1===================================================
2A Tour Through TREE_RCU's Data Structures [LWN.net]
3===================================================
4
5December 18, 2016
6
7This article was contributed by Paul E. McKenney
8
9Introduction
10============
11
12This document describes RCU's major data structures and their relationship
13to each other.
14
15Data-Structure Relationships
16============================
17
18RCU is for all intents and purposes a large state machine, and its
19data structures maintain the state in such a way as to allow RCU readers
20to execute extremely quickly, while also processing the RCU grace periods
21requested by updaters in an efficient and extremely scalable fashion.
22The efficiency and scalability of RCU updaters is provided primarily
23by a combining tree, as shown below:
24
25.. kernel-figure:: BigTreeClassicRCU.svg
26
27This diagram shows an enclosing ``rcu_state`` structure containing a tree
28of ``rcu_node`` structures. Each leaf node of the ``rcu_node`` tree has up
29to 16 ``rcu_data`` structures associated with it, so that there are
30``NR_CPUS`` number of ``rcu_data`` structures, one for each possible CPU.
31This structure is adjusted at boot time, if needed, to handle the common
32case where ``nr_cpu_ids`` is much less than ``NR_CPUs``.
33For example, a number of Linux distributions set ``NR_CPUs=4096``,
34which results in a three-level ``rcu_node`` tree.
35If the actual hardware has only 16 CPUs, RCU will adjust itself
36at boot time, resulting in an ``rcu_node`` tree with only a single node.
37
38The purpose of this combining tree is to allow per-CPU events
39such as quiescent states, dyntick-idle transitions,
40and CPU hotplug operations to be processed efficiently
41and scalably.
42Quiescent states are recorded by the per-CPU ``rcu_data`` structures,
43and other events are recorded by the leaf-level ``rcu_node``
44structures.
45All of these events are combined at each level of the tree until finally
46grace periods are completed at the tree's root ``rcu_node``
47structure.
48A grace period can be completed at the root once every CPU
49(or, in the case of ``CONFIG_PREEMPT_RCU``, task)
50has passed through a quiescent state.
51Once a grace period has completed, record of that fact is propagated
52back down the tree.
53
54As can be seen from the diagram, on a 64-bit system
55a two-level tree with 64 leaves can accommodate 1,024 CPUs, with a fanout
56of 64 at the root and a fanout of 16 at the leaves.
57
58+-----------------------------------------------------------------------+
59| **Quick Quiz**:                                                       |
60+-----------------------------------------------------------------------+
61| Why isn't the fanout at the leaves also 64?                           |
62+-----------------------------------------------------------------------+
63| **Answer**:                                                           |
64+-----------------------------------------------------------------------+
65| Because there are more types of events that affect the leaf-level     |
66| ``rcu_node`` structures than further up the tree. Therefore, if the   |
67| leaf ``rcu_node`` structures have fanout of 64, the contention on     |
68| these structures' ``->structures`` becomes excessive. Experimentation |
69| on a wide variety of systems has shown that a fanout of 16 works well |
70| for the leaves of the ``rcu_node`` tree.                              |
71|                                                                       |
72| Of course, further experience with systems having hundreds or         |
73| thousands of CPUs may demonstrate that the fanout for the non-leaf    |
74| ``rcu_node`` structures must also be reduced. Such reduction can be   |
75| easily carried out when and if it proves necessary. In the meantime,  |
76| if you are using such a system and running into contention problems   |
77| on the non-leaf ``rcu_node`` structures, you may use the              |
78| ``CONFIG_RCU_FANOUT`` kernel configuration parameter to reduce the    |
79| non-leaf fanout as needed.                                            |
80|                                                                       |
81| Kernels built for systems with strong NUMA characteristics might      |
82| also need to adjust ``CONFIG_RCU_FANOUT`` so that the domains of      |
83| the ``rcu_node`` structures align with hardware boundaries.           |
84| However, there has thus far been no need for this.                    |
85+-----------------------------------------------------------------------+
86
87If your system has more than 1,024 CPUs (or more than 512 CPUs on a
8832-bit system), then RCU will automatically add more levels to the tree.
89For example, if you are crazy enough to build a 64-bit system with
9065,536 CPUs, RCU would configure the ``rcu_node`` tree as follows:
91
92.. kernel-figure:: HugeTreeClassicRCU.svg
93
94RCU currently permits up to a four-level tree, which on a 64-bit system
95accommodates up to 4,194,304 CPUs, though only a mere 524,288 CPUs for
9632-bit systems. On the other hand, you can set both
97``CONFIG_RCU_FANOUT`` and ``CONFIG_RCU_FANOUT_LEAF`` to be as small as
982, which would result in a 16-CPU test using a 4-level tree. This can be
99useful for testing large-system capabilities on small test machines.
100
101This multi-level combining tree allows us to get most of the performance
102and scalability benefits of partitioning, even though RCU grace-period
103detection is inherently a global operation. The trick here is that only
104the last CPU to report a quiescent state into a given ``rcu_node``
105structure need advance to the ``rcu_node`` structure at the next level
106up the tree. This means that at the leaf-level ``rcu_node`` structure,
107only one access out of sixteen will progress up the tree. For the
108internal ``rcu_node`` structures, the situation is even more extreme:
109Only one access out of sixty-four will progress up the tree. Because the
110vast majority of the CPUs do not progress up the tree, the lock
111contention remains roughly constant up the tree. No matter how many CPUs
112there are in the system, at most 64 quiescent-state reports per grace
113period will progress all the way to the root ``rcu_node`` structure,
114thus ensuring that the lock contention on that root ``rcu_node``
115structure remains acceptably low.
116
117In effect, the combining tree acts like a big shock absorber, keeping
118lock contention under control at all tree levels regardless of the level
119of loading on the system.
120
121RCU updaters wait for normal grace periods by registering RCU callbacks,
122either directly via ``call_rcu()`` or indirectly via
123``synchronize_rcu()`` and friends. RCU callbacks are represented by
124``rcu_head`` structures, which are queued on ``rcu_data`` structures
125while they are waiting for a grace period to elapse, as shown in the
126following figure:
127
128.. kernel-figure:: BigTreePreemptRCUBHdyntickCB.svg
129
130This figure shows how ``TREE_RCU``'s and ``PREEMPT_RCU``'s major data
131structures are related. Lesser data structures will be introduced with
132the algorithms that make use of them.
133
134Note that each of the data structures in the above figure has its own
135synchronization:
136
137#. Each ``rcu_state`` structures has a lock and a mutex, and some fields
138   are protected by the corresponding root ``rcu_node`` structure's lock.
139#. Each ``rcu_node`` structure has a spinlock.
140#. The fields in ``rcu_data`` are private to the corresponding CPU,
141   although a few can be read and written by other CPUs.
