On Thu, Apr 29, 2021 at 05:06:05PM +0100, Steven Price wrote: > On 27/04/2021 18:43, Catalin Marinas wrote: > > On Fri, Apr 16, 2021 at 04:43:04PM +0100, Steven Price wrote: > > > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h > > > index e17b96d0e4b5..cf4b52a33b3c 100644 > > > --- a/arch/arm64/include/asm/pgtable.h > > > +++ b/arch/arm64/include/asm/pgtable.h > > > @@ -312,7 +312,7 @@ static inline void set_pte_at(struct mm_struct *mm, unsigned long addr, > > > __sync_icache_dcache(pte); > > > if (system_supports_mte() && > > > - pte_present(pte) && pte_tagged(pte) && !pte_special(pte)) > > > + pte_present(pte) && (pte_val(pte) & PTE_USER) && !pte_special(pte)) > > > > I would add a pte_user() macro here or, if we restore the tags only when > > the page is readable, use pte_access_permitted(pte, false). Also add a > > comment why we do this. > > pte_access_permitted() looks like it describes what we want (user space can > access the memory). I'll add the following comment: > > /* > * If the PTE would provide user space will access to the tags I think drop "will". > * associated with it then ensure that the MTE tags are synchronised. > * Exec-only mappings don't expose tags (instruction fetches don't > * check tags). > */ Sounds fine. > > There's also the pte_user_exec() case which may not have the PTE_USER > > set (exec-only permission) but I don't think it matters. We don't do tag > > checking on instruction fetches, so if the user adds a PROT_READ to it, > > it would go through set_pte_at() again. I'm not sure KVM does anything > > special with exec-only mappings at stage 2, I suspect they won't be > > accessible by the guest (but needs checking). > > It comes down to the behaviour of get_user_pages(). AFAICT that will fail if > the memory is exec-only, so no stage 2 mapping will be created. Which of > course means the guest can't do anything with that memory. That certainly > seems like the only sane behaviour even without MTE. That's my understanding as well. The get_user_pages_fast() path uses pte_access_permitted() and should return false. The slower get_user_pages() relies on checking the vma flags and it checks for VM_READ. -- Catalin _______________________________________________ kvmarm mailing list kvmarm@xxxxxxxxxxxxxxxxxxxxx https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
