aboutsummaryrefslogtreecommitdiffstats
path: root/kernel/time/Kconfig
diff options
context:
space:
mode:
Diffstat (limited to 'kernel/time/Kconfig')
-rw-r--r--kernel/time/Kconfig27
1 files changed, 27 insertions, 0 deletions
diff --git a/kernel/time/Kconfig b/kernel/time/Kconfig
index c7d2fd67799e..3381f098070f 100644
--- a/kernel/time/Kconfig
+++ b/kernel/time/Kconfig
@@ -157,6 +157,33 @@ config NO_HZ_FULL_SYSIDLE
157 157
158 Say N if you are unsure. 158 Say N if you are unsure.
159 159
160config NO_HZ_FULL_SYSIDLE_SMALL
161 int "Number of CPUs above which large-system approach is used"
162 depends on NO_HZ_FULL_SYSIDLE
163 range 1 NR_CPUS
164 default 8
165 help
166 The full-system idle detection mechanism takes a lazy approach
167 on large systems, as is required to attain decent scalability.
168 However, on smaller systems, scalability is not anywhere near as
169 large a concern as is energy efficiency. The sysidle subsystem
170 therefore uses a fast but non-scalable algorithm for small
171 systems and a lazier but scalable algorithm for large systems.
172 This Kconfig parameter defines the number of CPUs in the largest
173 system that will be considered to be "small".
174
175 The default value will be fine in most cases. Battery-powered
176 systems that (1) enable NO_HZ_FULL_SYSIDLE, (2) have larger
177 numbers of CPUs, and (3) are suffering from battery-lifetime
178 problems due to long sysidle latencies might wish to experiment
179 with larger values for this Kconfig parameter. On the other
180 hand, they might be even better served by disabling NO_HZ_FULL
181 entirely, given that NO_HZ_FULL is intended for HPC and
182 real-time workloads that at present do not tend to be run on
183 battery-powered systems.
184
185 Take the default if you are unsure.
186
160config NO_HZ 187config NO_HZ
161 bool "Old Idle dynticks config" 188 bool "Old Idle dynticks config"
162 depends on !ARCH_USES_GETTIMEOFFSET && GENERIC_CLOCKEVENTS 189 depends on !ARCH_USES_GETTIMEOFFSET && GENERIC_CLOCKEVENTS