linux-next20080314 build fails with !CONFIG_PM

PATCH New Kamalesh Babulal 4 participants 11 messages Mar 14, 2008 – Mar 15, 2008

Subsystems (4)

Driver core Suspend to RAM Hibernation Power management core

lore

Share & tools

lore

Kamalesh Babulal

author

2008-03-14 07:57 UTC #0 lore raw

The next-20080314 tree build fails

drivers/serial/serial_core.c: In function `uart_add_one_port':
drivers/serial/serial_core.c:2359: error: invalid lvalue in assignment make[2]: *** [drivers/serial/serial_core.o] Error 1

The config # CONFIG_PM was not set.The code which is causing the build failure is

	device_can_wakeup(tty_dev) = 1;

when the CONFIG_PM is set the macro is preprocessed as

#define device_can_wakeup(dev) \
        ((dev)->power.can_wakeup)

and when not set, it becomes 0 = 1

#define device_can_wakeup(dev) 0

Signature · 4 lines

Thanks & Regards,
Kamalesh Babulal,
Linux Technology Center,
IBM, ISTL.

Andrew Morton

2008-03-14 22:07 UTC #1 lore raw

Quoted 11 lines · On Fri, 14 Mar 2008 13:27:10 +0530 Kamalesh Babulal…

The next-20080314 tree build fails
drivers/serial/serial_core.c: In function `uart_add_one_port':
drivers/serial/serial_core.c:2359: error: invalid lvalue in assignment make[2]: *** [drivers/serial/serial_core.o] Error 1
The config # CONFIG_PM was not set.The code which is causing the build failure is
	device_can_wakeup(tty_dev) = 1;

Quoted 9 lines

when the CONFIG_PM is set the macro is preprocessed as
#define device_can_wakeup(dev) \
        ((dev)->power.can_wakeup)
and when not set, it becomes 0 = 1
#define device_can_wakeup(dev) 0

Caused by Alan's "PM: make wakeup flags available whenever CONFIG_PM is set" which I assume Len merged.

Sorry, but that code should be dragged out and shot. Doing this:

Quoted 1 line

	device_can_wakeup(tty_dev) = 1;

is just gross and stupid. It looks daft, it isn't C and it *requires* that device_can_wakeup() be implemented as a macro, which totally busts any concept of abstraction.

The code previously had quite reasonable accessors:

 #define device_set_wakeup_enable(dev,val) \
        ((dev)->power.should_wakeup = !!(val))
 #define device_may_wakeup(dev) \
        (device_can_wakeup(dev) && (dev)->power.should_wakeup)

can we please go back to that scheme? Please also convert all these magical macros into inlined C functions. One reason is that this:

+#define device_init_wakeup(dev,val) \
+ do { \
+ device_can_wakeup(dev) = !!(val); \
+ device_set_wakeup_enable(dev,val); \
+ } while(0)

is buggy. It evaluates its arguments multiple times and will cause failures when passed expressions which have side-effects.

Alan, please also pass all future patches through checkpatch - there is no need to be adding trivial coding-style errors in this day and age.

Thanks.

Rafael J. Wysocki

2008-03-14 22:38 UTC #2 lore raw

Quoted 14 lines · Andrew Morton

Quoted 11 lines · On Fri, 14 Mar 2008 13:27:10 +0530 Kamalesh Babulal…
The next-20080314 tree build fails
drivers/serial/serial_core.c: In function `uart_add_one_port':
drivers/serial/serial_core.c:2359: error: invalid lvalue in assignment make[2]: *** [drivers/serial/serial_core.o] Error 1
The config # CONFIG_PM was not set.The code which is causing the build failure is
	device_can_wakeup(tty_dev) = 1;

Quoted 8 lines

Quoted 8 lines
when the CONFIG_PM is set the macro is preprocessed as
#define device_can_wakeup(dev) \
        ((dev)->power.can_wakeup)
and when not set, it becomes 0 = 1
#define device_can_wakeup(dev) 0

Kamalesh, thanks for reporting the problem.

Quoted 2 lines

Caused by Alan's "PM: make wakeup flags available whenever CONFIG_PM is set" which I assume Len merged.

It's in the Greg's tree actually.

Quoted 25 lines

Sorry, but that code should be dragged out and shot. Doing this:
Quoted 1 line
	device_can_wakeup(tty_dev) = 1;
is just gross and stupid. It looks daft, it isn't C and it *requires* that device_can_wakeup() be implemented as a macro, which totally busts any concept of abstraction.
The code previously had quite reasonable accessors:
 #define device_set_wakeup_enable(dev,val) \
        ((dev)->power.should_wakeup = !!(val))
 #define device_may_wakeup(dev) \
        (device_can_wakeup(dev) && (dev)->power.should_wakeup)
can we please go back to that scheme? Please also convert all these magical macros into inlined C functions. One reason is that this:
(quoted diff) +5 -0
+#define device_init_wakeup(dev,val) \
+       do { \
+               device_can_wakeup(dev) = !!(val); \
+               device_set_wakeup_enable(dev,val); \
+       } while(0)

Quoted 7 lines

is buggy. It evaluates its arguments multiple times and will cause failures when passed expressions which have side-effects.
Alan, please also pass all future patches through checkpatch - there is no need to be adding trivial coding-style errors in this day and age.
Thanks.

Well, Greg please drop the "PM: make wakeup flags available whenever CONFIG_PM is set" and Alan please resubmit the patch with the problems pointed out by Andrew fixed.

In fact I'd even prefer it to be two patches, one that moves should_wakeup and the related macros out of the CONFIG_PM_SLEEP #ifdef, because that's what fixes the problem described in the changelog, and one that makes the remaining changes with a separate changelog.

Thanks,
Rafael

Alan Stern

2008-03-15 21:53 UTC #3 lore raw

cover PM wakeup flags revisited

Quoted 9 lines · Rafael J. Wysocki

Quoted 7 lines · Andrew Morton
Sorry, but that code should be dragged out and shot. Doing this:
Quoted 1 line
	device_can_wakeup(tty_dev) = 1;
is just gross and stupid. It looks daft, it isn't C and it *requires* that device_can_wakeup() be implemented as a macro, which totally busts any concept of abstraction.

It seems to have been a one-time occurrence. This patch series fixes it.

Quoted 10 lines

Quoted 10 lines
The code previously had quite reasonable accessors:
 #define device_set_wakeup_enable(dev,val) \
        ((dev)->power.should_wakeup = !!(val))
 #define device_may_wakeup(dev) \
        (device_can_wakeup(dev) && (dev)->power.should_wakeup)
can we please go back to that scheme? Please also convert all these magical macros into inlined C functions. One reason is that this:

Quoted 8 lines

Quoted 8 lines
(quoted diff) +5 -0
+#define device_init_wakeup(dev,val) \
+       do { \
+               device_can_wakeup(dev) = !!(val); \
+               device_set_wakeup_enable(dev,val); \
+       } while(0)
is buggy. It evaluates its arguments multiple times and will cause failures when passed expressions which have side-effects.

This series converts the macros to inline functions.

Quoted 2 lines

Quoted 2 lines
Alan, please also pass all future patches through checkpatch - there is no need to be adding trivial coding-style errors in this day and age.

Rest assurred that checkpath is now an integral part of my normal workflow. All the patches in this series pass with flying colors.

Quoted 8 lines

Well, Greg please drop the "PM: make wakeup flags available whenever CONFIG_PM is set" and Alan please resubmit the patch with the problems pointed out by Andrew fixed.
In fact I'd even prefer it to be two patches, one that moves should_wakeup and the related macros out of the CONFIG_PM_SLEEP #ifdef, because that's what fixes the problem described in the changelog, and one that makes the remaining changes with a separate changelog.

Done.

The 1/3 patch fixes the unfortunate code in the serial-core driver.

The 2/3 patch moves the macros out from under CONFIG_PM_SLEEP.

The 3/3 patch converts the macros to inline functions and creates a new pm_wakeup.h file for them. They can't remain in pm.h, because they depend on the definition of struct device -- and the compiler hasn't seen that yet when pm.h is read.

It's not clear why the can_wakeup accessors are written to work even when CONFIG_PM isn't set. Nevertheless, the patches retain that behavior.

Alan Stern

2008-03-15 21:53 UTC #4 lore raw

PATCH 1/3 Fix misuse of wakeup flag accessors in serial core

In reply to Rafael J. Wysocki

This patch (as1059) fixes a mistake in the way the serial core initializes a device's wakeup settings. It should use the accessor routine instead of relying on a macro producing an lvalue.

Signed-off-by Alan Stern <[email protected]>

---

drivers/serial/serial_core.c +1 -1

Index: usb-2.6/drivers/serial/serial_core.c
===================================================================
@@ -2356,7 +2356,7 @@ int uart_add_one_port(struct uart_driver
 	 */
 	tty_dev = tty_register_device(drv->tty_driver, port->line, port->dev);
 	if (likely(!IS_ERR(tty_dev))) {
-		device_can_wakeup(tty_dev) = 1;
+		device_init_wakeup(tty_dev, 1);
 		device_set_wakeup_enable(tty_dev, 0);
 	} else
 		printk(KERN_ERR "Cannot register tty device on line %d\n",

Alan Stern

2008-03-15 21:53 UTC #5 lore raw

cover PM wakeup flags revisited

In reply to Rafael J. Wysocki

Quoted 9 lines · Rafael J. Wysocki

Quoted 7 lines · Andrew Morton
Sorry, but that code should be dragged out and shot. Doing this:
Quoted 1 line
	device_can_wakeup(tty_dev) = 1;
is just gross and stupid. It looks daft, it isn't C and it *requires* that device_can_wakeup() be implemented as a macro, which totally busts any concept of abstraction.

It seems to have been a one-time occurrence. This patch series fixes it.

Quoted 10 lines

Quoted 10 lines
The code previously had quite reasonable accessors:
 #define device_set_wakeup_enable(dev,val) \
        ((dev)->power.should_wakeup = !!(val))
 #define device_may_wakeup(dev) \
        (device_can_wakeup(dev) && (dev)->power.should_wakeup)
can we please go back to that scheme? Please also convert all these magical macros into inlined C functions. One reason is that this:

Quoted 8 lines

Quoted 8 lines
(quoted diff) +5 -0
+#define device_init_wakeup(dev,val) \
+       do { \
+               device_can_wakeup(dev) = !!(val); \
+               device_set_wakeup_enable(dev,val); \
+       } while(0)
is buggy. It evaluates its arguments multiple times and will cause failures when passed expressions which have side-effects.

This series converts the macros to inline functions.

Quoted 2 lines

Quoted 2 lines
Alan, please also pass all future patches through checkpatch - there is no need to be adding trivial coding-style errors in this day and age.

Rest assurred that checkpath is now an integral part of my normal workflow. All the patches in this series pass with flying colors.

Quoted 8 lines

Well, Greg please drop the "PM: make wakeup flags available whenever CONFIG_PM is set" and Alan please resubmit the patch with the problems pointed out by Andrew fixed.
In fact I'd even prefer it to be two patches, one that moves should_wakeup and the related macros out of the CONFIG_PM_SLEEP #ifdef, because that's what fixes the problem described in the changelog, and one that makes the remaining changes with a separate changelog.

Done.

The 1/3 patch fixes the unfortunate code in the serial-core driver.

The 2/3 patch moves the macros out from under CONFIG_PM_SLEEP.

It's not clear why the can_wakeup accessors are written to work even when CONFIG_PM isn't set. Nevertheless, the patches retain that behavior.

Alan Stern

2008-03-15 21:53 UTC #6 lore raw

PATCH 1/3 Fix misuse of wakeup flag accessors in serial core

In reply to Rafael J. Wysocki

This patch (as1059) fixes a mistake in the way the serial core initializes a device's wakeup settings. It should use the accessor routine instead of relying on a macro producing an lvalue.

