v4.18-rc2 on droid4: screen turns red

After longer inactivity, attempt to wake up the screen results in screen being completely red. Not too useful. I can still connect over ssh.

This happens all the time:

Jun 30 11:34:08 devuan kernel: [  143.605499] ------------[ cut here
]------------
Jun 30 11:34:08 devuan kernel: [  143.610168] WARNING: CPU: 1 PID:
2220 at drivers/gpu/drm/omapdrm/
dss/hdmi4.c:667 hdmi_audio_stop+0x5c/0x68
Jun 30 11:34:08 devuan kernel: [  143.619873] Modules linked in:
Jun 30 11:34:08 devuan kernel: [  143.622955] CPU: 1 PID: 2220 Comm:
pulseaudio Not tainted 4.18.0-
rc2-88315-g3452793-dirty #756
Jun 30 11:34:08 devuan kernel: [  143.631591] Hardware name: Generic
OMAP4 (Flattened Device Tree)
Jun 30 11:34:08 devuan kernel: [  143.637634] [<c010f6cc>]
(unwind_backtrace) from [<c010ba50>] (sh
ow_stack+0x10/0x14)
Jun 30 11:34:08 devuan kernel: [  143.645446] [<c010ba50>]
(show_stack) from [<c086d7ec>] (dump_sta
ck+0x8c/0xac)
Jun 30 11:34:08 devuan kernel: [  143.652740] [<c086d7ec>]
(dump_stack) from [<c012d364>] (__warn+0
xe8/0x114)
Jun 30 11:34:08 devuan kernel: [  143.659729] [<c012d364>] (__warn)
from [<c012d3cc>] (warn_slowpath_null+0x3c/0x48)
Jun 30 11:34:08 devuan kernel: [  143.667358] [<c012d3cc>]
(warn_slowpath_null) from [<c047b7dc>] (hdmi_audio_stop+0x5c/0x68)
Jun 30 11:34:08 devuan kernel: [  143.675750] [<c047b7dc>]
(hdmi_audio_stop) from [<c06a9824>] (hdmi_dai_trigger+0x6c/0x88)
Jun 30 11:34:08 devuan kernel: [  143.683990] [<c06a9824>]
(hdmi_dai_trigger) from [<c06990d0>] (soc_pcm_trigger+0xf0/0x130)
Jun 30 11:34:08 devuan kernel: [  143.692291] [<c06990d0>]
(soc_pcm_trigger) from [<c067a9f4>] (snd_pcm_do_stop+0x54/0x58)
Jun 30 11:34:08 devuan kernel: [  143.700439] [<c067a9f4>]
(snd_pcm_do_stop) from [<c067a7a4>] (snd_pcm_action_single+0x38/0x78)
Jun 30 11:34:08 devuan kernel: [  143.709075] [<c067a7a4>]
(snd_pcm_action_single) from [<c067c468>] (snd_pcm_drop+0x48/0x94)
Jun 30 11:34:08 devuan kernel: [  143.717468] [<c067c468>]
(snd_pcm_drop) from [<c067d508>] (snd_pcm_ioctl+0x870/0x1180)
Jun 30 11:34:08 devuan kernel: [  143.725433] [<c067d508>]
(snd_pcm_ioctl) from [<c0215944>] (do_vfs_ioctl+0x80/0x960)
Jun 30 11:34:08 devuan kernel: [  143.733215] [<c0215944>]
(do_vfs_ioctl) from [<c0216280>] (ksys_ioctl+0x5c/0x80)
Jun 30 11:34:08 devuan kernel: [  143.740661] [<c0216280>]
(ksys_ioctl) from [<c0101000>] (ret_fast_syscall+0x0/0x54)
Jun 30 11:34:08 devuan kernel: [  143.748352] Exception
stack(0xeb8fffa8 to 0xeb8ffff0)
Jun 30 11:34:08 devuan kernel: [  143.753448] ffa0:

005129b0 b1971394 00000015 00004143 00000000 00512960

This is probably unrelated:

[13215.125732] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13245.205535] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13252.969940] power_supply battery: driver failed to report
`power_now' property: -22
[13275.285369] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13291.993652] Charging, 4350000 uV, 532000 uA
[13292.001708] Charging, 4350000 uV, 532000 uA
[13305.364837] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13335.445037] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13365.524291] cpcap-usb-phy cpcap-usb-phy.0: connected to USB host
[13372.940093] power_supply battery: driver failed to report
`power_now' property: -22
[13383.554412] g_ether gadget: packet filter 0e
[13383.558715] g_ether gadget: ecm req21.43 v000e i0000 l0
[13383.580261] g_ether gadget: notify connect true
[13383.612243] g_ether gadget: notify speed 425984000
[13387.213256] cpcap_adc cpcap_adc.0: error reading ADC: -110
[13387.218780] cpcap_battery cpcap_battery.0:
cpcap_battery_get_voltage failed: -110
[13387.523254] cpcap_adc cpcap_adc.0: error reading ADC: -110
[13387.528778] cpcap_battery cpcap_battery.0:

cpcap_battery_get_current failed: -110

Ideas welcome.

Best regards,
Pavel

Hi!

Quoted 3 lines

After longer inactivity, attempt to wake up the screen results in screen being completely red. Not too useful. I can still connect over ssh.

Still there in -rc4 :-(. But happens only once in a few days for me...

When X is shut down, text console works ok.

If I get my spare droid4 to boot again I'll send it to Prague so you can compare & see if it is somewhat deterministic ; )

Hi!

Quoted 3 lines

After longer inactivity, attempt to wake up the screen results in screen being completely red. Not too useful. I can still connect over ssh.

Still there in -rc4 :-(. But happens only once in a few days for me...

When X is shut down, text console works ok.

Quoted 7 lines

Quoted 3 lines

After longer inactivity, attempt to wake up the screen results in screen being completely red. Not too useful. I can still connect over ssh.

Still there in -rc4 :-(. But happens only once in a few days for me...

When X is shut down, text console works ok.

FYI, I've been chasing an occasional hang with lapdock and it might be related to the usb phy switching between usb and uart mode spontaneously. And then the usb traffic can cause a sysrq trigger potentially when the glitch happens. Not sure yet why it happens, it might be related to lost cpcap interrupts still.

Anyways, so far looks like doing the following may have prevented it. Well at least for about four days of uptime now:

# echo 0 > /proc/sys/kernel/sysrq

And no red screens here yet either. Care to check if this might somehow also prevent your red screen issue?