[PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[Kalle Jokiniemi]

These patches fix the issue where the wdtimer blocks CORE idle
transitions on n900 (OMAP3). The root cause was that SMART idle
mode in wdtimer did not allow the CORE idle transition to happen.

v2: updated commit message on patch 1/2, Cc'd Wim, and added
comments to code about possible HW bug in wdtimer2.

Wim, please check patch number one and add your acked by.

I propose that we push these upstream via linux-omap-pm,
as the patches deal with PM and need to be applied together
to get the desired functionality.

Paul Walmsley (2):
  Watchdog: omap_wdt: add fine grain runtime-pm
  OMAP3: wdtimer: Fix CORE idle transition

 arch/arm/mach-omap2/omap_hwmod_3xxx_data.c |    1 +
 drivers/watchdog/omap_wdt.c                |   25 +++++++++++++++++++++++--
 2 files changed, 24 insertions(+), 2 deletions(-)

[PATCH 1/2] Watchdog: omap_wdt: add fine grain runtime-pm

[Kalle Jokiniemi]

From: Paul Walmsley <paul@pwsan.com>

The omap_wdt should only be in full active state when the
registers are being accessed. Otherwise the device can be
on lower power mode.

This patch is based on a patch created by Kalle Jokiniemi:
https://patchwork.kernel.org/patch/618231/
which is itself based on a patch created by Atal
Shargorodsky: http://lkml.org/lkml/2009/3/10/266.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Kalle Jokiniemi <kalle.jokiniemi@nokia.com>
Tested-by: Kalle Jokiniemi <kalle.jokiniemi@nokia.com>
Cc: Wim Van Sebroeck <wim@iguana.be>
---
 drivers/watchdog/omap_wdt.c |   25 +++++++++++++++++++++++--
 1 files changed, 23 insertions(+), 2 deletions(-)

diff --git a/drivers/watchdog/omap_wdt.c b/drivers/watchdog/omap_wdt.c
index 3dd4971..2b4acb8 100644
--- a/drivers/watchdog/omap_wdt.c
+++ b/drivers/watchdog/omap_wdt.c
@@ -124,6 +124,8 @@ static void omap_wdt_set_timeout(struct omap_wdt_dev *wdev)
 	u32 pre_margin = GET_WLDR_VAL(timer_margin);
 	void __iomem *base = wdev->base;
 
+	pm_runtime_get_sync(wdev->dev);
+
 	/* just count up at 32 KHz */
 	while (__raw_readl(base + OMAP_WATCHDOG_WPS) & 0x04)
 		cpu_relax();
@@ -131,6 +133,8 @@ static void omap_wdt_set_timeout(struct omap_wdt_dev *wdev)
 	__raw_writel(pre_margin, base + OMAP_WATCHDOG_LDR);
 	while (__raw_readl(base + OMAP_WATCHDOG_WPS) & 0x04)
 		cpu_relax();
+
+	pm_runtime_put_sync(wdev->dev);
 }
 
 /*
@@ -160,6 +164,8 @@ static int omap_wdt_open(struct inode *inode, struct file *file)
 	omap_wdt_ping(wdev); /* trigger loading of new timeout value */
 	omap_wdt_enable(wdev);
 
+	pm_runtime_put_sync(wdev->dev);
+
 	return nonseekable_open(inode, file);
 }
 
@@ -171,6 +177,7 @@ static int omap_wdt_release(struct inode *inode, struct file *file)
 	 *      Shut off the timer unless NOWAYOUT is defined.
 	 */
 #ifndef CONFIG_WATCHDOG_NOWAYOUT
+	pm_runtime_get_sync(wdev->dev);
 
 	omap_wdt_disable(wdev);
 
@@ -190,9 +197,11 @@ static ssize_t omap_wdt_write(struct file *file, const char __user *data,
 
 	/* Refresh LOAD_TIME. */
 	if (len) {
+		pm_runtime_get_sync(wdev->dev);
 		spin_lock(&wdt_lock);
 		omap_wdt_ping(wdev);
 		spin_unlock(&wdt_lock);
+		pm_runtime_put_sync(wdev->dev);
 	}
 	return len;
 }
@@ -224,15 +233,18 @@ static long omap_wdt_ioctl(struct file *file, unsigned int cmd,
 			return put_user(omap_prcm_get_reset_sources(),
 					(int __user *)arg);
 	case WDIOC_KEEPALIVE:
+		pm_runtime_get_sync(wdev->dev);
 		spin_lock(&wdt_lock);
 		omap_wdt_ping(wdev);
 		spin_unlock(&wdt_lock);
+		pm_runtime_put_sync(wdev->dev);
 		return 0;
 	case WDIOC_SETTIMEOUT:
 		if (get_user(new_margin, (int __user *)arg))
 			return -EFAULT;
 		omap_wdt_adjust_timeout(new_margin);
 
