Btrfs: fix typo in fallocate to make it honor actual size
authorJosef Bacik <josef@redhat.com>
Mon, 22 Nov 2010 18:50:32 +0000 (18:50 +0000)
committerChris Mason <chris.mason@oracle.com>
Sat, 27 Nov 2010 17:59:16 +0000 (12:59 -0500)
There is a typo in __btrfs_prealloc_file_range() where we set the i_size to
actual_len/cur_offset, and then just set it to cur_offset again, and do the same
with btrfs_ordered_update_i_size().  This fixes it back to keeping i_size in a
local variable and then updating i_size properly.  Tested this with

xfs_io -F -f -c "falloc 0 1" -c "pwrite 0 1" foo

stat'ing foo gives us a size of 1 instead of 4096 like it was.  Thanks,

Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
fs/btrfs/inode.c

index 37cc1776a5d765f1132686694a114e528dce25dd..0058fb3c2561a46b5710a2dcad5354ad1cac5c49 100644 (file)
@@ -7002,6 +7002,7 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
        struct btrfs_root *root = BTRFS_I(inode)->root;
        struct btrfs_key ins;
        u64 cur_offset = start;
+       u64 i_size;
        int ret = 0;
        bool own_trans = true;
 
@@ -7043,11 +7044,11 @@ static int __btrfs_prealloc_file_range(struct inode *inode, int mode,
                    (actual_len > inode->i_size) &&
                    (cur_offset > inode->i_size)) {
                        if (cur_offset > actual_len)
-                               i_size_write(inode, actual_len);
+                               i_size = actual_len;
                        else
-                               i_size_write(inode, cur_offset);
-                       i_size_write(inode, cur_offset);
-                       btrfs_ordered_update_i_size(inode, cur_offset, NULL);
+                               i_size = cur_offset;
+                       i_size_write(inode, i_size);
+                       btrfs_ordered_update_i_size(inode, i_size, NULL);
                }
 
                ret = btrfs_update_inode(trans, root, inode);