Currently RPC performance metrics that tabulate elapsed time use
jiffies time values. This is problematic on systems that use slow
jiffies (for instance 100HZ systems built for paravirtualized
environments). It is also a problem for computing precise latency
statistics for advanced network transports, such as InfiniBand,
that can have round-trip latencies significanly faster than a single
clock tick.
For the RPC client, adopt the high resolution time stamp mechanism
already used by the network layer and blktrace: ktime.
We use ktime format time stamps for all internal computations, and
convert to milliseconds for presentation. As a result, we need only
addition operations in the performance critical paths; multiply/divide
is required only for presentation.
We could report RTT metrics in microseconds. In fact the mountstats
format is versioned to accomodate exactly this kind of interface
improvement.
For now, however, we'll stay with millisecond precision for
presentation to maintain backwards compatibility with the handful of
currently deployed user space tools. At a later point, we'll move to
an API such as BDI_STATS where a finer timestamp precision can be
reported.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Also have it return an ERR_PTR(-ENOMEM) instead of a null pointer.
Reviewed-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Move struct rpc_buffer's definition into a sunrpc.h, a common, internal
header file, in preparation for supporting the nfsv4.1 backchannel.
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfs41: sunrpc: #include <linux/net.h> from sunrpc.h]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
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>
We should probably not be testing any flags after we've cleared the
RPC_TASK_RUNNING flag, since rpc_make_runnable() is then free to assign the
rpc_task to another workqueue, which may then destroy it.
We can fix any races with rpc_make_runnable() by ensuring that we only
clear the RPC_TASK_RUNNING flag while holding the rpc_wait_queue->lock that
the task is supposed to be sleeping on (and then checking whether or not
the task really is sleeping).
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Push it into those callback functions that actually need it.
Note that all the NFS operations use their own locking, so don't need the
BKL. Ditto for the rpcbind client.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Try to make the comment here a little more clear and concise.
Also, this macro definition seems unnecessary.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The current RPCAUTH_LOOKUP_ROOTCREDS flag only works for AUTH_SYS
authentication, and then only as a special case in the code. This patch
removes the auth_sys special casing, and replaces it with generic code.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Now that we've tightened up the locking rules for RPC queue wakeups, we can
remove the RCU-safe kfree calls...
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
This is designed to replace the timeout timer in the individual rpc_tasks.
By putting the timer function in the wait queue, we will eventually be able
to reduce the total number of timers in use by the RPC subsystem.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
An audit of the current RPC timeout functions shows that they don't really
ever need to run in the softirq context. As long as the softirq is
able to signal that the wakeup is due to a timeout (which it can do by
setting task->tk_status to -ETIMEDOUT) then the callback functions can just
run as standard task->tk_callback functions (in the rpciod/process
context).
The only possible border-line case would be xprt_timer() for the case of
UDP, when the callback is used to reduce the size of the transport
congestion window. In testing, however, the effect of moving that update
to a callback would appear to be minor.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
In all cases where we currently use rpc_wake_up_task(), we almost always
know on which waitqueue the rpc_task is actually sleeping. This will allows
us to simplify the queue locking in a future patch.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
All RPC timeout callback functions are expected to wake the task up. We can
enforce this by moving the wakeup back into rpc_run_timer.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
A lot of the work done by the rpc_release() callback is inappropriate for
rpciod as it will often involve things like starting a new rpc call in
order to clean up state after an interrupted NFSv4 open() call, or
calls to mntput(), etc.
This patch allows the caller of rpc_run_task() to specify that the
rpc_release callback should run on a different workqueue than the default
rpciod_workqueue.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* 'task_killable' of git://git.kernel.org/pub/scm/linux/kernel/git/willy/misc: (22 commits)
Remove commented-out code copied from NFS
NFS: Switch from intr mount option to TASK_KILLABLE
Add wait_for_completion_killable
Add wait_event_killable
Add schedule_timeout_killable
Use mutex_lock_killable in vfs_readdir
Add mutex_lock_killable
Use lock_page_killable
Add lock_page_killable
Add fatal_signal_pending
Add TASK_WAKEKILL
exit: Use task_is_*
signal: Use task_is_*
sched: Use task_contributes_to_load, TASK_ALL and TASK_NORMAL
ptrace: Use task_is_*
power: Use task_is_*
wait: Use TASK_NORMAL
proc/base.c: Use task_is_*
proc/array.c: Use TASK_REPORT
perfmon: Use task_is_*
...
Fixed up conflicts in NFS/sunrpc manually..
We want the default scheduling priority (priority == 0) to remain
RPC_PRIORITY_NORMAL.
Also ensure that the priority wait queue scheduling is per process id
instead of sometimes being per thread, and sometimes being per inode.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The sunrpc client exports are not meant to be part of any official kernel
API: they can change at the drop of a hat. Mark them as internal functions
using EXPORT_SYMBOL_GPL.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Many-many code in the kernel initialized the timer->function
and timer->data together with calling init_timer(timer). There
is already a helper for this. Use it for networking code.
The patch is HUGE, but makes the code 130 lines shorter
(98 insertions(+), 228 deletions(-)).
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
By using the TASK_KILLABLE infrastructure, we can get rid of the 'intr'
mount option. We have to use _killable everywhere instead of _interruptible
as we get rid of rpc_clnt_sigmask/sigunmask.
Signed-off-by: Liam R. Howlett <howlett@gmail.com>
Signed-off-by: Matthew Wilcox <willy@linux.intel.com>
The task_struct->pid member is going to be deprecated, so start
using the helpers (task_pid_nr/task_pid_vnr/task_pid_nr_ns) in
the kernel.
The first thing to start with is the pid, printed to dmesg - in
this case we may safely use task_pid_nr(). Besides, printks produce
more (much more) than a half of all the explicit pid usage.
[akpm@linux-foundation.org: git-drm went and changed lots of stuff]
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Dave Airlie <airlied@linux.ie>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
SUNRPC: add EXPORT_SYMBOL_GPL for generic transport functions
As a preface to allowing arbitrary transport modules to be loaded
dynamically, add EXPORT_SYMBOL_GPL for all generic transport functions
that a transport implementation might want to use.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Acked-by: Tom Talpey <tmt@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The commit 4ada539ed7 lead to the unpleasant
possibility of an asynchronous rpc_task being required to call
rpciod_down() when it is complete. This again means that the rpciod
workqueue may get to call destroy_workqueue on itself -> hang...
Change rpciod_up/rpciod_down to just get/put the module, and then
create/destroy the workqueues on module load/unload.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Slab destructors were no longer supported after Christoph's
c59def9f22 change. They've been
BUGs for both slab and slub, and slob never supported them
either.
This rips out support for the dtor pointer from kmem_cache_create()
completely and fixes up every single callsite in the kernel (there were
about 224, not including the slab allocator definitions themselves,
or the documentation references).
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
Instead of taking the mutex every time we just need to increment/decrement
rpciod_users, we can optmise by using atomic_inc_not_zero and
atomic_dec_and_test.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Replace it with explicit calls to rpc_shutdown_client() or
rpc_destroy_client() (for the case of asynchronous calls).
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
rpciod_running is not used at all, but due to the way DECLARE_MUTEX_LOCKED
works we don't get a warning for it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
While the comment says:
* To prevent rpciod from hanging, this allocator never sleeps,
* returning NULL if the request cannot be serviced immediately.
The function does not actually check for NULL pointers being returned.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Use a cleaner method to find the size of an rpc_buffer. This actually
works on x86-64!
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Commit c5a4dd8b7c introduced the following
compiler warnings:
net/sunrpc/sched.c:766: warning: format '%u' expects type 'unsigned int', but argument 3 has type 'size_t'
net/sunrpc/sched.c:785: warning: format '%u' expects type 'unsigned int', but argument 2 has type 'size_t'
- Use %zu to format size_t
- Kill 2 useless casts
Signed-off-by: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently rpc_malloc sets req->rq_buffer internally. Make this a more
generic interface: return a pointer to the new buffer (or NULL) and
make the caller set req->rq_buffer and req->rq_bufsize. This looks much
more like kmalloc and eliminates the side effects.
To fix a potential deadlock, this patch also replaces GFP_NOFS with
GFP_NOWAIT in rpc_malloc. This prevents async RPCs from sleeping outside
the RPC's task scheduler while allocating their buffer.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The tk_pid field is an unsigned short. The proper print format specifier for
that type is %5u, not %4d.
Also clean up some miscellaneous print formatting nits.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The error values are already propagated through task->tk_status, and
none of the callers check one without checking the other, so we can
drop the return value.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Fix the Oops in http://bugzilla.linux-nfs.org/show_bug.cgi?id=138
We shouldn't be calling rpc_release_task() for tasks that are not active.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Replace all uses of kmem_cache_t with struct kmem_cache.
The patch was generated using the following script:
#!/bin/sh
#
# Replace one string by another in all the kernel sources.
#
set -e
for file in `find * -name "*.c" -o -name "*.h"|xargs grep -l $1`; do
quilt add $file
sed -e "1,\$s/$1/$2/g" $file >/tmp/$$
mv /tmp/$$ $file
quilt refresh
done
The script was run like this
sh replace kmem_cache_t "struct kmem_cache"
Signed-off-by: Christoph Lameter <clameter@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
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>
Pass the work_struct pointer to the work function rather than context data.
The work function can use container_of() to work out the data.
For the cases where the container of the work_struct may go away the moment the
pending bit is cleared, it is made possible to defer the release of the
structure by deferring the clearing of the pending bit.
To make this work, an extra flag is introduced into the management side of the
work_struct. This governs auto-release of the structure upon execution.
Ordinarily, the work queue executor would release the work_struct for further
scheduling or deallocation by clearing the pending bit prior to jumping to the
work function. This means that, unless the driver makes some guarantee itself
that the work_struct won't go away, the work function may not access anything
else in the work_struct or its container lest they be deallocated.. This is a
problem if the auxiliary data is taken away (as done by the last patch).
However, if the pending bit is *not* cleared before jumping to the work
function, then the work function *may* access the work_struct and its container
with no problems. But then the work function must itself release the
work_struct by calling work_release().
In most cases, automatic release is fine, so this is the default. Special
initiators exist for the non-auto-release case (ending in _NAR).
Signed-Off-By: David Howells <dhowells@redhat.com>
* Rougly half of callers already do it by not checking return value
* Code in drivers/acpi/osl.c does the following to be sure:
(void)kmem_cache_destroy(cache);
* Those who check it printk something, however, slab_error already printed
the name of failed cache.
* XFS BUGs on failed kmem_cache_destroy which is not the decision
low-level filesystem driver should make. Converted to ignore.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
include/linux/sunrpc/clnt.h already includes include/linux/sunrpc/xprt.h.
We can remove xprt.h from source files that already include clnt.h.
Likewise include/linux/sunrpc/timer.h.
Test plan:
Compile kernel with CONFIG_NFS enabled.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The previous patches removed the last user of RPC child tasks, so we can
remove support for child tasks from net/sunrpc/sched.c now.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Modify well over a dozen mempool users to call mempool_create_slab_pool()
rather than calling mempool_create() with extra arguments, saving about 30
lines of code and increasing readability.
Signed-off-by: Matthew Dobson <colpatch@us.ibm.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Semaphore to mutex conversion.
The conversion was generated via scripts, and the result was validated
automatically via a script as well.
Signed-off-by: Arjan van de Ven <arjan@infradead.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently this will not happen if we exit before rpc_new_task() was called.
Also fix up rpc_run_task() to do the same (for consistency).
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Account for various things that occur while an RPC task is executed.
Separate timers for RPC round trip and RPC execution time show how
long RPC requests wait in queue before being sent. Eventually these
will be accumulated at xprt_release time in one place where they can
be viewed from userland.
Test plan:
Compile kernel with CONFIG_NFS enabled.
Signed-off-by: Chuck Lever <cel@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
RPC wait queue length will eventually be exported to userland via the RPC
iostats interface.
Test plan:
Compile kernel with CONFIG_NFS enabled.
Signed-off-by: Chuck Lever <cel@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
In rpc_wake_up() and rpc_wake_up_status(), it is possible for the call to
__rpc_wake_up_task() to fail if another thread happens to be calling
rpc_wake_up_task() on the same rpc_task.
Problem noticed by Bruno Faccini.
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>
Fix the syntax of some kernel-doc comments
Signed-off-by: Martin Waitz <tali@admingilde.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Add RPC client transport switch support for replacing buffer management
on a per-transport basis.
In the current IPv4 socket transport implementation, RPC buffers are
allocated as needed for each RPC message that is sent. Some transport
implementations may choose to use pre-allocated buffers for encoding,
sending, receiving, and unmarshalling RPC messages, however. For
transports capable of direct data placement, the buffers can be carved
out of a pre-registered area of memory rather than from a slab cache.
Test-plan:
Millions of fsx operations. Performance characterization with "sio" and
"iozone". Use oprofile and other tools to look for significant regression
in CPU utilization.
Signed-off-by: Chuck Lever <cel@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The NFSv4 model requires us to complete all RPC calls that might
establish state on the server whether or not the user wants to
interrupt it. We may also need to schedule new work (including
new RPC calls) in order to cancel the new state.
The asynchronous RPC model will allow us to ensure that RPC calls
always complete, but in order to allow for "synchronous" RPC, we
want to add the ability to wait for completion.
The waits are, of course, interruptible.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Shrink the RPC task structure. Instead of storing separate pointers
for task->tk_exit and task->tk_release, put them in a structure.
Also pass the user data pointer as a parameter instead of passing it via
task->tk_calldata. This enables us to nest callbacks.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
- added typedef unsigned int __nocast gfp_t;
- replaced __nocast uses for gfp flags with gfp_t - it gives exactly
the same warnings as far as sparse is concerned, doesn't change
generated code (from gcc point of view we replaced unsigned int with
typedef) and documents what's going on far better.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Fix nocast sparse warnings:
net/rxrpc/call.c:2013:25: warning: implicit cast to nocast type
net/rxrpc/connection.c:538:46: warning: implicit cast to nocast type
net/sunrpc/sched.c:730:36: warning: implicit cast to nocast type
net/sunrpc/sched.c:734:56: warning: implicit cast to nocast type
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch puts mostly read only data in the right section
(read_mostly), to help sharing of these data between CPUS without
memory ping pongs.
On one of my production machine, tcp_statistics was sitting in a
heavily modified cache line, so *every* SNMP update had to force a
reload.
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the task->tk_exit() wants to restart the RPC call after delaying
then the current RPC code will clobber the timer by calling
rpc_delete_timer() immediately after re-entering the loop in
__rpc_execute().
Problem noticed by Oleg Nesterov <oleg@tv-sign.ru>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
Chuck Lever