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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
|
|
|
* Read-Copy Update mechanism for mutual exclusion (tree-based version)
|
|
|
|
* Internal non-public definitions that provide either classic
|
2011-03-02 22:15:15 +01:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
*
|
|
|
|
* 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, write to the Free Software
|
|
|
|
* Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
|
|
|
|
*
|
|
|
|
* Copyright Red Hat, 2009
|
|
|
|
* Copyright IBM Corporation, 2009
|
|
|
|
*
|
|
|
|
* Author: Ingo Molnar <mingo@elte.hu>
|
|
|
|
* Paul E. McKenney <paulmck@linux.vnet.ibm.com>
|
|
|
|
*/
|
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
#include <linux/delay.h>
|
2012-08-20 06:35:53 +02:00
|
|
|
#include <linux/gfp.h>
|
2012-06-12 02:39:43 +02:00
|
|
|
#include <linux/oom.h>
|
2012-07-16 12:42:38 +02:00
|
|
|
#include <linux/smpboot.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>
2009-08-22 22:56:52 +02:00
|
|
|
|
2011-08-19 20:39:11 +02:00
|
|
|
#define RCU_KTHREAD_PRIO 1
|
|
|
|
|
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
#define RCU_BOOST_PRIO CONFIG_RCU_BOOST_PRIO
|
|
|
|
#else
|
|
|
|
#define RCU_BOOST_PRIO RCU_KTHREAD_PRIO
|
|
|
|
#endif
|
|
|
|
|
2012-08-20 06:35:53 +02:00
|
|
|
#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? */
|
|
|
|
static bool rcu_nocb_poll; /* Offload kthread are to poll. */
|
|
|
|
module_param(rcu_nocb_poll, bool, 0444);
|
|
|
|
static char __initdata nocb_buf[NR_CPUS * 5];
|
|
|
|
#endif /* #ifdef CONFIG_RCU_NOCB_CPU */
|
|
|
|
|
2010-04-13 23:19:23 +02:00
|
|
|
/*
|
|
|
|
* Check the RCU kernel configuration parameters and print informative
|
|
|
|
* messages about anything out of the ordinary. If you like #ifdef, you
|
|
|
|
* will love this function.
|
|
|
|
*/
|
|
|
|
static void __init rcu_bootup_announce_oddness(void)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_RCU_TRACE
|
|
|
|
printk(KERN_INFO "\tRCU debugfs-based tracing is enabled.\n");
|
|
|
|
#endif
|
|
|
|
#if (defined(CONFIG_64BIT) && CONFIG_RCU_FANOUT != 64) || (!defined(CONFIG_64BIT) && CONFIG_RCU_FANOUT != 32)
|
|
|
|
printk(KERN_INFO "\tCONFIG_RCU_FANOUT set to non-default value of %d\n",
|
|
|
|
CONFIG_RCU_FANOUT);
|
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_RCU_FANOUT_EXACT
|
|
|
|
printk(KERN_INFO "\tHierarchical RCU autobalancing is disabled.\n");
|
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_RCU_FAST_NO_HZ
|
|
|
|
printk(KERN_INFO
|
|
|
|
"\tRCU dyntick-idle grace-period acceleration is enabled.\n");
|
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_PROVE_RCU
|
|
|
|
printk(KERN_INFO "\tRCU lockdep checking is enabled.\n");
|
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_RCU_TORTURE_TEST_RUNNABLE
|
|
|
|
printk(KERN_INFO "\tRCU torture testing starts during boot.\n");
|
|
|
|
#endif
|
2010-08-30 18:52:50 +02:00
|
|
|
#if defined(CONFIG_TREE_PREEMPT_RCU) && !defined(CONFIG_RCU_CPU_STALL_VERBOSE)
|
2012-01-16 22:29:10 +01:00
|
|
|
printk(KERN_INFO "\tDump stacks of tasks blocking RCU-preempt GP.\n");
|
|
|
|
#endif
|
|
|
|
#if defined(CONFIG_RCU_CPU_STALL_INFO)
|
|
|
|
printk(KERN_INFO "\tAdditional per-CPU info printed with stalls.\n");
|
2010-04-13 23:19:23 +02:00
|
|
|
#endif
|
|
|
|
#if NUM_RCU_LVL_4 != 0
|
2012-06-16 03:16:00 +02:00
|
|
|
printk(KERN_INFO "\tFour-level hierarchy is enabled.\n");
|
2010-04-13 23:19:23 +02:00
|
|
|
#endif
|
2012-04-24 00:52:53 +02:00
|
|
|
if (rcu_fanout_leaf != CONFIG_RCU_FANOUT_LEAF)
|
|
|
|
printk(KERN_INFO "\tExperimental boot-time adjustment of leaf fanout to %d.\n", rcu_fanout_leaf);
|
2012-05-09 06:00:28 +02:00
|
|
|
if (nr_cpu_ids != NR_CPUS)
|
|
|
|
printk(KERN_INFO "\tRCU restricting CPUs from NR_CPUS=%d to nr_cpu_ids=%d.\n", NR_CPUS, nr_cpu_ids);
|
2012-08-20 06:35:53 +02:00
|
|
|
#ifdef CONFIG_RCU_NOCB_CPU
|
|
|
|
if (have_rcu_nocb_mask) {
|
|
|
|
if (cpumask_test_cpu(0, rcu_nocb_mask)) {
|
|
|
|
cpumask_clear_cpu(0, rcu_nocb_mask);
|
|
|
|
pr_info("\tCPU 0: illegal no-CBs CPU (cleared).\n");
|
|
|
|
}
|
|
|
|
cpulist_scnprintf(nocb_buf, sizeof(nocb_buf), rcu_nocb_mask);
|
|
|
|
pr_info("\tExperimental no-CBs CPUs: %s.\n", nocb_buf);
|
|
|
|
if (rcu_nocb_poll)
|
|
|
|
pr_info("\tExperimental polled no-CBs CPUs.\n");
|
|
|
|
}
|
|
|
|
#endif /* #ifdef CONFIG_RCU_NOCB_CPU */
|
2010-04-13 23:19:23 +02:00
|
|
|
}
|
|
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
#ifdef CONFIG_TREE_PREEMPT_RCU
|
|
|
|
|
2012-05-29 08:26:01 +02:00
|
|
|
struct rcu_state rcu_preempt_state =
|
|
|
|
RCU_STATE_INITIALIZER(rcu_preempt, call_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>
2009-08-22 22:56:52 +02:00
|
|
|
DEFINE_PER_CPU(struct rcu_data, rcu_preempt_data);
|
2011-02-07 21:47:15 +01:00
|
|
|
static struct rcu_state *rcu_state = &rcu_preempt_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>
2009-08-22 22:56:52 +02:00
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
static int rcu_preempted_readers_exp(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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
|
|
|
* Tell them what RCU they are running.
|
|
|
|
*/
|
2009-11-11 20:28:06 +01:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
2011-03-02 22:15:15 +01:00
|
|
|
printk(KERN_INFO "Preemptible hierarchical RCU implementation.\n");
|
2010-04-13 23:19:23 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return the number of RCU-preempt batches processed thus far
|
|
|
|
* for debug and statistics.
|
|
|
|
*/
|
|
|
|
long rcu_batches_completed_preempt(void)
|
|
|
|
{
|
|
|
|
return rcu_preempt_state.completed;
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_batches_completed_preempt);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return the number of RCU batches processed thus far for debug & stats.
|
|
|
|
*/
|
|
|
|
long rcu_batches_completed(void)
|
|
|
|
{
|
|
|
|
return rcu_batches_completed_preempt();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_batches_completed);
|
|
|
|
|
2010-01-05 00:09:10 +01:00
|
|
|
/*
|
|
|
|
* Force a quiescent state for preemptible RCU.
|
|
|
|
*/
|
|
|
|
void rcu_force_quiescent_state(void)
|
|
|
|
{
|
2012-06-23 02:06:26 +02:00
|
|
|
force_quiescent_state(&rcu_preempt_state);
|
2010-01-05 00:09:10 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_force_quiescent_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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
* 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 02:37:01 +02:00
|
|
|
*
|
|
|
|
* Unlike the other rcu_*_qs() functions, callers to this function
|
|
|
|
* must disable irqs in order to protect the assignment to
|
|
|
|
* ->rcu_read_unlock_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>
2009-08-22 22:56:52 +02:00
|
|
|
*/
|
2009-09-13 18:15:10 +02:00
|
|
|
static void rcu_preempt_qs(int 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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
struct rcu_data *rdp = &per_cpu(rcu_preempt_data, cpu);
|
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 02:37:01 +02:00
|
|
|
|
rcu: Simplify quiescent-state accounting
There is often a delay between the time that a CPU passes through a
quiescent state and the time that this quiescent state is reported to the
RCU core. It is quite possible that the grace period ended before the
quiescent state could be reported, for example, some other CPU might have
deduced that this CPU passed through dyntick-idle mode. It is critically
important that quiescent state be counted only against the grace period
that was in effect at the time that the quiescent state was detected.
Previously, this was handled by recording the number of the last grace
period to complete when passing through a quiescent state. The RCU
core then checks this number against the current value, and rejects
the quiescent state if there is a mismatch. However, one additional
possibility must be accounted for, namely that the quiescent state was
recorded after the prior grace period completed but before the current
grace period started. In this case, the RCU core must reject the
quiescent state, but the recorded number will match. This is handled
when the CPU becomes aware of a new grace period -- at that point,
it invalidates any prior quiescent state.
This works, but is a bit indirect. The new approach records the current
grace period, and the RCU core checks to see (1) that this is still the
current grace period and (2) that this grace period has not yet ended.
This approach simplifies reasoning about correctness, and this commit
changes over to this new approach.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-27 09:17:43 +02:00
|
|
|
if (rdp->passed_quiesce == 0)
|
rcu: Add grace-period, quiescent-state, and call_rcu trace events
Add trace events to record grace-period start and end, quiescent states,
CPUs noticing grace-period start and end, grace-period initialization,
call_rcu() invocation, tasks blocking in RCU read-side critical sections,
tasks exiting those same critical sections, force_quiescent_state()
detection of dyntick-idle and offline CPUs, CPUs entering and leaving
dyntick-idle mode (except from NMIs), CPUs coming online and going
offline, and CPUs being kicked for staying in dyntick-idle mode for too
long (as in many weeks, even on 32-bit systems).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
rcu: Add the rcu flavor to callback trace events
The earlier trace events for registering RCU callbacks and for invoking
them did not include the RCU flavor (rcu_bh, rcu_preempt, or rcu_sched).
This commit adds the RCU flavor to those trace events.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-25 15:36:56 +02:00
|
|
|
trace_rcu_grace_period("rcu_preempt", rdp->gpnum, "cpuqs");
|
rcu: Simplify quiescent-state accounting
There is often a delay between the time that a CPU passes through a
quiescent state and the time that this quiescent state is reported to the
RCU core. It is quite possible that the grace period ended before the
quiescent state could be reported, for example, some other CPU might have
deduced that this CPU passed through dyntick-idle mode. It is critically
important that quiescent state be counted only against the grace period
that was in effect at the time that the quiescent state was detected.
Previously, this was handled by recording the number of the last grace
period to complete when passing through a quiescent state. The RCU
core then checks this number against the current value, and rejects
the quiescent state if there is a mismatch. However, one additional
possibility must be accounted for, namely that the quiescent state was
recorded after the prior grace period completed but before the current
grace period started. In this case, the RCU core must reject the
quiescent state, but the recorded number will match. This is handled
when the CPU becomes aware of a new grace period -- at that point,
it invalidates any prior quiescent state.
This works, but is a bit indirect. The new approach records the current
grace period, and the RCU core checks to see (1) that this is still the
current grace period and (2) that this grace period has not yet ended.
This approach simplifies reasoning about correctness, and this commit
changes over to this new approach.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-27 09:17:43 +02:00
|
|
|
rdp->passed_quiesce = 1;
|
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 02:37:01 +02:00
|
|
|
current->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
2009-09-13 18:15:10 +02:00
|
|
|
* 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
|
2010-11-30 06:56:39 +01:00
|
|
|
* 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.
|
2009-09-13 18:15:10 +02:00
|
|
|
*
|
|
|
|
* Caller 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>
2009-08-22 22:56:52 +02:00
|
|
|
*/
|
2012-07-02 16:08:42 +02:00
|
|
|
static void rcu_preempt_note_context_switch(int 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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
struct task_struct *t = current;
|
2009-09-13 18:15:10 +02:00
|
|
|
unsigned long 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>
2009-08-22 22:56:52 +02:00
|
|
|
struct rcu_data *rdp;
|
|
|
|
struct rcu_node *rnp;
|
|
|
|
|
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 06:14:35 +02:00
|
|
|
if (t->rcu_read_lock_nesting > 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>
2009-08-22 22:56:52 +02:00
|
|
|
(t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) {
|
|
|
|
|
|
|
|
/* Possibly blocking in an RCU read-side critical section. */
|
2012-07-02 16:08:42 +02:00
|
|
|
rdp = per_cpu_ptr(rcu_preempt_state.rda, 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>
2009-08-22 22:56:52 +02:00
|
|
|
rnp = rdp->mynode;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_lock_irqsave(&rnp->lock, 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>
2009-08-22 22:56:52 +02:00
|
|
|
t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BLOCKED;
|
2009-08-28 00:00:12 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If this CPU has already checked in, then this task
|
|
|
|
* will hold up the next grace period rather than the
|
|
|
|
* current grace period. Queue the task accordingly.
|
|
|
|
* If the task is queued for the current grace period
|
|
|
|
* (i.e., this CPU has not yet passed through a quiescent
|
|
|
|
* state for the current grace period), then as long
|
|
|
|
* as that task remains queued, the current grace period
|
2010-11-30 06:56:39 +01:00
|
|
|
* cannot end. Note that there is some uncertainty as
|
|
|
|
* to exactly when the current grace period started.
|
|
|
|
* We take a conservative approach, which can result
|
|
|
|
* in unnecessarily waiting on tasks that started very
|
|
|
|
* slightly after the current grace period began. C'est
|
|
|
|
* la vie!!!
|
2009-09-13 18:15:09 +02:00
|
|
|
*
|
|
|
|
* But first, note that the current CPU must still be
|
|
|
|
* on line!
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
*/
|
2009-09-13 18:15:09 +02:00
|
|
|
WARN_ON_ONCE((rdp->grpmask & rnp->qsmaskinit) == 0);
|
2009-09-18 18:50:18 +02:00
|
|
|
WARN_ON_ONCE(!list_empty(&t->rcu_node_entry));
|
2010-11-30 06:56:39 +01:00
|
|
|
if ((rnp->qsmask & rdp->grpmask) && rnp->gp_tasks != NULL) {
|
|
|
|
list_add(&t->rcu_node_entry, rnp->gp_tasks->prev);
|
|
|
|
rnp->gp_tasks = &t->rcu_node_entry;
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
if (rnp->boost_tasks != NULL)
|
|
|
|
rnp->boost_tasks = rnp->gp_tasks;
|
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
2010-11-30 06:56:39 +01:00
|
|
|
} else {
|
|
|
|
list_add(&t->rcu_node_entry, &rnp->blkd_tasks);
|
|
|
|
if (rnp->qsmask & rdp->grpmask)
|
|
|
|
rnp->gp_tasks = &t->rcu_node_entry;
|
|
|
|
}
|
rcu: Add grace-period, quiescent-state, and call_rcu trace events
Add trace events to record grace-period start and end, quiescent states,
CPUs noticing grace-period start and end, grace-period initialization,
call_rcu() invocation, tasks blocking in RCU read-side critical sections,
tasks exiting those same critical sections, force_quiescent_state()
detection of dyntick-idle and offline CPUs, CPUs entering and leaving
dyntick-idle mode (except from NMIs), CPUs coming online and going
offline, and CPUs being kicked for staying in dyntick-idle mode for too
long (as in many weeks, even on 32-bit systems).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
rcu: Add the rcu flavor to callback trace events
The earlier trace events for registering RCU callbacks and for invoking
them did not include the RCU flavor (rcu_bh, rcu_preempt, or rcu_sched).
