f2fs: split journal cache from curseg cache
PATCH v2 2/7 f2fs: split journal cache from curseg cache
In curseg cache, f2fs caches two different parts:
- datas of current summay block, i.e. summary entries, footer info.
- journal info, i.e. sparse nat/sit entries or io stat info.
With this approach, 1) it may cause higher lock contention when we access or update both of the parts of cache since we use the same mutex lock curseg_mutex to protect the cache. 2) current summary block with last journal info will be writebacked into device as a normal summary block when flushing, however, we treat journal info as valid one only in current summary, so most normal summary blocks contain junk journal data, it wastes remaining space of summary block.
So, in order to fix above issues, we split curseg cache into two parts:
a) current summary block, protected by original mutex lock curseg_mutex
b) journal cache, protected by newly introduced r/w semaphore journal_rwsem
When loading curseg cache during ->mount, we store summary info and journal info into different caches; When doing checkpoint, we combine datas of two cache into current summary block for persisting.
Signed-off-by Chao Yu <[email protected]>
---
v2:
- description of commit log was not so detailed, update it.
fs/f2fs/checkpoint.c | 2 +- fs/f2fs/node.c | 26 ++++++++-------- fs/f2fs/segment.c | 85 +++++++++++++++++++++++++++++++++++++--------------- fs/f2fs/segment.h | 2 ++ 4 files changed, 77 insertions(+), 38 deletions(-)
fs/f2fs/checkpoint.c +1 -1
@@ -1051,7 +1051,7 @@ static int do_checkpoint(struct f2fs_sb_info *sbi, struct cp_control *cpc)if (sb->s_bdev->bd_part) kbytes_written +=BD_PART_WRITTEN (sbi); - seg_i->sum_blk->journal.info.kbytes_written =cpu_to_le64 (kbytes_written); + seg_i->journal->info.kbytes_written =cpu_to_le64 (kbytes_written);if (__remain_node_summaries (cpc->reason)) {write_node_summaries (sbi, start_blk);
fs/f2fs/node.c +13 -13
@@ -354,7 +354,7 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni) {struct f2fs_nm_info *nm_i =NM_I (sbi);struct curseg_info *curseg =CURSEG_I (sbi,CURSEG_HOT_DATA ); -struct f2fs_journal *journal = &curseg->sum_blk->journal; +struct f2fs_journal *journal = curseg->journal;nid_t start_nid =START_NID (nid);struct f2fs_nat_block *nat_blk;struct page *page =NULL ;
@@ -381,13 +381,13 @@ void get_node_info(struct f2fs_sb_info *sbi, nid_t nid, struct node_info *ni)down_write (&nm_i->nat_tree_lock);/* Check current segment summary */ -mutex_lock (&curseg->curseg_mutex); +down_read (&curseg->journal_rwsem); i =lookup_journal_in_cursum (journal,NAT_JOURNAL , nid,0 );if (i >=0 ) { ne =nat_in_journal (journal, i);node_info_from_raw_nat (ni, &ne); } -mutex_unlock (&curseg->curseg_mutex); +up_read (&curseg->journal_rwsem);if (i >=0 ) goto cache;
@@ -1613,7 +1613,7 @@ static void build_free_nids(struct f2fs_sb_info *sbi) {struct f2fs_nm_info *nm_i =NM_I (sbi);struct curseg_info *curseg =CURSEG_I (sbi,CURSEG_HOT_DATA ); -struct f2fs_journal *journal = &curseg->sum_blk->journal; +struct f2fs_journal *journal = curseg->journal;int i =0 ;nid_t nid = nm_i->next_scan_nid;
@@ -1645,7 +1645,7 @@ static void build_free_nids(struct f2fs_sb_info *sbi) nm_i->next_scan_nid = nid;/* find free nids from current sum_pages */ -mutex_lock (&curseg->curseg_mutex); +down_read (&curseg->journal_rwsem);for (i =0 ; i <nats_in_cursum (journal); i++) {block_t addr;
@@ -1656,7 +1656,7 @@ static void build_free_nids(struct f2fs_sb_info *sbi)else remove_free_nid (nm_i, nid); } -mutex_unlock (&curseg->curseg_mutex); +up_read (&curseg->journal_rwsem);up_read (&nm_i->nat_tree_lock); ra_meta_pages(sbi, NAT_BLOCK_OFFSET(nm_i->next_scan_nid),
@@ -1920,10 +1920,10 @@ static void remove_nats_in_journal(struct f2fs_sb_info *sbi) {struct f2fs_nm_info *nm_i =NM_I (sbi);struct curseg_info *curseg =CURSEG_I (sbi,CURSEG_HOT_DATA ); -struct f2fs_journal *journal = &curseg->sum_blk->journal; +struct f2fs_journal *journal = curseg->journal;int i; -mutex_lock (&curseg->curseg_mutex); +down_write (&curseg->journal_rwsem);for (i =0 ; i <nats_in_cursum (journal); i++) {struct nat_entry *ne;struct f2fs_nat_entry raw_ne;
@@ -1939,7 +1939,7 @@ static void remove_nats_in_journal(struct f2fs_sb_info *sbi)__set_nat_cache_dirty (nm_i, ne); }update_nats_in_cursum (journal, -i); -mutex_unlock (&curseg->curseg_mutex); +up_write (&curseg->journal_rwsem); } static void __adjust_nat_entry_set(struct nat_entry_set *nes,
@@ -1964,7 +1964,7 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, struct nat_entry_set *set) {struct curseg_info *curseg =CURSEG_I (sbi,CURSEG_HOT_DATA ); -struct f2fs_journal *journal = &curseg->sum_blk->journal; +struct f2fs_journal *journal = curseg->journal;nid_t start_nid = set->set *NAT_ENTRY_PER_BLOCK ;bool to_journal = true;struct f2fs_nat_block *nat_blk;
@@ -1980,7 +1980,7 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, to_journal = false;if (to_journal) { -mutex_lock (&curseg->curseg_mutex); +down_write (&curseg->journal_rwsem); }else { page =get_next_nat_page (sbi, start_nid); nat_blk =page_address (page);
@@ -2013,7 +2013,7 @@ static void __flush_nat_entry_set(struct f2fs_sb_info *sbi, }if (to_journal) -mutex_unlock (&curseg->curseg_mutex); +up_write (&curseg->journal_rwsem);else f2fs_put_page (page,1 );
@@ -2030,7 +2030,7 @@ void flush_nat_entries(struct f2fs_sb_info *sbi) {struct f2fs_nm_info *nm_i =NM_I (sbi);struct curseg_info *curseg =CURSEG_I (sbi,CURSEG_HOT_DATA ); -struct f2fs_journal *journal = &curseg->sum_blk->journal; +struct f2fs_journal *journal = curseg->journal;struct nat_entry_set *setvec[SETVEC_SIZE ];struct nat_entry_set *set, *tmp;unsigned int found;
fs/f2fs/segment.c +61 -24
197 lines not on this page · raw message
fs/f2fs/segment.h +2 -0
@@ -258,6 +258,8 @@ struct victim_selection {struct curseg_info {struct mutex curseg_mutex;/* lock for consistency */ struct f2fs_summary_block *sum_blk;/* cached summary block */ +struct rw_semaphore journal_rwsem;/* protect journal area */ +struct f2fs_journal *journal;/* cached journal info */ unsigned char alloc_type;/* current allocation type */ unsigned int segno;/* current segment number */ unsigned short next_blkoff;/* next block offset to write */
Signature · 1 line
2.7.0