[PATCH v5 03/22] mm: Dont assume page-table invariance during faults

[Laurent Dufour]

From: Peter Zijlstra <peterz@xxxxxxxxxxxxx>

One of the side effects of speculating on faults (without holding
mmap_sem) is that we can race with free_pgtables() and therefore we
cannot assume the page-tables will stick around.

Remove the reliance on the pte pointer.

Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>

[Remove only if !CONFIG_SPF]
Signed-off-by: Laurent Dufour <ldufour@xxxxxxxxxxxxxxxxxx>
---
 mm/memory.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/mm/memory.c b/mm/memory.c
index 6632c9b357c9..b7a9baf3df8a 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2287,6 +2287,7 @@ int apply_to_page_range(struct mm_struct *mm, unsigned long addr,
 }
 EXPORT_SYMBOL_GPL(apply_to_page_range);
 
+#ifndef CONFIG_SPF
 /*
  * handle_pte_fault chooses page fault handler according to an entry which was
  * read non-atomically.  Before making any commitment, on those architectures
@@ -2296,7 +2297,7 @@ EXPORT_SYMBOL_GPL(apply_to_page_range);
  * and do_anonymous_page can safely check later on).
  */
 static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd,
-				pte_t *page_table, pte_t orig_pte)
+				 pte_t *page_table, pte_t orig_pte)
 {
 	int same = 1;
 #if defined(CONFIG_SMP) || defined(CONFIG_PREEMPT)
@@ -2310,6 +2311,7 @@ static inline int pte_unmap_same(struct mm_struct *mm, pmd_t *pmd,
 	pte_unmap(page_table);
 	return same;
 }
+#endif /* CONFIG_SPF */
 
 static inline void cow_user_page(struct page *dst, struct page *src, unsigned long va, struct vm_area_struct *vma)
 {
@@ -2871,11 +2873,14 @@ int do_swap_page(struct vm_fault *vmf)
 
 	if (vma_readahead)
 		page = swap_readahead_detect(vmf, &swap_ra);
+
+#ifndef CONFIG_SPF
 	if (!pte_unmap_same(vma->vm_mm, vmf->pmd, vmf->pte, vmf->orig_pte)) {
 		if (page)
 			put_page(page);
 		goto out;
 	}
+#endif
 
 	entry = pte_to_swp_entry(vmf->orig_pte);
 	if (unlikely(non_swap_entry(entry))) {
-- 
2.7.4