Kernel: Cache blocks_per_page in StorageDevice class

Instead of calculating blocks_per_page in every IO, cache it to save
CPU cycles as that value will not change after initialization.
This commit is contained in:
Pankaj Raghav 2022-01-29 11:51:21 +05:30 committed by Idan Horowitz
parent cf44d71edf
commit 3b27e28e67
2 changed files with 7 additions and 9 deletions

View file

@ -18,6 +18,7 @@ StorageDevice::StorageDevice(MajorNumber major, MinorNumber minor, size_t sector
: BlockDevice(major, minor, sector_size)
, m_early_storage_device_name(move(device_name))
, m_max_addressable_block(max_addressable_block)
, m_blocks_per_page(PAGE_SIZE / block_size())
{
}
@ -32,12 +33,10 @@ ErrorOr<size_t> StorageDevice::read(OpenFileDescription&, u64 offset, UserOrKern
size_t whole_blocks = len / block_size();
size_t remaining = len % block_size();
size_t blocks_per_page = PAGE_SIZE / block_size();
// PATAChannel will chuck a wobbly if we try to read more than PAGE_SIZE
// at a time, because it uses a single page for its DMA buffer.
if (whole_blocks >= blocks_per_page) {
whole_blocks = blocks_per_page;
if (whole_blocks >= m_blocks_per_page) {
whole_blocks = m_blocks_per_page;
remaining = 0;
}
@ -96,12 +95,10 @@ ErrorOr<size_t> StorageDevice::write(OpenFileDescription&, u64 offset, const Use
size_t whole_blocks = len / block_size();
size_t remaining = len % block_size();
size_t blocks_per_page = PAGE_SIZE / block_size();
// PATAChannel will chuck a wobbly if we try to write more than PAGE_SIZE
// at a time, because it uses a single page for its DMA buffer.
if (whole_blocks >= blocks_per_page) {
whole_blocks = blocks_per_page;
if (whole_blocks >= m_blocks_per_page) {
whole_blocks = m_blocks_per_page;
remaining = 0;
}

View file

@ -66,7 +66,8 @@ private:
// FIXME: Remove this method after figuring out another scheme for naming.
NonnullOwnPtr<KString> m_early_storage_device_name;
u64 m_max_addressable_block;
u64 m_max_addressable_block { 0 };
size_t m_blocks_per_page { 0 };
};
}