Re: [PATCH v1 3/4] acpi/x86: s2idle: call screen on and off as part of callbacks

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



> I will be here waiting for the result, when you have identified the reason of regression I
documented, include me in CC or reach me out privately with the work
ready to test.

Hi Dennis, as I said yesterday I would continue testing and I did. You
were right, there are two pieces to this puzzle.

Let's make something clear. Asus is not the only one using _DSM 3,4
this way. Asus is just the case where the deviation in behavior is
more clear. I suspect OneXplayer is using it as well for the
controller (USBP variable). And Ayaneo is notorious for unreliable
S0ix to the point where people recommend bios mods to manually force
S3 through PBS.

First one is that calling Display Off/On while the kernel is partly
suspended prevents the controller from coming back online. My patch
fixes that, and I am sure you can attest.

Second one is that asus plays an RGB suspend animation before
suspending the controller, which lasts for over 1 second. If the
device suspends before that, when it wakes up, it keeps playing the
animation the disconnection happens then. Since XInput does not
support selective suspend, this might lead to higher battery draw and,
as you noticed, being able to wake up the device with the controller.
For the Ally X, it means the controller takes around 2s more to wake
up (in addition to the animation quirk).

Thank you for catching that. I will update the patch series to take
that into account. If you want I can add a Reported-by for you, and if
you test the next revision out-of-band, I can also add Tested-by. In
addition, if you give me some details about the next firmware asus is
preparing, I can make sure the (pre-suspend) delay only applies to the
older firmware. There is no need for a delay post-suspend. If Asus is
removing the CSEE call for backwards compatibility, this might also
work.

I still think there is value to this patch, as it tries to study
Modern Standby behavior further. And I would suspect Asus is not the
only manufacturer being creative.

> do a request for comment on a proposal,

Yes, that will come after some time as well, but it will build on this patch.

Antheas




[Index of Archives]     [Linux Kernel Development]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux