system76
darp10: Add Darter Pro 10
- https://github.com/system76/coreboot/pull/219
- https://github.com/system76/ec/pull/451
- https://github.com/system76/firmware-update/pull/104
I mentioned this in the firmware issue as well, as it looks to impact both that PR, as well as firmware built from this PR.
While certifying the system, it appears that HDMI is not functioning. *randr doesn't indicate the presence of a display and there is no evidence in logs of a connection being detected. Several monitors and cable were testing, suggesting it is an actual Darp10 issue.
The HDMI is now correctly outputting.
However, I have run into the following issue around suspend/resume:
- I have the laptop, with the lid open, before me.
- I attach an HDMI display. Displays correctly adjust.
- I boot the system with the external HDMI attached, such as a reboot after updates.
- After the new boot, I close the laptop lid. Displays correctly adjust.
- I unplug the HDMI display.
- With lid closed, and no external display, the system correctly suspends.
- I attempt to open the lid. The system briefly resumes, but the display remains blank. The system suspends again shortly after.
At this point, I have to fully power the system off to render it usable again. Reattaching the HDMI display does not help.
Additional notes:
- This does not happen with displays attached over the Thunderbolt/DP connection.
- If the system is suspend with the lid closed and an HDMI display attached, and then resumed in that state by external keyboard input, there are no issues.
- If the system is suspend with the lid closed and an HDMI display, and then resumed by opening the laptop lid, there is no issue.
- If the system has an external HDMI display, and the lid closed; then the external display is removed, and the system suspends; and then the HDMI display is reattached: The system will correctly resume when woken with an external keyboard input.
- This only happens if the system is booted with the HDMI display attached. If the display is attached post-hoc, then these issues are not present.
Edit: I'm investigating this more, as it isn't perfectly consistent.
Okay: While I can get the HDMI/Suspend-Resume issue to trigger in bursts, I cannot get it to trigger reliably. I had two bouts of it today though, so it is often enough to be a concern and hopefully to be able to diagnose.
I believe it occurs when HDMI is attached at boot, but honestly, I do not remember what the condition was on the first tranche of times I saw the issue, though I know that attached-at-boot was the condition the second time.
The reason it concerns me is that users who use a Dual internal + external setup at a fixed location, like an office or home desk, are likely to close a laptop before unhooking it when transitioning to a on-the-move regime such as to give a presentation or to work from a remote location. There is a risk of hard data-loss with this problem as well, as anything not written-to-disk may be readily lost - for example, unsynced log entries that would normally be written out on resume are lost (which is annoying for diagnosing the issue).
For the moment, I'm continuing other certification test:
@crawfxrd I'm having issues with ethernet not working. lspci reports Ethernet controller: Intel Corporation Device 550a (rev 20)
However, nothing else seems to see the presence of an ethernet device. ip only returns results for the lo and wlp device types, and at best utilities like inxi are showing that there is something but suggests the correct driver is not present, which I somewhat doubt, unless the driver support post-dates our current kernel.
Any guidance on what to prod? I've harasses the sysfs entry fairly robustly, and I've confirmed that what appears to be the correct driver (the Intel e1000e driver) is present and loaded.
~The clock config in coreboot is most likely wrong.~ Fixed in latest push.
If you have ever flashed externally it won't work as the GbE firmware will be invalid and won't load.
e1000e 0000:00:1f.6: The NVM Checksum Is Not Valid