This commit adds the RCU flavor to those trace events.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-25 15:36:56 +02:00
|
|
|
trace_rcu_preempt_task(rdp->rsp->name,
|
|
|
|
t->pid,
|
|
|
|
(rnp->qsmask & rdp->grpmask)
|
|
|
|
? rnp->gpnum
|
|
|
|
: rnp->gpnum + 1);
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
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 06:14:35 +02:00
|
|
|
} else if (t->rcu_read_lock_nesting < 0 &&
|
|
|
|
t->rcu_read_unlock_special) {
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2009-09-18 18:50:18 +02:00
|
|
|
local_irq_save(flags);
|
2012-07-02 16:08:42 +02:00
|
|
|
rcu_preempt_qs(cpu);
|
2009-09-18 18:50:18 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
2009-09-23 18:50:41 +02:00
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2011-02-07 21:47:15 +01:00
|
|
|
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp)
|
2009-09-23 18:50:41 +02:00
|
|
|
{
|
2010-11-30 06:56:39 +01:00
|
|
|
return rnp->gp_tasks != NULL;
|
2009-09-23 18:50:41 +02:00
|
|
|
}
|
|
|
|
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
/*
|
|
|
|
* Record a quiescent state for all tasks that were previously queued
|
|
|
|
* on the specified rcu_node structure and that were blocking the current
|
|
|
|
* RCU grace period. The caller must hold the specified rnp->lock with
|
|
|
|
* irqs disabled, and this lock is released upon return, but irqs remain
|
|
|
|
* disabled.
|
|
|
|
*/
|
2009-12-02 21:10:13 +01:00
|
|
|
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags)
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
__releases(rnp->lock)
|
|
|
|
{
|
|
|
|
unsigned long mask;
|
|
|
|
struct rcu_node *rnp_p;
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
if (rnp->qsmask != 0 || rcu_preempt_blocked_readers_cgp(rnp)) {
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
return; /* Still need more quiescent states! */
|
|
|
|
}
|
|
|
|
|
|
|
|
rnp_p = rnp->parent;
|
|
|
|
if (rnp_p == NULL) {
|
|
|
|
/*
|
|
|
|
* Either there is only one rcu_node in the tree,
|
|
|
|
* or tasks were kicked up to root rcu_node due to
|
|
|
|
* CPUs going offline.
|
|
|
|
*/
|
2009-12-02 21:10:13 +01:00
|
|
|
rcu_report_qs_rsp(&rcu_preempt_state, flags);
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Report up the rest of the hierarchy. */
|
|
|
|
mask = rnp->grpmask;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock(&rnp->lock); /* irqs remain disabled. */
|
|
|
|
raw_spin_lock(&rnp_p->lock); /* irqs already disabled. */
|
2009-12-02 21:10:13 +01:00
|
|
|
rcu_report_qs_rnp(mask, &rcu_preempt_state, rnp_p, flags);
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
}
|
|
|
|
|
2010-11-30 06:56:39 +01:00
|
|
|
/*
|
|
|
|
* 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;
|
|
|
|
}
|
|
|
|
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2012-05-21 20:58:36 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
int empty;
|
2009-12-02 21:10:15 +01:00
|
|
|
int empty_exp;
|
2011-09-21 23:41:37 +02:00
|
|
|
int 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>
2009-08-22 22:56:52 +02:00
|
|
|
unsigned long flags;
|
2010-11-30 06:56:39 +01:00
|
|
|
struct list_head *np;
|
2011-08-04 16:55:34 +02:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
struct rt_mutex *rbmp = NULL;
|
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
struct rcu_node *rnp;
|
|
|
|
int special;
|
|
|
|
|
|
|
|
/* 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 critical section,
|
|
|
|
* let it know that we have done so.
|
|
|
|
*/
|
|
|
|
special = t->rcu_read_unlock_special;
|
|
|
|
if (special & RCU_READ_UNLOCK_NEED_QS) {
|
2009-09-13 18:15:10 +02:00
|
|
|
rcu_preempt_qs(smp_processor_id());
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Hardware IRQ handlers cannot block. */
|
2011-07-20 00:32:00 +02:00
|
|
|
if (in_irq() || in_serving_softirq()) {
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
local_irq_restore(flags);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Clean up if blocked during RCU read-side critical section. */
|
|
|
|
if (special & RCU_READ_UNLOCK_BLOCKED) {
|
|
|
|
t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_BLOCKED;
|
|
|
|
|
2009-08-27 23:58:16 +02:00
|
|
|
/*
|
|
|
|
* Remove this task from the list it blocked on. The
|
|
|
|
* task can migrate while we acquire the lock, but at
|
|
|
|
* most one time. So at most two passes through loop.
|
|
|
|
*/
|
|
|
|
for (;;) {
|
2009-08-28 00:00:12 +02:00
|
|
|
rnp = t->rcu_blocked_node;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_lock(&rnp->lock); /* irqs already disabled. */
|
2009-08-28 00:00:12 +02:00
|
|
|
if (rnp == t->rcu_blocked_node)
|
2009-08-27 23:58:16 +02:00
|
|
|
break;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock(&rnp->lock); /* irqs remain disabled. */
|
2009-08-27 23:58:16 +02:00
|
|
|
}
|
2011-02-07 21:47:15 +01:00
|
|
|
empty = !rcu_preempt_blocked_readers_cgp(rnp);
|
2009-12-02 21:10:15 +01:00
|
|
|
empty_exp = !rcu_preempted_readers_exp(rnp);
|
|
|
|
smp_mb(); /* ensure expedited fastpath sees end of RCU c-s. */
|
2010-11-30 06:56:39 +01:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
list_del_init(&t->rcu_node_entry);
|
2011-08-04 16:55:34 +02:00
|
|
|
t->rcu_blocked_node = NULL;
|
rcu: Add grace-period, quiescent-state, and call_rcu trace events
Add trace events to record grace-period start and end, quiescent states,
CPUs noticing grace-period start and end, grace-period initialization,
call_rcu() invocation, tasks blocking in RCU read-side critical sections,
tasks exiting those same critical sections, force_quiescent_state()
detection of dyntick-idle and offline CPUs, CPUs entering and leaving
dyntick-idle mode (except from NMIs), CPUs coming online and going
offline, and CPUs being kicked for staying in dyntick-idle mode for too
long (as in many weeks, even on 32-bit systems).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
rcu: Add the rcu flavor to callback trace events
The earlier trace events for registering RCU callbacks and for invoking
them did not include the RCU flavor (rcu_bh, rcu_preempt, or rcu_sched).
This commit adds the RCU flavor to those trace events.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-25 15:36:56 +02:00
|
|
|
trace_rcu_unlock_preempted_task("rcu_preempt",
|
|
|
|
rnp->gpnum, t->pid);
|
2010-11-30 06:56:39 +01:00
|
|
|
if (&t->rcu_node_entry == rnp->gp_tasks)
|
|
|
|
rnp->gp_tasks = np;
|
|
|
|
if (&t->rcu_node_entry == rnp->exp_tasks)
|
|
|
|
rnp->exp_tasks = np;
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
if (&t->rcu_node_entry == rnp->boost_tasks)
|
|
|
|
rnp->boost_tasks = np;
|
2011-08-04 16:55:34 +02:00
|
|
|
/* Snapshot/clear ->rcu_boost_mutex with rcu_node lock held. */
|
|
|
|
if (t->rcu_boost_mutex) {
|
|
|
|
rbmp = t->rcu_boost_mutex;
|
|
|
|
t->rcu_boost_mutex = NULL;
|
rcu: Fix RCU_BOOST race handling current->rcu_read_unlock_special
The RCU_BOOST commits for TREE_PREEMPT_RCU introduced an other-task
write to a new RCU_READ_UNLOCK_BOOSTED bit in the task_struct structure's
->rcu_read_unlock_special field, but, as noted by Steven Rostedt, without
correctly synchronizing all accesses to ->rcu_read_unlock_special.
This could result in bits in ->rcu_read_unlock_special being spuriously
set and cleared due to conflicting accesses, which in turn could result
in deadlocks between the rcu_node structure's ->lock and the scheduler's
rq and pi locks. These deadlocks would result from RCU incorrectly
believing that the just-ended RCU read-side critical section had been
preempted and/or boosted. If that RCU read-side critical section was
executed with either rq or pi locks held, RCU's ensuing (incorrect)
calls to the scheduler would cause the scheduler to attempt to once
again acquire the rq and pi locks, resulting in deadlock. More complex
deadlock cycles are also possible, involving multiple rq and pi locks
as well as locks from multiple rcu_node structures.
This commit fixes synchronization by creating ->rcu_boosted field in
task_struct that is accessed and modified only when holding the ->lock
in the rcu_node structure on which the task is queued (on that rcu_node
structure's ->blkd_tasks list). This results in tasks accessing only
their own current->rcu_read_unlock_special fields, making unsynchronized
access once again legal, and keeping the rcu_read_unlock() fastpath free
of atomic instructions and memory barriers.
The reason that the rcu_read_unlock() fastpath does not need to access
the new current->rcu_boosted field is that this new field cannot
be non-zero unless the RCU_READ_UNLOCK_BLOCKED bit is set in the
current->rcu_read_unlock_special field. Therefore, rcu_read_unlock()
need only test current->rcu_read_unlock_special: if that is zero, then
current->rcu_boosted must also be zero.
This bug does not affect TINY_PREEMPT_RCU because this implementation
of RCU accesses current->rcu_read_unlock_special with irqs disabled,
thus preventing races on the !SMP systems that TINY_PREEMPT_RCU runs on.
Maybe-reported-by: Dave Jones <davej@redhat.com>
Maybe-reported-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Reported-by: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Steven Rostedt <rostedt@goodmis.org>
2011-07-14 21:24:11 +02:00
|
|
|
}
|
2011-02-07 21:47:15 +01:00
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If this was the last task on the current list, and if
|
|
|
|
* we aren't waiting on any CPUs, report the quiescent state.
|
2011-09-21 23:41:37 +02:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
*/
|
2011-09-21 23:41:37 +02:00
|
|
|
empty_exp_now = !rcu_preempted_readers_exp(rnp);
|
rcu: Add grace-period, quiescent-state, and call_rcu trace events
Add trace events to record grace-period start and end, quiescent states,
CPUs noticing grace-period start and end, grace-period initialization,
call_rcu() invocation, tasks blocking in RCU read-side critical sections,
tasks exiting those same critical sections, force_quiescent_state()
detection of dyntick-idle and offline CPUs, CPUs entering and leaving
dyntick-idle mode (except from NMIs), CPUs coming online and going
offline, and CPUs being kicked for staying in dyntick-idle mode for too
long (as in many weeks, even on 32-bit systems).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
rcu: Add the rcu flavor to callback trace events
The earlier trace events for registering RCU callbacks and for invoking
them did not include the RCU flavor (rcu_bh, rcu_preempt, or rcu_sched).
This commit adds the RCU flavor to those trace events.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-25 15:36:56 +02:00
|
|
|
if (!empty && !rcu_preempt_blocked_readers_cgp(rnp)) {
|
|
|
|
trace_rcu_quiescent_state_report("preempt_rcu",
|
|
|
|
rnp->gpnum,
|
|
|
|
0, rnp->qsmask,
|
|
|
|
rnp->level,
|
|
|
|
rnp->grplo,
|
|
|
|
rnp->grphi,
|
|
|
|
!!rnp->gp_tasks);
|
2009-12-02 21:10:13 +01:00
|
|
|
rcu_report_unblock_qs_rnp(rnp, flags);
|
2012-06-28 17:08:25 +02:00
|
|
|
} else {
|
rcu: Add grace-period, quiescent-state, and call_rcu trace events
Add trace events to record grace-period start and end, quiescent states,
CPUs noticing grace-period start and end, grace-period initialization,
call_rcu() invocation, tasks blocking in RCU read-side critical sections,
tasks exiting those same critical sections, force_quiescent_state()
detection of dyntick-idle and offline CPUs, CPUs entering and leaving
dyntick-idle mode (except from NMIs), CPUs coming online and going
offline, and CPUs being kicked for staying in dyntick-idle mode for too
long (as in many weeks, even on 32-bit systems).
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
rcu: Add the rcu flavor to callback trace events
The earlier trace events for registering RCU callbacks and for invoking
them did not include the RCU flavor (rcu_bh, rcu_preempt, or rcu_sched).
This commit adds the RCU flavor to those trace events.
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-06-25 15:36:56 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2012-06-28 17:08:25 +02:00
|
|
|
}
|
2009-12-02 21:10:15 +01:00
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
/* Unboost if we were boosted. */
|
2011-08-04 16:55:34 +02:00
|
|
|
if (rbmp)
|
|
|
|
rt_mutex_unlock(rbmp);
|
2011-02-07 21:47:15 +01:00
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
/*
|
|
|
|
* If this was the last task on the expedited lists,
|
|
|
|
* then we need to report up the rcu_node hierarchy.
|
|
|
|
*/
|
2011-09-21 23:41:37 +02:00
|
|
|
if (!empty_exp && empty_exp_now)
|
2011-10-22 16:12:34 +02:00
|
|
|
rcu_report_exp_rnp(&rcu_preempt_state, 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>
2009-11-22 17:53:48 +01:00
|
|
|
} 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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-02-23 02:05:05 +01:00
|
|
|
#ifdef CONFIG_RCU_CPU_STALL_VERBOSE
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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;
|
|
|
|
|
2010-11-30 06:56:39 +01:00
|
|
|
raw_spin_lock_irqsave(&rnp->lock, flags);
|
2012-08-11 01:00:11 +02:00
|
|
|
if (!rcu_preempt_blocked_readers_cgp(rnp)) {
|
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
|
|
|
return;
|
|
|
|
}
|
2010-11-30 06:56:39 +01:00
|
|
|
t = list_entry(rnp->gp_tasks,
|
|
|
|
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(&rnp->lock, flags);
|
2010-02-23 02:05:05 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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);
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_CPU_STALL_VERBOSE */
|
|
|
|
|
|
|
|
static void rcu_print_detail_task_stall(struct rcu_state *rsp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_CPU_STALL_VERBOSE */
|
|
|
|
|
2012-01-16 22:29:10 +01:00
|
|
|
#ifdef CONFIG_RCU_CPU_STALL_INFO
|
|
|
|
|
|
|
|
static void rcu_print_task_stall_begin(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
printk(KERN_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)
|
|
|
|
{
|
|
|
|
printk(KERN_CONT "\n");
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_CPU_STALL_INFO */
|
|
|
|
|
|
|
|
static void rcu_print_task_stall_begin(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static void rcu_print_task_stall_end(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_CPU_STALL_INFO */
|
|
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
|
|
|
* Scan the current list of tasks blocked within RCU read-side critical
|
|
|
|
* sections, printing out the tid of each.
|
|
|
|
*/
|
2011-08-13 22:31:47 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
struct task_struct *t;
|
2011-08-13 22:31:47 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
if (!rcu_preempt_blocked_readers_cgp(rnp))
|
2011-08-13 22:31:47 +02:00
|
|
|
return 0;
|
2012-01-16 22:29:10 +01:00
|
|
|
rcu_print_task_stall_begin(rnp);
|
2010-11-30 06:56:39 +01:00
|
|
|
t = list_entry(rnp->gp_tasks,
|
|
|
|
struct task_struct, rcu_node_entry);
|
2011-08-13 22:31:47 +02:00
|
|
|
list_for_each_entry_continue(t, &rnp->blkd_tasks, rcu_node_entry) {
|
2012-01-16 22:29:10 +01:00
|
|
|
printk(KERN_CONT " P%d", t->pid);
|
2011-08-13 22:31:47 +02:00
|
|
|
ndetected++;
|
|
|
|
}
|
2012-01-16 22:29:10 +01:00
|
|
|
rcu_print_task_stall_end();
|
2011-08-13 22:31:47 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
2009-09-13 18:15:09 +02:00
|
|
|
/*
|
|
|
|
* 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.
|
2010-11-30 06:56:39 +01:00
|
|
|
*
|
|
|
|
* Also, if there are blocked tasks on the list, they automatically
|
|
|
|
* block the newly created grace period, so set up ->gp_tasks accordingly.
|
2009-09-13 18:15:09 +02:00
|
|
|
*/
|
|
|
|
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
|
|
|
|
{
|
2011-02-07 21:47:15 +01:00
|
|
|
WARN_ON_ONCE(rcu_preempt_blocked_readers_cgp(rnp));
|
2010-11-30 06:56:39 +01:00
|
|
|
if (!list_empty(&rnp->blkd_tasks))
|
|
|
|
rnp->gp_tasks = rnp->blkd_tasks.next;
|
2009-09-18 18:50:17 +02:00
|
|
|
WARN_ON_ONCE(rnp->qsmask);
|
2009-09-13 18:15:09 +02:00
|
|
|
}
|
|
|
|
|
2009-08-24 18:42:01 +02:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
2009-08-27 23:58:16 +02:00
|
|
|
/*
|
|
|
|
* Handle tasklist migration for case in which all CPUs covered by the
|
|
|
|
* specified rcu_node have gone offline. Move them up to the root
|
|
|
|
* rcu_node. The reason for not just moving them to the immediate
|
|
|
|
* parent is to remove the need for rcu_read_unlock_special() to
|
|
|
|
* make more than two attempts to acquire the target rcu_node's lock.
