This was originally added during early ext4 development to pin open directories, but it was never really used. I don't mind to remove it.
Cheers, Andreas
> On Mar 3, 2019, at 18:31, Jason Yan <yanaijie@xxxxxxxxxx> wrote:
>
> Friendly ping...
>
> Anyone can help to review this patch?
>
>> On 2019/2/27 10:20, Jason Yan wrote:
>> This function is never used from the beginning. Let's remove it.
>> Signed-off-by: Jason Yan <yanaijie@xxxxxxxxxx>
>> ---
>> fs/ext4/inode.c | 30 ------------------------------
>> 1 file changed, 30 deletions(-)
>> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
>> index 34d7e0703cc6..0060d08edd92 100644
>> --- a/fs/ext4/inode.c
>> +++ b/fs/ext4/inode.c
>> @@ -6089,36 +6089,6 @@ void ext4_dirty_inode(struct inode *inode, int flags)
>> return;
>> }
>> -#if 0
>> -/*
>> - * Bind an inode's backing buffer_head into this transaction, to prevent
>> - * it from being flushed to disk early. Unlike
>> - * ext4_reserve_inode_write, this leaves behind no bh reference and
>> - * returns no iloc structure, so the caller needs to repeat the iloc
>> - * lookup to mark the inode dirty later.
>> - */
>> -static int ext4_pin_inode(handle_t *handle, struct inode *inode)
>> -{
>> - struct ext4_iloc iloc;
>> -
>> - int err = 0;
>> - if (handle) {
>> - err = ext4_get_inode_loc(inode, &iloc);
>> - if (!err) {
>> - BUFFER_TRACE(iloc.bh, "get_write_access");
>> - err = jbd2_journal_get_write_access(handle, iloc.bh);
>> - if (!err)
>> - err = ext4_handle_dirty_metadata(handle,
>> - NULL,
>> - iloc.bh);
>> - brelse(iloc.bh);
>> - }
>> - }
>> - ext4_std_error(inode->i_sb, err);
>> - return err;
>> -}
>> -#endif
>> -
>> int ext4_change_inode_journal_flag(struct inode *inode, int val)
>> {
>> journal_t *journal;
>
