The RPC_TASK_KILLED flag should really not be set from another context
because it can clobber data in the struct task when task->tk_flags is
changed non-atomically.
Let's therefore swap out RPC_TASK_KILLED with an atomic flag, and add
a function to set that flag and safely wake up the task.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If a request transmission fails due to write space or slot unavailability
errors, but the queued task then gets transmitted before it has time to
process the error in call_transmit_status() or call_bc_transmit_status(),
we need to suppress the transmission error code to prevent it from leaking
out of the RPC layer.
Reported-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Tested-by: Chuck Lever <chuck.lever@oracle.com>
This reverts commit 009a82f643.
The ability to optimise here relies on compiler being able to optimise
away tail calls to avoid stack overflows. Unfortunately, we are seeing
reports of problems, so let's just revert.
Reported-by: Daniel Mack <daniel@zonque.org>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If the transport is still connected, then we do want to allow
RPC_SOFTCONN tasks to retry. They should time out if and only if
the connection is broken.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Now that we're using the xdr_stream functions to decode the header,
the test for the minimum reply length is redundant.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Handle the SYSTEM_ERR rpc error by retrying the RPC call as if it
were a garbage argument.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Ensure that when the "garbage args" case falls through, we do set
an error of EIO.
Fixes: a0584ee9ae ("SUNRPC: Use struct xdr_stream when decoding...")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
We must at minimum allocate enough memory to be able to see any auth
errors in the reply from the server.
Fixes: 2c94b8eca1 ("SUNRPC: Use au_rslack when computing reply...")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Before trying to bind a port, ensure we grab the send lock to
ensure that we don't change the port while another task is busy
transmitting requests.
The connect code already takes the send lock in xprt_connect(),
but it is harmless to take it before that.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
In cases where we know the task is not sleeping, try to optimise
away the indirect call to task->tk_action() by replacing it with
a direct call.
Only change tail calls, to allow gcc to perform tail call
elimination.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Before initiating transport actions that require putting the task to sleep,
such as rebinding or reconnecting, we should check whether or not the task
was already transmitted.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
The RPC task wakeup calls all check for RPC_IS_QUEUED() before taking any
locks. In addition, rpc_exit() already calls rpc_wake_up_queued_task().
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Fix a regression where soft and softconn requests are not timing out
as expected.
Fixes: 89f90fe1ad ("SUNRPC: Allow calls to xprt_transmit() to drain...")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Now that transmissions happen through a queue, we require the RPC tasks
to handle error conditions that may have been set while they were
sleeping. The back channel does not currently do this, but assumes
that any error condition happens during its own call to xprt_transmit().
The solution is to ensure that the back channel splits out the
error handling just like the forward channel does.
Fixes: 89f90fe1ad ("SUNRPC: Allow calls to xprt_transmit() to drain...")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If the socket is not connected, then we want to initiate a reconnect
rather that trying to transmit requests. If there is a large number
of requests queued and waiting for the lock in call_transmit(),
then it can take a while for one of the to loop back and retake
the lock in call_connect.
Fixes: 89f90fe1ad ("SUNRPC: Allow calls to xprt_transmit() to drain...")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
au_rslack is significantly smaller than (au_cslack << 2). Using
that value results in smaller receive buffers. In some cases this
eliminates an extra segment in Reply chunks (RPC/RDMA).
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Currently rpc_inline_rcv_pages() uses au_rslack to estimate the
size of the upper layer reply header. This is fine for auth flavors
where au_verfsize == au_rslack.
However, some auth flavors have more going on. krb5i for example has
two more words after the verifier, and another blob following the
RPC message. The calculation involving au_rslack pushes the upper
layer reply header too far into the rcv_buf.
au_rslack is still valuable: it's the amount of buffer space needed
for the reply, and is used when allocating the reply buffer. We'll
keep that.
