On Tue, Nov 29, 2022 at 09:15:21PM +0000, Oliver Upton wrote: > Hi Ricardo, > > Thanks for having a look. > > On Tue, Nov 29, 2022 at 12:52:12PM -0800, Ricardo Koller wrote: > > On Tue, Nov 29, 2022 at 07:19:44PM +0000, Oliver Upton wrote: > > [...] > > > > + ret = stage2_update_leaf_attrs(pgt, addr, 1, KVM_PTE_LEAF_ATTR_LO_S2_AF, 0, > > > + &pte, NULL, 0); > > > + if (!ret) > > > + dsb(ishst); > > > > At the moment, the only reason for stage2_update_leaf_attrs() to not > > update the PTE is if it's not valid: > > > > if (!kvm_pte_valid(pte)) > > return 0; > > > > I guess you could check that as well: > > > > + if (!ret || kvm_pte_valid(pte)) > > + dsb(ishst); > > Thanks for catching this. > > Instead of pivoting on the returned PTE value, how about we return > -EAGAIN from the early return in stage2_attr_walker()? It would better > match the pattern used elsewhere in the pgtable code. Bugh... Returning EAGAIN has some unfortunate consequences that I've missed until now... The stage2 attr walker is used to handle faults as well as range-based operations. In the former case, EAGAIN is sane as we retry execution but the latter is not. I stupidly got hung up on write protection not working as intended for some time. I think that callers into the page table walker should indicate whether or not the walk is to address a fault. If it is not, __kvm_pgtable_visit() and __kvm_pgtable_walk() should chug along instead of bailing for EAGAIN. Let me mess around with this and figure out what is least ugly. -- Thanks, Oliver
