f2fs: use rw_sem instead of fs_lock(locks mutex)

The fs_locks is used to block other ops(ex, recovery) when doing checkpoint.
And each other operate routine(besides checkpoint) needs to acquire a fs_lock,
there is a terrible problem here, if these are too many concurrency threads acquiring
fs_lock, so that they will block each other and may lead to some performance problem,
but this is not the phenomenon we want to see.
Though there are some optimization patches introduced to enhance the usage of fs_lock,
but the thorough solution is using a *rw_sem* to replace the fs_lock.
Checkpoint routine takes write_sem, and other ops take read_sem, so that we can block
other ops(ex, recovery) when doing checkpoint, and other ops will not disturb each other,
this can avoid the problem described above completely.
Because of the weakness of rw_sem, the above change may introduce a potential problem
that the checkpoint thread might get starved if other threads are intensively locking
the read semaphore for I/O.(Pointed out by Xu Jin)
In order to avoid this, a wait_list is introduced, the appending read semaphore ops
will be dropped into the wait_list if checkpoint thread is waiting for write semaphore,
and will be waked up when checkpoint thread gives up write semaphore.
Thanks to Kim's previous review and test, and will be very glad to see other guys'
performance tests about this patch.

V2:
  -fix the potential starvation problem.
  -use more suitable func name suggested by Xu Jin.

Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
[Jaegeuk Kim: adjust minor coding standard]
Signed-off-by: Jaegeuk Kim <jaegeuk.kim@samsung.com>

1 files changed, 17 insertions, 20 deletions

diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index 02c906971cc6..c80faa2fff39 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -35,18 +35,18 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	block_t old_blk_addr;
 	struct dnode_of_data dn;
-	int err, ilock;
+	int err;
 
 	f2fs_balance_fs(sbi);
 
 	sb_start_pagefault(inode->i_sb);
 
 	/* block allocation */
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, page->index, ALLOC_NODE);
 	if (err) {
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		goto out;
 	}
 
@@ -56,12 +56,12 @@ static int f2fs_vm_page_mkwrite(struct vm_area_struct *vma,
 		err = reserve_new_block(&dn);
 		if (err) {
 			f2fs_put_dnode(&dn);
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 			goto out;
 		}
 	}
 	f2fs_put_dnode(&dn);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	file_update_time(vma->vm_file);
 	lock_page(page);
@@ -270,7 +270,7 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	unsigned int blocksize = inode->i_sb->s_blocksize;
 	struct dnode_of_data dn;
 	pgoff_t free_from;
-	int count = 0, ilock = -1;
+	int count = 0;
 	int err;
 
 	trace_f2fs_truncate_blocks_enter(inode, from);
@@ -278,13 +278,13 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	free_from = (pgoff_t)
 			((from + blocksize - 1) >> (sbi->log_blocksize));
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	set_new_dnode(&dn, inode, NULL, NULL, 0);
 	err = get_dnode_of_data(&dn, free_from, LOOKUP_NODE);
 	if (err) {
 		if (err == -ENOENT)
 			goto free_next;
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 		trace_f2fs_truncate_blocks_exit(inode, err);
 		return err;
 	}
@@ -305,7 +305,7 @@ static int truncate_blocks(struct inode *inode, u64 from)
 	f2fs_put_dnode(&dn);
 free_next:
 	err = truncate_inode_blocks(inode, free_from);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	/* lastly zero out the first data page */
 	truncate_partial_data_page(inode, from);
@@ -416,16 +416,15 @@ static void fill_zero(struct inode *inode, pgoff_t index,
 {
 	struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
 	struct page *page;
-	int ilock;
 
 	if (!len)
 		return;
 
 	f2fs_balance_fs(sbi);
 
-	ilock = mutex_lock_op(sbi);
+	f2fs_lock_op(sbi);
 	page = get_new_data_page(inode, NULL, index, false);
-	mutex_unlock_op(sbi, ilock);
+	f2fs_unlock_op(sbi);
 
 	if (!IS_ERR(page)) {
 		wait_on_page_writeback(page);
@@ -484,7 +483,6 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len, int mode)
 			struct address_space *mapping = inode->i_mapping;
 			loff_t blk_start, blk_end;
 			struct f2fs_sb_info *sbi = F2FS_SB(inode->i_sb);
-			int ilock;
 
 			f2fs_balance_fs(sbi);
 
@@ -493,9 +491,9 @@ static int punch_hole(struct inode *inode, loff_t offset, loff_t len, int mode)
 			truncate_inode_pages_range(mapping, blk_start,
 					blk_end - 1);
 
-			ilock = mutex_lock_op(sbi);
+			f2fs_lock_op(sbi);
 			ret = truncate_hole(inode, pg_start, pg_end);
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 		}
 	}
 
@@ -529,13 +527,12 @@ static int expand_inode_data(struct inode *inode, loff_t offset,
 
 	for (index = pg_start; index <= pg_end; index++) {
 		struct dnode_of_data dn;
-		int ilock;
 
-		ilock = mutex_lock_op(sbi);
+		f2fs_lock_op(sbi);
 		set_new_dnode(&dn, inode, NULL, NULL, 0);
 		ret = get_dnode_of_data(&dn, index, ALLOC_NODE);
 		if (ret) {
-			mutex_unlock_op(sbi, ilock);
+			f2fs_unlock_op(sbi);
 			break;
 		}
 
@@ -543,12 +540,12 @@ static int expand_inode_data(struct inode *inode, loff_t offset,
 			ret = reserve_new_block(&dn);
 			if (ret) {
 				f2fs_put_dnode(&dn);
-				mutex_unlock_op(sbi, ilock);
+				f2fs_unlock_op(sbi);
 				break;
 			}
 		}
 		f2fs_put_dnode(&dn);
-		mutex_unlock_op(sbi, ilock);
+		f2fs_unlock_op(sbi);
 
 		if (pg_start == pg_end)
 			new_size = offset + len;