aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2018-02-09 09:40:33 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2018-02-09 09:40:33 -0800
commita051c14b8db35cb269e9d91e11fc3573b6f7475d (patch)
tree35d603e7619b2a64d55fd9529519eac629b24173 /Documentation
parentMerge tag 'nfsd-4.16' of git://linux-nfs.org/~bfields/linux (diff)
parentMerge branches 'pm-cpufreq', 'pm-cpuidle' and 'pm-domains' (diff)
downloadlinux-dev-a051c14b8db35cb269e9d91e11fc3573b6f7475d.tar.xz
linux-dev-a051c14b8db35cb269e9d91e11fc3573b6f7475d.zip
Merge tag 'pm-part2-4.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm
Pull more power management updates from Rafael Wysocki: "These are mostly fixes and cleanups and removal of the no longer needed at32ap-cpufreq driver. Specifics: - Drop the at32ap-cpufreq driver which is useless after the removal of the corresponding arch (Corentin LABBE). - Fix a regression from the 4.14 cycle in the APM idle driver by making it initialize the polling state properly (Rafael Wysocki). - Fix a crash on failing system suspend due to a missing check in the cpufreq core (Bo Yan). - Make the intel_pstate driver initialize the hardware-managed P-state control (HWP) feature on CPU0 upon resume from system suspend if HWP had been enabled before the system was suspended (Chen Yu). - Fix up the SCPI cpufreq driver after recent changes (Sudeep Holla, Wei Yongjun). - Avoid pointer subtractions during frequency table walks in cpufreq (Dominik Brodowski). - Avoid the check for ProcFeedback in ST/CZ in the cpufreq driver for AMD processors and add a MODULE_ALIAS for cpufreq on ARM IMX (Akshu Agrawal, Nicolas Chauvet). - Fix the prototype of swsusp_arch_resume() on x86 (Arnd Bergmann). - Fix up the parsing of power domains DT data (Ulf Hansson)" * tag 'pm-part2-4.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm: arm: imx: Add MODULE_ALIAS for cpufreq cpufreq: Add and use cpufreq_for_each_{valid_,}entry_idx() cpufreq: intel_pstate: Enable HWP during system resume on CPU0 cpufreq: scpi: fix error return code in scpi_cpufreq_init() x86: hibernate: fix swsusp_arch_resume() prototype PM / domains: Fix up domain-idle-states OF parsing cpufreq: scpi: fix static checker warning cdev isn't an ERR_PTR cpufreq: remove at32ap-cpufreq cpufreq: AMD: Ignore the check for ProcFeedback in ST/CZ x86: PM: Make APM idle driver initialize polling state cpufreq: Skip cpufreq resume if it's not suspended
Diffstat (limited to 'Documentation')
-rw-r--r--Documentation/cpu-freq/cpu-drivers.txt4
1 files changed, 4 insertions, 0 deletions
diff --git a/Documentation/cpu-freq/cpu-drivers.txt b/Documentation/cpu-freq/cpu-drivers.txt
index 434c49cc7330..61546ac578d6 100644
--- a/Documentation/cpu-freq/cpu-drivers.txt
+++ b/Documentation/cpu-freq/cpu-drivers.txt
@@ -291,3 +291,7 @@ For example:
/* Do something with pos */
pos->frequency = ...
}
+
+If you need to work with the position of pos within driver_freq_table,
+do not subtract the pointers, as it is quite costly. Instead, use the
+macros cpufreq_for_each_entry_idx() and cpufreq_for_each_valid_entry_idx().