On 03/12/2019 10:08, Jan Beulich wrote:
> On 29.11.2019 21:01, Igor Druzhinin wrote:
>> --- a/xen/arch/x86/cpu/common.c
>> +++ b/xen/arch/x86/cpu/common.c
>> @@ -54,7 +54,7 @@ static unsigned int forced_caps[NCAPINTS];
>>
>> DEFINE_PER_CPU(bool, full_gdt_loaded);
>>
>> -void __init setup_clear_cpu_cap(unsigned int cap)
>> +void setup_clear_cpu_cap(unsigned int cap)
>> {
>> const uint32_t *dfs;
>> unsigned int i;
>> @@ -83,7 +83,7 @@ void __init setup_clear_cpu_cap(unsigned int cap)
>> }
>> }
>>
>> -void __init setup_force_cpu_cap(unsigned int cap)
>> +void setup_force_cpu_cap(unsigned int cap)
>> {
>> if (__test_and_set_bit(cap, forced_caps))
>> return;
>
> The two functions are deliberately __init, as any call to them
> post-init is not going to take system-wide effect. These functions
> should really be __init_presmp, if we had something like this. No
> use of them on an AP boot path is going to affect the BSP, and
> hence will leave the system in an inconsistent state.
>
I agree with you and have a version where I just gate the corresponding
calls with (c == &boot_cpu_data). Removing __init was the approach
suggested by Andrew following the concern of potentially asymmetric
microcode in a system which I don't think would work anyway due to
the reasons you mentioned.
I will send the original approach as v2.
Igor
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xenproject.org
https://lists.xenproject.org/mailman/listinfo/xen-devel
