linux/kernel/rcu/tree_plugin.h

2604 lines
79 KiB
C
Raw Normal View History

rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Read-Copy Update mechanism for mutual exclusion (tree-based version)
* Internal non-public definitions that provide either classic
* or preemptible semantics.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation; either version 2 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, you can access it online at
* http://www.gnu.org/licenses/gpl-2.0.html.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*
* Copyright Red Hat, 2009
* Copyright IBM Corporation, 2009
*
* Author: Ingo Molnar <mingo@elte.hu>
* Paul E. McKenney <paulmck@linux.vnet.ibm.com>
*/
#include <linux/delay.h>
#include <linux/gfp.h>
#include <linux/oom.h>
#include <linux/sched/debug.h>
#include <linux/smpboot.h>
#include <uapi/linux/sched/types.h>
#include "../time/tick-internal.h"
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
#ifdef CONFIG_RCU_BOOST
#include "../locking/rtmutex_common.h"
/*
* Control variables for per-CPU and per-rcu_node kthreads. These
* handle all flavors of RCU.
*/
static DEFINE_PER_CPU(struct task_struct *, rcu_cpu_kthread_task);
DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status);
DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops);
DEFINE_PER_CPU(char, rcu_cpu_has_work);
#else /* #ifdef CONFIG_RCU_BOOST */
/*
* Some architectures do not define rt_mutexes, but if !CONFIG_RCU_BOOST,
* all uses are in dead code. Provide a definition to keep the compiler
* happy, but add WARN_ON_ONCE() to complain if used in the wrong place.
* This probably needs to be excluded from -rt builds.
*/
#define rt_mutex_owner(a) ({ WARN_ON_ONCE(1); NULL; })
#endif /* #else #ifdef CONFIG_RCU_BOOST */
#ifdef CONFIG_RCU_NOCB_CPU
static cpumask_var_t rcu_nocb_mask; /* CPUs to have callbacks offloaded. */
static bool have_rcu_nocb_mask; /* Was rcu_nocb_mask allocated? */
rcu: Make rcu_nocb_poll an early_param instead of module_param The as-documented rcu_nocb_poll will fail to enable this feature for two reasons. (1) there is an extra "s" in the documented name which is not in the code, and (2) since it uses module_param, it really is expecting a prefix, akin to "rcutree.fanout_leaf" and the prefix isn't documented. However, there are several reasons why we might not want to simply fix the typo and add the prefix: 1) we'd end up with rcutree.rcu_nocb_poll, and rather probably make a change to rcutree.nocb_poll 2) if we did #1, then the prefix wouldn't be consistent with the rcu_nocbs=<cpumap> parameter (i.e. one with, one without prefix) 3) the use of module_param in a header file is less than desired, since it isn't immediately obvious that it will get processed via rcutree.c and get the prefix from that (although use of module_param_named() could clarify that.) 4) the implied export of /sys/module/rcutree/parameters/rcu_nocb_poll data to userspace via module_param() doesn't really buy us anything, as it is read-only and we can tell if it is enabled already without it, since there is a printk at early boot telling us so. In light of all that, just change it from a module_param() to an early_setup() call, and worry about adding it to /sys later on if we decide to allow a dynamic setting of it. Also change the variable to be tagged as read_mostly, since it will only ever be fiddled with at most, once at boot. Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2012-12-21 05:19:22 +08:00
static bool __read_mostly rcu_nocb_poll; /* Offload kthread are to poll. */
#endif /* #ifdef CONFIG_RCU_NOCB_CPU */
/*
* Check the RCU kernel configuration parameters and print informative
* messages about anything out of the ordinary.
*/
static void __init rcu_bootup_announce_oddness(void)
{
if (IS_ENABLED(CONFIG_RCU_TRACE))
pr_info("\tRCU event tracing is enabled.\n");
if ((IS_ENABLED(CONFIG_64BIT) && RCU_FANOUT != 64) ||
(!IS_ENABLED(CONFIG_64BIT) && RCU_FANOUT != 32))
pr_info("\tCONFIG_RCU_FANOUT set to non-default value of %d\n",
RCU_FANOUT);
if (rcu_fanout_exact)
pr_info("\tHierarchical RCU autobalancing is disabled.\n");
if (IS_ENABLED(CONFIG_RCU_FAST_NO_HZ))
pr_info("\tRCU dyntick-idle grace-period acceleration is enabled.\n");
if (IS_ENABLED(CONFIG_PROVE_RCU))
pr_info("\tRCU lockdep checking is enabled.\n");
if (RCU_NUM_LVLS >= 4)
pr_info("\tFour(or more)-level hierarchy is enabled.\n");
if (RCU_FANOUT_LEAF != 16)
pr_info("\tBuild-time adjustment of leaf fanout to %d.\n",
RCU_FANOUT_LEAF);
if (rcu_fanout_leaf != RCU_FANOUT_LEAF)
pr_info("\tBoot-time adjustment of leaf fanout to %d.\n", rcu_fanout_leaf);
if (nr_cpu_ids != NR_CPUS)
pr_info("\tRCU restricting CPUs from NR_CPUS=%d to nr_cpu_ids=%u.\n", NR_CPUS, nr_cpu_ids);
#ifdef CONFIG_RCU_BOOST
pr_info("\tRCU priority boosting: priority %d delay %d ms.\n", kthread_prio, CONFIG_RCU_BOOST_DELAY);
#endif
if (blimit != DEFAULT_RCU_BLIMIT)
pr_info("\tBoot-time adjustment of callback invocation limit to %ld.\n", blimit);
if (qhimark != DEFAULT_RCU_QHIMARK)
pr_info("\tBoot-time adjustment of callback high-water mark to %ld.\n", qhimark);
if (qlowmark != DEFAULT_RCU_QLOMARK)
pr_info("\tBoot-time adjustment of callback low-water mark to %ld.\n", qlowmark);
if (jiffies_till_first_fqs != ULONG_MAX)
pr_info("\tBoot-time adjustment of first FQS scan delay to %ld jiffies.\n", jiffies_till_first_fqs);
if (jiffies_till_next_fqs != ULONG_MAX)
pr_info("\tBoot-time adjustment of subsequent FQS scan delay to %ld jiffies.\n", jiffies_till_next_fqs);
if (rcu_kick_kthreads)
pr_info("\tKick kthreads if too-long grace period.\n");
if (IS_ENABLED(CONFIG_DEBUG_OBJECTS_RCU_HEAD))
pr_info("\tRCU callback double-/use-after-free debug enabled.\n");
if (gp_preinit_delay)
pr_info("\tRCU debug GP pre-init slowdown %d jiffies.\n", gp_preinit_delay);
if (gp_init_delay)
pr_info("\tRCU debug GP init slowdown %d jiffies.\n", gp_init_delay);
if (gp_cleanup_delay)
pr_info("\tRCU debug GP init slowdown %d jiffies.\n", gp_cleanup_delay);
if (IS_ENABLED(CONFIG_RCU_EQS_DEBUG))
pr_info("\tRCU debug extended QS entry/exit.\n");
rcupdate_announce_bootup_oddness();
}
#ifdef CONFIG_PREEMPT_RCU
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
RCU_STATE_INITIALIZER(rcu_preempt, 'p', call_rcu);
static struct rcu_state *const rcu_state_p = &rcu_preempt_state;
static struct rcu_data __percpu *const rcu_data_p = &rcu_preempt_data;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
rcu: Don't migrate blocked tasks even if all corresponding CPUs offline When the last CPU associated with a given leaf rcu_node structure goes offline, something must be done about the tasks queued on that rcu_node structure. Each of these tasks has been preempted on one of the leaf rcu_node structure's CPUs while in an RCU read-side critical section that it have not yet exited. Handling these tasks is the job of rcu_preempt_offline_tasks(), which migrates them from the leaf rcu_node structure to the root rcu_node structure. Unfortunately, this migration has to be done one task at a time because each tasks allegiance must be shifted from the original leaf rcu_node to the root, so that future attempts to deal with these tasks will acquire the root rcu_node structure's ->lock rather than that of the leaf. Worse yet, this migration must be done with interrupts disabled, which is not so good for realtime response, especially given that there is no bound on the number of tasks on a given rcu_node structure's list. (OK, OK, there is a bound, it is just that it is unreasonably large, especially on 64-bit systems.) This was not considered a problem back when rcu_preempt_offline_tasks() was first written because realtime systems were assumed not to do CPU-hotplug operations while real-time applications were running. This assumption has proved of dubious validity given that people are starting to run multiple realtime applications on a single SMP system and that it is common practice to offline then online a CPU before starting its real-time application in order to clear extraneous processing off of that CPU. So we now need CPU hotplug operations to avoid undue latencies. This commit therefore avoids migrating these tasks, instead letting them be dequeued one by one from the original leaf rcu_node structure by rcu_read_unlock_special(). This means that the clearing of bits from the upper-level rcu_node structures must be deferred until the last such task has been dequeued, because otherwise subsequent grace periods won't wait on them. This commit has the beneficial side effect of simplifying the CPU-hotplug code for TREE_PREEMPT_RCU, especially in CONFIG_RCU_BOOST builds. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-11-01 03:56:16 +08:00
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
bool wake);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Tell them what RCU they are running.
*/
static void __init rcu_bootup_announce(void)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
pr_info("Preemptible hierarchical RCU implementation.\n");
rcu_bootup_announce_oddness();
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/* Flags for rcu_preempt_ctxt_queue() decision table. */
#define RCU_GP_TASKS 0x8
#define RCU_EXP_TASKS 0x4
#define RCU_GP_BLKD 0x2
#define RCU_EXP_BLKD 0x1
/*
* Queues a task preempted within an RCU-preempt read-side critical
* section into the appropriate location within the ->blkd_tasks list,
* depending on the states of any ongoing normal and expedited grace
* periods. The ->gp_tasks pointer indicates which element the normal
* grace period is waiting on (NULL if none), and the ->exp_tasks pointer
* indicates which element the expedited grace period is waiting on (again,
* NULL if none). If a grace period is waiting on a given element in the
* ->blkd_tasks list, it also waits on all subsequent elements. Thus,
* adding a task to the tail of the list blocks any grace period that is
* already waiting on one of the elements. In contrast, adding a task
* to the head of the list won't block any grace period that is already
* waiting on one of the elements.
*
* This queuing is imprecise, and can sometimes make an ongoing grace
* period wait for a task that is not strictly speaking blocking it.
* Given the choice, we needlessly block a normal grace period rather than
* blocking an expedited grace period.
*
* Note that an endless sequence of expedited grace periods still cannot
* indefinitely postpone a normal grace period. Eventually, all of the
* fixed number of preempted tasks blocking the normal grace period that are
* not also blocking the expedited grace period will resume and complete
* their RCU read-side critical sections. At that point, the ->gp_tasks
* pointer will equal the ->exp_tasks pointer, at which point the end of
* the corresponding expedited grace period will also be the end of the
* normal grace period.
*/
static void rcu_preempt_ctxt_queue(struct rcu_node *rnp, struct rcu_data *rdp)
__releases(rnp->lock) /* But leaves rrupts disabled. */
{
int blkd_state = (rnp->gp_tasks ? RCU_GP_TASKS : 0) +
(rnp->exp_tasks ? RCU_EXP_TASKS : 0) +
(rnp->qsmask & rdp->grpmask ? RCU_GP_BLKD : 0) +
(rnp->expmask & rdp->grpmask ? RCU_EXP_BLKD : 0);
struct task_struct *t = current;
lockdep_assert_held(&rnp->lock);
WARN_ON_ONCE(rdp->mynode != rnp);
WARN_ON_ONCE(rnp->level != rcu_num_lvls - 1);
/*
* Decide where to queue the newly blocked task. In theory,
* this could be an if-statement. In practice, when I tried
* that, it was quite messy.
*/
switch (blkd_state) {
case 0:
case RCU_EXP_TASKS:
case RCU_EXP_TASKS + RCU_GP_BLKD:
case RCU_GP_TASKS:
case RCU_GP_TASKS + RCU_EXP_TASKS:
/*
* Blocking neither GP, or first task blocking the normal
* GP but not blocking the already-waiting expedited GP.
* Queue at the head of the list to avoid unnecessarily
* blocking the already-waiting GPs.
*/
list_add(&t->rcu_node_entry, &rnp->blkd_tasks);
break;
case RCU_EXP_BLKD:
case RCU_GP_BLKD:
case RCU_GP_BLKD + RCU_EXP_BLKD:
case RCU_GP_TASKS + RCU_EXP_BLKD:
case RCU_GP_TASKS + RCU_GP_BLKD + RCU_EXP_BLKD:
case RCU_GP_TASKS + RCU_EXP_TASKS + RCU_GP_BLKD + RCU_EXP_BLKD:
/*
* First task arriving that blocks either GP, or first task
* arriving that blocks the expedited GP (with the normal
* GP already waiting), or a task arriving that blocks
* both GPs with both GPs already waiting. Queue at the
* tail of the list to avoid any GP waiting on any of the
* already queued tasks that are not blocking it.
*/
list_add_tail(&t->rcu_node_entry, &rnp->blkd_tasks);
break;
case RCU_EXP_TASKS + RCU_EXP_BLKD:
case RCU_EXP_TASKS + RCU_GP_BLKD + RCU_EXP_BLKD:
case RCU_GP_TASKS + RCU_EXP_TASKS + RCU_EXP_BLKD:
/*
* Second or subsequent task blocking the expedited GP.
* The task either does not block the normal GP, or is the
* first task blocking the normal GP. Queue just after
* the first task blocking the expedited GP.
*/
list_add(&t->rcu_node_entry, rnp->exp_tasks);
break;
case RCU_GP_TASKS + RCU_GP_BLKD:
case RCU_GP_TASKS + RCU_EXP_TASKS + RCU_GP_BLKD:
/*
* Second or subsequent task blocking the normal GP.
* The task does not block the expedited GP. Queue just
* after the first task blocking the normal GP.
*/
list_add(&t->rcu_node_entry, rnp->gp_tasks);
break;
default:
/* Yet another exercise in excessive paranoia. */
WARN_ON_ONCE(1);
break;
}
/*
* We have now queued the task. If it was the first one to
* block either grace period, update the ->gp_tasks and/or
* ->exp_tasks pointers, respectively, to reference the newly
* blocked tasks.
*/
if (!rnp->gp_tasks && (blkd_state & RCU_GP_BLKD))
rnp->gp_tasks = &t->rcu_node_entry;
if (!rnp->exp_tasks && (blkd_state & RCU_EXP_BLKD))
rnp->exp_tasks = &t->rcu_node_entry;
WARN_ON_ONCE(!(blkd_state & RCU_GP_BLKD) !=
!(rnp->qsmask & rdp->grpmask));
WARN_ON_ONCE(!(blkd_state & RCU_EXP_BLKD) !=
!(rnp->expmask & rdp->grpmask));
raw_spin_unlock_rcu_node(rnp); /* interrupts remain disabled. */
/*
* Report the quiescent state for the expedited GP. This expedited
* GP should not be able to end until we report, so there should be
* no need to check for a subsequent expedited GP. (Though we are
* still in a quiescent state in any case.)
*/
if (blkd_state & RCU_EXP_BLKD &&
t->rcu_read_unlock_special.b.exp_need_qs) {
t->rcu_read_unlock_special.b.exp_need_qs = false;
rcu_report_exp_rdp(rdp->rsp, rdp, true);
} else {
WARN_ON_ONCE(t->rcu_read_unlock_special.b.exp_need_qs);
}
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Record a preemptible-RCU quiescent state for the specified CPU. Note
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
* that this just means that the task currently running on the CPU is
* not in a quiescent state. There might be any number of tasks blocked
* while in an RCU read-side critical section.
rcu: refactor RCU's context-switch handling The addition of preemptible RCU to treercu resulted in a bit of confusion and inefficiency surrounding the handling of context switches for RCU-sched and for RCU-preempt. For RCU-sched, a context switch is a quiescent state, pure and simple, just like it always has been. For RCU-preempt, a context switch is in no way a quiescent state, but special handling is required when a task blocks in an RCU read-side critical section. However, the callout from the scheduler and the outer loop in ksoftirqd still calls something named rcu_sched_qs(), whose name is no longer accurate. Furthermore, when rcu_check_callbacks() notes an RCU-sched quiescent state, it ends up unnecessarily (though harmlessly, aside from the performance hit) enqueuing the current task if it happens to be running in an RCU-preempt read-side critical section. This not only increases the maximum latency of scheduler_tick(), it also needlessly increases the overhead of the next outermost rcu_read_unlock() invocation. This patch addresses this situation by separating the notion of RCU's context-switch handling from that of RCU-sched's quiescent states. The context-switch handling is covered by rcu_note_context_switch() in general and by rcu_preempt_note_context_switch() for preemptible RCU. This permits rcu_sched_qs() to handle quiescent states and only quiescent states. It also reduces the maximum latency of scheduler_tick(), though probably by much less than a microsecond. Finally, it means that tasks within preemptible-RCU read-side critical sections avoid incurring the overhead of queuing unless there really is a context switch. Suggested-by: Lai Jiangshan <laijs@cn.fujitsu.com> Acked-by: Lai Jiangshan <laijs@cn.fujitsu.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <peterz@infradead.org>
2010-04-02 08:37:01 +08:00
*
* As with the other rcu_*_qs() functions, callers to this function
* must disable preemption.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
static void rcu_preempt_qs(void)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
RCU_LOCKDEP_WARN(preemptible(), "rcu_preempt_qs() invoked with preemption enabled!!!\n");
if (__this_cpu_read(rcu_data_p->cpu_no_qs.s)) {
trace_rcu_grace_period(TPS("rcu_preempt"),
__this_cpu_read(rcu_data_p->gpnum),
TPS("cpuqs"));
__this_cpu_write(rcu_data_p->cpu_no_qs.b.norm, false);
barrier(); /* Coordinate with rcu_preempt_check_callbacks(). */
current->rcu_read_unlock_special.b.need_qs = false;
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* We have entered the scheduler, and the current task might soon be
* context-switched away from. If this task is in an RCU read-side
* critical section, we will no longer be able to rely on the CPU to
* record that fact, so we enqueue the task on the blkd_tasks list.
* The task will dequeue itself when it exits the outermost enclosing
* RCU read-side critical section. Therefore, the current grace period
* cannot be permitted to complete until the blkd_tasks list entries
* predating the current grace period drain, in other words, until
* rnp->gp_tasks becomes NULL.
