netfs: do not unlock and put the folio twice

check_write_begin() will unlock and put the folio when return
non-zero.  So we should avoid unlocking and putting it twice in
netfs layer.

Change the way ->check_write_begin() works in the following two ways:

 (1) Pass it a pointer to the folio pointer, allowing it to unlock and put
     the folio prior to doing the stuff it wants to do, provided it clears
     the folio pointer.

 (2) Change the return values such that 0 with folio pointer set means
     continue, 0 with folio pointer cleared means re-get and all error
     codes indicating an error (no special treatment for -EAGAIN).

[ bagasdotme: use Sphinx code text syntax for *foliop pointer ]

Cc: stable@vger.kernel.org
Link: https://tracker.ceph.com/issues/56423
Link: https://lore.kernel.org/r/cf169f43-8ee7-8697-25da-0204d1b4343e@redhat.com
Co-developed-by: David Howells <dhowells@redhat.com>
Signed-off-by: Xiubo Li <xiubli@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Bagas Sanjaya <bagasdotme@gmail.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>

3 files changed, 17 insertions, 13 deletions

diff --git a/fs/afs/file.c b/fs/afs/file.c
index 42118a4f3383..d1cfb235c4b9 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -375,7 +375,7 @@ static int afs_begin_cache_operation(struct netfs_io_request *rreq)
 }
 
 static int afs_check_write_begin(struct file *file, loff_t pos, unsigned len,
-				 struct folio *folio, void **_fsdata)
+				 struct folio **foliop, void **_fsdata)
 {
 	struct afs_vnode *vnode = AFS_FS_I(file_inode(file));
 
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 6dee88815491..d6e5916138e4 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -63,7 +63,7 @@
 	 (CONGESTION_ON_THRESH(congestion_kb) >> 2))
 
 static int ceph_netfs_check_write_begin(struct file *file, loff_t pos, unsigned int len,
-					struct folio *folio, void **_fsdata);
+					struct folio **foliop, void **_fsdata);
 
 static inline struct ceph_snap_context *page_snap_context(struct page *page)
 {
@@ -1288,18 +1288,19 @@ ceph_find_incompatible(struct page *page)
 }
 
 static int ceph_netfs_check_write_begin(struct file *file, loff_t pos, unsigned int len,
-					struct folio *folio, void **_fsdata)
+					struct folio **foliop, void **_fsdata)
 {
 	struct inode *inode = file_inode(file);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_snap_context *snapc;
 
-	snapc = ceph_find_incompatible(folio_page(folio, 0));
+	snapc = ceph_find_incompatible(folio_page(*foliop, 0));
 	if (snapc) {
 		int r;
 
-		folio_unlock(folio);
-		folio_put(folio);
+		folio_unlock(*foliop);
+		folio_put(*foliop);
+		*foliop = NULL;
 		if (IS_ERR(snapc))
 			return PTR_ERR(snapc);
 
diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index 42f892c5712e..0ce535852151 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -319,8 +319,9 @@ zero_out:
  * conflicting writes once the folio is grabbed and locked.  It is passed a
  * pointer to the fsdata cookie that gets returned to the VM to be passed to
  * write_end.  It is permitted to sleep.  It should return 0 if the request
- * should go ahead; unlock the folio and return -EAGAIN to cause the folio to
- * be regot; or return an error.
+ * should go ahead or it may return an error.  It may also unlock and put the
+ * folio, provided it sets ``*foliop`` to NULL, in which case a return of 0
+ * will cause the folio to be re-got and the process to be retried.
  *
  * The calling netfs must initialise a netfs context contiguous to the vfs
  * inode before calling this.
@@ -348,13 +349,13 @@ retry:
 
 	if (ctx->ops->check_write_begin) {
 		/* Allow the netfs (eg. ceph) to flush conflicts. */
-		ret = ctx->ops->check_write_begin(file, pos, len, folio, _fsdata);
+		ret = ctx->ops->check_write_begin(file, pos, len, &folio, _fsdata);
 		if (ret < 0) {
 			trace_netfs_failure(NULL, NULL, ret, netfs_fail_check_write_begin);
-			if (ret == -EAGAIN)
-				goto retry;
 			goto error;
 		}
+		if (!folio)
+			goto retry;
 	}
 
 	if (folio_test_uptodate(folio))
@@ -416,8 +417,10 @@ have_folio_no_wait:
 error_put:
 	netfs_put_request(rreq, false, netfs_rreq_trace_put_failed);
 error:
-	folio_unlock(folio);
-	folio_put(folio);
+	if (folio) {
+		folio_unlock(folio);
+		folio_put(folio);
+	}
 	_leave(" = %d", ret);
 	return ret;
 }