Nick Bulischeck

Thanks for your issue and reference to the mailing list. I'll have to move the module iteration code from https://github.com/nbulischeck/tyton/blob/master/src/module_list.c#L21 to grab module addresses manually now that `__module_address` is unexported.

For my future reference: 1. [__module_address calls mod_find](https://elixir.bootlin.com/linux/v5.10.14/source/kernel/module.c#L4528) 2. [mod_find does a list_for_each_entry_rcu on the module list calling within_module on each one](https://elixir.bootlin.com/linux/v5.10.14/source/kernel/module.c#L222) 3. [within_module checks the init and core sections...

@blshkv Thanks for the suggestion! I wholeheartedly agree. I actually saw this done recently in another kernel-related project and thought it was an interesting idea. I should have time to...

@blshkv If you want, you can try out the new branch here: https://github.com/nbulischeck/tyton/tree/procfs-status This creates a file at /proc/tyton and will report back the current analysis when you read it....

Based on the trace, I'm guessing that the strcmp here https://github.com/nbulischeck/tyton/blob/procfs-status/src/network_hooks.c#L29 somehow ends up with an invalid rbnode (which seems unlikely) or the act of you catting the file while...

Can you check that this command succeeds completely? `ls -la /proc/self/net/{tcp,tcp6,udp,udp6,udplite,udplite6}`

Alright then I'm assuming it's the rb_entry nodes. This really shouldn't be a problem, but I hope that my recent push has a solution. Can you git pull the latest...

Pushed a new commit. I just took a dive through 4.20.6 and nothing indicates this should fail. I'm on 4.20.7 on Arch and it works perfectly fine. I pushed a...

Perfect! Thank you for the steps to reproduce. My reason for saying I am going to spin up a Pentoo VM was so that I could stop pushing half-baked _potential_...

@blshkv If you have free time, can you confirm if the recently merged patch fixes your issue?