There's no need to be fooling with XPT_BUSY now that all the threads
are gone.
The list_del_init() here could execute at the same time as the
svc_xprt_enqueue()'s list_add_tail(), with undefined results. We don't
really care at this point, but it might result in a spurious
list-corruption warning or something.
And svc_close() isn't adding any value; just call svc_delete_xprt()
directly.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Follow up on b48fa6b991 by moving all the
svc_xprt_received() calls for the main xprt to one place. The clearing
of XPT_BUSY here is critical to the correctness of the server, so I'd
prefer it to be obvious where we do it.
The only substantive result is moving svc_xprt_received() after
svc_receive_deferred(). Other than a (likely insignificant) delay
waking up the next thread, that should be harmless.
Also reshuffle the exit code a little to skip a few other steps that we
don't care about the in the svc_delete_xprt() case.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
There's no harm to doing this, since the only caller will immediately
call svc_enqueue() afterwards, ensuring we don't miss the remaining
deferred requests just because XPT_DEFERRED was briefly cleared.
But why not just do this the simple way?
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
The big kernel lock has been removed from all these files at some point,
leaving only the #include.
Remove this too as a cleanup.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If any xprt marked DEAD is also left BUSY for the rest of its life, then
the XPT_DEAD check here is superfluous--we'll get the same result from
the XPT_BUSY check just after.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
As long as DEAD exports are left BUSY, and svc_delete_xprt is called
only with BUSY held, then svc_delete_xprt() will never be called on an
xprt that is already DEAD.
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Once an xprt has been deleted, there's no reason to allow it to be
enqueued--at worst, that might cause the xprt to be re-added to some
global list, resulting in later corruption.
Also, note this leaves us with no need for the reference-count
manipulation here.
Reviewed-by: NeilBrown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Saves some lines of code and some branticks when reading one.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
NFSv4.1 needs warning when a client tcp connection goes down, if that
connection is being used as a backchannel, so that it can warn the
client that it has lost the backchannel connection.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
After this the socket creation in it knows the context.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
The transport representation should be per-net of course.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
This is done in order to facilitate getting the ip_map_cache from
which to put the ip_map.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
The current practice of waiting for cache updates by queueing the
whole request to be retried has (at least) two problems.
1/ With NFSv4, requests can be quite complex and re-trying a whole
request when a later part fails should only be a last-resort, not a
normal practice.
2/ Large requests, and in particular any 'write' request, will not be
queued by the current code and doing so would be undesirable.
In many cases only a very sort wait is needed before the cache gets
valid data.
So, providing the underlying transport permits it by setting
->thread_wait,
arrange to wait briefly for an upcall to be completed (as reflected in
the clearing of CACHE_PENDING).
If the short wait was not long enough and CACHE_PENDING is still set,
fall back on the old approach.
The 'thread_wait' value is set to 5 seconds when there are spare
threads, and 1 second when there are no spare threads.
These values are probably much higher than needed, but will ensure
some forward progress.
Note that as we only request an update for a non-valid item, and as
non-valid items are updated in place it is extremely unlikely that
cache_check will return -ETIMEDOUT. Normally cache_defer_req will
sleep for a short while and then find that the item is_valid.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
svc_xprt_received must be called when ->xpo_recvfrom has finished
receiving a message, so that the XPT_BUSY flag will be cleared and
if necessary, requeued for further work.
This call is currently made in each ->xpo_recvfrom function, often
from multiple different points. In each case it is the earliest point
on a particular path where it is known that the protection provided by
XPT_BUSY is no longer needed.
However there are (still) some error paths which do not call
svc_xprt_received, and requiring each ->xpo_recvfrom to make the call
does not encourage robustness.
So: move the svc_xprt_received call to be made just after the
call to ->xpo_recvfrom(), and move it of the various ->xpo_recvfrom
methods.
This means that it may not be called at the earliest possible instant,
but this is unlikely to be a measurable performance issue.
Note that there are still other calls to svc_xprt_received as it is
also needed when an xprt is newly created.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
svc_xprt_put() can call tcp_close(), which can sleep, so we shouldn't be
holding this lock.
In fact, only the xpt_list removal and the sv_tmpcnt decrement should
need the sv_lock here.
Reported-by: Mi Jinlong <mijinlong@cn.fujitsu.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
This reverts commit b0401d7253, which
moved svc_delete_xprt() outside of XPT_BUSY, and allowed it to be called
after svc_xpt_recived(), removing its last reference and destroying it
after it had already been queued for future processing.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
This reverts commit b292cf9ce7. The
commit that it attempted to patch up,
b0401d7253, was fundamentally wrong, and
will also be reverted.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The 'struct svc_deferred_req's on the xpt_deferred queue do not
own a reference to the owning xprt. This is seen in svc_revisit
which is where things are added to this queue. dr->xprt is set to
NULL and the reference to the xprt it put.
So when this list is cleaned up in svc_delete_xprt, we mustn't
put the reference.
Also, replace the 'for' with a 'while' which is arguably
simpler and more likely to compile efficiently.
Cc: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
write_ports() converts svc_create_xprt()'s ENOENT error return to
EPROTONOSUPPORT so that rpc.nfsd (in user space) can report an error
message that makes sense.
It turns out that several of the other kernel APIs rpc.nfsd use can
also return ENOENT from svc_create_xprt(), by way of lockd_up().
On the client side, an NFSv2 or NFSv3 mount request can also return
the result of lockd_up(). This error may also be returned during an
NFSv4 mount request, since the NFSv4 callback service uses
svc_create_xprt() to create the callback listener. An ENOENT error
return results in a confusing error message from the mount command.
Let's have svc_create_xprt() return EPROTONOSUPPORT instead of ENOENT.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Clean up: Bruce observed we have more or less common logic in each of
svc_create_xprt()'s callers: the check to create an IPv6 RPC listener
socket only if CONFIG_IPV6 is set. I'm about to add another case
that does just the same.
If we move the ifdefs into __svc_xpo_create(), then svc_create_xprt()
call sites can get rid of the "#ifdef" ugliness, and can use the same
logic with or without IPv6 support available in the kernel.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
* 'for-2.6.33' of git://linux-nfs.org/~bfields/linux:
sunrpc: fix peername failed on closed listener
nfsd: make sure data is on disk before calling ->fsync
nfsd: fix "insecure" export option
There're some warnings of "nfsd: peername failed (err 107)!"
socket error -107 means Transport endpoint is not connected.
This warning message was outputed by svc_tcp_accept() [net/sunrpc/svcsock.c],
when kernel_getpeername returns -107. This means socket might be CLOSED.
And svc_tcp_accept was called by svc_recv() [net/sunrpc/svc_xprt.c]
if (test_bit(XPT_LISTENER, &xprt->xpt_flags)) {
<snip>
newxpt = xprt->xpt_ops->xpo_accept(xprt);
<snip>
So this might happen when xprt->xpt_flags has both XPT_LISTENER and XPT_CLOSE.
Let's take a look at commit b0401d72, this commit has moved the close
processing after do recvfrom method, but this commit also introduces this
warnings, if the xpt_flags has both XPT_LISTENER and XPT_CLOSED, we should
close it, not accpet then close.
Signed-off-by: Xiaotian Feng <dfeng@redhat.com>
Cc: J. Bruce Fields <bfields@fieldses.org>
Cc: Neil Brown <neilb@suse.de>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
* 'for-2.6.33' of git://linux-nfs.org/~bfields/linux: (42 commits)
nfsd: remove pointless paths in file headers
nfsd: move most of nfsfh.h to fs/nfsd
nfsd: remove unused field rq_reffh
nfsd: enable V4ROOT exports
nfsd: make V4ROOT exports read-only
nfsd: restrict filehandles accepted in V4ROOT case
nfsd: allow exports of symlinks
nfsd: filter readdir results in V4ROOT case
nfsd: filter lookup results in V4ROOT case
nfsd4: don't continue "under" mounts in V4ROOT case
nfsd: introduce export flag for v4 pseudoroot
nfsd: let "insecure" flag vary by pseudoflavor
nfsd: new interface to advertise export features
nfsd: Move private headers to source directory
vfs: nfsctl.c un-used nfsd #includes
lockd: Remove un-used nfsd headers #includes
s390: remove un-used nfsd #includes
sparc: remove un-used nfsd #includes
parsic: remove un-used nfsd #includes
compat.c: Remove dependence on nfsd private headers
...
Not including net/atm/
Compiled tested x86 allyesconfig only
Added a > 80 column line or two, which I ignored.
Existing checkpatch plaints willfully, cheerfully ignored.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This reverts commit 59a252ff8c.
This helps in an entirely cached workload but not necessarily in
workloads that require waiting on disk.
Conflicts:
include/linux/sunrpc/svc.h
net/sunrpc/svc_xprt.c
Reported-by: Simon Kirby <sim@hostway.ca>
Tested-by: Jesper Krogh <jesper@krogh.cc>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
When the call direction is a reply, copy the xid and call direction into the
req->rq_private_buf.head[0].iov_base otherwise rpc_verify_header returns
rpc_garbage.
Signed-off-by: Rahul Iyer <iyer@netapp.com>
Signed-off-by: Mike Sager <sager@netapp.com>
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[get rid of CONFIG_NFSD_V4_1]
[sunrpc: refactoring of svc_tcp_recvfrom]
[nfsd41: sunrpc: create common send routine for the fore and the back channels]
[nfsd41: sunrpc: Use free_page() to free server backchannel pages]
[nfsd41: sunrpc: Document server backchannel locking]
[nfsd41: sunrpc: remove bc_connect_worker()]
[nfsd41: sunrpc: Define xprt_server_backchannel()[
[nfsd41: sunrpc: remove bc_close and bc_init_auto_disconnect dummy functions]
[nfsd41: sunrpc: eliminate unneeded switch statement in xs_setup_tcp()]
[nfsd41: sunrpc: Don't auto close the server backchannel connection]
[nfsd41: sunrpc: Remove unused functions]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfsd41: change bc_sock to bc_xprt]
[nfsd41: sunrpc: move struct rpc_buffer def into a common header file]
[nfsd41: sunrpc: use rpc_sleep in bc_send_request so not to block on mutex]
[removed cosmetic changes]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[sunrpc: add new xprt class for nfsv4.1 backchannel]
[sunrpc: v2.1 change handling of auto_close and init_auto_disconnect operations for the nfsv4.1 backchannel]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
[reverted more cosmetic leftovers]
[got rid of xprt_server_backchannel]
[separated "nfsd41: sunrpc: add new xprt class for nfsv4.1 backchannel"]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Cc: Trond Myklebust <trond.myklebust@netapp.com>
[sunrpc: change idle timeout value for the backchannel]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Acked-by: Trond Myklebust <trond.myklebust@netapp.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
sunrpc: "Move close processing to a single place"
(d7979ae4a0) moved the
close processing before the recvfrom method. This may
cause the close processing never to execute. So this
patch moves it to the right place.
Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
lock_kernel() in knfsd was replaced with a mutex. The later
commit 03cf6c9f49 ("knfsd:
add file to export stats about nfsd pools") did not follow
that change. This patch fixes the issue.
Also move the get and put of nfsd_serv to the open and close methods
(instead of start and stop methods) to allow atomic check and increment
of reference count in the open method (where we can still return an
error).
Signed-off-by: Ryusei Yamaguchi <mandel59@gmail.com>
Signed-off-by: Isaku Yamahata <yamahata@valinux.co.jp>
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Cc: Greg Banks <gnb@fmeh.org>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
* Remove smp_lock.h from files which don't need it (including some headers!)
* Add smp_lock.h to files which do need it
* Make smp_lock.h include conditional in hardirq.h
It's needed only for one kernel_locked() usage which is under CONFIG_PREEMPT
This will make hardirq.h inclusion cheaper for every PREEMPT=n config
(which includes allmodconfig/allyesconfig, BTW)
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The svc_xprt_names() function can overflow its buffer if it's so near
the end of the passed in buffer that the "name too long" string still
doesn't fit. Of course, it could never tell if it was near the end
of the passed in buffer, since its only caller passes in zero as the
buffer length.
Let's make this API a little safer.
Change svc_xprt_names() so it *always* checks for a buffer overflow,
and change its only caller to pass in the correct buffer length.
If svc_xprt_names() does overflow its buffer, it now fails with an
ENAMETOOLONG errno, instead of trying to write a message at the end
of the buffer. I don't like this much, but I can't figure out a clean
way that's always safe to return some of the names, *and* an
indication that the buffer was not long enough.
The displayed error when doing a 'cat /proc/fs/nfsd/portlist' is
"File name too long".
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Fix the following sparse warnings in net/sunrpc/svc_xprt.c.
warning: symbol 'svc_recv' was not declared. Should it be static?
warning: symbol 'svc_drop' was not declared. Should it be static?
warning: symbol 'svc_send' was not declared. Should it be static?
warning: symbol 'svc_close_all' was not declared. Should it be static?
