This code is never called from interrupt context; it's always run by either
a user thread or rpciod. So KM_SKB_SUNRPC_DATA is inappropriate here.
Thanks to Aimé Le Rouzic for capturing an oops which showed the kernel
taking an interrupt while we were in this piece of code, resulting in a
nested kmap_atomic(.,KM_SKB_SUNRPC_DATA) call from
xdr_partial_copy_from_skb().
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Dumping all this data to the logs is wasteful (even when debugging is turned
off), and creates too much output to be useful when it's turned on.
Fix a minor style bug or two while we're at it.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Don't wake up bind waiters if a task finds that another task is already
trying to bind.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Change the location where the rpc_xprt structure is allocated so each
transport implementation can allocate a private area from the same
chunk of memory.
Note also that xprt->ops->destroy, rather than xprt_destroy, is now
responsible for freeing rpc_xprt when the transport is destroyed.
Test plan:
Connectathon.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
All internal RPC client operations should no longer depend on the BKL,
however lockd and NFS callbacks may still require it.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Use RCU to ensure that we can safely call rpc_finish_wakeup after we've
called __rpc_do_wake_up_task. If not, there is a theoretical race, in which
the rpc_task finishes executing, and gets freed first.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The sunrpc scheduler contains a race condition that can let an RPC
task end up being neither running nor on any wait queue. The race takes
place between rpc_make_runnable (called from rpc_wake_up_task) and
__rpc_execute under the following condition:
First __rpc_execute calls tk_action which puts the task on some wait
queue. The task is dequeued by another process before __rpc_execute
continues its execution. While executing rpc_make_runnable exactly after
setting the task `running' bit and before clearing the `queued' bit
__rpc_execute picks up execution, clears `running' and subsequently
both functions fall through, both under the false assumption somebody
else took the job.
Swapping rpc_test_and_set_running with rpc_clear_queued in
rpc_make_runnable fixes that hole. This introduces another possible
race condition that can be handled by checking for `queued' after
setting the `running' bit.
Bug noticed on a 4-way x86_64 system under XEN with an NFSv4 server
on the same physical machine, apparently one of the few ways to hit
this race condition at all.
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Christophe Saout <christophe@saout.de>
Signed-off-by: Trond Myklebust <trond.myklebust@fys.uio.no>
auth_domain_put() forgot to unlock acquired spinlock.
Cc: Olaf Kirch <okir@monad.swb.de>
Cc: Andy Adamson <andros@citi.umich.edu>
Cc: J. Bruce Fields <bfields@citi.umich.edu>
Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Neil Brown <neilb@suse.de>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
A recent patch fixed a problem which would occur when the refcount on an
auth_domain reached zero. This problem has not been reported in practice
despite existing in two major kernel releases because the refcount can
never reach zero.
This patch fixes the problems that stop the refcount reaching zero.
1/ We were adding to the refcount when inserting in the hash table,
but only removing from the hashtable when the refcount reached zero.
Obviously it never would. So don't count the implied reference of
being in the hash table.
2/ There are two paths on which a socket can be destroyed. One called
svcauth_unix_info_release(). The other didn't. So when the other was
taken, we can lose a reference to an ip_map which in-turn holds a
reference to an auth_domain
So unify the exit paths into svc_sock_put. This highlights the fact
that svc_delete_socket has slightly odd semantics - it does not drop
a reference but probably should. Fixing this need a bit more
thought and testing.
Signed-off-by: Neil Brown <neilb@suse.de>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This patch is suitable for just about any 2.6 kernel. It should go in
2.6.19 and 2.6.18.2 and possible even the .17 and .16 stable series.
This is a long standing bug that seems to have only recently become
apparent, presumably due to increasing use of NFS over TCP - many
distros seem to be making it the default.
The SK_CONN bit gets set when a listening socket may be ready
for an accept, just as SK_DATA is set when data may be available.
