linux.git/include/linux/sunrpc/xprt.h, branch v2.6.31

nfs41: Rename rq_received to rq_reply_bytes_recvd

2009-06-17T21:11:40+00:00

The 'rq_received' member of 'struct rpc_rqst' is used to track when we
have received a reply to our request.  With v4.1, the backchannel
can now accept callback requests over the existing connection.  Rename
this field to make it clear that it is only used for tracking reply bytes
and not all bytes received on the connection.

Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy

nfs41: Add backchannel processing support to RPC state machine

2009-06-17T21:11:24+00:00

Adds rpc_run_bc_task() which is called by the NFS callback service to
process backchannel requests.  It performs similar work to rpc_run_task()
though "schedules" the backchannel task to be executed starting at the
call_trasmit state in the RPC state machine.

It also introduces some miscellaneous updates to the argument validation,
call_transmit, and transport cleanup functions to take into account
that there are now forechannel and backchannel tasks.

Backchannel requests do not carry an RPC message structure, since the
payload has already been XDR encoded using the existing NFSv4 callback
mechanism.

Introduce a new transmit state for the client to reply on to backchannel
requests.  This new state simply reserves the transport and issues the
reply.  In case of a connection related error, disconnects the transport and
drops the reply.  It requires the forechannel to re-establish the connection
and the server to retransmit the request, as stated in NFSv4.1 section
2.9.2 "Client and Server Transport Behavior".

Note: There is no need to loop attempting to reserve the transport.  If EAGAIN
is returned by xprt_prepare_transmit(), return with tk_status == 0,
setting tk_action to call_bc_transmit.  rpc_execute() will invoke it again
after the task is taken off the sleep queue.

[nfs41: rpc_run_bc_task() need not be exported outside RPC module]
[nfs41: New call_bc_transmit RPC state]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy 
[nfs41: Backchannel: No need to loop in call_bc_transmit()]
Signed-off-by: Andy Adamson 
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy 
[rpc_count_iostats incorrectly exits early]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy 
[Convert rpc_reply_expected() to inline function]
[Remove unnecessary BUG_ON()]
[Rename variable]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy

nfs41: New backchannel helper routines

2009-06-17T20:06:14+00:00

This patch introduces support to setup the callback xprt on the client side.
It allocates/ destroys the preallocated memory structures used to process
backchannel requests.

At setup time, xprt_setup_backchannel() is invoked to allocate one or
more rpc_rqst structures and substructures.  This ensures that they
are available when an RPC callback arrives.  The rpc_rqst structures
are maintained in a linked list attached to the rpc_xprt structure.
We keep track of the number of allocations so that they can be correctly
removed when the channel is destroyed.

When an RPC callback arrives, xprt_alloc_bc_request() is invoked to
obtain a preallocated rpc_rqst structure.  An rpc_xprt structure is
returned, and its RPC_BC_PREALLOC_IN_USE bit is set in
rpc_xprt->bc_flags.  The structure is removed from the the list
since it is now in use, and it will be later added back when its
user is done with it.

After the RPC callback replies, the rpc_rqst structure is returned
by invoking xprt_free_bc_request().  This clears the
RPC_BC_PREALLOC_IN_USE bit and adds it back to the list, allowing it
to be reused by a subsequent RPC callback request.

To be consistent with the reception of RPC messages, the backchannel requests
should be placed into the 'struct rpc_rqst' rq_rcv_buf, which is then in turn
copied to the 'struct rpc_rqst' rq_private_buf.

[nfs41: Preallocate rpc_rqst receive buffer for handling callbacks]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy 
[Update copyright notice and explain page allocation]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy

nfs41: client callback structures

2009-06-17T20:06:13+00:00

Adds new list of rpc_xprt structures, and a readers/writers lock to
protect the list.  The list is used to preallocate resources for
the backchannel during backchannel requests.  Callbacks are not
expected to cause significant latency, so only one callback will
be allowed at this time.