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
* 'for-2.6.30' of git://linux-nfs.org/~bfields/linux: (81 commits)
nfsd41: define nfsd4_set_statp as noop for !CONFIG_NFSD_V4
nfsd41: define NFSD_DRC_SIZE_SHIFT in set_max_drc
nfsd41: Documentation/filesystems/nfs41-server.txt
nfsd41: CREATE_EXCLUSIVE4_1
nfsd41: SUPPATTR_EXCLCREAT attribute
nfsd41: support for 3-word long attribute bitmask
nfsd: dynamically skip encoded fattr bitmap in _nfsd4_verify
nfsd41: pass writable attrs mask to nfsd4_decode_fattr
nfsd41: provide support for minor version 1 at rpc level
nfsd41: control nfsv4.1 svc via /proc/fs/nfsd/versions
nfsd41: add OPEN4_SHARE_ACCESS_WANT nfs4_stateid bmap
nfsd41: access_valid
nfsd41: clientid handling
nfsd41: check encode size for sessions maxresponse cached
nfsd41: stateid handling
nfsd: pass nfsd4_compound_state* to nfs4_preprocess_{state,seq}id_op
nfsd41: destroy_session operation
nfsd41: non-page DRC for solo sequence responses
nfsd41: Add a create session replay cache
nfsd41: create_session operation
...
On an NFSv4.1 server cache miss that causes an upcall, NFS4ERR_DELAY will be
returned. It is up to the NFSv4.1 client to resend only the operations that
have not been processed.
Initialize rq_usedeferral to 1 in svc_process(). It sill be turned off in
nfsd4_proc_compound() only when NFSv4.1 Sessions are used.
Note: this isn't an adequate solution on its own. It's acceptable as a way
to get some minimal 4.1 up and working, but we're going to have to find a
way to avoid returning DELAY in all common cases before 4.1 can really be
considered ready.
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfsd41: reverse rq_nodeferral negative logic]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[sunrpc: initialize rq_usedeferral]
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The sv_family field is going away. Pass a protocol family argument to
svc_create_xprt() instead of extracting the family from the passed-in
svc_serv struct.
Again, as this is a listener socket and not an address, we make this
new argument an "int" protocol family, instead of an "sa_family_t."
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: add documentating comment and use appropriate data types for
svc_find_xprt()'s arguments.
This also eliminates a mixed sign comparison: @port was an int, while
the return value of svc_xprt_local_port() is an unsigned short.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Add /proc/fs/nfsd/pool_stats to export to userspace various
statistics about the operation of rpc server thread pools.
This patch is based on a forward-ported version of
knfsd-add-pool-thread-stats which has been shipping in the SGI
"Enhanced NFS" product since 2006 and which was previously
posted:
http://article.gmane.org/gmane.linux.nfs/10375
It has also been updated thus:
* moved EXPORT_SYMBOL() to near the function it exports
* made the new struct struct seq_operations const
* used SEQ_START_TOKEN instead of ((void *)1)
* merged fix from SGI PV 990526 "sunrpc: use dprintk instead of
printk in svc_pool_stats_*()" by Harshula Jayasuriya.
* merged fix from SGI PV 964001 "Crash reading pool_stats before
nfsds are started".
Signed-off-by: Greg Banks <gnb@sgi.com>
Signed-off-by: Harshula Jayasuriya <harshula@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Avoid overloading the CPU scheduler with enormous load averages
when handling high call-rate NFS loads. When the knfsd bottom half
is made aware of an incoming call by the socket layer, it tries to
choose an nfsd thread and wake it up. As long as there are idle
threads, one will be woken up.
If there are lot of nfsd threads (a sensible configuration when
the server is disk-bound or is running an HSM), there will be many
more nfsd threads than CPUs to run them. Under a high call-rate
low service-time workload, the result is that almost every nfsd is
runnable, but only a handful are actually able to run. This situation
causes two significant problems:
1. The CPU scheduler takes over 10% of each CPU, which is robbing
the nfsd threads of valuable CPU time.
2. At a high enough load, the nfsd threads starve userspace threads
of CPU time, to the point where daemons like portmap and rpc.mountd
do not schedule for tens of seconds at a time. Clients attempting
to mount an NFS filesystem timeout at the very first step (opening
a TCP connection to portmap) because portmap cannot wake up from
select() and call accept() in time.
Disclaimer: these effects were observed on a SLES9 kernel, modern
kernels' schedulers may behave more gracefully.
The solution is simple: keep in each svc_pool a counter of the number
of threads which have been woken but have not yet run, and do not wake
any more if that count reaches an arbitrary small threshold.
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. The server was running 128 nfsds.
