check ATTR_SIZE contraints in inode_change_ok

Make sure we check the truncate constraints early on in ->setattr by adding
those checks to inode_change_ok.  Also clean up and document inode_change_ok
to make this obvious.

As a fallout we don't have to call inode_newsize_ok from simple_setsize and
simplify it down to a truncate_setsize which doesn't return an error.  This
simplifies a lot of setattr implementations and means we use truncate_setsize
almost everywhere.  Get rid of fat_setsize now that it's trivial and mark
ext2_setsize static to make the calling convention obvious.

Keep the inode_newsize_ok in vmtruncate for now as all callers need an
audit for its removal anyway.

Note: setattr code in ecryptfs doesn't call inode_change_ok at all and
needs a deeper audit, but that is left for later.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>

2 files changed, 9 insertions, 16 deletions

diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 12f445cee9f7..03ae894c45de 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -967,14 +967,15 @@ static int do_writepage(struct page *page, int len)
  * the page locked, and it locks @ui_mutex. However, write-back does take inode
  * @i_mutex, which means other VFS operations may be run on this inode at the
  * same time. And the problematic one is truncation to smaller size, from where
- * we have to call 'simple_setsize()', which first changes @inode->i_size, then
+ * we have to call 'truncate_setsize()', which first changes @inode->i_size, then
  * drops the truncated pages. And while dropping the pages, it takes the page
- * lock. This means that 'do_truncation()' cannot call 'simple_setsize()' with
+ * lock. This means that 'do_truncation()' cannot call 'truncate_setsize()' with
  * @ui_mutex locked, because it would deadlock with 'ubifs_writepage()'. This
  * means that @inode->i_size is changed while @ui_mutex is unlocked.
  *
- * XXX: with the new truncate the above is not true anymore, the simple_setsize
- * calls can be replaced with the individual components.
+ * XXX(truncate): with the new truncate sequence this is not true anymore,
+ * and the calls to truncate_setsize can be move around freely.  They should
+ * be moved to the very end of the truncate sequence.
  *
  * But in 'ubifs_writepage()' we have to guarantee that we do not write beyond
  * inode size. How do we do this if @inode->i_size may became smaller while we
@@ -1128,9 +1129,7 @@ static int do_truncation(struct ubifs_info *c, struct inode *inode,
                 budgeted = 0;
         }
 
-        err = simple_setsize(inode, new_size);
-        if (err)
-                goto out_budg;
+        truncate_setsize(inode, new_size);
 
         if (offset) {
                 pgoff_t index = new_size >> PAGE_CACHE_SHIFT;
@@ -1217,16 +1216,14 @@ static int do_setattr(struct ubifs_info *c, struct inode *inode,
 
         if (attr->ia_valid & ATTR_SIZE) {
                 dbg_gen("size %lld -> %lld", inode->i_size, new_size);
-                err = simple_setsize(inode, new_size);
-                if (err)
-                        goto out;
+                truncate_setsize(inode, new_size);
         }
 
         mutex_lock(&ui->ui_mutex);
         if (attr->ia_valid & ATTR_SIZE) {
                 /* Truncation changes inode [mc]time */
                 inode->i_mtime = inode->i_ctime = ubifs_current_time(inode);
-                /* 'simple_setsize()' changed @i_size, update @ui_size */
+                /* 'truncate_setsize()' changed @i_size, update @ui_size */
                 ui->ui_size = inode->i_size;
         }
 
@@ -1248,10 +1245,6 @@ static int do_setattr(struct ubifs_info *c, struct inode *inode,
         if (IS_SYNC(inode))
                 err = inode->i_sb->s_op->write_inode(inode, NULL);
         return err;
-
-out:
-        ubifs_release_budget(c, &req);
-        return err;
 }
 
 int ubifs_setattr(struct dentry *dentry, struct iattr *attr)
diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h
index 04310878f449..0c9876b396dd 100644
--- a/fs/ubifs/ubifs.h
+++ b/fs/ubifs/ubifs.h
@@ -379,7 +379,7 @@ struct ubifs_gced_idx_leb {
  * The @ui_size is a "shadow" variable for @inode->i_size and UBIFS uses
  * @ui_size instead of @inode->i_size. The reason for this is that UBIFS cannot
  * make sure @inode->i_size is always changed under @ui_mutex, because it
- * cannot call 'simple_setsize()' with @ui_mutex locked, because it would deadlock
+ * cannot call 'truncate_setsize()' with @ui_mutex locked, because it would deadlock
  * with 'ubifs_writepage()' (see file.c). All the other inode fields are
  * changed under @ui_mutex, so they do not need "shadow" fields. Note, one
  * could consider to rework locking and base it on "shadow" fields.