Blank screen issue

[xandy]

Some additional info from my side - I’ve written a system image to an SD card and tried to boot from it, just in case, but it’s still blank.

While I can’t be 100% sure what was going on I suppose it did boot from the card as SSH was not coming up.

[stephano]

@habasud

Here’s my dmesg output:

Thanks for that. Looks like we are dealing with a separate issue from @xandy.

I’ve changed some panel timings in the driver and the blank screen seems to be happening less now. I haven’t seen a blank screen in over a dozen reboots. The moment I swap back to the old driver the blank screen returns (after a reboot or two).

I’m going to do some more testing over the next week to determine if there’s a simple fix. Ideally some small driver timing tweaks can make the panel more reliable.

[stephano]

For the issue @habasud and I are seeing, I created a patch to attempt to get some debugging info out of the panel driver. The patch does seem to improve the behavior (how often the panel goes blank) but does not fix the issue.

I created an issue for tracking my progress on debugging this, however because the issue is so easily resolved with a power cycle of the system controller, I can’t say I’ll be spending a ton of time on this until after I’ve ported the reform-tools to Arch.

[xandy]

Hello @wakest!

Is there perhaps any development on how to proceed with this?

Thank you!

[aaa]

Don’t think it has been mentioned here. But my way to deal with persistent black screen when booting is to toggle off the power switch on the side, plug in the charger and then toggle the power switch on. Gives it a little jolt that seems to put things back in order. No idea if it’ll work for anyone else. I have to do it after every hardware change though.

[stephano]

Hey @aaa, thanks for your response.

Which hardware / OS are you using (i.MX8M Plus + stock Debian install + sd card / nvme / flash)?

@aaa

I have to do it after every hardware change though.

Not sure what you mean by “hardware change” but could you give us some details on how often it happens?

For example, I can say that I see this behavior at least every 10 reboots, and once it happens it requires at least 1 hard powerdown with the switch or OLED menu to make it work again.

[wakest]

Sorry it took so long for me to get back to you! I consulted my coworkers here at MNT and we agreed you should send it in for a warranty repair here at our studio. I see you also already started an email exchange with our support email so we can continue from there

[aaa]

I just started to use my Pocket again after a hiatus, so I don’t know. It happened a lot back when I had the brownout/sudden restarts problem. But that seems to be almost solved now.

Yesterday when I installed new parts I got the problem. But I think it happened back when I upgraded the batteries as well so could be every time I unplug the batteries. Anyway I have little useful data for this thread currently.

Modified and upgraded rk3588 Pocket with modified (KDE converted) Debian standard install.

[xandy]

Hi, thanks for the tip!

Don’t know if I’ve tried this before but did now and it didn’t help in my case unfortunately.

[wakest]

If you unplug the batteries you can’t turn on the device without installing a jumper. Can you explain in more detail the steps you are taking before you are having the screen boot up blank?

[wakest]

A detail I missed until seeing your most recent email, you mention having installed different batteries, have you tried putting the original batteries back in the device and seeing if the black screen issue persists? First flick off the security power switch on the side, install the old batteries and the maybe switch the switch back and forth a few times and then try to turn it on (while also plugged in to USB C PD power.

[xandy]

Yes I did try with the original batteries, but no change. Everything but the screen works with either set of batteries.

[wakest]

Did this issue start after you tried the new batteries or was it already sometimes happening before you tried them? Just thinking if there is some residual charge happening somewhere.

[xandy]

The issue was already happening, it’s been occasional but noticeable since at least about summer and I got the batteries in October I think. And then it was another ~4 months with the new batteries until the frequency of the issue started increasing rapidly.

[habasud]

In my case I think I can say, that most of the time that “blank screen issue” happens, when I boot with the power supply connected. I used my pocket on battery for some time and everthing was o.k. Today I had some blank screens while starting it on the wire. I’ll observe…

Edit: It might also be, that an additionaly connected usb audio IF causes the behaviour

[lislegaard]

I just wanted to chime in that I have also had this happen a few times on my new Pocket with RK3588. SSH still works it is only the screen that is blank with the backlight on. It seems to only happen when the charger is in and not every time.

[lislegaard]

Adding to this, yesteday this happened without charger, but with an USB keyboard attached, so it might be more related to USB rather than actual charging.

[stephano]

@lislegaard could you post your dmesg output?

My guess is that you have the Wisecoco (huai zhi) panel. If so that’s a driver issue we are working on fixing.

You can see the latest update here:

Pocket screen off after idle and can't get on again?