Profiling showed schedule() taking 6.7% of every CPU, and __wake_up()
taking 5.2%. This patch drops those contributions to 3.0% and 2.2%.
Load average was over 120 before the patch, and 20.9 after.
This patch is a forward-ported version of knfsd-avoid-nfsd-overload
which has been shipping in the SGI "Enhanced NFS" product since 2006.
It has been posted before:
http://article.gmane.org/gmane.linux.nfs/10374
Signed-off-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
A race between svc_revisit and svc_delete_xprt can result in
deferred requests holding references on a transport that can never be
recovered because dead transports are not enqueued for subsequent
processing.
Check for XPT_DEAD in revisit to clean up completing deferrals on a dead
transport and sweep a transport's deferred queue to do the same for queued
but unprocessed deferrals.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The rqstp structure has a pointer to a svc_deferred_req record
that is allocated when requests are deferred. This record is common
to all transports and can be freed in common code.
Move the kfree of the rq_deferred to the common svc_xprt_release
function.
This also fixes a memory leak in the RDMA transport which does not
kfree the dr structure in it's version of the xpo_release_rqst callback.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
svc_check_conn_limits() attempts to prevent denial of service attacks
by having the service close old connections once it reaches a
threshold. This threshold is based on the number of threads in the
service:
(serv->sv_nrthreads + 3) * 20
Once we reach this, we drop the oldest connections and a printk pops
to warn the admin that they should increase the number of threads.
Increasing the number of threads isn't an option however for services
like lockd. We don't want to eliminate this check entirely for such
services but we need some way to increase this limit.
This patch adds a sv_maxconn field to the svc_serv struct. When it's
set to 0, we use the current method to calculate the max number of
connections. RPC services can then set this on an as-needed basis.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Acked-by: Neil Brown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Teach svc_create_xprt() to use the correct ANY address for AF_INET6 based
RPC services.
No caller uses AF_INET6 yet.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Remove a redundant check for the XPT_DEAD bit in the svc_xprt_enqueue
function. This same bit is checked below while holding the pool lock
and prints a debug message if found to be dead.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
svc_recv() calls alloc_page(), and if it fails it does a 500ms
uninterruptible sleep and then reattempts. There doesn't seem to be any
real reason for this to be uninterruptible, so change it to an
interruptible sleep. Also check for kthread_stop() and signalled() after
setting the task state to avoid races that might lead to sleeping after
kthread_stop() wakes up the task.
I've done some very basic smoke testing with this, but obviously it's
hard to test the actual changes since this all depends on an
alloc_page() call failing.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
When using kthreads that call into svc_recv, we want to make sure that
they do not block there for a long time when we're trying to take down
the kthread.
This patch changes svc_recv() to check kthread_should_stop() at the same
places that it checks to see if it's signalled(). Also check just before
svc_recv() tries to schedule(). By making sure that we check it just
after setting the task state we can avoid having to use any locking or
signalling to ensure it doesn't block for a long time.
There's still a chance of a 500ms sleep if alloc_page() fails, but
that should be a rare occurrence and isn't a terribly long time in
the context of a kthread being taken down.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Create a transport independent version of the svc_sock_names function.
The toclose capability of the svc_sock_names service can be implemented
using the svc_xprt_find and svc_xprt_close services.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Update the write handler for the portlist file to allow creating new
listening endpoints on a transport. The general form of the string is:
<transport_name><space><port number>
For example:
echo "tcp 2049" > /proc/fs/nfsd/portlist
This is intended to support the creation of a listening endpoint for
RDMA transports without adding #ifdef code to the nfssvc.c file.
Transports can also be removed as follows:
'-'<transport_name><space><port number>
For example:
echo "-tcp 2049" > /proc/fs/nfsd/portlist
Attempting to add a listener with an invalid transport string results
in EPROTONOSUPPORT and a perror string of "Protocol not supported".
Attempting to remove an non-existent listener (.e.g. bad proto or port)
results in ENOTCONN and a perror string of
"Transport endpoint is not connected"
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Add a new svc function that allows a service to query whether a
transport instance has already been created. This is used in lockd
to determine whether or not a transport needs to be created when
a lockd instance is brought up.
Specifying 0 for the address family or port is effectively a wild-card,
and will result in matching the first transport in the service's list
that has a matching class name.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Add a file that when read lists the set of registered svc
transports.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Some transports have a header in front of the RPC header. The current
defer/revisit processing considers only the iov_len and arg_len to
determine how much to back up when saving the original request
to revisit. Add a field to the rqstp structure to save the size
of the transport header so svc_defer can correctly compute
the start of a request.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
This functionally trivial patch moves all of the transport independent
functions from the svcsock.c file to the transport independent svc_xprt.c
file.
