The theory was that we would need to grab the socket lock anyway, so we
might as well use it to gate the allocation of RPC slots for a TCP
socket.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Rather than forcing each and every RPC task to grab the socket write
lock in order to send itself, we allow whichever task is holding the
write lock to attempt to drain the entire transmit queue.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Avoid memory starvation by giving RPCs that are tagged with the
RPC_TASK_SWAPPER flag the highest priority.
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>
One of the intentions with the priority queues was to ensure that no
single process can hog the transport. The field task->tk_owner therefore
identifies the RPC call's origin, and is intended to allow the RPC layer
to organise queues for fairness.
This commit therefore modifies the transmit queue to group requests
by task->tk_owner, and ensures that we round robin among those groups.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Remove the checks for whether or not we need to transmit, and whether
or not a reply has been received. Those are already handled in
call_transmit() itself.
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
When we shift to using the transmit queue, then the task that holds the
write lock will not necessarily be the same as the one being transmitted.
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>
When storing a struct rpc_rqst on the slot allocation list, we currently
use the same field 'rq_list' as we use to store the request on the
receive queue. Since the structure is never on both lists at the same
time, this is OK.
However, for clarity, let's make that a union with different names for
the different lists so that we can more easily distinguish between
the two states.
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>
Rather than waking up the entire queue of RPC messages a second time,
just wake up the task that was put to sleep.
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>
Remove trailing whitespace and blank line at EOF
Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Krzysztof Kozlowski <krzk@kernel.org> reports that a heavy NFSv4
WRITE workload against a slow NFS server causes his Raspberry Pi
clients to stall. Krzysztof bisected it to commit 37ac86c3a7
("SUNRPC: Initialize rpc_rqst outside of xprt->reserve_lock") .
I was able to reproduce similar behavior and it appears that rarely
the RPC client layer is re-allocating an XID for an RPC that it has
already partially sent. This results in the client ignoring the
subsequent reply, which carries the original XID.
For various reasons, checking !req->rq_xmit_bytes_sent in
xprt_prepare_transmit is not a 100% reliable mechanism for
determining when a fresh XID is needed.
Trond's preference is to allocate the XID at the time each rpc_rqst
slot is initialized.
This patch should also address a gcc 4.1.2 complaint reported by
Geert Uytterhoeven <geert@linux-m68k.org>.
Reported-by: Krzysztof Kozlowski <krzk@kernel.org>
Fixes: 37ac86c3a7 ("SUNRPC: Initialize rpc_rqst outside of ... ")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Krzysztof Kozlowski <krzk@kernel.org>
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Refactor: xprtrdma needs to have better control over when RPCs are
awoken from the backlog queue, so replace xprt_free_slot with a
transport op callout.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
alloc_slot is a transport-specific op, but initializing an rpc_rqst
is common to all transports. In addition, the only part of initial-
izing an rpc_rqst that needs serialization is getting a fresh XID.
Move rpc_rqst initialization to common code in preparation for
adding a transport-specific alloc_slot to xprtrdma.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
If recording xprt->stat.max_slots is moved into xprt_alloc_slot,
then xprt->num_reqs is never manipulated outside
xprt->reserve_lock. There's no longer a need for xprt->num_reqs to
be atomic.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Some RPC transports have more overhead in their send_request
callouts than others. For example, for RPC-over-RDMA:
- Marshaling an RPC often has to DMA map the RPC arguments
- Registration methods perform memory registration as part of
marshaling
To capture just server and network latencies more precisely: when
sending a Call, capture the rq_xtime timestamp _after_ the transport
header has been marshaled.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Some RPC transports have more overhead in their reply handlers
than others. For example, for RPC-over-RDMA:
- RPC completion has to wait for memory invalidation, which is
not a part of the server/network round trip
- Recently a context switch was introduced into the reply handler,
which further artificially inflates the measure of RPC RTT
To capture just server and network latencies more precisely: when
receiving a reply, compute the RTT as soon as the XID is recognized
rather than at RPC completion time.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Since commit 33849792cb ("xprtrdma: Detect unreachable NFS/RDMA
servers more reliably"), the xprtrdma transport now has a ->timer
callout. But xprtrdma does not need to compute RTT data, only UDP
needs that. Move the xprt_update_rtt call into the UDP transport
implementation.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The response to a write_space notification is very latency sensitive,
so we should queue it to the lower latency xprtiod_workqueue. This
is something we already do for the other cases where an rpc task
holds the transport XPRT_LOCKED bitlock.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Track RPC timeouts: report the XID and the server address to match
the content of network capture.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Stable bugfixes:
- NFS: Avoid a BUG_ON() in nfs_commit_inode() by not waiting for a
commit in the case that there were no commit requests.