It also adds a pointer to the NFS callback service so that
requests can be directed to it for processing.

New callback members added to svc_serv. The NFSv4.1 callback service will
sleep on the svc_serv->svc_cb_waitq until new callback requests arrive.
The request will be queued in svc_serv->svc_cb_list. This patch adds this
list, the sleep queue and spinlock to svc_serv.

[nfs41: NFSv4.1 callback support]
Signed-off-by: Ricardo Labiaga 
Signed-off-by: Benny Halevy

SUNRPC: Fix the problem of EADDRNOTAVAIL syslog floods on reconnect

2009-05-02T23:35:08+00:00

See http://bugzilla.kernel.org/show_bug.cgi?id=13034

If the port gets into a TIME_WAIT state, then we cannot reconnect without
binding to a new port.

Tested-by: Petr Vandrovec 
Tested-by: Jean Delvare 
Signed-off-by: Trond Myklebust 
Signed-off-by: Linus Torvalds

SUNRPC: Add the equivalent of the linger and linger2 timeouts to RPC sockets

2009-03-19T19:17:34+00:00

This fixes a regression against FreeBSD servers as reported by Tomas
Kasparek. Apparently when using RPC over a TCP socket, the FreeBSD servers
don't ever react to the client closing the socket, and so commit
e06799f958bf7f9f8fae15f0c6f519953fb0257c (SUNRPC: Use shutdown() instead of
close() when disconnecting a TCP socket) causes the setup to hang forever
whenever the client attempts to close and then reconnect.

We break the deadlock by adding a 'linger2' style timeout to the socket,
after which, the client will abort the connection using a TCP 'RST'.

The default timeout is set to 15 seconds. A subsequent patch will put it
under user control by means of a systctl.

Signed-off-by: Trond Myklebust

SUNRPC: dynamically load RPC transport modules on-demand

2009-03-11T18:37:56+00:00

Provide an api to attempt to load any necessary kernel RPC
client transport module automatically. By convention, the
desired module name is "xprt"+"transport name". For example,
when NFS mounting with "-o proto=rdma", attempt to load the
"xprtrdma" module.

Signed-off-by: Tom Talpey 
Cc: Chuck Lever 
Signed-off-by: Trond Myklebust

sunrpc: get rid of rpc_rqst.rq_bufsize

2008-12-23T21:06:13+00:00

rq_bufsize is not used.

Signed-off-by: Mike Sager 
Signed-off-by: Benny Halevy 
Signed-off-by: Trond Myklebust

SUNRPC: Don't disconnect more than once if retransmitting NFSv4 requests

2008-04-19T20:55:12+00:00

NFSv4 requires us to ensure that we break the TCP connection before we're
allowed to retransmit a request. However in the case where we're
retransmitting several requests that have been sent on the same
connection, we need to ensure that we don't interfere with the attempt to
reconnect and/or break the connection again once it has been established.

We therefore introduce a 'connection' cookie that is bumped every time a
connection is broken. This allows requests to track if they need to force a
disconnection.

Signed-off-by: Trond Myklebust

SUNRPC: Fix up xprt_write_space()

2008-04-19T20:52:44+00:00

The rest of the networking layer uses SOCK_ASYNC_NOSPACE to signal whether
or not we have someone waiting for buffer memory. Convert the SUNRPC layer
to use the same idiom.
Remove the unlikely()s in xs_udp_write_space and xs_tcp_write_space. In
fact, the most common case will be that there is nobody waiting for buffer
space.

SOCK_NOSPACE is there to tell the TCP layer whether or not the cwnd was
limited by the application window. Ensure that we follow the same idiom as
the rest of the networking layer here too.

Finally, ensure that we clear SOCK_ASYNC_NOSPACE once we wake up, so that
write_space() doesn't keep waking things up on xprt->pending.

Signed-off-by: Trond Myklebust