[RFC,20/37] mm: implement and enable speculative fault handling in handle_pte_fault()

Commit Message

Michel Lespinasse April 7, 2021, 1:44 a.m. UTC

In handle_pte_fault(), allow speculative execution to proceed.

Use pte_spinlock() to validate the mmap sequence count when locking
the page table.

If speculative execution proceeds through do_wp_page(), ensure that we
end up in the wp_page_reuse() or wp_page_copy() paths, rather than
wp_pfn_shared() or wp_page_shared() (both unreachable as we only
handle anon vmas so far) or handle_userfault() (needs an explicit
abort to handle non-speculatively).

Signed-off-by: Michel Lespinasse <[e-mail]>
---
 mm/memory.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

Patch

diff --git a/mm/memory.c b/mm/memory.c
index 547d9d0ee962..fc555fae0844 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3103,6 +3103,7 @@  static vm_fault_t wp_pfn_shared(struct vm_fault *vmf)
 {
 	struct vm_area_struct *vma = vmf->vma;
 
+	VM_BUG_ON(vmf->flags & FAULT_FLAG_SPECULATIVE);
 	if (vma->vm_ops && vma->vm_ops->pfn_mkwrite) {
 		vm_fault_t ret;
 
@@ -3123,6 +3124,8 @@  static vm_fault_t wp_page_shared(struct vm_fault *vmf)
 	struct vm_area_struct *vma = vmf->vma;
 	vm_fault_t ret = VM_FAULT_WRITE;
 
+	VM_BUG_ON(vmf->flags & FAULT_FLAG_SPECULATIVE);
+
 	get_page(vmf->page);
 
 	if (vma->vm_ops && vma->vm_ops->page_mkwrite) {
@@ -3176,6 +3179,8 @@  static vm_fault_t do_wp_page(struct vm_fault *vmf)
 
 	if (userfaultfd_pte_wp(vma, *vmf->pte)) {
 		pte_unmap_unlock(vmf->pte, vmf->ptl);
+		if (vmf->flags & FAULT_FLAG_SPECULATIVE)
+			return VM_FAULT_RETRY;
 		return handle_userfault(vmf, VM_UFFD_WP);
 	}
 
@@ -4366,13 +4371,8 @@  static vm_fault_t handle_pte_fault(struct vm_fault *vmf)
 	if (pte_protnone(vmf->orig_pte) && vma_is_accessible(vmf->vma))
 		return do_numa_page(vmf);
 
-	if (vmf->flags & FAULT_FLAG_SPECULATIVE) {
-		pte_unmap(vmf->pte);
+	if (!pte_spinlock(vmf))
 		return VM_FAULT_RETRY;
-	}
-
-	vmf->ptl = pte_lockptr(vmf->vma->vm_mm, vmf->pmd);
-	spin_lock(vmf->ptl);
 	entry = vmf->orig_pte;
 	if (unlikely(!pte_same(*vmf->pte, entry))) {
 		update_mmu_tlb(vmf->vma, vmf->address, vmf->pte);