Add combined process tree+network event source #19
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Add a generic "net" source and event type. Attempting to use the source on any platform other than Linux results in a runtime error for now. The idea would be for other OS implementations to exist over time. - Provide proof-of-concept Linux eBPF program using CO-RE safe APIs. This only handles outbound IPv4 connections for now, but can be extended to collect more information. The program itself should be usable back to Linux 4.17 kernels, but I'm using the new-style BPF link syscalls, which came along in Linux 5.7 (from 2020). - Introduce Nix flake with necessary C/LLVM tooling. The LLVM/C toolchain is only needed when modifying the C program: .o files are automatically generated for both big-endian and little-endian architectures and are checked into version control, so casual contributors will not need this for making changes to reveald as a whole. For folks working on the C program, the Nix flake provides a consistent, reproducible environment for generating the .o files regardless of OS being used for development.
Due to eBPF restrictions, the size of arguments collected is relatively small. Further iterations of this code should hopefully provide better fidelity, but I wanted to demonstrate the approach first.
This allows for the source to combine multiple signals together so more logical signals can be sent to other processes. I also came up with a better way of passing the arguments back to userspace so the limits are larger while giving priority to the event stream.
This was referenced May 21, 2025
|
Hey @zombiezen ! Excited to try this out today. Looks like an internal package still needs to be pushed though, is that the case? I see it in the other branch, maybe I can pull it into this one. |
|
Ah whoops, I forgot to remove the reference to the network source that no longer exists. Try again now. |
Allows us to tie the process arguments to PIDs in a bounded manner. Also picked out a more reliable attach point for detecting forks.
|
Now even more exciting: network connect events are correlated with their process information. |
|
Test config file: {
"sources": {
"processes": {
"type": "processes",
},
},
"destinations": {
"printer": {
"type": "printer",
},
}
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This obsoletes #17 and #18. This rolls the network observation into the same source as the process tree observation and increases the argument buffer sizes. I'm going to do a fast-follow to correlate the exec arguments with the network calls once I figure out how to reliably detect process exits.