[PATCH 17/21] mm: Change return values of finish_mkwrite_fault()

[Jan Kara <jack@suse.cz>]

Currently finish_mkwrite_fault() returns 0 when PTE got changed before
we acquired PTE lock and VM_FAULT_WRITE when we succeeded in modifying
the PTE. This is somewhat confusing since 0 generally means success, it
is also inconsistent with finish_fault() which returns 0 on success.
Change finish_mkwrite_fault() to return 0 on success and VM_FAULT_NOPAGE
when PTE changed. Practically, there should be no behavioral difference
since we bail out from the fault the same way regardless whether we
return 0, VM_FAULT_NOPAGE, or VM_FAULT_WRITE. Also note that
VM_FAULT_WRITE has no effect for shared mappings since the only two
places that check it - KSM and GUP - care about private mappings only.
Generally the meaning of VM_FAULT_WRITE for shared mappings is not well
defined and we should probably clean that up.

Signed-off-by: Jan Kara <jack@suse.cz>
---
 mm/memory.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 1517ff91c743..b3bd6b6c6472 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2296,10 +2296,10 @@ int finish_mkwrite_fault(struct vm_fault *vmf)
 	 */
 	if (!pte_same(*vmf->pte, vmf->orig_pte)) {
 		pte_unmap_unlock(vmf->pte, vmf->ptl);
-		return 0;
+		return VM_FAULT_NOPAGE;
 	}
 	wp_page_reuse(vmf);
-	return VM_FAULT_WRITE;
+	return 0;
 }
 
 /*
@@ -2342,8 +2342,7 @@ static int wp_page_shared(struct vm_fault *vmf)
 			return tmp;
 		}
 		tmp = finish_mkwrite_fault(vmf);
-		if (unlikely(!tmp || (tmp &
-				      (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))) {
+		if (unlikely(tmp & (VM_FAULT_ERROR | VM_FAULT_NOPAGE))) {
 			unlock_page(vmf->page);
 			put_page(vmf->page);
 			return tmp;
-- 
2.6.6

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>