But, add a new field that can be used to properly estimate the
location of the upper layer header in each RPC reply, based on the
auth flavor in use.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Certain NFS results (eg. READLINK) might expect a data payload that
is not an exact multiple of 4 bytes. In this case, XDR encoding
is required to pad that payload so its length on the wire is a
multiple of 4 bytes. The constants that define the maximum size of
each NFS result do not appear to account for this extra word.
In each case where the data payload is to be received into pages:
- 1 word is added to the size of the receive buffer allocated by
call_allocate
- rpc_inline_rcv_pages subtracts 1 word from @hdrsize so that the
extra buffer space falls into the rcv_buf's tail iovec
- If buf->pagelen is word-aligned, an XDR pad is not needed and
is thus removed from the tail
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
prepare_reply_buffer() and its NFSv4 equivalents expose the details
of the RPC header and the auth slack values to upper layer
consumers, creating a layering violation, and duplicating code.
Remedy these issues by adding a new RPC client API that hides those
details from upper layers in a common helper function.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Modernize and harden the code path that parses an RPC Reply
message.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
- Recover some instruction count because I'm about to introduce a
few xdr_inline_decode call sites
- Replace dprintk() call sites with trace points
- Reduce the hot path so it fits in fewer cachelines
I've also renamed it rpc_decode_header() to match everything else
in the RPC client.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Modernize and harden the code path that constructs each RPC Call
message.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
tsh_size was added to accommodate transports that send a pre-amble
before each RPC message. However, this assumes the pre-amble is
fixed in size, which isn't true for some transports. That makes
tsh_size not very generic.
Also I'd like to make the estimation of RPC send and receive
buffer sizes more precise. tsh_size doesn't currently appear to be
accounted for at all by call_allocate.
Therefore let's just remove the tsh_size concept, and make the only
transports that have a non-zero tsh_size employ a direct approach.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
According to RFC2203, the RPCSEC_GSS sequence numbers are bounded to
an upper limit of MAXSEQ = 0x80000000. Ensure that we handle that
correctly.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When we resend a request, ensure that the 'rq_bytes_sent' is reset
to zero.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Multipathing: In case of NFSv3, rpc_clnt_test_and_add_xprt() adds
the xprt to xprt switch (i.e. xps) if rpc_call_null_helper() returns
success. But in case of NFSv4.1, it needs to do EXCHANGEID to verify
the path along with check for session trunking.
Add the xprt in nfs4_test_session_trunk() only when
nfs4_detect_session_trunking() returns success. Also release refcount
hold by rpc_clnt_setup_test_and_add_xprt().
Signed-off-by: Santosh kumar pradhan <santoshkumar.pradhan@wdc.com>
Tested-by: Suresh Jayaraman <suresh.jayaraman@wdc.com>
Reported-by: Aditya Agnihotri <aditya.agnihotri@wdc.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
SUNRPC has two sorts of credentials, both of which appear as
"struct rpc_cred".
There are "generic credentials" which are supplied by clients
such as NFS and passed in 'struct rpc_message' to indicate
which user should be used to authorize the request, and there
are low-level credentials such as AUTH_NULL, AUTH_UNIX, AUTH_GSS
which describe the credential to be sent over the wires.
This patch replaces all the generic credentials by 'struct cred'
pointers - the credential structure used throughout Linux.
For machine credentials, there is a special 'struct cred *' pointer
which is statically allocated and recognized where needed as
having a special meaning. A look-up of a low-level cred will
map this to a machine credential.
Signed-off-by: NeilBrown <neilb@suse.com>
Acked-by: J. Bruce Fields <bfields@redhat.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The credential passed in rpc_message.rpc_cred is always a
generic credential except in one instance.
When gss_destroying_context() calls rpc_call_null(), it passes
a specific credential that it needs to destroy.
In this case the RPC acts *on* the credential rather than
being authorized by it.
This special case deserves explicit support and providing that will
mean that rpc_message.rpc_cred is *always* generic, allowing
some optimizations.