- SUNRPC: Fix a race in the receive code path
Other fixes:
- NFS: Fix a deadlock in nfs client initialization
- xprtrdma: Fix a performance regression for small IOs
-----BEGIN PGP SIGNATURE-----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=XhEC
-----END PGP SIGNATURE-----
Merge tag 'nfs-for-4.15-3' of git://git.linux-nfs.org/projects/anna/linux-nfs
Pull NFS client fixes from Anna Schumaker:
"This has two stable bugfixes, one to fix a BUG_ON() when
nfs_commit_inode() is called with no outstanding commit requests and
another to fix a race in the SUNRPC receive codepath.
Additionally, there are also fixes for an NFS client deadlock and an
xprtrdma performance regression.
Summary:
Stable bugfixes:
- NFS: Avoid a BUG_ON() in nfs_commit_inode() by not waiting for a
commit in the case that there were no commit requests.
- SUNRPC: Fix a race in the receive code path
Other fixes:
- NFS: Fix a deadlock in nfs client initialization
- xprtrdma: Fix a performance regression for small IOs"
* tag 'nfs-for-4.15-3' of git://git.linux-nfs.org/projects/anna/linux-nfs:
SUNRPC: Fix a race in the receive code path
nfs: don't wait on commit in nfs_commit_inode() if there were no commit requests
xprtrdma: Spread reply processing over more CPUs
nfs: fix a deadlock in nfs client initialization
We must ensure that the call to rpc_sleep_on() in xprt_transmit() cannot
race with the call to xprt_complete_rqst().
Reported-by: Chuck Lever <chuck.lever@oracle.com>
Link: https://bugzilla.linux-nfs.org/show_bug.cgi?id=317
Fixes: ce7c252a8c ("SUNRPC: Add a separate spinlock to protect..")
Cc: stable@vger.kernel.org # 4.14+
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Stable bugfixes:
- Revalidate "." and ".." correctly on open
- Avoid RCU usage in tracepoints
- Fix ugly referral attributes
- Fix a typo in nomigration mount option
- Revert "NFS: Move the flock open mode check into nfs_flock()"
Features:
- Implement a stronger send queue accounting system for NFS over RDMA
- Switch some atomics to the new refcount_t type
Other bugfixes and cleanups:
- Clean up access mode bits
- Remove special-case revalidations in nfs_opendir()
- Improve invalidating NFS over RDMA memory for async operations that time out
- Handle NFS over RDMA replies with a worqueue
- Handle NFS over RDMA sends with a workqueue
- Fix up replaying interrupted requests
- Remove dead NFS over RDMA definitions
- Update NFS over RDMA copyright information
- Be more consistent with bool initialization and comparisons
- Mark expected switch fall throughs
- Various sunrpc tracepoint cleanups
- Fix various OPEN races
- Fix a typo in nfs_rename()
- Use common error handling code in nfs_lock_and_join_request()
- Check that some structures are properly cleaned up during net_exit()
- Remove net pointer from dprintk()s
-----BEGIN PGP SIGNATURE-----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=2Zc1
-----END PGP SIGNATURE-----
Merge tag 'nfs-for-4.15-1' of git://git.linux-nfs.org/projects/anna/linux-nfs
Pull NFS client updates from Anna Schumaker:
"Stable bugfixes:
- Revalidate "." and ".." correctly on open
- Avoid RCU usage in tracepoints
- Fix ugly referral attributes
- Fix a typo in nomigration mount option
- Revert "NFS: Move the flock open mode check into nfs_flock()"
Features:
- Implement a stronger send queue accounting system for NFS over RDMA
- Switch some atomics to the new refcount_t type
Other bugfixes and cleanups:
- Clean up access mode bits
- Remove special-case revalidations in nfs_opendir()
- Improve invalidating NFS over RDMA memory for async operations that
