Drop the default for performance_cx_lowest (ie, what to use when AC is connected) to ACPI C2.

ACPI C3 ends up doing a lot more work before entering sleep, some of which
requires grabbing a global ACPI hardware serialising mutex.

Because of this, the more CPU cores you have, the more that lock contends
under load, reaching close to the #1 lock contention (after VM, which is being
worked on.)

Tested:

* Sandy bridge Xeon, 2 socket * 8 core
* Ivy bridge Xeon v2, 2 socket * 8 core
* Westmere-EX, 4 socket * 10 core
* Ivybridge desktop
* Sandybridge mobile
* Ivybridge mobile

MFC after:	2 weeks
This commit is contained in:
Adrian Chadd 2015-04-27 19:18:51 +00:00
parent b3ff02bf85
commit 709c1a78cf
Notes: svn2git 2020-12-20 02:59:44 +00:00
svn path=/head/; revision=282110

View File

@ -648,7 +648,7 @@ devfs_system_ruleset="" # The name (NOT number) of a ruleset to apply to /dev
devfs_set_rulesets="" # A list of /mount/dev=ruleset_name settings to
# apply (must be mounted already, i.e. fstab(5))
devfs_load_rulesets="YES" # Enable to always load the default rulesets
performance_cx_lowest="Cmax" # Online CPU idle state
performance_cx_lowest="C2" # Online CPU idle state
performance_cpu_freq="NONE" # Online CPU frequency
economy_cx_lowest="Cmax" # Offline CPU idle state
economy_cpu_freq="NONE" # Offline CPU frequency