diff mbox series

[RFC,2/8] fs/mpage: avoid negative shift for large blocksize

Message ID 20241113094727.1497722-3-mcgrof@kernel.org (mailing list archive)
State Not Applicable, archived
Headers show
Series enable bs > ps for block devices | expand

Commit Message

Luis R. Rodriguez Nov. 13, 2024, 9:47 a.m. UTC 
From: Hannes Reinecke <hare@kernel.org>

For large blocksizes the number of block bits is larger than PAGE_SHIFT,
so use shift to calculate the sector number from the page cache index.

With this in place we can now enable large folios on with buffer-heads.

Signed-off-by: Hannes Reinecke <hare@kernel.org>
---
 fs/mpage.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Matthew Wilcox (Oracle) Nov. 13, 2024, 2:06 p.m. UTC | #1 
On Wed, Nov 13, 2024 at 01:47:21AM -0800, Luis Chamberlain wrote:
> +++ b/fs/mpage.c
> @@ -181,7 +181,7 @@ static struct bio *do_mpage_readpage(struct mpage_readpage_args *args)
>  	if (folio_buffers(folio))
>  		goto confused;
>  
> -	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
> +	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);

	block_in_file = folio_pos(folio) >> blkbits;
?

> @@ -527,7 +527,7 @@ static int __mpage_writepage(struct folio *folio, struct writeback_control *wbc,
>  	 * The page has no buffers: map it to disk
>  	 */
>  	BUG_ON(!folio_test_uptodate(folio));
> -	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
> +	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);

Likewise.
Hannes Reinecke Nov. 14, 2024, 1:47 p.m. UTC | #2 
On 11/13/24 15:06, Matthew Wilcox wrote:
> On Wed, Nov 13, 2024 at 01:47:21AM -0800, Luis Chamberlain wrote:
>> +++ b/fs/mpage.c
>> @@ -181,7 +181,7 @@ static struct bio *do_mpage_readpage(struct mpage_readpage_args *args)
>>   	if (folio_buffers(folio))
>>   		goto confused;
>>   
>> -	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
>> +	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);
> 
> 	block_in_file = folio_pos(folio) >> blkbits;
> ?
> 
>> @@ -527,7 +527,7 @@ static int __mpage_writepage(struct folio *folio, struct writeback_control *wbc,
>>   	 * The page has no buffers: map it to disk
>>   	 */
>>   	BUG_ON(!folio_test_uptodate(folio));
>> -	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
>> +	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);
> 
> Likewise.

Yeah. /me not being able to find proper macros ...

Cheers,

Hannes
diff mbox series

Patch

diff --git a/fs/mpage.c b/fs/mpage.c
index 5b5989c7c2a0..ff76600380ca 100644
--- a/fs/mpage.c
+++ b/fs/mpage.c
@@ -181,7 +181,7 @@  static struct bio *do_mpage_readpage(struct mpage_readpage_args *args)
 	if (folio_buffers(folio))
 		goto confused;
 
-	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
+	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);
 	last_block = block_in_file + args->nr_pages * blocks_per_folio;
 	last_block_in_file = (i_size_read(inode) + blocksize - 1) >> blkbits;
 	if (last_block > last_block_in_file)
@@ -527,7 +527,7 @@  static int __mpage_writepage(struct folio *folio, struct writeback_control *wbc,
 	 * The page has no buffers: map it to disk
 	 */
 	BUG_ON(!folio_test_uptodate(folio));
-	block_in_file = (sector_t)folio->index << (PAGE_SHIFT - blkbits);
+	block_in_file = (sector_t)(((loff_t)folio->index << PAGE_SHIFT) >> blkbits);
 	/*
 	 * Whole page beyond EOF? Skip allocating blocks to avoid leaking
 	 * space.