Signed-off-by Alan Stern <[email protected]>

---

drivers/serial/serial_core.c +1 -1

Index: usb-2.6/drivers/serial/serial_core.c
===================================================================
@@ -2356,7 +2356,7 @@ int uart_add_one_port(struct uart_driver
 	 */
 	tty_dev = tty_register_device(drv->tty_driver, port->line, port->dev);
 	if (likely(!IS_ERR(tty_dev))) {
-		device_can_wakeup(tty_dev) = 1;
+		device_init_wakeup(tty_dev, 1);
 		device_set_wakeup_enable(tty_dev, 0);
 	} else
 		printk(KERN_ERR "Cannot register tty device on line %d\n",

Alan Stern

2008-03-15 21:53 UTC #7 lore raw

PATCH 2/3 PM: make wakeup flags available whenever CONFIG_PM is set (ver 2)

In reply to Rafael J. Wysocki

The various wakeup flags and their accessor macros in struct dev_pm_info should be available whenever CONFIG_PM is enabled, not just when CONFIG_PM_SLEEP is on. Otherwise remote wakeup won't always be configurable for runtime power management. This patch (as1056b) fixes the oversight.

Signed-off-by Alan Stern <[email protected]>

---

include/linux/pm.h +21 ~~-15~~

Index: usb-2.6/include/linux/pm.h
===================================================================
@@ -183,8 +183,8 @@ typedef struct pm_message {
 struct dev_pm_info {
 	pm_message_t		power_state;
 	unsigned		can_wakeup:1;
-#ifdef	CONFIG_PM_SLEEP
 	unsigned		should_wakeup:1;
+#ifdef	CONFIG_PM_SLEEP
 	struct list_head	entry;
 #endif
 };

@@ -197,11 +197,6 @@ extern void device_resume(void);
 extern int device_suspend(pm_message_t state);
 extern int device_prepare_suspend(pm_message_t state);
 
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
 extern void __suspend_report_result(const char *function, void *fn, int ret);
 
 #define suspend_report_result(fn, ret)					\

@@ -209,6 +204,24 @@ extern void __suspend_report_result(cons
 		__suspend_report_result(__FUNCTION__, fn, ret);		\
 	} while (0)
 
+#else /* !CONFIG_PM_SLEEP */
+
+static inline int device_suspend(pm_message_t state)
+{
+	return 0;
+}
+
+#define suspend_report_result(fn, ret) do { } while (0)
+
+#endif /* !CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+
+#define device_set_wakeup_enable(dev,val) \
+	((dev)->power.should_wakeup = !!(val))
+#define device_may_wakeup(dev) \
+	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
+
 /*
  * Platform hook to activate device wakeup capability, if that's not already
  * handled by enable_irq_wake() etc.

@@ -223,24 +236,17 @@ static inline int call_platform_enable_w
 	return 0;
 }
 
-#else /* !CONFIG_PM_SLEEP */
-
-static inline int device_suspend(pm_message_t state)
-{
-	return 0;
-}
+#else /* !CONFIG_PM */
 
 #define device_set_wakeup_enable(dev,val)	do{}while(0)
 #define device_may_wakeup(dev)			(0)
 
-#define suspend_report_result(fn, ret) do { } while (0)
-
 static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
 {
 	return 0;
 }
 
-#endif /* !CONFIG_PM_SLEEP */
+#endif /* !CONFIG_PM */
 
 /* changes to device_may_wakeup take effect on the next pm state change.
  * by default, devices should wakeup if they can.

drivers/base/power/main.c +0 -2

Index: usb-2.6/drivers/base/power/main.c
===================================================================
@@ -56,8 +56,6 @@ static DEFINE_MUTEX(dpm_list_mtx);
 
 static DECLARE_RWSEM(pm_sleep_rwsem);
 
-int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
 /**
  *	device_pm_add - add a device to the list of active devices
  *	@dev:	Device to be added to the list

drivers/base/power/sysfs.c +2 -0

Index: usb-2.6/drivers/base/power/sysfs.c
===================================================================
@@ -6,6 +6,8 @@
 #include <linux/string.h>
 #include "power.h"
 
+int (*platform_enable_wakeup)(struct device *dev, int is_on);
+
 
 /*
  *	wakeup - Report/change current wakeup option for device

Alan Stern

2008-03-15 21:53 UTC #8 lore raw

PATCH 2/3 PM: make wakeup flags available whenever CONFIG_PM is set (ver 2)

