Posts by escalade
-
-
-
-
-
-
-
Ah, yeah it could be the switch to iris. I was under the impression that the iris driver was considered in good shape, but I suppose this could be a corner case. Put the following in /storage/.config/profile.d/mesa.conf:
That will set the variable for the sway process and tell Mesa to use the old driver.
-
Code
2020-02-26 20:00:38 - [backend/drm/drm.c:1305] Found connector 'DP-1' 2020-02-26 20:00:38 - [backend/drm/drm.c:1453] Requesting modeset for 'HDMI-A-1' 2020-02-26 20:00:38 - [backend/drm/drm.c:697] Modesetting 'HDMI-A-1' with '1366x768@59790 mHz' 2020-02-26 20:00:38 - [backend/drm/util.c:210] Unable to add DRM framebuffer: Invalid argument 2020-02-26 20:00:38 - [backend/drm/drm.c:611] Page-flip failed with primary FB modifiers enabled, retrying without modifiers 2020-02-26 20:00:38 - [backend/drm/util.c:210] Unable to add DRM framebuffer: Invalid argument 2020-02-26 20:00:38 - [backend/drm/drm.c:620] Failed to initialize renderer on connector 'HDMI-A-1': initial page-flip failed 2020-02-26 20:00:38 - [backend/drm/drm.c:703] Failed to initialize renderer for plane 2020-02-26 20:00:38 - [sway/config/output.c:398] Failed to modeset output HDMI-A-1
Something weird going on, it seems to find your connector at DP-1 but tries to set up HDMI-A-1 instead. What do you mean bootloop, does it reboot? Does the network interface go down?
Try replacing /storage/.config/sway/config with a copy of /etc/sway/config, then add this to the end:
output DP-1 resolution 1920x1080
BTW, is your main GPU the integrated Intel HD Graphics 2500? I suppose maybe it's too old and lacking something in way of GLES compatibility, but then again I see no sign of that. Perhaps the connector issue can be fixed by config.
-
-
You can see the version tag when you log in through SSH, perhaps there is an issue writing to your filesystem for some reason? Is the upgrade succcessful at all? You can try hitting "escape" when the upgrade boot animation is running to see what's going on in the background. I'm interested to see the log files from the actual issue as well, you can upload them to pastebin with the command pastebinit < logfile.
-
Great to hear! In other news, the XU4 images have a functional desktop as well now. RPi is next
Oh, and worth to mention the generic builds has the 330% vulkan mesa improvement for gen7/haswell. Happy gaming. -
-
-
-
Seems like you are on an old release as well Generic.x86_64-devel-20200220125702-28eddc2. Latest is
RetroELEC-Generic.x86_64-9.2-devel-20200225010434-28eddc2.img.gz md5sum
66b9f295f70544861b7f75c859753621.
Check that your /etc/sway/config contains this in the end:
Code
Display More# Start LavaLauncher exec lavalauncher \ -b /usr/share/icons/lava/firefox.png "firefox.sh --new-window" \ -b /usr/share/icons/lava/emulationstation.png emulationstation.sh \ -b /usr/share/icons/lava/retroarch.png retroarch.sh \ -b /usr/share/icons/lava/kodi.png es-kodi.sh \ -b /usr/share/icons/lava/wifi.png cmst \ -b /usr/share/icons/lava/rpcs3.png rpcs3.sh \ -b /usr/share/icons/lava/ps2.png pcsx2.sh \ -b /usr/share/icons/lava/plex.png "firefox.sh --new-window http://localhost:32400/web" \ -b /usr/share/icons/lava/gamepad.png antimicro \ -b /usr/share/icons/lava/transmission.png "firefox.sh --new-window http://localhost:9091" \ -b /usr/share/icons/lava/youtube.png "firefox.sh --new-window http://www.youtube.com" \ -b /usr/share/icons/lava/netflix.png "firefox.sh --new-window http://www.netflix.com" \ -b /usr/share/icons/lava/terminal.png terminal \ -b /usr/share/icons/lava/power.png "swaynag -t warning -m 'Do you really want to shutdown?' -B Reboot 'systemctl reboot' -B Shutdown 'systemctl poweroff' -B Suspend 'systemctl suspend'" \ -s 70
-
-
Please try from a clean install (use a USB key or something), because I can't reproduce. I can't see any trace of what's going in your logs either, everything logs to it's own file in /var/log which should be in the zip file. I see them when I test here but not in your logs. Be sure not to reboot when creating the log files as they are deleted on reboot.
What's stopping you from dropping the core into /tmp/cores manually?
-