On Wed, 2018-07-11 at 11:06 +0200, Peter Zijlstra wrote:

On Tue, Jul 10, 2018 at 04:06:25PM -0700, Dave Hansen wrote:

quoted

On 07/10/2018 03:26 PM, Yu-cheng Yu wrote:

quoted

+	if (is_shstk_mapping(vma->vm_flags))
+		entry = pte_mkdirty_shstk(entry);
+	else
+		entry = pte_mkdirty(entry);
+
+	entry = maybe_mkwrite(entry, vma);
 	if (ptep_set_access_flags(vma, vmf->address, vmf->pte,
entry, 1))
 		update_mmu_cache(vma, vmf->address, vmf->pte);
 	pte_unmap_unlock(vmf->pte, vmf->ptl);

@@ -2526,7 +2532,11 @@ static int wp_page_copy(struct vm_fault

*vmf)
 		}
 		flush_cache_page(vma, vmf->address,
pte_pfn(vmf->orig_pte));
 		entry = mk_pte(new_page, vma->vm_page_prot);
-		entry = maybe_mkwrite(pte_mkdirty(entry), vma);
+		if (is_shstk_mapping(vma->vm_flags))
+			entry = pte_mkdirty_shstk(entry);
+		else
+			entry = pte_mkdirty(entry);
+		entry = maybe_mkwrite(entry, vma);

Do we want to lift this hunk of code and put it elsewhere?  Maybe:

	entry = pte_set_vma_features(entry, vma);

and then:

pte_t pte_set_vma_features(pte_t entry, struct vm_area_struct)
{
		/*
		 * Shadow stack PTEs are always dirty and always
		 * writable.  They have a different encoding for
		 * this than normal PTEs, though.
		 */
		if (is_shstk_mapping(vma->vm_flags))
			entry = pte_mkdirty_shstk(entry);
		else
			entry = pte_mkdirty(entry);

		entry = maybe_mkwrite(entry, vma);

	return entry;
}

Yes, that wants a helper like that. Not sold on the name, but
whatever.

Is there any way we can hide all the shadow stack magic in arch
code?

We use is_shstk_mapping() only to determine PAGE_DIRTY_SW or
PAGE_DIRTY_HW should be set in a PTE.  One way to remove this shadow
stack code from generic code is changing pte_mkdirty(pte) to
pte_mkdirty(pte, vma), and in the arch code we handle shadow stack.
Is this acceptable?

Thanks,
Yu-cheng