So add "tk_op_cred" to rpc_task and "rpc_op_cred" to the setup data.
Use this to pass the cred down from rpc_call_null(), and have
rpcauth_bindcred() notice it and bind it in place.
Credit to kernel test robot <fengguang.wu@intel.com> for finding
a bug in earlier version of this patch.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
In almost all cases the credential stored in rpc_message.rpc_cred
is a "generic" credential. One of the two expections is when an
AUTH_NULL credential is used such as for RPC ping requests.
To improve consistency, don't pass an explicit credential in
these cases, but instead pass NULL and set a task flag,
similar to RPC_TASK_ROOTCREDS, which requests that NULL credentials
be used by default.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When NFS creates a machine credential, it is a "generic" credential,
not tied to any auth protocol, and is really just a container for
the princpal name.
This doesn't get linked to a genuine credential until rpcauth_bindcred()
is called.
The lookup always succeeds, so various places that test if the machine
credential is NULL, are pointless.
As a step towards getting rid of generic credentials, this patch gets
rid of generic machine credentials. The nfs_client and rpc_client
just hold a pointer to a constant principal name.
When a machine credential is wanted, a special static 'struct rpc_cred'
pointer is used. rpcauth_bindcred() recognizes this, finds the
principal from the client, and binds the correct credential.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
When the socket is closed, we need to call xprt_disconnect_done() in order
to clean up the XPRT_WRITE_SPACE flag, and wake up the sleeping tasks.
However, we also want to ensure that we don't wake them up before the socket
is closed, since that would cause thundering herd issues with everyone
piling up to retransmit before the TCP shutdown dance has completed.
Only the task that holds XPRT_LOCKED needs to wake up early in order to
allow the close to complete.
Reported-by: Dave Wysochanski <dwysocha@redhat.com>
Reported-by: Scott Mayhew <smayhew@redhat.com>
Cc: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Tested-by: Chuck Lever <chuck.lever@oracle.com>
If we retransmit an RPC request, we currently end up clobbering the
value of req->rq_rcv_buf.bvec that was allocated by the initial call to
xprt_request_prepare(req).
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If a task failed to get the write lock in the call to xprt_connect(), then
it will be queued on xprt->sending. In that case, it is possible for it
to get transmitted before the call to call_connect_status(), in which
case it needs to be handled by call_transmit_status() instead.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Add a bvec array to struct xdr_buf, and have the client allocate it
when we need to receive data into pages.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Treat socket write space handling in the same way we now treat transport
congestion: by denying the XPRT_LOCK until the transport signals that it
has free buffer space.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Both RDMA and UDP transports require the request to get a "congestion control"
credit before they can be transmitted. Right now, this is done when
the request locks the socket. We'd like it to happen when a request attempts
to be transmitted for the first time.
In order to support retransmission of requests that already hold such
credits, we also want to ensure that they get queued first, so that we
don't deadlock with requests that have yet to obtain a credit.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Fix up the back channel code to recognise that it has already been
transmitted, so does not need to be called again.
Also ensure that we set req->rq_task.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Allow the caller in clnt.c to call into the code to wait for a reply
after calling xprt_transmit(). Again, the reason is that the backchannel
code does not need this functionality.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Separate out the action of adding a request to the reply queue so that the
backchannel code can simply skip calling it altogether.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Currently, we grab the socket bit lock before we allow the message
to be XDR encoded. That significantly slows down the transmission
rate, since we serialise on a potentially blocking operation.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
If a message has been encoded using RPCSEC_GSS, the server is
maintaining a window of sequence numbers that it considers valid.
The client should normally be tracking that window, and needs to
verify that the sequence number used by the message being transmitted
still lies inside the window of validity.
So far, we've been able to assume this condition would be realised
automatically, since the client has been encoding the message only
after taking the socket lock. Once we change that condition, we
will need the explicit check.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
