Re: [PATCH 3/6] x86/hyper-v: reenlightenment notifications support

[Vitaly Kuznetsov]

"Michael Kelley (EOSG)" <Michael.H.Kelley@xxxxxxxxxxxxx> writes:

> On Fri, Dec 08, 2017 at 11:49:57AM +0100, Vitaly Kuznetsov wrote:
>
>> -----Original Message-----
>> From: Vitaly Kuznetsov [mailto:vkuznets@xxxxxxxxxx]
>> Sent: Friday, December 8, 2017 2:50 AM
>> To: kvm@xxxxxxxxxxxxxxx; x86@xxxxxxxxxx
>> Cc: Paolo Bonzini <pbonzini@xxxxxxxxxx>; Radim KrÄmÃÅ <rkrcmar@xxxxxxxxxx>; Thomas
>> Gleixner <tglx@xxxxxxxxxxxxx>; Ingo Molnar <mingo@xxxxxxxxxx>; H. Peter Anvin
>> <hpa@xxxxxxxxx>; KY Srinivasan <kys@xxxxxxxxxxxxx>; Haiyang Zhang
>> <haiyangz@xxxxxxxxxxxxx>; Stephen Hemminger <sthemmin@xxxxxxxxxxxxx>; Michael
>> Kelley (EOSG) <Michael.H.Kelley@xxxxxxxxxxxxx>; Andy Lutomirski <luto@xxxxxxxxxx>;
>> Mohammed Gamal <mmorsy@xxxxxxxxxx>; Cathy Avery <cavery@xxxxxxxxxx>; linux-
>> kernel@xxxxxxxxxxxxxxx; devel@xxxxxxxxxxxxxxxxxxxxxx
>> Subject: [PATCH 3/6] x86/hyper-v: reenlightenment notifications support
>> 
>> Hyper-V supports Live Migration notification. This is supposed to be used in conjunction with
>> TSC emulation: when we are migrated to a host with different TSC frequency for some short
>> period host emulates our accesses to TSC and sends us an interrupt to notify about the event.
>> When we're done updating everything we can disable TSC emulation and everything will start
>> working fast again.
>> 
>> We didn't need these notifications before as Hyper-V guests are not supposed to use TSC as a
>> clocksource: in Linux we even mark it as unstable on boot. Guests normally use 'tsc page'
>> clocksouce and host updates its values on migrations automatically.
>> 
>> Things change when we want to run nested virtualization: even when we pass through PV
>> clocksources (kvm-clock or tsc page) to our guests we need to know TSC frequency and when it
>> changes.
>> 
>> Hyper-V Top Level Functional Specification (as of v5.0b) wrongly specifies
>> EAX:BIT(12) of CPUID:0x40000009 as the feature identification bit. The right one to check is
>> EAX:BIT(13) of CPUID:0x40000003. I was assured that the fix in on the way.
>> 
>> Signed-off-by: Vitaly Kuznetsov <vkuznets@xxxxxxxxxx>
>
> [snip]
>
>> diff --git a/arch/x86/include/asm/irq_vectors.h
>> b/arch/x86/include/asm/irq_vectors.h
>> index 67421f649cfa..e71c1120426b 100644
>> --- a/arch/x86/include/asm/irq_vectors.h
>> +++ b/arch/x86/include/asm/irq_vectors.h
>> @@ -103,7 +103,12 @@
>>  #endif
>> 
>>  #define MANAGED_IRQ_SHUTDOWN_VECTOR	0xef
>> -#define LOCAL_TIMER_VECTOR		0xee
>> +
>> +#if IS_ENABLED(CONFIG_HYPERV)
>> +#define HYPERV_REENLIGHTENMENT_VECTOR	0xee
>> +#endif
>> +
>> +#define LOCAL_TIMER_VECTOR		0xed
>> 
>>  #define NR_VECTORS			 256
>
> [snip]
>
> Since you are pre-allocating a new vector, would you want to update the irq_cpustat_t
> data structure and your interrupt handler to count the occurrences of these interrupts,
> and update arch_show_interrupts() to show the count?  Then cat /proc/interrupts
> will show the count.   The reenlightenment interrupts will presumably be rare, but so
> are some of the others that are already counted and displayed, and it seems like
> consistency should be maintained.

I could do that.

The problem with adding this entry to /proc/interrupts, as I see it,
is that everyone who has CONFIG_HYPERV enabled in their distro will see
one additional line which 99,9% of the time will be just '0'. I,
however, see some value in seeing that L1 migration occured.

I'll add a separate patch to the series and let x86 maintainers decide
if it's worthy. Thanks,

-- 
  Vitaly