142
143It is important to note that different data structures can have very
144different ideas about the state of RCU at any given time. For but one
145example, awareness of the start or end of a given RCU grace period
146propagates slowly through the data structures. This slow propagation is
147absolutely necessary for RCU to have good read-side performance. If this
148balkanized implementation seems foreign to you, one useful trick is to
149consider each instance of these data structures to be a different
150person, each having the usual slightly different view of reality.
151
152The general role of each of these data structures is as follows:
153
154#. ``rcu_state``: This structure forms the interconnection between the
155   ``rcu_node`` and ``rcu_data`` structures, tracks grace periods,
156   serves as short-term repository for callbacks orphaned by CPU-hotplug
157   events, maintains ``rcu_barrier()`` state, tracks expedited
158   grace-period state, and maintains state used to force quiescent
159   states when grace periods extend too long,
160#. ``rcu_node``: This structure forms the combining tree that propagates
161   quiescent-state information from the leaves to the root, and also
162   propagates grace-period information from the root to the leaves. It
163   provides local copies of the grace-period state in order to allow
164   this information to be accessed in a synchronized manner without
165   suffering the scalability limitations that would otherwise be imposed
166   by global locking. In ``CONFIG_PREEMPT_RCU`` kernels, it manages the
167   lists of tasks that have blocked while in their current RCU read-side
168   critical section. In ``CONFIG_PREEMPT_RCU`` with
169   ``CONFIG_RCU_BOOST``, it manages the per-\ ``rcu_node``
170   priority-boosting kernel threads (kthreads) and state. Finally, it
171   records CPU-hotplug state in order to determine which CPUs should be
172   ignored during a given grace period.
173#. ``rcu_data``: This per-CPU structure is the focus of quiescent-state
174   detection and RCU callback queuing. It also tracks its relationship
175   to the corresponding leaf ``rcu_node`` structure to allow
176   more-efficient propagation of quiescent states up the ``rcu_node``
177   combining tree. Like the ``rcu_node`` structure, it provides a local
178   copy of the grace-period information to allow for-free synchronized
179   access to this information from the corresponding CPU. Finally, this
180   structure records past dyntick-idle state for the corresponding CPU
181   and also tracks statistics.
182#. ``rcu_head``: This structure represents RCU callbacks, and is the
183   only structure allocated and managed by RCU users. The ``rcu_head``
184   structure is normally embedded within the RCU-protected data
185   structure.
186
187If all you wanted from this article was a general notion of how RCU's
188data structures are related, you are done. Otherwise, each of the
189following sections give more details on the ``rcu_state``, ``rcu_node``
190and ``rcu_data`` data structures.
191
192The ``rcu_state`` Structure
193~~~~~~~~~~~~~~~~~~~~~~~~~~~
194
195The ``rcu_state`` structure is the base structure that represents the
196state of RCU in the system. This structure forms the interconnection
197between the ``rcu_node`` and ``rcu_data`` structures, tracks grace
198periods, contains the lock used to synchronize with CPU-hotplug events,
199and maintains state used to force quiescent states when grace periods
200extend too long,
201
202A few of the ``rcu_state`` structure's fields are discussed, singly and
203in groups, in the following sections. The more specialized fields are
204covered in the discussion of their use.
205
206Relationship to rcu_node and rcu_data Structures
207''''''''''''''''''''''''''''''''''''''''''''''''
208
209This portion of the ``rcu_state`` structure is declared as follows:
210
211::
212
213     1   struct rcu_node node[NUM_RCU_NODES];
214     2   struct rcu_node *level[NUM_RCU_LVLS + 1];
215     3   struct rcu_data __percpu *rda;
216
217+-----------------------------------------------------------------------+
218| **Quick Quiz**:                                                       |
219+-----------------------------------------------------------------------+
220| Wait a minute! You said that the ``rcu_node`` structures formed a     |
221| tree, but they are declared as a flat array! What gives?              |
222+-----------------------------------------------------------------------+
223| **Answer**:                                                           |
224+-----------------------------------------------------------------------+
225| The tree is laid out in the array. The first node In the array is the |
226| head, the next set of nodes in the array are children of the head     |
227| node, and so on until the last set of nodes in the array are the      |
228| leaves.                                                               |
229| See the following diagrams to see how this works.                     |
230+-----------------------------------------------------------------------+
231
232The ``rcu_node`` tree is embedded into the ``->node[]`` array as shown
233in the following figure:
234
235.. kernel-figure:: TreeMapping.svg
236
237One interesting consequence of this mapping is that a breadth-first
238traversal of the tree is implemented as a simple linear scan of the
239array, which is in fact what the ``rcu_for_each_node_breadth_first()``
240macro does. This macro is used at the beginning and ends of grace
241periods.
242
243Each entry of the ``->level`` array references the first ``rcu_node``
244structure on the corresponding level of the tree, for example, as shown
245below:
246
247.. kernel-figure:: TreeMappingLevel.svg
248
249The zero\ :sup:`th` element of the array references the root
250``rcu_node`` structure, the first element references the first child of
251the root ``rcu_node``, and finally the second element references the
252first leaf ``rcu_node`` structure.
253
254For whatever it is worth, if you draw the tree to be tree-shaped rather
255than array-shaped, it is easy to draw a planar representation:
256
257.. kernel-figure:: TreeLevel.svg
258
259Finally, the ``->rda`` field references a per-CPU pointer to the
260corresponding CPU's ``rcu_data`` structure.
261
262All of these fields are constant once initialization is complete, and
263therefore need no protection.
264
265Grace-Period Tracking
266'''''''''''''''''''''
267
268This portion of the ``rcu_state`` structure is declared as follows:
269
270::
271
272     1   unsigned long gp_seq;
273
274RCU grace periods are numbered, and the ``->gp_seq`` field contains the
275current grace-period sequence number. The bottom two bits are the state
276of the current grace period, which can be zero for not yet started or
277one for in progress. In other words, if the bottom two bits of
278``->gp_seq`` are zero, then RCU is idle. Any other value in the bottom
279two bits indicates that something is broken. This field is protected by
280the root ``rcu_node`` structure's ``->lock`` field.
281
282There are ``->gp_seq`` fields in the ``rcu_node`` and ``rcu_data``
283structures as well. The fields in the ``rcu_state`` structure represent
284the most current value, and those of the other structures are compared
285in order to detect the beginnings and ends of grace periods in a
286distributed fashion. The values flow from ``rcu_state`` to ``rcu_node``
287(down the tree from the root to the leaves) to ``rcu_data``.
288
289Miscellaneous
290'''''''''''''
291
292This portion of the ``rcu_state`` structure is declared as follows:
293
294::
295
296     1   unsigned long gp_max;
297     2   char abbr;
298     3   char *name;
299
300The ``->gp_max`` field tracks the duration of the longest grace period
301in jiffies. It is protected by the root ``rcu_node``'s ``->lock``.
302
303The ``->name`` and ``->abbr`` fields distinguish between preemptible RCU
304(“rcu_preempt” and “p”) and non-preemptible RCU (“rcu_sched” and “s”).
305These fields are used for diagnostic and tracing purposes.