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
* Returns true if there were tasks blocking the current RCU grace
|
|
|
|
* period.
|
2009-08-27 23:58:16 +02:00
|
|
|
*
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
* Returns 1 if there was previously a task blocking the current grace
|
|
|
|
* period on the specified rcu_node structure.
|
|
|
|
*
|
2009-08-27 23:58:16 +02:00
|
|
|
* The caller must hold rnp->lock with irqs disabled.
|
|
|
|
*/
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
|
|
|
|
struct rcu_node *rnp,
|
|
|
|
struct rcu_data *rdp)
|
2009-08-27 23:58:16 +02:00
|
|
|
{
|
|
|
|
struct list_head *lp;
|
|
|
|
struct list_head *lp_root;
|
2009-12-02 21:10:15 +01:00
|
|
|
int retval = 0;
|
2009-08-27 23:58:16 +02:00
|
|
|
struct rcu_node *rnp_root = rcu_get_root(rsp);
|
2010-11-30 06:56:39 +01:00
|
|
|
struct task_struct *t;
|
2009-08-27 23:58:16 +02:00
|
|
|
|
2009-08-28 00:00:12 +02:00
|
|
|
if (rnp == rnp_root) {
|
|
|
|
WARN_ONCE(1, "Last CPU thought to be offlined?");
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
return 0; /* Shouldn't happen: at least one CPU online. */
|
2009-08-28 00:00:12 +02:00
|
|
|
}
|
2010-11-30 06:56:39 +01:00
|
|
|
|
|
|
|
/* If we are on an internal node, complain bitterly. */
|
|
|
|
WARN_ON_ONCE(rnp != rdp->mynode);
|
2009-08-27 23:58:16 +02:00
|
|
|
|
|
|
|
/*
|
2010-11-30 06:56:39 +01:00
|
|
|
* Move tasks up to root rcu_node. Don't try to get fancy for
|
|
|
|
* this corner-case operation -- just put this node's tasks
|
|
|
|
* at the head of the root node's list, and update the root node's
|
|
|
|
* ->gp_tasks and ->exp_tasks pointers to those of this node's,
|
|
|
|
* if non-NULL. This might result in waiting for more tasks than
|
|
|
|
* absolutely necessary, but this is a good performance/complexity
|
|
|
|
* tradeoff.
|
2009-08-27 23:58:16 +02:00
|
|
|
*/
|
2012-01-31 02:02:47 +01:00
|
|
|
if (rcu_preempt_blocked_readers_cgp(rnp) && rnp->qsmask == 0)
|
2009-12-02 21:10:15 +01:00
|
|
|
retval |= RCU_OFL_TASKS_NORM_GP;
|
|
|
|
if (rcu_preempted_readers_exp(rnp))
|
|
|
|
retval |= RCU_OFL_TASKS_EXP_GP;
|
2010-11-30 06:56:39 +01:00
|
|
|
lp = &rnp->blkd_tasks;
|
|
|
|
lp_root = &rnp_root->blkd_tasks;
|
|
|
|
while (!list_empty(lp)) {
|
|
|
|
t = list_entry(lp->next, typeof(*t), rcu_node_entry);
|
|
|
|
raw_spin_lock(&rnp_root->lock); /* irqs already disabled */
|
|
|
|
list_del(&t->rcu_node_entry);
|
|
|
|
t->rcu_blocked_node = rnp_root;
|
|
|
|
list_add(&t->rcu_node_entry, lp_root);
|
|
|
|
if (&t->rcu_node_entry == rnp->gp_tasks)
|
|
|
|
rnp_root->gp_tasks = rnp->gp_tasks;
|
|
|
|
if (&t->rcu_node_entry == rnp->exp_tasks)
|
|
|
|
rnp_root->exp_tasks = rnp->exp_tasks;
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
if (&t->rcu_node_entry == rnp->boost_tasks)
|
|
|
|
rnp_root->boost_tasks = rnp->boost_tasks;
|
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
2010-11-30 06:56:39 +01:00
|
|
|
raw_spin_unlock(&rnp_root->lock); /* irqs still disabled */
|
2009-08-27 23:58:16 +02:00
|
|
|
}
|
2011-02-07 21:47:15 +01:00
|
|
|
|
2012-07-27 22:41:47 +02:00
|
|
|
rnp->gp_tasks = NULL;
|
|
|
|
rnp->exp_tasks = NULL;
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
2012-07-27 22:41:47 +02:00
|
|
|
rnp->boost_tasks = NULL;
|
rcu: Improve boost selection when moving tasks to root rcu_node
The rcu_preempt_offline_tasks() moves all tasks queued on a given leaf
rcu_node structure to the root rcu_node, which is done when the last CPU
corresponding the the leaf rcu_node structure goes offline. Now that
RCU-preempt's synchronize_rcu_expedited() implementation blocks CPU-hotplug
operations during the initialization of each rcu_node structure's
->boost_tasks pointer, rcu_preempt_offline_tasks() can do a better job
of setting the root rcu_node's ->boost_tasks pointer.
The key point is that rcu_preempt_offline_tasks() runs as part of the
CPU-hotplug process, so that a concurrent synchronize_rcu_expedited()
is guaranteed to either have not started on the one hand (in which case
there is no boosting on behalf of the expedited grace period) or to be
completely initialized on the other (in which case, in the absence of
other priority boosting, all ->boost_tasks pointers will be initialized).
Therefore, if rcu_preempt_offline_tasks() finds that the ->boost_tasks
pointer is equal to the ->exp_tasks pointer, it can be sure that it is
correctly placed.
In the case where there was boosting ongoing at the time that the
synchronize_rcu_expedited() function started, different nodes might start
boosting the tasks blocking the expedited grace period at different times.
In this mixed case, the root node will either be boosting tasks for
the expedited grace period already, or it will start as soon as it gets
done boosting for the normal grace period -- but in this latter case,
the root node's tasks needed to be boosted in any case.
This commit therefore adds a check of the ->boost_tasks pointer against
the ->exp_tasks pointer to the list that prevents updating ->boost_tasks.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2012-07-31 23:09:49 +02:00
|
|
|
/*
|
|
|
|
* In case root is being boosted and leaf was not. Make sure
|
|
|
|
* that we boost the tasks blocking the current grace period
|
|
|
|
* in this case.
|
|
|
|
*/
|
2011-02-07 21:47:15 +01:00
|
|
|
raw_spin_lock(&rnp_root->lock); /* irqs already disabled */
|
|
|
|
if (rnp_root->boost_tasks != NULL &&
|
rcu: Improve boost selection when moving tasks to root rcu_node
The rcu_preempt_offline_tasks() moves all tasks queued on a given leaf
rcu_node structure to the root rcu_node, which is done when the last CPU
corresponding the the leaf rcu_node structure goes offline. Now that
RCU-preempt's synchronize_rcu_expedited() implementation blocks CPU-hotplug
operations during the initialization of each rcu_node structure's
->boost_tasks pointer, rcu_preempt_offline_tasks() can do a better job
of setting the root rcu_node's ->boost_tasks pointer.
The key point is that rcu_preempt_offline_tasks() runs as part of the
CPU-hotplug process, so that a concurrent synchronize_rcu_expedited()
is guaranteed to either have not started on the one hand (in which case
there is no boosting on behalf of the expedited grace period) or to be
completely initialized on the other (in which case, in the absence of
other priority boosting, all ->boost_tasks pointers will be initialized).
Therefore, if rcu_preempt_offline_tasks() finds that the ->boost_tasks
pointer is equal to the ->exp_tasks pointer, it can be sure that it is
correctly placed.
In the case where there was boosting ongoing at the time that the
synchronize_rcu_expedited() function started, different nodes might start
boosting the tasks blocking the expedited grace period at different times.
In this mixed case, the root node will either be boosting tasks for
the expedited grace period already, or it will start as soon as it gets
done boosting for the normal grace period -- but in this latter case,
the root node's tasks needed to be boosted in any case.
This commit therefore adds a check of the ->boost_tasks pointer against
the ->exp_tasks pointer to the list that prevents updating ->boost_tasks.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Reviewed-by: Josh Triplett <josh@joshtriplett.org>
2012-07-31 23:09:49 +02:00
|
|
|
rnp_root->boost_tasks != rnp_root->gp_tasks &&
|
|
|
|
rnp_root->boost_tasks != rnp_root->exp_tasks)
|
2011-02-07 21:47:15 +01:00
|
|
|
rnp_root->boost_tasks = rnp_root->gp_tasks;
|
|
|
|
raw_spin_unlock(&rnp_root->lock); /* irqs still disabled */
|
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
|
|
|
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
return retval;
|
2009-08-27 23:58:16 +02:00
|
|
|
}
|
|
|
|
|
2012-01-07 20:03:57 +01:00
|
|
|
#endif /* #ifdef CONFIG_HOTPLUG_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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
|
|
|
* 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(int cpu)
|
|
|
|
{
|
|
|
|
struct task_struct *t = current;
|
|
|
|
|
|
|
|
if (t->rcu_read_lock_nesting == 0) {
|
2009-09-13 18:15:10 +02:00
|
|
|
rcu_preempt_qs(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>
2009-08-22 22:56:52 +02:00
|
|
|
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 06:14:35 +02:00
|
|
|
if (t->rcu_read_lock_nesting > 0 &&
|
|
|
|
per_cpu(rcu_preempt_data, cpu).qs_pending)
|
2009-09-13 18:15:10 +02:00
|
|
|
t->rcu_read_unlock_special |= RCU_READ_UNLOCK_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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
2011-06-16 00:47:09 +02:00
|
|
|
#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 07:26:25 +02:00
|
|
|
static void rcu_preempt_do_callbacks(void)
|
|
|
|
{
|
|
|
|
rcu_do_batch(&rcu_preempt_state, &__get_cpu_var(rcu_preempt_data));
|
|
|
|
}
|
|
|
|
|
2011-06-16 00:47:09 +02:00
|
|
|
#endif /* #ifdef CONFIG_RCU_BOOST */
|
|
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Queue a preemptible-RCU callback for invocation after a grace period.
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
*/
|
|
|
|
void call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu))
|
|
|
|
{
|
2012-08-20 06:35:53 +02:00
|
|
|
__call_rcu(head, func, &rcu_preempt_state, -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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(call_rcu);
|
|
|
|
|
2012-01-06 23:11:30 +01:00
|
|
|
/*
|
|
|
|
* Queue an RCU callback for lazy invocation after a grace period.
|
|
|
|
* This will likely be later named something like "call_rcu_lazy()",
|
|
|
|
* but this change will require some way of tagging the lazy RCU
|
|
|
|
* callbacks in the list of pending callbacks. Until then, this
|
|
|
|
* function may only be called from __kfree_rcu().
|
|
|
|
*/
|
|
|
|
void kfree_call_rcu(struct rcu_head *head,
|
|
|
|
void (*func)(struct rcu_head *rcu))
|
|
|
|
{
|
2012-08-20 06:35:53 +02:00
|
|
|
__call_rcu(head, func, &rcu_preempt_state, -1, 1);
|
2012-01-06 23:11:30 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(kfree_call_rcu);
|
|
|
|
|
2009-11-22 17:53:50 +01:00
|
|
|
/**
|
|
|
|
* 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
|
2010-07-09 02:38:59 +02:00
|
|
|
* 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.
|
2012-10-23 22:47:01 +02:00
|
|
|
*
|
|
|
|
* See the description of synchronize_sched() for more detailed information
|
|
|
|
* on memory ordering guarantees.
|
2009-11-22 17:53:50 +01:00
|
|
|
*/
|
|
|
|
void synchronize_rcu(void)
|
|
|
|
{
|
2012-01-04 22:30:33 +01:00
|
|
|
rcu_lockdep_assert(!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");
|
2009-11-22 17:53:50 +01:00
|
|
|
if (!rcu_scheduler_active)
|
|
|
|
return;
|
2012-10-05 08:59:15 +02:00
|
|
|
if (rcu_expedited)
|
|
|
|
synchronize_rcu_expedited();
|
|
|
|
else
|
|
|
|
wait_rcu_gp(call_rcu);
|
2009-11-22 17:53:50 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(synchronize_rcu);
|
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
static DECLARE_WAIT_QUEUE_HEAD(sync_rcu_preempt_exp_wq);
|
2012-07-24 01:03:51 +02:00
|
|
|
static unsigned long sync_rcu_preempt_exp_count;
|
2009-12-02 21:10:15 +01:00
|
|
|
static DEFINE_MUTEX(sync_rcu_preempt_exp_mutex);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return non-zero if there are any tasks in RCU read-side critical
|
|
|
|
* sections blocking the current preemptible-RCU expedited grace period.
|
|
|
|
* If there is no preemptible-RCU expedited grace period currently in
|
|
|
|
* progress, returns zero unconditionally.
|
|
|
|
*/
|
|
|
|
static int rcu_preempted_readers_exp(struct rcu_node *rnp)
|
|
|
|
{
|
2010-11-30 06:56:39 +01:00
|
|
|
return rnp->exp_tasks != NULL;
|
2009-12-02 21:10:15 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* return non-zero if there is no RCU expedited grace period in progress
|
|
|
|
* for the specified rcu_node structure, in other words, if all CPUs and
|
|
|
|
* tasks covered by the specified rcu_node structure have done their bit
|
|
|
|
* for the current expedited grace period. Works only for preemptible
|
|
|
|
* RCU -- other RCU implementation use other means.
|
|
|
|
*
|
|
|
|
* Caller must hold sync_rcu_preempt_exp_mutex.
|
|
|
|
*/
|
|
|
|
static int sync_rcu_preempt_exp_done(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
return !rcu_preempted_readers_exp(rnp) &&
|
|
|
|
ACCESS_ONCE(rnp->expmask) == 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Report the exit from RCU read-side critical section for the last task
|
|
|
|
* that queued itself during or before the current expedited preemptible-RCU
|
|
|
|
* grace period. This event is reported either to the rcu_node structure on
|
|
|
|
* which the task was queued or to one of that rcu_node structure's ancestors,
|
|
|
|
* recursively up the tree. (Calm down, calm down, we do the recursion
|
|
|
|
* iteratively!)
|
|
|
|
*
|
2011-10-22 16:12:34 +02:00
|
|
|
* Most callers will set the "wake" flag, but the task initiating the
|
|
|
|
* expedited grace period need not wake itself.
|
|
|
|
*
|
2009-12-02 21:10:15 +01:00
|
|
|
* Caller must hold sync_rcu_preempt_exp_mutex.
|
|
|
|
*/
|
2011-10-22 16:12:34 +02:00
|
|
|
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
|
|
|
|
bool wake)
|
2009-12-02 21:10:15 +01:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
unsigned long mask;
|
|
|
|
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_lock_irqsave(&rnp->lock, flags);
|
2009-12-02 21:10:15 +01:00
|
|
|
for (;;) {
|
2011-07-17 11:05:49 +02:00
|
|
|
if (!sync_rcu_preempt_exp_done(rnp)) {
|
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2009-12-02 21:10:15 +01:00
|
|
|
break;
|
2011-07-17 11:05:49 +02:00
|
|
|
}
|
2009-12-02 21:10:15 +01:00
|
|
|
if (rnp->parent == NULL) {
|
2011-07-17 11:05:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2011-10-22 16:12:34 +02:00
|
|
|
if (wake)
|
|
|
|
wake_up(&sync_rcu_preempt_exp_wq);
|
2009-12-02 21:10:15 +01:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
mask = rnp->grpmask;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock(&rnp->lock); /* irqs remain disabled */
|
2009-12-02 21:10:15 +01:00
|
|
|
rnp = rnp->parent;
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_lock(&rnp->lock); /* irqs already disabled */
|
2009-12-02 21:10:15 +01:00
|
|
|
rnp->expmask &= ~mask;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Snapshot the tasks blocking the newly started preemptible-RCU expedited
|
|
|
|
* grace period for the specified rcu_node structure. If there are no such
|
|
|
|
* tasks, report it up the rcu_node hierarchy.
|
|
|
|
*
|
2012-10-08 19:54:03 +02:00
|
|
|
* Caller must hold sync_rcu_preempt_exp_mutex and must exclude
|
|
|
|
* CPU hotplug operations.
|
2009-12-02 21:10:15 +01:00
|
|
|
*/
|
|
|
|
static void
|
|
|
|
sync_rcu_preempt_exp_init(struct rcu_state *rsp, struct rcu_node *rnp)
|
|
|
|
{
|
2011-05-05 06:43:49 +02:00
|
|
|
unsigned long flags;
|
2010-11-30 06:56:39 +01:00
|
|
|
int must_wait = 0;
|
2009-12-02 21:10:15 +01:00
|
|
|
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_lock_irqsave(&rnp->lock, flags);
|
2012-06-28 17:08:25 +02:00
|
|
|
if (list_empty(&rnp->blkd_tasks)) {
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2012-06-28 17:08:25 +02:00
|
|
|
} else {
|
2010-11-30 06:56:39 +01:00
|
|
|
rnp->exp_tasks = rnp->blkd_tasks.next;
|
2011-05-05 06:43:49 +02:00
|
|
|
rcu_initiate_boost(rnp, flags); /* releases rnp->lock */
|
2010-11-30 06:56:39 +01:00
|
|
|
must_wait = 1;
|
|
|
|
}
|
2009-12-02 21:10:15 +01:00
|
|
|
if (!must_wait)
|
2011-10-22 16:12:34 +02:00
|
|
|
rcu_report_exp_rnp(rsp, rnp, false); /* Don't wake self. */
|
2009-12-02 21:10:15 +01:00
|
|
|
}
|
|
|
|
|
2012-01-31 23:00:41 +01:00
|
|
|
/**
|
|
|
|
* synchronize_rcu_expedited - Brute-force RCU grace period
|
|
|
|
*
|
|
|
|
* Wait for an RCU-preempt grace period, but expedite it. The basic
|
|
|
|
* idea is to invoke synchronize_sched_expedited() to push all the tasks to
|
|
|
|
* the ->blkd_tasks lists and wait for this list to drain. This consumes
|
|
|
|
* significant time on all CPUs and is unfriendly to real-time workloads,
|
|
|
|
* so is thus not recommended for any sort of common-case code.
|
|
|
|
* In fact, if you are using synchronize_rcu_expedited() in a loop,
|
|
|
|
* please restructure your code to batch your updates, and then Use a
|
|
|
|
* single synchronize_rcu() instead.
|
|
|
|
*
|
|
|
|
* Note that it is illegal to call this function while holding any lock
|
|
|
|
* that is acquired by a CPU-hotplug notifier. And yes, it is also illegal
|
|
|
|
* to call this function from a CPU-hotplug notifier. Failing to observe
|
|
|
|
* these restriction will result in deadlock.
|
2009-10-14 19:15:56 +02:00
|
|
|
*/
|
|
|
|
void synchronize_rcu_expedited(void)
|
|
|
|
{
|
2009-12-02 21:10:15 +01:00
|
|
|
unsigned long flags;
|
|
|
|
struct rcu_node *rnp;
|
|
|
|
struct rcu_state *rsp = &rcu_preempt_state;
|
2012-07-24 01:03:51 +02:00
|
|
|
unsigned long snap;
|
2009-12-02 21:10:15 +01:00
|
|
|
int trycount = 0;
|
|
|
|
|
|
|
|
smp_mb(); /* Caller's modifications seen first by other CPUs. */
|
|
|
|
snap = ACCESS_ONCE(sync_rcu_preempt_exp_count) + 1;
|
|
|
|
smp_mb(); /* Above access cannot bleed into critical section. */
|
|
|
|
|
2012-07-31 02:19:25 +02:00
|
|
|
/*
|
|
|
|
* Block CPU-hotplug operations. This means that any CPU-hotplug
|
|
|
|
* operation that finds an rcu_node structure with tasks in the
|
|
|
|
* process of being boosted will know that all tasks blocking
|
|
|
|
* this expedited grace period will already be in the process of
|
|
|
|
* being boosted. This simplifies the process of moving tasks
|
|
|
|
* from leaf to root rcu_node structures.
|
|
|
|
*/
|
|
|
|
get_online_cpus();
|
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
/*
|
|
|
|
* Acquire lock, falling back to synchronize_rcu() if too many
|
|
|
|
* lock-acquisition failures. Of course, if someone does the
|
|
|
|
* expedited grace period for us, just leave.
|
|
|
|
*/
|
|
|
|
while (!mutex_trylock(&sync_rcu_preempt_exp_mutex)) {
|
2012-07-31 02:19:25 +02:00
|
|
|
if (ULONG_CMP_LT(snap,
|
|
|
|
ACCESS_ONCE(sync_rcu_preempt_exp_count))) {
|
|
|
|
put_online_cpus();
|
|
|
|
goto mb_ret; /* Others did our work for us. */
|
|
|
|
}
|
2012-06-28 17:08:25 +02:00
|
|
|
if (trycount++ < 10) {
|
2009-12-02 21:10:15 +01:00
|
|
|
udelay(trycount * num_online_cpus());
|
2012-06-28 17:08:25 +02:00
|
|
|
} else {
|
2012-07-31 02:19:25 +02:00
|
|
|
put_online_cpus();
|
2012-10-05 08:59:15 +02:00
|
|
|
wait_rcu_gp(call_rcu);
|
2009-12-02 21:10:15 +01:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
2012-07-31 02:19:25 +02:00
|
|
|
if (ULONG_CMP_LT(snap, ACCESS_ONCE(sync_rcu_preempt_exp_count))) {
|
|
|
|
put_online_cpus();
|
2009-12-02 21:10:15 +01:00
|
|
|
goto unlock_mb_ret; /* Others did our work for us. */
|
2012-07-31 02:19:25 +02:00
|
|
|
}
|
2009-12-02 21:10:15 +01:00
|
|
|
|
2010-11-30 06:56:39 +01:00
|
|
|
/* force all RCU readers onto ->blkd_tasks lists. */
|
2009-12-02 21:10:15 +01:00
|
|
|
synchronize_sched_expedited();
|
|
|
|
|
|
|
|
/* Initialize ->expmask for all non-leaf rcu_node structures. */
|
|
|
|
rcu_for_each_nonleaf_node_breadth_first(rsp, rnp) {
|
2012-07-31 02:19:25 +02:00
|
|
|
raw_spin_lock_irqsave(&rnp->lock, flags);
|
2009-12-02 21:10:15 +01:00
|
|
|
rnp->expmask = rnp->qsmaskinit;
|
2012-07-31 02:19:25 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2009-12-02 21:10:15 +01:00
|
|
|
}
|
|
|
|
|
2010-11-30 06:56:39 +01:00
|
|
|
/* Snapshot current state of ->blkd_tasks lists. */
|
2009-12-02 21:10:15 +01:00
|
|
|
rcu_for_each_leaf_node(rsp, rnp)
|
|
|
|
sync_rcu_preempt_exp_init(rsp, rnp);
|
|
|
|
if (NUM_RCU_NODES > 1)
|
|
|
|
sync_rcu_preempt_exp_init(rsp, rcu_get_root(rsp));
|
|
|
|
|
2012-07-31 02:19:25 +02:00
|
|
|
put_online_cpus();
|
2009-12-02 21:10:15 +01:00
|
|
|
|
2010-11-30 06:56:39 +01:00
|
|
|
/* Wait for snapshotted ->blkd_tasks lists to drain. */
|
2009-12-02 21:10:15 +01:00
|
|
|
rnp = rcu_get_root(rsp);
|
|
|
|
wait_event(sync_rcu_preempt_exp_wq,
|
|
|
|
sync_rcu_preempt_exp_done(rnp));
|
|
|
|
|
|
|
|
/* Clean up and exit. */
|
|
|
|
smp_mb(); /* ensure expedited GP seen before counter increment. */
|
|
|
|
ACCESS_ONCE(sync_rcu_preempt_exp_count)++;
|
|
|
|
unlock_mb_ret:
|
|
|
|
mutex_unlock(&sync_rcu_preempt_exp_mutex);
|
|
|
|
mb_ret:
|
|
|
|
smp_mb(); /* ensure subsequent action seen after grace period. */
|
2009-10-14 19:15:56 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);
|
|
|
|
|
2009-10-07 06:48:17 +02:00
|
|
|
/**
|
|
|
|
* rcu_barrier - Wait until all in-flight call_rcu() callbacks complete.
|
2012-10-23 22:47:01 +02:00
|
|
|
*
|
|
|
|
* 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.
|
2009-10-07 06:48:17 +02:00
|
|
|
*/
|
|
|
|
void rcu_barrier(void)
|
|
|
|
{
|
2012-05-29 08:26:01 +02:00
|
|
|
_rcu_barrier(&rcu_preempt_state);
|
2009-10-07 06:48:17 +02:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_barrier);
|
|
|
|
|
2009-09-23 18:50:42 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Initialize preemptible RCU's state structures.
|
2009-09-23 18:50:42 +02:00
|
|
|
*/
|
|
|
|
static void __init __rcu_init_preempt(void)
|
|
|
|
{
|
2010-06-28 10:25:04 +02:00
|
|
|
rcu_init_one(&rcu_preempt_state, &rcu_preempt_data);
|
2009-09-23 18:50:42 +02:00
|
|
|
}
|
|
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
static struct rcu_state *rcu_state = &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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
|
|
|
* Tell them what RCU they are running.
|
|
|
|
*/
|
2009-11-11 20:28:06 +01:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
printk(KERN_INFO "Hierarchical RCU implementation.\n");
|
2010-04-13 23:19:23 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Return the number of RCU batches processed thus far for debug & stats.
|
|
|
|
*/
|
|
|
|
long rcu_batches_completed(void)
|
|
|
|
{
|
|
|
|
return rcu_batches_completed_sched();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_batches_completed);
|
|
|
|
|
2010-01-05 00:09:10 +01:00
|
|
|
/*
|
|
|
|
* Force a quiescent state for RCU, which, because there is no preemptible
|
|
|
|
* RCU, becomes the same as rcu-sched.
|
|
|
|
*/
|
|
|
|
void rcu_force_quiescent_state(void)
|
|
|
|
{
|
|
|
|
rcu_sched_force_quiescent_state();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_force_quiescent_state);
|
|
|
|
|
2012-07-02 16:08:42 +02:00
|
|
|
/*
|
|
|
|
* Because preemptible RCU does not exist, we never have to check for
|
|
|
|
* CPUs being in quiescent states.
|
|
|
|
*/
|
|
|
|
static void rcu_preempt_note_context_switch(int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2009-09-23 18:50:41 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, there are never any preempted
|
2009-09-23 18:50:41 +02:00
|
|
|
* RCU readers.
|
|
|
|
*/
|
2011-02-07 21:47:15 +01:00
|
|
|
static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp)
|
2009-09-23 18:50:41 +02:00
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
|
|
|
/* Because preemptible RCU does not exist, no quieting of tasks. */
|
2009-12-02 21:10:13 +01:00
|
|
|
static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags)
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
{
|
2010-02-23 02:05:02 +01:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
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>
2009-11-22 17:53:48 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #ifdef CONFIG_HOTPLUG_CPU */
|
|
|
|
|
2010-02-23 02:05:05 +01:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, we never have to check for
|
2010-02-23 02:05:05 +01:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
* tasks blocked within RCU read-side critical sections.
|
|
|
|
*/
|
2011-08-13 22:31:47 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
2011-08-13 22:31:47 +02:00
|
|
|
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>
2009-08-22 22:56:52 +02:00
|
|
|
}
|
|
|
|
|
2009-09-13 18:15:09 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because there is no preemptible RCU, there can be no readers blocked,
|
2009-09-18 18:50:19 +02:00
|
|
|
* so there is no need to check for blocked tasks. So check only for
|
|
|
|
* bogus qsmask values.
|
2009-09-13 18:15:09 +02:00
|
|
|
*/
|
|
|
|
static void rcu_preempt_check_blocked_tasks(struct rcu_node *rnp)
|
|
|
|
{
|
2009-09-18 18:50:19 +02:00
|
|
|
WARN_ON_ONCE(rnp->qsmask);
|
2009-09-13 18:15:09 +02:00
|
|
|
}
|
|
|
|
|
2009-08-24 18:42:01 +02:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
2009-08-27 23:58:16 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, it never needs to migrate
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
* tasks that were blocked within RCU read-side critical sections, and
|
|
|
|
* such non-existent tasks cannot possibly have been blocking the current
|
|
|
|
* grace period.
|
2009-08-27 23:58:16 +02:00
|
|
|
*/
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
static int rcu_preempt_offline_tasks(struct rcu_state *rsp,
|
|
|
|
struct rcu_node *rnp,
|
|
|
|
struct rcu_data *rdp)
|
2009-08-27 23:58:16 +02:00
|
|
|
{
|
rcu: Fix TREE_PREEMPT_RCU CPU_HOTPLUG bad-luck hang
If the following sequence of events occurs, then
TREE_PREEMPT_RCU will hang waiting for a grace period to
complete, eventually OOMing the system:
o A TREE_PREEMPT_RCU build of the kernel is booted on a system
with more than 64 physical CPUs present (32 on a 32-bit system).
Alternatively, a TREE_PREEMPT_RCU build of the kernel is booted
with RCU_FANOUT set to a sufficiently small value that the
physical CPUs populate two or more leaf rcu_node structures.
o A task is preempted in an RCU read-side critical section
while running on a CPU corresponding to a given leaf rcu_node
structure.
o All CPUs corresponding to this same leaf rcu_node structure
record quiescent states for the current grace period.
o All of these same CPUs go offline (hence the need for enough
physical CPUs to populate more than one leaf rcu_node structure).
This causes the preempted task to be moved to the root rcu_node
structure.
At this point, there is nothing left to cause the quiescent
state to be propagated up the rcu_node tree, so the current
grace period never completes.
The simplest fix, especially after considering the deadlock
possibilities, is to detect this situation when the last CPU is
offlined, and to set that CPU's ->qsmask bit in its leaf
rcu_node structure. This will cause the next invocation of
force_quiescent_state() to end the grace period.
Without this fix, this hang can be triggered in an hour or so on
some machines with rcutorture and random CPU onlining/offlining.
With this fix, these same machines pass a full 10 hours of this
sort of abuse.
Signed-off-by: Paul E. McKenney <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: <20091015162614.GA19131@linux.vnet.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-10-15 18:26:14 +02:00
|
|
|
return 0;
|
2009-08-27 23:58:16 +02:00
|
|
|
}
|
|
|
|
|
2012-01-07 20:03:57 +01:00
|
|
|
#endif /* #ifdef CONFIG_HOTPLUG_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>
2009-08-22 22:56:52 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* 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>
2009-08-22 22:56:52 +02:00
|
|
|
* to check.
|
|
|
|
*/
|
2009-09-23 18:50:42 +02:00
|
|
|
static void rcu_preempt_check_callbacks(int 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>
2009-08-22 22:56:52 +02:00
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2012-01-06 23:11:30 +01:00
|
|
|
/*
|
|
|
|
* Queue an RCU callback for lazy invocation after a grace period.
|
|
|
|
* This will likely be later named something like "call_rcu_lazy()",
|
|
|
|
* but this change will require some way of tagging the lazy RCU
|
|
|
|
* callbacks in the list of pending callbacks. Until then, this
|
|
|
|
* function may only be called from __kfree_rcu().
|
|
|
|
*
|
|
|
|
* Because there is no preemptible RCU, we use RCU-sched instead.
|
|
|
|
*/
|
|
|
|
void kfree_call_rcu(struct rcu_head *head,
|
|
|
|
void (*func)(struct rcu_head *rcu))
|
|
|
|
{
|
2012-08-20 06:35:53 +02:00
|
|
|
__call_rcu(head, func, &rcu_sched_state, -1, 1);
|
2012-01-06 23:11:30 +01:00
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(kfree_call_rcu);
|
|
|
|
|
2009-10-14 19:15:56 +02:00
|
|
|
/*
|
|
|
|
* Wait for an rcu-preempt grace period, but make it happen quickly.
|
2011-03-02 22:15:15 +01:00
|
|
|
* But because preemptible RCU does not exist, map to rcu-sched.
|
2009-10-14 19:15:56 +02:00
|
|
|
*/
|
|
|
|
void synchronize_rcu_expedited(void)
|
|
|
|
{
|
|
|
|
synchronize_sched_expedited();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(synchronize_rcu_expedited);
|
|
|
|
|
2009-12-02 21:10:15 +01:00
|
|
|
#ifdef CONFIG_HOTPLUG_CPU
|
|
|
|
|
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, there is never any need to
|
2009-12-02 21:10:15 +01:00
|
|
|
* report on tasks preempted in RCU read-side critical sections during
|
|
|
|
* expedited RCU grace periods.
|
|
|
|
*/
|
2011-10-22 16:12:34 +02:00
|
|
|
static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp,
|
|
|
|
bool wake)
|
2009-12-02 21:10:15 +01:00
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #ifdef CONFIG_HOTPLUG_CPU */
|
|
|
|
|
2009-10-07 06:48:17 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, rcu_barrier() is just
|
2009-10-07 06:48:17 +02:00
|
|
|
* another name for rcu_barrier_sched().
|
|
|
|
*/
|
|
|
|
void rcu_barrier(void)
|
|
|
|
{
|
|
|
|
rcu_barrier_sched();
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(rcu_barrier);
|
|
|
|
|
2009-09-23 18:50:42 +02:00
|
|
|
/*
|
2011-03-02 22:15:15 +01:00
|
|
|
* Because preemptible RCU does not exist, it need not be initialized.
|
2009-09-23 18:50:42 +02:00
|
|
|
*/
|
|
|
|
static void __init __rcu_init_preempt(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>
2009-08-22 22:56:52 +02:00
|
|
|
#endif /* #else #ifdef CONFIG_TREE_PREEMPT_RCU */
|
2010-02-23 02:04:59 +01:00
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
#ifdef CONFIG_RCU_BOOST
|
|
|
|
|
|
|
|
#include "rtmutex_common.h"
|
|
|
|
|
2011-02-22 22:42:43 +01:00
|
|
|
#ifdef CONFIG_RCU_TRACE
|
|
|
|
|
|
|
|
static void rcu_initiate_boost_trace(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
if (list_empty(&rnp->blkd_tasks))
|
|
|
|
rnp->n_balk_blkd_tasks++;
|
|
|
|
else if (rnp->exp_tasks == NULL && rnp->gp_tasks == NULL)
|
|
|
|
rnp->n_balk_exp_gp_tasks++;
|
|
|
|
else if (rnp->gp_tasks != NULL && rnp->boost_tasks != NULL)
|
|
|
|
rnp->n_balk_boost_tasks++;
|
|
|
|
else if (rnp->gp_tasks != NULL && rnp->qsmask != 0)
|
|
|
|
rnp->n_balk_notblocked++;
|
|
|
|
else if (rnp->gp_tasks != NULL &&
|
2011-05-02 12:46:10 +02:00
|
|
|
ULONG_CMP_LT(jiffies, rnp->boost_time))
|
2011-02-22 22:42:43 +01:00
|
|
|
rnp->n_balk_notyet++;
|
|
|
|
else
|
|
|
|
rnp->n_balk_nos++;
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_TRACE */
|
|
|
|
|
|
|
|
static void rcu_initiate_boost_trace(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_TRACE */
|
|
|
|
|
2012-07-16 12:42:35 +02:00
|
|
|
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);
|
|
|
|
}
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
/*
|
|
|
|
* 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 rt_mutex mtx;
|
|
|
|
struct task_struct *t;
|
|
|
|
struct list_head *tb;
|
|
|
|
|
|
|
|
if (rnp->exp_tasks == NULL && rnp->boost_tasks == NULL)
|
|
|
|
return 0; /* Nothing left to boost. */
|
|
|
|
|
|
|
|
raw_spin_lock_irqsave(&rnp->lock, 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(&rnp->lock, 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.
|
|
|
|
*/
|
2011-02-22 22:42:43 +01:00
|
|
|
if (rnp->exp_tasks != NULL) {
|
2011-02-07 21:47:15 +01:00
|
|
|
tb = rnp->exp_tasks;
|
2011-02-22 22:42:43 +01:00
|
|
|
rnp->n_exp_boosts++;
|
|
|
|
} else {
|
2011-02-07 21:47:15 +01:00
|
|
|
tb = rnp->boost_tasks;
|
2011-02-22 22:42:43 +01:00
|
|
|
rnp->n_normal_boosts++;
|
|
|
|
}
|
|
|
|
rnp->n_tasks_boosted++;
|
2011-02-07 21:47:15 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* 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(&mtx, t);
|
|
|
|
t->rcu_boost_mutex = &mtx;
|
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
|
|
|
rt_mutex_lock(&mtx); /* Side effect: boosts task t's priority. */
|
|
|
|
rt_mutex_unlock(&mtx); /* Keep lockdep happy. */
|
|
|
|
|
2011-12-09 23:43:47 +01:00
|
|
|
return ACCESS_ONCE(rnp->exp_tasks) != NULL ||
|
|
|
|
ACCESS_ONCE(rnp->boost_tasks) != NULL;
|
2011-02-07 21:47:15 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Priority-boosting kthread. One per leaf rcu_node and one for the
|
|
|
|
* root rcu_node.
|
|
|
|
*/
|
|
|
|
static int rcu_boost_kthread(void *arg)
|
|
|
|
{
|
|
|
|
struct rcu_node *rnp = (struct rcu_node *)arg;
|
|
|
|
int spincnt = 0;
|
|
|
|
int more2boost;
|
|
|
|
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("Start boost kthread@init");
|
2011-02-07 21:47:15 +01:00
|
|
|
for (;;) {
|
2011-03-30 02:48:28 +02:00
|
|
|
rnp->boost_kthread_status = RCU_KTHREAD_WAITING;
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("End boost kthread@rcu_wait");
|
2011-05-21 01:06:29 +02:00
|
|
|
rcu_wait(rnp->boost_tasks || rnp->exp_tasks);
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("Start boost kthread@rcu_wait");
|
2011-03-30 02:48:28 +02:00
|
|
|
rnp->boost_kthread_status = RCU_KTHREAD_RUNNING;
|
2011-02-07 21:47:15 +01:00
|
|
|
more2boost = rcu_boost(rnp);
|
|
|
|
if (more2boost)
|
|
|
|
spincnt++;
|
|
|
|
else
|
|
|
|
spincnt = 0;
|
|
|
|
if (spincnt > 10) {
|
2012-07-16 12:42:35 +02:00
|
|
|
rnp->boost_kthread_status = RCU_KTHREAD_YIELDING;
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("End boost kthread@rcu_yield");
|
2012-07-16 12:42:35 +02:00
|
|
|
schedule_timeout_interruptible(2);
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("Start boost kthread@rcu_yield");
|
2011-02-07 21:47:15 +01:00
|
|
|
spincnt = 0;
|
|
|
|
}
|
|
|
|
}
|
2011-05-05 06:43:49 +02:00
|
|
|
/* NOTREACHED */
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("End boost kthread@notreached");
|
2011-02-07 21:47:15 +01:00
|
|
|
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.
|
|
|
|
*
|
2012-08-02 00:57:54 +02:00
|
|
|
* 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.
|
2011-02-07 21:47:15 +01:00
|
|
|
*/
|
2011-05-05 06:43:49 +02:00
|
|
|
static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
|
2011-02-07 21:47:15 +01:00
|
|
|
{
|
|
|
|
struct task_struct *t;
|
|
|
|
|
2011-02-22 22:42:43 +01:00
|
|
|
if (!rcu_preempt_blocked_readers_cgp(rnp) && rnp->exp_tasks == NULL) {
|
|
|
|
rnp->n_balk_exp_gp_tasks++;
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2011-02-07 21:47:15 +01:00
|
|
|
return;
|
2011-02-22 22:42:43 +01:00
|
|
|
}
|
2011-02-07 21:47:15 +01:00
|
|
|
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;
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2011-02-07 21:47:15 +01:00
|
|
|
t = rnp->boost_kthread_task;
|
2012-07-16 12:42:35 +02:00
|
|
|
if (t)
|
|
|
|
rcu_wake_cond(t, rnp->boost_kthread_status);
|
2011-05-05 06:43:49 +02:00
|
|
|
} else {
|
2011-02-22 22:42:43 +01:00
|
|
|
rcu_initiate_boost_trace(rnp);
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
|
|
|
}
|
2011-02-07 21:47:15 +01:00
|
|
|
}
|
|
|
|
|
2011-06-16 00:47:09 +02:00
|
|
|
/*
|
|
|
|
* 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);
|
2011-06-17 01:02:54 +02:00
|
|
|
if (__this_cpu_read(rcu_cpu_kthread_task) != NULL &&
|
2012-07-16 12:42:35 +02:00
|
|
|
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));
|
|
|
|
}
|
2011-06-16 00:47:09 +02:00
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
|
|
|
|
2011-11-30 00:57:13 +01:00
|
|
|
/*
|
|
|
|
* Is the current CPU running the RCU-callbacks kthread?
|
|
|
|
* Caller must have preemption disabled.
|
|
|
|
*/
|
|
|
|
static bool rcu_is_callbacks_kthread(void)
|
|
|
|
{
|
|
|
|
return __get_cpu_var(rcu_cpu_kthread_task) == current;
|
|
|
|
}
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
#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 __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp,
|
2012-07-16 12:42:35 +02:00
|
|
|
struct rcu_node *rnp)
|
2011-02-07 21:47:15 +01:00
|
|
|
{
|
2012-07-16 12:42:35 +02:00
|
|
|
int rnp_index = rnp - &rsp->node[0];
|
2011-02-07 21:47:15 +01:00
|
|
|
unsigned long flags;
|
|
|
|
struct sched_param sp;
|
|
|
|
struct task_struct *t;
|
|
|
|
|
|
|
|
if (&rcu_preempt_state != rsp)
|
|
|
|
return 0;
|
2012-07-16 12:42:35 +02:00
|
|
|
|
|
|
|
if (!rcu_scheduler_fully_active || rnp->qsmaskinit == 0)
|
|
|
|
return 0;
|
|
|
|
|
2011-06-16 00:47:09 +02:00
|
|
|
rsp->boost = 1;
|
2011-02-07 21:47:15 +01:00
|
|
|
if (rnp->boost_kthread_task != NULL)
|
|
|
|
return 0;
|
|
|
|
t = kthread_create(rcu_boost_kthread, (void *)rnp,
|
2011-08-19 20:39:11 +02:00
|
|
|
"rcub/%d", rnp_index);
|
2011-02-07 21:47:15 +01:00
|
|
|
if (IS_ERR(t))
|
|
|
|
return PTR_ERR(t);
|
|
|
|
raw_spin_lock_irqsave(&rnp->lock, flags);
|
|
|
|
rnp->boost_kthread_task = t;
|
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2011-08-19 20:39:11 +02:00
|
|
|
sp.sched_priority = RCU_BOOST_PRIO;
|
2011-02-07 21:47:15 +01:00
|
|
|
sched_setscheduler_nocheck(t, SCHED_FIFO, &sp);
|
2011-05-31 05:38:55 +02:00
|
|
|
wake_up_process(t); /* get to TASK_INTERRUPTIBLE quickly. */
|
2011-02-07 21:47:15 +01:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2011-06-16 17:26:32 +02:00
|
|
|
static void rcu_kthread_do_work(void)
|
|
|
|
{
|
|
|
|
rcu_do_batch(&rcu_sched_state, &__get_cpu_var(rcu_sched_data));
|
|
|
|
rcu_do_batch(&rcu_bh_state, &__get_cpu_var(rcu_bh_data));
|
|
|
|
rcu_preempt_do_callbacks();
|
|
|
|
}
|
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
static void rcu_cpu_kthread_setup(unsigned int cpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
|
|
|
struct sched_param sp;
|
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
sp.sched_priority = RCU_KTHREAD_PRIO;
|
|
|
|
sched_setscheduler_nocheck(current, SCHED_FIFO, &sp);
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
static void rcu_cpu_kthread_park(unsigned int cpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
2012-07-16 12:42:38 +02:00
|
|
|
per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU;
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
static int rcu_cpu_kthread_should_run(unsigned int cpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
2012-07-16 12:42:38 +02:00
|
|
|
return __get_cpu_var(rcu_cpu_has_work);
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Per-CPU kernel thread that invokes RCU callbacks. This replaces the
|
2011-06-21 10:29:39 +02:00
|
|
|
* RCU softirq used in flavors and configurations of RCU that do not
|
|
|
|
* support RCU priority boosting.
|
2011-06-16 17:26:32 +02:00
|
|
|
*/
|
2012-07-16 12:42:38 +02:00
|
|
|
static void rcu_cpu_kthread(unsigned int cpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
2012-07-16 12:42:38 +02:00
|
|
|
unsigned int *statusp = &__get_cpu_var(rcu_cpu_kthread_status);
|
|
|
|
char work, *workp = &__get_cpu_var(rcu_cpu_has_work);
|
|
|
|
int spincnt;
|
2011-06-16 17:26:32 +02:00
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
for (spincnt = 0; spincnt < 10; spincnt++) {
|
2011-06-22 07:43:26 +02:00
|
|
|
trace_rcu_utilization("Start CPU kthread@rcu_wait");
|
2011-06-16 17:26:32 +02:00
|
|
|
local_bh_disable();
|
|
|
|
*statusp = RCU_KTHREAD_RUNNING;
|
2012-07-16 12:42:38 +02:00
|
|
|
this_cpu_inc(rcu_cpu_kthread_loops);
|
|
|
|
local_irq_disable();
|
2011-06-16 17:26:32 +02:00
|
|
|
work = *workp;
|
|
|
|
*workp = 0;
|
2012-07-16 12:42:38 +02:00
|
|
|
local_irq_enable();
|
2011-06-16 17:26:32 +02:00
|
|
|
if (work)
|
|
|
|
rcu_kthread_do_work();
|
|
|
|
local_bh_enable();
|
2012-07-16 12:42:38 +02:00
|
|
|
if (*workp == 0) {
|
|
|
|
trace_rcu_utilization("End CPU kthread@rcu_wait");
|
|
|
|
*statusp = RCU_KTHREAD_WAITING;
|
|
|
|
return;
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
}
|
2012-07-16 12:42:38 +02:00
|
|
|
*statusp = RCU_KTHREAD_YIELDING;
|
|
|
|
trace_rcu_utilization("Start CPU kthread@rcu_yield");
|
|
|
|
schedule_timeout_interruptible(2);
|
|
|
|
trace_rcu_utilization("End CPU kthread@rcu_yield");
|
|
|
|
*statusp = RCU_KTHREAD_WAITING;
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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.
|
|
|
|
*/
|
2012-07-16 12:42:35 +02:00
|
|
|
static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
2012-07-16 12:42:35 +02:00
|
|
|
struct task_struct *t = rnp->boost_kthread_task;
|
|
|
|
unsigned long mask = rnp->qsmaskinit;
|
2011-06-16 17:26:32 +02:00
|
|
|
cpumask_var_t cm;
|
|
|
|
int cpu;
|
|
|
|
|
2012-07-16 12:42:35 +02:00
|
|
|
if (!t)
|
2011-06-16 17:26:32 +02:00
|
|
|
return;
|
2012-07-16 12:42:35 +02:00
|
|
|
if (!zalloc_cpumask_var(&cm, GFP_KERNEL))
|
2011-06-16 17:26:32 +02:00
|
|
|
return;
|
|
|
|
for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask >>= 1)
|
|
|
|
if ((mask & 0x1) && cpu != outgoingcpu)
|
|
|
|
cpumask_set_cpu(cpu, cm);
|
|
|
|
if (cpumask_weight(cm) == 0) {
|
|
|
|
cpumask_setall(cm);
|
|
|
|
for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++)
|
|
|
|
cpumask_clear_cpu(cpu, cm);
|
|
|
|
WARN_ON_ONCE(cpumask_weight(cm) == 0);
|
|
|
|
}
|
2012-07-16 12:42:35 +02:00
|
|
|
set_cpus_allowed_ptr(t, cm);
|
2011-06-16 17:26:32 +02:00
|
|
|
free_cpumask_var(cm);
|
|
|
|
}
|
|
|
|
|
2012-07-16 12:42:38 +02:00
|
|
|
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,
|
|
|
|
};
|
2011-06-16 17:26:32 +02:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Spawn all kthreads -- called as soon as the scheduler is running.
|
|
|
|
*/
|
|
|
|
static int __init rcu_spawn_kthreads(void)
|
|
|
|
{
|
|
|
|
struct rcu_node *rnp;
|
2012-07-16 12:42:35 +02:00
|
|
|
int cpu;
|
2011-06-16 17:26:32 +02:00
|
|
|
|
2011-07-11 00:57:35 +02:00
|
|
|
rcu_scheduler_fully_active = 1;
|
2012-07-16 12:42:38 +02:00
|
|
|
for_each_possible_cpu(cpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
per_cpu(rcu_cpu_has_work, cpu) = 0;
|
2012-07-16 12:42:38 +02:00
|
|
|
BUG_ON(smpboot_register_percpu_thread(&rcu_cpu_thread_spec));
|
2011-06-16 17:26:32 +02:00
|
|
|
rnp = rcu_get_root(rcu_state);
|
2012-07-16 12:42:35 +02:00
|
|
|
(void)rcu_spawn_one_boost_kthread(rcu_state, rnp);
|
2011-06-16 17:26:32 +02:00
|
|
|
if (NUM_RCU_NODES > 1) {
|
|
|
|
rcu_for_each_leaf_node(rcu_state, rnp)
|
2012-07-16 12:42:35 +02:00
|
|
|
(void)rcu_spawn_one_boost_kthread(rcu_state, rnp);
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
early_initcall(rcu_spawn_kthreads);
|
|
|
|
|
|
|
|
static void __cpuinit rcu_prepare_kthreads(int cpu)
|
|
|
|
{
|
|
|
|
struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, cpu);
|
|
|
|
struct rcu_node *rnp = rdp->mynode;
|
|
|
|
|
|
|
|
/* Fire up the incoming CPU's kthread and leaf rcu_node kthread. */
|
2012-07-16 12:42:38 +02:00
|
|
|
if (rcu_scheduler_fully_active)
|
2012-07-16 12:42:35 +02:00
|
|
|
(void)rcu_spawn_one_boost_kthread(rcu_state, rnp);
|
2011-06-16 17:26:32 +02:00
|
|
|
}
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
#else /* #ifdef CONFIG_RCU_BOOST */
|
|
|
|
|
2011-05-05 06:43:49 +02:00
|
|
|
static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags)
|
2011-02-07 21:47:15 +01:00
|
|
|
{
|
2011-05-05 06:43:49 +02:00
|
|
|
raw_spin_unlock_irqrestore(&rnp->lock, flags);
|
2011-02-07 21:47:15 +01:00
|
|
|
}
|
|
|
|
|
2011-06-16 00:47:09 +02:00
|
|
|
static void invoke_rcu_callbacks_kthread(void)
|
2011-02-07 21:47:15 +01:00
|
|
|
{
|
2011-06-16 00:47:09 +02:00
|
|
|
WARN_ON_ONCE(1);
|
2011-02-07 21:47:15 +01:00
|
|
|
}
|
|
|
|
|
2011-11-30 00:57:13 +01:00
|
|
|
static bool rcu_is_callbacks_kthread(void)
|
|
|
|
{
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
static void rcu_preempt_boost_start_gp(struct rcu_node *rnp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2012-07-16 12:42:35 +02:00
|
|
|
static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu)
|
2011-06-16 17:26:32 +02:00
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2011-07-11 00:57:35 +02:00
|
|
|
static int __init rcu_scheduler_really_started(void)
|
|
|
|
{
|
|
|
|
rcu_scheduler_fully_active = 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
early_initcall(rcu_scheduler_really_started);
|
|
|
|
|
2011-06-16 17:26:32 +02:00
|
|
|
static void __cpuinit rcu_prepare_kthreads(int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2011-02-07 21:47:15 +01:00
|
|
|
#endif /* #else #ifdef CONFIG_RCU_BOOST */
|
|
|
|
|
2010-02-23 02:04:59 +01:00
|
|
|
#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.
|
|
|
|
*
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
* Because we not have RCU_FAST_NO_HZ, just check whether this CPU needs
|
|
|
|
* any flavor of RCU.
|
2010-02-23 02:04:59 +01:00
|
|
|
*/
|
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 01:41:44 +02:00
|
|
|
int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies)
|
2010-02-23 02:04:59 +01:00
|
|
|
{
|
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 01:41:44 +02:00
|
|
|
*delta_jiffies = ULONG_MAX;
|
2011-11-02 14:54:54 +01:00
|
|
|
return rcu_cpu_has_callbacks(cpu);
|
|
|
|
}
|
|
|
|
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
/*
|
|
|
|
* Because we do not have RCU_FAST_NO_HZ, don't bother initializing for it.
|
|
|
|
*/
|
|
|
|
static void rcu_prepare_for_idle_init(int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Because we do not have RCU_FAST_NO_HZ, don't bother cleaning up
|
|
|
|
* after it.
|
|
|
|
*/
|
|
|
|
static void rcu_cleanup_after_idle(int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2011-11-02 14:54:54 +01:00
|
|
|
/*
|
2012-01-16 22:29:10 +01:00
|
|
|
* Do the idle-entry grace-period work, which, because CONFIG_RCU_FAST_NO_HZ=n,
|
2011-11-02 14:54:54 +01:00
|
|
|
* is nothing.
|
|
|
|
*/
|
|
|
|
static void rcu_prepare_for_idle(int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2012-02-28 20:02:21 +01:00
|
|
|
/*
|
|
|
|
* 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)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
2010-02-23 02:04:59 +01:00
|
|
|
#else /* #if !defined(CONFIG_RCU_FAST_NO_HZ) */
|
|
|
|
|
2011-12-01 00:41:14 +01:00
|
|
|
/*
|
|
|
|
* 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_FLUSHES gives the maximum number of times that we will attempt
|
|
|
|
* to satisfy RCU. Beyond this point, it is better to incur a periodic
|
|
|
|
* scheduling-clock interrupt than to loop through the state machine
|
|
|
|
* at full power.
|
|
|
|
* RCU_IDLE_OPT_FLUSHES gives the number of RCU_IDLE_FLUSHES that are
|
|
|
|
* optional if RCU does not need anything immediately from this
|
|
|
|
* CPU, even if this CPU still has RCU callbacks queued. The first
|
|
|
|
* times through the state machine are mandatory: we need to give
|
|
|
|
* the state machine a chance to communicate a quiescent state
|
|
|
|
* to the RCU core.
|
|
|
|
* 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!
|
2012-01-10 23:13:24 +01:00
|
|
|
* 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.
|
2011-12-01 00:41:14 +01:00
|
|
|
*
|
|
|
|
* 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_FLUSHES 5 /* Number of dyntick-idle tries. */
|
|
|
|
#define RCU_IDLE_OPT_FLUSHES 3 /* Optional dyntick-idle tries. */
|
2012-06-05 05:45:10 +02:00
|
|
|
#define RCU_IDLE_GP_DELAY 4 /* Roughly one grace period. */
|
2012-01-10 23:13:24 +01:00
|
|
|
#define RCU_IDLE_LAZY_GP_DELAY (6 * HZ) /* Roughly six seconds. */
|
2011-12-01 00:41:14 +01:00
|
|
|
|
2012-06-24 19:15:02 +02:00
|
|
|
extern int tick_nohz_enabled;
|
|
|
|
|
2012-01-06 23:11:30 +01:00
|
|
|
/*
|
|
|
|
* Does the specified flavor of RCU have non-lazy callbacks pending on
|
|
|
|
* the specified CPU? Both RCU flavor and CPU are specified by the
|
|
|
|
* rcu_data structure.
|
|
|
|
*/
|
|
|
|
static bool __rcu_cpu_has_nonlazy_callbacks(struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
return rdp->qlen != rdp->qlen_lazy;
|
|
|
|
}
|
|
|
|
|
|
|
|
#ifdef CONFIG_TREE_PREEMPT_RCU
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Are there non-lazy RCU-preempt callbacks? (There cannot be if there
|
|
|
|
* is no RCU-preempt in the kernel.)
|
|
|
|
*/
|
|
|
|
static bool rcu_preempt_cpu_has_nonlazy_callbacks(int cpu)
|
|
|
|
{
|
|
|
|
struct rcu_data *rdp = &per_cpu(rcu_preempt_data, cpu);
|
|
|
|
|
|
|
|
return __rcu_cpu_has_nonlazy_callbacks(rdp);
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_TREE_PREEMPT_RCU */
|
|
|
|
|
|
|
|
static bool rcu_preempt_cpu_has_nonlazy_callbacks(int cpu)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* else #ifdef CONFIG_TREE_PREEMPT_RCU */
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Does any flavor of RCU have non-lazy callbacks on the specified CPU?
|
|
|
|
*/
|
|
|
|
static bool rcu_cpu_has_nonlazy_callbacks(int cpu)
|
|
|
|
{
|
|
|
|
return __rcu_cpu_has_nonlazy_callbacks(&per_cpu(rcu_sched_data, cpu)) ||
|
|
|
|
__rcu_cpu_has_nonlazy_callbacks(&per_cpu(rcu_bh_data, cpu)) ||
|
|
|
|
rcu_preempt_cpu_has_nonlazy_callbacks(cpu);
|
|
|
|
}
|
|
|
|
|
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 01:41:44 +02:00
|
|
|
/*
|
|
|
|
* Allow the CPU to enter dyntick-idle mode if either: (1) There are no
|
|
|
|
* callbacks on this CPU, (2) this CPU has not yet attempted to enter
|
|
|
|
* dyntick-idle mode, or (3) this CPU is in the process of attempting to
|
|
|
|
* enter dyntick-idle mode. Otherwise, if we have recently tried and failed
|
|
|
|
* to enter dyntick-idle mode, we refuse to try to enter it. After all,
|
|
|
|
* it is better to incur scheduling-clock interrupts than to spin
|
|
|
|
* continuously for the same time duration!
|
|
|
|
*
|
|
|
|
* The delta_jiffies argument is used to store the time when RCU is
|
|
|
|
* going to need the CPU again if it still has callbacks. The reason
|
|
|
|
* for this is that rcu_prepare_for_idle() might need to post a timer,
|
|
|
|
* but if so, it will do so after tick_nohz_stop_sched_tick() has set
|
|
|
|
* the wakeup time for this CPU. This means that RCU's timer can be
|
|
|
|
* delayed until the wakeup time, which defeats the purpose of posting
|
|
|
|
* a timer.
|
|
|
|
*/
|
|
|
|
int rcu_needs_cpu(int cpu, unsigned long *delta_jiffies)
|
|
|
|
{
|
|
|
|
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
|
|
|
|
|
|
|
|
/* Flag a new idle sojourn to the idle-entry state machine. */
|
|
|
|
rdtp->idle_first_pass = 1;
|
|
|
|
/* If no callbacks, RCU doesn't need the CPU. */
|
|
|
|
if (!rcu_cpu_has_callbacks(cpu)) {
|
|
|
|
*delta_jiffies = ULONG_MAX;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
if (rdtp->dyntick_holdoff == jiffies) {
|
|
|
|
/* RCU recently tried and failed, so don't try again. */
|
|
|
|
*delta_jiffies = 1;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
/* Set up for the possibility that RCU will post a timer. */
|
2012-06-05 05:45:10 +02:00
|
|
|
if (rcu_cpu_has_nonlazy_callbacks(cpu)) {
|
|
|
|
*delta_jiffies = round_up(RCU_IDLE_GP_DELAY + jiffies,
|
|
|
|
RCU_IDLE_GP_DELAY) - jiffies;
|
|
|
|
} else {
|
|
|
|
*delta_jiffies = jiffies + RCU_IDLE_LAZY_GP_DELAY;
|
|
|
|
*delta_jiffies = round_jiffies(*delta_jiffies) - jiffies;
|
|
|
|
}
|
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 01:41:44 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-04-30 23:16:19 +02:00
|
|
|
/*
|
|
|
|
* Handler for smp_call_function_single(). The only point of this
|
|
|
|
* handler is to wake the CPU up, so the handler does only tracing.
|
|
|
|
*/
|
|
|
|
void rcu_idle_demigrate(void *unused)
|
|
|
|
{
|
|
|
|
trace_rcu_prep_idle("Demigrate");
|
|
|
|
}
|
|
|
|
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
/*
|
|
|
|
* Timer handler used to force CPU to start pushing its remaining RCU
|
|
|
|
* callbacks in the case where it entered dyntick-idle mode with callbacks
|
|
|
|
* pending. The hander doesn't really need to do anything because the
|
|
|
|
* real work is done upon re-entry to idle, or by the next scheduling-clock
|
|
|
|
* interrupt should idle not be re-entered.
|
2012-04-30 23:16:19 +02:00
|
|
|
*
|
|
|
|
* One special case: the timer gets migrated without awakening the CPU
|
|
|
|
* on which the timer was scheduled on. In this case, we must wake up
|
|
|
|
* that CPU. We do so with smp_call_function_single().
