| Message ID | bc8baf98c9c9357423178d4fab6b945bcb959f1d.1722839158.git.wqu@suse.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | btrfs: make btrfs_is_subpage() to return false directly for 4K page size | expand |
On Mon, Aug 05, 2024 at 03:56:30PM +0930, Qu Wenruo wrote: > Btrfs only supports sectorsize 4K, 8K, 16K, 32K, 64K for now, thus for > systems with 4K page size, there is no way the fs is subpage (sectorsize > < PAGE_SIZE). > > So here we define btrfs_is_subpage() different according to the > PAGE_SIZE: > > - PAGE_SIZE > 4K > We may hit real subpage cases, define btrfs_is_subpage() as a regular > function and do the usual checks. > > - PAGE_SIZE == 4K (no smaller PAGE_SIZE support AFAIK) > There is no way the fs is subpage, so just define btrfs_is_subpage() > as an inline function which always return false. > > This saves some bytes for x86_64 debug builds: > > text data bss dec hex filename > Before: 1484452 168693 25776 1678921 199e49 fs/btrfs/btrfs.ko > After: 1476605 168445 25776 1670826 197eaa fs/btrfs/btrfs.ko The delta is 7847, not bad and it's on x86_64 so it'll affect most machines. > Signed-off-by: Qu Wenruo <wqu@suse.com> Reviewed-by: David Sterba <dsterba@suse.com>
On Mon, Aug 05, 2024 at 06:22:20PM +0200, David Sterba wrote: > On Mon, Aug 05, 2024 at 03:56:30PM +0930, Qu Wenruo wrote: > > Btrfs only supports sectorsize 4K, 8K, 16K, 32K, 64K for now, thus for > > systems with 4K page size, there is no way the fs is subpage (sectorsize > > < PAGE_SIZE). > > > > So here we define btrfs_is_subpage() different according to the > > PAGE_SIZE: > > > > - PAGE_SIZE > 4K > > We may hit real subpage cases, define btrfs_is_subpage() as a regular > > function and do the usual checks. > > > > - PAGE_SIZE == 4K (no smaller PAGE_SIZE support AFAIK) > > There is no way the fs is subpage, so just define btrfs_is_subpage() > > as an inline function which always return false. > > > > This saves some bytes for x86_64 debug builds: > > > > text data bss dec hex filename > > Before: 1484452 168693 25776 1678921 199e49 fs/btrfs/btrfs.ko > > After: 1476605 168445 25776 1670826 197eaa fs/btrfs/btrfs.ko > > The delta is 7847, not bad and it's on x86_64 so it'll affect most > machines. On my release build config it's text data bss dec hex filename 1339533 30572 16088 1386193 1526d1 pre/btrfs.ko 1332328 30372 16088 1378788 1509e4 post/btrfs.ko DELTA: -7205
diff --git a/fs/btrfs/subpage.c b/fs/btrfs/subpage.c index 8ddd5fcbeb93..631d96f1e905 100644 --- a/fs/btrfs/subpage.c +++ b/fs/btrfs/subpage.c @@ -64,6 +64,7 @@ * This means a slightly higher tree locking latency. */ +#if PAGE_SIZE > SZ_4K bool btrfs_is_subpage(const struct btrfs_fs_info *fs_info, struct address_space *mapping) { if (fs_info->sectorsize >= PAGE_SIZE) @@ -85,6 +86,7 @@ bool btrfs_is_subpage(const struct btrfs_fs_info *fs_info, struct address_space return true; return false; } +#endif void btrfs_init_subpage_info(struct btrfs_subpage_info *subpage_info, u32 sectorsize) { diff --git a/fs/btrfs/subpage.h b/fs/btrfs/subpage.h index 249396e118d0..5532cc4fac50 100644 --- a/fs/btrfs/subpage.h +++ b/fs/btrfs/subpage.h @@ -5,6 +5,7 @@ #include <linux/spinlock.h> #include <linux/atomic.h> +#include <linux/sizes.h> struct address_space; struct folio; @@ -88,7 +89,15 @@ enum btrfs_subpage_type { BTRFS_SUBPAGE_DATA, }; +#if PAGE_SIZE > SZ_4K bool btrfs_is_subpage(const struct btrfs_fs_info *fs_info, struct address_space *mapping); +#else +static inline bool btrfs_is_subpage(const struct btrfs_fs_info *fs_info, + struct address_space *mapping) +{ + return false; +} +#endif void btrfs_init_subpage_info(struct btrfs_subpage_info *subpage_info, u32 sectorsize); int btrfs_attach_subpage(const struct btrfs_fs_info *fs_info,
Btrfs only supports sectorsize 4K, 8K, 16K, 32K, 64K for now, thus for systems with 4K page size, there is no way the fs is subpage (sectorsize < PAGE_SIZE). So here we define btrfs_is_subpage() different according to the PAGE_SIZE: - PAGE_SIZE > 4K We may hit real subpage cases, define btrfs_is_subpage() as a regular function and do the usual checks. - PAGE_SIZE == 4K (no smaller PAGE_SIZE support AFAIK) There is no way the fs is subpage, so just define btrfs_is_subpage() as an inline function which always return false. This saves some bytes for x86_64 debug builds: text data bss dec hex filename Before: 1484452 168693 25776 1678921 199e49 fs/btrfs/btrfs.ko After: 1476605 168445 25776 1670826 197eaa fs/btrfs/btrfs.ko Signed-off-by: Qu Wenruo <wqu@suse.com> --- fs/btrfs/subpage.c | 2 ++ fs/btrfs/subpage.h | 9 +++++++++ 2 files changed, 11 insertions(+)