2008-01-23 08:49:59

by Yinghai Lu

Subject: [PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr

[PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr

Signed-off-by: Yinghai Lu <[email protected]>

diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c
index 2643a8f..b587166 100644
--- a/arch/x86/kernel/setup_64.c
+++ b/arch/x86/kernel/setup_64.c
@@ -349,7 +349,6 @@ void __init setup_arch(char **cmdline_p)
#ifdef CONFIG_NUMA
x86_cpu_to_node_map_early_ptr = (void *)&x86_cpu_to_node_map_init;
#endif
- x86_bios_cpu_apicid_early_ptr = (void *)&x86_bios_cpu_apicid_init;
#endif

#ifdef CONFIG_ACPI

2008-01-23 09:21:22

by Yinghai Lu

Subject: [PATCH] x86_64: should use array directly for early_ptr

[PATCH] x86_64: should use array directly for early_ptr

Signed-off-by: Yinghai Lu <[email protected]>

--- a/arch/x86/kernel/setup_64.c 2008-01-23 01:21:06.000000000 -0800
+++ b/arch/x86/kernel/setup_64.c 2008-01-23 01:21:24.000000000 -0800
@@ -344,10 +344,10 @@

#ifdef CONFIG_SMP
/* setup to use the early static init tables during kernel startup */
- x86_cpu_to_apicid_early_ptr = (void *)&x86_cpu_to_apicid_init;
- x86_bios_cpu_apicid_early_ptr = (void *)&x86_bios_cpu_apicid_init;
+ x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init;
+ x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init;
#ifdef CONFIG_NUMA
- x86_cpu_to_node_map_early_ptr = (void *)&x86_cpu_to_node_map_init;
+ x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init;
#endif
#endif

2008-01-23 14:17:00

by Mike Travis

Subject: Re: [PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr

Yinghai Lu wrote:
> [PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr
>
> Signed-off-by: Yinghai Lu <[email protected]>
>
> diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c
> index 2643a8f..b587166 100644
> --- a/arch/x86/kernel/setup_64.c
> +++ b/arch/x86/kernel/setup_64.c
> @@ -349,7 +349,6 @@ void __init setup_arch(char **cmdline_p)
> #ifdef CONFIG_NUMA
> x86_cpu_to_node_map_early_ptr = (void *)&x86_cpu_to_node_map_init;
> #endif
> - x86_bios_cpu_apicid_early_ptr = (void *)&x86_bios_cpu_apicid_init;
> #endif
>
> #ifdef CONFIG_ACPI


Thanks Yinghai!

I wonder how this got past my testing (I do build and test !SMP and !NUMA
versions.) I'll check into this.

Regards,
Mike

2008-01-23 14:18:55

by Mike Travis

Subject: Re: [PATCH] x86_64: should use array directly for early_ptr

Yinghai Lu wrote:
> [PATCH] x86_64: should use array directly for early_ptr
>
> Signed-off-by: Yinghai Lu <[email protected]>
>
> --- a/arch/x86/kernel/setup_64.c 2008-01-23 01:21:06.000000000 -0800
> +++ b/arch/x86/kernel/setup_64.c 2008-01-23 01:21:24.000000000 -0800
> @@ -344,10 +344,10 @@
>
> #ifdef CONFIG_SMP
> /* setup to use the early static init tables during kernel startup */
> - x86_cpu_to_apicid_early_ptr = (void *)&x86_cpu_to_apicid_init;
> - x86_bios_cpu_apicid_early_ptr = (void *)&x86_bios_cpu_apicid_init;
> + x86_cpu_to_apicid_early_ptr = (void *)x86_cpu_to_apicid_init;
> + x86_bios_cpu_apicid_early_ptr = (void *)x86_bios_cpu_apicid_init;
> #ifdef CONFIG_NUMA
> - x86_cpu_to_node_map_early_ptr = (void *)&x86_cpu_to_node_map_init;
> + x86_cpu_to_node_map_early_ptr = (void *)x86_cpu_to_node_map_init;
> #endif
> #endif
>


Also Thanks!

-Mike

2008-01-23 14:37:30

by Mike Travis

Subject: Re: [PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr

Mike Travis wrote:
> Yinghai Lu wrote:
>> [PATCH] x86_64: remove duplicated line about x86_bios_cpu_apicid_early_ptr
>>
>> Signed-off-by: Yinghai Lu <[email protected]>
>>
>> diff --git a/arch/x86/kernel/setup_64.c b/arch/x86/kernel/setup_64.c
>> index 2643a8f..b587166 100644
>> --- a/arch/x86/kernel/setup_64.c
>> +++ b/arch/x86/kernel/setup_64.c
>> @@ -349,7 +349,6 @@ void __init setup_arch(char **cmdline_p)
>> #ifdef CONFIG_NUMA
>> x86_cpu_to_node_map_early_ptr = (void *)&x86_cpu_to_node_map_init;
>> #endif
>> - x86_bios_cpu_apicid_early_ptr = (void *)&x86_bios_cpu_apicid_init;
>> #endif
>>
>> #ifdef CONFIG_ACPI
>
>
> Thanks Yinghai!
>
> I wonder how this got past my testing (I do build and test !SMP and !NUMA
> versions.) I'll check into this.

Hmm, never mind. Coffee has cleared up my confusion...

>
> Regards,
> Mike