It is entirely possible for svc_tcp_accept to be called with neither
of these set. It doesn't happen often but there is a small race in
svc_sock_enqueue as SK_CONN and SK_DATA are tested outside the
spin_lock. They could be cleared immediately after the test and
before the lock is gained.
This normally shouldn't be a problem. The sockets are non-blocking so
trying to read() or accept() when ther is nothing to do is not a problem.
However: svc_tcp_recvfrom makes the decision "Should I accept() or
should I read()" based on whether SK_CONN is set or not. This usually
works but is not safe. The decision should be based on whether it is
a TCP_LISTEN socket or a TCP_CONNECTED socket.
Signed-off-by: Neil Brown <neilb@suse.de>
Cc: Adrian Bunk <bunk@stusta.de>
Cc: <stable@kernel.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Yes, this actually passed tests the way it was.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When submitting a request to a fast portmapper (such as the local rpcbind
daemon), the request can complete before the parent task is even queued up on
xprt->binding. Fix this by queuing before submitting the rpcbind request.
Test plan:
Connectathon locking test with UDP.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It is possible for the ->fopen callback from lockd into nfsd to find that an
answer cannot be given straight away (an upcall is needed) and so the request
has to be 'dropped', to be retried later. That error status is not currently
propagated back.
So:
Change nlm_fopen to return nlm error codes (rather than a private
protocol) and define a new nlm_drop_reply code.
Cause nlm_drop_reply to cause the rpc request to get rpc_drop_reply
when this error comes back.
Cause svc_process to drop a request which returns a status of
rpc_drop_reply.
[akpm@osdl.org: fix warning storm]
Cc: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
There is some confusion about the meaning of 'bufsz' for a sunrpc server.
In some cases it is the largest message that can be sent or received. In
other cases it is the largest 'payload' that can be included in a NFS
message.
In either case, it is not possible for both the request and the reply to be
this large. One of the request or reply may only be one page long, which
fits nicely with NFS.
So we remove 'bufsz' and replace it with two numbers: 'max_payload' and
'max_mesg'. Max_payload is the size that the server requests. It is used
by the server to check the max size allowed on a particular connection:
depending on the protocol a lower limit might be used.
max_mesg is the largest single message that can be sent or received. It is
calculated as the max_payload, rounded up to a multiple of PAGE_SIZE, and
with PAGE_SIZE added to overhead. Only one of the request and reply may be
this size. The other must be at most one page.
Cc: Greg Banks <gnb@sgi.com>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The rpc reply has multiple levels of error returns. The code here contributes
to the confusion by using "accept_statp" for a pointer to what the rfc (and
wireshark, etc.) refer to as the "reply_stat". (The confusion is compounded
by the fact that the rfc also has an "accept_stat" which follows the
reply_stat in the succesful case.)
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
If the request is denied after gss_accept was called, we shouldn't try to wrap
the reply. We were checking the accept_stat but not the reply_stat.
To check the reply_stat in _release, we need a pointer to before (rather than
after) the verifier, so modify body_start appropriately.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Factor out some common code from the integrity and privacy cases.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The NFSACL patches introduced support for multiple RPC services listening on
the same transport. However, only the first of these services was registered
with portmapper. This was perfectly fine for nfsacl, as you traditionally do
not want these to show up in a portmapper listing.
The patch below changes the default behavior to always register all services
listening on a given transport, but retains the old behavior for nfsacl
services.
Signed-off-by: Olaf Kirch <okir@suse.de>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Speed up high call-rate workloads by caching the struct ip_map for the peer on
the connected struct svc_sock instead of looking it up in the ip_map cache
hashtable on every call. This helps workloads using AUTH_SYS authentication
over TCP.
Testing was on a 4 CPU 4 NIC Altix using 4 IRIX clients, each with 16
synthetic client threads simulating an rsync (i.e. recursive directory
listing) workload reading from an i386 RH9 install image (161480 regular files
in 10841 directories) on the server. That tree is small enough to fill in the
server's RAM so no disk traffic was involved. This setup gives a sustained
call rate in excess of 60000 calls/sec before being CPU-bound on the server.
