James Tucker
James Tucker
> > I work on tailscale and wireguard. Ideally what I want is a way to explicitly exclude these interfaces from consideration. We can play whack-a-mole with the plethora of...
> > > > I work on tailscale and wireguard. Ideally what I want is a way to explicitly exclude these interfaces from consideration. We can play whack-a-mole with the...
Superseded by https://go-review.googlesource.com/c/sys/+/609295
We're seeing quite a bit of this again. I found a single client log with a large number of different code paths for this, on 1.70. It feels like enough...
for reference, internal link to the set of stacks i studied most closely this evening: https://paste.corp.ts.net/paste/54627a0d-93f6-42b4-a8c6-7c83d9e610b5. I also had a scan over the arm64 versions of poll write just in...
Hi @iGerman00 thank you for the report. It looks like we need to forcefully close DNS sockets when a user changes exit node to ensure that this won't happen. The...
Yup, I think this strategy will be fairly quick and easy. It certainly would work with gitosis as I just tested it. I haven't tested gitolite yet.
i did notice ``` 2024-08-27T15:03:19Z | starting forever-blocked frontend Peek for shared fate shutdown 2024-08-27T15:03:19Z | [unexpected] frontend Peek returned (1, ); continuing ``` in the original report, I'm not...
Tailscale is slow: `wg: Failed to write packets to TUN device: write /dev/net/tun: invalid argument`
Does this reproduce for you on `linux-mainline`? cc @jwhited
Tailscale is slow: `wg: Failed to write packets to TUN device: write /dev/net/tun: invalid argument`
I was able to partially reproduce this, specifically: - Using arch LTS `Linux archlinux 6.6.44-2-lts #1 SMP PREEMPT_DYNAMIC Sat, 03 Aug 2024 18:10:33 +0000 x86_64 GNU/Linux` - Observed iperf3 resulting...