Matthieu Baerts

@pabeni: thank you for having checked that! I don't see similar reports on lore after a quick search. I guess it would be better to forward this to netfilter ML...

@pabeni : Thank you for the link! @fw-strlen : Thank you for the fix! I'm closing this ticket, because it will likely not be automatically closed as this patch will...

@cpaasch : Just to be sure, that's with @fw-strlen's patch from your [syzkaller_export](https://github.com/multipath-tcp/mptcp_net-next/commits/syzkaller_export) branch, right?

@cpaasch : do you still have this issues? It looks like Florian's patch has been applied: https://lore.kernel.org/netdev/[email protected]/T/ Note that there might be a follow-up patch silencing a `WARN()`: https://lore.kernel.org/netdev/ZnypC7iOx_qtUH37@calendula/

I never saw this old issue in my syzkaller instances (started in April 2025). I suggest closing this issue.

> So our expectation is that an existing file is getting (silently) overwritten? Personally, I think it makes sense to fail in this case, but then with a better error...

> > > So our expectation is that an existing file is getting (silently) overwritten? > > > > > > Personally, I think it makes sense to fail in...

> > > > How should we handle the case that an dump is requested, but no output file is selected? I guess that's not even possible with our current...

From the last weekly meeting: - it looks like we can get that with a blocking IO socket → not normal for an userspace program, to be investigated - when...

The CI just had the issue with the **default** packet scheduler (**not** with a BPF packet scheduler): https://github.com/multipath-tcp/mptcp_net-next/actions/runs/10108350314 ``` # #169/1 mptcp/base:OK # #169/2 mptcp/mptcpify:OK # #169/3 mptcp/subflow:OK # create_netns:PASS:ip...