xref: /linux-6.15/kernel/rcu/Kconfig (revision 40c1278a)
1# SPDX-License-Identifier: GPL-2.0-only
2#
3# RCU-related configuration options
4#
5
6menu "RCU Subsystem"
7
8config TREE_RCU
9	bool
10	default y if SMP
11	help
12	  This option selects the RCU implementation that is
13	  designed for very large SMP system with hundreds or
14	  thousands of CPUs.  It also scales down nicely to
15	  smaller systems.
16
17config PREEMPT_RCU
18	bool
19	default y if PREEMPTION
20	select TREE_RCU
21	help
22	  This option selects the RCU implementation that is
23	  designed for very large SMP systems with hundreds or
24	  thousands of CPUs, but for which real-time response
25	  is also required.  It also scales down nicely to
26	  smaller systems.
27
28	  Select this option if you are unsure.
29
30config TINY_RCU
31	bool
32	default y if !PREEMPTION && !SMP
33	help
34	  This option selects the RCU implementation that is
35	  designed for UP systems from which real-time response
36	  is not required.  This option greatly reduces the
37	  memory footprint of RCU.
38
39config RCU_EXPERT
40	bool "Make expert-level adjustments to RCU configuration"
41	default n
42	help
43	  This option needs to be enabled if you wish to make
44	  expert-level adjustments to RCU configuration.  By default,
45	  no such adjustments can be made, which has the often-beneficial
46	  side-effect of preventing "make oldconfig" from asking you all
47	  sorts of detailed questions about how you would like numerous
48	  obscure RCU options to be set up.
49
50	  Say Y if you need to make expert-level adjustments to RCU.
51
52	  Say N if you are unsure.
53
54config SRCU
55	bool
56	help
57	  This option selects the sleepable version of RCU. This version
58	  permits arbitrary sleeping or blocking within RCU read-side critical
59	  sections.
60
61config TINY_SRCU
62	bool
63	default y if SRCU && TINY_RCU
64	help
65	  This option selects the single-CPU non-preemptible version of SRCU.
66
67config TREE_SRCU
68	bool
69	default y if SRCU && !TINY_RCU
70	help
71	  This option selects the full-fledged version of SRCU.
72
73config TASKS_RCU_GENERIC
74	def_bool TASKS_RCU || TASKS_RUDE_RCU || TASKS_TRACE_RCU
75	select SRCU
76	help
77	  This option enables generic infrastructure code supporting
78	  task-based RCU implementations.  Not for manual selection.
79
80config TASKS_RCU
81	def_bool 0
82	select IRQ_WORK
83	help
84	  This option enables a task-based RCU implementation that uses
85	  only voluntary context switch (not preemption!), idle, and
86	  user-mode execution as quiescent states.  Not for manual selection.
87
88config TASKS_RUDE_RCU
89	def_bool 0
90	select IRQ_WORK
91	help
92	  This option enables a task-based RCU implementation that uses
93	  only context switch (including preemption) and user-mode
94	  execution as quiescent states.  It forces IPIs and context
95	  switches on all online CPUs, including idle ones, so use
96	  with caution.
97
98config FORCE_TASKS_TRACE_RCU
99	bool "Force selection of Tasks Trace RCU"
100	depends on RCU_EXPERT
101	select TASKS_TRACE_RCU
102	default n
103	help
104	  This option enables a task-based RCU implementation that uses
105	  explicit rcu_read_lock_trace() read-side markers, and allows
106	  these readers to appear in the idle loop as well as on the
107	  CPU hotplug code paths.  It can force IPIs on online CPUs,
108	  including idle ones, so use with caution.  Not for manual
109	  selection in most cases.
110
111config TASKS_TRACE_RCU
112	bool
113	default n
114	select IRQ_WORK
115
116config RCU_STALL_COMMON
117	def_bool TREE_RCU
118	help
119	  This option enables RCU CPU stall code that is common between
120	  the TINY and TREE variants of RCU.  The purpose is to allow
121	  the tiny variants to disable RCU CPU stall warnings, while
122	  making these warnings mandatory for the tree variants.
123
124config RCU_NEED_SEGCBLIST
125	def_bool ( TREE_RCU || TREE_SRCU || TASKS_RCU_GENERIC )
126
127config RCU_FANOUT
128	int "Tree-based hierarchical RCU fanout value"
129	range 2 64 if 64BIT
130	range 2 32 if !64BIT
131	depends on TREE_RCU && RCU_EXPERT
132	default 64 if 64BIT
133	default 32 if !64BIT
134	help
135	  This option controls the fanout of hierarchical implementations
136	  of RCU, allowing RCU to work efficiently on machines with
137	  large numbers of CPUs.  This value must be at least the fourth
138	  root of NR_CPUS, which allows NR_CPUS to be insanely large.
139	  The default value of RCU_FANOUT should be used for production
140	  systems, but if you are stress-testing the RCU implementation
141	  itself, small RCU_FANOUT values allow you to test large-system
142	  code paths on small(er) systems.