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
*/
|
2012-04-30 23:16:19 +02:00
|
|
|
static void rcu_idle_gp_timer_func(unsigned long cpu_in)
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
{
|
2012-04-30 23:16:19 +02:00
|
|
|
int cpu = (int)cpu_in;
|
|
|
|
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
trace_rcu_prep_idle("Timer");
|
2012-04-30 23:16:19 +02:00
|
|
|
if (cpu != smp_processor_id())
|
|
|
|
smp_call_function_single(cpu, rcu_idle_demigrate, NULL, 0);
|
|
|
|
else
|
|
|
|
WARN_ON_ONCE(1); /* Getting here can hang the system... */
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Initialize the timer used to pull CPUs out of dyntick-idle mode.
|
|
|
|
*/
|
|
|
|
static void rcu_prepare_for_idle_init(int cpu)
|
|
|
|
{
|
2012-05-09 21:07:05 +02:00
|
|
|
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
|
|
|
|
|
|
|
|
rdtp->dyntick_holdoff = jiffies - 1;
|
|
|
|
setup_timer(&rdtp->idle_gp_timer, rcu_idle_gp_timer_func, cpu);
|
|
|
|
rdtp->idle_gp_timer_expires = jiffies - 1;
|
|
|
|
rdtp->idle_first_pass = 1;
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Clean up for exit from idle. Because we are exiting from idle, there
|
2012-05-09 21:07:05 +02:00
|
|
|
* is no longer any point to ->idle_gp_timer, so cancel it. This will
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
* do nothing if this timer is not active, so just cancel it unconditionally.
|
|
|
|
*/
|
|
|
|
static void rcu_cleanup_after_idle(int cpu)
|
|
|
|
{
|
2012-05-09 21:07:05 +02:00
|
|
|
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
|
|
|
|
|
|
|
|
del_timer(&rdtp->idle_gp_timer);
|
2012-02-24 00:58:29 +01:00
|
|
|
trace_rcu_prep_idle("Cleanup after idle");
|
2012-06-24 19:15:02 +02:00
|
|
|
rdtp->tick_nohz_enabled_snap = ACCESS_ONCE(tick_nohz_enabled);
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
}
|
|
|
|
|
2011-11-02 14:54:54 +01:00
|
|
|
/*
|
|
|
|
* Check to see if any RCU-related work can be done by the current CPU,
|
|
|
|
* and if so, schedule a softirq to get it done. This function is part
|
|
|
|
* of the RCU implementation; it is -not- an exported member of the RCU API.
|
2010-02-23 02:04:59 +01:00
|
|
|
*
|
2011-11-02 14:54:54 +01:00
|
|
|
* The idea is for the current CPU to clear out all work required by the
|
|
|
|
* RCU core for the current grace period, so that this CPU can be permitted
|
|
|
|
* to enter dyntick-idle mode. In some cases, it will need to be awakened
|
|
|
|
* at the end of the grace period by whatever CPU ends the grace period.
|
|
|
|
* This allows CPUs to go dyntick-idle more quickly, and to reduce the
|
|
|
|
* number of wakeups by a modest integer factor.
|
2010-02-27 01:38:56 +01:00
|
|
|
*
|
|
|
|
* Because it is not legal to invoke rcu_process_callbacks() with irqs
|
|
|
|
* disabled, we do one pass of force_quiescent_state(), then do a
|
2011-06-16 00:47:09 +02:00
|
|
|
* invoke_rcu_core() to cause rcu_process_callbacks() to be invoked
|
2012-05-09 21:07:05 +02:00
|
|
|
* later. The ->dyntick_drain field controls the sequencing.
|
2011-11-02 14:54:54 +01:00
|
|
|
*
|
|
|
|
* The caller must have disabled interrupts.
|
2010-02-23 02:04:59 +01:00
|
|
|
*/
|
2011-11-02 14:54:54 +01:00
|
|
|
static void rcu_prepare_for_idle(int cpu)
|
2010-02-23 02:04:59 +01:00
|
|
|
{
|
2012-03-15 20:16:26 +01:00
|
|
|
struct timer_list *tp;
|
2012-05-09 21:07:05 +02:00
|
|
|
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
|
2012-06-24 19:15:02 +02:00
|
|
|
int tne;
|
|
|
|
|
|
|
|
/* Handle nohz enablement switches conservatively. */
|
|
|
|
tne = ACCESS_ONCE(tick_nohz_enabled);
|
|
|
|
if (tne != rdtp->tick_nohz_enabled_snap) {
|
|
|
|
if (rcu_cpu_has_callbacks(cpu))
|
|
|
|
invoke_rcu_core(); /* force nohz to see update. */
|
|
|
|
rdtp->tick_nohz_enabled_snap = tne;
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
if (!tne)
|
|
|
|
return;
|
2012-03-15 20:16:26 +01:00
|
|
|
|
2012-06-28 21:33:51 +02:00
|
|
|
/* Adaptive-tick mode, where usermode execution is idle to RCU. */
|
|
|
|
if (!is_idle_task(current)) {
|
|
|
|
rdtp->dyntick_holdoff = jiffies - 1;
|
|
|
|
if (rcu_cpu_has_nonlazy_callbacks(cpu)) {
|
|
|
|
trace_rcu_prep_idle("User dyntick with callbacks");
|
|
|
|
rdtp->idle_gp_timer_expires =
|
|
|
|
round_up(jiffies + RCU_IDLE_GP_DELAY,
|
|
|
|
RCU_IDLE_GP_DELAY);
|
|
|
|
} else if (rcu_cpu_has_callbacks(cpu)) {
|
|
|
|
rdtp->idle_gp_timer_expires =
|
|
|
|
round_jiffies(jiffies + RCU_IDLE_LAZY_GP_DELAY);
|
|
|
|
trace_rcu_prep_idle("User dyntick with lazy callbacks");
|
|
|
|
} else {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
tp = &rdtp->idle_gp_timer;
|
|
|
|
mod_timer_pinned(tp, rdtp->idle_gp_timer_expires);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2012-02-28 20:02:21 +01:00
|
|
|
/*
|
|
|
|
* If this is an idle re-entry, for example, due to use of
|
|
|
|
* RCU_NONIDLE() or the new idle-loop tracing API within the idle
|
|
|
|
* loop, then don't take any state-machine actions, unless the
|
|
|
|
* momentary exit from idle queued additional non-lazy callbacks.
|
2012-05-09 21:07:05 +02:00
|
|
|
* Instead, repost the ->idle_gp_timer if this CPU has callbacks
|
2012-02-28 20:02:21 +01:00
|
|
|
* pending.
|
|
|
|
*/
|
2012-05-09 21:07:05 +02:00
|
|
|
if (!rdtp->idle_first_pass &&
|
|
|
|
(rdtp->nonlazy_posted == rdtp->nonlazy_posted_snap)) {
|
2012-03-15 20:16:26 +01:00
|
|
|
if (rcu_cpu_has_callbacks(cpu)) {
|
2012-05-09 21:07:05 +02:00
|
|
|
tp = &rdtp->idle_gp_timer;
|
|
|
|
mod_timer_pinned(tp, rdtp->idle_gp_timer_expires);
|
2012-03-15 20:16:26 +01:00
|
|
|
}
|
2012-02-28 20:02:21 +01:00
|
|
|
return;
|
|
|
|
}
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->idle_first_pass = 0;
|
|
|
|
rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted - 1;
|
2012-02-28 20:02:21 +01:00
|
|
|
|
2011-11-23 02:07:11 +01:00
|
|
|
/*
|
2011-11-23 05:43:02 +01:00
|
|
|
* If there are no callbacks on this CPU, enter dyntick-idle mode.
|
|
|
|
* Also reset state to avoid prejudicing later attempts.
|
2011-11-23 02:07:11 +01:00
|
|
|
*/
|
2011-11-02 14:54:54 +01:00
|
|
|
if (!rcu_cpu_has_callbacks(cpu)) {
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->dyntick_holdoff = jiffies - 1;
|
|
|
|
rdtp->dyntick_drain = 0;
|
2011-11-22 23:58:03 +01:00
|
|
|
trace_rcu_prep_idle("No callbacks");
|
2011-11-02 14:54:54 +01:00
|
|
|
return;
|
2010-04-26 06:04:29 +02:00
|
|
|
}
|
2011-11-23 02:07:11 +01:00
|
|
|
|
|
|
|
/*
|
|
|
|
* If in holdoff mode, just return. We will presumably have
|
|
|
|
* refrained from disabling the scheduling-clock tick.
|
|
|
|
*/
|
2012-05-09 21:07:05 +02:00
|
|
|
if (rdtp->dyntick_holdoff == jiffies) {
|
2011-11-22 23:58:03 +01:00
|
|
|
trace_rcu_prep_idle("In holdoff");
|
2011-11-02 14:54:54 +01:00
|
|
|
return;
|
2011-11-22 23:58:03 +01:00
|
|
|
}
|
2010-02-27 01:38:56 +01:00
|
|
|
|
2012-05-09 21:07:05 +02:00
|
|
|
/* Check and update the ->dyntick_drain sequencing. */
|
|
|
|
if (rdtp->dyntick_drain <= 0) {
|
2010-02-27 01:38:56 +01:00
|
|
|
/* First time through, initialize the counter. */
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->dyntick_drain = RCU_IDLE_FLUSHES;
|
|
|
|
} else if (rdtp->dyntick_drain <= RCU_IDLE_OPT_FLUSHES &&
|
2012-02-14 19:12:54 +01:00
|
|
|
!rcu_pending(cpu) &&
|
|
|
|
!local_softirq_pending()) {
|
rcu: Permit dyntick-idle with callbacks pending
The current implementation of RCU_FAST_NO_HZ prevents CPUs from entering
dyntick-idle state if they have RCU callbacks pending. Unfortunately,
this has the side-effect of often preventing them from entering this
state, especially if at least one other CPU is not in dyntick-idle state.
However, the resulting per-tick wakeup is wasteful in many cases: if the
CPU has already fully responded to the current RCU grace period, there
will be nothing for it to do until this grace period ends, which will
frequently take several jiffies.
This commit therefore permits a CPU that has done everything that the
current grace period has asked of it (rcu_pending() == 0) even if it
still as RCU callbacks pending. However, such a CPU posts a timer to
wake it up several jiffies later (6 jiffies, based on experience with
grace-period lengths). This wakeup is required to handle situations
that can result in all CPUs being in dyntick-idle mode, thus failing
to ever complete the current grace period. If a CPU wakes up before
the timer goes off, then it cancels that timer, thus avoiding spurious
wakeups.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
2011-11-28 21:28:34 +01:00
|
|
|
/* Can we go dyntick-idle despite still having callbacks? */
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->dyntick_drain = 0;
|
|
|
|
rdtp->dyntick_holdoff = jiffies;
|
2012-05-06 04:10:35 +02:00
|
|
|
if (rcu_cpu_has_nonlazy_callbacks(cpu)) {
|
|
|
|
trace_rcu_prep_idle("Dyntick with callbacks");
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->idle_gp_timer_expires =
|
2012-06-05 05:45:10 +02:00
|
|
|
round_up(jiffies + RCU_IDLE_GP_DELAY,
|
|
|
|
RCU_IDLE_GP_DELAY);
|
2012-05-06 04:10:35 +02:00
|
|
|
} else {
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->idle_gp_timer_expires =
|
2012-06-05 05:45:10 +02:00
|
|
|
round_jiffies(jiffies + RCU_IDLE_LAZY_GP_DELAY);
|
2012-05-06 04:10:35 +02:00
|
|
|
trace_rcu_prep_idle("Dyntick with lazy callbacks");
|
|
|
|
}
|
2012-05-09 21:07:05 +02:00
|
|
|
tp = &rdtp->idle_gp_timer;
|
|
|
|
mod_timer_pinned(tp, rdtp->idle_gp_timer_expires);
|
|
|
|
rdtp->nonlazy_posted_snap = rdtp->nonlazy_posted;
|
2011-12-01 00:41:14 +01:00
|
|
|
return; /* Nothing more to do immediately. */
|
2012-05-09 21:07:05 +02:00
|
|
|
} else if (--(rdtp->dyntick_drain) <= 0) {
|
2010-02-27 01:38:56 +01:00
|
|
|
/* We have hit the limit, so time to give up. */
|
2012-05-09 21:07:05 +02:00
|
|
|
rdtp->dyntick_holdoff = jiffies;
|
2011-11-22 23:58:03 +01:00
|
|
|
trace_rcu_prep_idle("Begin holdoff");
|
2011-11-02 14:54:54 +01:00
|
|
|
invoke_rcu_core(); /* Force the CPU out of dyntick-idle. */
|
|
|
|
return;
|
2010-02-27 01:38:56 +01:00
|
|
|
}
|
|
|
|
|
2011-11-02 14:54:54 +01:00
|
|
|
/*
|
|
|
|
* Do one step of pushing the remaining RCU callbacks through
|
|
|
|
* the RCU core state machine.
|
|
|
|
*/
|
|
|
|
#ifdef CONFIG_TREE_PREEMPT_RCU
|
|
|
|
if (per_cpu(rcu_preempt_data, cpu).nxtlist) {
|
|
|
|
rcu_preempt_qs(cpu);
|
2012-06-23 02:06:26 +02:00
|
|
|
force_quiescent_state(&rcu_preempt_state);
|
2011-11-02 14:54:54 +01:00
|
|
|
}
|
|
|
|
#endif /* #ifdef CONFIG_TREE_PREEMPT_RCU */
|
2010-02-27 01:38:56 +01:00
|
|
|
if (per_cpu(rcu_sched_data, cpu).nxtlist) {
|
|
|
|
rcu_sched_qs(cpu);
|
2012-06-23 02:06:26 +02:00
|
|
|
force_quiescent_state(&rcu_sched_state);
|
2010-02-27 01:38:56 +01:00
|
|
|
}
|
|
|
|
if (per_cpu(rcu_bh_data, cpu).nxtlist) {
|
|
|
|
rcu_bh_qs(cpu);
|
2012-06-23 02:06:26 +02:00
|
|
|
force_quiescent_state(&rcu_bh_state);
|
2010-02-23 02:04:59 +01:00
|
|
|
}
|
|
|
|
|
2011-11-22 23:58:03 +01:00
|
|
|
/*
|
|
|
|
* If RCU callbacks are still pending, RCU still needs this CPU.
|
|
|
|
* So try forcing the callbacks through the grace period.
|
|
|
|
*/
|
2011-11-23 06:08:13 +01:00
|
|
|
if (rcu_cpu_has_callbacks(cpu)) {
|
2011-11-22 23:58:03 +01:00
|
|
|
trace_rcu_prep_idle("More callbacks");
|
2011-06-16 00:47:09 +02:00
|
|
|
invoke_rcu_core();
|
2012-06-28 17:08:25 +02:00
|
|
|
} else {
|
2011-11-22 23:58:03 +01:00
|
|
|
trace_rcu_prep_idle("Callbacks drained");
|
2012-06-28 17:08:25 +02:00
|
|
|
}
|
2010-02-23 02:04:59 +01:00
|
|
|
}
|
|
|
|
|
2012-02-28 20:02:21 +01:00
|
|
|
/*
|
2012-05-04 00:38:10 +02:00
|
|
|
* 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().
|
2012-02-28 20:02:21 +01:00
|
|
|
*/
|
|
|
|
static void rcu_idle_count_callbacks_posted(void)
|
|
|
|
{
|
2012-05-09 21:07:05 +02:00
|
|
|
__this_cpu_add(rcu_dynticks.nonlazy_posted, 1);
|
2012-02-28 20:02:21 +01:00
|
|
|
}
|
|
|
|
|
2012-06-12 02:39:43 +02:00
|
|
|
/*
|
|
|
|
* 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 = __this_cpu_ptr(rsp->rda);
|
|
|
|
if (rdp->qlen_lazy != 0) {
|
|
|
|
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);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Prevent premature wakeup: ensure that all increments happen
|
|
|
|
* before there is a chance of the counter reaching zero.
