Re: Thoughts of AMX KVM support based on latest kernel

From: Nakajima, Jun
Date: Thu Nov 18 2021 - 18:17:49 EST

On Nov 17, 2021, at 4:53 AM, Paolo Bonzini <pbonzini@xxxxxxxxxx> wrote:
> On 11/17/21 11:15, Tian, Kevin wrote:
>> We are not sure whether such trick is worthwhile, since a sane
>> guest shouldn't set XFD[AMX]=1 before storing the AMX state. This
>> is why we want to seek SDM change to mark out that the software
>> should not assume XTILEDATA is still valid when XFD[AMX]=1.
> Okay, I just don't want it to be called out as virtualization specific.
> It doesn't have to happen in current processors, but it should be architecturally valid behavior to clear the processor's state as soon as a bit in XFD is set to 1.
> Paolo

We recommend that "system software initialize AMX state _before_ doing so" (below). Also, I think what the “creative” guest is doing is "lazy restore”, and "This approach will not operate correctly for a variety of reasons."


System software may disable use of Intel AMX by clearing XCR0[18:17], by clearing CR4.OSXSAVE, or by setting
IA32_XFD[18]. It is recommended that system software initialize AMX state (e.g., by executing TILERELEASE)
before doing so. This is because maintaining AMX state in a non-initialized state may have negative power and performance implications.

System software should not use XFD to implement a “lazy restore” approach to management of the XTILEDATA
state component. This approach will not operate correctly for a variety of reasons. One is that the LDTILECFG and TILERELEASE instructions initialize XTILEDATA and do not cause an #NM exception. Another is that an execution of XSAVE by a user thread will save XTILEDATA as initialized instead of the data expected by the user thread.