- 11 Feb, 2015 1 commit
-
-
Paul E. McKenney authored
If the scheduling-clock interrupt sets the current tasks need_qs flag, but if the current CPU passes through a quiescent state in the meantime, then rcu_preempt_qs() will fail to clear the need_qs flag, which can fool RCU into thinking that additional rcu_read_unlock_special() processing is needed. This commit therefore clears the need_qs flag before checking for additional processing. For this problem to occur, we need rcu_preempt_data.passed_quiesce equal to true and current->rcu_read_unlock_special.b.need_qs also equal to true. This condition can occur as follows: 1. CPU 0 is aware of the current preemptible RCU grace period, but has not yet passed through a quiescent state. Among other things, this means that rcu_preempt_data.passed_quiesce is false. 2. Task A running on CPU 0 enters a preemptible RCU read-side critical section. 3. CPU 0 takes a scheduling-clock interrupt, which notices the RCU read-side critical section and the need for a quiescent state, and thus sets current->rcu_read_unlock_special.b.need_qs to true. 4. Task A is preempted, enters the scheduler, eventually invoking rcu_preempt_note_context_switch() which in turn invokes rcu_preempt_qs(). Because rcu_preempt_data.passed_quiesce is false, control enters the body of the "if" statement, which sets rcu_preempt_data.passed_quiesce to true. 5. At this point, CPU 0 takes an interrupt. The interrupt handler contains an RCU read-side critical section, and the rcu_read_unlock() notes that current->rcu_read_unlock_special is nonzero, and thus invokes rcu_read_unlock_special(). 6. Once in rcu_read_unlock_special(), the fact that current->rcu_read_unlock_special.b.need_qs is true becomes apparent, so rcu_read_unlock_special() invokes rcu_preempt_qs(). Recursively, given that we interrupted out of that same function in the preceding step. 7. Because rcu_preempt_data.passed_quiesce is now true, rcu_preempt_qs() does nothing, and simply returns. 8. Upon return to rcu_read_unlock_special(), it is noted that current->rcu_read_unlock_special is still nonzero (because the interrupted rcu_preempt_qs() had not yet gotten around to clearing current->rcu_read_unlock_special.b.need_qs). 9. Execution proceeds to the WARN_ON_ONCE(), which notes that we are in an interrupt handler and thus duly splats. The solution, as noted above, is to make rcu_read_unlock_special() clear out current->rcu_read_unlock_special.b.need_qs after calling rcu_preempt_qs(). The interrupted rcu_preempt_qs() will clear it again, but this is harmless. The worst that happens is that we clobber another attempt to set this field, but this is not a problem because we just got done reporting a quiescent state. Reported-by:
Sasha Levin <sasha.levin@oracle.com> Signed-off-by:
Paul E. McKenney <paulmck@linux.vnet.ibm.com> [ paulmck: Fix embarrassing build bug noted by Sasha Levin. ] Tested-by:
-
- 16 Jan, 2015 6 commits
-
-
Paul E. McKenney authored
Merge branches 'doc.2015.01.07a', 'fixes.2015.01.15a', 'preempt.2015.01.06a', 'srcu.2015.01.06a', 'stall.2015.01.16a' and 'torture.2015.01.11a' into HEAD doc.2015.01.07a: Documentation updates. fixes.2015.01.15a: Miscellaneous fixes. preempt.2015.01.06a: Changes to handling of lists of preempted tasks. srcu.2015.01.06a: SRCU updates. stall.2015.01.16a: RCU CPU stall-warning updates and fixes. torture.2015.01.11a: RCU torture-test updates and fixes.
-
Paul E. McKenney authored
The current tiny RCU stall-warning code assumes that the jiffies counter starts at zero, however, it is sometimes initialized to other values, for example, -30,000. This commit therefore changes rcu_init() to invoke reset_cpu_stall_ticks() for both flavors of RCU to initialize the stall-warning times properly at boot. Signed-off-by: -
Miroslav Benes authored
The tiny RCU CPU stall detection depends on *rcp->curtail not being NULL. It is however a tail pointer and thus NULL by definition. Instead we should check rcp->rcucblist for the presence of pending callbacks which need to be processed. With this fix INFO about the stall is printed and jiffies_stall (jiffies at next stall) correctly updated. Note that the check for pending callback is necessary to avoid spurious warnings if there are no pendings callbacks. Signed-off-by:
Miroslav Benes <mbenes@suse.cz> [ paulmck: Fused identical "if" statements, ported to -rcu. ] Signed-off-by:
-
Paul E. McKenney authored
This commit adds a message that is printed if the relevant grace-period kthread has not been able to run for the two seconds preceding the stall warning. (The two seconds is double the maximum interval between successive bouts of quiescent-state forcing.) Signed-off-by: -
Paul E. McKenney authored
Although cond_resched_rcu_qs() only applies to TASKS_RCU, it is used in places where it would be useful for it to apply to the normal RCU flavors, rcu_preempt, rcu_sched, and rcu_bh. This is especially the case for workloads that aggressively overload the system, particularly those that generate large numbers of RCU updates on systems running NO_HZ_FULL CPUs. This commit therefore communicates quiescent states from cond_resched_rcu_qs() to the normal RCU flavors. Note that it is unfortunately necessary to leave the old ->passed_quiesce mechanism in place to allow quiescent states that apply to only one flavor to be recorded. (Yes, we could decrement ->rcu_qs_ctr_snap in that case, but that is not so good for debugging of RCU internals.) In addition, if one of the RCU flavor's grace period has stalled, this will invoke rcu_momentary_dyntick_idle(), resulting in a heavy-weight quiescent state visible from other CPUs. Reported-by:
Dave Jones <davej@redhat.com> Signed-off-by:
-
Paul E. McKenney authored
Recent testing has shown that under heavy load, running RCU's grace-period kthreads at real-time priority can improve performance (according to 0day test robot) and reduce the incidence of RCU CPU stall warnings. However, most systems do just fine with the default non-realtime priorities for these kthreads, and it does not make sense to expose the entire user base to any risk stemming from this change, given that this change is of use only to a few users running extremely heavy workloads. Therefore, this commit allows users to specify realtime priorities for the grace-period kthreads, but leaves them running SCHED_OTHER by default. The realtime priority may be specified at build time via the RCU_KTHREAD_PRIO Kconfig parameter, or at boot time via the rcutree.kthread_prio parameter. Either way, 0 says to continue the default SCHED_OTHER behavior and values from 1-99 specify that priority of SCHED_FIFO behavior. Note that a value of 0 is not permitted when the RCU_BOOST Kconfig parameter is specified. Signed-off-by:
-
- 14 Jan, 2015 2 commits
-
-
Paul E. McKenney authored
Simplify run_ksoftirqd() by using the new cond_resched_rcu_qs() function that conditionally reschedules, but unconditionally supplies an RCU quiescent state. This commit is separate from the previous commit by Calvin Owens because Calvin's approach can be backported, while this commit cannot be. The reason that this commit cannot be backported is that cond_resched_rcu_qs() does not always provide the needed quiescent state in earlier kernels. Signed-off-by: -
Calvin Owens authored
While debugging an issue with excessive softirq usage, I encountered the following note in commit 3e339b5d ("softirq: Use hotplug thread infrastructure"): [ paulmck: Call rcu_note_context_switch() with interrupts enabled. ] ...but despite this note, the patch still calls RCU with IRQs disabled. This seemingly innocuous change caused a significant regression in softirq CPU usage on the sending side of a large TCP transfer (~1 GB/s): when introducing 0.01% packet loss, the softirq usage would jump to around 25%, spiking as high as 50%. Before the change, the usage would never exceed 5%. Moving the call to rcu_note_context_switch() after the cond_sched() call, as it was originally before the hotplug patch, completely eliminated this problem. Signed-off-by:
Calvin Owens <calvinowens@fb.com> Cc: stable@vger.kernel.org Signed-off-by:
-
- 11 Jan, 2015 12 commits
-
-
Paul E. McKenney authored
Signed-off-by: -
Paul E. McKenney authored
A system misconfiguration that prevents qemu from running at all (for example, a missing dynamically linked library) will keep the console.log file from the previous run. This can fool the developer into thinking that this failed run actually completed correctly. This commit therefore overwrites the console.log file just before launching qemu. Signed-off-by: -
Paul E. McKenney authored
More recent qemu implementations really want "-enable-kvm", and the "-soundhw pcspk" makes the script a bit less dependent on odd audio libraries being installed. This commit therefore adds both to the default qemu command line. Signed-off-by: -
Paul E. McKenney authored
The mpstat command recently added the %gnice column, which messes up the cpu2use.sh script's idle-CPU calculations. This commit therefore uses $NF instead of $12 to select the last (%idle) column. Signed-off-by: -
Paul E. McKenney authored
Currently, rcutorture's Reader Batch checks measure from the end of the previous grace period to the end of the current one. This commit tightens up these checks by measuring from the start and end of the same grace period. This involves adding rcu_batches_started() and friends corresponding to the existing rcu_batches_completed() and friends. We leave SRCU alone for the moment, as it does not yet have a way of tracking both ends of its grace periods. Signed-off-by: -
Paul E. McKenney authored
Signed-off-by: -
Paul E. McKenney authored
Now that the return type of rcu_batches_completed() and friends matches that of the rcu_torture_ops structure's ->completed field, the wrapper functions can be deleted. This commit carries out that deletion, while also wiring "sched"'s ->completed field to rcu_batches_completed_sched(). Signed-off-by: -
Paul E. McKenney authored
A bug in rcutorture has caused it to ignore completed batches. In preparation for fixing that bug, this commit provides TINY_RCU with the required rcu_batches_completed_sched(). Signed-off-by: -
Paul E. McKenney authored
The counter returned by the various ->completed functions is subject to overflow, which means that subtracting two such counters might result in overflow, which invokes undefined behavior in the C standard. This commit therefore changes these functions and variables to unsigned to avoid this undefined behavior. Signed-off-by: -
Paul E. McKenney authored
Signed-off-by: -
Paul E. McKenney authored
Long ago, the various ->completed fields were of type long, but now are unsigned long due to signed-integer-overflow concerns. However, the various _batches_completed() functions remained of type long, even though their only purpose in life is to return the corresponding ->completed field. This patch cleans this up by changing these functions' return types to unsigned long. Signed-off-by: -
Paul E. McKenney authored
Normal rcutorture checking overestimates grace periods somewhat due to the fact that there is a delay from a grace-period request until the start of the corresponding grace period and another delay from the end of that grace period to notification of the requestor. This means that rcutorture's detection of RCU bugs is less sensitive than it might be. It turns out that rcutorture also checks the underlying grace-period "completed" counter (displayed in Reader Batch output), which in theory allows rcutorture to do exact checks. In practice, memory misordering (by both compiler and CPU) can result in false positives. However, experience on x86 shows that these false positives are quite rare, occurring less than one time per 1,000 hours of testing. This commit therefore does the exact checking, giving a warning if any Reader Batch blows happen, and flagging an error if they happen more often than once every three hours in long tests. Signed-off-by:
-
- 07 Jan, 2015 2 commits
-
-
Davidlohr Bueso authored
Signed-off-by:
Davidlohr Bueso <dbueso@suse.de> Signed-off-by:
-
Paul E. McKenney authored
This commit documents the fact that it is not safe to use bitfields as shared variables in synchronization algorithms. It also documents that CPUs must be able to concurrently load from and store to adjacent one-byte and two-byte variables, which is in fact required by the C11 standard (Section 3.14). Signed-off-by:
-
- 06 Jan, 2015 17 commits
-
-
Paul E. McKenney authored
Subtle race conditions can result if a CPU stays in dyntick-idle mode long enough for the ->gpnum and ->completed fields to wrap. For example, consider the following sequence of events: o CPU 1 encounters a quiescent state while waiting for grace period 5 to complete, but then enters dyntick-idle mode. o While CPU 1 is in dyntick-idle mode, the grace-period counters wrap around so that the grace period number is now 4. o Just as CPU 1 exits dyntick-idle mode, grace period 4 completes and grace period 5 begins. o The quiescent state that CPU 1 passed through during the old grace period 5 looks like it applies to the new grace period 5. Therefore, the new grace period 5 completes without CPU 1 having passed through a quiescent state. This could clearly be a fatal surprise to any long-running RCU read-side critical section that happened to be running on CPU 1 at the time. At one time, this was not a problem, given that it takes significant time for the grace-period counters to overflow even on 32-bit systems. However, with the advent of NO_HZ_FULL and SMP embedded systems, arbitrarily long idle periods are now becoming quite feasible. It is therefore time to close this race. This commit therefore avoids this race condition by having the quiescent-state forcing code detect when a CPU is falling too far behind, and setting a new rcu_data field ->gpwrap when this happens. Whenever this new ->gpwrap field is set, the CPU's ->gpnum and ->completed fields are known to be untrustworthy, and can be ignored, along with any associated quiescent states. Signed-off-by: -
Paul E. McKenney authored
The current RCU CPU stall warning code will print "Stall ended before state dump start" any time that the stall-warning code is triggered on a CPU that has already reported a quiescent state for the current grace period and if all quiescent states have been reported for the current grace period. However, a true stall can result in these symptoms, for example, by preventing RCU's grace-period kthreads from ever running This commit therefore checks for this condition, reporting the end of the stall only if one of the grace-period counters has actually advanced. Otherwise, it reports the last time that the grace-period kthread made meaningful progress. (In normal situations, the grace-period kthread should make meaningful progress at least every jiffies_till_next_fqs jiffies.) Reported-by:
-
Paul E. McKenney authored
One way that an RCU CPU stall warning can happen is if the grace-period kthread is not allowed to execute. One proxy for this kthread's forward progress is the number of force-quiescent-state (fqs) scans. This commit therefore adds the number of fqs scans to the RCU CPU stall warning printouts when CONFIG_RCU_CPU_STALL_INFO=y. Signed-off-by: -
Paul E. McKenney authored
The current rcutorture scripting checks for actual stalls (via the "Call Trace:" check), but fails to spot the case where a stall ends just as it is being detected. This commit therefore adds a check for this case. Signed-off-by: -
Paul E. McKenney authored
The RCU_CPU_STALL_INFO code has been in for quite some time, and has proven reliable. This commit therefore enables it by default. Signed-off-by: -
Pranith Kumar authored
SRCU is not necessary to be compiled by default in all cases. For tinification efforts not compiling SRCU unless necessary is desirable. The current patch tries to make compiling SRCU optional by introducing a new Kconfig option CONFIG_SRCU which is selected when any of the components making use of SRCU are selected. If we do not select CONFIG_SRCU, srcu.o will not be compiled at all. text data bss dec hex filename 2007 0 0 2007 7d7 kernel/rcu/srcu.o Size of arch/powerpc/boot/zImage changes from text data bss dec hex filename 831552 64180 23944 919676 e087c arch/powerpc/boot/zImage : before 829504 64180 23952 917636 e0084 arch/powerpc/boot/zImage : after so the savings are about ~2000 bytes. Signed-off-by:
Pranith Kumar <bobby.prani@gmail.com> CC: Paul E. McKenney <paulmck@linux.vnet.ibm.com> CC: Josh Triplett <josh@joshtriplett.org> CC: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by:
-
Paul E. McKenney authored
The DEFINE_SRCU() and DEFINE_STATIC_SRCU() definitions are quite similar, so this commit combines them, saving a bit of code and removing redundancy. Signed-off-by: -
Paul E. McKenney authored
When rcutorture used only the low-order 32 bits of the grace-period number, it was not a problem for SRCU to use a 32-bit completed field. However, rcutorture now uses the full 64 bits on 64-bit systems, so this commit converts SRCU's ->completed field to unsigned long so as to provide 64 bits on 64-bit systems. Signed-off-by: -
Paul E. McKenney authored
The RCU callback lists are initialized in both rcu_boot_init_percpu_data() and rcu_init_percpu_data(). The former is intended for initializing immutable data, so this commit removes the initialization from rcu_boot_init_percpu_data() and leaves it in rcu_init_percpu_data(). This change prepares for permitting callbacks to be queued very early in boot. Signed-off-by: -
Paul E. McKenney authored
Now that blocked tasks are no longer migrated to the root rcu_node structure, there is no need to scan the root rcu_node structure for blocked tasks stalling the current grace period. This commit therefore removes this scan. Signed-off-by: -
Lai Jiangshan authored
The patch dfeb9765 ("Allow post-unlock reference for rt_mutex") ensured rcu-boost safe even the rt_mutex has post-unlock reference. But rt_mutex allowing post-unlock reference is definitely a bug and it was fixed by the commit 27e35715 ("rtmutex: Plug slow unlock race"). This fix made the previous patch (dfeb9765) useless. And even worse, the priority-inversion introduced by the the previous patch still exists. rcu_read_unlock_special() { rt_mutex_unlock(&rnp->boost_mtx); /* Priority-Inversion: * the current task had been deboosted and preempted as a low * priority task immediately, it could wait long before reschedule in, * and the rcu-booster also waits on this low priority task and sleeps. * This priority-inversion makes rcu-booster can't work * as expected. */ complete(&rnp->boost_completion); } Just revert the patch to avoid it. Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by:
Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by:
-
Paul E. McKenney authored
The rcu_cleanup_dead_cpu() function (called after a CPU has gone completely offline) has not reported a quiescent state because there was probably at least one synchronize_rcu() between the time the CPU went offline and the CPU_DEAD notifier, and this would have detected the CPU's offline state via quiescent-state forcing. However, the plan is for CPUs to take themselves offline, at which point it makes sense for them to report their own quiescent state. This commit makes this change in preparation for the new CPU-hotplug setup. Signed-off-by: -
Paul E. McKenney authored
When rcu_boost_kthread_setaffinity() sees that all CPUs for a given rcu_node structure are now offline, it affinities the corresponding RCU-boost ("rcub") kthread away from those CPUs. This is pointless because the kthread cannot run on those offline CPUs in any case. This commit therefore removes this unneeded code. Signed-off-by: -
Paul E. McKenney authored
Because there is no longer any preempted tasks on the root rcu_node, and because there is no longer ever an rcub kthread for the root rcu_node, this commit drops the code in force_qs_rnp() that attempts to awaken the non-existent root rcub kthread. This is strictly a performance enhancement, removing a root rcu_node ->lock acquisition and release along with some tests in rcu_initiate_boost(), ending with the test that notes that there is no rcub kthread. Signed-off-by: -
Paul E. McKenney authored
Now that offlining CPUs no longer moves leaf rcu_node structures' ->blkd_tasks lists to the root, there is no way for the root rcu_node structure's ->blkd_task list to be nonempty, unless the root node is also the sole leaf node. This commit therefore refrains from creating an rcub kthread for the root rcu_node structure unless it is also the sole leaf. Signed-off-by: -
Paul E. McKenney authored
Given that there is now arcu_preempt_has_tasks() function that checks to see if the ->blkd_tasks list is non-empty, this commit makes use of it. Signed-off-by: -
Paul E. McKenney authored
Now that we are not migrating callbacks, there is no need to hold the ->orphan_lock across the the ->qsmaskinit bit-clearing process. This commit therefore releases ->orphan_lock immediately after adopting the orphaned RCU callbacks. Signed-off-by:
-
