If the initial attempt at trunking detection using the krb5i auth flavor
fails with -EACCES, -NFS4ERR_CLID_INUSE, or -NFS4ERR_WRONGSEC, then the
NFS client tries again using auth_sys, cloning the rpc_clnt in the
process. If this second attempt at trunking detection succeeds, then
the resulting nfs_client->cl_rpcclient winds up having cl_max_connect=0
and subsequent attempts to add additional transport connections to the
rpc_clnt will fail with a message similar to the following being logged:
[502044.312640] SUNRPC: reached max allowed number (0) did not add
transport to server: 192.168.122.3
Signed-off-by: Scott Mayhew <smayhew@redhat.com>
Fixes: dc48e0abee ("SUNRPC enforce creation of no more than max_connect xprts")
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Ensure that the gssproxy client connects to the server from the gssproxy
daemon process context so that the AF_LOCAL socket connection is done
using the correct path and namespaces.
Fixes: 1d658336b0 ("SUNRPC: Add RPC based upcall mechanism for RPCGSS auth")
Cc: stable@vger.kernel.org
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
This reverts commit 892de36fd4.
The gssproxy server is unresponsive when it calls into the kernel to
start the upcall service, so it will not reply to our RPC ping at all.
Reported-by: "J.Bruce Fields" <bfields@fieldses.org>
Fixes: 892de36fd4 ("SUNRPC: Ensure gss-proxy connects on setup")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
For reasons best known to the author, gss-proxy does not implement a
NULL procedure, and returns RPC_PROC_UNAVAIL. However we still want to
ensure that we connect to the service at setup time.
So add a quirk-flag specially for this case.
Fixes: 1d658336b0 ("SUNRPC: Add RPC based upcall mechanism for RPCGSS auth")
Cc: stable@vger.kernel.org
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
A relocated task must release its previous transport.
Fixes: 82ee41b85c ("SUNRPC don't resend a task on an offlined transport")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If the call to rpc_alloc_task() fails, then ensure that the calldata is
released, and that rpc_run_task() and rpc_run_bc_task() bail out early.
Reported-by: NeilBrown <neilb@suse.de>
Fixes: 910ad38697 ("NFS: Fix memory allocation in rpc_alloc_task()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
We need to handle ENFILE, ENOBUFS, and ENOMEM, because
xprt_wake_pending_tasks() can be called with any one of these due to
socket creation failures.
Fixes: b61d59fffd ("SUNRPC: xs_tcp_connect_worker{4,6}: merge common code")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Both call_transmit() and call_bc_transmit() can now return ENOMEM, so
let's make sure that we handle the errors gracefully.
Fixes: 0472e47660 ("SUNRPC: Convert socket page send code to use iov_iter()")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If ->request_prepare() detects an error, it sets ->rq_task->tk_status.
This is easy for callers to ignore.
The only caller is xprt_request_enqueue_receive() and it does ignore the
error, as does call_encode() which calls it. This can result in a
request being queued to receive a reply without an allocated receive buffer.
So instead of setting rq_task->tk_status, return an error, and store in
->tk_status only in call_encode();
The call to xprt_request_enqueue_receive() is now earlier in
call_encode(), where the error can still be handled.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
When a task is being retried, due to an NFS error, if the assigned
transport has been put offline and the task is relocatable pick a new
transport.
Fixes: 6f081693e7 ("sunrpc: remove an offlined xprt using sysfs")
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If we are swapping over NFSv4, we may not be able to allocate memory to
start the state-manager thread at the time when we need it.
So keep it always running when swap is enabled, and just signal it to
start.
This requires updating and testing the cl_swapper count on the root
rpc_clnt after following all ->cl_parent links.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
rpc tasks can be marked as RPC_TASK_SWAPPER. This causes GFP_MEMALLOC
to be used for some allocations. This is needed in some cases, but not
in all where it is currently provided, and in some where it isn't
provided.
Currently *all* tasks associated with a rpc_client on which swap is
enabled get the flag and hence some GFP_MEMALLOC support.
GFP_MEMALLOC is provided for ->buf_alloc() but only swap-writes need it.
However xdr_alloc_bvec does not get GFP_MEMALLOC - though it often does
need it.
xdr_alloc_bvec is called while the XPRT_LOCK is held. If this blocks,
then it blocks all other queued tasks. So this allocation needs
GFP_MEMALLOC for *all* requests, not just writes, when the xprt is used
for any swap writes.
