The common case: There are 13 to 14 actions per RPC, and tk_callback
is non-NULL in only one of them. There's no need to store a NULL in
the tk_callback field during each FSM step.

This slightly improves throughput results in dbench and other multi-
threaded benchmarks on my two-socket client on 56Gb InfiniBand, but
will probably be inconsequential on slower systems.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

This shows up in every RPC:

     kworker/4:1-19772 [004]  3467.373443: rpc_task_run_action:  task:4711@2 flags=0e81 state=0005 status=0 action=call_status
     kworker/4:1-19772 [004]  3467.373444: rpc_task_run_action:  task:4711@2 flags=0e81 state=0005 status=0 action=call_status

What's actually going on is that the first iteration of the RPC
scheduler is invoking the function in tk_callback (in this case,
xprt_timer), then invoking call_status on the next iteration.

Feeding do_action, rather than tk_action, to the "task_run_action"
trace point will now always display the correct FSM step.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Track RPC timeouts: report the XID and the server address to match
the content of network capture.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Fix kernel-doc warnings in net/sunrpc/xprtrdma/ .

net/sunrpc/xprtrdma/verbs.c:1575: warning: No description found for parameter 'count'
net/sunrpc/xprtrdma/verbs.c:1575: warning: Excess function parameter 'min_reqs' description in 'rpcrdma_ep_post_extra_recv'

net/sunrpc/xprtrdma/backchannel.c:288: warning: No description found for parameter 'r_xprt'
net/sunrpc/xprtrdma/backchannel.c:288: warning: Excess function parameter 'xprt' description in 'rpcrdma_bc_receive_call'
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

The contents of seg->mr_len changed when ->ro_map stopped returning
the full chunk length in the first segment. Count the full length of
each Write chunk, not the length of the first segment (which now can
only be as large as a page).

Fixes: 9d6b0409 ("xprtrdma: Place registered MWs on a ... ")
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

This includes decoding Write and Reply chunks, and fixing up inline
payloads.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up: Code review suggested that a common bit of code can be
placed into a helper function, and this gives us fewer places to
stick an "I DMA unmapped something" trace point.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up: struct rpcrdma_mw was named after Memory Windows, but
xprtrdma no longer supports a Memory Window registration mode.
Rename rpcrdma_mw and its fields to reduce confusion and make
the code more sensible to read.

Renaming "mw" was suggested by Tom Talpey, the author of the
original xprtrdma implementation. It's a good idea, but I haven't
done this until now because it's a huge diffstat for no benefit
other than code readability.

However, I'm about to introduce static trace points that expose
a few of xprtrdma's internal data structures. They should make sense
in the trace report, and it's reasonable to treat trace points as a
kernel API contract which might be difficult to change later.

While I'm churning things up, two additional changes:
- rename variables unhelpfully called "r" to "mr", to improve code
  clarity, and
- rename the MR-related helper functions using the form
  "rpcrdma_mr_<verb>", to be consistent with other areas of the
  code.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up: Over time, the industry has adopted the term "frwr"
instead of "frmr". The term "frwr" is now more widely recognized.

For the past couple of years I've attempted to add new code using
"frwr" , but there still remains plenty of older code that still
uses "frmr". Replace all usage of "frmr" to avoid confusion.

While we're churning code, rename variables unhelpfully called "f"
to "frwr", to improve code clarity.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

No need for the overhead of atomically setting and clearing this bit
flag for every use of a pre-allocated backchannel rpc_rqst. These
are a distinct pool of rpc_rqsts that are used only for callback
operations, so it is safe to simply leave the bit set.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up. @rqst is set up differently for backchannel Replies. For
example, rqst->rq_task and task->tk_client are both NULL. So it is
easier to understand and maintain this code path if it is separated.

Also, we can get rid of the confusing rl_connect_cookie hack in
rpcrdma_bc_receive_call.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Since commit 5a6d1db4 ("SUNRPC: Add a transport-specific private
field in rpc_rqst"), the rpc_rqst's for RPC-over-RDMA backchannel
operations leave rq_buffer set to NULL.

xprt_release does not invoke ->op->buf_free when rq_buffer is NULL.
The RPCRDMA_REQ_F_BACKCHANNEL check in xprt_rdma_free is therefore
redundant because xprt_rdma_free is not invoked for backchannel
requests.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up. This logic is related to marshaling the request, and I'd
like to keep everything that touches req->rl_registered close
together, for CPU cache efficiency.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up a harmless oversight. xprtrdma's ->set_port method has
never properly supported IPv6.

This issue has never been a problem because NFS/RDMA mounts have
always required "port=20049", thus so far, rpcbind is not invoked
for these mounts.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Save more space in struct rpcrdma_xprt by removing the redundant
"addr" field from struct rpcrdma_create_data_internal. Wherever
we have rpcrdma_xprt, we also have the rpc_xprt, which has a
sockaddr_storage field with the same content.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

This makes the address strings available for debugging messages in
earlier stages of transport set up.

The first benefit is to get rid of the single-use rep_remote_addr
field, saving 128+ bytes in struct rpcrdma_ep.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up. Remove fields that should have been removed by
commit b3221d6a ("xprtrdma: Remove logic that constructs
RDMA_MSGP type calls").
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Clean up.

Commit b5f0afbe ("xprtrdma: Per-connection pad optimization")
should have removed this.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Refactoring change: Remote Invalidation is particular to the memory
registration mode that is use. Use a callout instead of a generic
function to handle Remote Invalidation.

This gets rid of the 8-byte flags field in struct rpcrdma_mw, of
which only a single bit flag has been allocated.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

The rpcrdma_req is not shared yet, and its associated Send hasn't
been posted, thus RMW should be safe. There's no need for the
expense of a lock cycle here.