+		pm_runtime_get_sync(wdev->dev);
 		spin_lock(&wdt_lock);
 		omap_wdt_disable(wdev);
 		omap_wdt_set_timeout(wdev);
@@ -240,6 +252,7 @@ static long omap_wdt_ioctl(struct file *file, unsigned int cmd,
 
 		omap_wdt_ping(wdev);
 		spin_unlock(&wdt_lock);
+		pm_runtime_put_sync(wdev->dev);
 		/* Fall */
 	case WDIOC_GETTIMEOUT:
 		return put_user(timer_margin, (int __user *)arg);
@@ -345,8 +358,11 @@ static void omap_wdt_shutdown(struct platform_device *pdev)
 {
 	struct omap_wdt_dev *wdev = platform_get_drvdata(pdev);
 
-	if (wdev->omap_wdt_users)
+	if (wdev->omap_wdt_users) {
+		pm_runtime_get_sync(wdev->dev);
 		omap_wdt_disable(wdev);
+		pm_runtime_put_sync(wdev->dev);
+	}
 }
 
 static int __devexit omap_wdt_remove(struct platform_device *pdev)
@@ -381,8 +397,11 @@ static int omap_wdt_suspend(struct platform_device *pdev, pm_message_t state)
 {
 	struct omap_wdt_dev *wdev = platform_get_drvdata(pdev);
 
-	if (wdev->omap_wdt_users)
+	if (wdev->omap_wdt_users) {
+		pm_runtime_get_sync(wdev->dev);
 		omap_wdt_disable(wdev);
+		pm_runtime_put_sync(wdev->dev);
+	}
 
 	return 0;
 }
@@ -392,8 +411,10 @@ static int omap_wdt_resume(struct platform_device *pdev)
 	struct omap_wdt_dev *wdev = platform_get_drvdata(pdev);
 
 	if (wdev->omap_wdt_users) {
+		pm_runtime_get_sync(wdev->dev);
 		omap_wdt_enable(wdev);
 		omap_wdt_ping(wdev);
+		pm_runtime_put_sync(wdev->dev);
 	}
 
 	return 0;
-- 
1.7.1

[PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition

[Kalle Jokiniemi]

From: Paul Walmsley <paul@pwsan.com>

The HW superwised smart idle for wdtimer in OMAP3 prevents
CORE power domain idle transitions. Disable it by swithing
to SW supervised transitions.

This could be a hardware bug in the OMAP3 wdtimer2 block.

Signed-off-by: Kalle Jokiniemi <kalle.jokiniemi@nokia.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Benoit Cousson <b-cousson@ti.com>
---
 arch/arm/mach-omap2/omap_hwmod_3xxx_data.c |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c
index 196a834..ebd3309 100644
--- a/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c
+++ b/arch/arm/mach-omap2/omap_hwmod_3xxx_data.c
@@ -1293,6 +1293,11 @@ static struct omap_hwmod omap3xxx_wd_timer2_hwmod = {
 	.slaves		= omap3xxx_wd_timer2_slaves,
 	.slaves_cnt	= ARRAY_SIZE(omap3xxx_wd_timer2_slaves),
 	.omap_chip	= OMAP_CHIP_INIT(CHIP_IS_OMAP3430),
+	/*
+	 * XXX: Use software supervised mode, HW supervised smartidle seems to
+	 * block CORE power domain idle transitions. Maybe a HW bug in wdt2?
+	 */
+	.flags		= HWMOD_SWSUP_SIDLE,
 };
 
 /* UART common */
-- 
1.7.1

Re: [PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition

[Paul Walmsley]

On Thu, 10 Mar 2011, Kalle Jokiniemi wrote:

> From: Paul Walmsley <paul@pwsan.com>
> 
> The HW superwised smart idle for wdtimer in OMAP3 prevents
> CORE power domain idle transitions. Disable it by swithing
> to SW supervised transitions.
> 
> This could be a hardware bug in the OMAP3 wdtimer2 block.
> 
> Signed-off-by: Kalle Jokiniemi <kalle.jokiniemi@nokia.com>
> Signed-off-by: Paul Walmsley <paul@pwsan.com>
> Cc: Benoit Cousson <b-cousson@ti.com>

I've queued this one for 2.6.39, since it seems pretty clear that
target smart-idle mode doesn't work with this IP block.

