Re: [Xen-devel] [PATCH 2/2] x86/vmx: conditionally disable LBR support due to TSX format quirk

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 10.08.16 at 17:47, <msw@xxxxxxxx> wrote:
> On Wed, Aug 10, 2016 at 06:34:10AM -0600, Jan Beulich wrote:
>> >>> On 10.08.16 at 08:59, <msw@xxxxxxxx> wrote:
>> > --- a/xen/arch/x86/hvm/vmx/vmx.c
>> > +++ b/xen/arch/x86/hvm/vmx/vmx.c
>> > @@ -2576,8 +2576,22 @@ static const struct lbr_info 
>> > *last_branch_msr_get(void)
>> >          /* Haswell */
>> >          case 60: case 63: case 69: case 70:
>> >          /* Broadwell */
>> > -        case 61: case 71: case 79: case 86:
>> > +        case 61: case 71: case 79: case 86: {
>> > +            u64 caps;
>> > +            bool_t tsx_support = boot_cpu_has(X86_FEATURE_HLE) ||
>> > +                                 boot_cpu_has(X86_FEATURE_RTM);
>> > +
>> > +            rdmsrl(MSR_IA32_PERF_CAPABILITIES, caps);
>> 
>> This is guarded by a X86_FEATURE_PDCM check in Linux - why
>> would we not need the same here?
> 
> You're right, it should be. It also seems to be missing from
> core2_vpmu_init().

Feel free to take the liberty to fix it there at once (but please
briefly mention this as an independent change in the description).

>> Also I think this RDMSR should be performed once at boot, not every
>> time we come here.
> 
> I thought you might say that. It didn't seem obviously right to put
> this in boot_cpu_data -- is that what you suggest?

Why boot_cpu_data? Just an ordinary (possibly per-CPU) static
in vmx.c.

Jan


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