306
307The ``rcu_node`` Structure
308~~~~~~~~~~~~~~~~~~~~~~~~~~
309
310The ``rcu_node`` structures form the combining tree that propagates
311quiescent-state information from the leaves to the root and also that
312propagates grace-period information from the root down to the leaves.
313They provides local copies of the grace-period state in order to allow
314this information to be accessed in a synchronized manner without
315suffering the scalability limitations that would otherwise be imposed by
316global locking. In ``CONFIG_PREEMPT_RCU`` kernels, they manage the lists
317of tasks that have blocked while in their current RCU read-side critical
318section. In ``CONFIG_PREEMPT_RCU`` with ``CONFIG_RCU_BOOST``, they
319manage the per-\ ``rcu_node`` priority-boosting kernel threads
320(kthreads) and state. Finally, they record CPU-hotplug state in order to
321determine which CPUs should be ignored during a given grace period.
322
323The ``rcu_node`` structure's fields are discussed, singly and in groups,
324in the following sections.
325
326Connection to Combining Tree
327''''''''''''''''''''''''''''
328
329This portion of the ``rcu_node`` structure is declared as follows:
330
331::
332
333     1   struct rcu_node *parent;
334     2   u8 level;
335     3   u8 grpnum;
336     4   unsigned long grpmask;
337     5   int grplo;
338     6   int grphi;
339
340The ``->parent`` pointer references the ``rcu_node`` one level up in the
341tree, and is ``NULL`` for the root ``rcu_node``. The RCU implementation
342makes heavy use of this field to push quiescent states up the tree. The
343``->level`` field gives the level in the tree, with the root being at
344level zero, its children at level one, and so on. The ``->grpnum`` field
345gives this node's position within the children of its parent, so this
346number can range between 0 and 31 on 32-bit systems and between 0 and 63
347on 64-bit systems. The ``->level`` and ``->grpnum`` fields are used only
348during initialization and for tracing. The ``->grpmask`` field is the
349bitmask counterpart of ``->grpnum``, and therefore always has exactly
350one bit set. This mask is used to clear the bit corresponding to this
351``rcu_node`` structure in its parent's bitmasks, which are described
352later. Finally, the ``->grplo`` and ``->grphi`` fields contain the
353lowest and highest numbered CPU served by this ``rcu_node`` structure,
354respectively.
355
356All of these fields are constant, and thus do not require any
357synchronization.
358
359Synchronization
360'''''''''''''''
361
362This field of the ``rcu_node`` structure is declared as follows:
363
364::
365
366     1   raw_spinlock_t lock;
367
368This field is used to protect the remaining fields in this structure,
369unless otherwise stated. That said, all of the fields in this structure
370can be accessed without locking for tracing purposes. Yes, this can
371result in confusing traces, but better some tracing confusion than to be
372heisenbugged out of existence.
373
374.. _grace-period-tracking-1:
375
376Grace-Period Tracking
377'''''''''''''''''''''
378
379This portion of the ``rcu_node`` structure is declared as follows:
380
381::
382
383     1   unsigned long gp_seq;
384     2   unsigned long gp_seq_needed;
385
386The ``rcu_node`` structures' ``->gp_seq`` fields are the counterparts of
387the field of the same name in the ``rcu_state`` structure. They each may
388lag up to one step behind their ``rcu_state`` counterpart. If the bottom
389two bits of a given ``rcu_node`` structure's ``->gp_seq`` field is zero,
390then this ``rcu_node`` structure believes that RCU is idle.
391
392The ``>gp_seq`` field of each ``rcu_node`` structure is updated at the
393beginning and the end of each grace period.
394
395The ``->gp_seq_needed`` fields record the furthest-in-the-future grace
396period request seen by the corresponding ``rcu_node`` structure. The
397request is considered fulfilled when the value of the ``->gp_seq`` field
398equals or exceeds that of the ``->gp_seq_needed`` field.
399
400+-----------------------------------------------------------------------+
401| **Quick Quiz**:                                                       |
402+-----------------------------------------------------------------------+
403| Suppose that this ``rcu_node`` structure doesn't see a request for a  |
404| very long time. Won't wrapping of the ``->gp_seq`` field cause        |
405| problems?                                                             |
406+-----------------------------------------------------------------------+
407| **Answer**:                                                           |
408+-----------------------------------------------------------------------+
409| No, because if the ``->gp_seq_needed`` field lags behind the          |
410| ``->gp_seq`` field, the ``->gp_seq_needed`` field will be updated at  |
411| the end of the grace period. Modulo-arithmetic comparisons therefore  |
412| will always get the correct answer, even with wrapping.               |
413+-----------------------------------------------------------------------+
414
415Quiescent-State Tracking
416''''''''''''''''''''''''
417
418These fields manage the propagation of quiescent states up the combining
419tree.
420
421This portion of the ``rcu_node`` structure has fields as follows:
422
423::
424
425     1   unsigned long qsmask;
426     2   unsigned long expmask;
427     3   unsigned long qsmaskinit;
428     4   unsigned long expmaskinit;
429
430The ``->qsmask`` field tracks which of this ``rcu_node`` structure's
431children still need to report quiescent states for the current normal
432grace period. Such children will have a value of 1 in their
433corresponding bit. Note that the leaf ``rcu_node`` structures should be
434thought of as having ``rcu_data`` structures as their children.
435Similarly, the ``->expmask`` field tracks which of this ``rcu_node``
436structure's children still need to report quiescent states for the
437current expedited grace period. An expedited grace period has the same
438conceptual properties as a normal grace period, but the expedited
439implementation accepts extreme CPU overhead to obtain much lower
440grace-period latency, for example, consuming a few tens of microseconds
441worth of CPU time to reduce grace-period duration from milliseconds to
442tens of microseconds. The ``->qsmaskinit`` field tracks which of this
443``rcu_node`` structure's children cover for at least one online CPU.
444This mask is used to initialize ``->qsmask``, and ``->expmaskinit`` is
445used to initialize ``->expmask`` and the beginning of the normal and
446expedited grace periods, respectively.
447
448+-----------------------------------------------------------------------+
449| **Quick Quiz**:                                                       |
450+-----------------------------------------------------------------------+
451| Why are these bitmasks protected by locking? Come on, haven't you     |
452| heard of atomic instructions???                                       |
453+-----------------------------------------------------------------------+
454| **Answer**:                                                           |
455+-----------------------------------------------------------------------+
456| Lockless grace-period computation! Such a tantalizing possibility!    |
457| But consider the following sequence of events:                        |
458|                                                                       |
459| #. CPU 0 has been in dyntick-idle mode for quite some time. When it   |
460|    wakes up, it notices that the current RCU grace period needs it to |
461|    report in, so it sets a flag where the scheduling clock interrupt  |
462|    will find it.                                                      |
463| #. Meanwhile, CPU 1 is running ``force_quiescent_state()``, and       |
464|    notices that CPU 0 has been in dyntick idle mode, which qualifies  |
465|    as an extended quiescent state.                                    |
466| #. CPU 0's scheduling clock interrupt fires in the middle of an RCU   |
467|    read-side critical section, and notices that the RCU core needs    |
468|    something, so commences RCU softirq processing.                    |
469| #. CPU 0's softirq handler executes and is just about ready to report |
470|    its quiescent state up the ``rcu_node`` tree.                      |
471| #. But CPU 1 beats it to the punch, completing the current grace      |
472|    period and starting a new one.                                     |
473| #. CPU 0 now reports its quiescent state for the wrong grace period.  |
474|    That grace period might now end before the RCU read-side critical  |
475|    section. If that happens, disaster will ensue.                     |
476|                                                                       |
477| So the locking is absolutely required in order to coordinate clearing |
478| of the bits with updating of the grace-period sequence number in      |
479| ``->gp_seq``.                                                         |
480+-----------------------------------------------------------------------+
481
482Blocked-Task Management
483'''''''''''''''''''''''
484
485``PREEMPT_RCU`` allows tasks to be preempted in the midst of their RCU
486read-side critical sections, and these tasks must be tracked explicitly.
487The details of exactly why and how they are tracked will be covered in a
488separate article on RCU read-side processing. For now, it is enough to
489know that the ``rcu_node`` structure tracks them.
490
491::
492
493     1   struct list_head blkd_tasks;
494     2   struct list_head *gp_tasks;
495     3   struct list_head *exp_tasks;
496     4   bool wait_blkd_tasks;
497
498The ``->blkd_tasks`` field is a list header for the list of blocked and
499preempted tasks. As tasks undergo context switches within RCU read-side
500critical sections, their ``task_struct`` structures are enqueued (via
501the ``task_struct``'s ``->rcu_node_entry`` field) onto the head of the
502``->blkd_tasks`` list for the leaf ``rcu_node`` structure corresponding
503to the CPU on which the outgoing context switch executed. As these tasks
504later exit their RCU read-side critical sections, they remove themselves
505from the list. This list is therefore in reverse time order, so that if
506one of the tasks is blocking the current grace period, all subsequent
507tasks must also be blocking that same grace period. Therefore, a single
508pointer into this list suffices to track all tasks blocking a given
509grace period. That pointer is stored in ``->gp_tasks`` for normal grace
510periods and in ``->exp_tasks`` for expedited grace periods. These last
511two fields are ``NULL`` if either there is no grace period in flight or
512if there are no blocked tasks preventing that grace period from
513completing. If either of these two pointers is referencing a task that
514removes itself from the ``->blkd_tasks`` list, then that task must
515advance the pointer to the next task on the list, or set the pointer to
516``NULL`` if there are no subsequent tasks on the list.
517
518For example, suppose that tasks T1, T2, and T3 are all hard-affinitied
519to the largest-numbered CPU in the system. Then if task T1 blocked in an
520RCU read-side critical section, then an expedited grace period started,
521then task T2 blocked in an RCU read-side critical section, then a normal
522grace period started, and finally task 3 blocked in an RCU read-side
523critical section, then the state of the last leaf ``rcu_node``
524structure's blocked-task list would be as shown below:
525
526.. kernel-figure:: blkd_task.svg
527
528Task T1 is blocking both grace periods, task T2 is blocking only the
529normal grace period, and task T3 is blocking neither grace period. Note
530that these tasks will not remove themselves from this list immediately
531upon resuming execution. They will instead remain on the list until they
532execute the outermost ``rcu_read_unlock()`` that ends their RCU
533read-side critical section.
534
535The ``->wait_blkd_tasks`` field indicates whether or not the current
536grace period is waiting on a blocked task.
537
538Sizing the ``rcu_node`` Array
539'''''''''''''''''''''''''''''
540
541The ``rcu_node`` array is sized via a series of C-preprocessor
542expressions as follows:
543
544::
545
546    1 #ifdef CONFIG_RCU_FANOUT
547    2 #define RCU_FANOUT CONFIG_RCU_FANOUT
548    3 #else
549    4 # ifdef CONFIG_64BIT
550    5 # define RCU_FANOUT 64
551    6 # else
552    7 # define RCU_FANOUT 32
553    8 # endif
554    9 #endif
555   10
556   11 #ifdef CONFIG_RCU_FANOUT_LEAF
557   12 #define RCU_FANOUT_LEAF CONFIG_RCU_FANOUT_LEAF
558   13 #else
559   14 # ifdef CONFIG_64BIT
560   15 # define RCU_FANOUT_LEAF 64
561   16 # else
562   17 # define RCU_FANOUT_LEAF 32
563   18 # endif
564   19 #endif
565   20
566   21 #define RCU_FANOUT_1        (RCU_FANOUT_LEAF)
567   22 #define RCU_FANOUT_2        (RCU_FANOUT_1 * RCU_FANOUT)
568   23 #define RCU_FANOUT_3        (RCU_FANOUT_2 * RCU_FANOUT)
569   24 #define RCU_FANOUT_4        (RCU_FANOUT_3 * RCU_FANOUT)
570   25
571   26 #if NR_CPUS <= RCU_FANOUT_1
572   27 #  define RCU_NUM_LVLS        1
573   28 #  define NUM_RCU_LVL_0        1
574   29 #  define NUM_RCU_NODES        NUM_RCU_LVL_0
575   30 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0 }
576   31 #  define RCU_NODE_NAME_INIT  { "rcu_node_0" }
577   32 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0" }
578   33 #  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0" }
579   34 #elif NR_CPUS <= RCU_FANOUT_2
580   35 #  define RCU_NUM_LVLS        2
581   36 #  define NUM_RCU_LVL_0        1
582   37 #  define NUM_RCU_LVL_1        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
583   38 #  define NUM_RCU_NODES        (NUM_RCU_LVL_0 + NUM_RCU_LVL_1)
584   39 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1 }
585   40 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1" }
586   41 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1" }
587   42 #  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1" }
588   43 #elif NR_CPUS <= RCU_FANOUT_3
589   44 #  define RCU_NUM_LVLS        3
590   45 #  define NUM_RCU_LVL_0        1
591   46 #  define NUM_RCU_LVL_1        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
592   47 #  define NUM_RCU_LVL_2        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
593   48 #  define NUM_RCU_NODES        (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2)
594   49 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2 }
595   50 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1", "rcu_node_2" }
596   51 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2" }
597   52 #  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2" }
598   53 #elif NR_CPUS <= RCU_FANOUT_4
599   54 #  define RCU_NUM_LVLS        4
600   55 #  define NUM_RCU_LVL_0        1
601   56 #  define NUM_RCU_LVL_1        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_3)
602   57 #  define NUM_RCU_LVL_2        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_2)
603   58 #  define NUM_RCU_LVL_3        DIV_ROUND_UP(NR_CPUS, RCU_FANOUT_1)
604   59 #  define NUM_RCU_NODES        (NUM_RCU_LVL_0 + NUM_RCU_LVL_1 + NUM_RCU_LVL_2 + NUM_RCU_LVL_3)
605   60 #  define NUM_RCU_LVL_INIT    { NUM_RCU_LVL_0, NUM_RCU_LVL_1, NUM_RCU_LVL_2, NUM_RCU_LVL_3 }
606   61 #  define RCU_NODE_NAME_INIT  { "rcu_node_0", "rcu_node_1", "rcu_node_2", "rcu_node_3" }
607   62 #  define RCU_FQS_NAME_INIT   { "rcu_node_fqs_0", "rcu_node_fqs_1", "rcu_node_fqs_2", "rcu_node_fqs_3" }
608   63 #  define RCU_EXP_NAME_INIT   { "rcu_node_exp_0", "rcu_node_exp_1", "rcu_node_exp_2", "rcu_node_exp_3" }
609   64 #else
610   65 # error "CONFIG_RCU_FANOUT insufficient for NR_CPUS"
611   66 #endif
612
613The maximum number of levels in the ``rcu_node`` structure is currently
614limited to four, as specified by lines 21-24 and the structure of the
615subsequent “if” statement. For 32-bit systems, this allows
61616*32*32*32=524,288 CPUs, which should be sufficient for the next few
617years at least. For 64-bit systems, 16*64*64*64=4,194,304 CPUs is
618allowed, which should see us through the next decade or so. This
619four-level tree also allows kernels built with ``CONFIG_RCU_FANOUT=8``
620to support up to 4096 CPUs, which might be useful in very large systems
621having eight CPUs per socket (but please note that no one has yet shown
622any measurable performance degradation due to misaligned socket and
623``rcu_node`` boundaries). In addition, building kernels with a full four
624levels of ``rcu_node`` tree permits better testing of RCU's
625combining-tree code.
626
627The ``RCU_FANOUT`` symbol controls how many children are permitted at
628each non-leaf level of the ``rcu_node`` tree. If the
629``CONFIG_RCU_FANOUT`` Kconfig option is not specified, it is set based
630on the word size of the system, which is also the Kconfig default.
631
632The ``RCU_FANOUT_LEAF`` symbol controls how many CPUs are handled by
633each leaf ``rcu_node`` structure. Experience has shown that allowing a
634given leaf ``rcu_node`` structure to handle 64 CPUs, as permitted by the
635number of bits in the ``->qsmask`` field on a 64-bit system, results in
636excessive contention for the leaf ``rcu_node`` structures' ``->lock``
637fields. The number of CPUs per leaf ``rcu_node`` structure is therefore
638limited to 16 given the default value of ``CONFIG_RCU_FANOUT_LEAF``. If
639``CONFIG_RCU_FANOUT_LEAF`` is unspecified, the value selected is based
640on the word size of the system, just as for ``CONFIG_RCU_FANOUT``.
641Lines 11-19 perform this computation.
642
643Lines 21-24 compute the maximum number of CPUs supported by a
644single-level (which contains a single ``rcu_node`` structure),
645two-level, three-level, and four-level ``rcu_node`` tree, respectively,
646given the fanout specified by ``RCU_FANOUT`` and ``RCU_FANOUT_LEAF``.
647These numbers of CPUs are retained in the ``RCU_FANOUT_1``,
648``RCU_FANOUT_2``, ``RCU_FANOUT_3``, and ``RCU_FANOUT_4`` C-preprocessor
649variables, respectively.
650
651These variables are used to control the C-preprocessor ``#if`` statement
652spanning lines 26-66 that computes the number of ``rcu_node`` structures
653required for each level of the tree, as well as the number of levels
654required. The number of levels is placed in the ``NUM_RCU_LVLS``
655C-preprocessor variable by lines 27, 35, 44, and 54. The number of
656``rcu_node`` structures for the topmost level of the tree is always
657exactly one, and this value is unconditionally placed into
658``NUM_RCU_LVL_0`` by lines 28, 36, 45, and 55. The rest of the levels
659(if any) of the ``rcu_node`` tree are computed by dividing the maximum
660number of CPUs by the fanout supported by the number of levels from the
661current level down, rounding up. This computation is performed by
662lines 37, 46-47, and 56-58. Lines 31-33, 40-42, 50-52, and 62-63 create
663initializers for lockdep lock-class names. Finally, lines 64-66 produce
664an error if the maximum number of CPUs is too large for the specified
665fanout.
666
667The ``rcu_segcblist`` Structure
668~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
669
670The ``rcu_segcblist`` structure maintains a segmented list of callbacks
671as follows:
672
673::
674
675    1 #define RCU_DONE_TAIL        0
676    2 #define RCU_WAIT_TAIL        1
677    3 #define RCU_NEXT_READY_TAIL  2
678    4 #define RCU_NEXT_TAIL        3
679    5 #define RCU_CBLIST_NSEGS     4
680    6
681    7 struct rcu_segcblist {
682    8   struct rcu_head *head;
683    9   struct rcu_head **tails[RCU_CBLIST_NSEGS];
684   10   unsigned long gp_seq[RCU_CBLIST_NSEGS];
685   11   long len;
686   12   long len_lazy;
687   13 };
688
689The segments are as follows:
690
691#. ``RCU_DONE_TAIL``: Callbacks whose grace periods have elapsed. These
692   callbacks are ready to be invoked.
693#. ``RCU_WAIT_TAIL``: Callbacks that are waiting for the current grace
694   period. Note that different CPUs can have different ideas about which
695   grace period is current, hence the ``->gp_seq`` field.
696#. ``RCU_NEXT_READY_TAIL``: Callbacks waiting for the next grace period
697   to start.
698#. ``RCU_NEXT_TAIL``: Callbacks that have not yet been associated with a
699   grace period.
700
701The ``->head`` pointer references the first callback or is ``NULL`` if
702the list contains no callbacks (which is *not* the same as being empty).
703Each element of the ``->tails[]`` array references the ``->next``
704pointer of the last callback in the corresponding segment of the list,
705or the list's ``->head`` pointer if that segment and all previous
706segments are empty. If the corresponding segment is empty but some
707previous segment is not empty, then the array element is identical to
708its predecessor. Older callbacks are closer to the head of the list, and
709new callbacks are added at the tail. This relationship between the
710``->head`` pointer, the ``->tails[]`` array, and the callbacks is shown
711in this diagram:
712
713.. kernel-figure:: nxtlist.svg
714
715In this figure, the ``->head`` pointer references the first RCU callback
716in the list. The ``->tails[RCU_DONE_TAIL]`` array element references the
717``->head`` pointer itself, indicating that none of the callbacks is
718ready to invoke. The ``->tails[RCU_WAIT_TAIL]`` array element references
719callback CB 2's ``->next`` pointer, which indicates that CB 1 and CB 2
720are both waiting on the current grace period, give or take possible
721disagreements about exactly which grace period is the current one. The
722``->tails[RCU_NEXT_READY_TAIL]`` array element references the same RCU
723callback that ``->tails[RCU_WAIT_TAIL]`` does, which indicates that
724there are no callbacks waiting on the next RCU grace period. The
725``->tails[RCU_NEXT_TAIL]`` array element references CB 4's ``->next``
726pointer, indicating that all the remaining RCU callbacks have not yet
727been assigned to an RCU grace period. Note that the
728``->tails[RCU_NEXT_TAIL]`` array element always references the last RCU
729callback's ``->next`` pointer unless the callback list is empty, in
730which case it references the ``->head`` pointer.
731
732There is one additional important special case for the
733``->tails[RCU_NEXT_TAIL]`` array element: It can be ``NULL`` when this
734list is *disabled*. Lists are disabled when the corresponding CPU is
735offline or when the corresponding CPU's callbacks are offloaded to a
736kthread, both of which are described elsewhere.
737
738CPUs advance their callbacks from the ``RCU_NEXT_TAIL`` to the
739``RCU_NEXT_READY_TAIL`` to the ``RCU_WAIT_TAIL`` to the
740``RCU_DONE_TAIL`` list segments as grace periods advance.
741
742The ``->gp_seq[]`` array records grace-period numbers corresponding to
743the list segments. This is what allows different CPUs to have different
744ideas as to which is the current grace period while still avoiding
745premature invocation of their callbacks. In particular, this allows CPUs
746that go idle for extended periods to determine which of their callbacks
747are ready to be invoked after reawakening.
748
749The ``->len`` counter contains the number of callbacks in ``->head``,
750and the ``->len_lazy`` contains the number of those callbacks that are
751known to only free memory, and whose invocation can therefore be safely
752deferred.
753
754.. important::
755
756   It is the ``->len`` field that determines whether or
757   not there are callbacks associated with this ``rcu_segcblist``
758   structure, *not* the ``->head`` pointer. The reason for this is that all
759   the ready-to-invoke callbacks (that is, those in the ``RCU_DONE_TAIL``
760   segment) are extracted all at once at callback-invocation time
761   (``rcu_do_batch``), due to which ``->head`` may be set to NULL if there
762   are no not-done callbacks remaining in the ``rcu_segcblist``. If
763   callback invocation must be postponed, for example, because a
764   high-priority process just woke up on this CPU, then the remaining
765   callbacks are placed back on the ``RCU_DONE_TAIL`` segment and
766   ``->head`` once again points to the start of the segment. In short, the
767   head field can briefly be ``NULL`` even though the CPU has callbacks
768   present the entire time. Therefore, it is not appropriate to test the
769   ``->head`` pointer for ``NULL``.
770
771In contrast, the ``->len`` and ``->len_lazy`` counts are adjusted only
772after the corresponding callbacks have been invoked. This means that the
773``->len`` count is zero only if the ``rcu_segcblist`` structure really
774is devoid of callbacks. Of course, off-CPU sampling of the ``->len``
775count requires careful use of appropriate synchronization, for example,
776memory barriers. This synchronization can be a bit subtle, particularly
777in the case of ``rcu_barrier()``.
778
779The ``rcu_data`` Structure
780~~~~~~~~~~~~~~~~~~~~~~~~~~
781
782The ``rcu_data`` maintains the per-CPU state for the RCU subsystem. The
783fields in this structure may be accessed only from the corresponding CPU
784(and from tracing) unless otherwise stated. This structure is the focus
785of quiescent-state detection and RCU callback queuing. It also tracks
786its relationship to the corresponding leaf ``rcu_node`` structure to
787allow more-efficient propagation of quiescent states up the ``rcu_node``
788combining tree. Like the ``rcu_node`` structure, it provides a local
789copy of the grace-period information to allow for-free synchronized
790access to this information from the corresponding CPU. Finally, this
791structure records past dyntick-idle state for the corresponding CPU and
792also tracks statistics.
793
794The ``rcu_data`` structure's fields are discussed, singly and in groups,
795in the following sections.
796
797Connection to Other Data Structures
798'''''''''''''''''''''''''''''''''''
799
800This portion of the ``rcu_data`` structure is declared as follows:
801
802::
803
804     1   int cpu;
805     2   struct rcu_node *mynode;
806     3   unsigned long grpmask;
807     4   bool beenonline;
808
809The ``->cpu`` field contains the number of the corresponding CPU and the
810``->mynode`` field references the corresponding ``rcu_node`` structure.
811The ``->mynode`` is used to propagate quiescent states up the combining
812tree. These two fields are constant and therefore do not require
813synchronization.
814
815The ``->grpmask`` field indicates the bit in the ``->mynode->qsmask``
816corresponding to this ``rcu_data`` structure, and is also used when
817propagating quiescent states. The ``->beenonline`` flag is set whenever
818the corresponding CPU comes online, which means that the debugfs tracing
819need not dump out any ``rcu_data`` structure for which this flag is not
820set.
821
822Quiescent-State and Grace-Period Tracking
823'''''''''''''''''''''''''''''''''''''''''
824
825This portion of the ``rcu_data`` structure is declared as follows:
826
827::
828
829     1   unsigned long gp_seq;
830     2   unsigned long gp_seq_needed;
831     3   bool cpu_no_qs;
832     4   bool core_needs_qs;
833     5   bool gpwrap;
834
835The ``->gp_seq`` field is the counterpart of the field of the same name
836in the ``rcu_state`` and ``rcu_node`` structures. The
837``->gp_seq_needed`` field is the counterpart of the field of the same
838name in the rcu_node structure. They may each lag up to one behind their
839``rcu_node`` counterparts, but in ``CONFIG_NO_HZ_IDLE`` and
840``CONFIG_NO_HZ_FULL`` kernels can lag arbitrarily far behind for CPUs in
841dyntick-idle mode (but these counters will catch up upon exit from
842dyntick-idle mode). If the lower two bits of a given ``rcu_data``
843structure's ``->gp_seq`` are zero, then this ``rcu_data`` structure
844believes that RCU is idle.
845
846+-----------------------------------------------------------------------+
847| **Quick Quiz**:                                                       |
848+-----------------------------------------------------------------------+
849| All this replication of the grace period numbers can only cause       |
850| massive confusion. Why not just keep a global sequence number and be  |
851| done with it???                                                       |
852+-----------------------------------------------------------------------+
853| **Answer**:                                                           |
854+-----------------------------------------------------------------------+
855| Because if there was only a single global sequence numbers, there     |
856| would need to be a single global lock to allow safely accessing and   |
857| updating it. And if we are not going to have a single global lock, we |
858| need to carefully manage the numbers on a per-node basis. Recall from |
859| the answer to a previous Quick Quiz that the consequences of applying |
860| a previously sampled quiescent state to the wrong grace period are    |
861| quite severe.                                                         |
862+-----------------------------------------------------------------------+
863
864The ``->cpu_no_qs`` flag indicates that the CPU has not yet passed
865through a quiescent state, while the ``->core_needs_qs`` flag indicates
866that the RCU core needs a quiescent state from the corresponding CPU.