Profiling showed strcmp(), called from ip_map_match(), was taking 4.8% of each
CPU, and ip_map_lookup() was taking 2.9%. This patch drops both contribution
into the profile noise.
Note that the above result overstates this value of this patch for most
workloads. The synthetic clients are all using separate IP addresses, so
there are 64 entries in the ip_map cache hash. Because the kernel measured
contained the bug fixed in commit
commit 1f1e030bf7
and was running on 64bit little-endian machine, probably all of those 64
entries were on a single chain, thus increasing the cost of ip_map_lookup().
With a modern kernel you would need more clients to see the same amount of
performance improvement. This patch has helped to scale knfsd to handle a
deployment with 2000 NFS clients.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The limit over UDP remains at 32K. Also, make some of the apparently
arbitrary sizing constants clearer.
The biggest change here involves replacing NFSSVC_MAXBLKSIZE by a function of
the rqstp. This allows it to be different for different protocols (udp/tcp)
and also allows it to depend on the servers declared sv_bufsiz.
Note that we don't actually increase sv_bufsz for nfs yet. That comes next.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
.. by allocating the array of 'kvec' in 'struct svc_rqst'.
As we plan to increase RPCSVC_MAXPAGES from 8 upto 256, we can no longer
allocate an array of this size on the stack. So we allocate it in 'struct
svc_rqst'.
However svc_rqst contains (indirectly) an array of the same type and size
(actually several, but they are in a union). So rather than waste space, we
move those arrays out of the separately allocated union and into svc_rqst to
share with the kvec moved out of svc_tcp_recvfrom (various arrays are used at
different times, so there is no conflict).
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
We are planning to increase RPCSVC_MAXPAGES from about 8 to about 256. This
means we need to be a bit careful about arrays of size RPCSVC_MAXPAGES.
struct svc_rqst contains two such arrays. However the there are never more
that RPCSVC_MAXPAGES pages in the two arrays together, so only one array is
needed.
The two arrays are for the pages holding the request, and the pages holding
the reply. Instead of two arrays, we can simply keep an index into where the
first reply page is.
This patch also removes a number of small inline functions that probably
server to obscure what is going on rather than clarify it, and opencode the
needed functionality.
Also remove the 'rq_restailpage' variable as it is *always* 0. i.e. if the
response 'xdr' structure has a non-empty tail it is always in the same pages
as the head.
check counters are initilised and incr properly
check for consistant usage of ++ etc
maybe extra some inlines for common approach
general review
Signed-off-by: Neil Brown <neilb@suse.de>
Cc: Magnus Maatta <novell@kiruna.se>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Arrgg.. We cannot 'lockd_up' before 'svc_addsock' as we don't know the
protocol yet.... So switch it around again and save the name of the created
sockets so that it can be closed if lock_up fails.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The refcount that nfsd holds on lockd is based on the number of open sockets.
So when we close a socket, we should decrement the ref (with lockd_down).
Currently when a socket is closed via writing to the portlist file, that
doesn't happen.
So: make sure we get an error return if the socket that was requested does is
not found, and call lockd_down if it was.
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Many files include the filename at the beginning, serveral used a wrong one.
Signed-off-by: Uwe Zeisberger <Uwe_Zeisberger@digi.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Replace references to system_utsname to the per-process uts namespace
where appropriate. This includes things like uname.
Changes: Per Eric Biederman's comments, use the per-process uts namespace
for ELF_PLATFORM, sunrpc, and parts of net/ipv4/ipconfig.c
[jdike@addtoit.com: UML fix]
[clg@fr.ibm.com: cleanup]
[akpm@osdl.org: build fix]
Signed-off-by: Serge E. Hallyn <serue@us.ibm.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Andrey Savochkin <saw@sw.ru>
Signed-off-by: Cedric Le Goater <clg@fr.ibm.com>
Cc: Jeff Dike <jdike@addtoit.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Actually implement multiple pools. On NUMA machines, allocate a svc_pool per
NUMA node; on SMP a svc_pool per CPU; otherwise a single global pool. Enqueue
sockets on the svc_pool corresponding to the CPU on which the socket bh is run
(i.e. the NIC interrupt CPU). Threads have their cpu mask set to limit them
to the CPUs in the svc_pool that owns them.
This is the patch that allows an Altix to scale NFS traffic linearly
beyond 4 CPUs and 4 NICs.
Incorporates changes and feedback from Neil Brown, Trond Myklebust, and
Christoph Hellwig.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Currently knfsd keeps its own list of all nfsd threads in nfssvc.c; add a new
way of managing the list of all threads in a svc_serv. Add
svc_create_pooled() to allow creation of a svc_serv whose threads are managed
by the sunrpc code. Add svc_set_num_threads() to manage the number of threads
in a service, either per-pool or globally across the service.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Split out the list of idle threads and pending sockets from svc_serv into a
new svc_pool structure, and allocate a fixed number (in this patch, 1) of
pools per svc_serv. The new structure contains a lock which takes over
several of the duties of svc_serv->sv_lock, which is now relegated to
protecting only sv_tempsocks, sv_permsocks, and sv_tmpcnt in svc_serv.
The point is to move the hottest fields out of svc_serv and into svc_pool,
allowing a following patch to arrange for a svc_pool per NUMA node or per CPU.
This is a major step towards making the NFS server NUMA-friendly.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
The SK_BUSY bit in svc_sock->sk_flags ensures that we do not attempt to
enqueue a socket twice. Currently, setting and clearing the bit is protected
by svc_serv->sv_lock. As I intend to reduce the data that the lock protects
so it's not held when svc_sock_enqueue() tests and sets SK_BUSY, that test and
set needs to be atomic.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Convert the svc_sock->sk_reserved variable from an int protected by
svc_serv->sv_lock, to an atomic. This reduces (by 1) the number of places we
need to take the (effectively global) svc_serv->sv_lock.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Protect the svc_sock->sk_deferred list with a new lock svc_sock->sk_defer_lock
instead of svc_serv->sv_lock. Using the more fine-grained lock reduces the
number of places we need to take the svc_serv lock.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Convert the svc_sock->sk_inuse counter from an int protected by
svc_serv->sv_lock, to an atomic. This reduces the number of places we need to
take the (effectively global) svc_serv->sv_lock.
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Following are 11 patches from Greg Banks which combine to make knfsd more
Numa-aware. They reduce hitting on 'global' data structures, and create some
data-structures that can be node-local.
knfsd threads are bound to a particular node, and the thread to handle a new
request is chosen from the threads that are attach to the node that received
the interrupt.
The distribution of threads across nodes can be controlled by a new file in
the 'nfsd' filesystem, though the default approach of an even spread is
probably fine for most sites.
Some (old) numbers that show the efficacy of these patches: N == number of
NICs == number of CPUs == nmber of clients. Number of NUMA nodes == N/2
N Throughput, MiB/s CPU usage, % (max=N*100)
Before After Before After
--- ------ ---- ----- -----
4 312 435 350 228
6 500 656 501 418
8 562 804 690 589
This patch:
Move the aging of RPC/TCP connection sockets from the main svc_recv() loop to
a timer which uses a mark-and-sweep algorithm every 6 minutes. This reduces
the amount of work that needs to be done in the main RPC loop and the length
of time we need to hold the (effectively global) svc_serv->sv_lock.
[akpm@osdl.org: cleanup]
Signed-off-by: Greg Banks <gnb@melbourne.sgi.com>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
It isn't needed as it is available in rqstp->rq_server, and dropping it allows
some local vars to be dropped.
[akpm@osdl.org: build fix]
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Userspace should create and bind a socket (but not connectted) and write the
'fd' to portlist. This will cause the nfs server to listen on that socket.
To close a socket, the name of the socket - as read from 'portlist' can be
written to 'portlist' with a preceding '-'.
Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