time out
- Handle NFS over RDMA replies with a worqueue
- Handle NFS over RDMA sends with a workqueue
- Fix up replaying interrupted requests
- Remove dead NFS over RDMA definitions
- Update NFS over RDMA copyright information
- Be more consistent with bool initialization and comparisons
- Mark expected switch fall throughs
- Various sunrpc tracepoint cleanups
- Fix various OPEN races
- Fix a typo in nfs_rename()
- Use common error handling code in nfs_lock_and_join_request()
- Check that some structures are properly cleaned up during
net_exit()
- Remove net pointer from dprintk()s"
* tag 'nfs-for-4.15-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (62 commits)
NFS: Revert "NFS: Move the flock open mode check into nfs_flock()"
NFS: Fix typo in nomigration mount option
nfs: Fix ugly referral attributes
NFS: super: mark expected switch fall-throughs
sunrpc: remove net pointer from messages
nfs: remove net pointer from messages
sunrpc: exit_net cleanup check added
nfs client: exit_net cleanup check added
nfs/write: Use common error handling code in nfs_lock_and_join_requests()
NFSv4: Replace closed stateids with the "invalid special stateid"
NFSv4: nfs_set_open_stateid must not trigger state recovery for closed state
NFSv4: Check the open stateid when searching for expired state
NFSv4: Clean up nfs4_delegreturn_done
NFSv4: cleanup nfs4_close_done
NFSv4: Retry NFS4ERR_OLD_STATEID errors in layoutreturn
pNFS: Retry NFS4ERR_OLD_STATEID errors in layoutreturn-on-close
NFSv4: Don't try to CLOSE if the stateid 'other' field has changed
NFSv4: Retry CLOSE and DELEGRETURN on NFS4ERR_OLD_STATEID.
NFS: Fix a typo in nfs_rename()
NFSv4: Fix open create exclusive when the server reboots
...
In preparation to enabling -Wimplicit-fallthrough, mark switch cases
where we are expecting to fall through.
Signed-off-by: Gustavo A. R. Silva <garsilva@embeddedor.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Several conflicts here.
NFP driver bug fix adding nfp_netdev_is_nfp_repr() check to
nfp_fl_output() needed some adjustments because the code block is in
an else block now.
Parallel additions to net/pkt_cls.h and net/sch_generic.h
A bug fix in __tcp_retransmit_skb() conflicted with some of
the rbtree changes in net-next.
The tc action RCU callback fixes in 'net' had some overlap with some
of the recent tcf_block reworking.
Signed-off-by: David S. Miller <davem@davemloft.net>
The transport may need to flush transport connect and receive tasks
that are running on rpciod. In order to do so safely, we need to
ensure that the caller of cancel_work_sync() etc is not itself
running on rpciod.
Do so by running the destroy task from the system workqueue.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly.
Cc: Trond Myklebust <trond.myklebust@primarydata.com>
Cc: Anna Schumaker <anna.schumaker@netapp.com>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Cc: Jeff Layton <jlayton@poochiereds.net>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-nfs@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
We remove the request from the receive list before we call
xprt_wait_on_pinned_rqst(), and so we need to use list_del_init().
Otherwise, we will see list corruption when xprt_complete_rqst()
is called.
Reported-by: Emre Celebi <emre@primarydata.com>
Fixes: ce7c252a8c ("SUNRPC: Add a separate spinlock to protect...")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Adopt the use of xprt_pin_rqst to eliminate contention between
Call-side users of rb_lock and the use of rb_lock in
rpcrdma_reply_handler.