Similarly, if the transport is not connected, that will block all
requests including swap writes, so memory allocations should get
GFP_MEMALLOC if swap writes are possible.
So with this patch:
1/ we ONLY set RPC_TASK_SWAPPER for swap writes.
2/ __rpc_execute() sets PF_MEMALLOC while handling any task
with RPC_TASK_SWAPPER set, or when handling any task that
holds the XPRT_LOCKED lock on an xprt used for swap.
This removes the need for the RPC_IS_SWAPPER() test
in ->buf_alloc handlers.
3/ xprt_prepare_transmit() sets PF_MEMALLOC after locking
any task to a swapper xprt. __rpc_execute() will clear it.
3/ PF_MEMALLOC is set for all the connect workers.
Reviewed-by: Chuck Lever <chuck.lever@oracle.com> (for xprtrdma parts)
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
When memory is short, new worker threads cannot be created and we depend
on the minimum one rpciod thread to be able to handle everything. So it
must not block waiting for memory.
mempools are particularly a problem as memory can only be released back
to the mempool by an async rpc task running. If all available workqueue
threads are waiting on the mempool, no thread is available to return
anything.
lookup_cred() can block on a mempool or kmalloc - and this can cause
deadlocks. So add a new RPCAUTH_LOOKUP flag for async lookups and don't
block on memory. If the -ENOMEM gets back to call_refreshresult(), wait
a short while and try again. HZ>>4 is chosen as it is used elsewhere
for -ENOMEM retries.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
The sections which should not re-enter the filesystem are already
protected with memalloc_nofs_save/restore calls, so it is better to use
GFP_KERNEL in these calls to allow better performance for synchronous
RPC calls.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If the supplied argument doesn't specify the transport type, use the
type of the existing rpc clnt and its existing transport.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In rpc_task_set_client(), testing for a NULL clnt is not necessary, as
clnt should always be a valid pointer to a rpc_client.
Signed-off-by: Thiago Rafael Becker <trbecker@gmail.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If we are adding new transports via rpc_clnt_test_and_add_xprt()
then check if we've reached the limit. Currently only pnfs path
adds transports via that function but this is done in
preparation when the client would add new transports when
session trunking is detected. A warning is logged if the
limit is reached.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Currently, xprt_switch keeps a number of all xprts (xps_nxprts)
that were added to the switch regardless of whethere it's an
nconnect transport or a transport to a trunkable address.
Introduce a new counter to keep track of transports to unique
destination addresses per xprt_switch.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
There are now tools in the refcount library that allow us to convert the
client shutdown code.
Reported-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In some rare failure modes, the server is actually reading the
transport, but then just dropping the requests on the floor.
TCP_USER_TIMEOUT cannot detect that case.
Prevent such a stuck server from pinning client resources
indefinitely by ensuring that certain idempotent requests
(such as NULL) can time out even if the connection is still
operational.
Otherwise rpc_bind_new_program(), gss_destroy_cred(), or
rpc_clnt_test_and_add_xprt() can wait forever.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Make it use the rpc_null_call_helper() so that it can share the
new rpc_call_ops structure to be introduced in the next patch.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Once a transport has been put offline, this transport can be also
removed from the list of transports. Any tasks that have been stuck
on this transport would find the next available active transport
and be re-tried. This transport would be removed from the xprt_switch
list and freed.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
When an RPC client gets created it's first transport is special
and should be marked a main transport.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
An rpc client uses a transport switch and one ore more transports
associated with that switch. Since transports are shared among
rpc clients, create a symlink into the xprt_switch directory
instead of duplicating entries under each rpc client.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
These will eventually have files placed under them for sysfs operations.
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If sunrpc.tcp_max_slot_table_entries is small and there are tasks
on the backlog queue, then when a request completes it is freed and the
first task on the queue is woken. The expectation is that it will wake
and claim that request. However if it was a sync task and the waiting
process was killed at just that moment, it will wake and NOT claim the
request.
As long as TASK_CONGESTED remains set, requests can only be claimed by
tasks woken from the backlog, and they are woken only as requests are
freed, so when a task doesn't claim a request, no other task can ever
get that request until TASK_CONGESTED is cleared. Each time this
happens the number of available requests is decreased by one.
With a sufficiently high workload and sufficiently low setting of
max_slot (16 in the case where this was seen), TASK_CONGESTED can remain
set for an extended period, and the above scenario (of a process being
killed just as its task was woken) can repeat until no requests can be
allocated. Then traffic stops.
This patch addresses the problem by introducing a positive handover of a
request from a completing task to a backlog task - the request is never
freed when there is a backlog.
When a task is woken it might not already have a request attached in
which case it is *not* freed (as with current code) but is initialised
(if needed) and used. If it isn't used it will eventually be freed by
rpc_exit_task(). xprt_release() is enhanced to be able to correctly
release an uninitialised request.
Fixes: ba60eb25ff ("SUNRPC: Fix a livelock problem in the xprt->backlog queue")
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Fix a misplaced barrier in call_decode. The struct rpc_rqst is modified
as follows by xprt_complete_rqst:
req->rq_private_buf.len = copied;
/* Ensure all writes are done before we update */
/* req->rq_reply_bytes_recvd */
smp_wmb();
req->rq_reply_bytes_recvd = copied;
And currently read as follows by call_decode:
smp_rmb(); // misplaced
if (!req->rq_reply_bytes_recvd)
goto out;
req->rq_rcv_buf.len = req->rq_private_buf.len;
This patch places the smp_rmb after the if to ensure that
rq_reply_bytes_recvd and rq_private_buf.len are read in order.
Fixes: 9ba828861c ("SUNRPC: Don't try to parse incomplete RPC messages")
Signed-off-by: Baptiste Lepers <baptiste.lepers@gmail.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
I've hit some crashes that occur in the xprt_rdma_inject_disconnect
path. It appears that, for some provides, rdma_disconnect() can
take so long that the transport can disconnect and release its
hardware resources while rdma_disconnect() is still running,
resulting in a UAF in the provider.
The transport's fault injection method may depend on the stability
of transport data structures. That means it needs to be invoked
only from contexts that hold the transport write lock.
Fixes: 4a06825839 ("SUNRPC: Transport fault injection")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
rpc_prepare_reply_pages() currently expects the 'hdrsize' argument to
contain the length of the data that we expect to want placed in the head
kvec plus a count of 1 word of padding that is placed after the page data.
This is very confusing when trying to read the code, and sometimes leads
to callers adding an arbitrary value of '1' just in order to satisfy the
requirement (whether or not the page data actually needs such padding).
This patch aims to clarify the code by changing the 'hdrsize' argument
to remove that 1 word of padding. This means we need to subtract the
padding from all the existing callers.
Fixes: 02ef04e432 ("NFS: Account for XDR pad of buf->pages")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Time to remove dprintk call sites in here.
Regarding the rpc_bind_status tracepoint: It's friendlier to
administrators if they don't have to look up the error code to
figure out what went wrong. Replace trace_rpc_bind_status with a
set of tracepoints that report more specifically what the problem
was, and what RPC program/version was being queried.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up.
When enabled, this dprintk adds a line in /var/log/messages after
every RPC that reports the task ID (no connection to on the wire
XID values) and the RPC's result (no connection to the program,
operation, or the arguments and results).
Thus it's value is pretty low. Let's remove it.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: Replace dprintk call sites.
Note that rpc_call_rpcerror() already has a trace point, so perhaps
adding trace_rpc_refresh_status() isn't necessary. However, it does
report a particular category of error.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
For a long while we've wanted a tracepoint that fires when a major
timeout is reported in the system log. Such a tracepoint can be
attached to other actions that can take place when a timeout is
detected (eg, server or connection health assessment).
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
This trace event can be used to audit transport connections from the
client.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: The rpc_rpc_request tracepoint serves the same purpose.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up: The rpc_task_run_action tracepoint serves the same
purpose.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Introduce a tracepoint in call_allocate that reports the exact
sizes in the RPC buffer allocation request and the status of the
result. This helps catch problems with XDR buffer provisioning,
and replaces transport-specific debugging instrumentation.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Current behaviour: every time a v3 operation is re-sent to the server
we update (double) the timeout. There is no distinction between whether
or not the previous timer had expired before the re-sent happened.
Here's the scenario:
1. Client sends a v3 operation
2. Server RST-s the connection (prior to the timeout) (eg., connection
is immediately reset)
3. Client re-sends a v3 operation but the timeout is now 120sec.
As a result, an application sees 2mins pause before a retry in case
server again does not reply. Where as if a connection reset didn't
change the timeout value, the client would have re-tried (the 3rd
time) after 60secs.
Signed-off-by: Olga Kornievskaia <kolga@netapp.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Move the RPC_TASK_SOFTCONN flag into rpc_call_null_helper(). The
only minor behavior change is that it is now also set when
destroying GSS contexts.
This gives a better guarantee that gss_send_destroy_context() will
not hang for long if a connection cannot be established.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up.
All of rpc_call_null_helper() call sites assert RPC_TASK_SOFT, so
move that setting into rpc_call_null_helper() itself.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Clean up.
Commit a52458b48a ("NFS/NFSD/SUNRPC: replace generic creds with
'struct cred'.") made rpc_call_null_helper() set RPC_TASK_NULLCREDS
unconditionally. Therefore there's no need for
rpc_call_null_helper()'s call sites to set RPC_TASK_NULLCREDS.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The "create" tracepoint records parts of the rpc_create arguments,
and the shutdown tracepoint records when the rpc_clnt is about to
signal pending tasks and destroy auths.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
To help tie the recorded xdr_buf to a particular RPC transaction,
the client side version of this class should display task ID
information and the server side one should show the request's XID.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Add a tracepoint in another common exit point for failing RPCs.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Each rpc_client has a cl_clid which is allocated from a global ida, and
a debugfs directory which is named after cl_clid.
We're releasing the cl_clid before we free the debugfs directory named
after it. As soon as the cl_clid is released, that value is available
for another newly created client.
That leaves a window where another client may attempt to create a new
debugfs directory with the same name as the not-yet-deleted debugfs
directory from the dying client. Symptoms are log messages like
Directory 4 with parent 'rpc_clnt' already present!
Fixes: 7c4310ff56 "SUNRPC: defer slow parts of rpc_free_client() to a workqueue."
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Ensure that signalled ASYNC rpc_tasks exit immediately instead of
spinning until a timeout (or forever).
To avoid checking for the signal flag on every scheduler iteration,
the check is instead introduced in the client's finite state
machine.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Fixes: ae67bd3821 ("SUNRPC: Fix up task signalling")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Parts of rpc_free_client() were recently moved to
a separate rpc_free_clent_work(). This introduced
a use-after-free as rpc_clnt_remove_pipedir() calls
rpc_net_ns(), and that uses clnt->cl_xprt which has already
been freed.
So move the call to xprt_put() after the call to
rpc_clnt_remove_pipedir().
Reported-by: syzbot+22b5ef302c7c40d94ea8@syzkaller.appspotmail.com
Fixes: 7c4310ff56 ("SUNRPC: defer slow parts of rpc_free_client() to a workqueue.")
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
The rpciod workqueue is on the write-out path for freeing dirty memory,
so it is important that it never block waiting for memory to be
allocated - this can lead to a deadlock.
rpc_execute() - which is often called by an rpciod work item - calls
rcp_task_release_client() which can lead to rpc_free_client().
rpc_free_client() makes two calls which could potentially block wating
for memory allocation.
rpc_clnt_debugfs_unregister() calls into debugfs and will block while
any of the debugfs files are being accessed. In particular it can block
while any of the 'open' methods are being called and all of these use
malloc for one thing or another. So this can deadlock if the memory
allocation waits for NFS to complete some writes via rpciod.
rpc_clnt_remove_pipedir() can take the inode_lock() and while it isn't
obvious that memory allocations can happen while the lock it held, it is
safer to assume they might and to not let rpciod call
rpc_clnt_remove_pipedir().
So this patch moves these two calls (together with the final kfree() and
rpciod_down()) into a work-item to be run from the system work-queue.
rpciod can continue its important work, and the final stages of the free
can happen whenever they happen.
I have seen this deadlock on a 4.12 based kernel where debugfs used
synchronize_srcu() when removing objects. synchronize_srcu() requires a
workqueue and there were no free workther threads and none could be
allocated. While debugsfs no longer uses SRCU, I believe the deadlock
is still possible.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
rpc_clnt_test_and_add_xprt() invokes rpc_call_null_helper(), which
return the value of rpc_run_task() to "task". Since rpc_run_task() is
impossible to return an ERR pointer, there is no need to add the
IS_ERR() condition on "task" here. So we need to remove it.
Fixes: 7f55489058 ("SUNRPC: Allow addition of new transports to a struct rpc_clnt")
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>