143
144	  Select a specific number if testing RCU itself.
145	  Take the default if unsure.
146
147config RCU_FANOUT_LEAF
148	int "Tree-based hierarchical RCU leaf-level fanout value"
149	range 2 64 if 64BIT && !RCU_STRICT_GRACE_PERIOD
150	range 2 32 if !64BIT && !RCU_STRICT_GRACE_PERIOD
151	range 2 3 if RCU_STRICT_GRACE_PERIOD
152	depends on TREE_RCU && RCU_EXPERT
153	default 16 if !RCU_STRICT_GRACE_PERIOD
154	default 2 if RCU_STRICT_GRACE_PERIOD
155	help
156	  This option controls the leaf-level fanout of hierarchical
157	  implementations of RCU, and allows trading off cache misses
158	  against lock contention.  Systems that synchronize their
159	  scheduling-clock interrupts for energy-efficiency reasons will
160	  want the default because the smaller leaf-level fanout keeps
161	  lock contention levels acceptably low.  Very large systems
162	  (hundreds or thousands of CPUs) will instead want to set this
163	  value to the maximum value possible in order to reduce the
164	  number of cache misses incurred during RCU's grace-period
165	  initialization.  These systems tend to run CPU-bound, and thus
166	  are not helped by synchronized interrupts, and thus tend to
167	  skew them, which reduces lock contention enough that large
168	  leaf-level fanouts work well.  That said, setting leaf-level
169	  fanout to a large number will likely cause problematic
170	  lock contention on the leaf-level rcu_node structures unless
171	  you boot with the skew_tick kernel parameter.
172
173	  Select a specific number if testing RCU itself.
174
175	  Select the maximum permissible value for large systems, but
176	  please understand that you may also need to set the skew_tick
177	  kernel boot parameter to avoid contention on the rcu_node
178	  structure's locks.
179
180	  Take the default if unsure.
181
182config RCU_BOOST
183	bool "Enable RCU priority boosting"
184	depends on (RT_MUTEXES && PREEMPT_RCU && RCU_EXPERT) || PREEMPT_RT
185	default y if PREEMPT_RT
186	help
187	  This option boosts the priority of preempted RCU readers that
188	  block the current preemptible RCU grace period for too long.
189	  This option also prevents heavy loads from blocking RCU
190	  callback invocation.
191
192	  Say Y here if you are working with real-time apps or heavy loads
193	  Say N here if you are unsure.
194
195config RCU_BOOST_DELAY
196	int "Milliseconds to delay boosting after RCU grace-period start"
197	range 0 3000
198	depends on RCU_BOOST
199	default 500
200	help
201	  This option specifies the time to wait after the beginning of
202	  a given grace period before priority-boosting preempted RCU
203	  readers blocking that grace period.  Note that any RCU reader
204	  blocking an expedited RCU grace period is boosted immediately.
205
206	  Accept the default if unsure.
207
208config RCU_NOCB_CPU
209	bool "Offload RCU callback processing from boot-selected CPUs"
210	depends on TREE_RCU
211	depends on RCU_EXPERT || NO_HZ_FULL
212	default n
213	help
214	  Use this option to reduce OS jitter for aggressive HPC or
215	  real-time workloads.	It can also be used to offload RCU
216	  callback invocation to energy-efficient CPUs in battery-powered
217	  asymmetric multiprocessors.  The price of this reduced jitter
218	  is that the overhead of call_rcu() increases and that some
219	  workloads will incur significant increases in context-switch
220	  rates.
221
222	  This option offloads callback invocation from the set of CPUs
223	  specified at boot time by the rcu_nocbs parameter.  For each
224	  such CPU, a kthread ("rcuox/N") will be created to invoke
225	  callbacks, where the "N" is the CPU being offloaded, and where
226	  the "x" is "p" for RCU-preempt (PREEMPTION kernels) and "s" for
227	  RCU-sched (!PREEMPTION kernels).  Nothing prevents this kthread
228	  from running on the specified CPUs, but (1) the kthreads may be
229	  preempted between each callback, and (2) affinity or cgroups can
230	  be used to force the kthreads to run on whatever set of CPUs is
231	  desired.
232
233	  Say Y here if you need reduced OS jitter, despite added overhead.
234	  Say N here if you are unsure.
235
236config TASKS_TRACE_RCU_READ_MB
237	bool "Tasks Trace RCU readers use memory barriers in user and idle"
238	depends on RCU_EXPERT && TASKS_TRACE_RCU
239	default PREEMPT_RT || NR_CPUS < 8
240	help
241	  Use this option to further reduce the number of IPIs sent
242	  to CPUs executing in userspace or idle during tasks trace
243	  RCU grace periods.  Given that a reasonable setting of
244	  the rcupdate.rcu_task_ipi_delay kernel boot parameter
245	  eliminates such IPIs for many workloads, proper setting
246	  of this Kconfig option is important mostly for aggressive
247	  real-time installations and for battery-powered devices,
248	  hence the default chosen above.
249
250	  Say Y here if you hate IPIs.
251	  Say N here if you hate read-side memory barriers.
252	  Take the default if you are unsure.
253
254endmenu # "RCU Subsystem"
255