Re: [Xen-devel] [PATCH] x86/vPMU: don't blindly assume IA32_PERF_CAPABILITIES MSR exists

[Jan Beulich <jbeulich@xxxxxxxx>]

On 26.02.2020 11:09, Roger Pau Monné wrote:
> On Wed, Feb 26, 2020 at 10:19:19AM +0100, Jan Beulich wrote:
>> Just like VMX'es lbr_tsx_fixup_check() the respective CPUID bit should
>> be consulted first.
>>
>> Reported-by: Farrah Chen <farrah.chen@xxxxxxxxx>
>> Signed-off-by: Jan Beulich <jbeulich@xxxxxxxx>
>>
>> --- a/xen/arch/x86/cpu/vpmu_intel.c
>> +++ b/xen/arch/x86/cpu/vpmu_intel.c
>> @@ -900,7 +900,6 @@ int vmx_vpmu_initialise(struct vcpu *v)
>>  
>>  int __init core2_vpmu_init(void)
>>  {
>> -    u64 caps;
>>      unsigned int version = 0;
>>      unsigned int i;
>>  
>> @@ -932,8 +931,14 @@ int __init core2_vpmu_init(void)
>>  
>>      arch_pmc_cnt = core2_get_arch_pmc_count();
>>      fixed_pmc_cnt = core2_get_fixed_pmc_count();
>> -    rdmsrl(MSR_IA32_PERF_CAPABILITIES, caps);
>> -    full_width_write = (caps >> 13) & 1;
>> +
>> +    if ( cpu_has_pdcm )
>> +    {
>> +        uint64_t caps;
>> +
>> +        rdmsrl(MSR_IA32_PERF_CAPABILITIES, caps);
>> +        full_width_write = (caps >> 13) & 1;
> 
> Will PMU work without PDCM?
> 
> I've been grepping the Intel SDMs, but the only mention is that PDCM
> signal the availability of MSR_IA32_PERF_CAPABILITIES.

Well, there's no other use of the MSR afaics except for getting
the one bit here, so I assume it'll work.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel