Possible fix for some suspend issues

[Dwayne Slater]

On most Linux distros ath9k is disabled before suspend when doing pm-suspend. Android x86 does not do the same. This is most likely the cause of most suspend/resume freezes that most users experience. On my Acer Aspire v3-551 after disabling ath9k (using rmmod) and suspending I've observed a 100% suspend success rate, while enabled I've observed a 0% suspend success rate (on 6.0.0-rc2, 6.0.0-rc1 had more of a 50-60% success rate). I have no experience with building Android or how the code is structured but a rmmod on suspend and a modprobe on resume would resolve most problems.

[Chih-Wei Huang]

2016-08-30 1:29 GMT+08:00 Dwayne Slater <ds8...@gmail.com>:
> On most Linux distros ath9k is disabled before suspend when doing pm-suspend. Android x86 does not do the same. This is most likely the cause of most suspend/resume freezes that most users experience. On my Acer Aspire v3-551 after disabling ath9k (using rmmod) and suspending I've observed a 100% suspend success rate, while enabled I've observed a 0% suspend success rate (on 6.0.0-rc2, 6.0.0-rc1 had more of a 50-60% success rate). I have no experience with building Android or how the code is structured but a rmmod on suspend and a modprobe on resume would resolve most problems.

Thank you for the suggestion.
We'll try to integrate it.


--
Chih-Wei
Android-x86 project
http://www.android-x86.org

[Chih-Wei Huang]

I've pushed a patch to unload/reload
wifi driver on suspend/resume

https://sourceforge.net/p/android-x86/hardware_x86power/ci/338e1f465738acec2d2ca5f9ef2d390456a5d752/

Please test it.

On my Acer ES131 with ath9k driver,
it works as expected.
(though resuming on this device also work without this patch)

[Dwayne Slater]

Thank you, I'll test it later today

[Dwayne Slater]

If it isn't a convenience to you can you send me a patched version of the .so for Android 6.0?