*
* Caller must disable interrupts.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
static void rcu_preempt_note_context_switch(bool preempt)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
struct task_struct *t = current;
struct rcu_data *rdp;
struct rcu_node *rnp;
RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_preempt_note_context_switch() invoked with interrupts enabled!!!\n");
WARN_ON_ONCE(!preempt && t->rcu_read_lock_nesting > 0);
rcu: protect __rcu_read_unlock() against scheduler-using irq handlers The addition of RCU read-side critical sections within runqueue and priority-inheritance lock critical sections introduced some deadlock cycles, for example, involving interrupts from __rcu_read_unlock() where the interrupt handlers call wake_up(). This situation can cause the instance of __rcu_read_unlock() invoked from interrupt to do some of the processing that would otherwise have been carried out by the task-level instance of __rcu_read_unlock(). When the interrupt-level instance of __rcu_read_unlock() is called with a scheduler lock held from interrupt-entry/exit situations where in_irq() returns false, deadlock can result. This commit resolves these deadlocks by using negative values of the per-task ->rcu_read_lock_nesting counter to indicate that an instance of __rcu_read_unlock() is in flight, which in turn prevents instances from interrupt handlers from doing any special processing. This patch is inspired by Steven Rostedt's earlier patch that similarly made __rcu_read_unlock() guard against interrupt-mediated recursion (see https://lkml.org/lkml/2011/7/15/326), but this commit refines Steven's approach to avoid the need for preemption disabling on the __rcu_read_unlock() fastpath and to also avoid the need for manipulating a separate per-CPU variable. This patch avoids need for preempt_disable() by instead using negative values of the per-task ->rcu_read_lock_nesting counter. Note that nested rcu_read_lock()/rcu_read_unlock() pairs are still permitted, but they will never see ->rcu_read_lock_nesting go to zero, and will therefore never invoke rcu_read_unlock_special(), thus preventing them from seeing the RCU_READ_UNLOCK_BLOCKED bit should it be set in ->rcu_read_unlock_special. This patch also adds a check for ->rcu_read_unlock_special being negative in rcu_check_callbacks(), thus preventing the RCU_READ_UNLOCK_NEED_QS bit from being set should a scheduling-clock interrupt occur while __rcu_read_unlock() is exiting from an outermost RCU read-side critical section. Of course, __rcu_read_unlock() can be preempted during the time that ->rcu_read_lock_nesting is negative. This could result in the setting of the RCU_READ_UNLOCK_BLOCKED bit after __rcu_read_unlock() checks it, and would also result it this task being queued on the corresponding rcu_node structure's blkd_tasks list. Therefore, some later RCU read-side critical section would enter rcu_read_unlock_special() to clean up -- which could result in deadlock if that critical section happened to be in the scheduler where the runqueue or priority-inheritance locks were held. This situation is dealt with by making rcu_preempt_note_context_switch() check for negative ->rcu_read_lock_nesting, thus refraining from queuing the task (and from setting RCU_READ_UNLOCK_BLOCKED) if we are already exiting from the outermost RCU read-side critical section (in other words, we really are no longer actually in that RCU read-side critical section). In addition, rcu_preempt_note_context_switch() invokes rcu_read_unlock_special() to carry out the cleanup in this case, which clears out the ->rcu_read_unlock_special bits and dequeues the task (if necessary), in turn avoiding needless delay of the current RCU grace period and needless RCU priority boosting. It is still illegal to call rcu_read_unlock() while holding a scheduler lock if the prior RCU read-side critical section has ever had either preemption or irqs enabled. However, the common use case is legal, namely where then entire RCU read-side critical section executes with irqs disabled, for example, when the scheduler lock is held across the entire lifetime of the RCU read-side critical section. Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-07-18 12:14:35 +08:00
if (t->rcu_read_lock_nesting > 0 &&
!t->rcu_read_unlock_special.b.blocked) {
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/* Possibly blocking in an RCU read-side critical section. */
rdp = this_cpu_ptr(rcu_state_p->rda);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
rnp = rdp->mynode;
raw_spin_lock_rcu_node(rnp);
t->rcu_read_unlock_special.b.blocked = true;
t->rcu_blocked_node = rnp;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Verify the CPU's sanity, trace the preemption, and
* then queue the task as required based on the states
* of any ongoing and expedited grace periods.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
rcu: Process offlining and onlining only at grace-period start Races between CPU hotplug and grace periods can be difficult to resolve, so the ->onoff_mutex is used to exclude the two events. Unfortunately, this means that it is impossible for an outgoing CPU to perform the last bits of its offlining from its last pass through the idle loop, because sleeplocks cannot be acquired in that context. This commit avoids these problems by buffering online and offline events in a new ->qsmaskinitnext field in the leaf rcu_node structures. When a grace period starts, the events accumulated in this mask are applied to the ->qsmaskinit field, and, if needed, up the rcu_node tree. The special case of all CPUs corresponding to a given leaf rcu_node structure being offline while there are still elements in that structure's ->blkd_tasks list is handled using a new ->wait_blkd_tasks field. In this case, propagating the offline bits up the tree is deferred until the beginning of the grace period after all of the tasks have exited their RCU read-side critical sections and removed themselves from the list, at which point the ->wait_blkd_tasks flag is cleared. If one of that leaf rcu_node structure's CPUs comes back online before the list empties, then the ->wait_blkd_tasks flag is simply cleared. This of course means that RCU's notion of which CPUs are offline can be out of date. This is OK because RCU need only wait on CPUs that were online at the time that the grace period started. In addition, RCU's force-quiescent-state actions will handle the case where a CPU goes offline after the grace period starts. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2015-01-24 13:52:37 +08:00
WARN_ON_ONCE((rdp->grpmask & rcu_rnp_online_cpus(rnp)) == 0);
WARN_ON_ONCE(!list_empty(&t->rcu_node_entry));
trace_rcu_preempt_task(rdp->rsp->name,
t->pid,
(rnp->qsmask & rdp->grpmask)
? rnp->gpnum
: rnp->gpnum + 1);
rcu_preempt_ctxt_queue(rnp, rdp);
rcu: protect __rcu_read_unlock() against scheduler-using irq handlers The addition of RCU read-side critical sections within runqueue and priority-inheritance lock critical sections introduced some deadlock cycles, for example, involving interrupts from __rcu_read_unlock() where the interrupt handlers call wake_up(). This situation can cause the instance of __rcu_read_unlock() invoked from interrupt to do some of the processing that would otherwise have been carried out by the task-level instance of __rcu_read_unlock(). When the interrupt-level instance of __rcu_read_unlock() is called with a scheduler lock held from interrupt-entry/exit situations where in_irq() returns false, deadlock can result. This commit resolves these deadlocks by using negative values of the per-task ->rcu_read_lock_nesting counter to indicate that an instance of __rcu_read_unlock() is in flight, which in turn prevents instances from interrupt handlers from doing any special processing. This patch is inspired by Steven Rostedt's earlier patch that similarly made __rcu_read_unlock() guard against interrupt-mediated recursion (see https://lkml.org/lkml/2011/7/15/326), but this commit refines Steven's approach to avoid the need for preemption disabling on the __rcu_read_unlock() fastpath and to also avoid the need for manipulating a separate per-CPU variable. This patch avoids need for preempt_disable() by instead using negative values of the per-task ->rcu_read_lock_nesting counter. Note that nested rcu_read_lock()/rcu_read_unlock() pairs are still permitted, but they will never see ->rcu_read_lock_nesting go to zero, and will therefore never invoke rcu_read_unlock_special(), thus preventing them from seeing the RCU_READ_UNLOCK_BLOCKED bit should it be set in ->rcu_read_unlock_special. This patch also adds a check for ->rcu_read_unlock_special being negative in rcu_check_callbacks(), thus preventing the RCU_READ_UNLOCK_NEED_QS bit from being set should a scheduling-clock interrupt occur while __rcu_read_unlock() is exiting from an outermost RCU read-side critical section. Of course, __rcu_read_unlock() can be preempted during the time that ->rcu_read_lock_nesting is negative. This could result in the setting of the RCU_READ_UNLOCK_BLOCKED bit after __rcu_read_unlock() checks it, and would also result it this task being queued on the corresponding rcu_node structure's blkd_tasks list. Therefore, some later RCU read-side critical section would enter rcu_read_unlock_special() to clean up -- which could result in deadlock if that critical section happened to be in the scheduler where the runqueue or priority-inheritance locks were held. This situation is dealt with by making rcu_preempt_note_context_switch() check for negative ->rcu_read_lock_nesting, thus refraining from queuing the task (and from setting RCU_READ_UNLOCK_BLOCKED) if we are already exiting from the outermost RCU read-side critical section (in other words, we really are no longer actually in that RCU read-side critical section). In addition, rcu_preempt_note_context_switch() invokes rcu_read_unlock_special() to carry out the cleanup in this case, which clears out the ->rcu_read_unlock_special bits and dequeues the task (if necessary), in turn avoiding needless delay of the current RCU grace period and needless RCU priority boosting. It is still illegal to call rcu_read_unlock() while holding a scheduler lock if the prior RCU read-side critical section has ever had either preemption or irqs enabled. However, the common use case is legal, namely where then entire RCU read-side critical section executes with irqs disabled, for example, when the scheduler lock is held across the entire lifetime of the RCU read-side critical section. Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-07-18 12:14:35 +08:00
} else if (t->rcu_read_lock_nesting < 0 &&
t->rcu_read_unlock_special.s) {
rcu: protect __rcu_read_unlock() against scheduler-using irq handlers The addition of RCU read-side critical sections within runqueue and priority-inheritance lock critical sections introduced some deadlock cycles, for example, involving interrupts from __rcu_read_unlock() where the interrupt handlers call wake_up(). This situation can cause the instance of __rcu_read_unlock() invoked from interrupt to do some of the processing that would otherwise have been carried out by the task-level instance of __rcu_read_unlock(). When the interrupt-level instance of __rcu_read_unlock() is called with a scheduler lock held from interrupt-entry/exit situations where in_irq() returns false, deadlock can result. This commit resolves these deadlocks by using negative values of the per-task ->rcu_read_lock_nesting counter to indicate that an instance of __rcu_read_unlock() is in flight, which in turn prevents instances from interrupt handlers from doing any special processing. This patch is inspired by Steven Rostedt's earlier patch that similarly made __rcu_read_unlock() guard against interrupt-mediated recursion (see https://lkml.org/lkml/2011/7/15/326), but this commit refines Steven's approach to avoid the need for preemption disabling on the __rcu_read_unlock() fastpath and to also avoid the need for manipulating a separate per-CPU variable. This patch avoids need for preempt_disable() by instead using negative values of the per-task ->rcu_read_lock_nesting counter. Note that nested rcu_read_lock()/rcu_read_unlock() pairs are still permitted, but they will never see ->rcu_read_lock_nesting go to zero, and will therefore never invoke rcu_read_unlock_special(), thus preventing them from seeing the RCU_READ_UNLOCK_BLOCKED bit should it be set in ->rcu_read_unlock_special. This patch also adds a check for ->rcu_read_unlock_special being negative in rcu_check_callbacks(), thus preventing the RCU_READ_UNLOCK_NEED_QS bit from being set should a scheduling-clock interrupt occur while __rcu_read_unlock() is exiting from an outermost RCU read-side critical section. Of course, __rcu_read_unlock() can be preempted during the time that ->rcu_read_lock_nesting is negative. This could result in the setting of the RCU_READ_UNLOCK_BLOCKED bit after __rcu_read_unlock() checks it, and would also result it this task being queued on the corresponding rcu_node structure's blkd_tasks list. Therefore, some later RCU read-side critical section would enter rcu_read_unlock_special() to clean up -- which could result in deadlock if that critical section happened to be in the scheduler where the runqueue or priority-inheritance locks were held. This situation is dealt with by making rcu_preempt_note_context_switch() check for negative ->rcu_read_lock_nesting, thus refraining from queuing the task (and from setting RCU_READ_UNLOCK_BLOCKED) if we are already exiting from the outermost RCU read-side critical section (in other words, we really are no longer actually in that RCU read-side critical section). In addition, rcu_preempt_note_context_switch() invokes rcu_read_unlock_special() to carry out the cleanup in this case, which clears out the ->rcu_read_unlock_special bits and dequeues the task (if necessary), in turn avoiding needless delay of the current RCU grace period and needless RCU priority boosting. It is still illegal to call rcu_read_unlock() while holding a scheduler lock if the prior RCU read-side critical section has ever had either preemption or irqs enabled. However, the common use case is legal, namely where then entire RCU read-side critical section executes with irqs disabled, for example, when the scheduler lock is held across the entire lifetime of the RCU read-side critical section. Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-07-18 12:14:35 +08:00
/*
* Complete exit from RCU read-side critical section on
* behalf of preempted instance of __rcu_read_unlock().
*/
rcu_read_unlock_special(t);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Either we were not in an RCU read-side critical section to
* begin with, or we have now recorded that critical section
* globally. Either way, we can now note a quiescent state
* for this CPU. Again, if we were in an RCU read-side critical
* section, and if that critical section was blocking the current
* grace period, then the fact that the task has been enqueued
* means that we continue to block the current grace period.
*/
rcu_preempt_qs();
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Check for preempted RCU readers blocking the current grace period
* for the specified rcu_node structure. If the caller needs a reliable
* answer, it must hold the rcu_node's ->lock.
*/
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp)
{
return rnp->gp_tasks != NULL;
}
/*
* Advance a ->blkd_tasks-list pointer to the next entry, instead
* returning NULL if at the end of the list.
*/
static struct list_head *rcu_next_node_entry(struct task_struct *t,
struct rcu_node *rnp)
{
struct list_head *np;
np = t->rcu_node_entry.next;
if (np == &rnp->blkd_tasks)
np = NULL;
return np;
}
/*
* Return true if the specified rcu_node structure has tasks that were
* preempted within an RCU read-side critical section.
*/
static bool rcu_preempt_has_tasks(struct rcu_node *rnp)
{
return !list_empty(&rnp->blkd_tasks);
}
rcu: Fix grace-period-stall bug on large systems with CPU hotplug 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: 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: <12589088301770-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-23 00:53:48 +08:00
/*
* Handle special cases during rcu_read_unlock(), such as needing to
* notify RCU core processing or task having blocked during the RCU
* read-side critical section.
*/
void rcu_read_unlock_special(struct task_struct *t)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
bool empty_exp;
bool empty_norm;
bool empty_exp_now;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
unsigned long flags;
struct list_head *np;
bool drop_boost_mutex = false;
struct rcu_data *rdp;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
struct rcu_node *rnp;
union rcu_special special;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/* NMI handlers cannot block and cannot safely manipulate state. */
if (in_nmi())
return;
local_irq_save(flags);
/*
* If RCU core is waiting for this CPU to exit its critical section,
* report the fact that it has exited. Because irqs are disabled,
* t->rcu_read_unlock_special cannot change.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
special = t->rcu_read_unlock_special;
if (special.b.need_qs) {
rcu_preempt_qs();
rcu: Clear need_qs flag to prevent splat 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: Sasha Levin <sasha.levin@oracle.com>
2015-01-23 14:47:14 +08:00
t->rcu_read_unlock_special.b.need_qs = false;
if (!t->rcu_read_unlock_special.s) {
local_irq_restore(flags);
return;
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Respond to a request for an expedited grace period, but only if
* we were not preempted, meaning that we were running on the same
* CPU throughout. If we were preempted, the exp_need_qs flag
* would have been cleared at the time of the first preemption,
* and the quiescent state would be reported when we were dequeued.
*/
if (special.b.exp_need_qs) {
WARN_ON_ONCE(special.b.blocked);
t->rcu_read_unlock_special.b.exp_need_qs = false;
rdp = this_cpu_ptr(rcu_state_p->rda);
rcu_report_exp_rdp(rcu_state_p, rdp, true);
if (!t->rcu_read_unlock_special.s) {
local_irq_restore(flags);
return;
}
}
/* Hardware IRQ handlers cannot block, complain if they get here. */
if (in_irq() || in_serving_softirq()) {
lockdep_rcu_suspicious(__FILE__, __LINE__,
"rcu_read_unlock() from irq or softirq with blocking in critical section!!!\n");
pr_alert("->rcu_read_unlock_special: %#x (b: %d, enq: %d nq: %d)\n",
t->rcu_read_unlock_special.s,
t->rcu_read_unlock_special.b.blocked,
t->rcu_read_unlock_special.b.exp_need_qs,
t->rcu_read_unlock_special.b.need_qs);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
local_irq_restore(flags);
return;
}
/* Clean up if blocked during RCU read-side critical section. */
if (special.b.blocked) {
t->rcu_read_unlock_special.b.blocked = false;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Remove this task from the list it blocked on. The task
* now remains queued on the rcu_node corresponding to the
* CPU it first blocked on, so there is no longer any need
* to loop. Retain a WARN_ON_ONCE() out of sheer paranoia.
*/
rnp = t->rcu_blocked_node;
raw_spin_lock_rcu_node(rnp); /* irqs already disabled. */
WARN_ON_ONCE(rnp != t->rcu_blocked_node);
WARN_ON_ONCE(rnp->level != rcu_num_lvls - 1);
empty_norm = !rcu_preempt_blocked_readers_cgp(rnp);
empty_exp = sync_rcu_preempt_exp_done(rnp);
smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
np = rcu_next_node_entry(t, rnp);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
list_del_init(&t->rcu_node_entry);
t->rcu_blocked_node = NULL;
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_unlock_preempted_task(TPS("rcu_preempt"),
rnp->gpnum, t->pid);
if (&t->rcu_node_entry == rnp->gp_tasks)
rnp->gp_tasks = np;
if (&t->rcu_node_entry == rnp->exp_tasks)
rnp->exp_tasks = np;
if (IS_ENABLED(CONFIG_RCU_BOOST)) {
/* Snapshot ->boost_mtx ownership w/rnp->lock held. */
drop_boost_mutex = rt_mutex_owner(&rnp->boost_mtx) == t;
if (&t->rcu_node_entry == rnp->boost_tasks)
rnp->boost_tasks = np;
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* If this was the last task on the current list, and if
* we aren't waiting on any CPUs, report the quiescent state.
rcu: Avoid RCU-preempt expedited grace-period botch Because rcu_read_unlock_special() samples rcu_preempted_readers_exp(rnp) after dropping rnp->lock, the following sequence of events is possible: 1. Task A exits its RCU read-side critical section, and removes itself from the ->blkd_tasks list, releases rnp->lock, and is then preempted. Task B remains on the ->blkd_tasks list, and blocks the current expedited grace period. 2. Task B exits from its RCU read-side critical section and removes itself from the ->blkd_tasks list. Because it is the last task blocking the current expedited grace period, it ends that expedited grace period. 3. Task A resumes, and samples rcu_preempted_readers_exp(rnp) which of course indicates that nothing is blocking the nonexistent expedited grace period. Task A is again preempted. 4. Some other CPU starts an expedited grace period. There are several tasks blocking this expedited grace period queued on the same rcu_node structure that Task A was using in step 1 above. 5. Task A examines its state and incorrectly concludes that it was the last task blocking the expedited grace period on the current rcu_node structure. It therefore reports completion up the rcu_node tree. 6. The expedited grace period can then incorrectly complete before the tasks blocked on this same rcu_node structure exit their RCU read-side critical sections. Arbitrarily bad things happen. This commit therefore takes a snapshot of rcu_preempted_readers_exp(rnp) prior to dropping the lock, so that only the last task thinks that it is the last task, thus avoiding the failure scenario laid out above. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2011-09-22 05:41:37 +08:00
* Note that rcu_report_unblock_qs_rnp() releases rnp->lock,
* so we must take a snapshot of the expedited state.
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
empty_exp_now = sync_rcu_preempt_exp_done(rnp);
if (!empty_norm && !rcu_preempt_blocked_readers_cgp(rnp)) {
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_quiescent_state_report(TPS("preempt_rcu"),
rnp->gpnum,
0, rnp->qsmask,
rnp->level,
rnp->grplo,
rnp->grphi,
!!rnp->gp_tasks);
rcu_report_unblock_qs_rnp(rcu_state_p, rnp, flags);
} else {
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
}
/* Unboost if we were boosted. */
if (IS_ENABLED(CONFIG_RCU_BOOST) && drop_boost_mutex)
rt_mutex_unlock(&rnp->boost_mtx);
/*
* If this was the last task on the expedited lists,
* then we need to report up the rcu_node hierarchy.
*/
rcu: Avoid RCU-preempt expedited grace-period botch Because rcu_read_unlock_special() samples rcu_preempted_readers_exp(rnp) after dropping rnp->lock, the following sequence of events is possible: 1. Task A exits its RCU read-side critical section, and removes itself from the ->blkd_tasks list, releases rnp->lock, and is then preempted. Task B remains on the ->blkd_tasks list, and blocks the current expedited grace period. 2. Task B exits from its RCU read-side critical section and removes itself from the ->blkd_tasks list. Because it is the last task blocking the current expedited grace period, it ends that expedited grace period. 3. Task A resumes, and samples rcu_preempted_readers_exp(rnp) which of course indicates that nothing is blocking the nonexistent expedited grace period. Task A is again preempted. 4. Some other CPU starts an expedited grace period. There are several tasks blocking this expedited grace period queued on the same rcu_node structure that Task A was using in step 1 above. 5. Task A examines its state and incorrectly concludes that it was the last task blocking the expedited grace period on the current rcu_node structure. It therefore reports completion up the rcu_node tree. 6. The expedited grace period can then incorrectly complete before the tasks blocked on this same rcu_node structure exit their RCU read-side critical sections. Arbitrarily bad things happen. This commit therefore takes a snapshot of rcu_preempted_readers_exp(rnp) prior to dropping the lock, so that only the last task thinks that it is the last task, thus avoiding the failure scenario laid out above. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2011-09-22 05:41:37 +08:00
if (!empty_exp && empty_exp_now)
rcu_report_exp_rnp(rcu_state_p, rnp, true);
rcu: Fix grace-period-stall bug on large systems with CPU hotplug 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: 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: <12589088301770-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-23 00:53:48 +08:00
} else {
local_irq_restore(flags);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
}
/*
* Dump detailed information for all tasks blocking the current RCU
* grace period on the specified rcu_node structure.
*/
static void rcu_print_detail_task_stall_rnp(struct rcu_node *rnp)
{
unsigned long flags;
struct task_struct *t;
raw_spin_lock_irqsave_rcu_node(rnp, flags);
if (!rcu_preempt_blocked_readers_cgp(rnp)) {
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
return;
}
t = list_entry(rnp->gp_tasks->prev,
struct task_struct, rcu_node_entry);
list_for_each_entry_continue(t, &rnp->blkd_tasks, rcu_node_entry)
sched_show_task(t);
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
}
/*
* Dump detailed information for all tasks blocking the current RCU
* grace period.
*/
static void rcu_print_detail_task_stall(struct rcu_state *rsp)
{
struct rcu_node *rnp = rcu_get_root(rsp);
rcu_print_detail_task_stall_rnp(rnp);
rcu_for_each_leaf_node(rsp, rnp)
rcu_print_detail_task_stall_rnp(rnp);
}
static void rcu_print_task_stall_begin(struct rcu_node *rnp)
{
pr_err("\tTasks blocked on level-%d rcu_node (CPUs %d-%d):",
rnp->level, rnp->grplo, rnp->grphi);
}
static void rcu_print_task_stall_end(void)
{
pr_cont("\n");
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Scan the current list of tasks blocked within RCU read-side critical
* sections, printing out the tid of each.
*/
static int rcu_print_task_stall(struct rcu_node *rnp)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
struct task_struct *t;
int ndetected = 0;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
if (!rcu_preempt_blocked_readers_cgp(rnp))
return 0;
rcu_print_task_stall_begin(rnp);
t = list_entry(rnp->gp_tasks->prev,
struct task_struct, rcu_node_entry);
list_for_each_entry_continue(t, &rnp->blkd_tasks, rcu_node_entry) {
pr_cont(" P%d", t->pid);
ndetected++;
}
rcu_print_task_stall_end();
return ndetected;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Scan the current list of tasks blocked within RCU read-side critical
* sections, printing out the tid of each that is blocking the current
* expedited grace period.
*/
static int rcu_print_task_exp_stall(struct rcu_node *rnp)
{
struct task_struct *t;
int ndetected = 0;
if (!rnp->exp_tasks)
return 0;
t = list_entry(rnp->exp_tasks->prev,
struct task_struct, rcu_node_entry);
list_for_each_entry_continue(t, &rnp->blkd_tasks, rcu_node_entry) {
pr_cont(" P%d", t->pid);
ndetected++;
}
return ndetected;
}
/*
* Check that the list of blocked tasks for the newly completed grace
* period is in fact empty. It is a serious bug to complete a grace
* period that still has RCU readers blocked! This function must be
* invoked -before- updating this rnp's ->gpnum, and the rnp's ->lock
* must be held by the caller.
*
* Also, if there are blocked tasks on the list, they automatically
* block the newly created grace period, so set up ->gp_tasks accordingly.
*/
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
{
struct task_struct *t;
RCU_LOCKDEP_WARN(preemptible(), "rcu_preempt_check_blocked_tasks() invoked with preemption enabled!!!\n");
WARN_ON_ONCE(rcu_preempt_blocked_readers_cgp(rnp));
if (rcu_preempt_has_tasks(rnp)) {
rnp->gp_tasks = rnp->blkd_tasks.next;
t = container_of(rnp->gp_tasks, struct task_struct,
rcu_node_entry);
trace_rcu_unlock_preempted_task(TPS("rcu_preempt-GPS"),
rnp->gpnum, t->pid);
}
WARN_ON_ONCE(rnp->qsmask);
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Check for a quiescent state from the current CPU. When a task blocks,
* the task is recorded in the corresponding CPU's rcu_node structure,
* which is checked elsewhere.
*
* Caller must disable hard irqs.
