That reminds me of something else, actually. Although I have enabled your PPA, I am still using tlp instead of power-profiles-daemon. Now I wonder if that’s significant enough to cause this issue…
It’s way past bedtime here, but I will give it a try tomorrow. Thanks.
OK, I had a chance to test this now, but unfortunately the problem persists.
I installed @Mario_Limonciello’s power-profiles-daemon, set /sys/class/drm/card0/device/power_dpm_force_performance_level to high, but nothing works (except my USB-C hub).
Interestingly, xrandr reports the following:
Screen 0: minimum 320 x 200, current 2256 x 1504, maximum 16384 x 16384
eDP connected 2256x1504+0+0 (normal left inverted right x axis y axis) 285mm x 190mm
2256x1504 60.00*+ 48.00
1920x1200 60.00
1920x1080 60.00
1600x1200 60.00
1680x1050 60.00
1280x1024 60.00
1440x900 60.00
1280x800 60.00
1280x720 60.00
1024x768 60.00
800x600 60.00
640x480 60.00
DisplayPort-0 disconnected (normal left inverted right x axis y axis)
DisplayPort-1 disconnected (normal left inverted right x axis y axis)
DisplayPort-2 disconnected (normal left inverted right x axis y axis)
DisplayPort-3 disconnected (normal left inverted right x axis y axis)
DisplayPort-4 disconnected (normal left inverted right x axis y axis)
DisplayPort-5 disconnected (normal left inverted right x axis y axis)
DisplayPort-6 disconnected (normal left inverted right x axis y axis)
DisplayPort-7 disconnected (normal left inverted right x axis y axis)
I took the output from lsusb and found that the HDMI adapter is plugged into /sys/bus/usb/devices/1-1/, so I did:
echo on | /sys/bus/usb/devices/1-1/power/control
It seems that this is all that’s needed to disable runtime PM on the USB interface? But unfortunately it didn’t work. I removed the HDMI cable and plugged it in again, to no avail.
I have a similar problem: display not detected.
With a FW16 AMD and display port expansion card.
Is there a debug setting for the amdgpu linux driver that will log messages about connect/disconnect events so one can diagnose this further? I am wondering if the cause might be swing or pre-emphasis negotiation not working well.
So, debug showing this negotiation would be helpful.
After some more testing with the live images, I was (accidentally) able to come up with a way to make the card work. If I remove the HDMI adapter and insert it again (with the computer still running), my monitor is recognized and things seem to work. This workaround worked with the Ubuntu 24.04 live image, but also with my Ubuntu Jammy system.
I did not have this problem with my previous Intel 12th gen configuration, so I’m thinking about opening an official ticket/support request with Framework.
FWIW, after a long exchange of emails with Framework support, it seems like the problem may be the fact that I’m using a 1st Gen HDMI adapter. At least we couldn’t find anything else that can explain what’s going on, and why disconnecting and reconnecting the adapter after boot fixes the issue.
One user seemed to think a kernel caused the module to get into a weird state or perhaps a certain controller on the mainboard. So unsure. Could it be preboot environment is doing something? I only have a 3rd gen one.
Edit: Wondering if you’ve done a full board reset?
OK, my 3rd Gen HDMI adapter arrived and I was able to confirm my suspicions: it works as expected.
My take from all this saga is this: 1st Gen HDMI adapters may not work with AMD Ryzen mainboards. Be careful if you’re upgrading your machine and have a 1st Gen HDMI adapter; it’s best to go for the 3rd Gen.
@parawizard, regarding the full board reset: yeah, I had done it as part of the troubleshooting I did with the Framework support team. It didn’t help.
The gen 1 module works fine if replugged after booting though right? I’m guessing pre Linux environment is putting it into a weird state. Seems like something framework will fix eventually if so.