867The ``->gpwrap`` field indicates that the corresponding CPU has remained
868idle for so long that the ``gp_seq`` counter is in danger of overflow,
869which will cause the CPU to disregard the values of its counters on its
870next exit from idle.
871
872RCU Callback Handling
873'''''''''''''''''''''
874
875In the absence of CPU-hotplug events, RCU callbacks are invoked by the
876same CPU that registered them. This is strictly a cache-locality
877optimization: callbacks can and do get invoked on CPUs other than the
878one that registered them. After all, if the CPU that registered a given
879callback has gone offline before the callback can be invoked, there
880really is no other choice.
881
882This portion of the ``rcu_data`` structure is declared as follows:
883
884::
885
886    1 struct rcu_segcblist cblist;
887    2 long qlen_last_fqs_check;
888    3 unsigned long n_cbs_invoked;
889    4 unsigned long n_nocbs_invoked;
890    5 unsigned long n_cbs_orphaned;
891    6 unsigned long n_cbs_adopted;
892    7 unsigned long n_force_qs_snap;
893    8 long blimit;
894
895The ``->cblist`` structure is the segmented callback list described
896earlier. The CPU advances the callbacks in its ``rcu_data`` structure
897whenever it notices that another RCU grace period has completed. The CPU
898detects the completion of an RCU grace period by noticing that the value
899of its ``rcu_data`` structure's ``->gp_seq`` field differs from that of
900its leaf ``rcu_node`` structure. Recall that each ``rcu_node``
901structure's ``->gp_seq`` field is updated at the beginnings and ends of
902each grace period.
903
904The ``->qlen_last_fqs_check`` and ``->n_force_qs_snap`` coordinate the
905forcing of quiescent states from ``call_rcu()`` and friends when
906callback lists grow excessively long.
907
908The ``->n_cbs_invoked``, ``->n_cbs_orphaned``, and ``->n_cbs_adopted``
909fields count the number of callbacks invoked, sent to other CPUs when
910this CPU goes offline, and received from other CPUs when those other
911CPUs go offline. The ``->n_nocbs_invoked`` is used when the CPU's
912callbacks are offloaded to a kthread.
913
914Finally, the ``->blimit`` counter is the maximum number of RCU callbacks
915that may be invoked at a given time.
916
917Dyntick-Idle Handling
918'''''''''''''''''''''
919
920This portion of the ``rcu_data`` structure is declared as follows:
921
922::
923
924     1   int dynticks_snap;
925     2   unsigned long dynticks_fqs;
926
927The ``->dynticks_snap`` field is used to take a snapshot of the
928corresponding CPU's dyntick-idle state when forcing quiescent states,
929and is therefore accessed from other CPUs. Finally, the
930``->dynticks_fqs`` field is used to count the number of times this CPU
931is determined to be in dyntick-idle state, and is used for tracing and
932debugging purposes.
933
934This portion of the rcu_data structure is declared as follows:
935
936::
937
938     1   long dynticks_nesting;
939     2   long dynticks_nmi_nesting;
940     3   atomic_t dynticks;
941     4   bool rcu_need_heavy_qs;
942     5   bool rcu_urgent_qs;
943
944These fields in the rcu_data structure maintain the per-CPU dyntick-idle
945state for the corresponding CPU. The fields may be accessed only from
946the corresponding CPU (and from tracing) unless otherwise stated.
947
948The ``->dynticks_nesting`` field counts the nesting depth of process
949execution, so that in normal circumstances this counter has value zero
950or one. NMIs, irqs, and tracers are counted by the
951``->dynticks_nmi_nesting`` field. Because NMIs cannot be masked, changes
952to this variable have to be undertaken carefully using an algorithm
953provided by Andy Lutomirski. The initial transition from idle adds one,
954and nested transitions add two, so that a nesting level of five is
955represented by a ``->dynticks_nmi_nesting`` value of nine. This counter
956can therefore be thought of as counting the number of reasons why this
957CPU cannot be permitted to enter dyntick-idle mode, aside from
958process-level transitions.
959
960However, it turns out that when running in non-idle kernel context, the
961Linux kernel is fully capable of entering interrupt handlers that never
962exit and perhaps also vice versa. Therefore, whenever the
963``->dynticks_nesting`` field is incremented up from zero, the
964``->dynticks_nmi_nesting`` field is set to a large positive number, and
965whenever the ``->dynticks_nesting`` field is decremented down to zero,
966the the ``->dynticks_nmi_nesting`` field is set to zero. Assuming that
967the number of misnested interrupts is not sufficient to overflow the
968counter, this approach corrects the ``->dynticks_nmi_nesting`` field
969every time the corresponding CPU enters the idle loop from process
970context.
971
972The ``->dynticks`` field counts the corresponding CPU's transitions to
973and from either dyntick-idle or user mode, so that this counter has an
974even value when the CPU is in dyntick-idle mode or user mode and an odd
975value otherwise. The transitions to/from user mode need to be counted
976for user mode adaptive-ticks support (see timers/NO_HZ.txt).
977
978The ``->rcu_need_heavy_qs`` field is used to record the fact that the
979RCU core code would really like to see a quiescent state from the
980corresponding CPU, so much so that it is willing to call for
981heavy-weight dyntick-counter operations. This flag is checked by RCU's
982context-switch and ``cond_resched()`` code, which provide a momentary
983idle sojourn in response.
984
985Finally, the ``->rcu_urgent_qs`` field is used to record the fact that
986the RCU core code would really like to see a quiescent state from the
987corresponding CPU, with the various other fields indicating just how
988badly RCU wants this quiescent state. This flag is checked by RCU's
989context-switch path (``rcu_note_context_switch``) and the cond_resched
990code.
991
992+-----------------------------------------------------------------------+
993| **Quick Quiz**:                                                       |
994+-----------------------------------------------------------------------+
995| Why not simply combine the ``->dynticks_nesting`` and                 |
996| ``->dynticks_nmi_nesting`` counters into a single counter that just   |
997| counts the number of reasons that the corresponding CPU is non-idle?  |
998+-----------------------------------------------------------------------+
999| **Answer**:                                                           |
1000+-----------------------------------------------------------------------+
1001| Because this would fail in the presence of interrupts whose handlers  |
1002| never return and of handlers that manage to return from a made-up     |
1003| interrupt.                                                            |
1004+-----------------------------------------------------------------------+
1005
1006Additional fields are present for some special-purpose builds, and are
1007discussed separately.
1008
1009The ``rcu_head`` Structure
1010~~~~~~~~~~~~~~~~~~~~~~~~~~
1011
1012Each ``rcu_head`` structure represents an RCU callback. These structures
1013are normally embedded within RCU-protected data structures whose
1014algorithms use asynchronous grace periods. In contrast, when using
1015algorithms that block waiting for RCU grace periods, RCU users need not
1016provide ``rcu_head`` structures.