Probably we'll need to do the same thing for OMAP2420/2430, but I'm not   
going to touch those right now, since we're so close to the merge window.


- Paul

Re: [PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[Wim Van Sebroeck]

Hi All,

> These patches fix the issue where the wdtimer blocks CORE idle
> transitions on n900 (OMAP3). The root cause was that SMART idle
> mode in wdtimer did not allow the CORE idle transition to happen.
> 
> v2: updated commit message on patch 1/2, Cc'd Wim, and added
> comments to code about possible HW bug in wdtimer2.
> 
> Wim, please check patch number one and add your acked by.
> 
> I propose that we push these upstream via linux-omap-pm,
> as the patches deal with PM and need to be applied together
> to get the desired functionality.
> 
> Paul Walmsley (2):
>   Watchdog: omap_wdt: add fine grain runtime-pm
>   OMAP3: wdtimer: Fix CORE idle transition
> 
>  arch/arm/mach-omap2/omap_hwmod_3xxx_data.c |    1 +
>  drivers/watchdog/omap_wdt.c                |   25 +++++++++++++++++++++++--
>  2 files changed, 24 insertions(+), 2 deletions(-)

Acked-by: Wim Van Sebroeck <wim@iguana.be>

linux-omap-pm is indeed the best tree for getting this upstream.

Kind regards,
Wim.

Re: [PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[Paul Walmsley]

Hi Wim, Kevin,

On Thu, 10 Mar 2011, Wim Van Sebroeck wrote:

> > These patches fix the issue where the wdtimer blocks CORE idle
> > transitions on n900 (OMAP3). The root cause was that SMART idle
> > mode in wdtimer did not allow the CORE idle transition to happen.
> > 
> > v2: updated commit message on patch 1/2, Cc'd Wim, and added
> > comments to code about possible HW bug in wdtimer2.
> > 
> > Wim, please check patch number one and add your acked by.
> > 
> > I propose that we push these upstream via linux-omap-pm,
> > as the patches deal with PM and need to be applied together
> > to get the desired functionality.
> > 
> > Paul Walmsley (2):
> >   Watchdog: omap_wdt: add fine grain runtime-pm
> >   OMAP3: wdtimer: Fix CORE idle transition
> > 
> >  arch/arm/mach-omap2/omap_hwmod_3xxx_data.c |    1 +
> >  drivers/watchdog/omap_wdt.c                |   25 +++++++++++++++++++++++--
> >  2 files changed, 24 insertions(+), 2 deletions(-)
> 
> Acked-by: Wim Van Sebroeck <wim@iguana.be>

Thanks.
 
> linux-omap-pm is indeed the best tree for getting this upstream.

Kevin, do you want to take the drivers/watchdog/omap_wdt.c patch?  The 
first patch is queued in my branch currently.


- Paul

Re: [PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[Kevin Hilman]

Paul Walmsley <paul@pwsan.com> writes:

> Hi Wim, Kevin,
>
> On Thu, 10 Mar 2011, Wim Van Sebroeck wrote:
>
>> > These patches fix the issue where the wdtimer blocks CORE idle
>> > transitions on n900 (OMAP3). The root cause was that SMART idle
>> > mode in wdtimer did not allow the CORE idle transition to happen.
>> > 
>> > v2: updated commit message on patch 1/2, Cc'd Wim, and added
>> > comments to code about possible HW bug in wdtimer2.
>> > 
>> > Wim, please check patch number one and add your acked by.
>> > 
>> > I propose that we push these upstream via linux-omap-pm,
>> > as the patches deal with PM and need to be applied together
>> > to get the desired functionality.
>> > 
>> > Paul Walmsley (2):
>> >   Watchdog: omap_wdt: add fine grain runtime-pm
>> >   OMAP3: wdtimer: Fix CORE idle transition
>> > 
>> >  arch/arm/mach-omap2/omap_hwmod_3xxx_data.c |    1 +
>> >  drivers/watchdog/omap_wdt.c                |   25 +++++++++++++++++++++++--
>> >  2 files changed, 24 insertions(+), 2 deletions(-)
>> 
>> Acked-by: Wim Van Sebroeck <wim@iguana.be>
>
> Thanks.
>  
>> linux-omap-pm is indeed the best tree for getting this upstream.
>
> Kevin, do you want to take the drivers/watchdog/omap_wdt.c patch?  The 
> first patch is queued in my branch currently.

