On Wed, Nov 06, 2019 at 04:58:23PM +1100, Paul Mackerras wrote: > On Mon, Nov 04, 2019 at 09:47:56AM +0530, Bharata B Rao wrote: > > - After the guest becomes secure, when we handle a page fault of a page > > belonging to SVM in HV, send that page to UV via UV_PAGE_IN. > > - Whenever a page is unmapped on the HV side, inform UV via UV_PAGE_INVAL. > > - Ensure all those routines that walk the secondary page tables of > > the guest don't do so in case of secure VM. For secure guest, the > > active secondary page tables are in secure memory and the secondary > > page tables in HV are freed when guest becomes secure. > > Why do we free the page tables? Just to save a little memory? It > feels like it would make things more fragile. I guess we could just leave the page tables around and they would get populated again if and when the guest is reset (i,e., when it goes back to non-secure mode) However it appeared cleaner to cleanup the page tables given that aren't in user any longer. > > Also, I don't see where the freeing gets done in this patch. There isn't a very good reason for freeing code to be not part of this patch. I just put that in reset patch (6/8) where there is code for reinitializing page tables again. Regards, Bharata.
