Injects EDEADLK conditions at pseudo-random interval, with
exponential backoff up to UINT_MAX (to ensure that every lock
operation still completes in a reasonable time).
This way we can test the wound slowpath even for ww mutex users
where contention is never expected, and the ww deadlock
avoidance algorithm is only needed for correctness against
malicious userspace. An example would be protecting kernel
modesetting properties, which thanks to single-threaded X isn't
really expected to contend, ever.
I've looked into using the CONFIG_FAULT_INJECTION
infrastructure, but decided against it for two reasons:
- EDEADLK handling is mandatory for ww mutex users and should
never affect the outcome of a syscall. This is in contrast to -ENOMEM
injection. So fine configurability isn't required.
- The fault injection framework only allows to set a simple
probability for failure. Now the probability that a ww mutex acquire
stage with N locks will never complete (due to too many injected
EDEADLK backoffs) is zero. But the expected number of ww_mutex_lock
operations for the completely uncontended case would be O(exp(N)).
The per-acuiqire ctx exponential backoff solution choosen here only
results in O(log N) overhead due to injection and so O(log N * N)
lock operations. This way we can fail with high probability (and so
have good test coverage even for fancy backoff and lock acquisition
paths) without running into patalogical cases.
Note that EDEADLK will only ever be injected when we managed to
acquire the lock. This prevents any behaviour changes for users
which rely on the EALREADY semantics.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: dri-devel@lists.freedesktop.org
Cc: linaro-mm-sig@lists.linaro.org
Cc: rostedt@goodmis.org
Cc: daniel@ffwll.ch
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20130620113117.4001.21681.stgit@patser
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The current mutex spinning code (with MUTEX_SPIN_ON_OWNER option
turned on) allow multiple tasks to spin on a single mutex
concurrently. A potential problem with the current approach is
that when the mutex becomes available, all the spinning tasks
will try to acquire the mutex more or less simultaneously. As a
result, there will be a lot of cacheline bouncing especially on
systems with a large number of CPUs.
This patch tries to reduce this kind of contention by putting
the mutex spinners into a queue so that only the first one in
the queue will try to acquire the mutex. This will reduce
contention and allow all the tasks to move forward faster.
The queuing of mutex spinners is done using an MCS lock based
implementation which will further reduce contention on the mutex
cacheline than a similar ticket spinlock based implementation.
This patch will add a new field into the mutex data structure
for holding the MCS lock. This expands the mutex size by 8 bytes
for 64-bit system and 4 bytes for 32-bit system. This overhead
will be avoid if the MUTEX_SPIN_ON_OWNER option is turned off.
The following table shows the jobs per minute (JPM) scalability
data on an 8-node 80-core Westmere box with a 3.7.10 kernel. The
numactl command is used to restrict the running of the fserver
workloads to 1/2/4/8 nodes with hyperthreading off.
+-----------------+-----------+-----------+-------------+----------+
| Configuration | Mean JPM | Mean JPM | Mean JPM | % Change |
| | w/o patch | patch 1 | patches 1&2 | 1->1&2 |
+-----------------+------------------------------------------------+
| | User Range 1100 - 2000 |
+-----------------+------------------------------------------------+
| 8 nodes, HT off | 227972 | 227237 | 305043 | +34.2% |
| 4 nodes, HT off | 393503 | 381558 | 394650 | +3.4% |
| 2 nodes, HT off | 334957 | 325240 | 338853 | +4.2% |
| 1 node , HT off | 198141 | 197972 | 198075 | +0.1% |
+-----------------+------------------------------------------------+
| | User Range 200 - 1000 |
+-----------------+------------------------------------------------+
| 8 nodes, HT off | 282325 | 312870 | 332185 | +6.2% |
| 4 nodes, HT off | 390698 | 378279 | 393419 | +4.0% |
| 2 nodes, HT off | 336986 | 326543 | 340260 | +4.2% |
| 1 node , HT off | 197588 | 197622 | 197582 | 0.0% |
+-----------------+-----------+-----------+-------------+----------+
At low user range 10-100, the JPM differences were within +/-1%.
So they are not that interesting.
The fserver workload uses mutex spinning extensively. With just
the mutex change in the first patch, there is no noticeable
change in performance. Rather, there is a slight drop in
performance. This mutex spinning patch more than recovers the
lost performance and show a significant increase of +30% at high
user load with the full 8 nodes. Similar improvements were also
seen in a 3.8 kernel.
The table below shows the %time spent by different kernel
functions as reported by perf when running the fserver workload
at 1500 users with all 8 nodes.
+-----------------------+-----------+---------+-------------+
| Function | % time | % time | % time |
| | w/o patch | patch 1 | patches 1&2 |
+-----------------------+-----------+---------+-------------+
| __read_lock_failed | 34.96% | 34.91% | 29.14% |
| __write_lock_failed | 10.14% | 10.68% | 7.51% |
| mutex_spin_on_owner | 3.62% | 3.42% | 2.33% |
| mspin_lock | N/A | N/A | 9.90% |
| __mutex_lock_slowpath | 1.46% | 0.81% | 0.14% |
| _raw_spin_lock | 2.25% | 2.50% | 1.10% |
+-----------------------+-----------+---------+-------------+
The fserver workload for an 8-node system is dominated by the
contention in the read/write lock. Mutex contention also plays a
role. With the first patch only, mutex contention is down (as
shown by the __mutex_lock_slowpath figure) which help a little
bit. We saw only a few percents improvement with that.
By applying patch 2 as well, the single mutex_spin_on_owner
figure is now split out into an additional mspin_lock figure.
The time increases from 3.42% to 11.23%. It shows a great
reduction in contention among the spinners leading to a 30%
improvement. The time ratio 9.9/2.33=4.3 indicates that there
are on average 4+ spinners waiting in the spin_lock loop for
each spinner in the mutex_spin_on_owner loop. Contention in
other locking functions also go down by quite a lot.
The table below shows the performance change of both patches 1 &
2 over patch 1 alone in other AIM7 workloads (at 8 nodes,
hyperthreading off).
+--------------+---------------+----------------+-----------------+
| Workload | mean % change | mean % change | mean % change |
| | 10-100 users | 200-1000 users | 1100-2000 users |
+--------------+---------------+----------------+-----------------+
| alltests | 0.0% | -0.8% | +0.6% |
| five_sec | -0.3% | +0.8% | +0.8% |
| high_systime | +0.4% | +2.4% | +2.1% |
| new_fserver | +0.1% | +14.1% | +34.2% |
| shared | -0.5% | -0.3% | -0.4% |
| short | -1.7% | -9.8% | -8.3% |
+--------------+---------------+----------------+-----------------+
The short workload is the only one that shows a decline in
performance probably due to the spinner locking and queuing
overhead.
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Reviewed-by: Davidlohr Bueso <davidlohr.bueso@hp.com>
Acked-by: Rik van Riel <riel@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Chandramouleeswaran Aswin <aswin@hp.com>
Cc: Norton Scott J <scott.norton@hp.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Clark Williams <williams@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1366226594-5506-4-git-send-email-Waiman.Long@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In the __mutex_lock_common() function, an initial entry into
the lock slow path will cause two atomic_xchg instructions to be
issued. Together with the atomic decrement in the fast path, a
total of three atomic read-modify-write instructions will be
issued in rapid succession. This can cause a lot of cache
bouncing when many tasks are trying to acquire the mutex at the
same time.
This patch will reduce the number of atomic_xchg instructions
used by checking the counter value first before issuing the
instruction. The atomic_read() function is just a simple memory
read. The atomic_xchg() function, on the other hand, can be up
to 2 order of magnitude or even more in cost when compared with
atomic_read(). By using atomic_read() to check the value first
before calling atomic_xchg(), we can avoid a lot of unnecessary
cache coherency traffic. The only downside with this change is
that a task on the slow path will have a tiny bit less chance of
getting the mutex when competing with another task in the fast
path.
The same is true for the atomic_cmpxchg() function in the
mutex-spin-on-owner loop. So an atomic_read() is also performed
before calling atomic_cmpxchg().
The mutex locking and unlocking code for the x86 architecture
can allow any negative number to be used in the mutex count to
indicate that some tasks are waiting for the mutex. I am not so
sure if that is the case for the other architectures. So the
default is to avoid atomic_xchg() if the count has already been
set to -1. For x86, the check is modified to include all
negative numbers to cover a larger case.
The following table shows the jobs per minutes (JPM) scalability
data on an 8-node 80-core Westmere box with a 3.7.10 kernel. The
numactl command is used to restrict the running of the
high_systime workloads to 1/2/4/8 nodes with hyperthreading on
and off.
+-----------------+-----------+------------+----------+
| Configuration | Mean JPM | Mean JPM | % Change |
| | w/o patch | with patch | |
+-----------------+-----------------------------------+
| | User Range 1100 - 2000 |
+-----------------+-----------------------------------+
| 8 nodes, HT on | 36980 | 148590 | +301.8% |
| 8 nodes, HT off | 42799 | 145011 | +238.8% |
| 4 nodes, HT on | 61318 | 118445 | +51.1% |
| 4 nodes, HT off | 158481 | 158592 | +0.1% |
| 2 nodes, HT on | 180602 | 173967 | -3.7% |
| 2 nodes, HT off | 198409 | 198073 | -0.2% |
| 1 node , HT on | 149042 | 147671 | -0.9% |
| 1 node , HT off | 126036 | 126533 | +0.4% |
+-----------------+-----------------------------------+
| | User Range 200 - 1000 |
+-----------------+-----------------------------------+
| 8 nodes, HT on | 41525 | 122349 | +194.6% |
| 8 nodes, HT off | 49866 | 124032 | +148.7% |
| 4 nodes, HT on | 66409 | 106984 | +61.1% |
| 4 nodes, HT off | 119880 | 130508 | +8.9% |
| 2 nodes, HT on | 138003 | 133948 | -2.9% |
| 2 nodes, HT off | 132792 | 131997 | -0.6% |
| 1 node , HT on | 116593 | 115859 | -0.6% |
| 1 node , HT off | 104499 | 104597 | +0.1% |
+-----------------+------------+-----------+----------+
At low user range 10-100, the JPM differences were within +/-1%.
So they are not that interesting.
AIM7 benchmark run has a pretty large run-to-run variance due to
random nature of the subtests executed. So a difference of less
than +-5% may not be really significant.
This patch improves high_systime workload performance at 4 nodes
and up by maintaining transaction rates without significant
drop-off at high node count. The patch has practically no
impact on 1 and 2 nodes system.
The table below shows the percentage time (as reported by perf
record -a -s -g) spent on the __mutex_lock_slowpath() function
by the high_systime workload at 1500 users for 2/4/8-node
configurations with hyperthreading off.
+---------------+-----------------+------------------+---------+
| Configuration | %Time w/o patch | %Time with patch | %Change |
+---------------+-----------------+------------------+---------+
| 8 nodes | 65.34% | 0.69% | -99% |
| 4 nodes | 8.70% | 1.02% | -88% |
| 2 nodes | 0.41% | 0.32% | -22% |
+---------------+-----------------+------------------+---------+
It is obvious that the dramatic performance improvement at 8
nodes was due to the drastic cut in the time spent within the
__mutex_lock_slowpath() function.
The table below show the improvements in other AIM7 workloads
(at 8 nodes, hyperthreading off).
+--------------+---------------+----------------+-----------------+
| Workload | mean % change | mean % change | mean % change |
| | 10-100 users | 200-1000 users | 1100-2000 users |
+--------------+---------------+----------------+-----------------+
| alltests | +0.6% | +104.2% | +185.9% |
| five_sec | +1.9% | +0.9% | +0.9% |
| fserver | +1.4% | -7.7% | +5.1% |
| new_fserver | -0.5% | +3.2% | +3.1% |
| shared | +13.1% | +146.1% | +181.5% |
| short | +7.4% | +5.0% | +4.2% |
+--------------+---------------+----------------+-----------------+
Signed-off-by: Waiman Long <Waiman.Long@hp.com>
Reviewed-by: Davidlohr Bueso <davidlohr.bueso@hp.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Chandramouleeswaran Aswin <aswin@hp.com>
Cc: Norton: Scott J <scott.norton@hp.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Clark Williams <williams@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1366226594-5506-3-git-send-email-Waiman.Long@hp.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The changed files were only including linux/module.h for the
EXPORT_SYMBOL infrastructure, and nothing else. Revector them
onto the isolated export header for faster compile times.
Nothing to see here but a whole lot of instances of:
-#include <linux/module.h>
+#include <linux/export.h>
This commit is only changing the kernel dir; next targets
will probably be mm, fs, the arch dirs, etc.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
The spinning mutex implementation uses cpu_relax() in busy loops as a
compiler barrier. Depending on the architecture, cpu_relax() may do more
than needed in this specific mutex spin loops. On System z we also give
up the time slice of the virtual cpu in cpu_relax(), which prevents
effective spinning on the mutex.
This patch replaces cpu_relax() in the spinning mutex code with
arch_mutex_cpu_relax(), which can be defined by each architecture that
selects HAVE_ARCH_MUTEX_CPU_RELAX. The default is still cpu_relax(), so
this patch should not affect other architectures than System z for now.
Signed-off-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1290437256.7455.4.camel@thinkpad>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Fix kernel-doc notation in linux/mutex.h and kernel/mutex.c,
then add these 2 files to the kernel-locking docbook as the
Mutex API reference chapter.
Add one API function to mutex-design.txt and correct a typo in
that file.
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
LKML-Reference: <20100902154816.6cc2f9ad.randy.dunlap@oracle.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Currently, we can hit a nasty case with optimistic
spinning on mutexes:
CPU A tries to take a mutex, while holding the BKL
CPU B tried to take the BLK while holding the mutex
This looks like a AB-BA scenario but in practice, is
allowed and happens due to the auto-release on
schedule() nature of the BKL.
In that case, the optimistic spinning code can get us
into a situation where instead of going to sleep, A
will spin waiting for B who is spinning waiting for
A, and the only way out of that loop is the
need_resched() test in mutex_spin_on_owner().
This patch fixes it by completely disabling spinning
if we own the BKL. This adds one more detail to the
extensive list of reasons why it's a bad idea for
kernel code to be holding the BKL.
Signed-off-by: Tony Breeds <tony@bakeyournoodle.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: <stable@kernel.org>
LKML-Reference: <20100519054636.GC12389@ozlabs.org>
[ added an unlikely() attribute to the branch ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Merge reason: we moved a mutex.h commit that originated from the
perfcounters tree into core/locking - but now merge
back that branch to solve a merge artifact and to
pick up cleanups of this commit that happened in
core/locking.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
