I guess you could also test the current kernel and Nvidia driver on Wayland to see if that route is workable. This seem a less janky workaround than freezing very important packages such as the kernel and your graphics driver.
The problem occurs before Wayland or anything even loads! I’ve tested other distros with the same Zen kernel and the same NVIDIA driver and it occurs there as well. I don’t think this is Garuda specific…I think it’s potentially a wider problem…I’m seeing tons of people with similar issues on a variety of other distros.
Curiously, it doesn’t happen on my sons rig with an older NVIDIA card though (GTX 970).
2 Likes
Ya, sorry about that suggestion. Obviously, this is occurring far earlier in the boot sequence. All I can suggest as an alternative to freezing the kernel is perhaps test an alternate such as linux-hardened. I have found the linux-hardened kernel sometimes worked when the LTS and current Linux kernel both have some serious issue.
1 Like
I’m poor–only have Intel–but the intel-ucode has to be loaded early in the process.
I wonder if your rig might need a similar process with amd-ucode or whatever.?
Ok, so I’ve updated to Kernel 6.6.4 as it became available. Still the same problem persists.
I don’t think this problem is specific to Garuda, I can see posts all over the place with the same issue. Problem appears to be the latest NVIDIA driver.
Switching kernels does not help.
The system does eventually boot, but it takes ages…so far this problem doesn’t appear to stop anything working, it’s just a massive pain in the arse because you can’t tell whether the system is going to boot at all. Occasionally it doesn’t.
I guess we’re waiting for a new NVIDIA driver to drop that hopefully fixes this.
Switching to the nouveau/open driver does help, but obviously that is not a long term fix.
Same thing happens with a clean install for me. So performing a clean install probably won’t help either.
For me, a clean Manjaro does not appear to have the same issue. So if you’re desperate to update and can’t live without a fix for a while, switch to Manjaro or something. I can see that people are having this same issue with Ubuntu and OpenSUSE and a couple of other distros…but Manjaro seems to have dodged this bug for now, somehow.
2 Likes
Well, annoying little bug, just glad to see it was not just my system. Ill give it a week or two and try updating. I just rolled back a snap and life is good as long as I do not update at this moment in time.
That is almost exactly what I’m seeing.
The same temporary boot hangup is happening to me that @hammerhead is describing on here. Before I offend any of the OG forum jannies (im just joking Mods I love y’all for real thank you for your assistance tons of people benefit from your work that you will never hear from, but yall are providing an excellent reference resource) is it cool to post my inxi as a reply? GTX 1080, i5 11th Gen Rocket 
lake, Zen if not
1 Like
I usually hate manually installing Nvidia drivers. It is almost always bugger. I wonder what would happen if / or if anyone has tested it with just the latest drivers from their website:
|Version:|535.129.03|
|Release Date:|2023.10.31|
They hold back their Arch updates from several weeks to a month (or even more) when major issues arise. In all likelihood the bug has simply been delayed by Manjaro, not fixed by the use of magical Manjaro green pixie dust. 
7 Likes
I’m not suggesting that Manjoro has fixed the bug…but I tested using the same Zen kernel (6.6.4) and NVIDIA driver versions and the problem isn’t there…so it’s not a case of them rolling out a fix, it’s a case of them not rolling out the bug…whatever it is.
I’d rather stick with Garuda…but I need to work, so flipping to Manjaro temporarily is the only way forward right now.
Interesting question. I can test this at some point…but manually installing NVIDIA drivers is a maintenance nightmare rolling forward…at least it used to be…I haven’t manually installed an NVIDIA driver for years…it used to be a real pain every time there was a kernel update.
I don’t mind, the more info the better!
I’ve actually just noticed that nvidia-dkms is running a couple of minor versions behind the nvidia package…could this be significant?
I am using the nvidia-dkms package not the nvidia package.
One of the updates that went in before the issue was the nvidia-settings package that puts nvidia-settings ahead of the nvidia-dkms package.
Hi there, just gonna drop some input as I had flip event errors. Things loaded, but it was really slow. And you seem to also be using multiple monitors like myself.
I’d resigned myself to living with the long boot times, but then one time when I loaded up a game, my third monitor went wonky, lost it’s identifier (went to default port name), and could only use fallback resolutions. I really didn’t want this so delved into it and long story short…
The faulty monitor had a bad HDMI port, and the EDID data was buggered. I unplugged it, and the was no longer any frame event errors, and boot times went from 90+ seconds to about ten. I ended up using an HDMI to DVI adapter so the DVI port on the monitor worked, and things have been smooth since.
I was using windows before and hadn’t had any issues, so I wasn’t really expecting to find a hardware fault in my monitors.
2 Likes
check out forum.level1techs.com and look through all the “kvm” posts on there. it is ASTONISHING how few devices(monitors, keyboards, mice, 'insert random usb device") are ACTUALLY compliant with the relevant spec. poor wendell thought he could just make a nice kvm that worked, then fell down a rabbit hole that would make alice cringe.
hell not even cables are good and safe.
2 Likes
Yes, I have a third monitor that acts a bit wonky and unplugging it seemed to “sort of” help…I’ve noticed issues with it in the past, but nothing that has ever led to a dodgy boot up.
Multiple monitors on NVIDIA cards under Linux has always been a fairly frustrating experience.
1 Like
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.