云微

The key problem is that `bpf_prog_attach_uprobe_with_override` is not supported by kernel. It's a self-define type only in bpftime, so we choose a type for that. Maybe we should choose a...

We have examples using dpdk or possible using afxdp to run the network functions in xdp. So part of the answer is yes, bpftime can help deploy these NAT to...

You can find some code here and we will release more later: https://github.com/eunomia-bpf/XDP-eBPF-in-DPDK

Seems like this is fixed?

There are several configs to improve the performance of bpftime: - Use JIT when running the eBPF program. The JIT will be enabled by default in the future after more...

https://github.com/eunomia-bpf/bpftime/pull/438 DEFAULT_MAX_FD is configuable after this PR.

Hi @Yifei-Zhang-Deeproute, thanks for the detailed report and for trying both the tutorial memleak and the `example/malloc` pair. **Short answer:** bpftime reduces *per-probe* overhead (you already saw 9–16× on the...

Ok! We will add it.

by the way, maybe you can also look at https://github.com/eunomia-bpf/bpftime/tree/master/tools%2Fbpftimetool