- 03 Dec, 2009 3 commits
-
-
Paul E. McKenney authored
Implement an synchronize_rcu_expedited() for preemptible RCU that actually is expedited. This uses synchronize_sched_expedited() to force all threads currently running in a preemptible-RCU read-side critical section onto the appropriate ->blocked_tasks[] list, then takes a snapshot of all of these lists and waits for them to drain. Signed-off-by:
Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: josh@joshtriplett.org Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com LKML-Reference: <1259784616158-git-send-email-> Signed-off-by:
Ingo Molnar <mingo@elte.hu>
-
Paul E. McKenney authored
Enable a fourth level of rcu_node hierarchy for TREE_RCU and TREE_PREEMPT_RCU. This is for stress-testing and experiemental purposes only, although in theory this would enable 16,777,216 CPUs on 64-bit systems, though only 1,048,576 CPUs on 32-bit systems. Normal experimental use of this fourth level will normally set CONFIG_RCU_FANOUT=2, requiring a 16-CPU system, though the more adventurous (and more fortunate) experimenters may wish to chose CONFIG_RCU_FANOUT=3 for 81-CPU systems or even CONFIG_RCU_FANOUT=4 for 256-CPU systems. Signed-off-by:
Josh Triplett <josh@joshtriplett.org> Acked-by:
Lai Jiangshan <laijs@cn.fujitsu.com> Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org Cc: Valdis.Kletnieks@vt.edu Cc: dhowells@redhat.com LKML-Reference: <12597846161257-git-send-email-> Signed-off-by:
-
Paul E. McKenney authored
The number of "quiet" functions has grown recently, and the names are no longer very descriptive. The point of all of these functions is to do some portion of the task of reporting a quiescent state, so rename them accordingly: o cpu_quiet() becomes rcu_report_qs_rdp(), which reports a quiescent state to the per-CPU rcu_data structure. If this turns out to be a new quiescent state for this grace period, then rcu_report_qs_rnp() will be invoked to propagate the quiescent state up the rcu_node hierarchy. o cpu_quiet_msk() becomes rcu_report_qs_rnp(), which reports a quiescent state for a given CPU (or possibly a set of CPUs) up the rcu_node hierarchy. o cpu_quiet_msk_finish() becomes rcu_report_qs_rsp(), which reports a full set of quiescent states to the global rcu_state structure. o task_quiet() becomes rcu_report_unblock_qs_rnp(), which reports a quiescent state due to a task exiting an RCU read-side critical section that had previously blocked in that same critical section. As indicated by the new name, this type of quiescent state is reported up the rcu_node hierarchy (using rcu_report_qs_rnp() to do so). Signed-off-by:
-
- 22 Nov, 2009 3 commits
-
-
Paul E. McKenney authored
Remove #ifdefs from kernel/rcupdate.c and include/linux/rcupdate.h by moving code to include/linux/rcutiny.h, include/linux/rcutree.h, and kernel/rcutree.c. Also remove some definitions that are no longer used. Signed-off-by:
-
Paul E. McKenney authored
The functions rcu_init() is a wrapper for __rcu_init(), and also sets up the CPU-hotplug notifier for rcu_barrier_cpu_hotplug(). But TINY_RCU doesn't need CPU-hotplug notification, and the rcu_barrier_cpu_hotplug() is a simple wrapper for rcu_cpu_notify(). So push rcu_init() out to kernel/rcutree.c and kernel/rcutiny.c and get rid of the wrapper function rcu_barrier_cpu_hotplug(). Signed-off-by:
-
Paul E. McKenney authored
When the last CPU of a given leaf rcu_node structure goes offline, all of the tasks queued on that leaf rcu_node structure (due to having blocked in their current RCU read-side critical sections) are requeued onto the root rcu_node structure. This requeuing is carried out by rcu_preempt_offline_tasks(). However, it is possible that these queued tasks are the only thing preventing the leaf rcu_node structure from reporting a quiescent state up the rcu_node hierarchy. Unfortunately, the old code would fail to do this reporting, resulting in a grace-period stall given the following sequence of events: 1. Kernel built for more than 32 CPUs on 32-bit systems or for more than 64 CPUs on 64-bit systems, so that there is more than one rcu_node structure. (Or CONFIG_RCU_FANOUT is artificially set to a number smaller than CONFIG_NR_CPUS.) 2. The kernel is built with CONFIG_TREE_PREEMPT_RCU. 3. A task running on a CPU associated with a given leaf rcu_node structure blocks while in an RCU read-side critical section -and- that CPU has not yet passed through a quiescent state for the current RCU grace period. This will cause the task to be queued on the leaf rcu_node's blocked_tasks[] array, in particular, on the element of this array corresponding to the current grace period. 4. Each of the remaining CPUs corresponding to this same leaf rcu_node structure pass through a quiescent state. However, the task is still in its RCU read-side critical section, so these quiescent states cannot be reported further up the rcu_node hierarchy. Nevertheless, all bits in the leaf rcu_node structure's ->qsmask field are now zero. 5. Each of the remaining CPUs go offline. (The events in step #4 and #5 can happen in any order as long as each CPU passes through a quiescent state before going offline.) 6. When the last CPU goes offline, __rcu_offline_cpu() will invoke rcu_preempt_offline_tasks(), which will move the task to the root rcu_node structure, but without reporting a quiescent state up the rcu_node hierarchy (and this failure to report a quiescent state is the bug). But because this leaf rcu_node structure's ->qsmask field is already zero and its ->block_tasks[] entries are all empty, force_quiescent_state() will skip this rcu_node structure. Therefore, grace periods are now hung. This patch abstracts some code out of rcu_read_unlock_special(), calling the result task_quiet() by analogy with cpu_quiet(), and invokes task_quiet() from both rcu_read_lock_special() and __rcu_offline_cpu(). Invoking task_quiet() from __rcu_offline_cpu() reports the quiescent state up the rcu_node hierarchy, fixing the bug. This ends up requiring a separate lock_class_key per level of the rcu_node hierarchy, which this patch also provides. Signed-off-by:
-
- 14 Nov, 2009 2 commits
-
-
Paul E. McKenney authored
Now that there are both ->gpnum and ->completed fields in the rcu_node structure, __rcu_pending() should check rdp->gpnum and rdp->completed against rnp->gpnum and rdp->completed, respectively, instead of the prior comparison against the rcu_state fields rsp->gpnum and rsp->completed. Given the old comparison, __rcu_pending() could return 1, resulting in a needless raise_softirq(RCU_SOFTIRQ). This useless work would happen if RCU responded to a scheduling-clock interrupt after the rcu_state fields had been updated, but before the rcu_node fields had been updated. Changing the comparison from the rcu_state fields to the rcu_node fields prevents this useless work from happening. Signed-off-by:
-
Paul E. McKenney authored
Now that a copy of the rsp->completed flag is available in all rcu_node structures, make full use of it. It is still legitimate to access rsp->completed while holding the root rcu_node structure's lock, however. Also, tighten up force_quiescent_state()'s checks for end of current grace period. Signed-off-by:
-
- 13 Nov, 2009 2 commits
-
-
Paul E. McKenney authored
The force_quiescent_state() function also took a snapshot of the ->completed field, which was as obnoxious as it was in rcu_sched_qs() and friends. So snapshot ->gpnum-1. Also, since the dyntick_record_completed() and dyntick_recall_completed() functions are now simple assignments that are independent of CONFIG_NO_HZ, and since their names are now misleading, get rid of them. Signed-off-by:
-
Paul E. McKenney authored
An earlier fix for a race resulted in a situation where the CPUs other than the CPU that detected the end of the grace period would not process their callbacks until the next grace period started. This means that these other CPUs would unnecessarily demand that an extra grace period be started. This patch eliminates this extra grace period and speeds callback processing by propagating rsp->completed to the rcu_node structures in the case where the CPU detecting the end of the grace period sees no reason to start a new grace period. Signed-off-by:
-
- 11 Nov, 2009 1 commit
-
-
Paul E. McKenney authored
Because rcu_bootup_announce() is used only at boot time, mark it as __init, presumably so that its memory can be reclaimed. Suggested-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
- 10 Nov, 2009 8 commits
-
-
Paul E. McKenney authored
The rdp->passed_quiesc_completed fields are used to properly associate the recorded quiescent state with a grace period. It is OK to wrongly associate a given quiescent state with a preceding grace period, but it is fatal to associate a given quiescent state with a grace period that begins after the quiescent state occurred. Grace periods are numbered, and the following fields track them: o ->gpnum is the number of the grace period currently in progress, or the number of the last grace period to complete if no grace period is currently in progress. o ->completed is the number of the last grace period to have completed. These two fields are equal if there is no grace period in progress, otherwise ->gpnum is one greater than ->completed. But the rdp->passed_quiesc_completed field compared against ->completed, and if equal, the quiescent state is presumed to count against the current grace period. The earlier code copied rdp->completed to rdp->passed_quiesc_completed, which has been made to work, but is error-prone. In contrast, copying one less than rdp->gpnum is guaranteed safe, because rdp->gpnum is not incremented until after the start of the corresponding grace period. At the end of the grace period, when ->completed has incremented, then any quiescent periods recorded previously will be discarded. Signed-off-by:
-
Paul E. McKenney authored
This field is used whether or not CONFIG_NO_HZ is set, so the old name of ->dynticks_completed is quite misleading. Change to ->completed_fqs, given that it the value that force_quiescent_state() is trying to drive the ->completed field away from. Signed-off-by:
-
Paul E. McKenney authored
This patch adds a counter increment to enable tasks to actually take the synchronize_sched_expedited() function's fastpath. Signed-off-by:
-
Paul E. McKenney authored
Some variants of gcc are reputed to dislike forward references to functions declared "inline". Remove the "inline" keyword from such functions. Signed-off-by:
-
Paul E. McKenney authored
Impose a clear locking design on the note_new_gpnum() function's use of the ->gpnum counter. This is done by updating rdp->gpnum only from the corresponding leaf rcu_node structure's rnp->gpnum field, and even then only under the protection of that same rcu_node structure's ->lock field. Performance and scalability are maintained using a form of double-checked locking, and excessive spinning is avoided by use of the spin_trylock() function. The use of spin_trylock() is safe due to the fact that CPUs who fail to acquire this lock will try again later. The hierarchical nature of the rcu_node data structure limits contention (which could be limited further if need be using the RCU_FANOUT kernel parameter). Without this patch, obscure but quite possible races could result in a quiescent state that occurred during one grace period to be accounted to the following grace period, causing this following grace period to end prematurely. Not good! Signed-off-by:
-
Paul E. McKenney authored
Impose a clear locking design on the rcu_process_gp_end() function's use of the ->completed counter. This is done by creating a ->completed field in the rcu_node structure, which can safely be accessed under the protection of that structure's lock. Performance and scalability are maintained by using a form of double-checked locking, so that rcu_process_gp_end() only acquires the leaf rcu_node structure's ->lock if a grace period has recently ended. This fix reduces rcutorture failure rate by at least two orders of magnitude under heavy stress with force_quiescent_state() being invoked artificially often. Without this fix, unsynchronized access to the ->completed field can cause rcu_process_gp_end() to advance callbacks whose grace period has not yet expired. (Bad idea!) Signed-off-by:
-
Paul E. McKenney authored
Impose a clear locking design on non-NO_HZ handling of the ->completed counter. This increases the distance between the RCU and the CPU-hotplug mechanisms. Signed-off-by:
-
Ingo Molnar authored
Merge reason: Pick up RCU fixlet to base further commits on. Signed-off-by:
-
- 02 Nov, 2009 3 commits
-
-
Lai Jiangshan authored
Currently, rcu_irq_exit() is invoked only for CONFIG_NO_HZ, while rcu_irq_enter() is invoked unconditionally. This patch moves rcu_irq_exit() out from under CONFIG_NO_HZ so that the calls are balanced. This patch has no effect on the behavior of the kernel because both rcu_irq_enter() and rcu_irq_exit() are empty for !CONFIG_NO_HZ, but the code is easier to understand if the calls are obviously balanced in all cases. Signed-off-by:
-
Paul E. McKenney authored
Very long RCU read-side critical sections (50 milliseconds or so) can cause a race between force_quiescent_state() and rcu_start_gp() as follows on kernel builds with multi-level rcu_node hierarchies: 1. CPU 0 calls force_quiescent_state(), sees that there is a grace period in progress, and acquires ->fsqlock. 2. CPU 1 detects the end of the grace period, and so cpu_quiet_msk_finish() sets rsp->completed to rsp->gpnum. This operation is carried out under the root rnp->lock, but CPU 0 has not yet acquired that lock. Note that rsp->signaled is still RCU_SAVE_DYNTICK from the last grace period. 3. CPU 1 calls rcu_start_gp(), but no one wants a new grace period, so it drops the root rnp->lock and returns. 4. CPU 0 acquires the root rnp->lock and picks up rsp->completed and rsp->signaled, then drops rnp->lock. It then enters the RCU_SAVE_DYNTICK leg of the switch statement. 5. CPU 2 invokes call_rcu(), and now needs a new grace period. It calls rcu_start_gp(), which acquires the root rnp->lock, sets rsp->signaled to RCU_GP_INIT (too bad that CPU 0 is already in the RCU_SAVE_DYNTICK leg of the switch statement!) and starts initializing the rcu_node hierarchy. If there are multiple levels to the hierarchy, it will drop the root rnp->lock and initialize the lower levels of the hierarchy. 6. CPU 0 notes that rsp->completed has not changed, which permits both CPU 2 and CPU 0 to try updating it concurrently. If CPU 0's update prevails, later calls to force_quiescent_state() can count old quiescent states against the new grace period, which can in turn result in premature ending of grace periods. Not good. This patch adds an RCU_GP_IDLE state for rsp->signaled that is set initially at boot time and any time a grace period ends. This prevents CPU 0 from getting into the workings of force_quiescent_state() in step 4. Additional locking and checks prevent the concurrent update of rsp->signaled in step 6. Signed-off-by: -
Thomas Gleixner authored
Ingo triggered the following warning: WARNING: at lib/debugobjects.c:255 debug_print_object+0x42/0x50() Hardware name: System Product Name ODEBUG: init active object type: timer_list Modules linked in: Pid: 2619, comm: dmesg Tainted: G W 2.6.32-rc5-tip+ #5298 Call Trace: [<81035443>] warn_slowpath_common+0x6a/0x81 [<8120e483>] ? debug_print_object+0x42/0x50 [<81035498>] warn_slowpath_fmt+0x29/0x2c [<8120e483>] debug_print_object+0x42/0x50 [<8120ec2a>] __debug_object_init+0x279/0x2d7 [<8120ecb3>] debug_object_init+0x13/0x18 [<810409d2>] init_timer_key+0x17/0x6f [<81041526>] free_uid+0x50/0x6c [<8104ed2d>] put_cred_rcu+0x61/0x72 [<81067fac>] rcu_do_batch+0x70/0x121 debugobjects warns about an enqueued timer being initialized. If CONFIG_USER_SCHED=y the user management code uses delayed work to remove the user from the hash table and tear down the sysfs objects. free_uid is called from RCU and initializes/schedules delayed work if the usage count of the user_struct is 0. The init/schedule happens outside of the uidhash_lock protected region which allows a concurrent caller of find_user() to reference the about to be destroyed user_struct w/o preventing the work from being scheduled. If the next free_uid call happens before the work timer expired then the active timer is initialized and the work scheduled again. The race was introduced in commit 5cb350ba (sched: group scheduling, sysfs tunables) and made more prominent by commit 3959214f (sched: delayed cleanup of user_struct) Move the init/schedule_delayed_work inside of the uidhash_lock protected region to prevent the race. Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Acked-by:
Dhaval Giani <dhaval@linux.vnet.ibm.com> Cc: Paul E. McKenney <paulmck@us.ibm.com> Cc: Kay Sievers <kay.sievers@vrfy.org> Cc: stable@kernel.org
-
- 28 Oct, 2009 1 commit
-
-
Thomas Gleixner authored
The requeue_pi path doesn't use unqueue_me() (and the racy lock_ptr == NULL test) nor does it use the wake_list of futex_wake() which where the reason for commit 41890f2 (futex: Handle spurious wake up) See debugging discussing on LKML Message-ID: <4AD4080C.20703@us.ibm.com> The changes in this fix to the wait_requeue_pi path were considered to be a likely unecessary, but harmless safety net. But it turns out that due to the fact that for unknown $@#!*( reasons EWOULDBLOCK is defined as EAGAIN we built an endless loop in the code path which returns correctly EWOULDBLOCK. Spurious wakeups in wait_requeue_pi code path are unlikely so we do the easy solution and return EWOULDBLOCK^WEAGAIN to user space and let it deal with the spurious wakeup. Cc: Darren Hart <dvhltc@us.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: John Stultz <johnstul@linux.vnet.ibm.com> Cc: Dinakar Guniguntala <dino@in.ibm.com> LKML-Reference: <4AE23C74.1090502@us.ibm.com> Cc: stable@kernel.org Signed-off-by:
-
- 27 Oct, 2009 1 commit
-
-
Paul E. McKenney authored
Some compilers are happy with "#elif CONFIG_RCU_TINY", while others strongly prefer "#elif defined(CONFIG_RCU_TINY)". Change to the latter to make more compilers happy. Signed-off-by:
-
- 26 Oct, 2009 7 commits
-
-
Peter Zijlstra authored
Use lockdep_set_class() to simplify the code and to avoid any additional overhead in the !LOCKDEP case. Also move the definition of rcu_root_class into kernel/rcutree.c, as suggested by Lai Jiangshan. Signed-off-by:
Peter Zijlstra <peterz@infradead.org> Signed-off-by:
-
Ingo Molnar authored
No change in functionality - just straighten out a few small stylistic details. Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: David Howells <dhowells@redhat.com> Cc: Josh Triplett <josh@joshtriplett.org> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: mathieu.desnoyers@polymtl.ca Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org Cc: Valdis.Kletnieks@vt.edu Cc: avi@redhat.com Cc: mtosatti@redhat.com LKML-Reference: <12565226351355-git-send-email-> Signed-off-by: -
Paul E. McKenney authored
Make rcutorture list the available torture_type values when it doesn't like the one specified. Signed-off-by:
-
Paul E. McKenney authored
Signed-off-by:
-
Paul E. McKenney authored
Adds the "srcu_expedited" torture type, and also renames sched_ops_sync to sched_sync_ops for consistency while we are in this file. Signed-off-by:
-
Paul E. McKenney authored
This patch creates a synchronize_srcu_expedited() that uses synchronize_sched_expedited() where synchronize_srcu() uses synchronize_sched(). The synchronize_srcu() and synchronize_srcu_expedited() functions become one-liners that pass synchronize_sched() or synchronize_sched_expedited(), repectively, to a new __synchronize_srcu() function. While in the file, move the EXPORT_SYMBOL_GPL()s to immediately follow the corresponding functions. Requested-by:
Avi Kivity <avi@redhat.com> Tested-by:
Marcelo Tosatti <mtosatti@redhat.com> Signed-off-by:
-
Paul E. McKenney authored
This patch is a version of RCU designed for !SMP provided for a small-footprint RCU implementation. In particular, the implementation of synchronize_rcu() is extremely lightweight and high performance. It passes rcutorture testing in each of the four relevant configurations (combinations of NO_HZ and PREEMPT) on x86. This saves about 1K bytes compared to old Classic RCU (which is no longer in mainline), and more than three kilobytes compared to Hierarchical RCU (updated to 2.6.30): CONFIG_TREE_RCU: text data bss dec filename 183 4 0 187 kernel/rcupdate.o 2783 520 36 3339 kernel/rcutree.o 3526 Total (vs 4565 for v7) CONFIG_TREE_PREEMPT_RCU: text data bss dec filename 263 4 0 267 kernel/rcupdate.o 4594 776 52 5422 kernel/rcutree.o 5689 Total (6155 for v7) CONFIG_TINY_RCU: text data bss dec filename 96 4 0 100 kernel/rcupdate.o 734 24 0 758 kernel/rcutiny.o 858 Total (vs 848 for v7) The above is for x86. Your mileage may vary on other platforms. Further compression is possible, but is being procrastinated. Changes from v7 (http://lkml.org/lkml/2009/10/9/388) o Apply Lai Jiangshan's review comments (aside from might_sleep() in synchronize_sched(), which is covered by SMP builds). o Fix up expedited primitives. Changes from v6 (http://lkml.org/lkml/2009/9/23/293). o Forward ported to put it into the 2.6.33 stream. o Added lockdep support. o Make lightweight rcu_barrier. Changes from v5 (http://lkml.org/lkml/2009/6/23/12). o Ported to latest pre-2.6.32 merge window kernel. - Renamed rcu_qsctr_inc() to rcu_sched_qs(). - Renamed rcu_bh_qsctr_inc() to rcu_bh_qs(). - Provided trivial rcu_cpu_notify(). - Provided trivial exit_rcu(). - Provided trivial rcu_needs_cpu(). - Fixed up the rcu_*_enter/exit() functions in linux/hardirq.h. o Removed the dependence on EMBEDDED, with a view to making TINY_RCU default for !SMP at some time in the future. o Added (trivial) support for expedited grace periods. Changes from v4 (http://lkml.org/lkml/2009/5/2/91) include: o Squeeze the size down a bit further by removing the ->completed field from struct rcu_ctrlblk. o This permits synchronize_rcu() to become the empty function. Previous concerns about rcutorture were unfounded, as rcutorture correctly handles a constant value from rcu_batches_completed() and rcu_batches_completed_bh(). Changes from v3 (http://lkml.org/lkml/2009/3/29/221) include: o Changed rcu_batches_completed(), rcu_batches_completed_bh() rcu_enter_nohz(), rcu_exit_nohz(), rcu_nmi_enter(), and rcu_nmi_exit(), to be static inlines, as suggested by David Howells. Doing this saves about 100 bytes from rcutiny.o. (The numbers between v3 and this v4 of the patch are not directly comparable, since they are against different versions of Linux.) Changes from v2 (http://lkml.org/lkml/2009/2/3/333) include: o Fix whitespace issues. o Change short-circuit "||" operator to instead be "+" in order to fix performance bug noted by "kraai" on LWN. (http://lwn.net/Articles/324348/) Changes from v1 (http://lkml.org/lkml/2009/1/13/440) include: o This version depends on EMBEDDED as well as !SMP, as suggested by Ingo. o Updated rcu_needs_cpu() to unconditionally return zero, permitting the CPU to enter dynticks-idle mode at any time. This works because callbacks can be invoked upon entry to dynticks-idle mode. o Paul is now OK with this being included, based on a poll at the Kernel Miniconf at linux.conf.au, where about ten people said that they cared about saving 900 bytes on single-CPU systems. o Applies to both mainline and tip/core/rcu. Signed-off-by:
David Howells <dhowells@redhat.com> Acked-by:
-
- 16 Oct, 2009 1 commit
-
-
Darren Hart authored
When requeuing tasks from one futex to another, the reference held by the requeued task to the original futex location needs to be dropped eventually. Dropping the reference may ultimately lead to a call to "iput_final" and subsequently call into filesystem- specific code - which may be non-atomic. It is therefore safer to defer this drop operation until after the futex_hash_bucket spinlock has been dropped. Originally-From: Helge Bahmann <hcb@chaoticmind.net> Signed-off-by:
Darren Hart <dvhltc@us.ibm.com> Cc: <stable@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Dinakar Guniguntala <dino@in.ibm.com> Cc: John Stultz <johnstul@linux.vnet.ibm.com> Cc: Sven-Thorsten Dietrich <sdietrich@novell.com> Cc: John Kacur <jkacur@redhat.com> LKML-Reference: <4AD7A298.5040802@us.ibm.com> Signed-off-by:
-
- 15 Oct, 2009 6 commits
-
-
Paul E. McKenney authored
If the following sequence of events occurs, then TREE_PREEMPT_RCU will hang waiting for a grace period to complete, eventually OOMing the system: o A TREE_PREEMPT_RCU build of the kernel is booted on a system with more than 64 physical CPUs present (32 on a 32-bit system). Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted with RCU_FANOUT set to a sufficiently small value that the physical CPUs populate two or more leaf rcu_node structures. o A task is preempted in an RCU read-side critical section while running on a CPU corresponding to a given leaf rcu_node structure. o All CPUs corresponding to this same leaf rcu_node structure record quiescent states for the current grace period. o All of these same CPUs go offline (hence the need for enough physical CPUs to populate more than one leaf rcu_node structure). This causes the preempted task to be moved to the root rcu_node structure. At this point, there is nothing left to cause the quiescent state to be propagated up the rcu_node tree, so the current grace period never completes. The simplest fix, especially after considering the deadlock possibilities, is to detect this situation when the last CPU is offlined, and to set that CPU's ->qsmask bit in its leaf rcu_node structure. This will cause the next invocation of force_quiescent_state() to end the grace period. Without this fix, this hang can be triggered in an hour or so on some machines with rcutorture and random CPU onlining/offlining. With this fix, these same machines pass a full 10 hours of this sort of abuse. Signed-off-by:
-
Paul E. McKenney authored
Signed-off-by:
-
Paul E. McKenney authored
o Remove the CONFIG_PREEMPT_RCU documentation since this config option has now been removed. o Change the now-incorrect references to "rcu" labels to instead be "rcu_sched". o Add notes stating that CONFIG_TREE_PREEMPT_RCU kernels will have additional "rcu_preempt" output. o Note the new "oqlen" field in the rcuhier output (for RCU callbacks orphaned by an offlined CPU). Signed-off-by:
-
Paul E. McKenney authored
Signed-off-by:
-
Paul E. McKenney authored
For the short term, map synchronize_rcu_expedited() to synchronize_rcu() for TREE_PREEMPT_RCU and to synchronize_sched_expedited() for TREE_RCU. Longer term, there needs to be a real expedited grace period for TREE_PREEMPT_RCU, but candidate patches to date are considerably more complex and intrusive. Signed-off-by:
-
Paul E. McKenney authored
As the number of callbacks on a given CPU rises, invoke force_quiescent_state() only every blimit number of callbacks (defaults to 10,000), and even then only if no other CPU has invoked force_quiescent_state() in the meantime. This should fix the performance regression reported by Nick. Reported-by:
Nick Piggin <npiggin@suse.de> Signed-off-by:
-
- 14 Oct, 2009 1 commit
-
-
Darren Hart authored
If userspace tries to perform a requeue_pi on a non-requeue_pi waiter, it will find the futex_q->requeue_pi_key to be NULL and OOPS. Check for NULL in match_futex() instead of doing explicit NULL pointer checks on all call sites. While match_futex(NULL, NULL) returning false is a little odd, it's still correct as we expect valid key references. Signed-off-by:
-
- 13 Oct, 2009 1 commit
-
-
Thomas Gleixner authored
The futex code does not handle spurious wake up in futex_wait and futex_wait_requeue_pi. The code assumes that any wake up which was not caused by futex_wake / requeue or by a timeout was caused by a signal wake up and returns one of the syscall restart error codes. In case of a spurious wake up the signal delivery code which deals with the restart error codes is not invoked and we return that error code to user space. That causes applications which actually check the return codes to fail. Blaise reported that on preempt-rt a python test program run into a exception trap. -rt exposed that due to a built in spurious wake up accelerator :) Solve this by checking signal_pending(current) in the wake up path and handle the spurious wake up case w/o returning to user space. Reported-by:
Blaise Gassend <blaise@willowgarage.com> Debugged-by:
-