Go ahead and merge with the first patch.

Acked-by: Kevin Hilman <khilman@ti.com>

Re: [PATCH v2 0/2] OMAP3: wdtimer: Fix CORE idle

[Paul Walmsley]

On Thu, 10 Mar 2011, Kevin Hilman wrote:

> Go ahead and merge with the first patch.
> 
> Acked-by: Kevin Hilman <khilman@ti.com>

Done.


- Paul

RE: [PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition

[kalle.jokiniemi]

Hi Paul,

> -----Original Message-----
> From: ext Paul Walmsley [mailto:paul@pwsan.com]
> Sent: 10. maaliskuuta 2011 16:53
> > From: Paul Walmsley <paul@pwsan.com>
> >
> > The HW superwised smart idle for wdtimer in OMAP3 prevents CORE
> power
> > domain idle transitions. Disable it by swithing to SW supervised
> > transitions.
> >
> > This could be a hardware bug in the OMAP3 wdtimer2 block.
> >
> > Signed-off-by: Kalle Jokiniemi <kalle.jokiniemi@nokia.com>
> > Signed-off-by: Paul Walmsley <paul@pwsan.com>
> > Cc: Benoit Cousson <b-cousson@ti.com>
> 
> I've queued this one for 2.6.39, since it seems pretty clear that target smart-
> idle mode doesn't work with this IP block.

What about the 1/2 patch? Is that going in, and to which kernel version (our integrator would like to put in a note when we can remove the back ported patch)

- Kalle

> 
> Probably we'll need to do the same thing for OMAP2420/2430, but I'm not
> going to touch those right now, since we're so close to the merge window.
> 
> 
> - Paul

RE: [PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition

[Paul Walmsley]

Hello Kalle,

On Mon, 14 Mar 2011, kalle.jokiniemi@nokia.com wrote:

> What about the 1/2 patch? Is that going in, and to which kernel version 
> (our integrator would like to put in a note when we can remove the back 
> ported patch)

It's queued for Linus for the upcoming merge window:

http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-2.6.git;a=commit;h=0503add9d251db22ba2f610fb8d9b7743a9786da

Thread here:

http://www.spinics.net/lists/linux-omap/msg48381.html

If there aren't further comments on the watchdog patch, and Linus is okay
with the branch, it should appear in 2.6.39.



- Paul

RE: [PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition

[kalle.jokiniemi]

Hi,

> -----Original Message-----
> From: linux-omap-owner@vger.kernel.org [mailto:linux-omap-
> owner@vger.kernel.org] On Behalf Of ext Paul Walmsley
> Sent: 14. maaliskuuta 2011 21:27
> To: Jokiniemi Kalle (Nokia-MS/Tampere)
> Cc: linux-watchdog@vger.kernel.org; wim@iguana.be; khilman@ti.com;
> linux-omap@vger.kernel.org; tony@atomide.com; jhnikula@gmail.com;
> Koskinen Ilkka (Nokia-MS/Tampere); b-cousson@ti.com
> Subject: RE: [PATCH 2/2] OMAP3: wdtimer: Fix CORE idle transition
> 
> Hello Kalle,
> 
> On Mon, 14 Mar 2011, kalle.jokiniemi@nokia.com wrote:
> 
> > What about the 1/2 patch? Is that going in, and to which kernel
> > version (our integrator would like to put in a note when we can remove
> > the back ported patch)
> 
> It's queued for Linus for the upcoming merge window:
> 
> http://git.kernel.org/?p=linux/kernel/git/tmlind/linux-omap-
> 2.6.git;a=commit;h=0503add9d251db22ba2f610fb8d9b7743a9786da
> 
> Thread here:
> 
> http://www.spinics.net/lists/linux-omap/msg48381.html
> 
> If there aren't further comments on the watchdog patch, and Linus is okay
> with the branch, it should appear in 2.6.39.

Thanks for the confirmation.

- Kalle


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