btrfs: make free space cache size consistent across different PAGE_SIZE

Currently free space cache inode size is determined by two factors:

- block group size
- PAGE_SIZE

This means, for the same sized block groups, with different PAGE_SIZE,
it will result in different inode sizes.

This will not be a good thing for subpage support, so change the
requirement for PAGE_SIZE to sectorsize.

Now for the same 4K sectorsize btrfs, it should result the same inode
size no matter what the PAGE_SIZE is.

Signed-off-by: Qu Wenruo <wqu@suse.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit is contained in:
Qu Wenruo 2021-04-13 14:23:14 +08:00 committed by David Sterba
parent 8df507cbb5
commit 0044ae11e8
1 changed files with 9 additions and 9 deletions

View File

@ -2505,7 +2505,7 @@ static int cache_save_setup(struct btrfs_block_group *block_group,
struct extent_changeset *data_reserved = NULL; struct extent_changeset *data_reserved = NULL;
u64 alloc_hint = 0; u64 alloc_hint = 0;
int dcs = BTRFS_DC_ERROR; int dcs = BTRFS_DC_ERROR;
u64 num_pages = 0; u64 cache_size = 0;
int retries = 0; int retries = 0;
int ret = 0; int ret = 0;
@ -2617,20 +2617,20 @@ again:
* taking up quite a bit since it's not folded into the other space * taking up quite a bit since it's not folded into the other space
* cache. * cache.
*/ */
num_pages = div_u64(block_group->length, SZ_256M); cache_size = div_u64(block_group->length, SZ_256M);
if (!num_pages) if (!cache_size)
num_pages = 1; cache_size = 1;
num_pages *= 16; cache_size *= 16;
num_pages *= PAGE_SIZE; cache_size *= fs_info->sectorsize;
ret = btrfs_check_data_free_space(BTRFS_I(inode), &data_reserved, 0, ret = btrfs_check_data_free_space(BTRFS_I(inode), &data_reserved, 0,
num_pages); cache_size);
if (ret) if (ret)
goto out_put; goto out_put;
ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, num_pages, ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, cache_size,
num_pages, num_pages, cache_size, cache_size,
&alloc_hint); &alloc_hint);
/* /*
* Our cache requires contiguous chunks so that we don't modify a bunch * Our cache requires contiguous chunks so that we don't modify a bunch