f2fs: fix to avoid data update racing between GC and DIO
Datas in file can be operated by GC and DIO simultaneously, so we will
face race case as below:
For write case:
Thread A Thread B
- generic_file_direct_write
- invalidate_inode_pages2_range
- f2fs_direct_IO
- do_blockdev_direct_IO
- do_direct_IO
- get_more_blocks
- f2fs_gc
- do_garbage_collect
- gc_data_segment
- move_data_page
- do_write_data_page
migrate data block to new block address
- dio_bio_submit
update user data to old block address
For read case:
Thread A Thread B
- generic_file_direct_write
- invalidate_inode_pages2_range
- f2fs_direct_IO
- do_blockdev_direct_IO
- do_direct_IO
- get_more_blocks
- f2fs_balance_fs
- f2fs_gc
- do_garbage_collect
- gc_data_segment
- move_data_page
- do_write_data_page
migrate data block to new block address
- write_checkpoint
- do_checkpoint
- clear_prefree_segments
- f2fs_issue_discard
discard old block adress
- dio_bio_submit
update user buffer from obsolete block address
In order to fix this, for one file, we should let DIO and GC getting exclusion
against with each other.
Signed-off-by: Chao Yu <yuchao0@huawei.com>
Signed-off-by: Jaegeuk Kim <jaegeuk@kernel.org>
Conflicts:
fs/f2fs/data.c
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 2b06036..d307a38 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -1756,8 +1756,10 @@
trace_f2fs_direct_IO_enter(inode, offset, count, rw);
+ down_read(&F2FS_I(inode)->dio_rwsem[rw]);
err = blockdev_direct_IO(rw, iocb, inode, iov, offset, nr_segs,
get_data_block_dio);
+ up_read(&F2FS_I(inode)->dio_rwsem[rw]);
if (err < 0 && (rw & WRITE)) {
if (err > 0)
set_inode_flag(inode, FI_UPDATE_WRITE);
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 6636beb..e6c7a143 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -506,6 +506,7 @@
struct list_head inmem_pages; /* inmemory pages managed by f2fs */
struct mutex inmem_lock; /* lock for inmemory pages */
struct extent_tree *extent_tree; /* cached extent_tree entry */
+ struct rw_semaphore dio_rwsem[2];/* avoid racing between dio and gc */
};
static inline void get_extent_info(struct extent_info *ext,
diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c
index 6f021a5..0d8daa6 100644
--- a/fs/f2fs/gc.c
+++ b/fs/f2fs/gc.c
@@ -755,12 +755,32 @@
/* phase 3 */
inode = find_gc_inode(gc_list, dni.ino);
if (inode) {
+ struct f2fs_inode_info *fi = F2FS_I(inode);
+ bool locked = false;
+
+ if (S_ISREG(inode->i_mode)) {
+ if (!down_write_trylock(&fi->dio_rwsem[READ]))
+ continue;
+ if (!down_write_trylock(
+ &fi->dio_rwsem[WRITE])) {
+ up_write(&fi->dio_rwsem[READ]);
+ continue;
+ }
+ locked = true;
+ }
+
start_bidx = start_bidx_of_node(nofs, inode)
+ ofs_in_node;
if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode))
move_encrypted_block(inode, start_bidx);
else
move_data_page(inode, start_bidx, gc_type);
+
+ if (locked) {
+ up_write(&fi->dio_rwsem[WRITE]);
+ up_write(&fi->dio_rwsem[READ]);
+ }
+
stat_inc_data_blk_count(sbi, 1, gc_type);
}
}
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 3ef3b3c..7b5d7aa 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -568,6 +568,8 @@
INIT_LIST_HEAD(&fi->gdirty_list);
INIT_LIST_HEAD(&fi->inmem_pages);
mutex_init(&fi->inmem_lock);
+ init_rwsem(&fi->dio_rwsem[READ]);
+ init_rwsem(&fi->dio_rwsem[WRITE]);
/* Will be used by directory only */
fi->i_dir_level = F2FS_SB(sb)->dir_level;