Hi,
CC drivers/power/power_supply_core.o
In file included from drivers/power/power_supply_core.c:18:
include/linux/power_supply.h:125: error: field 'changed_work' has incomplete type
drivers/power/power_supply_core.c: In function 'power_supply_changed_work':
drivers/power/power_supply_core.c:25: warning: type defaults to 'int' in declaration of '__mptr'
drivers/power/power_supply_core.c:25: warning: initialization from incompatible pointer type
drivers/power/power_supply_core.c: In function 'power_supply_changed':
drivers/power/power_supply_core.c:57: warning: implicit declaration of function 'schedule_work'
drivers/power/power_supply_core.c: In function 'power_supply_register':
drivers/power/power_supply_core.c:103: warning: implicit declaration of function 'INIT_WORK'
drivers/power/power_supply_core.c: In function 'power_supply_unregister':
drivers/power/power_supply_core.c:128: warning: implicit declaration of function 'flush_scheduled_work'
make[3]: *** [drivers/power/power_supply_core.o] Error 1
make[2]: *** [drivers/power] Error 2
make[1]: *** [drivers] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.21-mm2.agp'
make: *** [debian/stamp-build-kernel] Error 2
CONFIG_POWER_SUPPLY=y
# CONFIG_POWER_SUPPLY_DEBUG is not set
# CONFIG_PDA_POWER is not set
# CONFIG_BATTERY_DS2760 is not set
# CONFIG_BATTERY_OLPC is not set
P3/700 UP, gcc 4.1.3.
Care to add a #include <linux/workqueue.h> or so? ;)
(worked for me when adding to include/linux/power_supply.h)
Ugh, such a pain following ever-outdated and always-broken -mm versions ;)
Andreas Mohr
On Wed, May 09, 2007 at 11:08:20PM +0200, Andreas Mohr wrote:
> Hi,
>
> CC drivers/power/power_supply_core.o
> In file included from drivers/power/power_supply_core.c:18:
> include/linux/power_supply.h:125: error: field 'changed_work' has incomplete type
> drivers/power/power_supply_core.c: In function 'power_supply_changed_work':
> drivers/power/power_supply_core.c:25: warning: type defaults to 'int' in declaration of '__mptr'
> drivers/power/power_supply_core.c:25: warning: initialization from incompatible pointer type
> drivers/power/power_supply_core.c: In function 'power_supply_changed':
> drivers/power/power_supply_core.c:57: warning: implicit declaration of function 'schedule_work'
> drivers/power/power_supply_core.c: In function 'power_supply_register':
> drivers/power/power_supply_core.c:103: warning: implicit declaration of function 'INIT_WORK'
> drivers/power/power_supply_core.c: In function 'power_supply_unregister':
> drivers/power/power_supply_core.c:128: warning: implicit declaration of function 'flush_scheduled_work'
> make[3]: *** [drivers/power/power_supply_core.o] Error 1
> make[2]: *** [drivers/power] Error 2
> make[1]: *** [drivers] Error 2
> make[1]: Leaving directory `/usr/src/linux-2.6.21-mm2.agp'
> make: *** [debian/stamp-build-kernel] Error 2
>
> CONFIG_POWER_SUPPLY=y
> # CONFIG_POWER_SUPPLY_DEBUG is not set
> # CONFIG_PDA_POWER is not set
> # CONFIG_BATTERY_DS2760 is not set
> # CONFIG_BATTERY_OLPC is not set
>
> P3/700 UP, gcc 4.1.3.
>
> Care to add a #include <linux/workqueue.h> or so? ;)
> (worked for me when adding to include/linux/power_supply.h)
>
> Ugh, such a pain following ever-outdated and always-broken -mm versions ;)
>
> Andreas Mohr
http://lkml.org/lkml/2007/5/9/163
Already fixed in battery2-2.6.git. But thanks anyway for reporting. ;-)
I really do test builds prior commiting them, but headers changed
recently, and now including workqueue.h should be explicit. Wondering why
this worked prior, though. ;-)
Thanks,
--
Anton Vorontsov
email: [email protected]
backup email: [email protected]
irc://irc.freenode.org/bd2
On Thu, May 10, 2007 at 01:27:42AM +0400, Anton Vorontsov wrote:
> On Wed, May 09, 2007 at 11:08:20PM +0200, Andreas Mohr wrote:
> > Care to add a #include <linux/workqueue.h> or so? ;)
> > (worked for me when adding to include/linux/power_supply.h)
> >
> > Ugh, such a pain following ever-outdated and always-broken -mm versions ;)
> >
> > Andreas Mohr
>
> http://lkml.org/lkml/2007/5/9/163
Crap, I did grep for a changed_work report but came up empty thus reported it.
Hmm, incomplete LKML mbox here?
> I really do test builds prior commiting them, but headers changed
> recently, and now including workqueue.h should be explicit. Wondering why
> this worked prior, though. ;-)
Right, several header cleanup efforts recently (which is a good thing).
Note to self: always reply to version announce mail when reporting problems...
Andreas Mohr