On 24.03.23 17:56, Borislav Petkov wrote:
> On Mon, Mar 06, 2023 at 05:34:19PM +0100, Juergen Gross wrote:
>> Modern CPUs all share the same MTRR interface implemented via
>> generic_mtrr_ops.
>>
>> At several places in MTRR code this generic interface is deduced via
>> is_cpu(INTEL) tests, which is only working due to X86_VENDOR_INTEL
>> being 0 (the is_cpu() macro is testing mtrr_if->vendor, which isn't
>> explicitly set in generic_mtrr_ops).
>>
>> Fix that by replacing the is_cpu(INTEL) tests with testing for mtrr_if
>> to be &generic_mtrr_ops.
> 
> Two things:
> 
> * is_cpu() checks also whether mtrr_if is set. And we don't set it for
> all vendors. I wanted to replace that thing with a vendor check recently
> but there's that little issue.

The is_cpu() checks are either in functions reachable only with mtrr_if being
set, or are testing for INTEL, which is replaced by the test of mtrr_if being
&generic_mtrr_ops as written in the commit message.

> I guess for the cases where we have the generic MTRR implementation, we
> can safely assume that mtrr_if is set. Which leads me to the second
> thing:
> 
> * If you're going to test for &generic_mtrr_ops, then you can just as
> well do
> 
> 	cpu_feature_enabled(X86_FEATURE_MTRR)
> 
> which is a lot more telling.

Yes, I think this is true.

> 
>> diff --git a/arch/x86/kernel/cpu/mtrr/mtrr.c b/arch/x86/kernel/cpu/mtrr/mtrr.c
>> index 5fe62ee0361b..0c83990501f5 100644
>> --- a/arch/x86/kernel/cpu/mtrr/mtrr.c
>> +++ b/arch/x86/kernel/cpu/mtrr/mtrr.c
>> @@ -108,14 +108,12 @@ static int have_wrcomb(void)
>>   /*  This function returns the number of variable MTRRs  */
>>   static void __init set_num_var_ranges(bool use_generic)
>>   {
>> -	unsigned long config = 0, dummy;
>> +	unsigned long config, dummy;
>>   
>>   	if (use_generic)
>>   		rdmsr(MSR_MTRRcap, config, dummy);
>> -	else if (is_cpu(AMD) || is_cpu(HYGON))
>> -		config = 2;
>> -	else if (is_cpu(CYRIX) || is_cpu(CENTAUR))
>> -		config = 8;
>> +	else
>> +		config = mtrr_if->var_regs;
>>   
>>   	num_var_ranges = config & 0xff;
>>   }
> 
> Since you're touching this function, you might simply expand its body in
> its only call site in mtrr_bp_init(), put a comment above the expanded
> code and remove that function.

Okay.


Juergen