Skip to contentSkip to conversation

rtc-cmos: fix suspend/resume

PATCH In review Daniel Drake 4 participants 5 messages Dec 19, 2010 – Dec 29, 2010 Real time clock (RTC)

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/1 rtc-cmos: fix suspend/resume

From: Paul Fox <[email protected]>

rtc-cmos was setting suspend/resume hooks at the device_driver level. However, the platform bus code (drivers/base/platform.c) only looks for resume hooks at the dev_pm_ops level, or within the platform_driver.

Switch rtc_cmos to use dev_pm_ops so that suspend/resume code is executed again.

---
 drivers/rtc/rtc-cmos.c |   16 +++++++++-------
 1 files changed, 9 insertions(+), 7 deletions(-)

v2: incorporate feedback from Rafael J. Wysocki, fix tabs, make a bit more consistent with typical SIMPLE_DEV_PM_OPS users.

v3: remove const keyword already set by macro, thanks to Rafael

drivers/rtc/rtc-cmos.c +9 -7
index 5856167..dd8242d 100644
@@ -36,6 +36,7 @@
 #include <linux/platform_device.h>
 #include <linux/mod_devicetable.h>
 #include <linux/log2.h>
+#include <linux/pm.h>
 
 /* this is for "generic access to PC-style RTC" using CMOS_READ/CMOS_WRITE */
 #include <asm-generic/rtc.h>
@@ -850,7 +851,7 @@ static void __exit cmos_do_remove(struct device *dev)
 
 #ifdef	CONFIG_PM
 
-static int cmos_suspend(struct device *dev, pm_message_t mesg)
+static int cmos_suspend(struct device *dev)
 {
 	struct cmos_rtc	*cmos = dev_get_drvdata(dev);
 	unsigned char	tmp;
@@ -898,7 +899,7 @@ static int cmos_suspend(struct device *dev, pm_message_t mesg)
  */
 static inline int cmos_poweroff(struct device *dev)
 {
-	return cmos_suspend(dev, PMSG_HIBERNATE);
+	return cmos_suspend(dev);
 }
 
 static int cmos_resume(struct device *dev)
@@ -945,9 +946,9 @@ static int cmos_resume(struct device *dev)
 	return 0;
 }
 
+static SIMPLE_DEV_PM_OPS(cmos_pm_ops, cmos_suspend, cmos_resume);
+
 #else
-#define	cmos_suspend	NULL
-#define	cmos_resume	NULL
 
 static inline int cmos_poweroff(struct device *dev)
 {
@@ -1077,7 +1078,7 @@ static void __exit cmos_pnp_remove(struct pnp_dev *pnp)
 
 static int cmos_pnp_suspend(struct pnp_dev *pnp, pm_message_t mesg)
 {
-	return cmos_suspend(&pnp->dev, mesg);
+	return cmos_suspend(&pnp->dev);
 }
 
 static int cmos_pnp_resume(struct pnp_dev *pnp)
@@ -1157,8 +1158,9 @@ static struct platform_driver cmos_platform_driver = {
 	.shutdown	= cmos_platform_shutdown,
 	.driver = {
 		.name		= (char *) driver_name,
-		.suspend	= cmos_suspend,
-		.resume		= cmos_resume,
+#ifdef CONFIG_PM
+		.pm		= &cmos_pm_ops,
+#endif
 	}
 };
 
Signature · 1 line
1.7.3.3
Acked-by on v3 1/1
#1 lore raw
Quoted 11 lines · Daniel Drake

From: Paul Fox <[email protected]>

rtc-cmos was setting suspend/resume hooks at the device_driver level. However, the platform bus code (drivers/base/platform.c) only looks for resume hooks at the dev_pm_ops level, or within the platform_driver.

Switch rtc_cmos to use dev_pm_ops so that suspend/resume code is executed again.

Quoted 73 lines
---
 drivers/rtc/rtc-cmos.c |   16 +++++++++-------
 1 files changed, 9 insertions(+), 7 deletions(-)

v2: incorporate feedback from Rafael J. Wysocki, fix tabs, make a bit more consistent with typical SIMPLE_DEV_PM_OPS users.

v3: remove const keyword already set by macro, thanks to Rafael

drivers/rtc/rtc-cmos.c +9 -7
index 5856167..dd8242d 100644
@@ -36,6 +36,7 @@
 #include <linux/platform_device.h>
 #include <linux/mod_devicetable.h>
 #include <linux/log2.h>
+#include <linux/pm.h>
 
 /* this is for "generic access to PC-style RTC" using CMOS_READ/CMOS_WRITE */
 #include <asm-generic/rtc.h>
@@ -850,7 +851,7 @@ static void __exit cmos_do_remove(struct device *dev)
 
 #ifdef	CONFIG_PM
 
-static int cmos_suspend(struct device *dev, pm_message_t mesg)
+static int cmos_suspend(struct device *dev)
 {
 	struct cmos_rtc	*cmos = dev_get_drvdata(dev);
 	unsigned char	tmp;
@@ -898,7 +899,7 @@ static int cmos_suspend(struct device *dev, pm_message_t mesg)
  */
 static inline int cmos_poweroff(struct device *dev)
 {
-	return cmos_suspend(dev, PMSG_HIBERNATE);
+	return cmos_suspend(dev);
 }
 
 static int cmos_resume(struct device *dev)
@@ -945,9 +946,9 @@ static int cmos_resume(struct device *dev)
 	return 0;
 }
 
