- Nov 21, 2012
-
-
Trond Myklebust authored
We can't send a SEQUENCE op unless the session is OK, so it is pointless to handle the CHECK_LEASE state before we've dealt with SESSION_RESET and BIND_CONN_TO_SESSION. Signed-off-by:Trond Myklebust <Trond.Myklebust@netapp.com>
-
Bryan Schumaker authored
If I mount an NFS v4.1 server to a single client multiple times and then run xfstests over each mountpoint I usually get the client into a state where recovery deadlocks. The server informs the client of a cb_path_down sequence error, the client then does a bind_connection_to_session and checks the status of the lease. I found that bind_connection_to_session sets the NFS4_SESSION_DRAINING flag on the client, but this flag is never unset before nfs4_check_lease() reaches nfs4_proc_sequence(). This causes the client to deadlock, halting all NFS activity to the server. nfs4_proc_sequence() is only called by the state manager, so I can change it to run in privileged mode to bypass the NFS4_SESSION_DRAINING check and avoid the deadlock. Signed-off-by:
Bryan Schumaker <bjschuma@netapp.com> Signed-off-by:
-
- Nov 05, 2012
-
-
Trond Myklebust authored
We can and should use the rpc_create_args and __rpc_clone_client() to change the program and version number on the resulting rpc_client. Signed-off-by: -
Weston Andros Adamson authored
Use WARN_ON_ONCE instead of calling BUG_ON and return -EINVAL when RPC_TASK_ASYNC flag is passed to rpc_call_sync. Signed-off-by:
Weston Andros Adamson <dros@netapp.com> Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE(). Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE(). Signed-off-by:
-
Weston Andros Adamson authored
Replace two BUG_ON() calls with WARN_ON_ONCE() and early returns. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE(). Signed-off-by:
-
Weston Andros Adamson authored
Instead of calling BUG_ON(), do a WARN_ON_ONCE() and return -EINVAL. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE() and truncate the encoded string if len > max. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE() and NULL return - the caller will handle this like a memory allocation failure. Signed-off-by:
-
Weston Andros Adamson authored
Replace bounds checking BUG_ON() with a WARN_ON_ONCE() and resetting the requested len to the max. Signed-off-by:
-
Weston Andros Adamson authored
Replace two bounds checking BUG_ON() calls with WARN_ON_ONCE() and resetting the requested size to RPCSVC_MAXPAGES. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with a WARN_ON_ONCE() and early return. Signed-off-by:
-
Weston Andros Adamson authored
Replace multiple BUG_ON() calls with WARN_ON_ONCE() and early return when sanity checking socket ownership (lock). The bind call will fail if the socket was unsuccessfully reclassified. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with a WARN() and early return. Signed-off-by:
-
Weston Andros Adamson authored
Replace two BUG_ON() calls checking the RPC_BC_PA_IN_USE flag with WARN_ON_ONCE(). Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON() - the condition is definitely a misuse of the API, but shouldn't cause a crash. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE() in two parts of cache_read(). Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE(). The error condition is a simple ref counting sanity check and the following code will not free anything until final put. Signed-off-by:
-
Weston Andros Adamson authored
If max_reqs is 0, do nothing besides the usual dprintks. Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE() and clean up after inactive task. Signed-off-by:
-
Weston Andros Adamson authored
Remove redundant BUG_ON(). Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE(). Signed-off-by:
-
Weston Andros Adamson authored
Remove unneeded BUG_ON() Signed-off-by:
-
Weston Andros Adamson authored
Replace BUG_ON() with WARN_ON_ONCE() - rpc_run_bc_task calls rpc_init_task() then increments the tk_count, so this is a simple sanity check that if hit once would hit every time this code path is executed. Signed-off-by:
-
Weston Andros Adamson authored
Print a KERN_INFO message before rpc_d_lookup_sb returns NULL, like other error paths in that function. Signed-off-by:
-
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
- Convert the non-trivial ones into WARN_ON_ONCE(). - Remove the trivial refcounting BUGs Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
- Offset bound checks are done in the NFS client code. - So are filehandle size checks - The cookie length is a constant - The utsname()->nodename is already bounded Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
...and ensure that we set the return value for nfs_page_async_flush() to zero! (Reported-by: Dros Adamson) Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
Signed-off-by: -
Trond Myklebust authored
An EAGAIN return value would be unexpected, but there is no reason to BUG... Signed-off-by: -
Trond Myklebust authored
Convert the ones that are not trivial to check into WARN_ON_ONCE(). Remove checks for things such as NFS2_MAXPATHLEN, which are trivially done by the caller. Add a comment to the case of nfs3_xdr_enc_setacl3args. What is being done there is just wrong... Signed-off-by: -
Trond Myklebust authored
If the nfs_client fails to initialise correctly, then it will return an error condition. Signed-off-by:
-
