Re: [PATCH 7/6] fat: Support a truncate() for expanding size

[OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>]

Andrew Morton <akpm@osdl.org> writes:

>> The generic_cont_expand() is too generic.
>
> But can it be fixed??

Oh, probably we can...

>> If "size" is block boundary, generic_cont_expand() expands the
>> ->i_size to "size + 1", after it, the caller of it will truncate to
>> "size" by vmtruncate().
>
> Something like this?
>

[...]

> -int generic_cont_expand(struct inode *inode, loff_t size)
> +static int __generic_cont_expand(struct inode *inode, loff_t size, int dont_do_that)
>  {
>  	struct address_space *mapping = inode->i_mapping;
>  	struct page *page;
> @@ -2182,9 +2182,8 @@ int generic_cont_expand(struct inode *in
>  	** skip the prepare.  make sure we never send an offset for the start
>  	** of a block
>  	*/
> -	if ((offset & (inode->i_sb->s_blocksize - 1)) == 0) {
> +	if (!dont_do_that &&  (offset & (inode->i_sb->s_blocksize - 1)) == 0)
>  		offset++;

Yes. But, if size is the page boundary, index is different.

Probably something like the below. And I'd like to do vmtruncate()
if ->prepare_write() returns a error. The sync_page_range_nolock()
can do by caller, so not necessary.

Hmm, I'll rethink this at tonight (10 hours later), the result may be
same after all though.

Thanks.


if (!dont_do_that) {
	offset = (size & (PAGE_CACHE_SIZE-1)); /* Within page */

	/* ugh.  in prepare/commit_write, if from==to==start of block, we 
	** skip the prepare.  make sure we never send an offset for the start
	** of a block
	*/
	if ((offset & (inode->i_sb->s_blocksize - 1)) == 0) {
		offset++;
	}
	index = size >> PAGE_CACHE_SHIFT;
} else {
	/* calculate the stuff of last page */
	loff_t pos = size - 1;
	index = pos >> PAGE_CACHE_SHIFT;
	offset = (pos & (PAGE_CACHE_SIZE - 1)) + 1;
}
-- 
OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>