As I said in my message: same computer, same version of EndeavourOS up-to-date (3 installations where only the Window Manager differs), same DNS (resolv.conf), same connection, different behavior… I cannot explain that
I’m also experiencing this problem. New pages/tabs will not load, images won’t load but page will, page loads halfway. Sometimes a refresh will fix it, other times it will not. This may not do any good since the source of the problem was already found, but I wanted to add my experience with what I’m using and what I’ve tried.
Browser: Firefox 142.0.1 (64-bit)
EndeavourOS: 6.16.4 (rolling)
DNS: I have tried 9.9.9.9, 1.1.1.1, and 8.8.8.8 and it happens on all of them
Like @Moonbase59, I also have my DHCP pointed to my router for DNS queries, and the router has the DNS on WAN that I choose to use (tried all 3 and no resolve, clearing DNS cache each time).
Hope they fix that kernel issue real soon now. I found that on all machines here that prefer IPv6, I have no problems whatsoever.
For a real quick (and possibly not that definitive) check you could visit https://ipleak.net/ and see what your browser prefers:
1 Like
I seem to have the problem with LTS kernel as well.
While it should be fixed there, shouldn’t it?
Here’s what I’m currently running (LTS):
Linux lg-eos-kde 6.12.44-1-lts #1 SMP PREEMPT_DYNAMIC Thu, 28 Aug 2025 15:07:21 +0000 x86_64 GNU/Linux
And here are the same problems:
2 Likes
I had the same issue on the latest LTS kernel, I decided to rollback 6.12.41-1-lts and no longer have the issue.
1 Like
I do wonder if the EOS team could push a fix somehow… I mean, anyone can compile the kernel with the fix applied… (Provided they’re comfortable with the process and have time to do it.)
1 Like
Also downgrading to
Linux lg-eos-kde 6.16.1-arch1-1 #1 SMP PREEMPT_DYNAMIC Fri, 15 Aug 2025 16:04:43 +0000 x86_64 GNU/Linux
as suggested before also fixed the problems I mentioned.
3 Likes
The bug was introduced in 6.16.2, and at that time it also entered the LTS branch, which was at 6.12.43. Since then, it has not been officially fixed in 6.16.3, 6.16.4, or 6.12.44. The funny thing is that it’s very easy to fix. The few distros that have fixed it are CachyOS, Nobara, PikaOS…, which use their own custom kernels.
For my part, I compiled a corrected 6.16.3 and also a 6.16.4 on the same day it came out, with Oscar’s correction and working like a charm.
Browser default: IPv4 (69 ms) IPv6 test not reachable. (error)
Fallback: Fail (timeout)
That’s why, in my case, the IPv4 bug was so devastating, and I noticed it from the very first minute.
1 Like
I am on Linux LTS kernel 6.12.42-1-lts and with IPV4 I do not have any issue.
1 Like
So does anyone know when this will be fixed? I think it’s kind of ridiculous for the internet to not work for more than a week.
I guess maintainers are using IPv6 or recompiled the kernel for themselves or downgraded. But that shouldn’t be necessary. An OS shouldn’t be left majorly broken for this long. Especially when the fix was identified so soon.
1 Like
13 days and counting. And yes, I also find it quite unbelievable, as it’s a bug that literally breaks the internet for a very high percentage of Linux users.
1 Like
I’m sure the Arch devs are very burdened by the current attacks on their infrastructure – but it’s not like this is an arcane, eccentric, deep problem requiring hours of study. Anyone can compile a fixed kernel if they have the know-how, the confidence, and the time. It’s one patch. So why aren’t they pushing a quick hotfix?
Tomorrow it will be two weeks of this problem, during which there was a kernel update, which somehow missed this… that’s pretty insane.
Anyway, as I said, if the EOS devs could somehow push a fix just for us, I’d be very grateful! Won’t fix the wider issue, but at least we on this forum could happily engage in smug pollution for our choice of distro. 
3 Likes
I guess they wait for it to be fixed upstream in 6.16.5.
I had to attempt to update five times before all the repos managed to get a connection through and update.
Feels petty to keep posting in this thread to complain, but if we don’t make noise, our frustration isn’t known, so – apologies for being rude, but I am quite annoyed.
Two weeks of this – EOS could probably ship a hotfixed kernel… There’s probably a reason they don’t, I offer no judgement on that either way, I don’t know enough to say anything about it – but I mean, there’s not even a software news entry about this, so anyone who isn’t actively reading the forums etc. is naturally going to assume faulty hardware, bad system config, etc.
Sorry to keep ragging on the same points, but I don’t have the time or knowledge to apply a fix myself, and this is a deeply frustrating issue that has affected me every day it’s been going on. I am quite annoyed, so a certain amount of unreasonable pushiness is inevitable.
1 Like
Ever thought of downgrading the kernel ?
sudo -i
sudo downgrade linux
Select the latest 6.15 kernel and your done.
2 Likes
6.16.1 is OK too.
Think this has been said before, but using a rolling distro, where updates are upstream all the time, there is always the risk a package (or in this case a kernel) can contain bugs, or will not work the way that is expected on certain computers. If you are not willing to take the risk go with another (non rolling) distro. Or spend your energy on using some of the work arounds offered in this topic.
1 Like
I would guess it’s because Arch is kept as vanilla as possible. Has a patched kernel been released yet?