drm/i915: increase default latency constant (v2 w/comment)
[safe/jmp/linux-2.6] / fs / ocfs2 / namei.c
index 33464c6..8601f93 100644 (file)
@@ -118,7 +118,7 @@ static struct dentry *ocfs2_lookup(struct inode *dir, struct dentry *dentry,
        mlog(0, "find name %.*s in directory %llu\n", dentry->d_name.len,
             dentry->d_name.name, (unsigned long long)OCFS2_I(dir)->ip_blkno);
 
-       status = ocfs2_inode_lock(dir, NULL, 0);
+       status = ocfs2_inode_lock_nested(dir, NULL, 0, OI_LS_PARENT);
        if (status < 0) {
                if (status != -ENOENT)
                        mlog_errno(status);
@@ -636,7 +636,7 @@ static int ocfs2_link(struct dentry *old_dentry,
        if (S_ISDIR(inode->i_mode))
                return -EPERM;
 
-       err = ocfs2_inode_lock(dir, &parent_fe_bh, 1);
+       err = ocfs2_inode_lock_nested(dir, &parent_fe_bh, 1, OI_LS_PARENT);
        if (err < 0) {
                if (err != -ENOENT)
                        mlog_errno(err);
@@ -800,7 +800,8 @@ static int ocfs2_unlink(struct inode *dir,
                return -EPERM;
        }
 
-       status = ocfs2_inode_lock(dir, &parent_node_bh, 1);
+       status = ocfs2_inode_lock_nested(dir, &parent_node_bh, 1,
+                                        OI_LS_PARENT);
        if (status < 0) {
                if (status != -ENOENT)
                        mlog_errno(status);
@@ -978,7 +979,8 @@ static int ocfs2_double_lock(struct ocfs2_super *osb,
                        inode1 = tmpinode;
                }
                /* lock id2 */
-               status = ocfs2_inode_lock(inode2, bh2, 1);
+               status = ocfs2_inode_lock_nested(inode2, bh2, 1,
+                                                OI_LS_RENAME1);
                if (status < 0) {
                        if (status != -ENOENT)
                                mlog_errno(status);
@@ -987,7 +989,7 @@ static int ocfs2_double_lock(struct ocfs2_super *osb,
        }
 
        /* lock id1 */
-       status = ocfs2_inode_lock(inode1, bh1, 1);
+       status = ocfs2_inode_lock_nested(inode1, bh1, 1, OI_LS_RENAME2);
        if (status < 0) {
                /*
                 * An error return must mean that no cluster locks
@@ -1103,7 +1105,8 @@ static int ocfs2_rename(struct inode *old_dir,
         * won't have to concurrently downconvert the inode and the
         * dentry locks.
         */
-       status = ocfs2_inode_lock(old_inode, &old_inode_bh, 1);
+       status = ocfs2_inode_lock_nested(old_inode, &old_inode_bh, 1,
+                                        OI_LS_PARENT);
        if (status < 0) {
                if (status != -ENOENT)
                        mlog_errno(status);