Android: hdmi not suspended


I’m building a system with a hikey, a hdmi screen and Android running on it. It seems that when Android goes to sleep the screen is filled with black but does not enter sleep mode (no signal). I tried with another board that supports Android too, and the screen enters the sleep mode when Android goes to sleep.

Is there any way to change this behavior ?

Thanks for your help :slight_smile:

@damien, could you give us some details about the screen that you are using and the HiKey release? The kernel version would help as well please.

The screen is a Fujitsu B19-7 led (mac resolution 1280x1024) with a DVI-D and a VGA port.
The hikey is a 1GB model (is there any other information I can give about the board ?).
Android is running on top of a kernel 4.1.15 provided for AOSP (kernel/hikey-linaro - Git at Google), tag android-hikey-6.0.1_r0.1.

Yea. This is the expected behavior at the moment, as the drm layer doesn’t have the suspend/resume ops to trigger the DPMS modes.

I did have a patch to make it work last year, but it was a bit hackish and I never polished it up. I’ll take a look at reviving the patch, but you’ll likely have to rebuild AOSP/master to get it in the short term.

Oh Great !

I’m already building the AOSP from scratch so it’s not an issue. I’m really interested in your patch even if it’s a draft :wink:

Damien: Just fyi, spent some time on this last night and got the screen powering off, but unfortunately I’m not reliably getting it to come back on at the moment (at least not quickly enough).

I’ll continue working on it, but it might not be as fast a turnaround as I had hoped.

John: would it be ok for you to share the patch as is ? I’ll be interested to try working on it, at least try to understand what it does and maybe help you to revive it :slight_smile:

Damien: The current code (on top of the 4.4 AOSP hikey tree) is here:
john.stultz/android-dev.git - Android Development/Trivial Tree. WARNING: Branches may be rebased. dev/hdmi-suspend

I did have it working last year, but since then the drm/dsi driver was totally rewritten, so I’m not sure whats wrong now. Let me know if you find anything with it.

Hi John,

I built a kernel with your set of patches and I reached the state where it goes to sleep but almost never returns. Do you have any lead or documentation I could use to work on this ? I mean, I’m not familiar with DRM drivers but I can try to debug and do something :slight_smile: