Commit f7458a5d authored by Linus Torvalds's avatar Linus Torvalds

Merge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull RCU updates from Ingo Molnar:
 "The RCU changes in this cycle are:

   - Dynticks updates, consolidating open-coded counter accesses into a
     well-defined API

   - SRCU updates: Simplify algorithm, add formal verification

   - Documentation updates

   - Miscellaneous fixes

   - Torture-test updates

  Most of the diffstat comes from the relatively large documentation
  update"

* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (42 commits)
  srcu: Reduce probability of SRCU ->unlock_count[] counter overflow
  rcutorture: Add CBMC-based formal verification for SRCU
  srcu: Force full grace-period ordering
  srcu: Implement more-efficient reader counts
  rcu: Adjust FQS offline checks for exact online-CPU detection
  rcu: Check cond_resched_rcu_qs() state less often to reduce GP overhead
  rcu: Abstract extended quiescent state determination
  rcu: Abstract dynticks extended quiescent state enter/exit operations
  rcu: Add lockdep checks to synchronous expedited primitives
  rcu: Eliminate unused expedited_normal counter
  llist: Clarify comments about when locking is needed
  rcu: Fix comment in rcu_organize_nocb_kthreads()
  rcu: Enable RCU tracepoints by default to aid in debugging
  rcu: Make rcu_cpu_starting() use its "cpu" argument
  rcu: Add comment headers to expedited-grace-period counter functions
  rcu: Don't wake rcuc/X kthreads on NOCB CPUs
  rcu: Re-enable TASKS_RCU for User Mode Linux
  rcu: Once again use NMI-based stack traces in stall warnings
  rcu: Remove short-term CPU kicking
  rcu: Add long-term CPU kicking
  ...
parents 575260e3 a8709fa4
...@@ -4,7 +4,7 @@ ...@@ -4,7 +4,7 @@
<head><title>A Tour Through TREE_RCU's Data Structures [LWN.net]</title> <head><title>A Tour Through TREE_RCU's Data Structures [LWN.net]</title>
<meta HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"> <meta HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<p>January 27, 2016</p> <p>December 18, 2016</p>
<p>This article was contributed by Paul E.&nbsp;McKenney</p> <p>This article was contributed by Paul E.&nbsp;McKenney</p>
<h3>Introduction</h3> <h3>Introduction</h3>
...@@ -31,9 +31,6 @@ to each other. ...@@ -31,9 +31,6 @@ to each other.
Accessor Functions</a> Accessor Functions</a>
</ol> </ol>
At the end we have the
<a href="#Answers to Quick Quizzes">answers to the quick quizzes</a>.
<h3><a name="Data-Structure Relationships">Data-Structure Relationships</a></h3> <h3><a name="Data-Structure Relationships">Data-Structure Relationships</a></h3>
<p>RCU is for all intents and purposes a large state machine, and its <p>RCU is for all intents and purposes a large state machine, and its
......
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
...@@ -1480,7 +1480,7 @@ speed-of-light delays if nothing else. ...@@ -1480,7 +1480,7 @@ speed-of-light delays if nothing else.
<p> <p>
Furthermore, uncertainty about external state is inherent in many cases. Furthermore, uncertainty about external state is inherent in many cases.
For example, a pair of veternarians might use heartbeat to determine For example, a pair of veterinarians might use heartbeat to determine
whether or not a given cat was alive. whether or not a given cat was alive.
But how long should they wait after the last heartbeat to decide that But how long should they wait after the last heartbeat to decide that
the cat is in fact dead? the cat is in fact dead?
...@@ -1489,9 +1489,9 @@ mean that a relaxed cat would be considered to cycle between death ...@@ -1489,9 +1489,9 @@ mean that a relaxed cat would be considered to cycle between death
and life more than 100 times per minute. and life more than 100 times per minute.
Moreover, just as with human beings, a cat's heart might stop for Moreover, just as with human beings, a cat's heart might stop for
some period of time, so the exact wait period is a judgment call. some period of time, so the exact wait period is a judgment call.
One of our pair of veternarians might wait 30 seconds before pronouncing One of our pair of veterinarians might wait 30 seconds before pronouncing
the cat dead, while the other might insist on waiting a full minute. the cat dead, while the other might insist on waiting a full minute.
The two veternarians would then disagree on the state of the cat during The two veterinarians would then disagree on the state of the cat during
the final 30 seconds of the minute following the last heartbeat. the final 30 seconds of the minute following the last heartbeat.
<p> <p>
...@@ -1945,7 +1945,7 @@ guard against mishaps and misuse: ...@@ -1945,7 +1945,7 @@ guard against mishaps and misuse:
<ol> <ol>
<li> It is all too easy to forget to use <tt>rcu_read_lock()</tt> <li> It is all too easy to forget to use <tt>rcu_read_lock()</tt>
everywhere that it is needed, so kernels built with everywhere that it is needed, so kernels built with
<tt>CONFIG_PROVE_RCU=y</tt> will spat if <tt>CONFIG_PROVE_RCU=y</tt> will splat if
<tt>rcu_dereference()</tt> is used outside of an <tt>rcu_dereference()</tt> is used outside of an
RCU read-side critical section. RCU read-side critical section.
Update-side code can use <tt>rcu_dereference_protected()</tt>, Update-side code can use <tt>rcu_dereference_protected()</tt>,
...@@ -2421,7 +2421,7 @@ However, there are some restrictions on the code placed within ...@@ -2421,7 +2421,7 @@ However, there are some restrictions on the code placed within
<li> Blocking is prohibited. <li> Blocking is prohibited.
In practice, this is not a serious restriction given that idle In practice, this is not a serious restriction given that idle
tasks are prohibited from blocking to begin with. tasks are prohibited from blocking to begin with.
<li> Although nesting <tt>RCU_NONIDLE()</tt> is permited, they cannot <li> Although nesting <tt>RCU_NONIDLE()</tt> is permitted, they cannot
nest indefinitely deeply. nest indefinitely deeply.
However, given that they can be nested on the order of a million However, given that they can be nested on the order of a million
deep, even on 32-bit systems, this should not be a serious deep, even on 32-bit systems, this should not be a serious
...@@ -2885,7 +2885,7 @@ APIs for defining and initializing <tt>srcu_struct</tt> structures. ...@@ -2885,7 +2885,7 @@ APIs for defining and initializing <tt>srcu_struct</tt> structures.
<h3><a name="Tasks RCU">Tasks RCU</a></h3> <h3><a name="Tasks RCU">Tasks RCU</a></h3>
<p> <p>
Some forms of tracing use &ldquo;tramopolines&rdquo; to handle the Some forms of tracing use &ldquo;trampolines&rdquo; to handle the
binary rewriting required to install different types of probes. binary rewriting required to install different types of probes.
It would be good to be able to free old trampolines, which sounds It would be good to be able to free old trampolines, which sounds
like a job for some form of RCU. like a job for some form of RCU.
......
...@@ -237,7 +237,7 @@ o "ktl" is the low-order 16 bits (in hexadecimal) of the count of ...@@ -237,7 +237,7 @@ o "ktl" is the low-order 16 bits (in hexadecimal) of the count of
The output of "cat rcu/rcu_preempt/rcuexp" looks as follows: The output of "cat rcu/rcu_preempt/rcuexp" looks as follows:
s=21872 wd1=0 wd2=0 wd3=5 n=0 enq=0 sc=21872 s=21872 wd1=0 wd2=0 wd3=5 enq=0 sc=21872
These fields are as follows: These fields are as follows:
...@@ -249,9 +249,6 @@ o "wd1", "wd2", and "wd3" are the number of times that an attempt ...@@ -249,9 +249,6 @@ o "wd1", "wd2", and "wd3" are the number of times that an attempt
completed an expedited grace period that satisfies the attempted completed an expedited grace period that satisfies the attempted
request. "Our work is done." request. "Our work is done."
o "n" is number of times that a concurrent CPU-hotplug operation
forced a fallback to a normal grace period.
o "enq" is the number of quiescent states still outstanding. o "enq" is the number of quiescent states still outstanding.
o "sc" is the number of times that the attempt to start a o "sc" is the number of times that the attempt to start a
......
...@@ -3269,6 +3269,13 @@ ...@@ -3269,6 +3269,13 @@
Lazy RCU callbacks are those which RCU can Lazy RCU callbacks are those which RCU can
prove do nothing more than free memory. prove do nothing more than free memory.
rcutree.rcu_kick_kthreads= [KNL]
Cause the grace-period kthread to get an extra
wake_up() if it sleeps three times longer than
it should at force-quiescent-state time.
This wake_up() will be accompanied by a
WARN_ONCE() splat and an ftrace_dump().
rcuperf.gp_exp= [KNL] rcuperf.gp_exp= [KNL]
Measure performance of expedited synchronous Measure performance of expedited synchronous
grace-period primitives. grace-period primitives.
......
This diff is collapsed.
...@@ -3,28 +3,33 @@ ...@@ -3,28 +3,33 @@
/* /*
* Lock-less NULL terminated single linked list * Lock-less NULL terminated single linked list
* *
* If there are multiple producers and multiple consumers, llist_add * Cases where locking is not needed:
* can be used in producers and llist_del_all can be used in * If there are multiple producers and multiple consumers, llist_add can be
* consumers. They can work simultaneously without lock. But * used in producers and llist_del_all can be used in consumers simultaneously
* llist_del_first can not be used here. Because llist_del_first * without locking. Also a single consumer can use llist_del_first while
* depends on list->first->next does not changed if list->first is not * multiple producers simultaneously use llist_add, without any locking.
* changed during its operation, but llist_del_first, llist_add, *
* llist_add (or llist_del_all, llist_add, llist_add) sequence in * Cases where locking is needed:
* another consumer may violate that. * If we have multiple consumers with llist_del_first used in one consumer, and
* * llist_del_first or llist_del_all used in other consumers, then a lock is
* If there are multiple producers and one consumer, llist_add can be * needed. This is because llist_del_first depends on list->first->next not
* used in producers and llist_del_all or llist_del_first can be used * changing, but without lock protection, there's no way to be sure about that
* in the consumer. * if a preemption happens in the middle of the delete operation and on being
* * preempted back, the list->first is the same as before causing the cmpxchg in
* This can be summarized as follow: * llist_del_first to succeed. For example, while a llist_del_first operation
* is in progress in one consumer, then a llist_del_first, llist_add,
* llist_add (or llist_del_all, llist_add, llist_add) sequence in another
* consumer may cause violations.
*
* This can be summarized as follows:
* *
* | add | del_first | del_all * | add | del_first | del_all
* add | - | - | - * add | - | - | -
* del_first | | L | L * del_first | | L | L
* del_all | | | - * del_all | | | -
* *
* Where "-" stands for no lock is needed, while "L" stands for lock * Where, a particular row's operation can happen concurrently with a column's
* is needed. * operation, with "-" being no lock needed, while "L" being lock is needed.
* *
* The list entries deleted via llist_del_all can be traversed with * The list entries deleted via llist_del_all can be traversed with
* traversing function such as llist_for_each etc. But the list * traversing function such as llist_for_each etc. But the list
......
...@@ -1161,5 +1161,17 @@ do { \ ...@@ -1161,5 +1161,17 @@ do { \
ftrace_dump(oops_dump_mode); \ ftrace_dump(oops_dump_mode); \
} while (0) } while (0)
/*
* Place this after a lock-acquisition primitive to guarantee that
* an UNLOCK+LOCK pair acts as a full barrier. This guarantee applies
* if the UNLOCK and LOCK are executed by the same CPU or if the
* UNLOCK and LOCK operate on the same lock variable.
*/
#ifdef CONFIG_PPC
#define smp_mb__after_unlock_lock() smp_mb() /* Full ordering for lock. */
#else /* #ifdef CONFIG_PPC */
#define smp_mb__after_unlock_lock() do { } while (0)
#endif /* #else #ifdef CONFIG_PPC */
#endif /* __LINUX_RCUPDATE_H */ #endif /* __LINUX_RCUPDATE_H */
...@@ -27,6 +27,12 @@ ...@@ -27,6 +27,12 @@
#include <linux/cache.h> #include <linux/cache.h>
struct rcu_dynticks;
static inline int rcu_dynticks_snap(struct rcu_dynticks *rdtp)
{
return 0;
}
static inline unsigned long get_state_synchronize_rcu(void) static inline unsigned long get_state_synchronize_rcu(void)
{ {
return 0; return 0;
......
...@@ -33,9 +33,9 @@ ...@@ -33,9 +33,9 @@
#include <linux/rcupdate.h> #include <linux/rcupdate.h>
#include <linux/workqueue.h> #include <linux/workqueue.h>
struct srcu_struct_array { struct srcu_array {
unsigned long c[2]; unsigned long lock_count[2];
unsigned long seq[2]; unsigned long unlock_count[2];
}; };
struct rcu_batch { struct rcu_batch {
...@@ -46,7 +46,7 @@ struct rcu_batch { ...@@ -46,7 +46,7 @@ struct rcu_batch {
struct srcu_struct { struct srcu_struct {
unsigned long completed; unsigned long completed;
struct srcu_struct_array __percpu *per_cpu_ref; struct srcu_array __percpu *per_cpu_ref;
spinlock_t queue_lock; /* protect ->batch_queue, ->running */ spinlock_t queue_lock; /* protect ->batch_queue, ->running */
bool running; bool running;
/* callbacks just queued */ /* callbacks just queued */
...@@ -118,7 +118,7 @@ void process_srcu(struct work_struct *work); ...@@ -118,7 +118,7 @@ void process_srcu(struct work_struct *work);
* See include/linux/percpu-defs.h for the rules on per-CPU variables. * See include/linux/percpu-defs.h for the rules on per-CPU variables.
*/ */
#define __DEFINE_SRCU(name, is_static) \ #define __DEFINE_SRCU(name, is_static) \
static DEFINE_PER_CPU(struct srcu_struct_array, name##_srcu_array);\ static DEFINE_PER_CPU(struct srcu_array, name##_srcu_array);\
is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name) is_static struct srcu_struct name = __SRCU_STRUCT_INIT(name)
#define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */) #define DEFINE_SRCU(name) __DEFINE_SRCU(name, /* not static */)
#define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static) #define DEFINE_STATIC_SRCU(name) __DEFINE_SRCU(name, static)
......
...@@ -385,11 +385,11 @@ TRACE_EVENT(rcu_quiescent_state_report, ...@@ -385,11 +385,11 @@ TRACE_EVENT(rcu_quiescent_state_report,
/* /*
* Tracepoint for quiescent states detected by force_quiescent_state(). * Tracepoint for quiescent states detected by force_quiescent_state().
* These trace events include the type of RCU, the grace-period number * These trace events include the type of RCU, the grace-period number that
* that was blocked by the CPU, the CPU itself, and the type of quiescent * was blocked by the CPU, the CPU itself, and the type of quiescent state,
* state, which can be "dti" for dyntick-idle mode, "ofl" for CPU offline, * which can be "dti" for dyntick-idle mode, "ofl" for CPU offline, "kick"
* or "kick" when kicking a CPU that has been in dyntick-idle mode for * when kicking a CPU that has been in dyntick-idle mode for too long, or
* too long. * "rqc" if the CPU got a quiescent state via its rcu_qs_ctr.
*/ */
TRACE_EVENT(rcu_fqs, TRACE_EVENT(rcu_fqs,
......
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
CONFIG_RCU_TORTURE_TEST=y CONFIG_RCU_TORTURE_TEST=y
CONFIG_PRINTK_TIME=y CONFIG_PRINTK_TIME=y
CONFIG_RCU_TORTURE_TEST_SLOW_CLEANUP=y
CONFIG_RCU_TORTURE_TEST_SLOW_INIT=y
CONFIG_RCU_TORTURE_TEST_SLOW_PREINIT=y
This diff is collapsed.
This diff is collapsed.
This diff is collapsed.
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment