NFS: More fine grained attribute tracking

Currently, if the NFS_INO_INVALID_ATTR flag is set, for instance by
a call to nfs_post_op_update_inode_locked(), then it will not be cleared
until all the attributes have been revalidated. This means, for instance,
that NFSv4 writes will always force a full attribute revalidation.

Track the ctime, mtime, size and change attribute separately from the
other attributes so that we can have nfs_post_op_update_inode_locked()
set them correctly, and later have the cache consistency bitmask be
able to clear them.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>

fs/nfs/dir.c

@@ -1272,7 +1272,9 @@ static void nfs_drop_nlink(struct inode *inode)
 	/* drop the inode if we're reasonably sure this is the last link */
 	if (inode->i_nlink == 1)
 		clear_nlink(inode);
-	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR;
+	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+		| NFS_INO_INVALID_CTIME
+		| NFS_INO_INVALID_OTHER;
 	spin_unlock(&inode->i_lock);
 }
 

fs/nfs/nfs4proc.c

@@ -1045,7 +1045,9 @@ static void update_changeattr(struct inode *dir, struct nfs4_change_info *cinfo,
 	struct nfs_inode *nfsi = NFS_I(dir);
 
 	spin_lock(&dir->i_lock);
-	nfsi->cache_validity |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA;
+	nfsi->cache_validity |= NFS_INO_INVALID_CTIME
+		| NFS_INO_INVALID_MTIME
+		| NFS_INO_INVALID_DATA;
 	if (cinfo->atomic && cinfo->before == inode_peek_iversion_raw(dir)) {
 		nfsi->cache_validity &= ~NFS_INO_REVAL_PAGECACHE;
 		nfsi->attrtimeo_timestamp = jiffies;
@@ -5364,7 +5366,8 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl
 	 * so mark the attribute cache invalid.
 	 */
 	spin_lock(&inode->i_lock);
-	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_ATTR;
+	NFS_I(inode)->cache_validity |= NFS_INO_INVALID_CHANGE
+		| NFS_INO_INVALID_CTIME;
 	spin_unlock(&inode->i_lock);
 	nfs_access_zap_cache(inode);
 	nfs_zap_acl_cache(inode);

fs/nfs/write.c

@@ -1562,8 +1562,11 @@ static int nfs_writeback_done(struct rpc_task *task,
 	}
 
 	/* Deal with the suid/sgid bit corner case */
-	if (nfs_should_remove_suid(inode))
-		nfs_mark_for_revalidate(inode);
+	if (nfs_should_remove_suid(inode)) {
+		spin_lock(&inode->i_lock);
+		NFS_I(inode)->cache_validity |= NFS_INO_INVALID_OTHER;
+		spin_unlock(&inode->i_lock);
+	}
 	return 0;
 }
 

include/linux/nfs_fs.h

@@ -198,7 +198,6 @@ struct nfs_inode {
 /*
  * Cache validity bit flags
  */
-#define NFS_INO_INVALID_ATTR	BIT(0)		/* cached attrs are invalid */
 #define NFS_INO_INVALID_DATA	BIT(1)		/* cached data is invalid */
 #define NFS_INO_INVALID_ATIME	BIT(2)		/* cached atime is invalid */
 #define NFS_INO_INVALID_ACCESS	BIT(3)		/* cached access cred invalid */
@@ -206,6 +205,17 @@ struct nfs_inode {
 #define NFS_INO_REVAL_PAGECACHE	BIT(5)		/* must revalidate pagecache */
 #define NFS_INO_REVAL_FORCED	BIT(6)		/* force revalidation ignoring a delegation */
 #define NFS_INO_INVALID_LABEL	BIT(7)		/* cached label is invalid */
+#define NFS_INO_INVALID_CHANGE	BIT(8)		/* cached change is invalid */
+#define NFS_INO_INVALID_CTIME	BIT(9)		/* cached ctime is invalid */
+#define NFS_INO_INVALID_MTIME	BIT(10)		/* cached mtime is invalid */
+#define NFS_INO_INVALID_SIZE	BIT(11)		/* cached size is invalid */
+#define NFS_INO_INVALID_OTHER	BIT(12)		/* other attrs are invalid */
+
+#define NFS_INO_INVALID_ATTR	(NFS_INO_INVALID_CHANGE \
+		| NFS_INO_INVALID_CTIME \
+		| NFS_INO_INVALID_MTIME \
+		| NFS_INO_INVALID_SIZE \
+		| NFS_INO_INVALID_OTHER)	/* inode metadata is invalid */
 
 /*
  * Bit offsets in flags field
@@ -292,10 +302,11 @@ static inline void nfs_mark_for_revalidate(struct inode *inode)
 	struct nfs_inode *nfsi = NFS_I(inode);
 
 	spin_lock(&inode->i_lock);
-	nfsi->cache_validity |= NFS_INO_INVALID_ATTR |
-				NFS_INO_REVAL_PAGECACHE |
-				NFS_INO_INVALID_ACCESS |
-				NFS_INO_INVALID_ACL;
+	nfsi->cache_validity |= NFS_INO_REVAL_PAGECACHE
+		| NFS_INO_INVALID_ACCESS
+		| NFS_INO_INVALID_ACL
+		| NFS_INO_INVALID_CHANGE
+		| NFS_INO_INVALID_CTIME;
 	if (S_ISDIR(inode->i_mode))
 		nfsi->cache_validity |= NFS_INO_INVALID_DATA;
 	spin_unlock(&inode->i_lock);