On Thu, 2022-12-15 at 14:46 -0800, Isaku Yamahata wrote: > > Btw, as you mentioned in the changelog, private memory always has WB memory > > type, thus cannot be virtualized. Is it better to modify update_mtrr() to > > just > > return early if the gfn range is purely private? > > MTRR support in cpuid is fixed to 1, PAT in cpuid is native. > MTRR and PAT are supported on shared pages. > But none of those are mentioned in the changelog or whatever. They are hidden in the spec or in the later patches far away. ALso, the handling of "load/save IA32_PAT" in VMEXIT/VMENTRY VMCS control is different between TDX module and KVM. None of those are mentioned. All those make the patch review so hard. > > > IMHO the handling of MTRR/PAT virtualization for TDX guest deserves > > dedicated > > patch(es) to put them together so it's easier to review. Now the relevant > > parts > > spread in multiple independent patches (MSR handling, vt_get_mt_mask(), > > etc). > > Ok, let me check it. Yes. IMHO you should put all relevant parts together and make a clear changelog to justify the patch(es).
