The command only worked when running it inside the git repository.
Now, it works from everywhere, and the version is constructed using
the latest git tag and the SHA of the latest commit, which may help
with further debugging in future issues.
* Ability to reset scaling frequencies to default
If one of the scaling frequencies option is missing, either because
it's commented out or auto-cpufreq.conf is not being used, the default
frequency limits will be set.
TODO: perform setting only once per battery state change
* Ability to reset scaling frequencies to default 2.
Additional changes to optimize auto-cpufreq speed, by conditionally
performing frequency scaling settings.
* Add min/max allowed frequencies option (#309)
Setting the minimum/maximum allowed CPU frequencies is now possible
via 'scaling_min_freq' and 'scaling_max_freq' options in config
for both modes (charger and battery).
Values should be given in kHZ.
Example:
scaling_min_freq = 1800000
scaling_max_freq = 2000000
I also did autoformatting via 'black', using --line-length 100.
Code is now a bit tidier than before.
* Fixes for min/max allowed frequencies option
* Further modify min/max frequency option
* Fix temporary modification in min/max allowed frequency option
Co-authored-by: Akos Varady <akos.varady@ericsson.com>
* Started the solution to Snap and PopOS not working
Needs Snap detection on line 102/103
This code does NOT function properly unless that detection added
Added ability for the program to resume if the PopOS fix works
* Implemented changes as suggested by bobslept
* Implement python virtual environment
* added venv instead of using system env pip
* adjusted the unit file to startup the app from the venv
* created a wrapper script to call the app from the venv
* extended cleanup for venv and additional scripts
* refactored the is_running() function to find the process now that it
is called from a venv
* remove update_service_file since we changed the binary path to the venv location
* fix bug in argument handling; upgrade pip before installing python packages
* fix bug in syntax
* Renamed auto-cpufreq wrapper
* Change permissions to use --stats as user
* Changed init scripts to use wrapper
Co-authored-by: aroundthfur <velimir@foolcontrol.org>
* Fix for PopOS os_release problem
Fixes issue #276
* Added Pop!_OS detection
* Removed redundant abort code in Pop!_OS detection
* Added exit case for pressing 'N' or not pressing anything
Also fixed weird whitespace issue with backslash
* Changed to "Operation aborted by user" for 'N'
* Update auto_cpufreq/core.py
Co-authored-by: bobslept <38557801+bobslept@users.noreply.github.com>
Co-authored-by: bobslept <38557801+bobslept@users.noreply.github.com>
* check if charging in laptop with battery
* using both battery and ac adapter to check if charging
also battery status unknown is now recognized as charging
* daemon running msg was wrong (it was printing the install message). Fixed to a daemon running specific message.
* move footer after the daemon check func
* Add config file for setting preferred governor and turbo boost settings.
* Add an example config file.
* Display notice when using settings from config file when starting
* Add section about configuring to README.md
* Add --config parameter to set path for configuration file to use
```
❯ pacman -Qs auto-cpufreq
local/auto-cpufreq-git 1.6.0.r0.g0ce8311-1
Automatic CPU speed & power optimizer
```
With wildcard:
```
auto-cpufreq version:
error: package 'auto-cpufreq*' was not found
```
```
❯ pacman -Qi auto-cpufreq* | grep Version
zsh: no matches found: auto-cpufreq*
```
Without wildcard:
```
auto-cpufreq version:
Version : 1.6.0.r0.g0ce8311-1
```
```
❯ pacman -Qi auto-cpufreq | grep Version
Version : 1.6.0.r0.g0ce8311-1
```
It turns out it only works with the wildcard in my `pkgbuilds` directory where `auto-cpufreq-git` lives. Weird.
* Ignore Pycharm descriptors from VCS.
* Rename all occurances of `log` to `stats`.
* Rename log file to stats and move to /var/run
* Bring back --log switch with deprecation message.
If it detects the device is a laptop, it should default to the regular method of determining ac_state
If the device returns any other output, it should default to ac_state=true
continued work originally started in #158
