Re: x86: clean up non-smp usage of cpu maps

Re: x86: clean up non-smp usage of cpu maps

[Andrew Morton]

On Fri, 18 Apr 2008 17:36:55 GMT Linux Kernel Mailing List <linux-kernel@vger.kernel.org> wrote:

>     x86: clean up non-smp usage of cpu maps
>     
>     Cleanup references to the early cpu maps for the non-SMP configuration
>     and remove some functions called for SMP configurations only.
>     
> ...
>
> index 1179aa0..dc79409 100644
> --- a/arch/x86/kernel/setup.c
> +++ b/arch/x86/kernel/setup.c
> @@ -10,7 +10,7 @@
>  #include <asm/setup.h>
>  #include <asm/topology.h>
>  
> -#ifdef CONFIG_HAVE_SETUP_PER_CPU_AREA
> +#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_SMP)

This implies that we can have

	CONFIG_HAVE_SETUP_PER_CPU_AREA && !CONFIG_SMP

which a) appears to be true and b) seems rather dumb?

Re: x86: clean up non-smp usage of cpu maps

[Mike Travis]

Andrew Morton wrote:
> On Fri, 18 Apr 2008 17:36:55 GMT Linux Kernel Mailing List <linux-kernel@vger.kernel.org> wrote:
> 
>>     x86: clean up non-smp usage of cpu maps
>>     
>>     Cleanup references to the early cpu maps for the non-SMP configuration
>>     and remove some functions called for SMP configurations only.
>>     
>> ...
>>
>> index 1179aa0..dc79409 100644
>> --- a/arch/x86/kernel/setup.c
>> +++ b/arch/x86/kernel/setup.c
>> @@ -10,7 +10,7 @@
>>  #include <asm/setup.h>
>>  #include <asm/topology.h>
>>  
>> -#ifdef CONFIG_HAVE_SETUP_PER_CPU_AREA
>> +#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_SMP)
> 
> This implies that we can have
> 
> 	CONFIG_HAVE_SETUP_PER_CPU_AREA && !CONFIG_SMP
> 
> which a) appears to be true and b) seems rather dumb?

Yeah, I kind of picked that up in a code merge and didn't really look at it
too closely.  I should have checked the Kconfig file to make sure that it
couldn't inadvertently get set if SMP is not set.

Thanks,
Mike

Re: x86: clean up non-smp usage of cpu maps

[Andrew Morton]

On Fri, 18 Apr 2008 19:21:50 -0700 Mike Travis <travis@sgi.com> wrote:

> Andrew Morton wrote:
> > On Fri, 18 Apr 2008 17:36:55 GMT Linux Kernel Mailing List <linux-kernel@vger.kernel.org> wrote:
> > 
> >>     x86: clean up non-smp usage of cpu maps
> >>     
> >>     Cleanup references to the early cpu maps for the non-SMP configuration
> >>     and remove some functions called for SMP configurations only.
> >>     
> >> ...
> >>
> >> index 1179aa0..dc79409 100644
> >> --- a/arch/x86/kernel/setup.c
> >> +++ b/arch/x86/kernel/setup.c
> >> @@ -10,7 +10,7 @@
> >>  #include <asm/setup.h>
> >>  #include <asm/topology.h>
> >>  
> >> -#ifdef CONFIG_HAVE_SETUP_PER_CPU_AREA
> >> +#if defined(CONFIG_HAVE_SETUP_PER_CPU_AREA) && defined(CONFIG_SMP)
> > 
> > This implies that we can have
> > 
> > 	CONFIG_HAVE_SETUP_PER_CPU_AREA && !CONFIG_SMP
> > 
> > which a) appears to be true and b) seems rather dumb?
> 
> Yeah, I kind of picked that up in a code merge and didn't really look at it
> too closely.  I should have checked the Kconfig file to make sure that it
> couldn't inadvertently get set if SMP is not set.

config HAVE_SETUP_PER_CPU_AREA
	def_bool X86_64 || (X86_SMP && !X86_VOYAGER)

it _can_ get set at present on x86_64, which I assume is incorrect.