
Signature · 1 line
1.7.0.4
author
#1 lore raw

PATCH 2/4 Input: pxa27x_keypad bug fix for direct_key_mask

When direcct_key_num is 0, the mask should be 0. When direcct_key_num is 1, the mask should be 0b1.

Signed-off-by Chao Xie <[email protected]>
---
 drivers/input/keyboard/pxa27x_keypad.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
drivers/input/keyboard/pxa27x_keypad.c +1 -1
index 1c5752f..f9a76d4 100644
@@ -383,7 +383,7 @@ static void pxa27x_keypad_config(struct pxa27x_keypad *keypad)
 	if (pdata->direct_key_num > direct_key_num)
 		direct_key_num = pdata->direct_key_num;
 
-	keypad->direct_key_mask = ((2 << direct_key_num) - 1) & ~mask;
+	keypad->direct_key_mask = ((1 << direct_key_num) - 1) & ~mask;
 
 	/* enable direct key */
 	if (direct_key_num)
Signature · 1 line
1.7.0.4
author
#2 lore raw

PATCH 3/4 Input: pxa27x_keypad direct key may be low active

In reply to Chao Xie

KPDK_DK only indicates the pin level of direct key. So it is related to board, and low level may be active which indicates that a key is pressed.

Signed-off-by Chao Xie <[email protected]>
---
 arch/arm/plat-pxa/include/plat/pxa27x_keypad.h |    2 ++
 drivers/input/keyboard/pxa27x_keypad.c         |   10 +++++++++-
 2 files changed, 11 insertions(+), 1 deletions(-)
arch/arm/plat-pxa/include/plat/pxa27x_keypad.h +2 -0
index abcc36e..7ffb16b 100644
@@ -44,6 +44,8 @@ struct pxa27x_keypad_platform_data {
 	/* direct keys */
 	int		direct_key_num;
 	unsigned int	direct_key_map[MAX_DIRECT_KEY_NUM];
+	/* the key output may be low active */
+	int		direct_key_low_active;
 
 	/* rotary encoders 0 */
 	int		enable_rotary0;
drivers/input/keyboard/pxa27x_keypad.c +9 -1
index f9a76d4..b41ebda 100644
@@ -311,7 +311,15 @@ static void pxa27x_keypad_scan_direct(struct pxa27x_keypad *keypad)
 	if (pdata->enable_rotary0 || pdata->enable_rotary1)
 		pxa27x_keypad_scan_rotary(keypad);
 
-	new_state = KPDK_DK(kpdk) & keypad->direct_key_mask;
+	/*
+	 * The KPDR_DK only output the key pin level, so it relates to board,
+	 * and low level may be active.
+	 */
+	if (pdata->direct_key_low_active)
+		new_state = ~KPDK_DK(kpdk) & keypad->direct_key_mask;
+	else
+		new_state = KPDK_DK(kpdk) & keypad->direct_key_mask;
+
 	bits_changed = keypad->direct_key_state ^ new_state;
 
 	if (bits_changed == 0)
Signature · 1 line
1.7.0.4
author
#3 lore raw

PATCH 4/4 Input: pxa27x_keypad add choice to set direct_key_mask

In reply to Chao Xie

Direct keys usage may not start from KP_DKIN0, add a msk option to configure the specifics for platforms that can skip some keys.

Signed-off-by Chao Xie <[email protected]>
---
 arch/arm/plat-pxa/include/plat/pxa27x_keypad.h |    2 ++
 drivers/input/keyboard/pxa27x_keypad.c         |    7 ++++++-
 2 files changed, 8 insertions(+), 1 deletions(-)
arch/arm/plat-pxa/include/plat/pxa27x_keypad.h +2 -0
index 7ffb16b..5ce8d5e 100644
@@ -46,6 +46,8 @@ struct pxa27x_keypad_platform_data {
 	unsigned int	direct_key_map[MAX_DIRECT_KEY_NUM];
 	/* the key output may be low active */
 	int		direct_key_low_active;
+	/* give board a chance to choose the start direct key */
+	unsigned int	direct_key_mask;
 