+static SIMPLE_DEV_PM_OPS(cmos_pm_ops, cmos_suspend, cmos_resume);
+
 #else
-#define	cmos_suspend	NULL
-#define	cmos_resume	NULL
 
 static inline int cmos_poweroff(struct device *dev)
 {
@@ -1077,7 +1078,7 @@ static void __exit cmos_pnp_remove(struct pnp_dev *pnp)
 
 static int cmos_pnp_suspend(struct pnp_dev *pnp, pm_message_t mesg)
 {
-	return cmos_suspend(&pnp->dev, mesg);
+	return cmos_suspend(&pnp->dev);
 }
 
 static int cmos_pnp_resume(struct pnp_dev *pnp)
@@ -1157,8 +1158,9 @@ static struct platform_driver cmos_platform_driver = {
 	.shutdown	= cmos_platform_shutdown,
 	.driver = {
 		.name		= (char *) driver_name,
-		.suspend	= cmos_suspend,
-		.resume		= cmos_resume,
+#ifdef CONFIG_PM
+		.pm		= &cmos_pm_ops,
+#endif
 	}
 };
 
#2 lore raw
In reply to Daniel Drake
Quoted 11 lines · On Sun, 19 Dec 2010 19:08:28 +0000 (GMT) Daniel Drake <[email protected]> wrote:

From: Paul Fox <[email protected]>

rtc-cmos was setting suspend/resume hooks at the device_driver level. However, the platform bus code (drivers/base/platform.c) only looks for resume hooks at the dev_pm_ops level, or within the platform_driver.

Switch rtc_cmos to use dev_pm_ops so that suspend/resume code is executed again.

Quoted 8 lines
---
 drivers/rtc/rtc-cmos.c |   16 +++++++++-------
 1 files changed, 9 insertions(+), 7 deletions(-)

v2: incorporate feedback from Rafael J. Wysocki, fix tabs, make a bit more consistent with typical SIMPLE_DEV_PM_OPS users.

v3: remove const keyword already set by macro, thanks to Rafael

It's unclear what the user-visible effects of this bug were. Machine fails to suspend? RTC loses its brains on resume? Something else?

That's really important information for a bugfix's changelog. Please never omit it.

I'm going to assume that whatever-the-behaviour-is is fairly serious, and that we want this patch in 2.6.37. So I tagged it for backporting into 2.6.37.1, as we're getting pretty close to 2.6.37.

The patch also applies to 2.6.36. Is it needed there? And in earlier kernels?

#3 lore raw
Quoted 18 lines · andrew
Quoted 15 lines · On Sun, 19 Dec 2010 19:08:28 +0000 (GMT) Daniel Drake <[email protected]> wrote:

From: Paul Fox <[email protected]>

rtc-cmos was setting suspend/resume hooks at the device_driver level. However, the platform bus code (drivers/base/platform.c) only looks for resume hooks at the dev_pm_ops level, or within the platform_driver.

Switch rtc_cmos to use dev_pm_ops so that suspend/resume code is executed again.

---
 drivers/rtc/rtc-cmos.c |   16 +++++++++-------
 1 files changed, 9 insertions(+), 7 deletions(-)
Quoted 7 lines
Quoted 4 lines

v2: incorporate feedback from Rafael J. Wysocki, fix tabs, make a bit more consistent with typical SIMPLE_DEV_PM_OPS users.

v3: remove const keyword already set by macro, thanks to Rafael

It's unclear what the user-visible effects of this bug were. Machine fails to suspend? RTC loses its brains on resume? Something else?

the user visible symptom in our (XO laptop) case was that rtcwake would fail to wake the laptop. the RTC alarm would expire, but the wakeup wasn't unmasked.

as for severity, the impact may have been reduced because if i recall correctly, the bug only affected platforms with CONFIG_PNP disabled.

paul

Quoted 6 lines

That's really important information for a bugfix's changelog. Please never omit it.

Quoted 6 lines

I'm going to assume that whatever-the-behaviour-is is fairly serious, and that we want this patch in 2.6.37. So I tagged it for backporting into 2.6.37.1, as we're getting pretty close to 2.6.37.

The patch also applies to 2.6.36. Is it needed there? And in earlier kernels?

=---------------------
paul fox, [email protected]

#4 lore raw
In reply to Andrew Morton
Quoted 5 lines · Andrew Morton

It's unclear what the user-visible effects of this bug were.  Machine fails to suspend?  RTC loses its brains on resume?  Something else?

That's really important information for a bugfix's changelog.  Please never omit it.

Sorry about that.
After looking in more detail, I agree with Paul. This bug only affects setups where no PNP RTC is available, or where CONFIG_PNP is disabled. It also affects OLPC's coming-soon addition of an XO-specific RTC driver.

So, a clear and simple fix, but not a wide-reaching bug in the first place. 2.6.38 would be fine, no real need for -stable backporting IMO.

Thanks,
Daniel