Re: [edk2-devel] [PATCH] OvmfPkg: Set PciLib for TdxDxe driver

["Yao, Jiewen" <jiewen.yao@intel.com>]

Why does TdxDxe call TdxMailbox in an SEV platform?
Or why does TdxMailbox call SynchronizationLib in an SEV platform?

There are many places we can do CcProbe to stop action. Why we need do it in DSC?

In general, DSC lib override should be the last option, because it requires the platform integrator aware such tricky thing. I strongly discourage to use this, although it is architecturally support. 

I do believe we have many other ways to achieve the same goal.

Thank you
Yao Jiewen


> -----Original Message-----
> From: Xu, Min M <min.m.xu@intel.com>
> Sent: Tuesday, April 19, 2022 10:42 AM
> To: Yao, Jiewen <jiewen.yao@intel.com>; devel@edk2.groups.io
> Cc: Brijesh Singh <brijesh.singh@amd.com>; Aktas, Erdem
> <erdemaktas@google.com>; James Bottomley <jejb@linux.ibm.com>; Tom
> Lendacky <thomas.lendacky@amd.com>
> Subject: RE: [edk2-devel] [PATCH] OvmfPkg: Set PciLib for TdxDxe driver
> 
> On April 19, 2022 10:08 AM, Yao Jiewen wrote:
> >
> > If TdxDxe breaks SEV, should we skip the TdxDxe in SEV path?
> >
> > I don't understand why we need use Cfg8.
> >
> In TdxDxe driver we need to relocate APs and it requires the TdxMailboxLib.
> The lib chain is that TdxMailbox ->  SynchronizationLib -> TimerLib (which is
> DxeAcpiTimerLib) -> PciLib (Which was DxePciLibI440FxQ35).
> TdxDxe driver was loaded before AmdSevDxe so its lib constructors are called
> even it is in SEV-ES guest. In other words those lib constructor cannot be skipped
> in SEV-ES guest.
> 
> The analysis of the root cause (why DxePciLibI440FxQ35 trigger the broken) is in
> below link:
> https://edk2.groups.io/g/devel/message/88968
> 
> Thanks
> Min