In reply to Rafael J. Wysocki

Signed-off-by Alan Stern <[email protected]>

---

include/linux/pm.h +21 ~~-15~~

Index: usb-2.6/include/linux/pm.h
===================================================================
@@ -183,8 +183,8 @@ typedef struct pm_message {
 struct dev_pm_info {
 	pm_message_t		power_state;
 	unsigned		can_wakeup:1;
-#ifdef	CONFIG_PM_SLEEP
 	unsigned		should_wakeup:1;
+#ifdef	CONFIG_PM_SLEEP
 	struct list_head	entry;
 #endif
 };

@@ -197,11 +197,6 @@ extern void device_resume(void);
 extern int device_suspend(pm_message_t state);
 extern int device_prepare_suspend(pm_message_t state);
 
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
 extern void __suspend_report_result(const char *function, void *fn, int ret);
 
 #define suspend_report_result(fn, ret)					\

@@ -209,6 +204,24 @@ extern void __suspend_report_result(cons
 		__suspend_report_result(__FUNCTION__, fn, ret);		\
 	} while (0)
 
+#else /* !CONFIG_PM_SLEEP */
+
+static inline int device_suspend(pm_message_t state)
+{
+	return 0;
+}
+
+#define suspend_report_result(fn, ret) do { } while (0)
+
+#endif /* !CONFIG_PM_SLEEP */
+
+#ifdef CONFIG_PM
+
+#define device_set_wakeup_enable(dev,val) \
+	((dev)->power.should_wakeup = !!(val))
+#define device_may_wakeup(dev) \
+	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
+
 /*
  * Platform hook to activate device wakeup capability, if that's not already
  * handled by enable_irq_wake() etc.

@@ -223,24 +236,17 @@ static inline int call_platform_enable_w
 	return 0;
 }
 
-#else /* !CONFIG_PM_SLEEP */
-
-static inline int device_suspend(pm_message_t state)
-{
-	return 0;
-}
+#else /* !CONFIG_PM */
 
 #define device_set_wakeup_enable(dev,val)	do{}while(0)
 #define device_may_wakeup(dev)			(0)
 
-#define suspend_report_result(fn, ret) do { } while (0)
-
 static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
 {
 	return 0;
 }
 
-#endif /* !CONFIG_PM_SLEEP */
+#endif /* !CONFIG_PM */
 
 /* changes to device_may_wakeup take effect on the next pm state change.
  * by default, devices should wakeup if they can.

drivers/base/power/main.c +0 -2

Index: usb-2.6/drivers/base/power/main.c
===================================================================
@@ -56,8 +56,6 @@ static DEFINE_MUTEX(dpm_list_mtx);
 
 static DECLARE_RWSEM(pm_sleep_rwsem);
 
-int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
 /**
  *	device_pm_add - add a device to the list of active devices
  *	@dev:	Device to be added to the list

drivers/base/power/sysfs.c +2 -0

Index: usb-2.6/drivers/base/power/sysfs.c
===================================================================
@@ -6,6 +6,8 @@
 #include <linux/string.h>
 #include "power.h"
 
+int (*platform_enable_wakeup)(struct device *dev, int is_on);
+
 
 /*
  *	wakeup - Report/change current wakeup option for device

Alan Stern

2008-03-15 21:54 UTC #9 lore raw

PATCH 3/3 PM: convert wakeup flag accessors to inline functions

In reply to Rafael J. Wysocki

This patch (as1058) improves the wakeup macros in include/linux/pm.h. All but the trivial ones are converted to inline routines, which requires moving them to a separate header file since they depend on the definition of struct device.

