Re: [PATCH] x86, mce: fix __init annotations

[Yong Wang <yong.y.wang@linux.intel.com>]

On Thu, Nov 12, 2009 at 03:52:40PM +0900, Hidetoshi Seto wrote:
> Hidetoshi Seto wrote:
> > Yong Wang wrote:
> >> @@ -268,7 +268,7 @@ void mcheck_intel_therm_init(void)
> >>  		lvtthmr_init = apic_read(APIC_LVTTHMR);
> >>  }
> >>  
> >> -void intel_init_thermal(struct cpuinfo_x86 *c)
> >> +void __init intel_init_thermal(struct cpuinfo_x86 *c)
> >>  {
> >>  	unsigned int cpu = smp_processor_id();
> >>  	int tm2 = 0;
> > 
> > But I think this could be called on hot-plugged AP.
> > Should be __cpuinit ?
> > 
> > void __cpuinit mcheck_cpu_init(struct cpuinfo_x86 *c)
> >  ->  static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c)
> >   ->  void mce_intel_feature_init(struct cpuinfo_x86 *c)
> >    ->  void intel_init_thermal(struct cpuinfo_x86 *c)
> 
> I found it is called from resume path, so neither __init nor __cpuinit
> will be applicable...
> 
> static int mce_resume(struct sys_device *dev)
>  -> static void __mcheck_cpu_init_vendor(struct cpuinfo_x86 *c)
>   ->  void mce_intel_feature_init(struct cpuinfo_x86 *c)
>    ->  void intel_init_thermal(struct cpuinfo_x86 *c)
> 
> Following patch is based on tip:perf/mce.
> 
> 
> Thanks,
> H.Seto
> 
> ===
> 
> Subject: [PATCH] x86, mce: fix __init annotations
> 
> The intel_init_thermal() is called from resume path, so it cannot be
> marked as __init.
> 
> OTOH mce_banks_init() is only called from __mcheck_cpu_cap_init()
> which is marked as __cpuinit, so it can be also marked as __cpuinit.
> 
> Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
> ---
>  arch/x86/kernel/cpu/mcheck/mce.c         |    4 ++--
>  arch/x86/kernel/cpu/mcheck/therm_throt.c |    2 +-
>  2 files changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
> index 0d41020..5f277ca 100644
> --- a/arch/x86/kernel/cpu/mcheck/mce.c
> +++ b/arch/x86/kernel/cpu/mcheck/mce.c
> @@ -1201,7 +1201,7 @@ int mce_notify_irq(void)
>  }
>  EXPORT_SYMBOL_GPL(mce_notify_irq);
>  
> -static int mce_banks_init(void)
> +static int __cpuinit __mcheck_cpu_mce_banks_init(void)
>  {
>  	int i;
>  
> @@ -1242,7 +1242,7 @@ static int __cpuinit __mcheck_cpu_cap_init(void)
>  	WARN_ON(banks != 0 && b != banks);
>  	banks = b;
>  	if (!mce_banks) {
> -		int err = mce_banks_init();
> +		int err = __mcheck_cpu_mce_banks_init();
>  
>  		if (err)
>  			return err;
> diff --git a/arch/x86/kernel/cpu/mcheck/therm_throt.c b/arch/x86/kernel/cpu/mcheck/therm_throt.c
> index 8a73d5c..4fef985 100644
> --- a/arch/x86/kernel/cpu/mcheck/therm_throt.c
> +++ b/arch/x86/kernel/cpu/mcheck/therm_throt.c
> @@ -268,7 +268,7 @@ void __init mcheck_intel_therm_init(void)
>  		lvtthmr_init = apic_read(APIC_LVTTHMR);
>  }
>  
> -void __init intel_init_thermal(struct cpuinfo_x86 *c)
> +void intel_init_thermal(struct cpuinfo_x86 *c)
>  {
>  	unsigned int cpu = smp_processor_id();
>  	int tm2 = 0;
> 

Oh, yeah. We should only mark mcheck_intel_therm_init as __init for the
thermal part. Thanks for spotting.