Re: [syzbot] #syz test

[syzbot]

For archival purposes, forwarding an incoming command email to
linux-kernel@xxxxxxxxxxxxxxx, syzkaller-bugs@xxxxxxxxxxxxxxxx.

***

Subject: #syz test
Author: gshahrouzi@xxxxxxxxx

Try to remove device but fails
Reinitialize reference counting for write operations for device
Shut down the filesystem
Make the filesystem read only
	Read-write bit is not set so it skips trying to make the filesystem read-only, implying that the filesystem is already in read-only mode.
Encounter error because write operations are still assigned to a device
When the filesystem is starting up, it tries setting it to set the filesystem to read-only but it’s already set.
When failing to remove a device, it tries giving back write permissions to a device when the filesystem is read-only (not sure if this is nonsensical or not).
Since a device has write permissions, it tries to remove them but fails because the filesystem is in read-only mode.
So the fix here is to prevent it from giving the device write permissions if the filesystem is in read-only mode.

Signed-off-by: Gabriel Shahrouzi <gshahrouzi@xxxxxxxxx>
---
 fs/bcachefs/super.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/bcachefs/super.c b/fs/bcachefs/super.c
index b79e80a435e09..788e870bfef6a 100644
--- a/fs/bcachefs/super.c
+++ b/fs/bcachefs/super.c
@@ -1757,7 +1757,8 @@ int bch2_dev_remove(struct bch_fs *c, struct bch_dev *ca, int flags)
 	up_write(&c->state_lock);
 	return 0;
 err:
-	if (ca->mi.state == BCH_MEMBER_STATE_rw &&
+	if (test_bit(BCH_FS_rw, &c->flags) &&
+	    ca->mi.state == BCH_MEMBER_STATE_rw &&
 	    !percpu_ref_is_zero(&ca->io_ref[READ]))
 		__bch2_dev_read_write(c, ca);
 	up_write(&c->state_lock);
-- 
2.43.0