[PATCH] f2fs: don't allow any writes on aborted atomic writes

[Jaegeuk Kim]

In order to prevent abusing atomic writes by abnormal users, we've added a
threshold, 20% over memory footprint, which disallows further atomic writes.
Previously, however, SQLite doesn't know the files became normal, so that
it could write stale data and commit on revoked normal database file.

Once f2fs detects such the abnormal behavior, this patch tries to avoid further
writes in write_begin().

Signed-off-by: Jaegeuk Kim <jaegeuk@xxxxxxxxxx>
---
 fs/f2fs/data.c | 3 ++-
 fs/f2fs/file.c | 7 ++++++-
 2 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 22dd00c6e241..02ec2603725f 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2295,7 +2295,8 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 	trace_f2fs_write_begin(inode, pos, len, flags);
 
 	if (f2fs_is_atomic_file(inode) &&
-			!f2fs_available_free_memory(sbi, INMEM_PAGES)) {
+			(is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST) ||
+			!f2fs_available_free_memory(sbi, INMEM_PAGES))) {
 		err = -ENOMEM;
 		drop_atomic = true;
 		goto fail;
diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c
index ff2cb8fb6934..c2c47f3248c4 100644
--- a/fs/f2fs/file.c
+++ b/fs/f2fs/file.c
@@ -1708,8 +1708,11 @@ static int f2fs_ioc_start_atomic_write(struct file *filp)
 
 	down_write(&F2FS_I(inode)->i_gc_rwsem[WRITE]);
 
-	if (f2fs_is_atomic_file(inode))
+	if (f2fs_is_atomic_file(inode)) {
+		if (is_inode_flag_set(inode, FI_ATOMIC_REVOKE_REQUEST))
+			ret = -EINVAL;
 		goto out;
+	}
 
 	ret = f2fs_convert_inline_inode(inode);
 	if (ret)
@@ -1871,6 +1874,8 @@ static int f2fs_ioc_abort_volatile_write(struct file *filp)
 		ret = f2fs_do_sync_file(filp, 0, LLONG_MAX, 0, true);
 	}
 
+	clear_inode_flag(inode, FI_ATOMIC_REVOKE_REQUEST);
+
 	inode_unlock(inode);
 
 	mnt_drop_write_file(filp);
-- 
2.17.0.441.gb46fe60e1d-goog