1017
1018The ``rcu_head`` structure has fields as follows:
1019
1020::
1021
1022     1   struct rcu_head *next;
1023     2   void (*func)(struct rcu_head *head);
1024
1025The ``->next`` field is used to link the ``rcu_head`` structures
1026together in the lists within the ``rcu_data`` structures. The ``->func``
1027field is a pointer to the function to be called when the callback is
1028ready to be invoked, and this function is passed a pointer to the
1029``rcu_head`` structure. However, ``kfree_rcu()`` uses the ``->func``
1030field to record the offset of the ``rcu_head`` structure within the
1031enclosing RCU-protected data structure.
1032
1033Both of these fields are used internally by RCU. From the viewpoint of
1034RCU users, this structure is an opaque “cookie”.
1035
1036+-----------------------------------------------------------------------+
1037| **Quick Quiz**:                                                       |
1038+-----------------------------------------------------------------------+
1039| Given that the callback function ``->func`` is passed a pointer to    |
1040| the ``rcu_head`` structure, how is that function supposed to find the |
1041| beginning of the enclosing RCU-protected data structure?              |
1042+-----------------------------------------------------------------------+
1043| **Answer**:                                                           |
1044+-----------------------------------------------------------------------+
1045| In actual practice, there is a separate callback function per type of |
1046| RCU-protected data structure. The callback function can therefore use |
1047| the ``container_of()`` macro in the Linux kernel (or other            |
1048| pointer-manipulation facilities in other software environments) to    |
1049| find the beginning of the enclosing structure.                        |
1050+-----------------------------------------------------------------------+
1051
1052RCU-Specific Fields in the ``task_struct`` Structure
1053~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
1054
1055The ``CONFIG_PREEMPT_RCU`` implementation uses some additional fields in
1056the ``task_struct`` structure:
1057
1058::
1059
1060    1 #ifdef CONFIG_PREEMPT_RCU
1061    2   int rcu_read_lock_nesting;
1062    3   union rcu_special rcu_read_unlock_special;
1063    4   struct list_head rcu_node_entry;
1064    5   struct rcu_node *rcu_blocked_node;
1065    6 #endif /* #ifdef CONFIG_PREEMPT_RCU */
1066    7 #ifdef CONFIG_TASKS_RCU
1067    8   unsigned long rcu_tasks_nvcsw;
1068    9   bool rcu_tasks_holdout;
1069   10   struct list_head rcu_tasks_holdout_list;
1070   11   int rcu_tasks_idle_cpu;
1071   12 #endif /* #ifdef CONFIG_TASKS_RCU */
1072
1073The ``->rcu_read_lock_nesting`` field records the nesting level for RCU
1074read-side critical sections, and the ``->rcu_read_unlock_special`` field
1075is a bitmask that records special conditions that require
1076``rcu_read_unlock()`` to do additional work. The ``->rcu_node_entry``
1077field is used to form lists of tasks that have blocked within
1078preemptible-RCU read-side critical sections and the
1079``->rcu_blocked_node`` field references the ``rcu_node`` structure whose
1080list this task is a member of, or ``NULL`` if it is not blocked within a
1081preemptible-RCU read-side critical section.
1082
1083The ``->rcu_tasks_nvcsw`` field tracks the number of voluntary context
1084switches that this task had undergone at the beginning of the current
1085tasks-RCU grace period, ``->rcu_tasks_holdout`` is set if the current
1086tasks-RCU grace period is waiting on this task,
1087``->rcu_tasks_holdout_list`` is a list element enqueuing this task on
1088the holdout list, and ``->rcu_tasks_idle_cpu`` tracks which CPU this
1089idle task is running, but only if the task is currently running, that
1090is, if the CPU is currently idle.
1091
1092Accessor Functions
1093~~~~~~~~~~~~~~~~~~
1094
1095The following listing shows the ``rcu_get_root()``,
1096``rcu_for_each_node_breadth_first`` and ``rcu_for_each_leaf_node()``
1097function and macros:
1098
1099::
1100
1101     1 static struct rcu_node *rcu_get_root(struct rcu_state *rsp)
1102     2 {
1103     3   return &rsp->node[0];
1104     4 }
1105     5
1106     6 #define rcu_for_each_node_breadth_first(rsp, rnp) \
1107     7   for ((rnp) = &(rsp)->node[0]; \
1108     8        (rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
1109     9
1110    10 #define rcu_for_each_leaf_node(rsp, rnp) \
1111    11   for ((rnp) = (rsp)->level[NUM_RCU_LVLS - 1]; \
1112    12        (rnp) < &(rsp)->node[NUM_RCU_NODES]; (rnp)++)
1113
1114The ``rcu_get_root()`` simply returns a pointer to the first element of
1115the specified ``rcu_state`` structure's ``->node[]`` array, which is the
1116root ``rcu_node`` structure.
1117
1118As noted earlier, the ``rcu_for_each_node_breadth_first()`` macro takes
1119advantage of the layout of the ``rcu_node`` structures in the
1120``rcu_state`` structure's ``->node[]`` array, performing a breadth-first
1121traversal by simply traversing the array in order. Similarly, the
1122``rcu_for_each_leaf_node()`` macro traverses only the last part of the
1123array, thus traversing only the leaf ``rcu_node`` structures.
1124
1125+-----------------------------------------------------------------------+
1126| **Quick Quiz**:                                                       |
1127+-----------------------------------------------------------------------+
1128| What does ``rcu_for_each_leaf_node()`` do if the ``rcu_node`` tree    |
1129| contains only a single node?                                          |
1130+-----------------------------------------------------------------------+
1131| **Answer**:                                                           |
1132+-----------------------------------------------------------------------+
1133| In the single-node case, ``rcu_for_each_leaf_node()`` traverses the   |
1134| single node.                                                          |
1135+-----------------------------------------------------------------------+
1136
1137Summary
1138~~~~~~~
1139
1140So the state of RCU is represented by an ``rcu_state`` structure, which
1141contains a combining tree of ``rcu_node`` and ``rcu_data`` structures.
1142Finally, in ``CONFIG_NO_HZ_IDLE`` kernels, each CPU's dyntick-idle state
1143is tracked by dynticks-related fields in the ``rcu_data`` structure. If
1144you made it this far, you are well prepared to read the code
1145walkthroughs in the other articles in this series.
1146
1147Acknowledgments
1148~~~~~~~~~~~~~~~
1149
1150I owe thanks to Cyrill Gorcunov, Mathieu Desnoyers, Dhaval Giani, Paul
1151Turner, Abhishek Srivastava, Matt Kowalczyk, and Serge Hallyn for
1152helping me get this document into a more human-readable state.
1153
1154Legal Statement
1155~~~~~~~~~~~~~~~
1156
1157This work represents the view of the author and does not necessarily
1158represent the view of IBM.
1159
1160Linux is a registered trademark of Linus Torvalds.
1161
1162Other company, product, and service names may be trademarks or service
1163marks of others.
1164