*/
static void rcu_preempt_check_callbacks(void)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
struct task_struct *t = current;
if (t->rcu_read_lock_nesting == 0) {
rcu_preempt_qs();
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
return;
}
rcu: protect __rcu_read_unlock() against scheduler-using irq handlers The addition of RCU read-side critical sections within runqueue and priority-inheritance lock critical sections introduced some deadlock cycles, for example, involving interrupts from __rcu_read_unlock() where the interrupt handlers call wake_up(). This situation can cause the instance of __rcu_read_unlock() invoked from interrupt to do some of the processing that would otherwise have been carried out by the task-level instance of __rcu_read_unlock(). When the interrupt-level instance of __rcu_read_unlock() is called with a scheduler lock held from interrupt-entry/exit situations where in_irq() returns false, deadlock can result. This commit resolves these deadlocks by using negative values of the per-task ->rcu_read_lock_nesting counter to indicate that an instance of __rcu_read_unlock() is in flight, which in turn prevents instances from interrupt handlers from doing any special processing. This patch is inspired by Steven Rostedt's earlier patch that similarly made __rcu_read_unlock() guard against interrupt-mediated recursion (see https://lkml.org/lkml/2011/7/15/326), but this commit refines Steven's approach to avoid the need for preemption disabling on the __rcu_read_unlock() fastpath and to also avoid the need for manipulating a separate per-CPU variable. This patch avoids need for preempt_disable() by instead using negative values of the per-task ->rcu_read_lock_nesting counter. Note that nested rcu_read_lock()/rcu_read_unlock() pairs are still permitted, but they will never see ->rcu_read_lock_nesting go to zero, and will therefore never invoke rcu_read_unlock_special(), thus preventing them from seeing the RCU_READ_UNLOCK_BLOCKED bit should it be set in ->rcu_read_unlock_special. This patch also adds a check for ->rcu_read_unlock_special being negative in rcu_check_callbacks(), thus preventing the RCU_READ_UNLOCK_NEED_QS bit from being set should a scheduling-clock interrupt occur while __rcu_read_unlock() is exiting from an outermost RCU read-side critical section. Of course, __rcu_read_unlock() can be preempted during the time that ->rcu_read_lock_nesting is negative. This could result in the setting of the RCU_READ_UNLOCK_BLOCKED bit after __rcu_read_unlock() checks it, and would also result it this task being queued on the corresponding rcu_node structure's blkd_tasks list. Therefore, some later RCU read-side critical section would enter rcu_read_unlock_special() to clean up -- which could result in deadlock if that critical section happened to be in the scheduler where the runqueue or priority-inheritance locks were held. This situation is dealt with by making rcu_preempt_note_context_switch() check for negative ->rcu_read_lock_nesting, thus refraining from queuing the task (and from setting RCU_READ_UNLOCK_BLOCKED) if we are already exiting from the outermost RCU read-side critical section (in other words, we really are no longer actually in that RCU read-side critical section). In addition, rcu_preempt_note_context_switch() invokes rcu_read_unlock_special() to carry out the cleanup in this case, which clears out the ->rcu_read_unlock_special bits and dequeues the task (if necessary), in turn avoiding needless delay of the current RCU grace period and needless RCU priority boosting. It is still illegal to call rcu_read_unlock() while holding a scheduler lock if the prior RCU read-side critical section has ever had either preemption or irqs enabled. However, the common use case is legal, namely where then entire RCU read-side critical section executes with irqs disabled, for example, when the scheduler lock is held across the entire lifetime of the RCU read-side critical section. Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-07-18 12:14:35 +08:00
if (t->rcu_read_lock_nesting > 0 &&
__this_cpu_read(rcu_data_p->core_needs_qs) &&
__this_cpu_read(rcu_data_p->cpu_no_qs.b.norm))
t->rcu_read_unlock_special.b.need_qs = true;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
#ifdef CONFIG_RCU_BOOST
rcu: Use softirq to address performance regression Commit a26ac2455ffcf3(rcu: move TREE_RCU from softirq to kthread) introduced performance regression. In an AIM7 test, this commit degraded performance by about 40%. The commit runs rcu callbacks in a kthread instead of softirq. We observed high rate of context switch which is caused by this. Out test system has 64 CPUs and HZ is 1000, so we saw more than 64k context switch per second which is caused by RCU's per-CPU kthread. A trace showed that most of the time the RCU per-CPU kthread doesn't actually handle any callbacks, but instead just does a very small amount of work handling grace periods. This means that RCU's per-CPU kthreads are making the scheduler do quite a bit of work in order to allow a very small amount of RCU-related processing to be done. Alex Shi's analysis determined that this slowdown is due to lock contention within the scheduler. Unfortunately, as Peter Zijlstra points out, the scheduler's real-time semantics require global action, which means that this contention is inherent in real-time scheduling. (Yes, perhaps someone will come up with a workaround -- otherwise, -rt is not going to do well on large SMP systems -- but this patch will work around this issue in the meantime. And "the meantime" might well be forever.) This patch therefore re-introduces softirq processing to RCU, but only for core RCU work. RCU callbacks are still executed in kthread context, so that only a small amount of RCU work runs in softirq context in the common case. This should minimize ksoftirqd execution, allowing us to skip boosting of ksoftirqd for CONFIG_RCU_BOOST=y kernels. Signed-off-by: Shaohua Li <shaohua.li@intel.com> Tested-by: "Alex,Shi" <alex.shi@intel.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-14 13:26:25 +08:00
static void rcu_preempt_do_callbacks(void)
{
rcu_do_batch(rcu_state_p, this_cpu_ptr(rcu_data_p));
rcu: Use softirq to address performance regression Commit a26ac2455ffcf3(rcu: move TREE_RCU from softirq to kthread) introduced performance regression. In an AIM7 test, this commit degraded performance by about 40%. The commit runs rcu callbacks in a kthread instead of softirq. We observed high rate of context switch which is caused by this. Out test system has 64 CPUs and HZ is 1000, so we saw more than 64k context switch per second which is caused by RCU's per-CPU kthread. A trace showed that most of the time the RCU per-CPU kthread doesn't actually handle any callbacks, but instead just does a very small amount of work handling grace periods. This means that RCU's per-CPU kthreads are making the scheduler do quite a bit of work in order to allow a very small amount of RCU-related processing to be done. Alex Shi's analysis determined that this slowdown is due to lock contention within the scheduler. Unfortunately, as Peter Zijlstra points out, the scheduler's real-time semantics require global action, which means that this contention is inherent in real-time scheduling. (Yes, perhaps someone will come up with a workaround -- otherwise, -rt is not going to do well on large SMP systems -- but this patch will work around this issue in the meantime. And "the meantime" might well be forever.) This patch therefore re-introduces softirq processing to RCU, but only for core RCU work. RCU callbacks are still executed in kthread context, so that only a small amount of RCU work runs in softirq context in the common case. This should minimize ksoftirqd execution, allowing us to skip boosting of ksoftirqd for CONFIG_RCU_BOOST=y kernels. Signed-off-by: Shaohua Li <shaohua.li@intel.com> Tested-by: "Alex,Shi" <alex.shi@intel.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-14 13:26:25 +08:00
}
#endif /* #ifdef CONFIG_RCU_BOOST */
/**
* call_rcu() - Queue an RCU callback for invocation after a grace period.
* @head: structure to be used for queueing the RCU updates.
* @func: actual callback function to be invoked after the grace period
*
* The callback function will be invoked some time after a full grace
* period elapses, in other words after all pre-existing RCU read-side
* critical sections have completed. However, the callback function
* might well execute concurrently with RCU read-side critical sections
* that started after call_rcu() was invoked. RCU read-side critical
* sections are delimited by rcu_read_lock() and rcu_read_unlock(),
* and may be nested.
*
* Note that all CPUs must agree that the grace period extended beyond
* all pre-existing RCU read-side critical section. On systems with more
* than one CPU, this means that when "func()" is invoked, each CPU is
* guaranteed to have executed a full memory barrier since the end of its
* last RCU read-side critical section whose beginning preceded the call
* to call_rcu(). It also means that each CPU executing an RCU read-side
* critical section that continues beyond the start of "func()" must have
* executed a memory barrier after the call_rcu() but before the beginning
* of that RCU read-side critical section. Note that these guarantees
* include CPUs that are offline, idle, or executing in user mode, as
* well as CPUs that are executing in the kernel.
*
* Furthermore, if CPU A invoked call_rcu() and CPU B invoked the
* resulting RCU callback function "func()", then both CPU A and CPU B are
* guaranteed to execute a full memory barrier during the time interval
* between the call to call_rcu() and the invocation of "func()" -- even
* if CPU A and CPU B are the same CPU (but again only if the system has
* more than one CPU).
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
*/
void call_rcu(struct rcu_head *head, rcu_callback_t func)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
__call_rcu(head, func, rcu_state_p, -1, 0);
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
EXPORT_SYMBOL_GPL(call_rcu);
/**
* synchronize_rcu - wait until a grace period has elapsed.
*
* Control will return to the caller some time after a full grace
* period has elapsed, in other words after all currently executing RCU
* read-side critical sections have completed. Note, however, that
* upon return from synchronize_rcu(), the caller might well be executing
* concurrently with new RCU read-side critical sections that began while
* synchronize_rcu() was waiting. RCU read-side critical sections are
* delimited by rcu_read_lock() and rcu_read_unlock(), and may be nested.
*
* See the description of synchronize_sched() for more detailed
* information on memory-ordering guarantees. However, please note
* that -only- the memory-ordering guarantees apply. For example,
* synchronize_rcu() is -not- guaranteed to wait on things like code
* protected by preempt_disable(), instead, synchronize_rcu() is -only-
* guaranteed to wait on RCU read-side critical sections, that is, sections
* of code protected by rcu_read_lock().
*/
void synchronize_rcu(void)
{
RCU_LOCKDEP_WARN(lock_is_held(&rcu_bh_lock_map) ||
lock_is_held(&rcu_lock_map) ||
lock_is_held(&rcu_sched_lock_map),
"Illegal synchronize_rcu() in RCU read-side critical section");
rcu: Narrow early boot window of illegal synchronous grace periods The current preemptible RCU implementation goes through three phases during bootup. In the first phase, there is only one CPU that is running with preemption disabled, so that a no-op is a synchronous grace period. In the second mid-boot phase, the scheduler is running, but RCU has not yet gotten its kthreads spawned (and, for expedited grace periods, workqueues are not yet running. During this time, any attempt to do a synchronous grace period will hang the system (or complain bitterly, depending). In the third and final phase, RCU is fully operational and everything works normally. This has been OK for some time, but there has recently been some synchronous grace periods showing up during the second mid-boot phase. This code worked "by accident" for awhile, but started failing as soon as expedited RCU grace periods switched over to workqueues in commit 8b355e3bc140 ("rcu: Drive expedited grace periods from workqueue"). Note that the code was buggy even before this commit, as it was subject to failure on real-time systems that forced all expedited grace periods to run as normal grace periods (for example, using the rcu_normal ksysfs parameter). The callchain from the failure case is as follows: early_amd_iommu_init() |-> acpi_put_table(ivrs_base); |-> acpi_tb_put_table(table_desc); |-> acpi_tb_invalidate_table(table_desc); |-> acpi_tb_release_table(...) |-> acpi_os_unmap_memory |-> acpi_os_unmap_iomem |-> acpi_os_map_cleanup |-> synchronize_rcu_expedited The kernel showing this callchain was built with CONFIG_PREEMPT_RCU=y, which caused the code to try using workqueues before they were initialized, which did not go well. This commit therefore reworks RCU to permit synchronous grace periods to proceed during this mid-boot phase. This commit is therefore a fix to a regression introduced in v4.9, and is therefore being put forward post-merge-window in v4.10. This commit sets a flag from the existing rcu_scheduler_starting() function which causes all synchronous grace periods to take the expedited path. The expedited path now checks this flag, using the requesting task to drive the expedited grace period forward during the mid-boot phase. Finally, this flag is updated by a core_initcall() function named rcu_exp_runtime_mode(), which causes the runtime codepaths to be used. Note that this arrangement assumes that tasks are not sent POSIX signals (or anything similar) from the time that the first task is spawned through core_initcall() time. Fixes: 8b355e3bc140 ("rcu: Drive expedited grace periods from workqueue") Reported-by: "Zheng, Lv" <lv.zheng@intel.com> Reported-by: Borislav Petkov <bp@alien8.de> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Stan Kain <stan.kain@gmail.com> Tested-by: Ivan <waffolz@hotmail.com> Tested-by: Emanuel Castelo <emanuel.castelo@gmail.com> Tested-by: Bruno Pesavento <bpesavento@infinito.it> Tested-by: Borislav Petkov <bp@suse.de> Tested-by: Frederic Bezies <fredbezies@gmail.com> Cc: <stable@vger.kernel.org> # 4.9.0-
2017-01-10 18:28:26 +08:00
if (rcu_scheduler_active == RCU_SCHEDULER_INACTIVE)
return;
if (rcu_gp_is_expedited())
synchronize_rcu_expedited();
else
wait_rcu_gp(call_rcu);
}
EXPORT_SYMBOL_GPL(synchronize_rcu);
/**
* rcu_barrier - Wait until all in-flight call_rcu() callbacks complete.
*
* Note that this primitive does not necessarily wait for an RCU grace period
* to complete. For example, if there are no RCU callbacks queued anywhere
* in the system, then rcu_barrier() is within its rights to return
* immediately, without waiting for anything, much less an RCU grace period.
*/
void rcu_barrier(void)
{
_rcu_barrier(rcu_state_p);
}
EXPORT_SYMBOL_GPL(rcu_barrier);
/*
* Initialize preemptible RCU's state structures.
*/
static void __init __rcu_init_preempt(void)
{
rcu_init_one(rcu_state_p);
}
/*
* Check for a task exiting while in a preemptible-RCU read-side
* critical section, clean up if so. No need to issue warnings,
* as debug_check_no_locks_held() already does this if lockdep
* is enabled.
*/
void exit_rcu(void)
{
struct task_struct *t = current;
if (likely(list_empty(&current->rcu_node_entry)))
return;
t->rcu_read_lock_nesting = 1;
barrier();
t->rcu_read_unlock_special.b.blocked = true;
__rcu_read_unlock();
}
#else /* #ifdef CONFIG_PREEMPT_RCU */
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
static struct rcu_state *const rcu_state_p = &rcu_sched_state;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Tell them what RCU they are running.
*/
static void __init rcu_bootup_announce(void)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
pr_info("Hierarchical RCU implementation.\n");
rcu_bootup_announce_oddness();
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Because preemptible RCU does not exist, we never have to check for
* CPUs being in quiescent states.
*/
static void rcu_preempt_note_context_switch(bool preempt)
{
}
/*
* Because preemptible RCU does not exist, there are never any preempted
* RCU readers.
*/
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp)
{
return 0;
}
/*
* Because there is no preemptible RCU, there can be no readers blocked.
*/
static bool rcu_preempt_has_tasks(struct rcu_node *rnp)
rcu: Fix grace-period-stall bug on large systems with CPU hotplug 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: 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: <12589088301770-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-23 00:53:48 +08:00
{
return false;
rcu: Fix grace-period-stall bug on large systems with CPU hotplug 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: 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: <12589088301770-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-11-23 00:53:48 +08:00
}
/*
* Because preemptible RCU does not exist, we never have to check for
* tasks blocked within RCU read-side critical sections.
*/
static void rcu_print_detail_task_stall(struct rcu_state *rsp)
{
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Because preemptible RCU does not exist, we never have to check for
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
* tasks blocked within RCU read-side critical sections.
*/
static int rcu_print_task_stall(struct rcu_node *rnp)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
return 0;
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
}
/*
* Because preemptible RCU does not exist, we never have to check for
* tasks blocked within RCU read-side critical sections that are
* blocking the current expedited grace period.
*/
static int rcu_print_task_exp_stall(struct rcu_node *rnp)
{
return 0;
}
/*
* Because there is no preemptible RCU, there can be no readers blocked,
* so there is no need to check for blocked tasks. So check only for
* bogus qsmask values.
*/
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
{
WARN_ON_ONCE(rnp->qsmask);
}
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
/*
* Because preemptible RCU does not exist, it never has any callbacks
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
* to check.
*/
static void rcu_preempt_check_callbacks(void)
rcu: Merge preemptable-RCU functionality into hierarchical RCU Create a kernel/rcutree_plugin.h file that contains definitions for preemptable RCU (or, under the #else branch of the #ifdef, empty definitions for the classic non-preemptable semantics). These definitions fit into plugins defined in kernel/rcutree.c for this purpose. This variant of preemptable RCU uses a new algorithm whose read-side expense is roughly that of classic hierarchical RCU under CONFIG_PREEMPT. This new algorithm's update-side expense is similar to that of classic hierarchical RCU, and, in absence of read-side preemption or blocking, is exactly that of classic hierarchical RCU. Perhaps more important, this new algorithm has a much simpler implementation, saving well over 1,000 lines of code compared to mainline's implementation of preemptable RCU, which will hopefully be retired in favor of this new algorithm. The simplifications are obtained by maintaining per-task nesting state for running tasks, and using a simple lock-protected algorithm to handle accounting when tasks block within RCU read-side critical sections, making use of lessons learned while creating numerous user-level RCU implementations over the past 18 months. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: laijs@cn.fujitsu.com Cc: dipankar@in.ibm.com Cc: akpm@linux-foundation.org Cc: mathieu.desnoyers@polymtl.ca Cc: josht@linux.vnet.ibm.com Cc: dvhltc@us.ibm.com Cc: niv@us.ibm.com Cc: peterz@infradead.org Cc: rostedt@goodmis.org LKML-Reference: <12509746134003-git-send-email-> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-08-23 04:56:52 +08:00
{
}
/*
* Because preemptible RCU does not exist, rcu_barrier() is just
* another name for rcu_barrier_sched().
*/
void rcu_barrier(void)
{
rcu_barrier_sched();
}
EXPORT_SYMBOL_GPL(rcu_barrier);
/*
* Because preemptible RCU does not exist, it need not be initialized.
*/
static void __init __rcu_init_preempt(void)
{
}
/*
* Because preemptible RCU does not exist, tasks cannot possibly exit
* while in preemptible RCU read-side critical sections.
*/
void exit_rcu(void)
{
}
#endif /* #else #ifdef CONFIG_PREEMPT_RCU */
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
#ifdef CONFIG_RCU_BOOST
#include "../locking/rtmutex_common.h"
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
static void rcu_wake_cond(struct task_struct *t, int status)
{
/*
* If the thread is yielding, only wake it when this
* is invoked from idle
*/
if (status != RCU_KTHREAD_YIELDING || is_idle_task(current))
wake_up_process(t);
}
/*
* Carry out RCU priority boosting on the task indicated by ->exp_tasks
* or ->boost_tasks, advancing the pointer to the next task in the
* ->blkd_tasks list.
*
* Note that irqs must be enabled: boosting the task can block.
* Returns 1 if there are more tasks needing to be boosted.
*/
static int rcu_boost(struct rcu_node *rnp)
{
unsigned long flags;
struct task_struct *t;
struct list_head *tb;
if (READ_ONCE(rnp->exp_tasks) == NULL &&
READ_ONCE(rnp->boost_tasks) == NULL)
return 0; /* Nothing left to boost. */
raw_spin_lock_irqsave_rcu_node(rnp, flags);
/*
* Recheck under the lock: all tasks in need of boosting
* might exit their RCU read-side critical sections on their own.
*/
if (rnp->exp_tasks == NULL && rnp->boost_tasks == NULL) {
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
return 0;
}
/*
* Preferentially boost tasks blocking expedited grace periods.
* This cannot starve the normal grace periods because a second
* expedited grace period must boost all blocked tasks, including
* those blocking the pre-existing normal grace period.
*/
if (rnp->exp_tasks != NULL) {
tb = rnp->exp_tasks;
rnp->n_exp_boosts++;
} else {
tb = rnp->boost_tasks;
rnp->n_normal_boosts++;
}
rnp->n_tasks_boosted++;
/*
* We boost task t by manufacturing an rt_mutex that appears to
* be held by task t. We leave a pointer to that rt_mutex where
* task t can find it, and task t will release the mutex when it
* exits its outermost RCU read-side critical section. Then
* simply acquiring this artificial rt_mutex will boost task
* t's priority. (Thanks to tglx for suggesting this approach!)
*
* Note that task t must acquire rnp->lock to remove itself from
* the ->blkd_tasks list, which it will do from exit() if from
* nowhere else. We therefore are guaranteed that task t will
* stay around at least until we drop rnp->lock. Note that
* rnp->lock also resolves races between our priority boosting
* and task t's exiting its outermost RCU read-side critical
* section.
*/
t = container_of(tb, struct task_struct, rcu_node_entry);
rt_mutex_init_proxy_locked(&rnp->boost_mtx, t);
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
/* Lock only for side effect: boosts task t's priority. */
rt_mutex_lock(&rnp->boost_mtx);
rt_mutex_unlock(&rnp->boost_mtx); /* Then keep lockdep happy. */
return READ_ONCE(rnp->exp_tasks) != NULL ||
READ_ONCE(rnp->boost_tasks) != NULL;
}
/*
* Priority-boosting kthread, one per leaf rcu_node.
*/
static int rcu_boost_kthread(void *arg)
{
struct rcu_node *rnp = (struct rcu_node *)arg;
int spincnt = 0;
int more2boost;
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("Start boost kthread@init"));
for (;;) {
rnp->boost_kthread_status = RCU_KTHREAD_WAITING;
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("End boost kthread@rcu_wait"));
rcu_wait(rnp->boost_tasks || rnp->exp_tasks);
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("Start boost kthread@rcu_wait"));
rnp->boost_kthread_status = RCU_KTHREAD_RUNNING;
more2boost = rcu_boost(rnp);
if (more2boost)
spincnt++;
else
spincnt = 0;
if (spincnt > 10) {
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
rnp->boost_kthread_status = RCU_KTHREAD_YIELDING;
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("End boost kthread@rcu_yield"));
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
schedule_timeout_interruptible(2);
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("Start boost kthread@rcu_yield"));
spincnt = 0;
}
}
/* NOTREACHED */
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("End boost kthread@notreached"));
return 0;
}
/*
* Check to see if it is time to start boosting RCU readers that are
* blocking the current grace period, and, if so, tell the per-rcu_node
* kthread to start boosting them. If there is an expedited grace
* period in progress, it is always time to boost.
*
* The caller must hold rnp->lock, which this function releases.
* The ->boost_kthread_task is immortal, so we don't need to worry
* about it going away.
*/
static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
__releases(rnp->lock)
{
struct task_struct *t;
lockdep_assert_held(&rnp->lock);
if (!rcu_preempt_blocked_readers_cgp(rnp) && rnp->exp_tasks == NULL) {
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
return;
}
if (rnp->exp_tasks != NULL ||
(rnp->gp_tasks != NULL &&
rnp->boost_tasks == NULL &&
rnp->qsmask == 0 &&
ULONG_CMP_GE(jiffies, rnp->boost_time))) {
if (rnp->exp_tasks == NULL)
rnp->boost_tasks = rnp->gp_tasks;
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
t = rnp->boost_kthread_task;
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
if (t)
rcu_wake_cond(t, rnp->boost_kthread_status);
} else {
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
}
}
/*
* Wake up the per-CPU kthread to invoke RCU callbacks.
