Btrfs: __btrfs_std_error() logic should be consistent w/out CONFIG_PRINTK defined
authorAnand Jain <anand.jain@oracle.com>
Fri, 14 Aug 2015 10:32:52 +0000 (18:32 +0800)
committerDavid Sterba <dsterba@suse.com>
Tue, 29 Sep 2015 14:30:00 +0000 (16:30 +0200)
commit57d816a15ba2c2690c57635134bc01cf4da4623c
tree9457a7ff7bb5bf280abdbbc8d85c971d96ba7c84
parent92fc03fbdcbe2523be3f7e6b8e95fee9563a10d2
Btrfs: __btrfs_std_error() logic should be consistent w/out CONFIG_PRINTK defined

error handling logic behaves differently with or without
CONFIG_PRINTK defined, since there are two copies of the same
function which a bit of different logic

One, when CONFIG_PRINTK is defined, code is

__btrfs_std_error(..)
{
::
       save_error_info(fs_info);
       if (sb->s_flags & MS_BORN)
               btrfs_handle_error(fs_info);
}

and two when CONFIG_PRINTK is not defined, the code is

__btrfs_std_error(..)
{
::
       if (sb->s_flags & MS_BORN) {
               save_error_info(fs_info);
               btrfs_handle_error(fs_info);
        }
}

I doubt if this was intentional ? and appear to have caused since
we maintain two copies of the same function and they got diverged
with commits.

Now to decide which logic is correct reviewed changes as below,

 533574c6bc30cf526cc1c41bde050c854a945efb
Commit added two copies of this function

 cf79ffb5b79e8a2b587fbf218809e691bb396c98
Commit made change to only one copy of the function and to the
copy when CONFIG_PRINTK is defined.

To fix this, instead of maintaining two copies of same function
approach, maintain single function, and just put the extra
portion of the code under CONFIG_PRINTK define.

This patch just does that. And keeps code of with CONFIG_PRINTK
defined.

Signed-off-by: Anand Jain <anand.jain@oracle.com>
Signed-off-by: David Sterba <dsterba@suse.com>
fs/btrfs/super.c