In addition the following formatting changes were made:
- White space cleanup
- Function signatures on single line
- The inline directive was removed
- Lines over 80 columns were reformatted
- The term 'socket' was changed to 'transport' in comments
- The SMP comment was moved and updated.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Move the svc transport list logic into common transport creation code.
Refactor this code path to make the flow of control easier to read.
Move the setting and clearing of the BUSY_BIT during transport creation
to common code.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
This patch moves the transport sockaddr to the svc_xprt
structure. Convenience functions are added to set and
get the local and remote addresses of a transport from
the transport provider as well as determine the length
of a sockaddr.
A transport is responsible for setting the xpt_local
and xpt_remote addresses in the svc_xprt structure as
part of transport creation and xpo_accept processing. This
cannot be done in a generic way and in fact varies
between TCP, UDP and RDMA. A set of xpo_ functions
(e.g. getlocalname, getremotename) could have been
added but this would have resulted in additional
caching and copying of the addresses around. Note that
the xpt_local address should also be set on listening
endpoints; for TCP/RDMA this is done as part of
endpoint creation.
For connected transports like TCP and RDMA, the addresses
never change and can be set once and copied into the
rqstp structure for each request. For UDP, however, the
local and remote addresses may change for each request. In
this case, the address information is obtained from the
UDP recvmsg info and copied into the rqstp structure from
there.
A svc_xprt_local_port function was also added that returns
the local port given a transport. This is used by
svc_create_xprt when returning the port associated with
a newly created transport, and later when creating a
generic find transport service to check if a service is
already listening on a given port.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
This patch moves the transport independent sk_deferred list to the svc_xprt
structure and updates the svc_deferred_req structure to keep pointers to
svc_xprt's directly. The deferral processing code is also moved out of the
transport dependent recvfrom functions and into the generic svc_recv path.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Move the authinfo cache to svc_xprt. This allows both the TCP and RDMA
transports to share this logic. A flag bit is used to determine if
auth information is to be cached or not. Previously, this code looked
at the transport protocol.
I've also changed the spin_lock/unlock logic so that a lock is not taken for
transports that are not caching auth info.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Move the sk_mutex field to the transport independent svc_xprt structure.
Now all the fields that svc_send touches are transport neutral. Change the
svc_send function to use the transport independent svc_xprt directly instead
of the transport dependent svc_sock structure.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Move sk_list and sk_ready to svc_xprt. This involves close because these
lists are walked by svcs when closing all their transports. So I combined
the moving of these lists to svc_xprt with making close transport independent.
The svc_force_sock_close has been changed to svc_close_all and takes a list
as an argument. This removes some svc internals knowledge from the svcs.
This code races with module removal and transport addition.
Thanks to Simon Holm Thøgersen for a compile fix.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Cc: Simon Holm Thøgersen <odie@cs.aau.dk>
This is another incremental change that moves transport independent
fields from svc_sock to the svc_xprt structure. The changes
should be functionally null.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Change the atomic_t reference count to a kref and move it to the
transport indepenent svc_xprt structure. Change the reference count
wrapper names to be generic.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The svc_create_xprt function is a transport independent version
of the svc_makesock function.
Since transport instance creation contains transport dependent and
independent components, add an xpo_create transport function. The
transport implementation of this function allocates the memory for the
endpoint, implements the transport dependent initialization logic, and
calls svc_xprt_init to initialize the transport independent field (svc_xprt)
in it's data structure.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
The transport class (svc_xprt_class) represents a type of transport, e.g.
udp, tcp, rdma. A transport class has a unique name and a set of transport
operations kept in the svc_xprt_ops structure.
A transport class can be dynamically registered and unregisterd. The
svc_xprt_class represents the module that implements the transport
type and keeps reference counts on the module to avoid unloading while
there are active users.
The endpoint (svc_xprt) is a generic, transport independent endpoint that can
be used to send and receive data for an RPC service. It inherits it's
operations from the transport class.
A transport driver module registers and unregisters itself with svc sunrpc
by calling svc_reg_xprt_class, and svc_unreg_xprt_class respectively.
Signed-off-by: Tom Tucker <tom@opengridcomputing.com>
Acked-by: Neil Brown <neilb@suse.de>
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Reviewed-by: Greg Banks <gnb@sgi.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
J. Bruce Fields