RE: [PATCH v7 3/4] fat: permit to return phy block number by fibmap in fallocated region

[Namjae Jeon <namjae.jeon@samsung.com>]

> 
> Hi,
> 
> Sorry for long delay.
> 
> > +static int fat_get_block_bmap(struct inode *inode, sector_t iblock,
> > +		struct buffer_head *bh_result, int create)
> > +{
> > +	struct super_block *sb = inode->i_sb;
> > +	unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits;
> > +	int err;
> > +	sector_t bmap, last_block;
> > +	unsigned long mapped_blocks;
> > +
> > +	BUG_ON(create != 0);
> > +
> > +	last_block = inode->i_blocks >> (sb->s_blocksize_bits - 9);
> > +
> > +	if (iblock >= last_block)
> > +		return 0;
> > +
> > +	err = fat_get_mapped_cluster(inode, iblock, last_block, &mapped_blocks,
> > +		&bmap);
> > +	if (err)
> > +		return err;
> 
Hi OGAWA.
> Probably, we are missing rootdir check here? If it is fat12/fat16,
> rootdir has to treat as special, because it doesn't have cluster chain.
> 
> So, how about to pass flags to fat_bmap()? For example,
> 
> 	/* Check EOF for excluding bmap() ioctl path */
> 	static int is_exceed_eof()
> 	{
> 		...
> 	}
> 
> 	fat_bmap(..., from_bmap)
> 	{
> 		...
> 
> 		if (!from_bmap && is_exceed_eof(inode, sector, create))
> 			return 0;
> 
> 		...
> 	}
> 
> or such.
Right, root dir check is missing, we need to put it here.
And I will change it as you suggest.

Thanks for your review:)
> 
> Thanks.
> --
> OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>