Fixes: 0ba6f370 ("xprtrdma: Refactor rpcrdma_deferred_completion")
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

The backchannel code uses rpcrdma_recv_buffer_put to add new reps
to the free rep list. This also decrements rb_recv_count, which
spoofs the receive overrun logic in rpcrdma_buffer_get_rep.

Commit 9b06688b ("xprtrdma: Fix additional uses of
spin_lock_irqsave(rb_lock)") replaced the original open-coded
list_add with a call to rpcrdma_recv_buffer_put(), but then a year
later, commit 05c97466 ("xprtrdma: Fix receive buffer
accounting") added rep accounting to rpcrdma_recv_buffer_put.
It was an oversight to let the backchannel continue to use this
function.

The fix this, let's combine the "add to free list" logic with
rpcrdma_create_rep.

Also, do not allocate RPCRDMA_MAX_BC_REQUESTS rpcrdma_reps in
rpcrdma_buffer_create and then allocate additional rpcrdma_reps in
rpcrdma_bc_setup_reps. Allocating the extra reps during backchannel
set-up is sufficient.

Fixes: 05c97466 ("xprtrdma: Fix receive buffer accounting")
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

This leak has been around forever, and is exceptionally rare.

EINVAL causes mount to fail with "an incorrect mount option was
specified" although it's not likely that one of the mount
options is incorrect. Instead, return ENODEV in this case, as this
appears to be an issue with system or device configuration rather
than a specific mount option.

Some obsolete comments are also removed.
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Move static key increments to the beginning of the init function
so they pair 1:1 with decrements in ingress/clsact_destroy,
which is called in case ingress/clsact_init fails.

Fixes: 6529eaba ("net: sched: introduce tcf block infractructure")
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NCong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since in qdisc_create, the destroy op is called when init fails, we
don't do cleanup in init and leave it up to destroy.
This fixes use-after-free when trying to put already freed block.

Fixes: 6e40cf2d ("net: sched: use extended variants of block_get/put in ingress and clsact qdiscs")
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NCong Wang <xiyou.wangcong@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

We must ensure that the call to rpc_sleep_on() in xprt_transmit() cannot
race with the call to xprt_complete_rqst().
Reported-by: NChuck Lever <chuck.lever@oracle.com>
Link: https://bugzilla.linux-nfs.org/show_bug.cgi?id=317
Fixes: ce7c252a ("SUNRPC: Add a separate spinlock to protect..")
Cc: stable@vger.kernel.org # 4.14+
Reviewed-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NTrond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

Commit d8f532d2 ("xprtrdma: Invoke rpcrdma_reply_handler
directly from RECV completion") introduced a performance regression
for NFS I/O small enough to not need memory registration. In multi-
threaded benchmarks that generate primarily small I/O requests,
IOPS throughput is reduced by nearly a third. This patch restores
the previous level of throughput.

Because workqueues are typically BOUND (in particular ib_comp_wq,
nfsiod_workqueue, and rpciod_workqueue), NFS/RDMA workloads tend
to aggregate on the CPU that is handling Receive completions.

The usual approach to addressing this problem is to create a QP
and CQ for each CPU, and then schedule transactions on the QP
for the CPU where you want the transaction to complete. The
transaction then does not require an extra context switch during
completion to end up on the same CPU where the transaction was
started.

This approach doesn't work for the Linux NFS/RDMA client because
currently the Linux NFS client does not support multiple connections
per client-server pair, and the RDMA core API does not make it
straightforward for ULPs to determine which CPU is responsible for
handling Receive completions for a CQ.

So for the moment, record the CPU number in the rpcrdma_req before
the transport sends each RPC Call. Then during Receive completion,
queue the RPC completion on that same CPU.

Additionally, move all RPC completion processing to the deferred
handler so that even RPCs with simple small replies complete on
the CPU that sent the corresponding RPC Call.

Fixes: d8f532d2 ("xprtrdma: Invoke rpcrdma_reply_handler ...")
Signed-off-by: NChuck Lever <chuck.lever@oracle.com>
Signed-off-by: NAnna Schumaker <Anna.Schumaker@Netapp.com>

If pskb_may_pull return failed, return PACKET_REJECT instead of -ENOMEM.

Fixes: 84e54fe0 ("gre: introduce native tunnel support for ERSPAN")
Cc: William Tu <u9012063@gmail.com>
Signed-off-by: NHaishuang Yan <yanhaishuang@cmss.chinamobile.com>
Acked-by: NWilliam Tu <u9012063@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Let RED utilize the new internal flag, TCQ_F_OFFLOADED,
to mark a given qdisc as offloaded instead of using a dedicated
indication.

Also, change internal logic into looking at said flag when possible.

Fixes: 602f3baf ("net_sch: red: Add offload ability to RED qdisc")
Signed-off-by: NYuval Mintz <yuvalm@mellanox.com>
Acked-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Qdiscs can be offloaded to HW, but current implementation isn't uniform.
Instead, qdiscs either pass information about offload status via their
TCA_OPTIONS or omit it altogether.

Introduce a new attribute - TCA_HW_OFFLOAD that would form a uniform
uAPI for the offloading status of qdiscs.
Signed-off-by: NYuval Mintz <yuvalm@mellanox.com>
Acked-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

skb_complete_tx_timestamp must ingest the skb it is passed. Call
kfree_skb if the skb cannot be enqueued.

Fixes: b245be1f ("net-timestamp: no-payload only sysctl")
Fixes: 9ac25fc0 ("net: fix socket refcounting in skb_complete_tx_timestamp()")
Reported-by: NRichard Cochran <richardcochran@gmail.com>
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Reviewed-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>