Using Fedora 39 Workstation Beta with Gnome 45, latest builds as of 10/14/2023 When you first boot the machine, it defaults to X11 with no option to select Wayland session. However, if you log out an log back in, it will correctly use Wayland. The issue persists every time you reboot the machine. Reproducible: Always Steps to Reproduce: 1. Power on Framework 13 AMD 2. X11 will be the default session 3. Log out 4. Log back into Wayland session Expected Results: Should default to Wayland
Hello people on Fedora project, Note that this issue was reported on the following thread on Framework forum. https://community.frame.work/t/fedora-39-official-fedora-39-thread/37465/15
Hello Adam, Could you help to fix this issue? I don't have Framework Laptop 13 AMD where the reported issue happens in Fedora Linux 39. I am just here to help. The component: gnome-desktop may be set by an assumption. You may assign this issue to a right person, and request a necessary info to the reporter, and change the current component to a proper one.
We would need logs to have any hope of figuring this out. Please attach the output of `journalctl -b` after a boot where you reproduce the whole issue (initial login which falls back to X, then a subsequent logout/login which gives you Wayland). It'd be great if you could note important times too (when the first login happens, when the logout happens, when the second login happens).
Created attachment 1993971 [details] journaltcl -b logs Journalctl -b logs as requested Boot at 13:01, Log into X11 around 13:02, Log out / Log into Wayland at 13:03
Can this be reproduced using BIOS 3.03? I've got a suspicion that the issue in BIOS 3.02 that causes MES timeouts (seen in your log) is the reason that Wayland crashes the first time.
@mario.limonciello I don't have access to the 3.03 BIOS yet, but I can test again when I do. I have also been running into intermittent amdgpu driver crashes (screen goes black and comes back) when launching applications, which may or may not freeze the entire system. Not sure if it's also related to the new BIOS, but happy to post logs for that as well if you think it's helpful.
If it happens in the new BIOS open another issue for it.
Late to the party, I've been heads down on a multitude of other testing tasks. 3.02, fresh install before updates, boots us to X. Confirmed. 3.02, ran dnf updates, still boots to X. Confirmed. 3.03 (internal testing branch of 3.03) fresh install before updates, boots to Wayland. 3.03 (internal testing branch of 3.03), boots to Wayland. Based on what Mario indicated was happening, looks to be on point.
So, is there anything for Fedora to do here? Is the 3.03 firmware update available to the public yet?
Unfortunately, BIOS 3.03 firmware is not downloadable to the public yet for Framework Laptop 13 AMD Batch 1 users who got the BIOS 3.02 firmware. If there will be update about it, we see it on the following page. https://community.frame.work/t/amd-ryzen-7040-series-bios-3-03-beta-release/37479 It seems that Framework Laptop 13 AMD Batch2 with BIOS 3.03 firmware is starting to ship to the users. So far I observed one person received the laptop. https://www.reddit.com/r/framework/comments/17cctvg/batch_2_shipped_yesterday_received_today/
> It seems that Framework Laptop 13 AMD Batch2 with BIOS 3.03 firmware is starting to ship to the users. I am observing Framework Laptop 13 AMD Batch 2 was shipped with BIOS 3.02 from some users' reports. https://community.frame.work/t/amd-batch-2-guild/29605/144
> So, is there anything for Fedora to do here? Is the 3.03 firmware update available to the public yet? It seems that the BIOS 3.03 firmware update is available to the public now. https://community.frame.work/t/amd-ryzen-7040-series-bios-3-03-and-driver-bundle-beta/38655
Jun Aruga, thanks for catching this up, it's appreciated. (In reply to Adam Williamson from comment #9) > So, is there anything for Fedora to do here? Is the 3.03 firmware update > available to the public yet? Nothing for Fedora to do here, this was a BIOS specific issue that was resolved with the latest BIOS release. Thanks Adam. This should be considered resolved.
3.03 here seems to be working fine. ;)