*/
static void invoke_rcu_callbacks_kthread(void)
{
unsigned long flags;
local_irq_save(flags);
__this_cpu_write(rcu_cpu_has_work, 1);
if (__this_cpu_read(rcu_cpu_kthread_task) != NULL &&
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
current != __this_cpu_read(rcu_cpu_kthread_task)) {
rcu_wake_cond(__this_cpu_read(rcu_cpu_kthread_task),
__this_cpu_read(rcu_cpu_kthread_status));
}
local_irq_restore(flags);
}
/*
* Is the current CPU running the RCU-callbacks kthread?
* Caller must have preemption disabled.
*/
static bool rcu_is_callbacks_kthread(void)
{
rcu: Replace __get_cpu_var() uses __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Other use cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : __get_cpu_var() always only does an address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&x), y, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Signed-off-by: Christoph Lameter <cl@linux.com> [ paulmck: Address conflicts. ] Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2013-09-01 04:34:10 +08:00
return __this_cpu_read(rcu_cpu_kthread_task) == current;
}
#define RCU_BOOST_DELAY_JIFFIES DIV_ROUND_UP(CONFIG_RCU_BOOST_DELAY * HZ, 1000)
/*
* Do priority-boost accounting for the start of a new grace period.
*/
static void rcu_preempt_boost_start_gp(struct rcu_node *rnp)
{
rnp->boost_time = jiffies + RCU_BOOST_DELAY_JIFFIES;
}
/*
* Create an RCU-boost kthread for the specified node if one does not
* already exist. We only create this kthread for preemptible RCU.
* Returns zero if all is well, a negated errno otherwise.
*/
static int rcu_spawn_one_boost_kthread(struct rcu_state *rsp,
rcu: Process offlining and onlining only at grace-period start Races between CPU hotplug and grace periods can be difficult to resolve, so the ->onoff_mutex is used to exclude the two events. Unfortunately, this means that it is impossible for an outgoing CPU to perform the last bits of its offlining from its last pass through the idle loop, because sleeplocks cannot be acquired in that context. This commit avoids these problems by buffering online and offline events in a new ->qsmaskinitnext field in the leaf rcu_node structures. When a grace period starts, the events accumulated in this mask are applied to the ->qsmaskinit field, and, if needed, up the rcu_node tree. The special case of all CPUs corresponding to a given leaf rcu_node structure being offline while there are still elements in that structure's ->blkd_tasks list is handled using a new ->wait_blkd_tasks field. In this case, propagating the offline bits up the tree is deferred until the beginning of the grace period after all of the tasks have exited their RCU read-side critical sections and removed themselves from the list, at which point the ->wait_blkd_tasks flag is cleared. If one of that leaf rcu_node structure's CPUs comes back online before the list empties, then the ->wait_blkd_tasks flag is simply cleared. This of course means that RCU's notion of which CPUs are offline can be out of date. This is OK because RCU need only wait on CPUs that were online at the time that the grace period started. In addition, RCU's force-quiescent-state actions will handle the case where a CPU goes offline after the grace period starts. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2015-01-24 13:52:37 +08:00
struct rcu_node *rnp)
{
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
int rnp_index = rnp - &rsp->node[0];
unsigned long flags;
struct sched_param sp;
struct task_struct *t;
if (rcu_state_p != rsp)
return 0;
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
rcu: Process offlining and onlining only at grace-period start Races between CPU hotplug and grace periods can be difficult to resolve, so the ->onoff_mutex is used to exclude the two events. Unfortunately, this means that it is impossible for an outgoing CPU to perform the last bits of its offlining from its last pass through the idle loop, because sleeplocks cannot be acquired in that context. This commit avoids these problems by buffering online and offline events in a new ->qsmaskinitnext field in the leaf rcu_node structures. When a grace period starts, the events accumulated in this mask are applied to the ->qsmaskinit field, and, if needed, up the rcu_node tree. The special case of all CPUs corresponding to a given leaf rcu_node structure being offline while there are still elements in that structure's ->blkd_tasks list is handled using a new ->wait_blkd_tasks field. In this case, propagating the offline bits up the tree is deferred until the beginning of the grace period after all of the tasks have exited their RCU read-side critical sections and removed themselves from the list, at which point the ->wait_blkd_tasks flag is cleared. If one of that leaf rcu_node structure's CPUs comes back online before the list empties, then the ->wait_blkd_tasks flag is simply cleared. This of course means that RCU's notion of which CPUs are offline can be out of date. This is OK because RCU need only wait on CPUs that were online at the time that the grace period started. In addition, RCU's force-quiescent-state actions will handle the case where a CPU goes offline after the grace period starts. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2015-01-24 13:52:37 +08:00
if (!rcu_scheduler_fully_active || rcu_rnp_online_cpus(rnp) == 0)
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
return 0;
rsp->boost = 1;
if (rnp->boost_kthread_task != NULL)
return 0;
t = kthread_create(rcu_boost_kthread, (void *)rnp,
"rcub/%d", rnp_index);
if (IS_ERR(t))
return PTR_ERR(t);
raw_spin_lock_irqsave_rcu_node(rnp, flags);
rnp->boost_kthread_task = t;
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
sp.sched_priority = kthread_prio;
sched_setscheduler_nocheck(t, SCHED_FIFO, &sp);
wake_up_process(t); /* get to TASK_INTERRUPTIBLE quickly. */
return 0;
}
static void rcu_kthread_do_work(void)
{
rcu: Replace __get_cpu_var() uses __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Other use cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : __get_cpu_var() always only does an address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&x), y, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Signed-off-by: Christoph Lameter <cl@linux.com> [ paulmck: Address conflicts. ] Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2013-09-01 04:34:10 +08:00
rcu_do_batch(&rcu_sched_state, this_cpu_ptr(&rcu_sched_data));
rcu_do_batch(&rcu_bh_state, this_cpu_ptr(&rcu_bh_data));
rcu_preempt_do_callbacks();
}
static void rcu_cpu_kthread_setup(unsigned int cpu)
{
struct sched_param sp;
sp.sched_priority = kthread_prio;
sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);
}
static void rcu_cpu_kthread_park(unsigned int cpu)
{
per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU;
}
static int rcu_cpu_kthread_should_run(unsigned int cpu)
{
rcu: Replace __get_cpu_var() uses __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Other use cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : __get_cpu_var() always only does an address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&x), y, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Signed-off-by: Christoph Lameter <cl@linux.com> [ paulmck: Address conflicts. ] Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2013-09-01 04:34:10 +08:00
return __this_cpu_read(rcu_cpu_has_work);
}
/*
* Per-CPU kernel thread that invokes RCU callbacks. This replaces the
* RCU softirq used in flavors and configurations of RCU that do not
* support RCU priority boosting.
*/
static void rcu_cpu_kthread(unsigned int cpu)
{
rcu: Replace __get_cpu_var() uses __get_cpu_var() is used for multiple purposes in the kernel source. One of them is address calculation via the form &__get_cpu_var(x). This calculates the address for the instance of the percpu variable of the current processor based on an offset. Other use cases are for storing and retrieving data from the current processors percpu area. __get_cpu_var() can be used as an lvalue when writing data or on the right side of an assignment. __get_cpu_var() is defined as : __get_cpu_var() always only does an address determination. However, store and retrieve operations could use a segment prefix (or global register on other platforms) to avoid the address calculation. this_cpu_write() and this_cpu_read() can directly take an offset into a percpu area and use optimized assembly code to read and write per cpu variables. This patch converts __get_cpu_var into either an explicit address calculation using this_cpu_ptr() or into a use of this_cpu operations that use the offset. Thereby address calcualtions are avoided and less registers are used when code is generated. At the end of the patchset all uses of __get_cpu_var have been removed so the macro is removed too. The patchset includes passes over all arches as well. Once these operations are used throughout then specialized macros can be defined in non -x86 arches as well in order to optimize per cpu access by f.e. using a global register that may be set to the per cpu base. Transformations done to __get_cpu_var() 1. Determine the address of the percpu instance of the current processor. DEFINE_PER_CPU(int, y); int *x = &__get_cpu_var(y); Converts to int *x = this_cpu_ptr(&y); 2. Same as #1 but this time an array structure is involved. DEFINE_PER_CPU(int, y[20]); int *x = __get_cpu_var(y); Converts to int *x = this_cpu_ptr(y); 3. Retrieve the content of the current processors instance of a per cpu variable. DEFINE_PER_CPU(int, u); int x = __get_cpu_var(y) Converts to int x = __this_cpu_read(y); 4. Retrieve the content of a percpu struct DEFINE_PER_CPU(struct mystruct, y); struct mystruct x = __get_cpu_var(y); Converts to memcpy(this_cpu_ptr(&x), y, sizeof(x)); 5. Assignment to a per cpu variable DEFINE_PER_CPU(int, y) __get_cpu_var(y) = x; Converts to this_cpu_write(y, x); 6. Increment/Decrement etc of a per cpu variable DEFINE_PER_CPU(int, y); __get_cpu_var(y)++ Converts to this_cpu_inc(y) Signed-off-by: Christoph Lameter <cl@linux.com> [ paulmck: Address conflicts. ] Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2013-09-01 04:34:10 +08:00
unsigned int *statusp = this_cpu_ptr(&rcu_cpu_kthread_status);
char work, *workp = this_cpu_ptr(&rcu_cpu_has_work);
int spincnt;
for (spincnt = 0; spincnt < 10; spincnt++) {
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("Start CPU kthread@rcu_wait"));
local_bh_disable();
*statusp = RCU_KTHREAD_RUNNING;
this_cpu_inc(rcu_cpu_kthread_loops);
local_irq_disable();
work = *workp;
*workp = 0;
local_irq_enable();
if (work)
rcu_kthread_do_work();
local_bh_enable();
if (*workp == 0) {
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("End CPU kthread@rcu_wait"));
*statusp = RCU_KTHREAD_WAITING;
return;
}
}
*statusp = RCU_KTHREAD_YIELDING;
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("Start CPU kthread@rcu_yield"));
schedule_timeout_interruptible(2);
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_utilization(TPS("End CPU kthread@rcu_yield"));
*statusp = RCU_KTHREAD_WAITING;
}
/*
* Set the per-rcu_node kthread's affinity to cover all CPUs that are
* served by the rcu_node in question. The CPU hotplug lock is still
* held, so the value of rnp->qsmaskinit will be stable.
*
* We don't include outgoingcpu in the affinity set, use -1 if there is
* no outgoing CPU. If there are no CPUs left in the affinity set,
* this function allows the kthread to execute on any CPU.
*/
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu)
{
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
struct task_struct *t = rnp->boost_kthread_task;
rcu: Process offlining and onlining only at grace-period start Races between CPU hotplug and grace periods can be difficult to resolve, so the ->onoff_mutex is used to exclude the two events. Unfortunately, this means that it is impossible for an outgoing CPU to perform the last bits of its offlining from its last pass through the idle loop, because sleeplocks cannot be acquired in that context. This commit avoids these problems by buffering online and offline events in a new ->qsmaskinitnext field in the leaf rcu_node structures. When a grace period starts, the events accumulated in this mask are applied to the ->qsmaskinit field, and, if needed, up the rcu_node tree. The special case of all CPUs corresponding to a given leaf rcu_node structure being offline while there are still elements in that structure's ->blkd_tasks list is handled using a new ->wait_blkd_tasks field. In this case, propagating the offline bits up the tree is deferred until the beginning of the grace period after all of the tasks have exited their RCU read-side critical sections and removed themselves from the list, at which point the ->wait_blkd_tasks flag is cleared. If one of that leaf rcu_node structure's CPUs comes back online before the list empties, then the ->wait_blkd_tasks flag is simply cleared. This of course means that RCU's notion of which CPUs are offline can be out of date. This is OK because RCU need only wait on CPUs that were online at the time that the grace period started. In addition, RCU's force-quiescent-state actions will handle the case where a CPU goes offline after the grace period starts. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2015-01-24 13:52:37 +08:00
unsigned long mask = rcu_rnp_online_cpus(rnp);
cpumask_var_t cm;
int cpu;
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
if (!t)
return;
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
if (!zalloc_cpumask_var(&cm, GFP_KERNEL))
return;
rcu: Correctly handle sparse possible cpus In many cases in the RCU tree code, we iterate over the set of cpus for a leaf node described by rcu_node::grplo and rcu_node::grphi, checking per-cpu data for each cpu in this range. However, if the set of possible cpus is sparse, some cpus described in this range are not possible, and thus no per-cpu region will have been allocated (or initialised) for them by the generic percpu code. Erroneous accesses to a per-cpu area for these !possible cpus may fault or may hit other data depending on the addressed generated when the erroneous per cpu offset is applied. In practice, both cases have been observed on arm64 hardware (the former being silent, but detectable with additional patches). To avoid issues resulting from this, we must iterate over the set of *possible* cpus for a given leaf node. This patch add a new helper, for_each_leaf_node_possible_cpu, to enable this. As iteration is often intertwined with rcu_node local bitmask manipulation, a new leaf_node_cpu_bit helper is added to make this simpler and more consistent. The RCU tree code is made to use both of these where appropriate. Without this patch, running reboot at a shell can result in an oops like: [ 3369.075979] Unable to handle kernel paging request at virtual address ffffff8008b21b4c [ 3369.083881] pgd = ffffffc3ecdda000 [ 3369.087270] [ffffff8008b21b4c] *pgd=00000083eca48003, *pud=00000083eca48003, *pmd=0000000000000000 [ 3369.096222] Internal error: Oops: 96000007 [#1] PREEMPT SMP [ 3369.101781] Modules linked in: [ 3369.104825] CPU: 2 PID: 1817 Comm: NetworkManager Tainted: G W 4.6.0+ #3 [ 3369.121239] task: ffffffc0fa13e000 ti: ffffffc3eb940000 task.ti: ffffffc3eb940000 [ 3369.128708] PC is at sync_rcu_exp_select_cpus+0x188/0x510 [ 3369.134094] LR is at sync_rcu_exp_select_cpus+0x104/0x510 [ 3369.139479] pc : [<ffffff80081109a8>] lr : [<ffffff8008110924>] pstate: 200001c5 [ 3369.146860] sp : ffffffc3eb9435a0 [ 3369.150162] x29: ffffffc3eb9435a0 x28: ffffff8008be4f88 [ 3369.155465] x27: ffffff8008b66c80 x26: ffffffc3eceb2600 [ 3369.160767] x25: 0000000000000001 x24: ffffff8008be4f88 [ 3369.166070] x23: ffffff8008b51c3c x22: ffffff8008b66c80 [ 3369.171371] x21: 0000000000000001 x20: ffffff8008b21b40 [ 3369.176673] x19: ffffff8008b66c80 x18: 0000000000000000 [ 3369.181975] x17: 0000007fa951a010 x16: ffffff80086a30f0 [ 3369.187278] x15: 0000007fa9505590 x14: 0000000000000000 [ 3369.192580] x13: ffffff8008b51000 x12: ffffffc3eb940000 [ 3369.197882] x11: 0000000000000006 x10: ffffff8008b51b78 [ 3369.203184] x9 : 0000000000000001 x8 : ffffff8008be4000 [ 3369.208486] x7 : ffffff8008b21b40 x6 : 0000000000001003 [ 3369.213788] x5 : 0000000000000000 x4 : ffffff8008b27280 [ 3369.219090] x3 : ffffff8008b21b4c x2 : 0000000000000001 [ 3369.224406] x1 : 0000000000000001 x0 : 0000000000000140 ... [ 3369.972257] [<ffffff80081109a8>] sync_rcu_exp_select_cpus+0x188/0x510 [ 3369.978685] [<ffffff80081128b4>] synchronize_rcu_expedited+0x64/0xa8 [ 3369.985026] [<ffffff80086b987c>] synchronize_net+0x24/0x30 [ 3369.990499] [<ffffff80086ddb54>] dev_deactivate_many+0x28c/0x298 [ 3369.996493] [<ffffff80086b6bb8>] __dev_close_many+0x60/0xd0 [ 3370.002052] [<ffffff80086b6d48>] __dev_close+0x28/0x40 [ 3370.007178] [<ffffff80086bf62c>] __dev_change_flags+0x8c/0x158 [ 3370.012999] [<ffffff80086bf718>] dev_change_flags+0x20/0x60 [ 3370.018558] [<ffffff80086cf7f0>] do_setlink+0x288/0x918 [ 3370.023771] [<ffffff80086d0798>] rtnl_newlink+0x398/0x6a8 [ 3370.029158] [<ffffff80086cee84>] rtnetlink_rcv_msg+0xe4/0x220 [ 3370.034891] [<ffffff80086e274c>] netlink_rcv_skb+0xc4/0xf8 [ 3370.040364] [<ffffff80086ced8c>] rtnetlink_rcv+0x2c/0x40 [ 3370.045663] [<ffffff80086e1fe8>] netlink_unicast+0x160/0x238 [ 3370.051309] [<ffffff80086e24b8>] netlink_sendmsg+0x2f0/0x358 [ 3370.056956] [<ffffff80086a0070>] sock_sendmsg+0x18/0x30 [ 3370.062168] [<ffffff80086a21cc>] ___sys_sendmsg+0x26c/0x280 [ 3370.067728] [<ffffff80086a30ac>] __sys_sendmsg+0x44/0x88 [ 3370.073027] [<ffffff80086a3100>] SyS_sendmsg+0x10/0x20 [ 3370.078153] [<ffffff8008085e70>] el0_svc_naked+0x24/0x28 Signed-off-by: Mark Rutland <mark.rutland@arm.com> Reported-by: Dennis Chen <dennis.chen@arm.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: Josh Triplett <josh@joshtriplett.org> Cc: Lai Jiangshan <jiangshanlai@gmail.com> Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com> Cc: Steve Capper <steve.capper@arm.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Will Deacon <will.deacon@arm.com> Cc: linux-kernel@vger.kernel.org Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2016-06-03 22:20:04 +08:00
for_each_leaf_node_possible_cpu(rnp, cpu)
if ((mask & leaf_node_cpu_bit(rnp, cpu)) &&
cpu != outgoingcpu)
cpumask_set_cpu(cpu, cm);
if (cpumask_weight(cm) == 0)
cpumask_setall(cm);
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
set_cpus_allowed_ptr(t, cm);
free_cpumask_var(cm);
}
static struct smp_hotplug_thread rcu_cpu_thread_spec = {
.store = &rcu_cpu_kthread_task,
.thread_should_run = rcu_cpu_kthread_should_run,
.thread_fn = rcu_cpu_kthread,
.thread_comm = "rcuc/%u",
.setup = rcu_cpu_kthread_setup,
.park = rcu_cpu_kthread_park,
};
/*
* Spawn boost kthreads -- called as soon as the scheduler is running.
*/
static void __init rcu_spawn_boost_kthreads(void)
{
struct rcu_node *rnp;
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
int cpu;
for_each_possible_cpu(cpu)
per_cpu(rcu_cpu_has_work, cpu) = 0;
BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec));
rcu_for_each_leaf_node(rcu_state_p, rnp)
(void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp);
}
static void rcu_prepare_kthreads(int cpu)
{
struct rcu_data *rdp = per_cpu_ptr(rcu_state_p->rda, cpu);
struct rcu_node *rnp = rdp->mynode;
/* Fire up the incoming CPU's kthread and leaf rcu_node kthread. */
if (rcu_scheduler_fully_active)
(void)rcu_spawn_one_boost_kthread(rcu_state_p, rnp);
}
#else /* #ifdef CONFIG_RCU_BOOST */
static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
__releases(rnp->lock)
{
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
}
static void invoke_rcu_callbacks_kthread(void)
{
WARN_ON_ONCE(1);
}
static bool rcu_is_callbacks_kthread(void)
{
return false;
}
static void rcu_preempt_boost_start_gp(struct rcu_node *rnp)
{
}
rcu: Yield simpler The rcu_yield() code is amazing. It's there to avoid starvation of the system when lots of (boosting) work is to be done. Now looking at the code it's functionality is: Make the thread SCHED_OTHER and very nice, i.e. get it out of the way Arm a timer with 2 ticks schedule() Now if the system goes idle the rcu task returns, regains SCHED_FIFO and plugs on. If the systems stays busy the timer fires and wakes a per node kthread which in turn makes the per cpu thread SCHED_FIFO and brings it back on the cpu. For the boosting thread the "make it FIFO" bit is missing and it just runs some magic boost checks. Now this is a lot of code with extra threads and complexity. It's way simpler to let the tasks when they detect overload schedule away for 2 ticks and defer the normal wakeup as long as they are in yielded state and the cpu is not idle. That solves the same problem and the only difference is that when the cpu goes idle it's not guaranteed that the thread returns right away, but it won't be longer out than two ticks, so no harm is done. If that's an issue than it is way simpler just to wake the task from idle as RCU has callbacks there anyway. Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: Namhyung Kim <namhyung@kernel.org> Reviewed-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/20120716103948.131256723@linutronix.de Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2012-07-16 18:42:35 +08:00
static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu)
{
}
static void __init rcu_spawn_boost_kthreads(void)
{
}
static void rcu_prepare_kthreads(int cpu)
{
}
#endif /* #else #ifdef CONFIG_RCU_BOOST */
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
#if !defined(CONFIG_RCU_FAST_NO_HZ)
/*
* Check to see if any future RCU-related work will need to be done
* by the current CPU, even if none need be done immediately, returning
* 1 if so. This function is part of the RCU implementation; it is -not-
* an exported member of the RCU API.
*
* Because we not have RCU_FAST_NO_HZ, just check whether this CPU needs
* any flavor of RCU.
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
*/
int rcu_needs_cpu(u64 basemono, u64 *nextevt)
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
{
*nextevt = KTIME_MAX;
return rcu_cpu_has_callbacks(NULL);
}
/*
* Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up
* after it.
*/
static void rcu_cleanup_after_idle(void)
{
}
/*
* Do the idle-entry grace-period work, which, because CONFIG_RCU_FAST_NO_HZ=n,
* is nothing.
*/
static void rcu_prepare_for_idle(void)
{
}
/*
* Don't bother keeping a running count of the number of RCU callbacks
* posted because CONFIG_RCU_FAST_NO_HZ=n.
*/
static void rcu_idle_count_callbacks_posted(void)
{
}
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
#else /* #if !defined(CONFIG_RCU_FAST_NO_HZ) */
/*
* This code is invoked when a CPU goes idle, at which point we want
* to have the CPU do everything required for RCU so that it can enter
* the energy-efficient dyntick-idle mode. This is handled by a
* state machine implemented by rcu_prepare_for_idle() below.
*
* The following three proprocessor symbols control this state machine:
*
* RCU_IDLE_GP_DELAY gives the number of jiffies that a CPU is permitted
* to sleep in dyntick-idle mode with RCU callbacks pending. This
* is sized to be roughly one RCU grace period. Those energy-efficiency
* benchmarkers who might otherwise be tempted to set this to a large
* number, be warned: Setting RCU_IDLE_GP_DELAY too high can hang your
* system. And if you are -that- concerned about energy efficiency,
* just power the system down and be done with it!
* RCU_IDLE_LAZY_GP_DELAY gives the number of jiffies that a CPU is
* permitted to sleep in dyntick-idle mode with only lazy RCU
* callbacks pending. Setting this too high can OOM your system.
*
* The values below work well in practice. If future workloads require
* adjustment, they can be converted into kernel config parameters, though
* making the state machine smarter might be a better option.
*/
#define RCU_IDLE_GP_DELAY 4 /* Roughly one grace period. */
#define RCU_IDLE_LAZY_GP_DELAY (6 * HZ) /* Roughly six seconds. */
static int rcu_idle_gp_delay = RCU_IDLE_GP_DELAY;
module_param(rcu_idle_gp_delay, int, 0644);
static int rcu_idle_lazy_gp_delay = RCU_IDLE_LAZY_GP_DELAY;
module_param(rcu_idle_lazy_gp_delay, int, 0644);
/*
* Try to advance callbacks for all flavors of RCU on the current CPU, but
* only if it has been awhile since the last time we did so. Afterwards,
* if there are any callbacks ready for immediate invocation, return true.
*/
static bool __maybe_unused rcu_try_advance_all_cbs(void)
{
bool cbs_ready = false;
struct rcu_data *rdp;
struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
struct rcu_node *rnp;
struct rcu_state *rsp;
/* Exit early if we advanced recently. */
if (jiffies == rdtp->last_advance_all)
return false;
rdtp->last_advance_all = jiffies;
for_each_rcu_flavor(rsp) {
rdp = this_cpu_ptr(rsp->rda);
rnp = rdp->mynode;
/*
* Don't bother checking unless a grace period has
* completed since we last checked and there are
* callbacks not yet ready to invoke.
*/
rcu: Handle gpnum/completed wrap while dyntick idle 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 <paulmck@linux.vnet.ibm.com>
2014-12-09 12:26:55 +08:00
if ((rdp->completed != rnp->completed ||
unlikely(READ_ONCE(rdp->gpwrap))) &&
rcu_segcblist_pend_cbs(&rdp->cblist))
note_gp_changes(rsp, rdp);
if (rcu_segcblist_ready_cbs(&rdp->cblist))
cbs_ready = true;
}
return cbs_ready;
}
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
/*
* Allow the CPU to enter dyntick-idle mode unless it has callbacks ready
* to invoke. If the CPU has callbacks, try to advance them. Tell the
* caller to set the timeout based on whether or not there are non-lazy
* callbacks.
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
*
* The caller must have disabled interrupts.
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
*/
int rcu_needs_cpu(u64 basemono, u64 *nextevt)
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
{
struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
unsigned long dj;
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_needs_cpu() invoked with irqs enabled!!!");
/* Snapshot to detect later posting of non-lazy callback. */
rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted;
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
/* If no callbacks, RCU doesn't need the CPU. */
if (!rcu_cpu_has_callbacks(&rdtp->all_lazy)) {
*nextevt = KTIME_MAX;
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
return 0;
}
/* Attempt to advance callbacks. */
if (rcu_try_advance_all_cbs()) {
/* Some ready to invoke, so initiate later invocation. */
invoke_rcu_core();
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
return 1;
}
rdtp->last_accelerate = jiffies;
/* Request timer delay depending on laziness, and round. */
if (!rdtp->all_lazy) {
dj = round_up(rcu_idle_gp_delay + jiffies,
rcu_idle_gp_delay) - jiffies;
} else {
dj = round_jiffies(rcu_idle_lazy_gp_delay + jiffies) - jiffies;
}
*nextevt = basemono + dj * TICK_NSEC;
rcu: Precompute RCU_FAST_NO_HZ timer offsets When a CPU is entering dyntick-idle mode, tick_nohz_stop_sched_tick() calls rcu_needs_cpu() see if RCU needs that CPU, and, if not, computes the next wakeup time based on the timer wheels. Only later, when actually entering the idle loop, rcu_prepare_for_idle() will be invoked. In some cases, rcu_prepare_for_idle() will post timers to wake the CPU back up. But all for naught: The next wakeup time for the CPU has already been computed, and posting a timer afterwards does not force that wakeup time to be recomputed. This means that rcu_prepare_for_idle()'s have no effect. This is not a problem on a busy system because something else will wake up the CPU soon enough. However, on lightly loaded systems, the CPU might stay asleep for a considerable length of time. If that CPU has a callback that the rest of the system is waiting on, the system might run very slowly or (in theory) even hang. This commit avoids this problem by having rcu_needs_cpu() give tick_nohz_stop_sched_tick() an estimate of when RCU will need the CPU to wake back up, which tick_nohz_stop_sched_tick() takes into account when programming the CPU's wakeup time. An alternative approach is for rcu_prepare_for_idle() to use hrtimers instead of normal timers, but timers are much more efficient than are hrtimers for frequently and repeatedly posting and cancelling a given timer, which is exactly what RCU_FAST_NO_HZ does. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Reported-by: Heiko Carstens <heiko.carstens@de.ibm.com> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com> Tested-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr>
2012-05-11 07:41:44 +08:00
return 0;
}
rcu: Make RCU_FAST_NO_HZ handle timer migration The current RCU_FAST_NO_HZ assumes that timers do not migrate unless a CPU goes offline, in which case it assumes that the CPU will have to come out of dyntick-idle mode (cancelling the timer) in order to go offline. This is important because when RCU_FAST_NO_HZ permits a CPU to enter dyntick-idle mode despite having RCU callbacks pending, it posts a timer on that CPU to force a wakeup on that CPU. This wakeup ensures that the CPU will eventually handle the end of the grace period, including invoking its RCU callbacks. However, Pascal Chapperon's test setup shows that the timer handler rcu_idle_gp_timer_func() really does get invoked in some cases. This is problematic because this can cause the CPU that entered dyntick-idle mode despite still having RCU callbacks pending to remain in dyntick-idle mode indefinitely, which means that its RCU callbacks might never be invoked. This situation can result in grace-period delays or even system hangs, which matches Pascal's observations of slow boot-up and shutdown (https://lkml.org/lkml/2012/4/5/142). See also the bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=806548 This commit therefore causes the "should never be invoked" timer handler rcu_idle_gp_timer_func() to use smp_call_function_single() to wake up the CPU for which the timer was intended, allowing that CPU to invoke its RCU callbacks in a timely manner. Reported-by: Pascal Chapperon <pascal.chapperon@wanadoo.fr> Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2012-05-01 05:16:19 +08:00
/*
* Prepare a CPU for idle from an RCU perspective. The first major task
* is to sense whether nohz mode has been enabled or disabled via sysfs.
* The second major task is to check to see if a non-lazy callback has
* arrived at a CPU that previously had only lazy callbacks. The third
* major task is to accelerate (that is, assign grace-period numbers to)
* any recently arrived callbacks.
*
* The caller must have disabled interrupts.
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
*/
static void rcu_prepare_for_idle(void)
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
{
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
bool needwake;
struct rcu_data *rdp;
struct rcu_dynticks *rdtp = this_cpu_ptr(&rcu_dynticks);
struct rcu_node *rnp;
struct rcu_state *rsp;
int tne;
RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_prepare_for_idle() invoked with irqs enabled!!!");
if (rcu_is_nocb_cpu(smp_processor_id()))
return;
/* Handle nohz enablement switches conservatively. */
tne = READ_ONCE(tick_nohz_active);
if (tne != rdtp->tick_nohz_enabled_snap) {
if (rcu_cpu_has_callbacks(NULL))
invoke_rcu_core(); /* force nohz to see update. */
rdtp->tick_nohz_enabled_snap = tne;
return;
}
if (!tne)
return;
/*
* If a non-lazy callback arrived at a CPU having only lazy
* callbacks, invoke RCU core for the side-effect of recalculating
* idle duration on re-entry to idle.
*/
if (rdtp->all_lazy &&
rdtp->nonlazy_posted != rdtp->nonlazy_posted_snap) {
rdtp->all_lazy = false;
rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted;
invoke_rcu_core();
return;
}
/*
* If we have not yet accelerated this jiffy, accelerate all
* callbacks on this CPU.
*/
if (rdtp->last_accelerate == jiffies)
return;
rdtp->last_accelerate = jiffies;
for_each_rcu_flavor(rsp) {
rdp = this_cpu_ptr(rsp->rda);
if (rcu_segcblist_pend_cbs(&rdp->cblist))
continue;
rnp = rdp->mynode;
raw_spin_lock_rcu_node(rnp); /* irqs already disabled. */
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
needwake = rcu_accelerate_cbs(rsp, rnp, rdp);
raw_spin_unlock_rcu_node(rnp); /* irqs remain disabled. */
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
if (needwake)
rcu_gp_kthread_wake(rsp);
}
}
/*
* Clean up for exit from idle. Attempt to advance callbacks based on
* any grace periods that elapsed while the CPU was idle, and if any
* callbacks are now ready to invoke, initiate invocation.
*/
static void rcu_cleanup_after_idle(void)
{
RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_cleanup_after_idle() invoked with irqs enabled!!!");
if (rcu_is_nocb_cpu(smp_processor_id()))
return;
if (rcu_try_advance_all_cbs())
invoke_rcu_core();
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
}
/*
* Keep a running count of the number of non-lazy callbacks posted
* on this CPU. This running counter (which is never decremented) allows
* rcu_prepare_for_idle() to detect when something out of the idle loop
* posts a callback, even if an equal number of callbacks are invoked.
* Of course, callbacks should only be posted from within a trace event
* designed to be called from idle or from within RCU_NONIDLE().
*/
static void rcu_idle_count_callbacks_posted(void)
{
__this_cpu_add(rcu_dynticks.nonlazy_posted, 1);
}
/*
* Data for flushing lazy RCU callbacks at OOM time.
*/
static atomic_t oom_callback_count;
static DECLARE_WAIT_QUEUE_HEAD(oom_callback_wq);
/*
* RCU OOM callback -- decrement the outstanding count and deliver the
* wake-up if we are the last one.
*/
static void rcu_oom_callback(struct rcu_head *rhp)
{
if (atomic_dec_and_test(&oom_callback_count))
wake_up(&oom_callback_wq);
}
/*
* Post an rcu_oom_notify callback on the current CPU if it has at
* least one lazy callback. This will unnecessarily post callbacks
* to CPUs that already have a non-lazy callback at the end of their
* callback list, but this is an infrequent operation, so accept some
* extra overhead to keep things simple.
*/
static void rcu_oom_notify_cpu(void *unused)
{
struct rcu_state *rsp;
struct rcu_data *rdp;
for_each_rcu_flavor(rsp) {
rdp = raw_cpu_ptr(rsp->rda);
if (rcu_segcblist_n_lazy_cbs(&rdp->cblist)) {
atomic_inc(&oom_callback_count);
rsp->call(&rdp->oom_head, rcu_oom_callback);
}
}
}
/*
* If low on memory, ensure that each CPU has a non-lazy callback.
* This will wake up CPUs that have only lazy callbacks, in turn
* ensuring that they free up the corresponding memory in a timely manner.
* Because an uncertain amount of memory will be freed in some uncertain
* timeframe, we do not claim to have freed anything.
*/
static int rcu_oom_notify(struct notifier_block *self,
unsigned long notused, void *nfreed)
{
int cpu;
/* Wait for callbacks from earlier instance to complete. */
wait_event(oom_callback_wq, atomic_read(&oom_callback_count) == 0);
smp_mb(); /* Ensure callback reuse happens after callback invocation. */
/*
* Prevent premature wakeup: ensure that all increments happen
* before there is a chance of the counter reaching zero.
*/
atomic_set(&oom_callback_count, 1);
for_each_online_cpu(cpu) {
smp_call_function_single(cpu, rcu_oom_notify_cpu, NULL, 1);
cond_resched_rcu_qs();
}
/* Unconditionally decrement: no need to wake ourselves up. */
atomic_dec(&oom_callback_count);
return NOTIFY_OK;
}
static struct notifier_block rcu_oom_nb = {
.notifier_call = rcu_oom_notify
};
static int __init rcu_register_oom_notifier(void)
{
register_oom_notifier(&rcu_oom_nb);
return 0;
}
early_initcall(rcu_register_oom_notifier);
rcu: Accelerate grace period if last non-dynticked CPU Currently, rcu_needs_cpu() simply checks whether the current CPU has an outstanding RCU callback, which means that the last CPU to go into dyntick-idle mode might wait a few ticks for the relevant grace periods to complete. However, if all the other CPUs are in dyntick-idle mode, and if this CPU is in a quiescent state (which it is for RCU-bh and RCU-sched any time that we are considering going into dyntick-idle mode), then the grace period is instantly complete. This patch therefore repeatedly invokes the RCU grace-period machinery in order to force any needed grace periods to complete quickly. It does so a limited number of times in order to prevent starvation by an RCU callback function that might pass itself to call_rcu(). However, if any CPU other than the current one is not in dyntick-idle mode, fall back to simply checking (with fix to bug noted by Lai Jiangshan). Also, take advantage of last grace-period forcing, the opportunity to do so noted by Steve Rostedt. And apply simplified #ifdef condition suggested by Frederic Weisbecker. 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: <1266887105-1528-15-git-send-email-paulmck@linux.vnet.ibm.com> Signed-off-by: Ingo Molnar <mingo@elte.hu>
2010-02-23 09:04:59 +08:00
#endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */
#ifdef CONFIG_RCU_FAST_NO_HZ
static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
{
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
unsigned long nlpd = rdtp->nonlazy_posted - rdtp->nonlazy_posted_snap;
sprintf(cp, "last_accelerate: %04lx/%04lx, nonlazy_posted: %ld, %c%c",
rdtp->last_accelerate & 0xffff, jiffies & 0xffff,
ulong2long(nlpd),
rdtp->all_lazy ? 'L' : '.',
rdtp->tick_nohz_enabled_snap ? '.' : 'D');
}
#else /* #ifdef CONFIG_RCU_FAST_NO_HZ */
static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
{
*cp = '\0';
}
#endif /* #else #ifdef CONFIG_RCU_FAST_NO_HZ */
/* Initiate the stall-info list. */
static void print_cpu_stall_info_begin(void)
{
pr_cont("\n");
}
/*
* Print out diagnostic information for the specified stalled CPU.
*
* If the specified CPU is aware of the current RCU grace period
* (flavor specified by rsp), then print the number of scheduling
* clock interrupts the CPU has taken during the time that it has
* been aware. Otherwise, print the number of RCU grace periods
* that this CPU is ignorant of, for example, "1" if the CPU was
* aware of the previous grace period.
*
* Also print out idle and (if CONFIG_RCU_FAST_NO_HZ) idle-entry info.
*/
static void print_cpu_stall_info(struct rcu_state *rsp, int cpu)
{
char fast_no_hz[72];
struct rcu_data *rdp = per_cpu_ptr(rsp->rda, cpu);
struct rcu_dynticks *rdtp = rdp->dynticks;
char *ticks_title;
unsigned long ticks_value;
if (rsp->gpnum == rdp->gpnum) {
ticks_title = "ticks this GP";
ticks_value = rdp->ticks_this_gp;
} else {
ticks_title = "GPs behind";
ticks_value = rsp->gpnum - rdp->gpnum;
}
print_cpu_stall_fast_no_hz(fast_no_hz, cpu);
pr_err("\t%d-%c%c%c: (%lu %s) idle=%03x/%llx/%d softirq=%u/%u fqs=%ld %s\n",
cpu,
"O."[!!cpu_online(cpu)],
"o."[!!(rdp->grpmask & rdp->mynode->qsmaskinit)],
"N."[!!(rdp->grpmask & rdp->mynode->qsmaskinitnext)],
ticks_value, ticks_title,
rcu_dynticks_snap(rdtp) & 0xfff,
rdtp->dynticks_nesting, rdtp->dynticks_nmi_nesting,
rdp->softirq_snap, kstat_softirqs_cpu(RCU_SOFTIRQ, cpu),
READ_ONCE(rsp->n_force_qs) - rsp->n_force_qs_gpstart,
fast_no_hz);
}
/* Terminate the stall-info list. */
static void print_cpu_stall_info_end(void)
{
pr_err("\t");
}
/* Zero ->ticks_this_gp for all flavors of RCU. */
static void zero_cpu_stall_ticks(struct rcu_data *rdp)
{
rdp->ticks_this_gp = 0;
rdp->softirq_snap = kstat_softirqs_cpu(RCU_SOFTIRQ, smp_processor_id());
}
/* Increment ->ticks_this_gp for all flavors of RCU. */
static void increment_cpu_stall_ticks(void)
{
struct rcu_state *rsp;
for_each_rcu_flavor(rsp)
raw_cpu_inc(rsp->rda->ticks_this_gp);
}
#ifdef CONFIG_RCU_NOCB_CPU
/*
* Offload callback processing from the boot-time-specified set of CPUs
* specified by rcu_nocb_mask. For each CPU in the set, there is a
* kthread created that pulls the callbacks from the corresponding CPU,
* waits for a grace period to elapse, and invokes the callbacks.
* The no-CBs CPUs do a wake_up() on their kthread when they insert
* a callback into any empty list, unless the rcu_nocb_poll boot parameter
* has been specified, in which case each kthread actively polls its
* CPU. (Which isn't so great for energy efficiency, but which does
* reduce RCU's overhead on that CPU.)
*
* This is intended to be used in conjunction with Frederic Weisbecker's
* adaptive-idle work, which would seriously reduce OS jitter on CPUs
* running CPU-bound user-mode computations.
*
* Offloading of callback processing could also in theory be used as
* an energy-efficiency measure because CPUs with no RCU callbacks
* queued are more aggressive about entering dyntick-idle mode.
*/
/* Parse the boot-time rcu_nocb_mask CPU list from the kernel parameters. */
static int __init rcu_nocb_setup(char *str)
{
alloc_bootmem_cpumask_var(&rcu_nocb_mask);
have_rcu_nocb_mask = true;
cpulist_parse(str, rcu_nocb_mask);
return 1;
}
__setup("rcu_nocbs=", rcu_nocb_setup);
rcu: Make rcu_nocb_poll an early_param instead of module_param The as-documented rcu_nocb_poll will fail to enable this feature for two reasons. (1) there is an extra "s" in the documented name which is not in the code, and (2) since it uses module_param, it really is expecting a prefix, akin to "rcutree.fanout_leaf" and the prefix isn't documented. However, there are several reasons why we might not want to simply fix the typo and add the prefix: 1) we'd end up with rcutree.rcu_nocb_poll, and rather probably make a change to rcutree.nocb_poll 2) if we did #1, then the prefix wouldn't be consistent with the rcu_nocbs=<cpumap> parameter (i.e. one with, one without prefix) 3) the use of module_param in a header file is less than desired, since it isn't immediately obvious that it will get processed via rcutree.c and get the prefix from that (although use of module_param_named() could clarify that.) 4) the implied export of /sys/module/rcutree/parameters/rcu_nocb_poll data to userspace via module_param() doesn't really buy us anything, as it is read-only and we can tell if it is enabled already without it, since there is a printk at early boot telling us so. In light of all that, just change it from a module_param() to an early_setup() call, and worry about adding it to /sys later on if we decide to allow a dynamic setting of it. Also change the variable to be tagged as read_mostly, since it will only ever be fiddled with at most, once at boot. Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2012-12-21 05:19:22 +08:00
static int __init parse_rcu_nocb_poll(char *arg)
{
rcu_nocb_poll = true;
rcu: Make rcu_nocb_poll an early_param instead of module_param The as-documented rcu_nocb_poll will fail to enable this feature for two reasons. (1) there is an extra "s" in the documented name which is not in the code, and (2) since it uses module_param, it really is expecting a prefix, akin to "rcutree.fanout_leaf" and the prefix isn't documented. However, there are several reasons why we might not want to simply fix the typo and add the prefix: 1) we'd end up with rcutree.rcu_nocb_poll, and rather probably make a change to rcutree.nocb_poll 2) if we did #1, then the prefix wouldn't be consistent with the rcu_nocbs=<cpumap> parameter (i.e. one with, one without prefix) 3) the use of module_param in a header file is less than desired, since it isn't immediately obvious that it will get processed via rcutree.c and get the prefix from that (although use of module_param_named() could clarify that.) 4) the implied export of /sys/module/rcutree/parameters/rcu_nocb_poll data to userspace via module_param() doesn't really buy us anything, as it is read-only and we can tell if it is enabled already without it, since there is a printk at early boot telling us so. In light of all that, just change it from a module_param() to an early_setup() call, and worry about adding it to /sys later on if we decide to allow a dynamic setting of it. Also change the variable to be tagged as read_mostly, since it will only ever be fiddled with at most, once at boot. Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2012-12-21 05:19:22 +08:00
return 0;
}
early_param("rcu_nocb_poll", parse_rcu_nocb_poll);
/*
* Wake up any no-CBs CPUs' kthreads that were waiting on the just-ended
* grace period.
*/
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq)
{
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
swake_up_all(sq);
}
/*
* Set the root rcu_node structure's ->need_future_gp field
* based on the sum of those of all rcu_node structures. This does
* double-count the root rcu_node structure's requests, but this
* is necessary to handle the possibility of a rcu_nocb_kthread()
* having awakened during the time that the rcu_node structures
* were being updated for the end of the previous grace period.
*/
static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq)
{
rnp->need_future_gp[(rnp->completed + 1) & 0x1] += nrq;
}
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp)
rcu: Do not call rcu_nocb_gp_cleanup() while holding rnp->lock rcu_nocb_gp_cleanup() is called while holding rnp->lock. Currently, this is okay because the wake_up_all() in rcu_nocb_gp_cleanup() will not enable the IRQs. lockdep is happy. By switching over using swait this is not true anymore. swake_up_all() enables the IRQs while processing the waiters. __do_softirq() can now run and will eventually call rcu_process_callbacks() which wants to grap nrp->lock. Let's move the rcu_nocb_gp_cleanup() call outside the lock before we switch over to swait. If we would hold the rnp->lock and use swait, lockdep reports following: ================================= [ INFO: inconsistent lock state ] 4.2.0-rc5-00025-g9a73ba0 #136 Not tainted --------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. rcu_preempt/8 [HC0[0]:SC0[0]:HE1:SE1] takes: (rcu_node_1){+.?...}, at: [<ffffffff811387c7>] rcu_gp_kthread+0xb97/0xeb0 {IN-SOFTIRQ-W} state was registered at: [<ffffffff81109b9f>] __lock_acquire+0xd5f/0x21e0 [<ffffffff8110be0f>] lock_acquire+0xdf/0x2b0 [<ffffffff81841cc9>] _raw_spin_lock_irqsave+0x59/0xa0 [<ffffffff81136991>] rcu_process_callbacks+0x141/0x3c0 [<ffffffff810b1a9d>] __do_softirq+0x14d/0x670 [<ffffffff810b2214>] irq_exit+0x104/0x110 [<ffffffff81844e96>] smp_apic_timer_interrupt+0x46/0x60 [<ffffffff81842e70>] apic_timer_interrupt+0x70/0x80 [<ffffffff810dba66>] rq_attach_root+0xa6/0x100 [<ffffffff810dbc2d>] cpu_attach_domain+0x16d/0x650 [<ffffffff810e4b42>] build_sched_domains+0x942/0xb00 [<ffffffff821777c2>] sched_init_smp+0x509/0x5c1 [<ffffffff821551e3>] kernel_init_freeable+0x172/0x28f [<ffffffff8182cdce>] kernel_init+0xe/0xe0 [<ffffffff8184231f>] ret_from_fork+0x3f/0x70 irq event stamp: 76 hardirqs last enabled at (75): [<ffffffff81841330>] _raw_spin_unlock_irq+0x30/0x60 hardirqs last disabled at (76): [<ffffffff8184116f>] _raw_spin_lock_irq+0x1f/0x90 softirqs last enabled at (0): [<ffffffff810a8df2>] copy_process.part.26+0x602/0x1cf0 softirqs last disabled at (0): [< (null)>] (null) other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(rcu_node_1); <Interrupt> lock(rcu_node_1); *** DEADLOCK *** 1 lock held by rcu_preempt/8: #0: (rcu_node_1){+.?...}, at: [<ffffffff811387c7>] rcu_gp_kthread+0xb97/0xeb0 stack backtrace: CPU: 0 PID: 8 Comm: rcu_preempt Not tainted 4.2.0-rc5-00025-g9a73ba0 #136 Hardware name: Dell Inc. PowerEdge R820/066N7P, BIOS 2.0.20 01/16/2014 0000000000000000 000000006d7e67d8 ffff881fb081fbd8 ffffffff818379e0 0000000000000000 ffff881fb0812a00 ffff881fb081fc38 ffffffff8110813b 0000000000000000 0000000000000001 ffff881f00000001 ffffffff8102fa4f Call Trace: [<ffffffff818379e0>] dump_stack+0x4f/0x7b [<ffffffff8110813b>] print_usage_bug+0x1db/0x1e0 [<ffffffff8102fa4f>] ? save_stack_trace+0x2f/0x50 [<ffffffff811087ad>] mark_lock+0x66d/0x6e0 [<ffffffff81107790>] ? check_usage_forwards+0x150/0x150 [<ffffffff81108898>] mark_held_locks+0x78/0xa0 [<ffffffff81841330>] ? _raw_spin_unlock_irq+0x30/0x60 [<ffffffff81108a28>] trace_hardirqs_on_caller+0x168/0x220 [<ffffffff81108aed>] trace_hardirqs_on+0xd/0x10 [<ffffffff81841330>] _raw_spin_unlock_irq+0x30/0x60 [<ffffffff810fd1c7>] swake_up_all+0xb7/0xe0 [<ffffffff811386e1>] rcu_gp_kthread+0xab1/0xeb0 [<ffffffff811089bf>] ? trace_hardirqs_on_caller+0xff/0x220 [<ffffffff81841341>] ? _raw_spin_unlock_irq+0x41/0x60 [<ffffffff81137c30>] ? rcu_barrier+0x20/0x20 [<ffffffff810d2014>] kthread+0x104/0x120 [<ffffffff81841330>] ? _raw_spin_unlock_irq+0x30/0x60 [<ffffffff810d1f10>] ? kthread_create_on_node+0x260/0x260 [<ffffffff8184231f>] ret_from_fork+0x3f/0x70 [<ffffffff810d1f10>] ? kthread_create_on_node+0x260/0x260 Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-5-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:40 +08:00
{
return &rnp->nocb_gp_wq[rnp->completed & 0x1];
}
static void rcu_init_one_nocb(struct rcu_node *rnp)
{
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
init_swait_queue_head(&rnp->nocb_gp_wq[0]);
init_swait_queue_head(&rnp->nocb_gp_wq[1]);
}
/* Is the specified CPU a no-CBs CPU? */
bool rcu_is_nocb_cpu(int cpu)
{
if (have_rcu_nocb_mask)
return cpumask_test_cpu(cpu, rcu_nocb_mask);
return false;
}
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/*
* Kick the leader kthread for this NOCB group. Caller holds ->nocb_lock
* and this function releases it.
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
*/
static void __wake_nocb_leader(struct rcu_data *rdp, bool force,
unsigned long flags)
__releases(rdp->nocb_lock)
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
{
struct rcu_data *rdp_leader = rdp->nocb_leader;
lockdep_assert_held(&rdp->nocb_lock);
if (!READ_ONCE(rdp_leader->nocb_kthread)) {
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
return;
}
if (rdp_leader->nocb_leader_sleep || force) {
/* Prior smp_mb__after_atomic() orders against prior enqueue. */
WRITE_ONCE(rdp_leader->nocb_leader_sleep, false);
del_timer(&rdp->nocb_timer);
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
smp_mb(); /* ->nocb_leader_sleep before swake_up(). */
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
swake_up(&rdp_leader->nocb_wq);
} else {
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
}
}
/*
* Kick the leader kthread for this NOCB group, but caller has not
* acquired locks.
*/
static void wake_nocb_leader(struct rcu_data *rdp, bool force)
{
unsigned long flags;
raw_spin_lock_irqsave(&rdp->nocb_lock, flags);
__wake_nocb_leader(rdp, force, flags);
}
/*
* Arrange to wake the leader kthread for this NOCB group at some
* future time when it is safe to do so.
*/
static void wake_nocb_leader_defer(struct rcu_data *rdp, int waketype,
const char *reason)
{
unsigned long flags;
raw_spin_lock_irqsave(&rdp->nocb_lock, flags);
if (rdp->nocb_defer_wakeup == RCU_NOCB_WAKE_NOT)
mod_timer(&rdp->nocb_timer, jiffies + 1);
WRITE_ONCE(rdp->nocb_defer_wakeup, waketype);
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, reason);
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
}
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
/*
* Does the specified CPU need an RCU callback for the specified flavor
* of rcu_barrier()?
*/
static bool rcu_nocb_cpu_needs_barrier(struct rcu_state *rsp, int cpu)
{
struct rcu_data *rdp = per_cpu_ptr(rsp->rda, cpu);
unsigned long ret;
#ifdef CONFIG_PROVE_RCU
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
struct rcu_head *rhp;
#endif /* #ifdef CONFIG_PROVE_RCU */
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
/*
* Check count of all no-CBs callbacks awaiting invocation.
* There needs to be a barrier before this function is called,
* but associated with a prior determination that no more
* callbacks would be posted. In the worst case, the first
* barrier in _rcu_barrier() suffices (but the caller cannot
* necessarily rely on this, not a substitute for the caller
* getting the concurrency design right!). There must also be
* a barrier between the following load an posting of a callback
* (if a callback is in fact needed). This is associated with an
* atomic_inc() in the caller.
*/
ret = atomic_long_read(&rdp->nocb_q_count);
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
#ifdef CONFIG_PROVE_RCU
rhp = READ_ONCE(rdp->nocb_head);
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
if (!rhp)
rhp = READ_ONCE(rdp->nocb_gp_head);
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
if (!rhp)
rhp = READ_ONCE(rdp->nocb_follower_head);
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
/* Having no rcuo kthread but CBs after scheduler starts is bad! */
if (!READ_ONCE(rdp->nocb_kthread) && rhp &&
rcu_scheduler_fully_active) {
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
/* RCU callback enqueued before CPU first came online??? */
pr_err("RCU: Never-onlined no-CBs CPU %d has CB %p\n",
cpu, rhp->func);
WARN_ON_ONCE(1);
}
#endif /* #ifdef CONFIG_PROVE_RCU */
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
return !!ret;
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
}
/*
* Enqueue the specified string of rcu_head structures onto the specified
* CPU's no-CBs lists. The CPU is specified by rdp, the head of the
* string by rhp, and the tail of the string by rhtp. The non-lazy/lazy
* counts are supplied by rhcount and rhcount_lazy.
*
* If warranted, also wake up the kthread servicing this CPUs queues.
*/
static void __call_rcu_nocb_enqueue(struct rcu_data *rdp,
struct rcu_head *rhp,
struct rcu_head **rhtp,
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
int rhcount, int rhcount_lazy,
unsigned long flags)
{
int len;
struct rcu_head **old_rhpp;
struct task_struct *t;
/* Enqueue the callback on the nocb list and update counts. */
atomic_long_add(rhcount, &rdp->nocb_q_count);
/* rcu_barrier() relies on ->nocb_q_count add before xchg. */
old_rhpp = xchg(&rdp->nocb_tail, rhtp);
WRITE_ONCE(*old_rhpp, rhp);
atomic_long_add(rhcount_lazy, &rdp->nocb_q_count_lazy);
smp_mb__after_atomic(); /* Store *old_rhpp before _wake test. */
/* If we are not being polled and there is a kthread, awaken it ... */
t = READ_ONCE(rdp->nocb_kthread);
if (rcu_nocb_poll || !t) {
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu,
TPS("WakeNotPoll"));
return;
}
len = atomic_long_read(&rdp->nocb_q_count);
if (old_rhpp == &rdp->nocb_head) {
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
if (!irqs_disabled_flags(flags)) {
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/* ... if queue was empty ... */
wake_nocb_leader(rdp, false);
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu,
TPS("WakeEmpty"));
} else {
wake_nocb_leader_defer(rdp, RCU_NOCB_WAKE,
TPS("WakeEmptyIsDeferred"));
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
}
rdp->qlen_last_fqs_check = 0;
} else if (len > rdp->qlen_last_fqs_check + qhimark) {
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/* ... or if many callbacks queued. */
if (!irqs_disabled_flags(flags)) {
wake_nocb_leader(rdp, true);
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu,
TPS("WakeOvf"));
} else {
wake_nocb_leader_defer(rdp, RCU_NOCB_WAKE,
TPS("WakeOvfIsDeferred"));
}
rdp->qlen_last_fqs_check = LONG_MAX / 2;
} else {
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, TPS("WakeNot"));
}
return;
}
/*
* This is a helper for __call_rcu(), which invokes this when the normal
* callback queue is inoperable. If this is not a no-CBs CPU, this
* function returns failure back to __call_rcu(), which can complain
* appropriately.
*
* Otherwise, this function queues the callback where the corresponding
* "rcuo" kthread can find it.
*/
static bool __call_rcu_nocb(struct rcu_data *rdp, struct rcu_head *rhp,
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
bool lazy, unsigned long flags)
{
if (!rcu_is_nocb_cpu(rdp->cpu))
return false;
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
__call_rcu_nocb_enqueue(rdp, rhp, &rhp->next, 1, lazy, flags);
if (__is_kfree_rcu_offset((unsigned long)rhp->func))
trace_rcu_kfree_callback(rdp->rsp->name, rhp,
(unsigned long)rhp->func,
-atomic_long_read(&rdp->nocb_q_count_lazy),
-atomic_long_read(&rdp->nocb_q_count));
else
trace_rcu_callback(rdp->rsp->name, rhp,
-atomic_long_read(&rdp->nocb_q_count_lazy),
-atomic_long_read(&rdp->nocb_q_count));
/*
* If called from an extended quiescent state with interrupts
* disabled, invoke the RCU core in order to allow the idle-entry
* deferred-wakeup check to function.
*/
if (irqs_disabled_flags(flags) &&
!rcu_is_watching() &&
cpu_online(smp_processor_id()))
invoke_rcu_core();
return true;
}
/*
* Adopt orphaned callbacks on a no-CBs CPU, or return 0 if this is
* not a no-CBs CPU.
*/
static bool __maybe_unused rcu_nocb_adopt_orphan_cbs(struct rcu_data *my_rdp,
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
struct rcu_data *rdp,
unsigned long flags)
{
RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_nocb_adopt_orphan_cbs() invoked with irqs enabled!!!");
if (!rcu_is_nocb_cpu(smp_processor_id()))
return false; /* Not NOCBs CPU, caller must migrate CBs. */
__call_rcu_nocb_enqueue(my_rdp, rcu_segcblist_head(&rdp->cblist),
rcu_segcblist_tail(&rdp->cblist),
rcu_segcblist_n_cbs(&rdp->cblist),
rcu_segcblist_n_lazy_cbs(&rdp->cblist), flags);
rcu_segcblist_init(&rdp->cblist);
rcu_segcblist_disable(&rdp->cblist);
return true;
}
/*
* If necessary, kick off a new grace period, and either way wait
* for a subsequent grace period to complete.
*/
static void rcu_nocb_wait_gp(struct rcu_data *rdp)
{
unsigned long c;
bool d;
unsigned long flags;
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
bool needwake;
struct rcu_node *rnp = rdp->mynode;
raw_spin_lock_irqsave_rcu_node(rnp, flags);
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
needwake = rcu_start_future_gp(rnp, rdp, &c);
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
rcu: Make callers awaken grace-period kthread The rcu_start_gp_advanced() function currently uses irq_work_queue() to defer wakeups of the RCU grace-period kthread. This deferring is necessary to avoid RCU-scheduler deadlocks involving the rcu_node structure's lock, meaning that RCU cannot call any of the scheduler's wake-up functions while holding one of these locks. Unfortunately, the second and subsequent calls to irq_work_queue() are ignored, and the first call will be ignored (aside from queuing the work item) if the scheduler-clock tick is turned off. This is OK for many uses, especially those where irq_work_queue() is called from an interrupt or softirq handler, because in those cases the scheduler-clock-tick state will be re-evaluated, which will turn the scheduler-clock tick back on. On the next tick, any deferred work will then be processed. However, this strategy does not always work for RCU, which can be invoked at process level from idle CPUs. In this case, the tick might never be turned back on, indefinitely defering a grace-period start request. Note that the RCU CPU stall detector cannot see this condition, because there is no RCU grace period in progress. Therefore, we can (and do!) see long tens-of-seconds stalls in grace-period handling. In theory, we could see a full grace-period hang, but rcutorture testing to date has seen only the tens-of-seconds stalls. Event tracing demonstrates that irq_work_queue() is being called repeatedly to no effect during these stalls: The "newreq" event appears repeatedly from a task that is not one of the grace-period kthreads. In theory, irq_work_queue() might be fixed to avoid this sort of issue, but RCU's requirements are unusual and it is quite straightforward to pass wake-up responsibility up through RCU's call chain, so that the wakeup happens when the offending locks are released. This commit therefore makes this change. The rcu_start_gp_advanced(), rcu_start_future_gp(), rcu_accelerate_cbs(), rcu_advance_cbs(), __note_gp_changes(), and rcu_start_gp() functions now return a boolean which indicates when a wake-up is needed. A new rcu_gp_kthread_wake() does the wakeup when it is necessary and safe to do so: No self-wakes, no wake-ups if the ->gp_flags field indicates there is no need (as in someone else did the wake-up before we got around to it), and no wake-ups before the grace-period kthread has been created. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2014-03-12 04:02:16 +08:00
if (needwake)
rcu_gp_kthread_wake(rdp->rsp);
/*
* Wait for the grace period. Do so interruptibly to avoid messing
* up the load average.
*/
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_future_gp(rnp, rdp, c, TPS("StartWait"));
for (;;) {
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
swait_event_interruptible(
rnp->nocb_gp_wq[c & 0x1],
(d = ULONG_CMP_GE(READ_ONCE(rnp->completed), c)));
if (likely(d))
break;
WARN_ON(signal_pending(current));
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_future_gp(rnp, rdp, c, TPS("ResumeWait"));
}
rcu: Have the RCU tracepoints use the tracepoint_string infrastructure Currently, RCU tracepoints save only a pointer to strings in the ring buffer. When displayed via the /sys/kernel/debug/tracing/trace file they are referenced like the printf "%s" that looks at the address in the ring buffer and prints out the string it points too. This requires that the strings are constant and persistent in the kernel. The problem with this is for tools like trace-cmd and perf that read the binary data from the buffers but have no access to the kernel memory to find out what string is represented by the address in the buffer. By using the tracepoint_string infrastructure, the RCU tracepoint strings can be exported such that userspace tools can map the addresses to the strings. # cat /sys/kernel/debug/tracing/printk_formats 0xffffffff81a4a0e8 : "rcu_preempt" 0xffffffff81a4a0f4 : "rcu_bh" 0xffffffff81a4a100 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437a6 : "rcu_sched" 0xffffffff818437a0 : "cpuqs" 0xffffffff818437b0 : "rcu_bh" 0xffffffff818437b7 : "Start context switch" 0xffffffff818437cc : "End context switch" 0xffffffff818437a0 : "cpuqs" [...] Now userspaces tools can display: rcu_utilization: Start context switch rcu_dyntick: Start 1 0 rcu_utilization: End context switch rcu_batch_start: rcu_preempt CBs=0/5 bl=10 rcu_dyntick: End 0 140000000000000 rcu_invoke_callback: rcu_preempt rhp=0xffff880071c0d600 func=proc_i_callback rcu_invoke_callback: rcu_preempt rhp=0xffff880077b5b230 func=__d_free rcu_dyntick: Start 140000000000000 0 rcu_invoke_callback: rcu_preempt rhp=0xffff880077563980 func=file_free_rcu rcu_batch_end: rcu_preempt CBs-invoked=3 idle=>c<>c<>c<>c< rcu_utilization: End RCU core rcu_grace_period: rcu_preempt 9741 start rcu_dyntick: Start 1 0 rcu_dyntick: End 0 140000000000000 rcu_dyntick: Start 140000000000000 0 Instead of: rcu_utilization: ffffffff81843110 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_batch_start: ffffffff81842f1d CBs=0/4 bl=10 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888aac0 func=file_free_rcu rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f95 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88006aeb4600 func=proc_i_callback rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f32 rcu_future_grace_period: ffffffff81842f1d 9939 9939 9940 0 0 3 ffffffff81842f3c rcu_invoke_callback: ffffffff81842f1d rhp=0xffff880071cb9fc0 func=__d_free rcu_grace_period: ffffffff81842f1d 9939 ffffffff81842f80 rcu_invoke_callback: ffffffff81842f1d rhp=0xffff88007888ae80 func=file_free_rcu rcu_batch_end: ffffffff81842f1d CBs-invoked=4 idle=>c<>c<>c<>c< rcu_utilization: ffffffff8184311f Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2013-07-13 05:18:47 +08:00
trace_rcu_future_gp(rnp, rdp, c, TPS("EndWait"));
smp_mb(); /* Ensure that CB invocation happens after GP end. */
}
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/*
* Leaders come here to wait for additional callbacks to show up.
* This function does not return until callbacks appear.
*/
static void nocb_leader_wait(struct rcu_data *my_rdp)
{
bool firsttime = true;
unsigned long flags;
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
bool gotcbs;
struct rcu_data *rdp;
struct rcu_head **tail;
wait_again:
/* Wait for callbacks to appear. */
if (!rcu_nocb_poll) {
trace_rcu_nocb_wake(my_rdp->rsp->name, my_rdp->cpu, TPS("Sleep"));
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
swait_event_interruptible(my_rdp->nocb_wq,
!READ_ONCE(my_rdp->nocb_leader_sleep));
raw_spin_lock_irqsave(&my_rdp->nocb_lock, flags);
my_rdp->nocb_leader_sleep = true;
WRITE_ONCE(my_rdp->nocb_defer_wakeup, RCU_NOCB_WAKE_NOT);
del_timer(&my_rdp->nocb_timer);
raw_spin_unlock_irqrestore(&my_rdp->nocb_lock, flags);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
} else if (firsttime) {
firsttime = false; /* Don't drown trace log with "Poll"! */
trace_rcu_nocb_wake(my_rdp->rsp->name, my_rdp->cpu, TPS("Poll"));
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
}
/*
* Each pass through the following loop checks a follower for CBs.
* We are our own first follower. Any CBs found are moved to
* nocb_gp_head, where they await a grace period.
*/
gotcbs = false;
smp_mb(); /* wakeup and _sleep before ->nocb_head reads. */
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
for (rdp = my_rdp; rdp; rdp = rdp->nocb_next_follower) {
rdp->nocb_gp_head = READ_ONCE(rdp->nocb_head);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (!rdp->nocb_gp_head)
continue; /* No CBs here, try next follower. */
/* Move callbacks to wait-for-GP list, which is empty. */
WRITE_ONCE(rdp->nocb_head, NULL);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
rdp->nocb_gp_tail = xchg(&rdp->nocb_tail, &rdp->nocb_head);
gotcbs = true;
}
/* No callbacks? Sleep a bit if polling, and go retry. */
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (unlikely(!gotcbs)) {
WARN_ON(signal_pending(current));
if (rcu_nocb_poll) {
schedule_timeout_interruptible(1);
} else {
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
trace_rcu_nocb_wake(my_rdp->rsp->name, my_rdp->cpu,
TPS("WokeEmpty"));
}
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
goto wait_again;
}
/* Wait for one grace period. */
rcu_nocb_wait_gp(my_rdp);
/* Each pass through the following loop wakes a follower, if needed. */
for (rdp = my_rdp; rdp; rdp = rdp->nocb_next_follower) {
if (!rcu_nocb_poll &&
READ_ONCE(rdp->nocb_head) &&
READ_ONCE(my_rdp->nocb_leader_sleep)) {
raw_spin_lock_irqsave(&my_rdp->nocb_lock, flags);
my_rdp->nocb_leader_sleep = false;/* No need to sleep.*/
raw_spin_unlock_irqrestore(&my_rdp->nocb_lock, flags);
}
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (!rdp->nocb_gp_head)
continue; /* No CBs, so no need to wake follower. */
/* Append callbacks to follower's "done" list. */
raw_spin_lock_irqsave(&rdp->nocb_lock, flags);
tail = rdp->nocb_follower_tail;
rdp->nocb_follower_tail = rdp->nocb_gp_tail;
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
*tail = rdp->nocb_gp_head;
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (rdp != my_rdp && tail == &rdp->nocb_follower_head) {
/* List was empty, so wake up the follower. */
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
swake_up(&rdp->nocb_wq);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
}
}
/* If we (the leader) don't have CBs, go wait some more. */
if (!my_rdp->nocb_follower_head)
goto wait_again;
}
/*
* Followers come here to wait for additional callbacks to show up.
* This function does not return until callbacks appear.
*/
static void nocb_follower_wait(struct rcu_data *rdp)
{
for (;;) {
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, TPS("FollowerSleep"));
swait_event_interruptible(rdp->nocb_wq,
READ_ONCE(rdp->nocb_follower_head));
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (smp_load_acquire(&rdp->nocb_follower_head)) {
/* ^^^ Ensure CB invocation follows _head test. */
return;
}
WARN_ON(signal_pending(current));
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, TPS("WokeEmpty"));
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
}
}
/*
* Per-rcu_data kthread, but only for no-CBs CPUs. Each kthread invokes
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
* callbacks queued by the corresponding no-CBs CPU, however, there is
* an optional leader-follower relationship so that the grace-period
* kthreads don't have to do quite so many wakeups.
*/
static int rcu_nocb_kthread(void *arg)
{
int c, cl;
unsigned long flags;
struct rcu_head *list;
struct rcu_head *next;
struct rcu_head **tail;
struct rcu_data *rdp = arg;
/* Each pass through this loop invokes one batch of callbacks */
for (;;) {
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/* Wait for callbacks. */
if (rdp->nocb_leader == rdp)
nocb_leader_wait(rdp);
else
nocb_follower_wait(rdp);
/* Pull the ready-to-invoke callbacks onto local list. */
raw_spin_lock_irqsave(&rdp->nocb_lock, flags);
list = rdp->nocb_follower_head;
rdp->nocb_follower_head = NULL;
tail = rdp->nocb_follower_tail;
rdp->nocb_follower_tail = &rdp->nocb_follower_head;
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
BUG_ON(!list);
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, TPS("WokeNonEmpty"));
/* Each pass through the following loop invokes a callback. */
trace_rcu_batch_start(rdp->rsp->name,
atomic_long_read(&rdp->nocb_q_count_lazy),
atomic_long_read(&rdp->nocb_q_count), -1);
c = cl = 0;
while (list) {
next = list->next;
/* Wait for enqueuing to complete, if needed. */
while (next == NULL && &list->next != tail) {
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu,
TPS("WaitQueue"));
schedule_timeout_interruptible(1);
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu,
TPS("WokeQueue"));
next = list->next;
}
debug_rcu_head_unqueue(list);
local_bh_disable();
if (__rcu_reclaim(rdp->rsp->name, list))
cl++;
c++;
local_bh_enable();
rcu: Fix soft lockup for rcu_nocb_kthread Carrying out the following steps results in a softlockup in the RCU callback-offload (rcuo) kthreads: 1. Connect to ixgbevf, and set the speed to 10Gb/s. 2. Use ifconfig to bring the nic up and down repeatedly. [ 317.005148] IPv6: ADDRCONF(NETDEV_CHANGE): eth2: link becomes ready [ 368.106005] BUG: soft lockup - CPU#1 stuck for 22s! [rcuos/1:15] [ 368.106005] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 368.106005] task: ffff88057dd8a220 ti: ffff88057dd9c000 task.ti: ffff88057dd9c000 [ 368.106005] RIP: 0010:[<ffffffff81579e04>] [<ffffffff81579e04>] fib_table_lookup+0x14/0x390 [ 368.106005] RSP: 0018:ffff88061fc83ce8 EFLAGS: 00000286 [ 368.106005] RAX: 0000000000000001 RBX: 00000000020155c0 RCX: 0000000000000001 [ 368.106005] RDX: ffff88061fc83d50 RSI: ffff88061fc83d70 RDI: ffff880036d11a00 [ 368.106005] RBP: ffff88061fc83d08 R08: 0000000000000001 R09: 0000000000000000 [ 368.106005] R10: ffff880036d11a00 R11: ffffffff819e0900 R12: ffff88061fc83c58 [ 368.106005] R13: ffffffff816154dd R14: ffff88061fc83d08 R15: 00000000020155c0 [ 368.106005] FS: 0000000000000000(0000) GS:ffff88061fc80000(0000) knlGS:0000000000000000 [ 368.106005] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 368.106005] CR2: 00007f8c2aee9c40 CR3: 000000057b222000 CR4: 00000000000407e0 [ 368.106005] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 368.106005] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 [ 368.106005] Stack: [ 368.106005] 00000000010000c0 ffff88057b766000 ffff8802e380b000 ffff88057af03e00 [ 368.106005] ffff88061fc83dc0 ffffffff815349a6 ffff88061fc83d40 ffffffff814ee146 [ 368.106005] ffff8802e380af00 00000000e380af00 ffffffff819e0900 020155c0010000c0 [ 368.106005] Call Trace: [ 368.106005] <IRQ> [ 368.106005] [ 368.106005] [<ffffffff815349a6>] ip_route_input_noref+0x516/0xbd0 [ 368.106005] [<ffffffff814ee146>] ? skb_release_data+0xd6/0x110 [ 368.106005] [<ffffffff814ee20a>] ? kfree_skb+0x3a/0xa0 [ 368.106005] [<ffffffff8153698f>] ip_rcv_finish+0x29f/0x350 [ 368.106005] [<ffffffff81537034>] ip_rcv+0x234/0x380 [ 368.106005] [<ffffffff814fd656>] __netif_receive_skb_core+0x676/0x870 [ 368.106005] [<ffffffff814fd868>] __netif_receive_skb+0x18/0x60 [ 368.106005] [<ffffffff814fe4de>] process_backlog+0xae/0x180 [ 368.106005] [<ffffffff814fdcb2>] net_rx_action+0x152/0x240 [ 368.106005] [<ffffffff81077b3f>] __do_softirq+0xef/0x280 [ 368.106005] [<ffffffff8161619c>] call_softirq+0x1c/0x30 [ 368.106005] <EOI> [ 368.106005] [ 368.106005] [<ffffffff81015d95>] do_softirq+0x65/0xa0 [ 368.106005] [<ffffffff81077174>] local_bh_enable+0x94/0xa0 [ 368.106005] [<ffffffff81114922>] rcu_nocb_kthread+0x232/0x370 [ 368.106005] [<ffffffff81098250>] ? wake_up_bit+0x30/0x30 [ 368.106005] [<ffffffff811146f0>] ? rcu_start_gp+0x40/0x40 [ 368.106005] [<ffffffff8109728f>] kthread+0xcf/0xe0 [ 368.106005] [<ffffffff810971c0>] ? kthread_create_on_node+0x140/0x140 [ 368.106005] [<ffffffff816147d8>] ret_from_fork+0x58/0x90 [ 368.106005] [<ffffffff810971c0>] ? kthread_create_on_node+0x140/0x140 ==================================cut here============================== It turns out that the rcuos callback-offload kthread is busy processing a very large quantity of RCU callbacks, and it is not reliquishing the CPU while doing so. This commit therefore adds an cond_resched_rcu_qs() within the loop to allow other tasks to run. Signed-off-by: Ding Tianhong <dingtianhong@huawei.com> [ paulmck: Substituted cond_resched_rcu_qs for cond_resched. ] Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2016-06-15 15:27:36 +08:00
cond_resched_rcu_qs();
list = next;
}
trace_rcu_batch_end(rdp->rsp->name, c, !!list, 0, 0, 1);
smp_mb__before_atomic(); /* _add after CB invocation. */
atomic_long_add(-c, &rdp->nocb_q_count);
atomic_long_add(-cl, &rdp->nocb_q_count_lazy);
rdp->n_nocbs_invoked += c;
}
return 0;
}
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
/* Is a deferred wakeup of rcu_nocb_kthread() required? */
static int rcu_nocb_need_deferred_wakeup(struct rcu_data *rdp)
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
{
return READ_ONCE(rdp->nocb_defer_wakeup);
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
}
/* Do a deferred wakeup of rcu_nocb_kthread(). */
static void do_nocb_deferred_wakeup_common(struct rcu_data *rdp)
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
{
unsigned long flags;
int ndw;
raw_spin_lock_irqsave(&rdp->nocb_lock, flags);
if (!rcu_nocb_need_deferred_wakeup(rdp)) {
raw_spin_unlock_irqrestore(&rdp->nocb_lock, flags);
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
return;
}
ndw = READ_ONCE(rdp->nocb_defer_wakeup);
WRITE_ONCE(rdp->nocb_defer_wakeup, RCU_NOCB_WAKE_NOT);
__wake_nocb_leader(rdp, ndw == RCU_NOCB_WAKE_FORCE, flags);
trace_rcu_nocb_wake(rdp->rsp->name, rdp->cpu, TPS("DeferredWake"));
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
}
/* Do a deferred wakeup of rcu_nocb_kthread() from a timer handler. */
static void do_nocb_deferred_wakeup_timer(unsigned long x)
{
do_nocb_deferred_wakeup_common((struct rcu_data *)x);
}
/*
* Do a deferred wakeup of rcu_nocb_kthread() from fastpath.
* This means we do an inexact common-case check. Note that if
* we miss, ->nocb_timer will eventually clean things up.
*/
static void do_nocb_deferred_wakeup(struct rcu_data *rdp)
{
if (rcu_nocb_need_deferred_wakeup(rdp))
do_nocb_deferred_wakeup_common(rdp);
}
void __init rcu_init_nohz(void)
{
int cpu;
bool need_rcu_nocb_mask = true;
struct rcu_state *rsp;
#if defined(CONFIG_NO_HZ_FULL)
if (tick_nohz_full_running && cpumask_weight(tick_nohz_full_mask))
need_rcu_nocb_mask = true;
#endif /* #if defined(CONFIG_NO_HZ_FULL) */
if (!have_rcu_nocb_mask && need_rcu_nocb_mask) {
if (!zalloc_cpumask_var(&rcu_nocb_mask, GFP_KERNEL)) {
pr_info("rcu_nocb_mask allocation failed, callback offloading disabled.\n");
return;
}
have_rcu_nocb_mask = true;
}
if (!have_rcu_nocb_mask)
return;
#if defined(CONFIG_NO_HZ_FULL)
if (tick_nohz_full_running)
cpumask_or(rcu_nocb_mask, rcu_nocb_mask, tick_nohz_full_mask);
#endif /* #if defined(CONFIG_NO_HZ_FULL) */
if (!cpumask_subset(rcu_nocb_mask, cpu_possible_mask)) {
pr_info("\tNote: kernel parameter 'rcu_nocbs=' contains nonexistent CPUs.\n");
cpumask_and(rcu_nocb_mask, cpu_possible_mask,
rcu_nocb_mask);
}
pr_info("\tOffload RCU callbacks from CPUs: %*pbl.\n",
cpumask_pr_args(rcu_nocb_mask));
if (rcu_nocb_poll)
pr_info("\tPoll for callbacks from no-CBs CPUs.\n");
for_each_rcu_flavor(rsp) {
for_each_cpu(cpu, rcu_nocb_mask)
init_nocb_callback_list(per_cpu_ptr(rsp->rda, cpu));
rcu_organize_nocb_kthreads(rsp);
}
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
}
/* Initialize per-rcu_data variables for no-CBs CPUs. */
static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp)
{
rdp->nocb_tail = &rdp->nocb_head;
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
init_swait_queue_head(&rdp->nocb_wq);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
rdp->nocb_follower_tail = &rdp->nocb_follower_head;
raw_spin_lock_init(&rdp->nocb_lock);
setup_timer(&rdp->nocb_timer, do_nocb_deferred_wakeup_timer,
(unsigned long)rdp);
}
/*
* If the specified CPU is a no-CBs CPU that does not already have its
* rcuo kthread for the specified RCU flavor, spawn it. If the CPUs are
* brought online out of order, this can require re-organizing the
* leader-follower relationships.
*/
static void rcu_spawn_one_nocb_kthread(struct rcu_state *rsp, int cpu)
{
struct rcu_data *rdp;
struct rcu_data *rdp_last;
struct rcu_data *rdp_old_leader;
struct rcu_data *rdp_spawn = per_cpu_ptr(rsp->rda, cpu);
struct task_struct *t;
/*
* If this isn't a no-CBs CPU or if it already has an rcuo kthread,
* then nothing to do.
*/
if (!rcu_is_nocb_cpu(cpu) || rdp_spawn->nocb_kthread)
return;
/* If we didn't spawn the leader first, reorganize! */
rdp_old_leader = rdp_spawn->nocb_leader;
if (rdp_old_leader != rdp_spawn && !rdp_old_leader->nocb_kthread) {
rdp_last = NULL;
rdp = rdp_old_leader;
do {
rdp->nocb_leader = rdp_spawn;
if (rdp_last && rdp != rdp_spawn)
rdp_last->nocb_next_follower = rdp;
if (rdp == rdp_spawn) {
rdp = rdp->nocb_next_follower;
} else {
rdp_last = rdp;
rdp = rdp->nocb_next_follower;
rdp_last->nocb_next_follower = NULL;
}
} while (rdp);
rdp_spawn->nocb_next_follower = rdp_old_leader;
}
/* Spawn the kthread for this CPU and RCU flavor. */
t = kthread_run(rcu_nocb_kthread, rdp_spawn,
"rcuo%c/%d", rsp->abbr, cpu);
BUG_ON(IS_ERR(t));
WRITE_ONCE(rdp_spawn->nocb_kthread, t);
}
/*
* If the specified CPU is a no-CBs CPU that does not already have its
* rcuo kthreads, spawn them.
*/
static void rcu_spawn_all_nocb_kthreads(int cpu)
{
struct rcu_state *rsp;
if (rcu_scheduler_fully_active)
for_each_rcu_flavor(rsp)
rcu_spawn_one_nocb_kthread(rsp, cpu);
}
/*
* Once the scheduler is running, spawn rcuo kthreads for all online
* no-CBs CPUs. This assumes that the early_initcall()s happen before
* non-boot CPUs come online -- if this changes, we will need to add
* some mutual exclusion.
*/
static void __init rcu_spawn_nocb_kthreads(void)
{
int cpu;
for_each_online_cpu(cpu)
rcu_spawn_all_nocb_kthreads(cpu);
}
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
/* How many follower CPU IDs per leader? Default of -1 for sqrt(nr_cpu_ids). */
static int rcu_nocb_leader_stride = -1;
module_param(rcu_nocb_leader_stride, int, 0444);
/*
* Initialize leader-follower relationships for all no-CBs CPU.
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
*/
static void __init rcu_organize_nocb_kthreads(struct rcu_state *rsp)
{
int cpu;
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
int ls = rcu_nocb_leader_stride;
int nl = 0; /* Next leader. */
struct rcu_data *rdp;
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
struct rcu_data *rdp_leader = NULL; /* Suppress misguided gcc warn. */
struct rcu_data *rdp_prev = NULL;
if (!have_rcu_nocb_mask)
return;
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (ls == -1) {
ls = int_sqrt(nr_cpu_ids);
rcu_nocb_leader_stride = ls;
}
/*
* Each pass through this loop sets up one rcu_data structure.
* Should the corresponding CPU come online in the future, then
* we will spawn the needed set of rcu_nocb_kthread() kthreads.
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
*/
for_each_cpu(cpu, rcu_nocb_mask) {
rdp = per_cpu_ptr(rsp->rda, cpu);
rcu: Parallelize and economize NOCB kthread wakeups An 80-CPU system with a context-switch-heavy workload can require so many NOCB kthread wakeups that the RCU grace-period kthreads spend several tens of percent of a CPU just awakening things. This clearly will not scale well: If you add enough CPUs, the RCU grace-period kthreads would get behind, increasing grace-period latency. To avoid this problem, this commit divides the NOCB kthreads into leaders and followers, where the grace-period kthreads awaken the leaders each of whom in turn awakens its followers. By default, the number of groups of kthreads is the square root of the number of CPUs, but this default may be overridden using the rcutree.rcu_nocb_leader_stride boot parameter. This reduces the number of wakeups done per grace period by the RCU grace-period kthread by the square root of the number of CPUs, but of course by shifting those wakeups to the leaders. In addition, because the leaders do grace periods on behalf of their respective followers, the number of wakeups of the followers decreases by up to a factor of two. Instead of being awakened once when new callbacks arrive and again at the end of the grace period, the followers are awakened only at the end of the grace period. For a numerical example, in a 4096-CPU system, the grace-period kthread would awaken 64 leaders, each of which would awaken its 63 followers at the end of the grace period. This compares favorably with the 79 wakeups for the grace-period kthread on an 80-CPU system. Reported-by: Rik van Riel <riel@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-25 00:26:11 +08:00
if (rdp->cpu >= nl) {
/* New leader, set up for followers & next leader. */
nl = DIV_ROUND_UP(rdp->cpu + 1, ls) * ls;
rdp->nocb_leader = rdp;
rdp_leader = rdp;
} else {
/* Another follower, link to previous leader. */
rdp->nocb_leader = rdp_leader;
rdp_prev->nocb_next_follower = rdp;
}
rdp_prev = rdp;
}
}
/* Prevent __call_rcu() from enqueuing callbacks on no-CBs CPUs */
static bool init_nocb_callback_list(struct rcu_data *rdp)
{
if (!rcu_is_nocb_cpu(rdp->cpu))
return false;
/* If there are early-boot callbacks, move them to nocb lists. */
if (!rcu_segcblist_empty(&rdp->cblist)) {
rdp->nocb_head = rcu_segcblist_head(&rdp->cblist);
rdp->nocb_tail = rcu_segcblist_tail(&rdp->cblist);
atomic_long_set(&rdp->nocb_q_count,
rcu_segcblist_n_cbs(&rdp->cblist));
atomic_long_set(&rdp->nocb_q_count_lazy,
rcu_segcblist_n_lazy_cbs(&rdp->cblist));
rcu_segcblist_init(&rdp->cblist);
}
rcu_segcblist_disable(&rdp->cblist);
return true;
}
#else /* #ifdef CONFIG_RCU_NOCB_CPU */
rcu: Make rcu_barrier() understand about missing rcuo kthreads Commit 35ce7f29a44a (rcu: Create rcuo kthreads only for onlined CPUs) avoids creating rcuo kthreads for CPUs that never come online. This fixes a bug in many instances of firmware: Instead of lying about their age, these systems instead lie about the number of CPUs that they have. Before commit 35ce7f29a44a, this could result in huge numbers of useless rcuo kthreads being created. It appears that experience indicates that I should have told the people suffering from this problem to fix their broken firmware, but I instead produced what turned out to be a partial fix. The missing piece supplied by this commit makes sure that rcu_barrier() knows not to post callbacks for no-CBs CPUs that have not yet come online, because otherwise rcu_barrier() will hang on systems having firmware that lies about the number of CPUs. It is tempting to simply have rcu_barrier() refuse to post a callback on any no-CBs CPU that does not have an rcuo kthread. This unfortunately does not work because rcu_barrier() is required to wait for all pending callbacks. It is therefore required to wait even for those callbacks that cannot possibly be invoked. Even if doing so hangs the system. Given that posting a callback to a no-CBs CPU that does not yet have an rcuo kthread can hang rcu_barrier(), It is tempting to report an error in this case. Unfortunately, this will result in false positives at boot time, when it is perfectly legal to post callbacks to the boot CPU before the scheduler has started, in other words, before it is legal to invoke rcu_barrier(). So this commit instead has rcu_barrier() avoid posting callbacks to CPUs having neither rcuo kthread nor pending callbacks, and has it complain bitterly if it finds CPUs having no rcuo kthread but some pending callbacks. And when rcu_barrier() does find CPUs having no rcuo kthread but pending callbacks, as noted earlier, it has no choice but to hang indefinitely. Reported-by: Yanko Kaneti <yaneti@declera.com> Reported-by: Jay Vosburgh <jay.vosburgh@canonical.com> Reported-by: Meelis Roos <mroos@linux.ee> Reported-by: Eric B Munson <emunson@akamai.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Tested-by: Eric B Munson <emunson@akamai.com> Tested-by: Jay Vosburgh <jay.vosburgh@canonical.com> Tested-by: Yanko Kaneti <yaneti@declera.com> Tested-by: Kevin Fenzi <kevin@scrye.com> Tested-by: Meelis Roos <mroos@linux.ee>
2014-10-28 00:15:54 +08:00
static bool rcu_nocb_cpu_needs_barrier(struct rcu_state *rsp, int cpu)
{
WARN_ON_ONCE(1); /* Should be dead code. */
return false;
}
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
static void rcu_nocb_gp_cleanup(struct swait_queue_head *sq)
{
}
static void rcu_nocb_gp_set(struct rcu_node *rnp, int nrq)
{
}
rcu: Use simple wait queues where possible in rcutree As of commit dae6e64d2bcfd ("rcu: Introduce proper blocking to no-CBs kthreads GP waits") the RCU subsystem started making use of wait queues. Here we convert all additions of RCU wait queues to use simple wait queues, since they don't need the extra overhead of the full wait queue features. Originally this was done for RT kernels[1], since we would get things like... BUG: sleeping function called from invalid context at kernel/rtmutex.c:659 in_atomic(): 1, irqs_disabled(): 1, pid: 8, name: rcu_preempt Pid: 8, comm: rcu_preempt Not tainted Call Trace: [<ffffffff8106c8d0>] __might_sleep+0xd0/0xf0 [<ffffffff817d77b4>] rt_spin_lock+0x24/0x50 [<ffffffff8106fcf6>] __wake_up+0x36/0x70 [<ffffffff810c4542>] rcu_gp_kthread+0x4d2/0x680 [<ffffffff8105f910>] ? __init_waitqueue_head+0x50/0x50 [<ffffffff810c4070>] ? rcu_gp_fqs+0x80/0x80 [<ffffffff8105eabb>] kthread+0xdb/0xe0 [<ffffffff8106b912>] ? finish_task_switch+0x52/0x100 [<ffffffff817e0754>] kernel_thread_helper+0x4/0x10 [<ffffffff8105e9e0>] ? __init_kthread_worker+0x60/0x60 [<ffffffff817e0750>] ? gs_change+0xb/0xb ...and hence simple wait queues were deployed on RT out of necessity (as simple wait uses a raw lock), but mainline might as well take advantage of the more streamline support as well. [1] This is a carry forward of work from v3.10-rt; the original conversion was by Thomas on an earlier -rt version, and Sebastian extended it to additional post-3.10 added RCU waiters; here I've added a commit log and unified the RCU changes into one, and uprev'd it to match mainline RCU. Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-6-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:41 +08:00
static struct swait_queue_head *rcu_nocb_gp_get(struct rcu_node *rnp)
rcu: Do not call rcu_nocb_gp_cleanup() while holding rnp->lock rcu_nocb_gp_cleanup() is called while holding rnp->lock. Currently, this is okay because the wake_up_all() in rcu_nocb_gp_cleanup() will not enable the IRQs. lockdep is happy. By switching over using swait this is not true anymore. swake_up_all() enables the IRQs while processing the waiters. __do_softirq() can now run and will eventually call rcu_process_callbacks() which wants to grap nrp->lock. Let's move the rcu_nocb_gp_cleanup() call outside the lock before we switch over to swait. If we would hold the rnp->lock and use swait, lockdep reports following: ================================= [ INFO: inconsistent lock state ] 4.2.0-rc5-00025-g9a73ba0 #136 Not tainted --------------------------------- inconsistent {IN-SOFTIRQ-W} -> {SOFTIRQ-ON-W} usage. rcu_preempt/8 [HC0[0]:SC0[0]:HE1:SE1] takes: (rcu_node_1){+.?...}, at: [<ffffffff811387c7>] rcu_gp_kthread+0xb97/0xeb0 {IN-SOFTIRQ-W} state was registered at: [<ffffffff81109b9f>] __lock_acquire+0xd5f/0x21e0 [<ffffffff8110be0f>] lock_acquire+0xdf/0x2b0 [<ffffffff81841cc9>] _raw_spin_lock_irqsave+0x59/0xa0 [<ffffffff81136991>] rcu_process_callbacks+0x141/0x3c0 [<ffffffff810b1a9d>] __do_softirq+0x14d/0x670 [<ffffffff810b2214>] irq_exit+0x104/0x110 [<ffffffff81844e96>] smp_apic_timer_interrupt+0x46/0x60 [<ffffffff81842e70>] apic_timer_interrupt+0x70/0x80 [<ffffffff810dba66>] rq_attach_root+0xa6/0x100 [<ffffffff810dbc2d>] cpu_attach_domain+0x16d/0x650 [<ffffffff810e4b42>] build_sched_domains+0x942/0xb00 [<ffffffff821777c2>] sched_init_smp+0x509/0x5c1 [<ffffffff821551e3>] kernel_init_freeable+0x172/0x28f [<ffffffff8182cdce>] kernel_init+0xe/0xe0 [<ffffffff8184231f>] ret_from_fork+0x3f/0x70 irq event stamp: 76 hardirqs last enabled at (75): [<ffffffff81841330>] _raw_spin_unlock_irq+0x30/0x60 hardirqs last disabled at (76): [<ffffffff8184116f>] _raw_spin_lock_irq+0x1f/0x90 softirqs last enabled at (0): [<ffffffff810a8df2>] copy_process.part.26+0x602/0x1cf0 softirqs last disabled at (0): [< (null)>] (null) other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(rcu_node_1); <Interrupt> lock(rcu_node_1); *** DEADLOCK *** 1 lock held by rcu_preempt/8: #0: (rcu_node_1){+.?...}, at: [<ffffffff811387c7>] rcu_gp_kthread+0xb97/0xeb0 stack backtrace: CPU: 0 PID: 8 Comm: rcu_preempt Not tainted 4.2.0-rc5-00025-g9a73ba0 #136 Hardware name: Dell Inc. PowerEdge R820/066N7P, BIOS 2.0.20 01/16/2014 0000000000000000 000000006d7e67d8 ffff881fb081fbd8 ffffffff818379e0 0000000000000000 ffff881fb0812a00 ffff881fb081fc38 ffffffff8110813b 0000000000000000 0000000000000001 ffff881f00000001 ffffffff8102fa4f Call Trace: [<ffffffff818379e0>] dump_stack+0x4f/0x7b [<ffffffff8110813b>] print_usage_bug+0x1db/0x1e0 [<ffffffff8102fa4f>] ? save_stack_trace+0x2f/0x50 [<ffffffff811087ad>] mark_lock+0x66d/0x6e0 [<ffffffff81107790>] ? check_usage_forwards+0x150/0x150 [<ffffffff81108898>] mark_held_locks+0x78/0xa0 [<ffffffff81841330>] ? _raw_spin_unlock_irq+0x30/0x60 [<ffffffff81108a28>] trace_hardirqs_on_caller+0x168/0x220 [<ffffffff81108aed>] trace_hardirqs_on+0xd/0x10 [<ffffffff81841330>] _raw_spin_unlock_irq+0x30/0x60 [<ffffffff810fd1c7>] swake_up_all+0xb7/0xe0 [<ffffffff811386e1>] rcu_gp_kthread+0xab1/0xeb0 [<ffffffff811089bf>] ? trace_hardirqs_on_caller+0xff/0x220 [<ffffffff81841341>] ? _raw_spin_unlock_irq+0x41/0x60 [<ffffffff81137c30>] ? rcu_barrier+0x20/0x20 [<ffffffff810d2014>] kthread+0x104/0x120 [<ffffffff81841330>] ? _raw_spin_unlock_irq+0x30/0x60 [<ffffffff810d1f10>] ? kthread_create_on_node+0x260/0x260 [<ffffffff8184231f>] ret_from_fork+0x3f/0x70 [<ffffffff810d1f10>] ? kthread_create_on_node+0x260/0x260 Signed-off-by: Daniel Wagner <daniel.wagner@bmw-carit.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Cc: linux-rt-users@vger.kernel.org Cc: Boqun Feng <boqun.feng@gmail.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> Link: http://lkml.kernel.org/r/1455871601-27484-5-git-send-email-wagi@monom.org Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
2016-02-19 16:46:40 +08:00
{
return NULL;
}
static void rcu_init_one_nocb(struct rcu_node *rnp)
{
}
static bool __call_rcu_nocb(struct rcu_data *rdp, struct rcu_head *rhp,
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
bool lazy, unsigned long flags)
{
return false;
}
static bool __maybe_unused rcu_nocb_adopt_orphan_cbs(struct rcu_data *my_rdp,
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
struct rcu_data *rdp,
unsigned long flags)
{
return false;
}
static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp)
{
}
static int rcu_nocb_need_deferred_wakeup(struct rcu_data *rdp)
rcu: Break call_rcu() deadlock involving scheduler and perf Dave Jones got the following lockdep splat: > ====================================================== > [ INFO: possible circular locking dependency detected ] > 3.12.0-rc3+ #92 Not tainted > ------------------------------------------------------- > trinity-child2/15191 is trying to acquire lock: > (&rdp->nocb_wq){......}, at: [<ffffffff8108ff43>] __wake_up+0x23/0x50 > > but task is already holding lock: > (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > which lock already depends on the new lock. > > > the existing dependency chain (in reverse order) is: > > -> #3 (&ctx->lock){-.-...}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff811500ff>] __perf_event_task_sched_out+0x2df/0x5e0 > [<ffffffff81091b83>] perf_event_task_sched_out+0x93/0xa0 > [<ffffffff81732052>] __schedule+0x1d2/0xa20 > [<ffffffff81732f30>] preempt_schedule_irq+0x50/0xb0 > [<ffffffff817352b6>] retint_kernel+0x26/0x30 > [<ffffffff813eed04>] tty_flip_buffer_push+0x34/0x50 > [<ffffffff813f0504>] pty_write+0x54/0x60 > [<ffffffff813e900d>] n_tty_write+0x32d/0x4e0 > [<ffffffff813e5838>] tty_write+0x158/0x2d0 > [<ffffffff811c4850>] vfs_write+0xc0/0x1f0 > [<ffffffff811c52cc>] SyS_write+0x4c/0xa0 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > -> #2 (&rq->lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff81733f90>] _raw_spin_lock+0x40/0x80 > [<ffffffff810980b2>] wake_up_new_task+0xc2/0x2e0 > [<ffffffff81054336>] do_fork+0x126/0x460 > [<ffffffff81054696>] kernel_thread+0x26/0x30 > [<ffffffff8171ff93>] rest_init+0x23/0x140 > [<ffffffff81ee1e4b>] start_kernel+0x3f6/0x403 > [<ffffffff81ee1571>] x86_64_start_reservations+0x2a/0x2c > [<ffffffff81ee1664>] x86_64_start_kernel+0xf1/0xf4 > > -> #1 (&p->pi_lock){-.-.-.}: > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff810979d1>] try_to_wake_up+0x31/0x350 > [<ffffffff81097d62>] default_wake_function+0x12/0x20 > [<ffffffff81084af8>] autoremove_wake_function+0x18/0x40 > [<ffffffff8108ea38>] __wake_up_common+0x58/0x90 > [<ffffffff8108ff59>] __wake_up+0x39/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111b8d>] call_rcu+0x1d/0x20 > [<ffffffff81093697>] cpu_attach_domain+0x287/0x360 > [<ffffffff81099d7e>] build_sched_domains+0xe5e/0x10a0 > [<ffffffff81efa7fc>] sched_init_smp+0x3b7/0x47a > [<ffffffff81ee1f4e>] kernel_init_freeable+0xf6/0x202 > [<ffffffff817200be>] kernel_init+0xe/0x190 > [<ffffffff8173d22c>] ret_from_fork+0x7c/0xb0 > > -> #0 (&rdp->nocb_wq){......}: > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 > > other info that might help us debug this: > > Chain exists of: > &rdp->nocb_wq --> &rq->lock --> &ctx->lock > > Possible unsafe locking scenario: > > CPU0 CPU1 > ---- ---- > lock(&ctx->lock); > lock(&rq->lock); > lock(&ctx->lock); > lock(&rdp->nocb_wq); > > *** DEADLOCK *** > > 1 lock held by trinity-child2/15191: > #0: (&ctx->lock){-.-...}, at: [<ffffffff81154c19>] perf_event_exit_task+0x109/0x230 > > stack backtrace: > CPU: 2 PID: 15191 Comm: trinity-child2 Not tainted 3.12.0-rc3+ #92 > ffffffff82565b70 ffff880070c2dbf8 ffffffff8172a363 ffffffff824edf40 > ffff880070c2dc38 ffffffff81726741 ffff880070c2dc90 ffff88022383b1c0 > ffff88022383aac0 0000000000000000 ffff88022383b188 ffff88022383b1c0 > Call Trace: > [<ffffffff8172a363>] dump_stack+0x4e/0x82 > [<ffffffff81726741>] print_circular_bug+0x200/0x20f > [<ffffffff810cb7ca>] __lock_acquire+0x191a/0x1be0 > [<ffffffff810c6439>] ? get_lock_stats+0x19/0x60 > [<ffffffff8100b2f4>] ? native_sched_clock+0x24/0x80 > [<ffffffff810cc243>] lock_acquire+0x93/0x200 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8173419b>] _raw_spin_lock_irqsave+0x4b/0x90 > [<ffffffff8108ff43>] ? __wake_up+0x23/0x50 > [<ffffffff8108ff43>] __wake_up+0x23/0x50 > [<ffffffff8110d4f8>] __call_rcu_nocb_enqueue+0xa8/0xc0 > [<ffffffff81111450>] __call_rcu+0x140/0x820 > [<ffffffff8109bc8f>] ? local_clock+0x3f/0x50 > [<ffffffff81111bb0>] kfree_call_rcu+0x20/0x30 > [<ffffffff81149abf>] put_ctx+0x4f/0x70 > [<ffffffff81154c3e>] perf_event_exit_task+0x12e/0x230 > [<ffffffff81056b8d>] do_exit+0x30d/0xcc0 > [<ffffffff810c9af5>] ? trace_hardirqs_on_caller+0x115/0x1e0 > [<ffffffff810c9bcd>] ? trace_hardirqs_on+0xd/0x10 > [<ffffffff8105893c>] do_group_exit+0x4c/0xc0 > [<ffffffff810589c4>] SyS_exit_group+0x14/0x20 > [<ffffffff8173d4e4>] tracesys+0xdd/0xe2 The underlying problem is that perf is invoking call_rcu() with the scheduler locks held, but in NOCB mode, call_rcu() will with high probability invoke the scheduler -- which just might want to use its locks. The reason that call_rcu() needs to invoke the scheduler is to wake up the corresponding rcuo callback-offload kthread, which does the job of starting up a grace period and invoking the callbacks afterwards. One solution (championed on a related problem by Lai Jiangshan) is to simply defer the wakeup to some point where scheduler locks are no longer held. Since we don't want to unnecessarily incur the cost of such deferral, the task before us is threefold: 1. Determine when it is likely that a relevant scheduler lock is held. 2. Defer the wakeup in such cases. 3. Ensure that all deferred wakeups eventually happen, preferably sooner rather than later. We use irqs_disabled_flags() as a proxy for relevant scheduler locks being held. This works because the relevant locks are always acquired with interrupts disabled. We may defer more often than needed, but that is at least safe. The wakeup deferral is tracked via a new field in the per-CPU and per-RCU-flavor rcu_data structure, namely ->nocb_defer_wakeup. This flag is checked by the RCU core processing. The __rcu_pending() function now checks this flag, which causes rcu_check_callbacks() to initiate RCU core processing at each scheduling-clock interrupt where this flag is set. Of course this is not sufficient because scheduling-clock interrupts are often turned off (the things we used to be able to count on!). So the flags are also checked on entry to any state that RCU considers to be idle, which includes both NO_HZ_IDLE idle state and NO_HZ_FULL user-mode-execution state. This approach should allow call_rcu() to be invoked regardless of what locks you might be holding, the key word being "should". Reported-by: Dave Jones <davej@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Peter Zijlstra <peterz@infradead.org>
2013-10-05 05:33:34 +08:00
{
return false;
}
static void do_nocb_deferred_wakeup(struct rcu_data *rdp)
{
}
static void rcu_spawn_all_nocb_kthreads(int cpu)
{
}
static void __init rcu_spawn_nocb_kthreads(void)
{
}
static bool init_nocb_callback_list(struct rcu_data *rdp)
{
return false;
}
#endif /* #else #ifdef CONFIG_RCU_NOCB_CPU */
rcu: Kick adaptive-ticks CPUs that are holding up RCU grace periods Adaptive-ticks CPUs inform RCU when they enter kernel mode, but they do not necessarily turn the scheduler-clock tick back on. This state of affairs could result in RCU waiting on an adaptive-ticks CPU running for an extended period in kernel mode. Such a CPU will never run the RCU state machine, and could therefore indefinitely extend the RCU state machine, sooner or later resulting in an OOM condition. This patch, inspired by an earlier patch by Frederic Weisbecker, therefore causes RCU's force-quiescent-state processing to check for this condition and to send an IPI to CPUs that remain in that state for too long. "Too long" currently means about three jiffies by default, which is quite some time for a CPU to remain in the kernel without blocking. The rcu_tree.jiffies_till_first_fqs and rcutree.jiffies_till_next_fqs sysfs variables may be used to tune "too long" if needed. Reported-by: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Chris Metcalf <cmetcalf@tilera.com> Cc: Christoph Lameter <cl@linux.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Gilad Ben Yossef <gilad@benyossef.com> Cc: Hakan Akkan <hakanakkan@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Kevin Hilman <khilman@linaro.org> Cc: Li Zhong <zhong@linux.vnet.ibm.com> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de>
2013-04-13 07:19:10 +08:00
/*
* An adaptive-ticks CPU can potentially execute in kernel mode for an
* arbitrarily long period of time with the scheduling-clock tick turned
* off. RCU will be paying attention to this CPU because it is in the
* kernel, but the CPU cannot be guaranteed to be executing the RCU state
* machine because the scheduling-clock tick has been disabled. Therefore,
* if an adaptive-ticks CPU is failing to respond to the current grace
* period and has not be idle from an RCU perspective, kick it.
*/
rcu: Reduce overhead of cond_resched() checks for RCU Commit ac1bea85781e (Make cond_resched() report RCU quiescent states) fixed a problem where a CPU looping in the kernel with but one runnable task would give RCU CPU stall warnings, even if the in-kernel loop contained cond_resched() calls. Unfortunately, in so doing, it introduced performance regressions in Anton Blanchard's will-it-scale "open1" test. The problem appears to be not so much the increased cond_resched() path length as an increase in the rate at which grace periods complete, which increased per-update grace-period overhead. This commit takes a different approach to fixing this bug, mainly by moving the RCU-visible quiescent state from cond_resched() to rcu_note_context_switch(), and by further reducing the check to a simple non-zero test of a single per-CPU variable. However, this approach requires that the force-quiescent-state processing send resched IPIs to the offending CPUs. These will be sent only once the grace period has reached an age specified by the boot/sysfs parameter rcutree.jiffies_till_sched_qs, or once the grace period reaches an age halfway to the point at which RCU CPU stall warnings will be emitted, whichever comes first. Reported-by: Dave Hansen <dave.hansen@intel.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Andi Kleen <ak@linux.intel.com> Cc: Christoph Lameter <cl@gentwo.org> Cc: Mike Galbraith <umgwanakikbuti@gmail.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org> [ paulmck: Made rcu_momentary_dyntick_idle() as suggested by the ktest build robot. Also fixed smp_mb() comment as noted by Oleg Nesterov. ] Merge with e552592e (Reduce overhead of cond_resched() checks for RCU) Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2014-06-21 07:49:01 +08:00
static void __maybe_unused rcu_kick_nohz_cpu(int cpu)
rcu: Kick adaptive-ticks CPUs that are holding up RCU grace periods Adaptive-ticks CPUs inform RCU when they enter kernel mode, but they do not necessarily turn the scheduler-clock tick back on. This state of affairs could result in RCU waiting on an adaptive-ticks CPU running for an extended period in kernel mode. Such a CPU will never run the RCU state machine, and could therefore indefinitely extend the RCU state machine, sooner or later resulting in an OOM condition. This patch, inspired by an earlier patch by Frederic Weisbecker, therefore causes RCU's force-quiescent-state processing to check for this condition and to send an IPI to CPUs that remain in that state for too long. "Too long" currently means about three jiffies by default, which is quite some time for a CPU to remain in the kernel without blocking. The rcu_tree.jiffies_till_first_fqs and rcutree.jiffies_till_next_fqs sysfs variables may be used to tune "too long" if needed. Reported-by: Frederic Weisbecker <fweisbec@gmail.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Reviewed-by: Josh Triplett <josh@joshtriplett.org> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Chris Metcalf <cmetcalf@tilera.com> Cc: Christoph Lameter <cl@linux.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Gilad Ben Yossef <gilad@benyossef.com> Cc: Hakan Akkan <hakanakkan@gmail.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Kevin Hilman <khilman@linaro.org> Cc: Li Zhong <zhong@linux.vnet.ibm.com> Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Thomas Gleixner <tglx@linutronix.de>
2013-04-13 07:19:10 +08:00
{
#ifdef CONFIG_NO_HZ_FULL
if (tick_nohz_full_cpu(cpu))
smp_send_reschedule(cpu);
#endif /* #ifdef CONFIG_NO_HZ_FULL */
}
nohz_full: Add rcu_dyntick data for scalable detection of all-idle state This commit adds fields to the rcu_dyntick structure that are used to detect idle CPUs. These new fields differ from the existing ones in that the existing ones consider a CPU executing in user mode to be idle, where the new ones consider CPUs executing in user mode to be busy. The handling of these new fields is otherwise quite similar to that for the exiting fields. This commit also adds the initialization required for these fields. So, why is usermode execution treated differently, with RCU considering it a quiescent state equivalent to idle, while in contrast the new full-system idle state detection considers usermode execution to be non-idle? It turns out that although one of RCU's quiescent states is usermode execution, it is not a full-system idle state. This is because the purpose of the full-system idle state is not RCU, but rather determining when accurate timekeeping can safely be disabled. Whenever accurate timekeeping is required in a CONFIG_NO_HZ_FULL kernel, at least one CPU must keep the scheduling-clock tick going. If even one CPU is executing in user mode, accurate timekeeping is requires, particularly for architectures where gettimeofday() and friends do not enter the kernel. Only when all CPUs are really and truly idle can accurate timekeeping be disabled, allowing all CPUs to turn off the scheduling clock interrupt, thus greatly improving energy efficiency. This naturally raises the question "Why is this code in RCU rather than in timekeeping?", and the answer is that RCU has the data and infrastructure to efficiently make this determination. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Frederic Weisbecker <fweisbec@gmail.com> Cc: Steven Rostedt <rostedt@goodmis.org> Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2013-06-22 03:34:33 +08:00
/*
* Is this CPU a NO_HZ_FULL CPU that should ignore RCU so that the
* grace-period kthread will do force_quiescent_state() processing?
* The idea is to avoid waking up RCU core processing on such a
* CPU unless the grace period has extended for too long.
*
* This code relies on the fact that all NO_HZ_FULL CPUs are also
* CONFIG_RCU_NOCB_CPU CPUs.
*/
static bool rcu_nohz_full_cpu(struct rcu_state *rsp)
{
#ifdef CONFIG_NO_HZ_FULL
if (tick_nohz_full_cpu(smp_processor_id()) &&
(!rcu_gp_in_progress(rsp) ||
ULONG_CMP_LT(jiffies, READ_ONCE(rsp->gp_start) + HZ)))
return true;
#endif /* #ifdef CONFIG_NO_HZ_FULL */
return false;
}
/*
* Bind the grace-period kthread for the sysidle flavor of RCU to the
* timekeeping CPU.
*/
static void rcu_bind_gp_kthread(void)
{
int __maybe_unused cpu;
if (!tick_nohz_full_enabled())
return;
housekeeping_affine(current);
}
/* Record the current task on dyntick-idle entry. */
static void rcu_dynticks_task_enter(void)
{
#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL)
WRITE_ONCE(current->rcu_tasks_idle_cpu, smp_processor_id());
#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */
}
/* Record no current task on dyntick-idle exit. */
static void rcu_dynticks_task_exit(void)
{
#if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL)
WRITE_ONCE(current->rcu_tasks_idle_cpu, -1);
#endif /* #if defined(CONFIG_TASKS_RCU) && defined(CONFIG_NO_HZ_FULL) */
}