Quick update on the display issue. I had seen this bug report, but my panel’s issue seemed different. A complete failure of the display, not 1/2 a display. @minute’s comment here peaked my interest. I thought I had tried this by hacking the firmware’s call to “turn_som_power_on” to call the reset even on warm boot. Since that didn’t work I assumed it wasn’t the issue. But looking into the call stack, that function isn’t invoked on warm boot. I hadn’t thought about wiring the reset into the …

[lislegaard]

@stephano

My guess is that you have the Wisecoco (huai zhi) panel.

Yeah looks like it right? I am a bit noob, but I think these are the relevant lines?

sudo dmesg | grep -iE "panel|drm|display|wisecoco"
[    0.172131] /dsi@fde30000: Fixed dependency cycle(s) with /dsi@fde30000/panel
[    0.172227] /dsi@fde30000/panel: Fixed dependency cycle(s) with /dsi@fde30000
[    2.496970] /dsi@fde30000: Fixed dependency cycle(s) with /dsi@fde30000/panel
[    2.497035] /dsi@fde30000/panel: Fixed dependency cycle(s) with /dsi@fde30000
[    2.497138] panel-mnt-pocket-reform fde30000.dsi.0: DSI eot/hfp/hbp/hsa disabled via device tree
[    2.497163] panel-mnt-pocket-reform fde30000.dsi.0: supply vddp not found, using dummy regulator
[    2.497249] panel-mnt-pocket-reform fde30000.dsi.0: supply iovcc not found, using dummy regulator
[    2.497366] panel-mnt-pocket-reform fde30000.dsi.0: error -ENOENT: cannot get reset-gpios 0
[    2.497386] panel-mnt-pocket-reform fde30000.dsi.0: error -ENOENT: cannot get dcdc-en-gpio 0
[    2.770318] rockchip-drm display-subsystem: bound fdd90000.vop (ops vop2_component_ops [rockchipdrm])
[    2.771338] rockchip-drm display-subsystem: bound fdea0000.hdmi (ops dw_hdmi_qp_rockchip_ops [rockchipdrm])
[    2.771411] rockchip-drm display-subsystem: bound fde30000.dsi (ops dw_mipi_dsi2_rockchip_ops [rockchipdrm])
[    2.809894] [drm] Initialized rockchip 1.0.0 for display-subsystem on minor 0
[    2.849831] panel-mnt-pocket-reform fde30000.dsi.0: [display] init in prepare...
[    2.873812] panel-mnt-pocket-reform fde30000.dsi.0: [mnt pocket reform display] read register 12a: 0x0
[    2.897783] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] not sending tables this time
[    2.997821] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] OK
[    3.077902] rockchip-drm display-subsystem: [drm] fb0: rockchipdrmfb frame buffer device
[    3.153901] panel-mnt-pocket-reform fde30000.dsi.0: [display] init in prepare...
[    3.204558] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] sent tables OK
[    3.311201] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] OK
[    6.047103] systemd[1]: modprobe@drm.service - Load Kernel Module drm skipped, unmet condition check ConditionKernelModuleLoaded=!drm
[    7.920300] panthor fb000000.gpu: [drm] clock rate = 280000000
[    7.962023] panthor fb000000.gpu: [drm] Mali-G610 id 0xa867 major 0x0 minor 0x0 status 0x5
[    7.962030] panthor fb000000.gpu: [drm] Features: L2:0x7120306 Tiler:0x809 Mem:0x301 MMU:0x2830 AS:0xff
[    7.962033] panthor fb000000.gpu: [drm] shader_present=0x50005 l2_present=0x1 tiler_present=0x1
[    7.995027] panthor fb000000.gpu: [drm] Firmware protected mode entry is not supported, ignoring
[    7.997918] panthor fb000000.gpu: [drm] Firmware git sha: 95a25d71030715381f33105394285e1dcc860a65
[    8.000654] panthor fb000000.gpu: [drm] CSF FW using interface v1.5.0, Features 0x0 Instrumentation features 0x71
[    8.003519] panthor fb000000.gpu: [drm] Using Transparent Hugepage
[    8.042678] [drm] Initialized panthor 1.8.0 for fb000000.gpu on minor 1
[    8.908302] reform2_lpc spi1.0: enabling backlight control for MNT Pocket Reform with Display Version 2.

[stephano]

Yes, that’s the V2 Panel:

[3.204558] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] sent tables OK
[3.311201] panel-mnt-pocket-reform fde30000.dsi.0: [display v2] OK

Keep an eye on that thread. I’ll post again once we’ve worked out a way to split out the drivers and update the firmware.