This replaces the mechanism introduced in 431af645cf ("xprtrdma:
Fix client lock-up after application signal fires").
Use recv_lock to quickly find the completing rqst, pin it, then
drop the lock. At that point invalidation and pull-up of the Reply
XDR can be done. Both are often expensive operations.
Finally, take recv_lock again to signal completion to the RPC
layer. It also protects adjustment of "cwnd".
This greatly reduces the amount of time a lock is held by the
reply handler. Comparing lock_stat results shows a marked decrease
in contention on rb_lock and recv_lock.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
[trond.myklebust@primarydata.com: Remove call to rpcrdma_buffer_put() from
the "out_norqst:" path in rpcrdma_reply_handler.]
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
This further reduces contention with the transport_lock, and allows us
to convert to using a non-bh-safe spinlock, since the list is now never
accessed from a bh context.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Instead add a mechanism to ensure that the request doesn't disappear
from underneath us while copying from the socket. We do this by
preventing xprt_release() from freeing the XDR buffers until the
flag RPC_TASK_MSG_RECV has been cleared from the request.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
In xprt_alloc_slot(), the spin lock is only needed to provide atomicity
between the atomic_add_unless() failure and the call to xprt_add_backlog().
We do not actually need to hold it across the memory allocation itself.
By dropping the lock, we can use a more resilient GFP_NOFS allocation,
just as we now do in the rest of the RPC client code.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
xprt_force_disconnect() is already invoked from the socket
transport. I want to invoke xprt_force_disconnect() from the
RPC-over-RDMA transport, which is a separate module from sunrpc.ko.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
The transport lock is needed to protect the xprt_adjust_cwnd() call
in xs_udp_timer, but it is not necessary for accessing the
rq_reply_bytes_recvd or tk_status fields. It is correct to sublimate
the lock into UDP's xs_udp_timer method, where it is required.
The ->timer method has to take the transport lock if needed, but it
can now sleep safely, or even call back into the RPC scheduler.
This is more a clean-up than a fix, but the "issue" was introduced
by my transport switch patches back in 2005.
Fixes: 46c0ee8bc4 ("RPC: separate xprt_timer implementations")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
xs_connect() contains an exponential backoff mechanism so the repeated
connection attempts are delayed by longer and longer amounts.
This is appropriate when the connection failed due to a timeout, but
it not appropriate when a definitive "no" answer is received. In such
cases, call_connect_status() imposes a minimum 3-second back-off, so
not having the exponetial back-off will never result in immediate
retries.
The current situation is a problem when the NFS server tries to
register with rpcbind but rpcbind isn't running. All connection
attempts are made on the same "xprt" and as the connection is never
"closed", the exponential back delays successive attempts to register,
or de-register, different protocols. This results in a multi-minute
delay with no benefit.
So, when call_connect_status() receives a definitive "no", use
xprt_conditional_disconnect() to cancel the previous connection attempt.
This will set XPRT_CLOSE_WAIT so that xprt->ops->close() calls xs_close()
which resets the reestablish_timeout.
To ensure xprt_conditional_disconnect() does the right thing, we
ensure that rq_connect_cookie is set before a connection attempt, and
allow xprt_conditional_disconnect() to complete even when the
transport is not fully connected.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
xprtrdma needs to allocate the Call and Reply buffers separately.
TBH, the reliance on using a single buffer for the pair of XDR
buffers is transport implementation-specific.
Instead of passing just the rq_buffer into the buf_free method, pass
the task structure and let buf_free take care of freeing both
XDR buffers at once.
There's a micro-optimization here. In the common case, both
xprt_release and the transport's buf_free method were checking if
rq_buffer was NULL. Now the check is done only once per RPC.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Anna Schumaker <Anna.Schumaker@Netapp.com>
Ensure that we don't forget to set up the disconnection timer for the
case when a connect request is fulfilled after the RPC request that
initiated it has timed out or been interrupted.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Use the low latency transport workqueue to process the task that is
next in line on the xprt->sending queue.
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Trond Myklebust