Messages in this thread | | | From | Lai Jiangshan <> | Date | Thu, 26 May 2022 17:15:12 +0800 | Subject | Re: [PATCH V2 4/7] KVM: X86/MMU: Activate special shadow pages and remove old logic |
| |
On Tue, May 17, 2022 at 8:16 AM David Matlack <dmatlack@google.com> wrote:
> > This comment does not explain why FNAME(fetch) might pass these values > to allocate special shadow pages. How about adjust it to something like > this: > > /* > * Initialize the guest walker with default values. These values will > * be used in cases where KVM shadows a guest page table structure > * with more levels than what the guest. For example, KVM shadows > * 2-level non-PAE guests with 3-level PAE paging. > * > * Note, the gfn is technically ignored for these special shadow > * pages, but it's more consistent to always pass 0 to > * kvm_mmu_get_page(). > */
The comment is copied into V3 with slightly changed.
> > > + for (i = 2; i < PT_MAX_FULL_LEVELS; i++) { > > s/2/PT32_ROOT_LEVEL/ >
Did in v3
Thanks Lai
| |