Re: [PATCH] acpi: fix pwr_btn_event_pending build error

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Sun, Jul 22, 2012 at 12:52:43AM +0300, Kirill A. Shutemov wrote:
> On Sat, Jul 21, 2012 at 02:41:56PM -0700, Paul E. McKenney wrote:
> > Hello!
> > 
> > A recent build for KVM on x86 resulted in the following build error:
> > 
> > drivers/acpi/sleep.c:60:13: warning: ‘pwr_btn_event_pending’ defined but not
> > +used [-Wunused-variable]
> > 
> > This patch silences this error.
> > 
> > Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
> > 
> > diff --git a/drivers/acpi/sleep.c b/drivers/acpi/sleep.c
> > index 8856102..18347aa 100644
> > --- a/drivers/acpi/sleep.c
> > +++ b/drivers/acpi/sleep.c
> > @@ -57,7 +57,9 @@ MODULE_PARM_DESC(gts, "Enable evaluation of _GTS on suspend.");
> >  MODULE_PARM_DESC(bfs, "Enable evaluation of _BFS on resume".);
> >  
> >  static u8 sleep_states[ACPI_S_STATE_COUNT];
> > +#ifdef CONFIG_ACPI_SLEEP
> >  static bool pwr_btn_event_pending;
> > +#endif /* #ifdef CONFIG_ACPI_SLEEP */
> 
> __maybe_unused ?

Are the compiler and linker smart enough to get rid of the storage in
case it really isn't used?  (Not that the storage for a bool is worth
worrying about, but...)

							Thanx, Paul