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