|
|
|
|
*/
|
|
|
|
atomic_set(&oom_callback_count, 1);
|
|
|
|
|
|
|
|
get_online_cpus();
|
|
|
|
for_each_online_cpu(cpu) {
|
|
|
|
smp_call_function_single(cpu, rcu_oom_notify_cpu, NULL, 1);
|
|
|
|
cond_resched();
|
|
|
|
}
|
|
|
|
put_online_cpus();
|
|
|
|
|
|
|
|
/* 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);
|
|
|
|
|
2010-02-23 02:04:59 +01:00
|
|
|
#endif /* #else #if !defined(CONFIG_RCU_FAST_NO_HZ) */
|
2012-01-16 22:29:10 +01:00
|
|
|
|
|
|
|
#ifdef CONFIG_RCU_CPU_STALL_INFO
|
|
|
|
|
|
|
|
#ifdef CONFIG_RCU_FAST_NO_HZ
|
|
|
|
|
|
|
|
static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
|
|
|
|
{
|
2012-05-09 21:07:05 +02:00
|
|
|
struct rcu_dynticks *rdtp = &per_cpu(rcu_dynticks, cpu);
|
|
|
|
struct timer_list *tltp = &rdtp->idle_gp_timer;
|
2012-09-21 19:41:50 +02:00
|
|
|
char c;
|
2012-01-16 22:29:10 +01:00
|
|
|
|
2012-09-21 19:41:50 +02:00
|
|
|
c = rdtp->dyntick_holdoff == jiffies ? 'H' : '.';
|
|
|
|
if (timer_pending(tltp))
|
|
|
|
sprintf(cp, "drain=%d %c timer=%lu",
|
|
|
|
rdtp->dyntick_drain, c, tltp->expires - jiffies);
|
|
|
|
else
|
|
|
|
sprintf(cp, "drain=%d %c timer not pending",
|
|
|
|
rdtp->dyntick_drain, c);
|
2012-01-16 22:29:10 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_FAST_NO_HZ */
|
|
|
|
|
|
|
|
static void print_cpu_stall_fast_no_hz(char *cp, int cpu)
|
|
|
|
{
|
2012-06-19 10:43:16 +02:00
|
|
|
*cp = '\0';
|
2012-01-16 22:29:10 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_FAST_NO_HZ */
|
|
|
|
|
|
|
|
/* Initiate the stall-info list. */
|
|
|
|
static void print_cpu_stall_info_begin(void)
|
|
|
|
{
|
|
|
|
printk(KERN_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);
|
|
|
|
printk(KERN_ERR "\t%d: (%lu %s) idle=%03x/%llx/%d %s\n",
|
|
|
|
cpu, ticks_value, ticks_title,
|
|
|
|
atomic_read(&rdtp->dynticks) & 0xfff,
|
|
|
|
rdtp->dynticks_nesting, rdtp->dynticks_nmi_nesting,
|
|
|
|
fast_no_hz);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Terminate the stall-info list. */
|
|
|
|
static void print_cpu_stall_info_end(void)
|
|
|
|
{
|
|
|
|
printk(KERN_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;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Increment ->ticks_this_gp for all flavors of RCU. */
|
|
|
|
static void increment_cpu_stall_ticks(void)
|
|
|
|
{
|
2012-08-10 22:55:03 +02:00
|
|
|
struct rcu_state *rsp;
|
|
|
|
|
|
|
|
for_each_rcu_flavor(rsp)
|
|
|
|
__this_cpu_ptr(rsp->rda)->ticks_this_gp++;
|
2012-01-16 22:29:10 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_CPU_STALL_INFO */
|
|
|
|
|
|
|
|
static void print_cpu_stall_info_begin(void)
|
|
|
|
{
|
|
|
|
printk(KERN_CONT " {");
|
|
|
|
}
|
|
|
|
|
|
|
|
static void print_cpu_stall_info(struct rcu_state *rsp, int cpu)
|
|
|
|
{
|
|
|
|
printk(KERN_CONT " %d", cpu);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void print_cpu_stall_info_end(void)
|
|
|
|
{
|
|
|
|
printk(KERN_CONT "} ");
|
|
|
|
}
|
|
|
|
|
|
|
|
static void zero_cpu_stall_ticks(struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static void increment_cpu_stall_ticks(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_CPU_STALL_INFO */
|
2012-08-20 06:35:53 +02:00
|
|
|
|
|
|
|
#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);
|
|
|
|
|
|
|
|
/* Is the specified CPU a no-CPUs CPU? */
|
|
|
|
static bool is_nocb_cpu(int cpu)
|
|
|
|
{
|
|
|
|
if (have_rcu_nocb_mask)
|
|
|
|
return cpumask_test_cpu(cpu, rcu_nocb_mask);
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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,
|
|
|
|
int rhcount, int rhcount_lazy)
|
|
|
|
{
|
|
|
|
int len;
|
|
|
|
struct rcu_head **old_rhpp;
|
|
|
|
struct task_struct *t;
|
|
|
|
|
|
|
|
/* Enqueue the callback on the nocb list and update counts. */
|
|
|
|
old_rhpp = xchg(&rdp->nocb_tail, rhtp);
|
|
|
|
ACCESS_ONCE(*old_rhpp) = rhp;
|
|
|
|
atomic_long_add(rhcount, &rdp->nocb_q_count);
|
|
|
|
atomic_long_add(rhcount_lazy, &rdp->nocb_q_count_lazy);
|
|
|
|
|
|
|
|
/* If we are not being polled and there is a kthread, awaken it ... */
|
|
|
|
t = ACCESS_ONCE(rdp->nocb_kthread);
|
|
|
|
if (rcu_nocb_poll | !t)
|
|
|
|
return;
|
|
|
|
len = atomic_long_read(&rdp->nocb_q_count);
|
|
|
|
if (old_rhpp == &rdp->nocb_head) {
|
|
|
|
wake_up(&rdp->nocb_wq); /* ... only if queue was empty ... */
|
|
|
|
rdp->qlen_last_fqs_check = 0;
|
|
|
|
} else if (len > rdp->qlen_last_fqs_check + qhimark) {
|
|
|
|
wake_up_process(t); /* ... or if many callbacks queued. */
|
|
|
|
rdp->qlen_last_fqs_check = LONG_MAX / 2;
|
|
|
|
}
|
|
|
|
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,
|
|
|
|
bool lazy)
|
|
|
|
{
|
|
|
|
|
|
|
|
if (!is_nocb_cpu(rdp->cpu))
|
|
|
|
return 0;
|
|
|
|
__call_rcu_nocb_enqueue(rdp, rhp, &rhp->next, 1, lazy);
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* 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_state *rsp,
|
|
|
|
struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
long ql = rsp->qlen;
|
|
|
|
long qll = rsp->qlen_lazy;
|
|
|
|
|
|
|
|
/* If this is not a no-CBs CPU, tell the caller to do it the old way. */
|
|
|
|
if (!is_nocb_cpu(smp_processor_id()))
|
|
|
|
return 0;
|
|
|
|
rsp->qlen = 0;
|
|
|
|
rsp->qlen_lazy = 0;
|
|
|
|
|
|
|
|
/* First, enqueue the donelist, if any. This preserves CB ordering. */
|
|
|
|
if (rsp->orphan_donelist != NULL) {
|
|
|
|
__call_rcu_nocb_enqueue(rdp, rsp->orphan_donelist,
|
|
|
|
rsp->orphan_donetail, ql, qll);
|
|
|
|
ql = qll = 0;
|
|
|
|
rsp->orphan_donelist = NULL;
|
|
|
|
rsp->orphan_donetail = &rsp->orphan_donelist;
|
|
|
|
}
|
|
|
|
if (rsp->orphan_nxtlist != NULL) {
|
|
|
|
__call_rcu_nocb_enqueue(rdp, rsp->orphan_nxtlist,
|
|
|
|
rsp->orphan_nxttail, ql, qll);
|
|
|
|
ql = qll = 0;
|
|
|
|
rsp->orphan_nxtlist = NULL;
|
|
|
|
rsp->orphan_nxttail = &rsp->orphan_nxtlist;
|
|
|
|
}
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* There must be at least one non-no-CBs CPU in operation at any given
|
|
|
|
* time, because no-CBs CPUs are not capable of initiating grace periods
|
|
|
|
* independently. This function therefore complains if the specified
|
|
|
|
* CPU is the last non-no-CBs CPU, allowing the CPU-hotplug system to
|
|
|
|
* avoid offlining the last such CPU. (Recursion is a wonderful thing,
|
|
|
|
* but you have to have a base case!)
|
|
|
|
*/
|
|
|
|
static bool nocb_cpu_expendable(int cpu)
|
|
|
|
{
|
|
|
|
cpumask_var_t non_nocb_cpus;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If there are no no-CB CPUs or if this CPU is not a no-CB CPU,
|
|
|
|
* then offlining this CPU is harmless. Let it happen.
|
|
|
|
*/
|
|
|
|
if (!have_rcu_nocb_mask || is_nocb_cpu(cpu))
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
/* If no memory, play it safe and keep the CPU around. */
|
|
|
|
if (!alloc_cpumask_var(&non_nocb_cpus, GFP_NOIO))
|
|
|
|
return 0;
|
|
|
|
cpumask_andnot(non_nocb_cpus, cpu_online_mask, rcu_nocb_mask);
|
|
|
|
cpumask_clear_cpu(cpu, non_nocb_cpus);
|
|
|
|
ret = !cpumask_empty(non_nocb_cpus);
|
|
|
|
free_cpumask_var(non_nocb_cpus);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Helper structure for remote registry of RCU callbacks.
|
|
|
|
* This is needed for when a no-CBs CPU needs to start a grace period.
|
|
|
|
* If it just invokes call_rcu(), the resulting callback will be queued,
|
|
|
|
* which can result in deadlock.
|
|
|
|
*/
|
|
|
|
struct rcu_head_remote {
|
|
|
|
struct rcu_head *rhp;
|
|
|
|
call_rcu_func_t *crf;
|
|
|
|
void (*func)(struct rcu_head *rhp);
|
|
|
|
};
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Register a callback as specified by the rcu_head_remote struct.
|
|
|
|
* This function is intended to be invoked via smp_call_function_single().
|
|
|
|
*/
|
|
|
|
static void call_rcu_local(void *arg)
|
|
|
|
{
|
|
|
|
struct rcu_head_remote *rhrp =
|
|
|
|
container_of(arg, struct rcu_head_remote, rhp);
|
|
|
|
|
|
|
|
rhrp->crf(rhrp->rhp, rhrp->func);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Set up an rcu_head_remote structure and the invoke call_rcu_local()
|
|
|
|
* on CPU 0 (which is guaranteed to be a non-no-CBs CPU) via
|
|
|
|
* smp_call_function_single().
|
|
|
|
*/
|
|
|
|
static void invoke_crf_remote(struct rcu_head *rhp,
|
|
|
|
void (*func)(struct rcu_head *rhp),
|
|
|
|
call_rcu_func_t crf)
|
|
|
|
{
|
|
|
|
struct rcu_head_remote rhr;
|
|
|
|
|
|
|
|
rhr.rhp = rhp;
|
|
|
|
rhr.crf = crf;
|
|
|
|
rhr.func = func;
|
|
|
|
smp_call_function_single(0, call_rcu_local, &rhr, 1);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Helper functions to be passed to wait_rcu_gp(), each of which
|
|
|
|
* invokes invoke_crf_remote() to register a callback appropriately.
|
|
|
|
*/
|
|
|
|
static void __maybe_unused
|
|
|
|
call_rcu_preempt_remote(struct rcu_head *rhp,
|
|
|
|
void (*func)(struct rcu_head *rhp))
|
|
|
|
{
|
|
|
|
invoke_crf_remote(rhp, func, call_rcu);
|
|
|
|
}
|
|
|
|
static void call_rcu_bh_remote(struct rcu_head *rhp,
|
|
|
|
void (*func)(struct rcu_head *rhp))
|
|
|
|
{
|
|
|
|
invoke_crf_remote(rhp, func, call_rcu_bh);
|
|
|
|
}
|
|
|
|
static void call_rcu_sched_remote(struct rcu_head *rhp,
|
|
|
|
void (*func)(struct rcu_head *rhp))
|
|
|
|
{
|
|
|
|
invoke_crf_remote(rhp, func, call_rcu_sched);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Per-rcu_data kthread, but only for no-CBs CPUs. Each kthread invokes
|
|
|
|
* callbacks queued by the corresponding no-CBs CPU.
|
|
|
|
*/
|
|
|
|
static int rcu_nocb_kthread(void *arg)
|
|
|
|
{
|
|
|
|
int c, cl;
|
|
|
|
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 (;;) {
|
|
|
|
/* If not polling, wait for next batch of callbacks. */
|
|
|
|
if (!rcu_nocb_poll)
|
|
|
|
wait_event(rdp->nocb_wq, rdp->nocb_head);
|
|
|
|
list = ACCESS_ONCE(rdp->nocb_head);
|
|
|
|
if (!list) {
|
|
|
|
schedule_timeout_interruptible(1);
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Extract queued callbacks, update counts, and wait
|
|
|
|
* for a grace period to elapse.
|
|
|
|
*/
|
|
|
|
ACCESS_ONCE(rdp->nocb_head) = NULL;
|
|
|
|
tail = xchg(&rdp->nocb_tail, &rdp->nocb_head);
|
|
|
|
c = atomic_long_xchg(&rdp->nocb_q_count, 0);
|
|
|
|
cl = atomic_long_xchg(&rdp->nocb_q_count_lazy, 0);
|
|
|
|
ACCESS_ONCE(rdp->nocb_p_count) += c;
|
|
|
|
ACCESS_ONCE(rdp->nocb_p_count_lazy) += cl;
|
|
|
|
wait_rcu_gp(rdp->rsp->call_remote);
|
|
|
|
|
|
|
|
/* Each pass through the following loop invokes a callback. */
|
|
|
|
trace_rcu_batch_start(rdp->rsp->name, cl, c, -1);
|
|
|
|
c = cl = 0;
|
|
|
|
while (list) {
|
|
|
|
next = list->next;
|
|
|
|
/* Wait for enqueuing to complete, if needed. */
|
|
|
|
while (next == NULL && &list->next != tail) {
|
|
|
|
schedule_timeout_interruptible(1);
|
|
|
|
next = list->next;
|
|
|
|
}
|
|
|
|
debug_rcu_head_unqueue(list);
|
|
|
|
local_bh_disable();
|
|
|
|
if (__rcu_reclaim(rdp->rsp->name, list))
|
|
|
|
cl++;
|
|
|
|
c++;
|
|
|
|
local_bh_enable();
|
|
|
|
list = next;
|
|
|
|
}
|
|
|
|
trace_rcu_batch_end(rdp->rsp->name, c, !!list, 0, 0, 1);
|
|
|
|
ACCESS_ONCE(rdp->nocb_p_count) -= c;
|
|
|
|
ACCESS_ONCE(rdp->nocb_p_count_lazy) -= cl;
|
|
|
|
rdp->n_cbs_invoked += c;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* 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;
|
|
|
|
init_waitqueue_head(&rdp->nocb_wq);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Create a kthread for each RCU flavor for each no-CBs CPU. */
|
|
|
|
static void __init rcu_spawn_nocb_kthreads(struct rcu_state *rsp)
|
|
|
|
{
|
|
|
|
int cpu;
|
|
|
|
struct rcu_data *rdp;
|
|
|
|
struct task_struct *t;
|
|
|
|
|
|
|
|
if (rcu_nocb_mask == NULL)
|
|
|
|
return;
|
|
|
|
for_each_cpu(cpu, rcu_nocb_mask) {
|
|
|
|
rdp = per_cpu_ptr(rsp->rda, cpu);
|
|
|
|
t = kthread_run(rcu_nocb_kthread, rdp, "rcuo%d", cpu);
|
|
|
|
BUG_ON(IS_ERR(t));
|
|
|
|
ACCESS_ONCE(rdp->nocb_kthread) = t;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Prevent __call_rcu() from enqueuing callbacks on no-CBs CPUs */
|
|
|
|
static void init_nocb_callback_list(struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
if (rcu_nocb_mask == NULL ||
|
|
|
|
!cpumask_test_cpu(rdp->cpu, rcu_nocb_mask))
|
|
|
|
return;
|
|
|
|
rdp->nxttail[RCU_NEXT_TAIL] = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Initialize the ->call_remote fields in the rcu_state structures. */
|
|
|
|
static void __init rcu_init_nocb(void)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_PREEMPT_RCU
|
|
|
|
rcu_preempt_state.call_remote = call_rcu_preempt_remote;
|
|
|
|
#endif /* #ifdef CONFIG_PREEMPT_RCU */
|
|
|
|
rcu_bh_state.call_remote = call_rcu_bh_remote;
|
|
|
|
rcu_sched_state.call_remote = call_rcu_sched_remote;
|
|
|
|
}
|
|
|
|
|
|
|
|
#else /* #ifdef CONFIG_RCU_NOCB_CPU */
|
|
|
|
|
|
|
|
static bool is_nocb_cpu(int cpu)
|
|
|
|
{
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool __call_rcu_nocb(struct rcu_data *rdp, struct rcu_head *rhp,
|
|
|
|
bool lazy)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool __maybe_unused rcu_nocb_adopt_orphan_cbs(struct rcu_state *rsp,
|
|
|
|
struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool nocb_cpu_expendable(int cpu)
|
|
|
|
{
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __init rcu_boot_init_nocb_percpu_data(struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __init rcu_spawn_nocb_kthreads(struct rcu_state *rsp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static void init_nocb_callback_list(struct rcu_data *rdp)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
static void __init rcu_init_nocb(void)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* #else #ifdef CONFIG_RCU_NOCB_CPU */
|