Signed-off-by Alan Stern <[email protected]>

---

include/linux/pm.h +1 ~~-45~~

Index: usb-2.6/include/linux/pm.h
===================================================================
@@ -211,54 +211,10 @@ static inline int device_suspend(pm_mess
 	return 0;
 }
 
-#define suspend_report_result(fn, ret) do { } while (0)
+#define suspend_report_result(fn, ret)		do {} while (0)
 
 #endif /* !CONFIG_PM_SLEEP */
 
-#ifdef CONFIG_PM
-
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
-/*
- * Platform hook to activate device wakeup capability, if that's not already
- * handled by enable_irq_wake() etc.
- * Returns zero on success, else negative errno
- */
-extern int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	if (platform_enable_wakeup)
-		return (*platform_enable_wakeup)(dev, is_on);
-	return 0;
-}
-
-#else /* !CONFIG_PM */
-
-#define device_set_wakeup_enable(dev,val)	do{}while(0)
-#define device_may_wakeup(dev)			(0)
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	return 0;
-}
-
-#endif /* !CONFIG_PM */
-
-/* changes to device_may_wakeup take effect on the next pm state change.
- * by default, devices should wakeup if they can.
- */
-#define device_can_wakeup(dev) \
-	((dev)->power.can_wakeup)
-#define device_init_wakeup(dev,val) \
-	do { \
-		device_can_wakeup(dev) = !!(val); \
-		device_set_wakeup_enable(dev,val); \
-	} while(0)
-
 /*
  * Global Power Management flags
  * Used to keep APM and ACPI from both being active

include/linux/device.h +3 -0

Index: usb-2.6/include/linux/device.h
===================================================================
@@ -475,6 +475,9 @@ struct device {
 	void	(*release)(struct device *dev);
 };
 
+/* Get the wakeup routines, which depend on struct device */
+#include <linux/pm_wakeup.h>
+
 #ifdef CONFIG_NUMA
 static inline int dev_to_node(struct device *dev)
 {

include/linux/pm_wakeup.h new file +90 -0

90 lines not on this page · raw message

Alan Stern

2008-03-15 21:54 UTC #10 lore raw

PATCH 3/3 PM: convert wakeup flag accessors to inline functions

In reply to Rafael J. Wysocki

Signed-off-by Alan Stern <[email protected]>

---

include/linux/pm.h +1 ~~-45~~

Index: usb-2.6/include/linux/pm.h
===================================================================
@@ -211,54 +211,10 @@ static inline int device_suspend(pm_mess
 	return 0;
 }
 
-#define suspend_report_result(fn, ret) do { } while (0)
+#define suspend_report_result(fn, ret)		do {} while (0)
 
 #endif /* !CONFIG_PM_SLEEP */
 
-#ifdef CONFIG_PM
-
-#define device_set_wakeup_enable(dev,val) \
-	((dev)->power.should_wakeup = !!(val))
-#define device_may_wakeup(dev) \
-	(device_can_wakeup(dev) && (dev)->power.should_wakeup)
-
-/*
- * Platform hook to activate device wakeup capability, if that's not already
- * handled by enable_irq_wake() etc.
- * Returns zero on success, else negative errno
- */
-extern int (*platform_enable_wakeup)(struct device *dev, int is_on);
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	if (platform_enable_wakeup)
-		return (*platform_enable_wakeup)(dev, is_on);
-	return 0;
-}
-
-#else /* !CONFIG_PM */
-
-#define device_set_wakeup_enable(dev,val)	do{}while(0)
-#define device_may_wakeup(dev)			(0)
-
-static inline int call_platform_enable_wakeup(struct device *dev, int is_on)
-{
-	return 0;
-}
-
-#endif /* !CONFIG_PM */
-
-/* changes to device_may_wakeup take effect on the next pm state change.
- * by default, devices should wakeup if they can.
- */
-#define device_can_wakeup(dev) \
-	((dev)->power.can_wakeup)
-#define device_init_wakeup(dev,val) \
-	do { \
-		device_can_wakeup(dev) = !!(val); \
-		device_set_wakeup_enable(dev,val); \
-	} while(0)
-
 /*
  * Global Power Management flags
  * Used to keep APM and ACPI from both being active

include/linux/device.h +3 -0

Index: usb-2.6/include/linux/device.h
===================================================================
@@ -475,6 +475,9 @@ struct device {
 	void	(*release)(struct device *dev);
 };
 
+/* Get the wakeup routines, which depend on struct device */
+#include <linux/pm_wakeup.h>
+
 #ifdef CONFIG_NUMA
 static inline int dev_to_node(struct device *dev)
 {

include/linux/pm_wakeup.h new file +90 -0

90 lines not on this page · raw message