The current quiescent-state detection algorithm is needlessly
complex. It records the grace-period number corresponding to
the quiescent state at the time of the quiescent state, which
works, but it seems better to simply erase any record of previous
quiescent states at the time that the CPU notices the new grace
period. This has the further advantage of removing another piece
of RCU for which lockless reasoning is required.
Therefore, this commit makes this change.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The synchronize_rcu_expedited() function disables interrupts across a
scan of all leaf rcu_node structures, which is not good for real-time
scheduling latency on large systems (hundreds or especially thousands
of CPUs). This commit therefore holds off CPU-hotplug operations using
get_online_cpus(), and removes the prior acquisiion of the ->onofflock
(which required disabling interrupts).
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
In the C language, signed overflow is undefined. It is true that
twos-complement arithmetic normally comes to the rescue, but if the
compiler can subvert this any time it has any information about the values
being compared. For example, given "if (a - b > 0)", if the compiler
has enough information to realize that (for example) the value of "a"
is positive and that of "b" is negative, the compiler is within its
rights to optimize to a simple "if (1)", which might not be what you want.
This commit therefore converts synchronize_rcu_expedited()'s work-done
detection counter from signed to unsigned.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Now that the rcu_node structures' ->completed fields are unconditionally
assigned at grace-period cleanup time, they should already have the
correct value for the new grace period at grace-period initialization
time. This commit therefore inserts a WARN_ON_ONCE() to verify this
invariant.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Preemption greatly raised the probability of certain types of race
conditions, so this commit adds an anti-heisenbug to greatly increase
the collision cross section, also known as the probability of occurrence.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The current approach to grace-period initialization is vulnerable to
extremely low-probability races. These races stem from the fact that
the old grace period is marked completed on the same traversal through
the rcu_node structure that is marking the start of the new grace period.
This means that some rcu_node structures will believe that the old grace
period is still in effect at the same time that other rcu_node structures
believe that the new grace period has already started.
These sorts of disagreements can result in too-short grace periods,
as shown in the following scenario:
1. CPU 0 completes a grace period, but needs an additional
grace period, so starts initializing one, initializing all
the non-leaf rcu_node structures and the first leaf rcu_node
structure. Because CPU 0 is both completing the old grace
period and starting a new one, it marks the completion of
the old grace period and the start of the new grace period
in a single traversal of the rcu_node structures.
Therefore, CPUs corresponding to the first rcu_node structure
can become aware that the prior grace period has completed, but
CPUs corresponding to the other rcu_node structures will see
this same prior grace period as still being in progress.
2. CPU 1 passes through a quiescent state, and therefore informs
the RCU core. Because its leaf rcu_node structure has already
been initialized, this CPU's quiescent state is applied to the
new (and only partially initialized) grace period.
3. CPU 1 enters an RCU read-side critical section and acquires
a reference to data item A. Note that this CPU believes that
its critical section started after the beginning of the new
grace period, and therefore will not block this new grace period.
4. CPU 16 exits dyntick-idle mode. Because it was in dyntick-idle
mode, other CPUs informed the RCU core of its extended quiescent
state for the past several grace periods. This means that CPU 16
is not yet aware that these past grace periods have ended. Assume
that CPU 16 corresponds to the second leaf rcu_node structure --
which has not yet been made aware of the new grace period.
5. CPU 16 removes data item A from its enclosing data structure
and passes it to call_rcu(), which queues a callback in the
RCU_NEXT_TAIL segment of the callback queue.
6. CPU 16 enters the RCU core, possibly because it has taken a
scheduling-clock interrupt, or alternatively because it has
more than 10,000 callbacks queued. It notes that the second
most recent grace period has completed (recall that because it
corresponds to the second as-yet-uninitialized rcu_node structure,
it cannot yet become aware that the most recent grace period has
completed), and therefore advances its callbacks. The callback
for data item A is therefore in the RCU_NEXT_READY_TAIL segment
of the callback queue.
7. CPU 0 completes initialization of the remaining leaf rcu_node
structures for the new grace period, including the structure
corresponding to CPU 16.
8. CPU 16 again enters the RCU core, again, possibly because it has
taken a scheduling-clock interrupt, or alternatively because
it now has more than 10,000 callbacks queued. It notes that
the most recent grace period has ended, and therefore advances
its callbacks. The callback for data item A is therefore in
the RCU_DONE_TAIL segment of the callback queue.
9. All CPUs other than CPU 1 pass through quiescent states. Because
CPU 1 already passed through its quiescent state, the new grace
period completes. Note that CPU 1 is still in its RCU read-side
critical section, still referencing data item A.
10. Suppose that CPU 2 wais the last CPU to pass through a quiescent
state for the new grace period, and suppose further that CPU 2
did not have any callbacks queued, therefore not needing an
additional grace period. CPU 2 therefore traverses all of the
rcu_node structures, marking the new grace period as completed,
but does not initialize a new grace period.
11. CPU 16 yet again enters the RCU core, yet again possibly because
it has taken a scheduling-clock interrupt, or alternatively
because it now has more than 10,000 callbacks queued. It notes
that the new grace period has ended, and therefore advances
its callbacks. The callback for data item A is therefore in
the RCU_DONE_TAIL segment of the callback queue. This means
that this callback is now considered ready to be invoked.
12. CPU 16 invokes the callback, freeing data item A while CPU 1
is still referencing it.
This scenario represents a day-zero bug for TREE_RCU. This commit
therefore ensures that the old grace period is marked completed in
all leaf rcu_node structures before a new grace period is marked
started in any of them.
That said, it would have been insanely difficult to force this race to
happen before the grace-period initialization process was preemptible.
Therefore, this commit is not a candidate for -stable.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Conflicts:
kernel/rcutree.c
The module parameters blimit, qhimark, and qlomark (and more
recently, rcu_fanout_leaf) have permission masks of zero, so
that their values are not visible from sysfs. This is unnecessary
and inconvenient to administrators who might like an easy way to
see what these values are on a running system. This commit therefore
sets their permission masks to 0444, allowing them to be read but
not written.
Reported-by: Rusty Russell <rusty@ozlabs.org>
Reported-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Although almost everyone is well-served by the defaults, some uses of RCU
benefit from shorter grace periods, while others benefit more from the
greater efficiency provided by longer grace periods. Situations requiring
a large number of grace periods to elapse (and wireshark startup has
been called out as an example of this) are helped by lower-latency
grace periods. Furthermore, in some embedded applications, people are
willing to accept a small degradation in update efficiency (due to there
being more of the shorter grace-period operations) in order to gain the
lower latency.
In contrast, those few systems with thousands of CPUs need longer grace
periods because the CPU overhead of a grace period rises roughly
linearly with the number of CPUs. Such systems normally do not make
much use of facilities that require large numbers of grace periods to
elapse, so this is a good tradeoff.
Therefore, this commit allows the durations to be controlled from sysfs.
There are two sysfs parameters, one named "jiffies_till_first_fqs" that
specifies the delay in jiffies from the end of grace-period initialization
until the first attempt to force quiescent states, and the other named
"jiffies_till_next_fqs" that specifies the delay (again in jiffies)
between subsequent attempts to force quiescent states. They both default
to three jiffies, which is compatible with the old hard-coded behavior.
At some future time, it may be possible to automatically increase the
grace-period length with the number of CPUs, but we do not yet have
sufficient data to do a good job. Preliminary data indicates that we
should add an addiitonal jiffy to each of the delays for every 200 CPUs
in the system, but more experimentation is needed. For now, the number
of systems with more than 1,000 CPUs is small enough that this can be
relegated to boot-time hand tuning.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Large systems running RCU_FAST_NO_HZ kernels see extreme memory
contention on the rcu_state structure's ->fqslock field. This
can be avoided by disabling RCU_FAST_NO_HZ, either at compile time
or at boot time (via the nohz kernel boot parameter), but large
systems will no doubt become sensitive to energy consumption.
This commit therefore uses a combining-tree approach to spread the
memory contention across new cache lines in the leaf rcu_node structures.
This can be thought of as a tournament lock that has only a try-lock
acquisition primitive.
The effect on small systems is minimal, because such systems have
an rcu_node "tree" consisting of a single node. In addition, this
functionality is not used on fastpaths.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Moving quiescent-state forcing into a kthread dispenses with the need
for the ->n_rp_need_fqs field, so this commit removes it.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
RCU quiescent-state forcing is currently carried out without preemption
points, which can result in excessive latency spikes on large systems
(many hundreds or thousands of CPUs). This patch therefore inserts
a voluntary preemption point into force_qs_rnp(), which should greatly
reduce the magnitude of these spikes.
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
As the first step towards allowing quiescent-state forcing to be
preemptible, this commit moves RCU quiescent-state forcing into the
same kthread that is now used to initialize and clean up after grace
periods. This is yet another step towards keeping scheduling
latency down to a dull roar.
Updated to change from raw_spin_lock_irqsave() to raw_spin_lock_irq()
and to remove the now-unused rcu_state structure fields as suggested by
Peter Zijlstra.
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The fields in the rcu_state structure that are protected by the
root rcu_node structure's ->lock can share a cache line with the
fields protected by ->onofflock. This can result in excessive
memory contention on large systems, so this commit applies
____cacheline_internodealigned_in_smp to the ->onofflock field in
order to segregate them.
Signed-off-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
In kernels built with CONFIG_RCU_FAST_NO_HZ=y, CPUs can accumulate a
large number of lazy callbacks, which as the name implies will be slow
to be invoked. This can be a problem on small-memory systems, where the
default 6-second sleep for CPUs having only lazy RCU callbacks could well
be fatal. This commit therefore installs an OOM hander that ensures that
every CPU with lazy callbacks has at least one non-lazy callback, in turn
ensuring timely advancement for these callbacks.
Updated to fix bug that disabled OOM killing, noted by Lai Jiangshan.
Updated to push the for_each_rcu_flavor() loop into rcu_oom_notify_cpu(),
thus reducing the number of IPIs, as suggested by Steven Rostedt. Also
to make the for_each_online_cpu() loop be preemptible. (Later, it might
be good to use smp_call_function(), as suggested by Peter Zijlstra.)
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Sasha Levin <levinsasha928@gmail.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Earlier versions of RCU invoked the RCU core from the CPU_DYING notifier
in order to note a quiescent state for the outgoing CPU. Because the
CPU is marked "offline" during the execution of the CPU_DYING notifiers,
the RCU core had to tolerate being invoked from an offline CPU. However,
commit b1420f1c (Make rcu_barrier() less disruptive) left only tracing
code in the CPU_DYING notifier, so the RCU core need no longer execute
on offline CPUs. This commit therefore enforces this restriction.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Then rcu_gp_kthread() function is too large and furthermore needs to
have the force_quiescent_state() code pulled in. This commit therefore
breaks up rcu_gp_kthread() into rcu_gp_init() and rcu_gp_cleanup().
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
RCU grace-period cleanup is currently carried out with interrupts
disabled, which can result in excessive latency spikes on large systems
(many hundreds or thousands of CPUs). This patch therefore makes the
RCU grace-period cleanup be preemptible, including voluntary preemption
points, which should eliminate those latency spikes. Similar spikes from
forcing of quiescent states will be dealt with similarly by later patches.
Updated to replace uses of spin_lock_irqsave() with spin_lock_irq(), as
suggested by Peter Zijlstra.
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
As a first step towards allowing grace-period cleanup to be preemptible,
this commit moves the RCU grace-period cleanup into the same kthread
that is now used to initialize grace periods. This is needed to keep
scheduling latency down to a dull roar.
[ paulmck: Get rid of stray spin_lock_irqsave() calls. ]
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
RCU grace-period initialization is currently carried out with interrupts
disabled, which can result in 200-microsecond latency spikes on systems
on which RCU has been configured for 4096 CPUs. This patch therefore
makes the RCU grace-period initialization be preemptible, which should
eliminate those latency spikes. Similar spikes from grace-period cleanup
and the forcing of quiescent states will be dealt with similarly by later
patches.
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
The next step in reducing RCU's grace-period initialization latency on
large systems will make this initialization preemptible. Unfortunately,
making the grace-period initialization subject to interrupts (let alone
preemption) exposes the following race on systems whose rcu_node tree
contains more than one node:
1. CPU 31 starts initializing the grace period, including the
first leaf rcu_node structures, and is then preempted.
2. CPU 0 refers to the first leaf rcu_node structure, and notes
that a new grace period has started. It passes through a
quiescent state shortly thereafter, and informs the RCU core
of this rite of passage.
3. CPU 0 enters an RCU read-side critical section, acquiring
a pointer to an RCU-protected data item.
4. CPU 31 takes an interrupt whose handler removes the data item
referenced by CPU 0 from the data structure, and registers an
RCU callback in order to free it.
5. CPU 31 resumes initializing the grace period, including its
own rcu_node structure. In invokes rcu_start_gp_per_cpu(),
which advances all callbacks, including the one registered
in #4 above, to be handled by the current grace period.
6. The remaining CPUs pass through quiescent states and inform
the RCU core, but CPU 0 remains in its RCU read-side critical
section, still referencing the now-removed data item.
7. The grace period completes and all the callbacks are invoked,
including the one that frees the data item that CPU 0 is still
referencing. Oops!!!
One way to avoid this race is to remove grace-period acceleration from
rcu_start_gp_per_cpu(). Now, the only reason for this acceleration was
to allow CPUs bringing RCU out of idle state to have their callbacks
invoked after only one grace period, rather than the two grace periods
that would otherwise be required. But this acceleration does not
work when RCU grace-period initialization is moved to a kthread because
the CPU posting the callback is no longer necessarily the CPU that is
initializing the resulting grace period.
This commit therefore removes this now-pointless (and soon to be dangerous)
grace-period acceleration, thus avoiding the above race.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
As the first step towards allowing grace-period initialization to be
preemptible, this commit moves the RCU grace-period initialization
into its own kthread. This is needed to keep large-system scheduling
latency at reasonable levels.
Also change raw_spin_lock_irqsave() to raw_spin_lock_irq() as suggested
by Peter Zijlstra in review comments.
Reported-by: Mike Galbraith <mgalbraith@suse.de>
Reported-by: Dimitri Sivanich <sivanich@sgi.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
Each grace period is supposed to have at least one callback waiting
for that grace period to complete. However, if CONFIG_NO_HZ=n, an
extra callback-free grace period is no big problem -- it will chew up
a tiny bit of CPU time, but it will complete normally. In contrast,
CONFIG_NO_HZ=y kernels have the potential for all the CPUs to go to
sleep indefinitely, in turn indefinitely delaying completion of the
callback-free grace period. Given that nothing is waiting on this grace
period, this is also not a problem.
That is, unless RCU CPU stall warnings are also enabled, as they are
in recent kernels. In this case, if a CPU wakes up after at least one
minute of inactivity, an RCU CPU stall warning will result. The reason
that no one noticed until quite recently is that most systems have enough
OS noise that they will never remain absolutely idle for a full minute.
But there are some embedded systems with cut-down userspace configurations
that consistently get into this situation.
All this begs the question of exactly how a callback-free grace period
gets started in the first place. This can happen due to the fact that
CPUs do not necessarily agree on which grace period is in progress.
If a CPU still believes that the grace period that just completed is
still ongoing, it will believe that it has callbacks that need to wait for
another grace period, never mind the fact that the grace period that they
were waiting for just completed. This CPU can therefore erroneously
decide to start a new grace period. Note that this can happen in
TREE_RCU and TREE_PREEMPT_RCU even on a single-CPU system: Deadlock
considerations mean that the CPU that detected the end of the grace
period is not necessarily officially informed of this fact for some time.
Once this CPU notices that the earlier grace period completed, it will
invoke its callbacks. It then won't have any callbacks left. If no
other CPU has any callbacks, we now have a callback-free grace period.
This commit therefore makes CPUs check more carefully before starting a
new grace period. This new check relies on an array of tail pointers
into each CPU's list of callbacks. If the CPU is up to date on which
grace periods have completed, it checks to see if any callbacks follow
the RCU_DONE_TAIL segment, otherwise it checks to see if any callbacks
follow the RCU_WAIT_TAIL segment. The reason that this works is that
the RCU_WAIT_TAIL segment will be promoted to the RCU_DONE_TAIL segment
as soon as the CPU is officially notified that the old grace period
has ended.
This change is to cpu_needs_another_gp(), which is called in a number
of places. The only one that really matters is in rcu_start_gp(), where
the root rcu_node structure's ->lock is held, which prevents any
other CPU from starting or completing a grace period, so that the
comparison that determines whether the CPU is missing the completion
of a grace period is stable.
Reported-by: Becky Bruce <bgillbruce@gmail.com>
Reported-by: Subodh Nijsure <snijsure@grid-net.com>
Reported-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Paul Walmsley <paul@pwsan.com> # OMAP3730, OMAP4430
Cc: stable@vger.kernel.org
Pull timer fix from Ingo Molnar:
"One more timekeeping fix for v3.6"
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
time: Fix timeekeping_get_ns overflow on 32bit systems
Pull workqueue / powernow-k8 fix from Tejun Heo:
"This is the fix for the bug where cpufreq/powernow-k8 was tripping
BUG_ON() in try_to_wake_up_local() by migrating workqueue worker to a
different CPU.
https://bugzilla.kernel.org/show_bug.cgi?id=47301
As discussed, the fix is now two parts - one to reimplement
work_on_cpu() so that it doesn't create a new kthread each time and
the actual fix which makes powernow-k8 use work_on_cpu() instead of
performing manual migration.
While pretty late in the merge cycle, both changes are on the safer
side. Jiri and I verified two existing users of work_on_cpu() and
Duncan confirmed that the powernow-k8 fix survived about 18 hours of
testing."
* 'for-3.6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
cpufreq/powernow-k8: workqueue user shouldn't migrate the kworker to another CPU
workqueue: reimplement work_on_cpu() using system_wq
The existing work_on_cpu() implementation is hugely inefficient. It
creates a new kthread, execute that single function and then let the
kthread die on each invocation.
Now that system_wq can handle concurrent executions, there's no
advantage of doing this. Reimplement work_on_cpu() using system_wq
which makes it simpler and way more efficient.
stable: While this isn't a fix in itself, it's needed to fix a
workqueue related bug in cpufreq/powernow-k8. AFAICS, this
shouldn't break other existing users.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Jiri Kosina <jkosina@suse.cz>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Len Brown <lenb@kernel.org>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: stable@vger.kernel.org
Pull another workqueue fix from Tejun Heo:
"Unfortunately, yet another late fix. This too is discovered and fixed
by Lai. This bug was introduced during this merge window by commit
25511a4776 ("workqueue: reimplement CPU online rebinding to handle
idle workers") which started using WORKER_REBIND flag for idle rebind
too.
The bug is relatively easy to trigger if the CPU rapidly goes through
off, on and then off (and stay off). The fix is on the safer side.
This hasn't been on linux-next yet but I'm pushing early so that it
can get more exposure before v3.6 release."
* 'for-3.6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
workqueue: always clear WORKER_REBIND in busy_worker_rebind_fn()
The kernel doesn't check the pid for negative values, so if you try to
write -2 to /proc/sys/kernel/ns_last_pid, you will get a kernel panic.
The crash happens because the next pid is -1, and alloc_pidmap() will
try to access to a nonexistent pidmap.
map = &pid_ns->pidmap[pid/BITS_PER_PAGE];
Signed-off-by: Andrew Vagin <avagin@openvz.org>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@parallels.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit 970e178985.
Nikolay Ulyanitsky reported thatthe 3.6-rc5 kernel has a 15-20%
performance drop on PostgreSQL 9.2 on his machine (running "pgbench").
Borislav Petkov was able to reproduce this, and bisected it to this
commit 970e178985 ("sched: Improve scalability via 'CPU buddies' ...")
apparently because the new single-idle-buddy model simply doesn't find
idle CPU's to reschedule on aggressively enough.
Mike Galbraith suspects that it is likely due to the user-mode spinlocks
in PostgreSQL not reacting well to preemption, but we don't really know
the details - I'll just revert the commit for now.
There are hopefully other approaches to improve scheduler scalability
without it causing these kinds of downsides.
Reported-by: Nikolay Ulyanitsky <lystor@gmail.com>
Bisected-by: Borislav Petkov <bp@alien8.de>
Acked-by: Mike Galbraith <efault@gmx.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
As Oleg pointed out in [0] uprobe should not use the ptrace interface
for enabling/disabling single stepping.
[0] http://lkml.kernel.org/r/20120730141638.GA5306@redhat.com
Add the new "__weak arch" helpers which simply call user_*_single_step()
as a preparation. This is only needed to not break the powerpc port, we
will fold this logic into arch_uprobe_pre/post_xol() hooks later.
We should also change handle_singlestep(), _disable_step(&uprobe->arch)
should be called before put_uprobe().
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
The wrong MMF_HAS_UPROBES doesn't really hurt, just it triggers
the "slow" and unnecessary handle_swbp() path if the task hits
the non-uprobe breakpoint.
So this patch changes find_active_uprobe() to check every valid
vma and clear MMF_HAS_UPROBES if no uprobes were found. This is
adds the slow O(n) path, but it is only called in unlikely case
when the task hits the normal breakpoint first time after
uprobe_unregister().
Note the "not strictly accurate" comment in mmf_recalc_uprobes().
We can fix this, we only need to teach vma_has_uprobes() to return
a bit more more info, but I am not sure this worth the trouble.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Add the new MMF_RECALC_UPROBES flag, it means that MMF_HAS_UPROBES
can be false positive after remove_breakpoint() or uprobe_munmap().
It is also set by uprobe_dup_mmap(), this is not optimal but simple.
We could add the new hook, uprobe_dup_vma(), to set MMF_HAS_UPROBES
only if the new mm actually has uprobes, but I don't think this
makes sense.
The next patch will use this flag to clear MMF_HAS_UPROBES.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Nobody plays with uprobes_tree/uprobes_treelock in interrupt context,
no need to disable irqs.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Pull perf fixes from Ingo Molnar:
"This tree includes various fixes"
Ingo really needs to improve on the whole "explain git pull" part.
"Various fixes" indeed.
* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
perf/hwpb: Invoke __perf_event_disable() if interrupts are already disabled
perf/x86: Enable Intel Cedarview Atom suppport
perf_event: Switch to internal refcount, fix race with close()
oprofile, s390: Fix uninitialized memory access when writing to oprofilefs
perf/x86: Fix microcode revision check for SNB-PEBS
Fix kprobes/x86 to support jprobes on ftrace-based kprobes.
Because of -mfentry support of ftrace, ftrace is now put
on the beginning of function where jprobes are put.
Originally ftrace-based kprobes doesn't support jprobe
because it will change regs->ip and ftrace doesn't support
changing IP and ftrace itself doesn't conflict jprobe.
However, ftrace -mfentry support moves mcount call on the
top of functions where jprobes are put. This means that
jprobe always conflicts with ftrace-based kprobe and fails.
This patch allows ftrace-based kprobes to support jprobes
by allowing to modify regs->ip and kprobes breakpoint
handler also allows to skip singlestepping because there
is a ftrace call (not an original instruction).
Link: http://lkml.kernel.org/r/20120905143125.10329.90836.stgit@localhost.localdomain
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Commit 56449f437 "tracing: make the trace clocks available generally",
in April 2009, made trace_clock available unconditionally, since
CONFIG_X86_DS used it too.
Commit faa4602e47 "x86, perf, bts, mm: Delete the never used BTS-ptrace code",
in March 2010, removed CONFIG_X86_DS, and now only CONFIG_RING_BUFFER (split
out from CONFIG_TRACING for general use) has a dependency on trace_clock. So,
only compile in trace_clock with CONFIG_RING_BUFFER or CONFIG_TRACING
enabled.
Link: http://lkml.kernel.org/r/20120903024513.GA19583@leaf
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Break out the DEBUG_SPINLOCK dependency (requires moving up
UNINLINE_SPIN_UNLOCK, as this was the only one in that block not
depending on that option).
Avoid putting values not selected into the resulting .config -
they are not useful for anything, make the output less legible,
and just consume space: Use "depends on" rather than directly
setting the default from the combined dependency values.
Signed-off-by: Jan Beulich <jbeulich@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/504DF2AC020000780009A2DF@nat28.tlf.novell.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Daniel Lezcano reported seeing multi-second stalls from
keyboard input on his T61 laptop when NOHZ and CPU_IDLE
were enabled on a 32bit kernel.
He bisected the problem down to commit
1e75fa8be9 ("time: Condense timekeeper.xtime into xtime_sec").
After reproducing this issue, I narrowed the problem down
to the fact that timekeeping_get_ns() returns a 64bit
nsec value that hasn't been accumulated. In some cases
this value was being then stored in timespec.tv_nsec
(which is a long).
On 32bit systems, with idle times larger then 4 seconds
(or less, depending on the value of xtime_nsec), the
returned nsec value would overflow 32bits. This limited
kept time from increasing, causing timers to not expire.
The fix is to make sure we don't directly store the
result of timekeeping_get_ns() into a tv_nsec field,
instead using a 64bit nsec value which can then be
added into the timespec via timespec_add_ns().
Reported-and-bisected-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Tested-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Acked-by: Prarit Bhargava <prarit@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Link: http://lkml.kernel.org/r/1347405963-35715-1-git-send-email-john.stultz@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Steve Rostedt asked for the merge of a single commit, into both
the RCU and the perf/tracing tree:
| Josh made a change to the tracing code that affects both the
| work Paul McKenney and I are currently doing. At the last
| Kernel Summit back in August, Linus said when such a case
| exists, it is best to make a separate branch based off of his
| tree and place the change there. This way, the repositories
| that need to share the change can both pull them in and the
| SHA1 will match for both. Whichever branch is pulled in first
| by Linus will also pull in the necessary change for the other
| branch as well.
Signed-off-by: Ingo Molnar <mingo@kernel.org>
It is considered good form to lock the lock you claim to be nested in.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
[ removed nest_lock arg to print_lock_nested_lock_not_held in favour
of hlock->nest_lock, also renamed the lock arg to hlock since its
a held_lock type ]
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/5051A9E7.5040501@canonical.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Heteregeneous ARM platform uses arch_scale_freq_power function
to reflect the relative capacity of each core
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1341826026-6504-6-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
There is no load_balancer to be selected now. It just sets the
state of the nohz tick to stop.
So rename the function, pass the 'cpu' as a parameter and then
remove the useless call from tick_nohz_restart_sched_tick().
[ s/set_nohz_tick_stopped/nohz_balance_enter_idle/g
s/clear_nohz_tick_stopped/nohz_balance_exit_idle/g ]
Signed-off-by: Alex Shi <alex.shi@intel.com>
Acked-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1347261059-24747-1-git-send-email-alex.shi@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Commit f319da0c68 ("sched: Fix load avg vs cpu-hotplug") was an
incomplete fix:
In particular, the problem is that at the point it calls
calc_load_migrate() nr_running := 1 (the stopper thread), so move the
call to CPU_DEAD where we're sure that nr_running := 0.
Also note that we can call calc_load_migrate() without serialization, we
know the state of rq is stable since its cpu is dead, and we modify the
global state using appropriate atomic ops.
Suggested-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1346882630.2600.59.camel@twins
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Now that the last architecture to use this has stopped doing so (ARM,
thanks Catalin!) we can remove this complexity from the scheduler
core.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Link: http://lkml.kernel.org/n/tip-g9p2a1w81xxbrze25v9zpzbf@git.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
On tickless systems, one CPU runs load balance for all idle CPUs.
The cpu_load of this CPU is updated before starting the load balance
of each other idle CPUs. We should instead update the cpu_load of
the balance_cpu.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Venkatesh Pallipadi <venki@google.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/1347509486-8688-1-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
ptrace_notify() and get_signal_to_deliver() do unnecessary things
before task_work_run():
1. smp_mb__after_clear_bit() is not needed, test_and_clear_bit()
implies mb().
2. And we do not need the barrier at all, in this case we only
care about the "synchronous" works added by the task itself.
3. No need to clear TIF_NOTIFY_RESUME, and we should not assume
task_works is the only user of this flag.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Al Viro <viro@zeniv.linux.org.uk>
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/20120826191217.GA4238@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
ed3e694d "move exit_task_work() past exit_files() et.al" destroyed
the add/exit synchronization we had, the caller itself should ensure
task_work_add() can't race with the exiting task.
However, this is not convenient/simple, and the only user which tries
to do this is buggy (see the next patch). Unless the task is current,
there is simply no way to do this in general.
Change exit_task_work()->task_work_run() to use the dummy "work_exited"
entry to let task_work_add() know it should fail.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Al Viro <viro@zeniv.linux.org.uk>
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/20120826191211.GA4228@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Change task_work's to use llist-like code to avoid pi_lock
in task_work_add(), this makes it useable under rq->lock.
task_work_cancel() and task_work_run() still use pi_lock
to synchronize with each other.
(This is in preparation for a deadlock fix.)
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Al Viro <viro@zeniv.linux.org.uk>
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/20120826191209.GA4221@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Pull workqueue fixes from Tejun Heo:
"It's later than I'd like but well the timing just didn't work out this
time.
There are three bug fixes. One from before 3.6-rc1 and two from the
new CPU hotplug code. Kudos to Lai for discovering all of them and
providing fixes.
* Atomicity bug when clearing a flag and setting another. The two
operation should have been atomic but wasn't. This bug has existed
for a long time but is unlikely to have actually happened. Fix is
safe. Marked for -stable.
* If CPU hotplug cycles happen back-to-back before workers finish the
previous cycle, the states could get out of sync and it could get
stuck. Fixed by waiting for workers to complete before finishing
hotplug cycle.
* While CPU hotplug is in progress, idle workers could be depleted
which can then lead to deadlock. I think both happening together
is highly unlikely but still better to fix it and the fix isn't too
scary.
There's another workqueue related regression which reported a few days
ago:
https://bugzilla.kernel.org/show_bug.cgi?id=47301
It's a bit of head scratcher but there is a semi-reliable reproduce
case, so I'm hoping to resolve it soonish."
* 'for-3.6-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
workqueue: fix possible idle worker depletion across CPU hotplug
workqueue: restore POOL_MANAGING_WORKERS
workqueue: fix possible deadlock in idle worker rebinding
workqueue: move WORKER_REBIND clearing in rebind_workers() to the end of the function
workqueue: UNBOUND -> REBIND morphing in rebind_workers() should be atomic
To simplify both normal and CPU hotplug paths, worker management is
prevented while CPU hoplug is in progress. This is achieved by CPU
hotplug holding the same exclusion mechanism used by workers to ensure
there's only one manager per pool.
If someone else seems to be performing the manager role, workers
proceed to execute work items. CPU hotplug using the same mechanism
can lead to idle worker depletion because all workers could proceed to
execute work items while CPU hotplug is in progress and CPU hotplug
itself wouldn't actually perform the worker management duty - it
doesn't guarantee that there's an idle worker left when it releases
management.
This idle worker depletion, under extreme circumstances, can break
forward-progress guarantee and thus lead to deadlock.
This patch fixes the bug by using separate mechanisms for manager
exclusion among workers and hotplug exclusion. For manager exclusion,
POOL_MANAGING_WORKERS which was restored by the previous patch is
used. pool->manager_mutex is now only used for exclusion between the
elected manager and CPU hotplug. The elected manager won't proceed
without holding pool->manager_mutex.
This ensures that the worker which won the manager position can't skip
managing while CPU hotplug is in progress. It will block on
manager_mutex and perform management after CPU hotplug is complete.
Note that hotplug may happen while waiting for manager_mutex. A
manager isn't either on idle or busy list and thus the hoplug code
can't unbind/rebind it. Make the manager handle its own un/rebinding.
tj: Updated comment and description.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
This patch restores POOL_MANAGING_WORKERS which was replaced by
pool->manager_mutex by 6037315269 "workqueue: use mutex for global_cwq
manager exclusion".
There's a subtle idle worker depletion bug across CPU hotplug events
and we need to distinguish an actual manager and CPU hotplug
preventing management. POOL_MANAGING_WORKERS will be used for the
former and manager_mutex the later.
This patch just lays POOL_MANAGING_WORKERS on top of the existing
manager_mutex and doesn't introduce any synchronization changes. The
next patch will update it.
Note that this patch fixes a non-critical anomaly where
too_many_workers() may return %true spuriously while CPU hotplug is in
progress. While the issue could schedule idle timer spuriously, it
didn't trigger any actual misbehavior.
tj: Rewrote patch description.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Currently, rebind_workers() and idle_worker_rebind() are two-way
interlocked. rebind_workers() waits for idle workers to finish
rebinding and rebound idle workers wait for rebind_workers() to finish
rebinding busy workers before proceeding.
Unfortunately, this isn't enough. The second wait from idle workers
is implemented as follows.
wait_event(gcwq->rebind_hold, !(worker->flags & WORKER_REBIND));
rebind_workers() clears WORKER_REBIND, wakes up the idle workers and
then returns. If CPU hotplug cycle happens again before one of the
idle workers finishes the above wait_event(), rebind_workers() will
repeat the first part of the handshake - set WORKER_REBIND again and
wait for the idle worker to finish rebinding - and this leads to
deadlock because the idle worker would be waiting for WORKER_REBIND to
clear.
This is fixed by adding another interlocking step at the end -
rebind_workers() now waits for all the idle workers to finish the
above WORKER_REBIND wait before returning. This ensures that all
rebinding steps are complete on all idle workers before the next
hotplug cycle can happen.
This problem was diagnosed by Lai Jiangshan who also posted a patch to
fix the issue, upon which this patch is based.
This is the minimal fix and further patches are scheduled for the next
merge window to simplify the CPU hotplug path.
Signed-off-by: Tejun Heo <tj@kernel.org>
Original-patch-by: Lai Jiangshan <laijs@cn.fujitsu.com>
LKML-Reference: <1346516916-1991-3-git-send-email-laijs@cn.fujitsu.com>
This doesn't make any functional difference and is purely to help the
next patch to be simpler.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Lai Jiangshan <laijs@cn.fujitsu.com>
The compiler may compile the following code into TWO write/modify
instructions.
worker->flags &= ~WORKER_UNBOUND;
worker->flags |= WORKER_REBIND;
so the other CPU may temporarily see worker->flags which doesn't have
either WORKER_UNBOUND or WORKER_REBIND set and perform local wakeup
prematurely.
Fix it by using single explicit assignment via ACCESS_ONCE().
Because idle workers have another WORKER_NOT_RUNNING flag, this bug
doesn't exist for them; however, update it to use the same pattern for
consistency.
tj: Applied the change to idle workers too and updated comments and
patch description a bit.
Signed-off-by: Lai Jiangshan <laijs@cn.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: stable@vger.kernel.org
While debugging a warning message on PowerPC while using hardware
breakpoints, it was discovered that when perf_event_disable is invoked
through hw_breakpoint_handler function with interrupts disabled, a
subsequent IPI in the code path would trigger a WARN_ON_ONCE message in
smp_call_function_single function.
This patch calls __perf_event_disable() when interrupts are already
disabled, instead of perf_event_disable().
Reported-by: Edjunior Barbosa Machado <emachado@linux.vnet.ibm.com>
Signed-off-by: K.Prasad <Prasad.Krishnan@gmail.com>
[naveen.n.rao@linux.vnet.ibm.com: v3: Check to make sure we target current task]
Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120802081635.5811.17737.stgit@localhost.localdomain
[ Fixed build error on MIPS. ]
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Don't mess with file refcounts (or keep a reference to file, for
that matter) in perf_event. Use explicit refcount of its own
instead. Deal with the race between the final reference to event
going away and new children getting created for it by use of
atomic_long_inc_not_zero() in inherit_event(); just have the
latter free what it had allocated and return NULL, that works
out just fine (children of siblings of something doomed are
created as singletons, same as if the child of leader had been
created and immediately killed).
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: stable@kernel.org
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20120820135925.GG23464@ZenIV.linux.org.uk
Signed-off-by: Ingo Molnar <mingo@kernel.org>
It's impossible to enter the else branch if we have set
skip_clock_update in task_yield_fair(), as yield_to_task_fair()
will directly return true after invoke task_yield_fair().
Signed-off-by: Michael Wang <wangyun@linux.vnet.ibm.com>
Acked-by: Mike Galbraith <efault@gmx.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/4FF2925A.9060005@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Unlike others, sched_migration_cost, sched_time_avg and
sched_shares_window doesn't have time unit as suffix. Add them.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1345083330-19486-1-git-send-email-namhyung@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Various sd->*_idx's are used for refering the rq's load average table
when selecting a cpu to run. However they can be set to any number
with sysctl knobs so that it can crash the kernel if something bad is
given. Fix it by limiting them into the actual range.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1345104204-8317-1-git-send-email-namhyung@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Commit beac4c7e4a ("sched: Remove AFFINE_WAKEUPS feature") removed
use of the flag but left the definition. Get rid of it.
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Link: http://lkml.kernel.org/r/1345090865-20851-1-git-send-email-namhyung@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Fix two kernel-doc warnings in kernel/sched/fair.c:
Warning(kernel/sched/fair.c:3660): Excess function parameter 'cpus' description in 'update_sg_lb_stats'
Warning(kernel/sched/fair.c:3806): Excess function parameter 'cpus' description in 'update_sd_lb_stats'
Signed-off-by: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/50303714.3090204@xenotime.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
migrate_tasks() uses _pick_next_task_rt() to get tasks from the
real-time runqueues to be migrated. When rt_rq is throttled
_pick_next_task_rt() won't return anything, in which case
migrate_tasks() can't move all threads over and gets stuck in an
infinite loop.
Instead unthrottle rt runqueues before migrating tasks.
Additionally: move unthrottle_offline_cfs_rqs() to rq_offline_fair()
Signed-off-by: Peter Boonstoppel <pboonstoppel@nvidia.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Turner <pjt@google.com>
Link: http://lkml.kernel.org/r/5FBF8E85CA34454794F0F7ECBA79798F379D3648B7@HQMAIL04.nvidia.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Azat Khuzhin reported high loadavg in Linux v3.6
After checking the upstream scheduler code, I found Peter's commit:
5167e8d541 sched/nohz: Rewrite and fix load-avg computation -- again
not fully applied, missing the call to calc_load_exit_idle().
After that idle exit in sampling window will always be calculated
to non-idle, and the load will be higher than normal.
This patch adds the missing call to calc_load_exit_idle().
Signed-off-by: Charles Wang <muming.wq@taobao.com>
Cc: stable@kernel.org
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1345449754-27130-1-git-send-email-muming.wq@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Rabik and Paul reported two different issues related to the same few
lines of code.
Rabik's issue is that the nr_uninterruptible migration code is wrong in
that he sees artifacts due to this (Rabik please do expand in more
detail).
Paul's issue is that this code as it stands relies on us using
stop_machine() for unplug, we all would like to remove this assumption
so that eventually we can remove this stop_machine() usage altogether.
The only reason we'd have to migrate nr_uninterruptible is so that we
could use for_each_online_cpu() loops in favour of
for_each_possible_cpu() loops, however since nr_uninterruptible() is the
only such loop and its using possible lets not bother at all.
The problem Rabik sees is (probably) caused by the fact that by
migrating nr_uninterruptible we screw rq->calc_load_active for both rqs
involved.
So don't bother with fancy migration schemes (meaning we now have to
keep using for_each_possible_cpu()) and instead fold any nr_active delta
after we migrate all tasks away to make sure we don't have any skewed
nr_active accounting.
Reported-by: Rakib Mullick <rakib.mullick@gmail.com>
Reported-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1345454817.23018.27.camel@twins
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Andreas Bombe reported that the added ktime_t overflow checking added to
timespec_valid in commit 4e8b14526c ("time: Improve sanity checking of
timekeeping inputs") was causing problems with X.org because it caused
timeouts larger then KTIME_T to be invalid.
Previously, these large timeouts would be clamped to KTIME_MAX and would
never expire, which is valid.
This patch splits the ktime_t overflow checking into a new
timespec_valid_strict function, and converts the timekeeping codes
internal checking to use this more strict function.
Reported-and-tested-by: Andreas Bombe <aeb@debian.org>
Cc: Zhouping Liu <zliu@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Signed-off-by: John Stultz <john.stultz@linaro.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Nobody does set_orig_insn(verify => false), and I think nobody will.
Remove this argument. IIUC set_orig_insn(verify => false) was needed
to single-step without xol area.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Now that we have uprobe_dup_mmap() we can fold uprobe_reset_state()
into the new hook and remove it. mmput()->uprobe_clear_state() can't
be called before dup_mmap().
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Add the new MMF_HAS_UPROBES flag. It is set by install_breakpoint()
and it is copied by dup_mmap(), uprobe_pre_sstep_notifier() checks
it to avoid the slow path if the task was never probed. Perhaps it
makes sense to check it in valid_vma(is_register => false) as well.
This needs the new dup_mmap()->uprobe_dup_mmap() hook. We can't use
uprobe_reset_state() or put MMF_HAS_UPROBES into MMF_INIT_MASK, we
need oldmm->mmap_sem to avoid the race with uprobe_register() or
mmap() from another thread.
Currently we never clear this bit, it can be false-positive after
uprobe_unregister() or uprobe_munmap() or if dup_mmap() hits the
probed VM_DONTCOPY vma. But this is fine correctness-wise and has
no effect unless the task hits the non-uprobe breakpoint.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
-EEXIST from install_breakpoint() no longer makes sense, all
callers should simply treat it as "success". Change the code
to return zero and simplify register_for_each_vma().
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Once install_breakpoint() fails uprobe_mmap() "ignores" all other
uprobes and returns the error.
It was never really needed to to stop after the first error, and
in fact it was always wrong at least in -ENOTSUPP case.
Change uprobe_mmap() to ignore the errors and always return 0.
This is not what we want in the long term, but until we teach
the callers to handle the failure it would be better to remove
the pointless complications. And this doesn't look too bad, the
only "reasonable" error is ENOMEM but in this case the caller
should be oom-killed in the likely case or the system has more
serious problems.
However it makes sense to stop if fatal_signal_pending() == T.
In particular this helps if the task was oom-killed.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
1. Kill dup_mmap()->uprobe_mmap(), it was only needed to calculate
new_mm->uprobes_state.count removed by the previous patch.
If the forking process has a pending uprobe (int3) in vma, it will
be copied by copy_page_range(), note that it checks vma->anon_vma
so "Don't copy ptes" is not possible after install_breakpoint()
which does anon_vma_prepare().
2. Remove is_swbp_at_addr() and "int count" in uprobe_mmap(). Again,
this was needed for uprobes_state.count.
As a side effect this fixes the bug pointed out by Srikar,
this code lacked the necessary put_uprobe().
3. uprobe_munmap() becomes a nop after the previous patch. Remove the
meaningless code but do not remove the helper, we will need it.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
uprobes_state->count is only needed to avoid the slow path in
uprobe_pre_sstep_notifier(). It is also checked in uprobe_munmap()
but ironically its only goal to decrement this counter. However,
it is very broken. Just some examples:
- uprobe_mmap() can race with uprobe_unregister() and wrongly
increment the counter if it hits the non-uprobe "int3". Note
that install_breakpoint() checks ->consumers first and returns
-EEXIST if it is NULL.
"atomic_sub() if error" in uprobe_mmap() looks obviously wrong
too.
- uprobe_munmap() can race with uprobe_register() and wrongly
decrement the counter by the same reason.
- Suppose an appication tries to increase the mmapped area via
sys_mremap(). vma_adjust() does uprobe_munmap(whole_vma) first,
this can nullify the counter temporarily and race with another
thread which can hit the bp, the application will be killed by
SIGTRAP.
- Suppose an application mmaps 2 consecutive areas in the same file
and one (or both) of these areas has uprobes. In the likely case
mmap_region()->vma_merge() suceeds. Like above, this leads to
uprobe_munmap/uprobe_mmap from vma_merge()->vma_adjust() but then
mmap_region() does another uprobe_mmap(resulting_vma) and doubles
the counter.
This patch only removes this counter and fixes the compile errors,
then we will try to cleanup the changed code and add something else
instead.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
by the time we get here (after we pass cleanup_ret) uprobe is always is
set. If it is NULL we leave very early in the code.
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Since read_opcode() reads from the referenced page and doesnt modify
the page contents nor the page attributes, there is no need to lock
the page.
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Pull timer fixes from Thomas Gleixner:
"Mostly small fixes for the fallout of the timekeeping overhaul in 3.6
along with stable fixes to address an accumulation problem and missing
sanity checks for RTC readouts and user space provided values."
* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
time: Avoid making adjustments if we haven't accumulated anything
time: Avoid potential shift overflow with large shift values
time: Fix casting issue in timekeeping_forward_now
time: Ensure we normalize the timekeeper in tk_xtime_add
time: Improve sanity checking of timekeeping inputs
The function graph has a test to check if the frame pointer is
corrupted, which can happen with various options of gcc with mcount.
But this is not an issue with -mfentry as -mfentry does not need nor use
frame pointers for function graph tracing.
Link: http://lkml.kernel.org/r/20120807194059.773895870@goodmis.org
Acked-by: H. Peter Anvin <hpa@linux.intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Thanks to Andi Kleen, gcc 4.6.0 now supports -mfentry for x86
(and hopefully soon for other archs). What this does is to have
the function profiler start at the beginning of the function
instead of after the stack is set up. As plain -pg (mcount) is
called after the stack is set up, and in some cases can have issues
with the function graph tracer. It also requires frame pointers to
be enabled.
The -mfentry now calls __fentry__ at the beginning of the function.
This allows for compiling without frame pointers and even has the
ability to access parameters if needed.
If the architecture and the compiler both support -mfentry then
use that instead.
Link: http://lkml.kernel.org/r/20120807194059.392617243@goodmis.org
Acked-by: H. Peter Anvin <hpa@linux.intel.com>
Acked-by: Ingo Molnar <mingo@kernel.org>
Cc: Michal Marek <mmarek@suse.cz>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
We still patch SMP instructions to UP variants if we boot with a
single CPU, but not at any other time. In particular, not if we
unplug CPUs to return to a single cpu.
Paul McKenney points out:
mean offline overhead is 6251/48=130.2 milliseconds.
If I remove the alternatives_smp_switch() from the offline
path [...] the mean offline overhead is 550/42=13.1 milliseconds
Basically, we're never going to get those 120ms back, and the
code is pretty messy.
We get rid of:
1) The "smp-alt-once" boot option. It's actually "smp-alt-boot", the
documentation is wrong. It's now the default.
2) The skip_smp_alternatives flag used by suspend.
3) arch_disable_nonboot_cpus_begin() and arch_disable_nonboot_cpus_end()
which were only used to set this one flag.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Cc: Paul McKenney <paul.mckenney@us.ibm.com>
Cc: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/87vcgwwive.fsf@rustcorp.com.au
Signed-off-by: Ingo Molnar <mingo@kernel.org>
If update_wall_time() is called and the current offset isn't large
enough to accumulate, avoid re-calling timekeeping_adjust which may
change the clock freq and can cause 1ns inconsistencies with
CLOCK_REALTIME_COARSE/CLOCK_MONOTONIC_COARSE.
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: stable@vger.kernel.org
Link: http://lkml.kernel.org/r/1345595449-34965-5-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Andreas Schwab noticed that the 1 << tk->shift could overflow if the
shift value was greater than 30, since 1 would be a 32bit long on
32bit architectures. This issue was introduced by 1e75fa8be (time:
Condense timekeeper.xtime into xtime_sec)
Use 1ULL instead to ensure we don't overflow on the shift.
Reported-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1345595449-34965-4-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
arch_gettimeoffset returns a u32 value which when shifted by tk->shift
can overflow. This issue was introduced with 1e75fa8be (time: Condense
timekeeper.xtime into xtime_sec)
Cast it to u64 first.
Signed-off-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1345595449-34965-3-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Andreas noticed problems with resume on specific hardware after commit
1e75fa8b (time: Condense timekeeper.xtime into xtime_sec) combined
with commit b44d50dca (time: Fix casting issue in tk_set_xtime and
tk_xtime_add)
After some digging I realized we aren't normalizing the timekeeper
after the add. Add the missing normalize call.
Reported-by: Andreas Schwab <schwab@linux-m68k.org>
Tested-by: Andreas Schwab <schwab@linux-m68k.org>
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Ingo Molnar <mingo@kernel.org>
Link: http://lkml.kernel.org/r/1345595449-34965-2-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Pull audit-tree fixes from Miklos Szeredi:
"The audit subsystem maintainers (Al and Eric) are not responding to
repeated resends. Eric did ack them a while ago, but no response
since then. So I'm sending these directly to you."
* 'audit-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/vfs:
audit: clean up refcounting in audit-tree
audit: fix refcounting in audit-tree
audit: don't free_chunk() after fsnotify_add_mark()
It seems commit 4a9d4b024a ("switch fput to task_work_add") re-
introduced the problem addressed in 944be0b224 ("close_files(): add
scheduling point")
If a server process with a lot of files (say 2 million tcp sockets) is
killed, we can spend a lot of time in task_work_run() and trigger a soft
lockup.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Timer internals are protected with irq-safe locks but timer execution
isn't, so a timer being dequeued for execution and its execution
aren't atomic against IRQs. This makes it impossible to wait for its
completion from IRQ handlers and difficult to shoot down a timer from
IRQ handlers.
This issue caused some issues for delayed_work interface. Because
there's no way to reliably shoot down delayed_work->timer from IRQ
handlers, __cancel_delayed_work() can't share the logic to steal the
target delayed_work with cancel_delayed_work_sync(), and can only
steal delayed_works which are on queued on timer. Similarly, the
pending mod_delayed_work() can't be used from IRQ handlers.
This patch adds a new timer flag TIMER_IRQSAFE, which makes the timer
to be executed without enabling IRQ after dequeueing such that its
dequeueing and execution are atomic against IRQ handlers.
This makes it safe to wait for the timer's completion from IRQ
handlers, for example, using del_timer_sync(). It can never be
executing on the local CPU and if executing on other CPUs it won't be
interrupted until done.
This will enable simplifying delayed_work cancel/mod interface.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: torvalds@linux-foundation.org
Cc: peterz@infradead.org
Link: http://lkml.kernel.org/r/1344449428-24962-5-git-send-email-tj@kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Over time, timer initializers became messy with unnecessarily
duplicated code which are inconsistently spread across timer.h and
timer.c.
This patch cleans up timer initializers.
* timer.c::__init_timer() is renamed to do_init_timer().
* __TIMER_INITIALIZER() added. It takes @flags and all initializers
are wrappers around it.
* init_timer[_on_stack]_key() now take @flags.
* __init_timer[_on_stack]() added. They take @flags and all init
macros are wrappers around them.
* __setup_timer[_on_stack]() added. It uses __init_timer() and takes
@flags. All setup macros are wrappers around the two.
Note that this patch doesn't add missing init/setup combinations -
e.g. init_timer_deferrable_on_stack(). Adding missing ones is
trivial.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: torvalds@linux-foundation.org
Cc: peterz@infradead.org
Link: http://lkml.kernel.org/r/1344449428-24962-4-git-send-email-tj@kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
To prepare for addition of another flag, generalize timer->base flags
handling.
* Rename from TBASE_*_FLAG to TIMER_* and make them LU constants.
* Define and use TIMER_FLAG_MASK for flags masking so that multiple
flags can be handled correctly.
* Don't dereference timer->base directly even if
!tbase_get_deferrable(). All two such places are already passed in
@base, so use it instead.
* Make sure tvec_base's alignment is large enough for timer->base
flags using BUILD_BUG_ON().
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: torvalds@linux-foundation.org
Cc: peterz@infradead.org
Link: http://lkml.kernel.org/r/1344449428-24962-2-git-send-email-tj@kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Export dummy_irq_chip to modules to allow them to do things such as
irq_set_chip_and_handler(virq,
&dummy_irq_chip,
handle_level_irq);
This fixes
ERROR: "dummy_irq_chip" [drivers/gpio/gpio-pcf857x.ko] undefined!
when gpio-pcf857x.c is being built as a module.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Greg KH <gregkh@linuxfoundation.org>
Link: http://lkml.kernel.org/r/871ujstrp6.wl%25kuninori.morimoto.gx@renesas.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Export irq_set_chip_and_handler_name() to modules to allow them to
do things such as
irq_set_chip_and_handler(....);
This fixes
ERROR: "irq_set_chip_and_handler_name" \
[drivers/gpio/gpio-pcf857x.ko] undefined!
when gpio-pcf857x.c is being built as a module.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: Linus Walleij <linus.walleij@linaro.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Cc: Greg KH <gregkh@linuxfoundation.org>
Link: http://lkml.kernel.org/r/873948trpk.wl%25kuninori.morimoto.gx@renesas.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
This patch fixes:
https://bugzilla.redhat.com/show_bug.cgi?id=843640
If mmap_region()->uprobe_mmap() fails, unmap_and_free_vma path
does unmap_region() but does not remove the soon-to-be-freed vma
from rb tree. Actually there are more problems but this is how
William noticed this bug.
Perhaps we could do do_munmap() + return in this case, but in
fact it is simply wrong to abort if uprobe_mmap() fails. Until
at least we move the !UPROBE_COPY_INSN code from
install_breakpoint() to uprobe_register().
For example, uprobe_mmap()->install_breakpoint() can fail if the
probed insn is not supported (remember, uprobe_register()
succeeds if nobody mmaps inode/offset), mmap() should not fail
in this case.
dup_mmap()->uprobe_mmap() is wrong too by the same reason,
fork() can race with uprobe_register() and fail for no reason if
it wins the race and does install_breakpoint() first.
And, if nothing else, both mmap_region() and dup_mmap() return
success if uprobe_mmap() fails. Change them to ignore the error
code from uprobe_mmap().
Reported-and-tested-by: William Cohen <wcohen@redhat.com>
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: <stable@vger.kernel.org> # v3.5
Cc: Anton Arapov <anton@redhat.com>
Cc: William Cohen <wcohen@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20120819171042.GB26957@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Paul E. McKenney