Re: [PATCH v1] exfat: mark straddling folio RO for 4K block size

[Chi Zhiling <chizhiling@163.com>]

On 10/9/26 10:46 AM, Yuezhang.Mo@sony.com wrote:
>>>    #include "exfat_raw.h"
>>>    #include "exfat_fs.h"
>>> @@ -91,6 +92,20 @@ static int exfat_cont_expand(struct inode *inode, loff_t size)
>>>         */
>>>        pagecache_isize_extended(inode, oldsize, inode->i_size);
>>>
>>> +     if (i_blocksize(inode) >= PAGE_SIZE && (oldsize & (PAGE_SIZE - 1))) {
>>> +             struct folio *folio;
>>> +
>>> +             folio = filemap_lock_folio(inode->i_mapping,
>>> +                             oldsize >> PAGE_SHIFT);
>>> +             if (!IS_ERR(folio)) {
>>> +                     if (folio_mkclean(folio))
>>> +                             folio_mark_dirty(folio);
>>> +
>>> +                     folio_unlock(folio);
>>> +                     folio_put(folio);
>>> +             }
>>> +     }
>>
>> I think fixing this issue specifically for the block size = 4K case may
>> not be thorough enough, because pagecache_isize_extended() may also fail
>> to work when the block size is smaller than 4K.
> 
> What condition might pagecache_isize_extended() fail to work when the block
> size is smaller than 4K?

I didn’t explain this clearly. What I meant is that the issue is not the 
block size, but the fact that VDL in exFAT is not block-aligned. 
pagecache_isize_extended is designed to work at block granularity, so it 
won’t work either if oldsize and newsize fall within the same block.

/* Page straddling @from will not have any hole block created? */
rounded_from = round_up(from, bsize);
if (to <= rounded_from || !(rounded_from & (PAGE_SIZE - 1)))
         return;

> 
>>
>> We should set the folio containing valid_size to RO whenever valid_size
>> != i_size and valid_size is not page-aligned, regardless of the block size.
> 
> We should set the folio containing oldsize to RO, rather than valid_size.
> This is because, when the folio containing oldsize was first set to RW, oldsize
> constrained valid_size, after the file size is subsequently expanded, that
> constraint changes, necessitating a re-setting of valid_size.

Yes, using oldsize is better.