btrfs: use precalculated sectorsize_bits from fs_info

We do a lot of calculations where we divide or multiply by sectorsize.
We also know and make sure that sectorsize is a power of two, so this
means all divisions can be turned to shifts and avoid eg. expensive
u64/u32 divisions.

The type is u32 as it's more register friendly on x86_64 compared to u8
and the resulting assembly is smaller (movzbl vs movl).

There's also superblock s_blocksize_bits but it's usually one more
pointer dereference farther than fs_info.

Reviewed-by: Johannes Thumshirn <johannes.thumshirn@wdc.com>
Signed-off-by: David Sterba <dsterba@suse.com>
This commit is contained in:
David Sterba
2020-07-01 20:45:04 +02:00
parent e940e9a7c7
commit ab108d992b
8 changed files with 17 additions and 13 deletions

View File

@@ -416,7 +416,7 @@ int convert_free_space_to_extents(struct btrfs_trans_handle *trans,
btrfs_mark_buffer_dirty(leaf);
btrfs_release_path(path);
nrbits = div_u64(block_group->length, block_group->fs_info->sectorsize);
nrbits = block_group->length >> block_group->fs_info->sectorsize_bits;
start_bit = find_next_bit_le(bitmap, nrbits, 0);
while (start_bit < nrbits) {
@@ -540,8 +540,8 @@ static void free_space_set_bits(struct btrfs_block_group *block_group,
end = found_end;
ptr = btrfs_item_ptr_offset(leaf, path->slots[0]);
first = div_u64(*start - found_start, fs_info->sectorsize);
last = div_u64(end - found_start, fs_info->sectorsize);
first = (*start - found_start) >> fs_info->sectorsize_bits;
last = (end - found_start) >> fs_info->sectorsize_bits;
if (bit)
extent_buffer_bitmap_set(leaf, ptr, first, last - first);
else