NFSv4: Ensure correct locking when accessing the 'lock_states' list (4b44b40e) · Commits · 戴 / test

There are currently 2 places in the state recovery code, where we do not take sufficient precautions before accessing the state->lock_states. In both cases, we should be holding the state->state_lock. Reported-by:

Pascal Bouchareine <pascal@gandi.net> Signed-off-by:

Trond Myklebust <Trond.Myklebust@netapp.com>

fs/nfs/nfs4state.c

+4 −0

Original line number	Diff line number	Diff line
		@@ -1156,11 +1156,13 @@ restart:
		if (status >= 0) {
		status = nfs4_reclaim_locks(state, ops);
		if (status >= 0) {
		spin_lock(&state->state_lock);
		list_for_each_entry(lock, &state->lock_states, ls_locks) {
		if (!(lock->ls_flags & NFS_LOCK_INITIALIZED))
		printk("%s: Lock reclaim failed!\n",
		__func__);
		}
		spin_unlock(&state->state_lock);
		nfs4_put_open_state(state);
		goto restart;
		}
		@@ -1224,10 +1226,12 @@ static void nfs4_clear_open_state(struct nfs4_state *state)
		clear_bit(NFS_O_RDONLY_STATE, &state->flags);
		clear_bit(NFS_O_WRONLY_STATE, &state->flags);
		clear_bit(NFS_O_RDWR_STATE, &state->flags);
		spin_lock(&state->state_lock);
		list_for_each_entry(lock, &state->lock_states, ls_locks) {
		lock->ls_seqid.flags = 0;
		lock->ls_flags &= ~NFS_LOCK_INITIALIZED;
		}
		spin_unlock(&state->state_lock);
		}

		static void nfs4_reset_seqids(struct nfs_server *server,

Admin message