Re: [Xen-devel] [PATCH v13 for-xen-4.5 17/21] x86/VPMU: Handle PMU interrupts for PV guests

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 13.10.14 at 21:10, <boris.ostrovsky@xxxxxxxxxx> wrote:
> On 10/13/2014 11:29 AM, Jan Beulich wrote:
>>>>> On 03.10.14 at 23:40, <boris.ostrovsky@xxxxxxxxxx> wrote:
>>>   int vpmu_do_wrmsr(unsigned int msr, uint64_t msr_content, uint64_t 
> supported)
>>>   {
>>> -    struct vpmu_struct *vpmu = vcpu_vpmu(current);
>>> +    struct vcpu *curr = current;
>>> +    struct vpmu_struct *vpmu = vcpu_vpmu(curr);
>>>   
>>>       if ( !(vpmu_mode & (XENPMU_MODE_SELF | XENPMU_MODE_HV)) )
>>>           return 0;
>>>   
>>>       if ( vpmu->arch_vpmu_ops && vpmu->arch_vpmu_ops->do_wrmsr )
>>> -        return vpmu->arch_vpmu_ops->do_wrmsr(msr, msr_content, supported);
>>> +    {
>>> +        int ret = vpmu->arch_vpmu_ops->do_wrmsr(msr, msr_content, 
>>> supported);
>>> +
>>> +        /*
>>> +         * We may have received a PMU interrupt during WRMSR handling
>>> +         * and since do_wrmsr may load VPMU context we should save
>>> +         * (and unload) it again.
>>> +         */
>> Is this btw true also when do_wrmsr() failed? Or could you not rather
>> handle ret != 0 before the following if() (or check ret == 0 together
>> with the other conditions)?
> 
> There is indeed no need to save/unload if do_wrmsr() failed but one of 
> conditions in the if() that you refer to (PMU_CACHED bit test) can only 
> happen if do_wrmsr() succeeded because the context could not have been 
> loaded in case of failure.

Ah, okay, that's fine then. I hope I'll remember when looking over
v14+.

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel