rw-semaphore is the only type of lock doing this ugliness of
exporting at the end of the file.
Signed-off-by: NDavidlohr Bueso <dbueso@suse.de>
Cc: dave@stgolabs.net
Cc: peterz@infradead.org
Link: http://lkml.kernel.org/r/1410500066-5909-1-git-send-email-dave@stgolabs.netSigned-off-by: NIngo Molnar <mingo@kernel.org>

The arch_mutex_cpu_relax() function, introduced by 34b133f8, is
hacky and ugly. It was added a few years ago to address the fact
that common cpu_relax() calls include yielding on s390, and thus
impact the optimistic spinning functionality of mutexes. Nowadays
we use this function well beyond mutexes: rwsem, qrwlock, mcs and
lockref. Since the macro that defines the call is in the mutex header,
any users must include mutex.h and the naming is misleading as well.

This patch (i) renames the call to cpu_relax_lowlatency  ("relax, but
only if you can do it with very low latency") and (ii) defines it in
each arch's asm/processor.h local header, just like for regular cpu_relax
functions. On all archs, except s390, cpu_relax_lowlatency is simply cpu_relax,
and thus we can take it out of mutex.h. While this can seem redundant,
I believe it is a good choice as it allows us to move out arch specific
logic from generic locking primitives and enables future(?) archs to
transparently define it, similarly to System Z.
Signed-off-by: NDavidlohr Bueso <davidlohr@hp.com>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Anton Blanchard <anton@samba.org>
Cc: Aurelien Jacquiot <a-jacquiot@ti.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Bharat Bhushan <r65777@freescale.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Chen Liqin <liqin.linux@gmail.com>
Cc: Chris Metcalf <cmetcalf@tilera.com>
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Chris Zankel <chris@zankel.net>
Cc: David Howells <dhowells@redhat.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Deepthi Dharwar <deepthi@linux.vnet.ibm.com>
Cc: Dominik Dingel <dingel@linux.vnet.ibm.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: Haavard Skinnemoen <hskinnemoen@gmail.com>
Cc: Hans-Christian Egtvedt <egtvedt@samfundet.no>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Hirokazu Takata <takata@linux-m32r.org>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: James E.J. Bottomley <jejb@parisc-linux.org>
Cc: James Hogan <james.hogan@imgtec.com>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Jesper Nilsson <jesper.nilsson@axis.com>
Cc: Joe Perches <joe@perches.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Joseph Myers <joseph@codesourcery.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Koichi Yasutake <yasutake.koichi@jp.panasonic.com>
Cc: Lennox Wu <lennox.wu@gmail.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mark Salter <msalter@redhat.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Max Filippov <jcmvbkbc@gmail.com>
Cc: Michael Neuling <mikey@neuling.org>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Mikael Starvik <starvik@axis.com>
Cc: Nicolas Pitre <nico@linaro.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Paul Burton <paul.burton@imgtec.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Qais Yousef <qais.yousef@imgtec.com>
Cc: Qiaowei Ren <qiaowei.ren@intel.com>
Cc: Rafael Wysocki <rafael.j.wysocki@intel.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Richard Kuo <rkuo@codeaurora.org>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Steven Miao <realmz6@gmail.com>
Cc: Steven Rostedt <srostedt@redhat.com>
Cc: Stratos Karafotis <stratosk@semaphore.gr>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Vasily Kulikov <segoon@openwall.com>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Vineet Gupta <Vineet.Gupta1@synopsys.com>
Cc: Waiman Long <Waiman.Long@hp.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: adi-buildroot-devel@lists.sourceforge.net
Cc: linux390@de.ibm.com
Cc: linux-alpha@vger.kernel.org
Cc: linux-am33-list@redhat.com
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-c6x-dev@linux-c6x.org
Cc: linux-cris-kernel@axis.com
Cc: linux-hexagon@vger.kernel.org
Cc: linux-ia64@vger.kernel.org
Cc: linux@lists.openrisc.net
Cc: linux-m32r-ja@ml.linux-m32r.org
Cc: linux-m32r@ml.linux-m32r.org
Cc: linux-m68k@lists.linux-m68k.org
Cc: linux-metag@vger.kernel.org
Cc: linux-mips@linux-mips.org
Cc: linux-parisc@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-s390@vger.kernel.org
Cc: linux-sh@vger.kernel.org
Cc: linux-xtensa@linux-xtensa.org
Cc: sparclinux@vger.kernel.org
Link: http://lkml.kernel.org/r/1404079773.2619.4.camel@buesod1.americas.hpqcorp.netSigned-off-by: NIngo Molnar <mingo@kernel.org>

Just like with mutexes (CONFIG_MUTEX_SPIN_ON_OWNER),
encapsulate the dependencies for rwsem optimistic spinning.
No logical changes here as it continues to depend on both
SMP and the XADD algorithm variant.
Signed-off-by: NDavidlohr Bueso <davidlohr@hp.com>
Acked-by: NJason Low <jason.low2@hp.com>
[ Also make it depend on ARCH_SUPPORTS_ATOMIC_RMW. ]
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1405112406-13052-2-git-send-email-davidlohr@hp.com
Cc: aswin@hp.com
Cc: Chris Mason <clm@fb.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Josef Bacik <jbacik@fusionio.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Waiman Long <Waiman.Long@hp.com>
Signed-off-by: NIngo Molnar <mingo@kernel.org>
Signed-off-by: NIngo Molnar <mingo@kernel.org>

Currently, we initialize the osq lock by directly setting the lock's values. It
would be preferable if we use an init macro to do the initialization like we do
with other locks.

This patch introduces and uses a macro and function for initializing the osq lock.
Signed-off-by: NJason Low <jason.low2@hp.com>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: Scott Norton <scott.norton@hp.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Waiman Long <waiman.long@hp.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Chris Mason <clm@fb.com>
Cc: Josef Bacik <jbacik@fusionio.com>
Link: http://lkml.kernel.org/r/1405358872-3732-4-git-send-email-jason.low2@hp.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

The cancellable MCS spinlock is currently used to queue threads that are
doing optimistic spinning. It uses per-cpu nodes, where a thread obtaining
the lock would access and queue the local node corresponding to the CPU that
it's running on. Currently, the cancellable MCS lock is implemented by using
pointers to these nodes.

In this patch, instead of operating on pointers to the per-cpu nodes, we
store the CPU numbers in which the per-cpu nodes correspond to in atomic_t.
A similar concept is used with the qspinlock.

By operating on the CPU # of the nodes using atomic_t instead of pointers
to those nodes, this can reduce the overhead of the cancellable MCS spinlock
by 32 bits (on 64 bit systems).
Signed-off-by: NJason Low <jason.low2@hp.com>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: Scott Norton <scott.norton@hp.com>
Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Waiman Long <waiman.long@hp.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Chris Mason <clm@fb.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Josef Bacik <jbacik@fusionio.com>
Link: http://lkml.kernel.org/r/1405358872-3732-3-git-send-email-jason.low2@hp.comSigned-off-by: NIngo Molnar <mingo@kernel.org>

Commit 4fc828e2 ("locking/rwsem: Support optimistic spinning")
introduced a major performance regression for workloads such as
xfs_repair which mix read and write locking of the mmap_sem across
many threads. The result was xfs_repair ran 5x slower on 3.16-rc2
than on 3.15 and using 20x more system CPU time.

Perf profiles indicate in some workloads that significant time can
be spent spinning on !owner. This is because we don't set the lock
owner when readers(s) obtain the rwsem.

In this patch, we'll modify rwsem_can_spin_on_owner() such that we'll
return false if there is no lock owner. The rationale is that if we
just entered the slowpath, yet there is no lock owner, then there is
a possibility that a reader has the lock. To be conservative, we'll
avoid spinning in these situations.

This patch reduced the total run time of the xfs_repair workload from
about 4 minutes 24 seconds down to approximately 1 minute 26 seconds,
back to close to the same performance as on 3.15.

Retesting of AIM7, which were some of the workloads used to test the
original optimistic spinning code, confirmed that we still get big
performance gains with optimistic spinning, even with this additional
regression fix. Davidlohr found that while the 'custom' workload took
a performance hit of ~-14% to throughput for >300 users with this
additional patch, the overall gain with optimistic spinning is
still ~+45%. The 'disk' workload even improved by ~+15% at >1000 users.
Tested-by: NDave Chinner <dchinner@redhat.com>
Acked-by: NDavidlohr Bueso <davidlohr@hp.com>
Signed-off-by: NJason Low <jason.low2@hp.com>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1404532172.2572.30.camel@j-VirtualBoxSigned-off-by: NIngo Molnar <mingo@kernel.org>

WARNING: line over 80 characters
#205: FILE: kernel/locking/rwsem-xadd.c:275:
+		old = cmpxchg(&sem->count, count, count + RWSEM_ACTIVE_WRITE_BIAS);

WARNING: line over 80 characters
#376: FILE: kernel/locking/rwsem-xadd.c:434:
+		 * If there were already threads queued before us and there are no

WARNING: line over 80 characters
#377: FILE: kernel/locking/rwsem-xadd.c:435:
+		 * active writers, the lock must be read owned; so we try to wake

total: 0 errors, 3 warnings, 417 lines checked
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/n/tip-pn6pslaplw031lykweojsn8c@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

We have reached the point where our mutexes are quite fine tuned
for a number of situations. This includes the use of heuristics
and optimistic spinning, based on MCS locking techniques.

Exclusive ownership of read-write semaphores are, conceptually,
just about the same as mutexes, making them close cousins. To
this end we need to make them both perform similarly, and
right now, rwsems are simply not up to it. This was discovered
by both reverting commit 4fc3f1d6 (mm/rmap, migration: Make
rmap_walk_anon() and try_to_unmap_anon() more scalable) and
similarly, converting some other mutexes (ie: i_mmap_mutex) to
rwsems. This creates a situation where users have to choose
between a rwsem and mutex taking into account this important
performance difference. Specifically, biggest difference between
both locks is when we fail to acquire a mutex in the fastpath,
optimistic spinning comes in to play and we can avoid a large
amount of unnecessary sleeping and overhead of moving tasks in
and out of wait queue. Rwsems do not have such logic.

This patch, based on the work from Tim Chen and I, adds support
for write-side optimistic spinning when the lock is contended.
It also includes support for the recently added cancelable MCS
locking for adaptive spinning. Note that is is only applicable
to the xadd method, and the spinlock rwsem variant remains intact.

Allowing optimistic spinning before putting the writer on the wait
queue reduces wait queue contention and provided greater chance
for the rwsem to get acquired. With these changes, rwsem is on par
with mutex. The performance benefits can be seen on a number of
workloads. For instance, on a 8 socket, 80 core 64bit Westmere box,
aim7 shows the following improvements in throughput:

 +--------------+---------------------+-----------------+
 |   Workload   | throughput-increase | number of users |
 +--------------+---------------------+-----------------+
 | alltests     | 20%                 | >1000           |
 | custom       | 27%, 60%            | 10-100, >1000   |
 | high_systime | 36%, 30%            | >100, >1000     |
 | shared       | 58%, 29%            | 10-100, >1000   |
 +--------------+---------------------+-----------------+

There was also improvement on smaller systems, such as a quad-core
x86-64 laptop running a 30Gb PostgreSQL (pgbench) workload for up
to +60% in throughput for over 50 clients. Additionally, benefits
were also noticed in exim (mail server) workloads. Furthermore, no
performance regression have been seen at all.

Based-on-work-from: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: NDavidlohr Bueso <davidlohr@hp.com>
[peterz: rej fixup due to comment patches, sched/rt.h header]
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Cc: Alex Shi <alex.shi@linaro.org>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Michel Lespinasse <walken@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Hurley <peter@hurleysoftware.com>
Cc: "Paul E.McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Aswin Chandramouleeswaran <aswin@hp.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: "Scott J Norton" <scott.norton@hp.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Chris Mason <clm@fb.com>
Cc: Josef Bacik <jbacik@fusionio.com>
Link: http://lkml.kernel.org/r/1399055055.6275.15.camel@buesod1.americas.hpqcorp.netSigned-off-by: NIngo Molnar <mingo@kernel.org>

It took me quite a while to understand how rwsem's count field
mainifested itself in different scenarios.

Add comments to provide a quick reference to the the rwsem's count
field for each scenario where readers and writers are contending
for the lock.

Hopefully it will be useful for future maintenance of the code and
for people to get up to speed on how the logic in the code works.
Signed-off-by: NTim Chen <tim.c.chen@linux.intel.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Alex Shi <alex.shi@linaro.org>
Cc: Michel Lespinasse <walken@google.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Peter Hurley <peter@hurleysoftware.com>
Cc: Paul E.McKenney <paulmck@linux.vnet.ibm.com>
Cc: Jason Low <jason.low2@hp.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/1399060437.2970.146.camel@schen9-DESKSigned-off-by: NIngo Molnar <mingo@kernel.org>

Mark the rwsem functions that can be called from assembler asmlinkage.
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Link: http://lkml.kernel.org/r/1391845930-28580-9-git-send-email-ak@linux.intel.comSigned-off-by: NH. Peter Anvin <hpa@linux.intel.com>

Notably: changed lib/rwsem* targets from lib- to obj-, no idea about
the ramifications of that.
Suggested-by: NIngo Molnar <mingo@kernel.org>
Signed-off-by: NPeter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/n/tip-g0kynfh5feriwc6p3h6kpbw6@git.kernel.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>

This patch tries to reduce the amount of cmpxchg calls in the writer
failed path by checking the counter value first before issuing the
instruction.  If ->count is not set to RWSEM_WAITING_BIAS then there is
no point wasting a cmpxchg call.

Furthermore, Michel states "I suppose it helps due to the case where
someone else steals the lock while we're trying to acquire
sem->wait_lock."

Two very different workloads and machines were used to see how this
patch improves throughput: pgbench on a quad-core laptop and aim7 on a
large 8 socket box with 80 cores.

Some results comparing Michel's fast-path write lock stealing
(tps-rwsem) on a quad-core laptop running pgbench:

  | db_size | clients  |  tps-rwsem     |   tps-patch  |
  +---------+----------+----------------+--------------+
  | 160 MB   |       1 |           6906 |         9153 | + 32.5
  | 160 MB   |       2 |          15931 |        22487 | + 41.1%
  | 160 MB   |       4 |          33021 |        32503 |
  | 160 MB   |       8 |          34626 |        34695 |
  | 160 MB   |      16 |          33098 |        34003 |
  | 160 MB   |      20 |          31343 |        31440 |
  | 160 MB   |      30 |          28961 |        28987 |
  | 160 MB   |      40 |          26902 |        26970 |
  | 160 MB   |      50 |          25760 |        25810 |
  ------------------------------------------------------
  | 1.6 GB   |       1 |           7729 |         7537 |
  | 1.6 GB   |       2 |          19009 |        23508 | + 23.7%
  | 1.6 GB   |       4 |          33185 |        32666 |
  | 1.6 GB   |       8 |          34550 |        34318 |
  | 1.6 GB   |      16 |          33079 |        32689 |
  | 1.6 GB   |      20 |          31494 |        31702 |
  | 1.6 GB   |      30 |          28535 |        28755 |
  | 1.6 GB   |      40 |          27054 |        27017 |
  | 1.6 GB   |      50 |          25591 |        25560 |
  ------------------------------------------------------
  | 7.6 GB   |       1 |           6224 |         7469 | + 20.0%
  | 7.6 GB   |       2 |          13611 |        12778 |
  | 7.6 GB   |       4 |          33108 |        32927 |
  | 7.6 GB   |       8 |          34712 |        34878 |
  | 7.6 GB   |      16 |          32895 |        33003 |
  | 7.6 GB   |      20 |          31689 |        31974 |
  | 7.6 GB   |      30 |          29003 |        28806 |
  | 7.6 GB   |      40 |          26683 |        26976 |
  | 7.6 GB   |      50 |          25925 |        25652 |
  ------------------------------------------------------

For the aim7 worloads, they overall improved on top of Michel's
patchset.  For full graphs on how the rwsem series plus this patch
behaves on a large 8 socket machine against a vanilla kernel:

  http://stgolabs.net/rwsem-aim7-results.tar.gzSigned-off-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Change explicit "signed long" declarations into plain "long" as suggested
by Peter Hurley.
Signed-off-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Reviewed-by: NMichel Lespinasse <walken@google.com>
Signed-off-by: NMichel Lespinasse <walken@google.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This change fixes a race condition where a reader might determine it
needs to block, but by the time it acquires the wait_lock the rwsem has
active readers and no queued waiters.

In this situation the reader can run in parallel with the existing
active readers; it does not need to block until the active readers
complete.

Thanks to Peter Hurley for noticing this possible race.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When we decide to wake up readers, we must first grant them as many read
locks as necessary, and then actually wake up all these readers.  But in
order to know how many read shares to grant, we must first count the
readers at the head of the queue.  This might take a while if there are
many readers, and we want to be protected against a writer stealing the
lock while we're counting.  To that end, we grant the first reader lock
before counting how many more readers are queued.

We also require some adjustments to the wake_type semantics.

RWSEM_WAKE_NO_ACTIVE used to mean that we had found the count to be
RWSEM_WAITING_BIAS, in which case the rwsem was known to be free as
nobody could steal it while we hold the wait_lock.  This doesn't make
sense once we implement fastpath write lock stealing, so we now use
RWSEM_WAKE_ANY in that case.

Similarly, when rwsem_down_write_failed found that a read lock was
active, it would use RWSEM_WAKE_READ_OWNED which signalled that new
readers could be woken without checking first that the rwsem was
available.  We can't do that anymore since the existing readers might
release their read locks, and a writer could steal the lock before we
wake up additional readers.  So, we have to use a new RWSEM_WAKE_READERS
value to indicate we only want to wake readers, but we don't currently
hold any read lock.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This is mostly for cleanup value:

- We don't need several gotos to handle the case where the first
  waiter is a writer. Two simple tests will do (and generate very
  similar code).

- In the remainder of the function, we know the first waiter is a reader,
  so we don't have to double check that. We can use do..while loops
  to iterate over the readers to wake (generates slightly better code).
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We can skip the initial trylock in rwsem_down_write_failed() if there
are known active lockers already, thus saving one likely-to-fail
cmpxchg.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Acked-by: NRik van Riel <riel@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In rwsem_down_write_failed(), if there are active locks after we wake up
(i.e.  the lock got stolen from us), skip taking the wait_lock and go
back to sleep immediately.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Acked-by: NRik van Riel <riel@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Using rwsem_atomic_update to try stealing the write lock forced us to
undo the adjustment in the failure path.  We can have simpler and faster
code by using cmpxchg instead.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Acked-by: NRik van Riel <riel@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Some small code simplifications can be achieved by doing more agressive
lock stealing:

- When rwsem_down_write_failed() notices that there are no active locks
  (and thus no thread to wake us if we decided to sleep), it used to wake
  the first queued process. However, stealing the lock is also sufficient
  to deal with this case, so we don't need this check anymore.

- In try_get_writer_sem(), we can steal the lock even when the first waiter
  is a reader. This is correct because the code path that wakes readers is
  protected by the wait_lock. As to the performance effects of this change,
  they are expected to be minimal: readers are still granted the lock
  (rather than having to acquire it themselves) when they reach the front
  of the wait queue, so we have essentially the same behavior as in
  rwsem-spinlock.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When waking writers, we never grant them the lock - instead, they have
to acquire it themselves when they run, and remove themselves from the
wait_list when they succeed.

As a result, we can do a few simplifications in rwsem_down_write_failed():

- We don't need to check for !waiter.task since __rwsem_do_wake() doesn't
  remove writers from the wait_list

- There is no point releaseing the wait_lock before entering the wait loop,
  as we will need to reacquire it immediately. We can change the loop so
  that the lock is always held at the start of each loop iteration.

- We don't need to get a reference on the task structure, since the task
  is responsible for removing itself from the wait_list. There is no risk,
  like in the rwsem_down_read_failed() case, that a task would wake up and
  exit (thus destroying its task structure) while __rwsem_do_wake() is
  still running - wait_lock protects against that.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When trying to acquire a read lock, the RWSEM_ACTIVE_READ_BIAS
adjustment doesn't cause other readers to block, so we never have to
worry about waking them back after canceling this adjustment in
rwsem_down_read_failed().

We also never want to steal the lock in rwsem_down_read_failed(), so we
don't have to grab the wait_lock either.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Remove the rwsem_down_failed_common function and replace it with two
identical copies of its code in rwsem_down_{read,write}_failed.

This is because we want to make different optimizations in
rwsem_down_{read,write}_failed; we are adding this pure-duplication
step as a separate commit in order to make it easier to check the
following steps.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This change reduces the size of the spinlocked and TASK_UNINTERRUPTIBLE
sections in rwsem_down_failed_common():

- We only need the sem->wait_lock to insert ourselves on the wait_list;
  the waiter node can be prepared outside of the wait_lock.

- The task state only needs to be set to TASK_UNINTERRUPTIBLE immediately
  before checking if we actually need to sleep; it doesn't need to protect
  the entire function.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We are not planning to add some new waiter flags, so we can convert the
waiter type into an enumeration.

Background: David Howells suggested I do this back when I tried adding
a new waiter type for unfair readers. However, I believe the cleanup
applies regardless of that use case.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Reviewed-by: NRik van Riel <riel@redhat.com>
Reviewed-by: NPeter Hurley <peter@hurleysoftware.com>
Acked-by: NDavidlohr Bueso <davidlohr.bueso@hp.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Commit 5a505085 ("mm/rmap: Convert the struct anon_vma::mutex
to an rwsem") changed struct anon_vma::mutex to an rwsem, which
caused aim7 fork_test performance to drop by 50%.

Yuanhan Liu did the following excellent analysis:

    https://lkml.org/lkml/2013/1/29/84

and found that the regression is caused by strict, serialized,
FIFO sequential write-ownership of rwsems. Ingo suggested
implementing opportunistic lock-stealing for the front writer
task in the waitqueue.

Yuanhan Liu implemented lock-stealing for spinlock-rwsems,
which indeed recovered much of the regression - confirming
the analysis that the main factor in the regression was the
FIFO writer-fairness of rwsems.

In this patch we allow lock-stealing to happen when the first
waiter is also writer. With that change in place the
aim7 fork_test performance is fully recovered on my
Intel NHM EP, NHM EX, SNB EP 2S and 4S test-machines.

Reported-by: lkp@linux.intel.com
Reported-by: NYuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: NAlex Shi <alex.shi@intel.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Michel Lespinasse <walken@google.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Anton Blanchard <anton@samba.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: paul.gortmaker@windriver.com
Link: https://lkml.org/lkml/2013/1/29/84
Link: http://lkml.kernel.org/r/1360069915-31619-1-git-send-email-alex.shi@intel.com
[ Small stylistic fixes, updated changelog. ]
Signed-off-by: NIngo Molnar <mingo@kernel.org>

For files only using THIS_MODULE and/or EXPORT_SYMBOL, map
them onto including export.h -- or if the file isn't even
using those, then just delete the include.  Fix up any implicit
include dependencies that were being masked by module.h along
the way.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

There is no reason to allow the lock protecting rwsems (the
ownerless variant) to be preemptible on -rt. Convert it to raw.

In mainline this change documents the low level nature of
the lock - otherwise there's no functional difference. Lockdep
and Sparse checking will work as usual.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Peter Zijlstra pointed out, that the only user of asmregparm (x86) is
compiling the kernel already with -mregparm=3. So the annotation of
the rwsem functions is redundant. Remove it.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: David Miller <davem@davemloft.net>
Cc: Chris Zankel <chris@zankel.net>
LKML-Reference: <alpine.LFD.2.00.1101262130450.31804@localhost6.localdomain6>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>

More code can be pushed from rwsem_down_read_failed and
rwsem_down_write_failed into rwsem_down_failed_common.

Following change adding down_read_critical infrastructure support also
enjoys having flags available in a register rather than having to fish it
out in the struct rwsem_waiter...
Signed-off-by: NMichel Lespinasse <walken@google.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This change addresses the following situation:

- Thread A acquires the rwsem for read
- Thread B tries to acquire the rwsem for write, notices there is already
  an active owner for the rwsem.
- Thread C tries to acquire the rwsem for read, notices that thread B already
  tried to acquire it.
- Thread C grabs the spinlock and queues itself on the wait queue.
- Thread B grabs the spinlock and queues itself behind C. At this point A is
  the only remaining active owner on the rwsem.

In this situation thread B could notice that it was the last active writer
on the rwsem, and decide to wake C to let it proceed in parallel with A
since they both only want the rwsem for read.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Previously each waiting thread added a bias of RWSEM_WAITING_BIAS.  With
this change, the bias is added only once to indicate that the wait list is
non-empty.

This has a few nice properties which will be used in following changes:
- when the spinlock is held and the waiter list is known to be non-empty,
  count < RWSEM_WAITING_BIAS  <=>  there is an active writer on that sem
- count == RWSEM_WAITING_BIAS  <=>  there are waiting threads and no
                                     active readers/writers on that sem
Signed-off-by: NMichel Lespinasse <walken@google.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In __rwsem_do_wake(), we can skip the active count check unless we come
there from up_xxxx().  Also when checking the active count, it is not
actually necessary to increment it; this allows us to get rid of the read
side undo code and simplify the calculation of the final rwsem count
adjustment once we've counted the reader threads to wake.

The basic observation is the following.  When there are waiter threads on
a rwsem and the spinlock is held, other threads can only increment the
active count by trying to grab the rwsem in down_xxxx().  However
down_xxxx() will notice there are waiter threads and take the down_failed
path, blocking to acquire the spinlock on the way there.  Therefore, a
thread observing an active count of zero with waiters queued and the
spinlock held, is protected against other threads acquiring the rwsem
until it wakes the last waiter or releases the spinlock.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This is in preparation for later changes in the series.

In __rwsem_do_wake(), the first queued waiter is checked first in order to
determine whether it's a writer or a reader.  The code paths diverge at
this point.  The code that checks and increments the rwsem active count is
duplicated on both sides - the point is that later changes in the series
will be able to independently modify both sides.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Acked-by: NDavid Howells <dhowells@redhat.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Suleiman Souhlal <suleiman@google.com>
Cc: Ying Han <yinghan@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

If there are no active threasd using a semaphore, it is always correct
to unqueue blocked threads.  This seems to be what was intended in the
undo code.

What was done instead, was to look for a sem count of zero - this is an
impossible situation, given that at least one thread is known to be
queued on the semaphore.  The code might be correct as written, but it's
hard to reason about and it's not what was intended (otherwise the goto
out would have been unconditional).

Go for checking the active count - the alternative is not worth the
headache.
Signed-off-by: NMichel Lespinasse <walken@google.com>
Signed-off-by: NDavid Howells <dhowells@redhat.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

introduce the "asmregparm" calling convention: for functions
implemented in assembly with a fixed regparm input parameters
calling convention.

mark the semaphore and rwsem slowpath functions with that.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>

This following commit

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=fdf8cb0909b531f9ae8f9b9d7e4eb35ba3505f07

un-inlined a low-level rwsem function, but did not mark it as __sched.
The result is that it now shows up as thread wchan (which also affects
/proc/profile stats).  The following simple patch fixes this by properly
marking rwsem_down_failed_common() as a __sched function.

Also in this patch, which is up for discussion, marks down_read() and
down_write() proper as __sched.  For profiling, it is pretty much
useless to know that a semaphore is beig help - it is necessary to know
_which_ one.  By going up another frame on the stack, the information
becomes much more useful.

In summary, the below change to lib/rwsem.c should be applied; the
changes to kernel/rwsem.c could be applied if other kernel hackers agree
with my proposal that down_read()/down_write() in the profile is not
enough.

[ akpm@linux-foundation.org: build fix ]
Signed-off-by: NLivio Soares <livio@eecg.toronto.edu>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

This annotation makes it possible to assign a subclass on lock init. This
annotation is meant to reduce the _nested() annotations by assigning a
default subclass.

One could do without this annotation and rely on lockdep_set_class()
exclusively, but that would require a manual stack of struct lock_class_key
objects.
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: NDmitry Torokhov <dtor@mail.ru>

Un-inlining rwsem_down_failed_common() (two callsites) reduced lib/rwsem.o
on my Athlon, gcc 4.1.2 from 5935 to 5480 Bytes (455 Bytes saved).

I thus guess that reduced icache footprint (and better function caching) is
worth more than any function call overhead.
Signed-off-by: NAndreas Mohr <andi@lisas.de>
Cc: David Howells <dhowells@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Use the lock validator framework to prove rwsem locking correctness.
Signed-off-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>