Re: [PATCH v6 04/11] x86: define IA32_FEATUE_CONTROL.SGX_LC

[Jarkko Sakkinen]

On Tue, Nov 28, 2017 at 10:28:03AM -0800, Sean Christopherson wrote:
> On Tue, 2017-11-28 at 09:16 -0800, Sean Christopherson wrote:
> > On Sat, 2017-11-25 at 21:29 +0200, Jarkko Sakkinen wrote:
> > > 
> > > When IA32_FEATURE_CONTROL.SGX_LC identifies that the root key for
> > > enclave signatures can be configured either by the OS or pre-boot
> > > firmware.
> > > 
> > > If this the case, IA32_SGXLEPUBKEYHASHn MSRs (0 < n < 4) can be used
> > > to
> > > set the SHA256 of the root key. IA32_FEATURE_CONTROL bit 17 controls
> > > whether the MSRs are writable by the OS. The pre-boot firmware can
> > > decided whether to  set this bit before setting
> > > IA32_FEATURE_CONTROL.LOCK.
> > The commit message (feature control bit) doesn't match the patch (CPUID
> > bit).
> 
> Also, assuming this message is destined for the commit that adds SGX_LC
> to feature control, I think it should first and foremost describe the
> hardware behavior.  The firmware vs OS interaction and use cases are
> valuable to document but IMO should come after the hardware description.
> 
> And though it's not documented in the SDM, I think it's worthwhile to
> describe the SGX activation sequence and its relationship with the SGX
> MSRs, e.g. the LE hash MSRs are writable prior to SGX activation.
> Without that information, it's unclear as to how the LE hash MSRs could
> be different than Intel's reset value.
> 
> So, maybe something like this?
> 
>     After SGX is activated[1] the IA32_SGXLEPUBKEYHASHn MSRs are writable
>     if and only if SGX_LC is set in the IA32_FEATURE_CONTROL MSR and the
>     IA32_FEATURE_CONTROL MSR is locked, otherwise they are read-only.
> 
>     For example, firmware can allow the OS to change the launch enclave
>     root key by setting IA32_FEATURE_CONTROL.SGX_LC, and thus give the
>     OS complete control over the enclaves it runs.  Alternatively,
>     firmware can clear IA32_FEATURE_CONTROL.SGX_LC to lock down the root
>     key and restrict the OS to running enclaves signed with the root key
>     or whitelisted/trusted by a launch enclave (which must be signed with
>     the root key).
> 
>     [1] SGX related bits in IA32_FEATURE_CONTROL cannot be set until SGX
>         is activated, e.g. by firmware.  SGX activation is triggered by
>         setting bit 0 in MSR 0x7a.  Until SGX is activated, the LE hash
>         MSRs are writable, e.g. to allow firmware to lock down the LE
>         root key with a non-Intel value.

Thanks I'll use this as a basis and move most of the crappy commit
message to the commit (with some editing) that defines the MSRs.

/Jarkko