 	/* rotary encoders 0 */
 	int		enable_rotary0;
drivers/input/keyboard/pxa27x_keypad.c +6 -1
index b41ebda..661892b 100644
@@ -391,7 +391,12 @@ static void pxa27x_keypad_config(struct pxa27x_keypad *keypad)
 	if (pdata->direct_key_num > direct_key_num)
 		direct_key_num = pdata->direct_key_num;
 
-	keypad->direct_key_mask = ((1 << direct_key_num) - 1) & ~mask;
+	/* Direct keys usage may not start from KP_DKIN0, check the platfrom
+	 * msk data to config the specific */
+	if (pdata->direct_key_mask)
+		keypad->direct_key_mask = pdata->direct_key_mask;
+	else
+		keypad->direct_key_mask = ((1 << direct_key_num) - 1) & ~mask;
 
 	/* enable direct key */
 	if (direct_key_num)
Signature · 1 line
1.7.0.4
#4 lore raw

PATCH 4/4 Input: pxa27x_keypad add choice to set direct_key_mask

Hello.

Quoted 4 lines · Chao Xie

Direct keys usage may not start from KP_DKIN0, add a msk option to configure the specifics for platforms that can skip some keys.

Signed-off-by Chao Xie<[email protected]>

[...]

drivers/input/keyboard/pxa27x_keypad.c static void pxa27x_keypad_config(struct pxa27x_keypad *keypad) · line 395

Quoted 11 lines
drivers/input/keyboard/pxa27x_keypad.c +2 -1
index b41ebda..661892b 100644
@@ -391,7 +391,12 @@ static void pxa27x_keypad_config(struct pxa27x_keypad *keypad)
  	if (pdata->direct_key_num>  direct_key_num)
  		direct_key_num = pdata->direct_key_num;
 
-	keypad->direct_key_mask = ((1<<  direct_key_num) - 1)&  ~mask;
+	/* Direct keys usage may not start from KP_DKIN0, check the platfrom
+	 * msk data to config the specific */

The preferred multi-line comment style is not this one. You got it right in
the previous patch. :-)

WBR, Sergei

#5 lore raw

PATCH 4/4 Input: pxa27x_keypad add choice to set direct_key_mask

Thanks. I will sent V2 patches.

-----Original Message-----
From: Sergei Shtylyov [mailto:sshtylyov at mvista.com]
Sent: Thursday, March 29, 2012 2:04 AM
To: Chao Xie
Cc: linux-input at vger.kernel.org; eric.y.miao at gmail.com; Haojian Zhuang; linux-arm-kernel at lists.infradead.org
Subject: Re: [PATCH 4/4] Input: pxa27x_keypad add choice to set direct_key_mask

Hello.

Quoted 4 lines · Chao Xie

Direct keys usage may not start from KP_DKIN0, add a msk option to configure the specifics for platforms that can skip some keys.

Signed-off-by Chao Xie<[email protected]>

[...]

drivers/input/keyboard/pxa27x_keypad.c static void pxa27x_keypad_config(struct pxa27x_keypad *keypad) · line 395

Quoted 11 lines
drivers/input/keyboard/pxa27x_keypad.c +2 -1
index b41ebda..661892b 100644
@@ -391,7 +391,12 @@ static void pxa27x_keypad_config(struct pxa27x_keypad *keypad)
  	if (pdata->direct_key_num>  direct_key_num)
  		direct_key_num = pdata->direct_key_num;
 
-	keypad->direct_key_mask = ((1<<  direct_key_num) - 1)&  ~mask;
+	/* Direct keys usage may not start from KP_DKIN0, check the platfrom
+	 * msk data to config the specific */

The preferred multi-line comment style is not this one. You got it right in
the previous patch. :-)

WBR, Sergei