Skip to content

Commit

Permalink
NFSv4: Fix deadlock between nfs4_evict_inode() and nfs4_opendata_get_…
Browse files Browse the repository at this point in the history
…inode()

If the inode is being evicted, but has to return a delegation first,
then it can cause a deadlock in the corner case where the server reboots
before the delegreturn completes, but while the call to iget5_locked() in
nfs4_opendata_get_inode() is waiting for the inode free to complete.
Since the open call still holds a session slot, the reboot recovery
cannot proceed.

In order to break the logjam, we can turn the delegation return into a
privileged operation for the case where we're evicting the inode. We
know that in that case, there can be no other state recovery operation
that conflicts.

Reported-by: zhangxiaoxu (A) <zhangxiaoxu5@huawei.com>
Fixes: 5fcdfac ("NFSv4: Return delegations synchronously in evict_inode")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
  • Loading branch information
Trond Myklebust committed Jun 3, 2021
1 parent d1b5c23 commit dfe1fe7
Show file tree
Hide file tree
Showing 2 changed files with 12 additions and 1 deletion.
1 change: 1 addition & 0 deletions fs/nfs/nfs4_fs.h
Original file line number Diff line number Diff line change
Expand Up @@ -205,6 +205,7 @@ struct nfs4_exception {
struct inode *inode;
nfs4_stateid *stateid;
long timeout;
unsigned char task_is_privileged : 1;
unsigned char delay : 1,
recovering : 1,
retry : 1;
Expand Down
12 changes: 11 additions & 1 deletion fs/nfs/nfs4proc.c
Original file line number Diff line number Diff line change
Expand Up @@ -589,6 +589,8 @@ int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_
goto out_retry;
}
if (exception->recovering) {
if (exception->task_is_privileged)
return -EDEADLOCK;
ret = nfs4_wait_clnt_recover(clp);
if (test_bit(NFS_MIG_FAILED, &server->mig_status))
return -EIO;
Expand All @@ -614,6 +616,8 @@ nfs4_async_handle_exception(struct rpc_task *task, struct nfs_server *server,
goto out_retry;
}
if (exception->recovering) {
if (exception->task_is_privileged)
return -EDEADLOCK;
rpc_sleep_on(&clp->cl_rpcwaitq, task, NULL);
if (test_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) == 0)
rpc_wake_up_queued_task(&clp->cl_rpcwaitq, task);
Expand Down Expand Up @@ -6417,6 +6421,7 @@ static void nfs4_delegreturn_done(struct rpc_task *task, void *calldata)
struct nfs4_exception exception = {
.inode = data->inode,
.stateid = &data->stateid,
.task_is_privileged = data->args.seq_args.sa_privileged,
};

if (!nfs4_sequence_done(task, &data->res.seq_res))
Expand Down Expand Up @@ -6540,7 +6545,6 @@ static int _nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
data = kzalloc(sizeof(*data), GFP_NOFS);
if (data == NULL)
return -ENOMEM;
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1, 0);

nfs4_state_protect(server->nfs_client,
NFS_SP4_MACH_CRED_CLEANUP,
Expand Down Expand Up @@ -6571,6 +6575,12 @@ static int _nfs4_proc_delegreturn(struct inode *inode, const struct cred *cred,
}
}

if (!data->inode)
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1,
1);
else
nfs4_init_sequence(&data->args.seq_args, &data->res.seq_res, 1,
0);
task_setup_data.callback_data = data;
msg.rpc_argp = &data->args;
msg.rpc_resp = &data->res;
Expand Down

0 comments on commit dfe1fe7

Please sign in to comment.