Re: [PATCH v2] x86: Clear feature bits disabled at compile-time

From: Maciej Wieczor-Retman
Date: Wed Jul 23 2025 - 12:18:25 EST


On 2025-07-23 at 07:08:43 -0700, H. Peter Anvin wrote:
>On 2025-07-23 02:22, Maciej Wieczor-Retman wrote:
>>
>> arch/x86/kernel/cpu/common.c | 12 ++++++++++++
>> arch/x86/tools/cpufeaturemasks.awk | 8 ++++++++
>> 2 files changed, 20 insertions(+)
>>
>> diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
>> index 77afca95cced..ba8b5fba8552 100644
>> --- a/arch/x86/kernel/cpu/common.c
>> +++ b/arch/x86/kernel/cpu/common.c
>> @@ -1709,6 +1709,16 @@ static void __init cpu_parse_early_param(void)
>> }
>> }
>>
>> +static __init void init_cpu_cap(struct cpuinfo_x86 *c)
>> +{
>> + int i;
>> +
>> + for (i = 0; i < NCAPINTS; i++) {
>> + cpu_caps_set[i] = REQUIRED_MASK(i);
>> + cpu_caps_cleared[i] = DISABLED_MASK(i);
>> + }
>> +}
>> +
>
>No... just use an static array initializer for cpu_caps_cleared[]. You
>don't even need to add any code at all.

Oh, right, that seems simpler

>
>Ironically enough, I actually had those macros generated in the original
>awk script version, but they weren't used.
>
>And you MUST NOT initialize cpu_caps_set[].

After thinking about it more I can see why doing that could be a problem, yes.

>What we *can* and probably should do is, at the very end of the process, verify
>that the final mask conforms to cpu_caps_set[] and print a nasty message and
>set the TAINT_CPU_OUT_OF_SPEC flag:
>
> add_taint(TAINT_CPU_OUT_OF_SPEC, LOCKDEP_STILL_OK)

Where would that end be though? End of identify_cpu()? apply_forced_caps() is
almost at the end of the function, and other functions don't seem to modify
x86_cpuinfo or cpu_caps_set[].

>... but that is a separate patch, and doesn't need to be backported.
>
>Xin send me a patch privately (attached) but it needs to have the
>REQUIRED_MASK_INIT_VALUES part removed and made into a proper patch.
>
>Xin didn't add an SoB on it, but you can use mine:
>
>Signed-off-by: H. Peter Anvin (Intel) <hpa@xxxxxxxxx>

Thanks, I'll work with that.

>
> -hpa



--
Kind regards
Maciej Wieczór-Retman