Closed R3zk0n closed 8 months ago
@R3zk0n thanks for reporting this! I’ll take a look asap. I’m in the middle of moving so I hope I get around to fixing this soon. PRs welcome :)
All good, will do my best to try to fix and then submit a PR, seems like the issue might be in the docker_runner.py
OK! I am back so i think ive worked out the issue,
it appears that the like_debugger isnt being tagged, and isnt installing gef correctly which cause it not be able to continue and use the remote debugging due gef-remote not being installed which makes sense, demostrated below
But mainly forcing the tagging of the container then installing gef, seems to make it work fine,
Not exactly where to make the edit to force the gef to be installed, since in theory it should already be done but i guess doing a execute to confirm gef is installed in the container if not bail out would be useful, trying to find where to add it but once i do ill do a commit! :)
The GEF "installation" happens in the .dockerfile_dbg
: https://github.com/0xricksanchez/like-dbg/blob/main/.dockerfile_dbg#L44. There are no sanity checks if the command did not crash 😬
Tried to use a bash script to replace the build but for some reason it runs in the wrong context, still working on it. :D
Describe the bug It appears like on a clean build of like-dbg, the like_debugger docker image isn't being tagged sometimes, manually forcing the tag seems to break the the gdb script to the vm-linux.
To Reproduce Steps to reproduce the behavior:
Expected behavior The script deploys and builds everything however often the docker image for the debugger isnt tagged, maybe a way to force the tagging to prevent this type of issues?
Screenshots
Desktop (please complete the following information): Fill me in on the specs of your environment:
Operating system (e.g.:
cat /etc/os-release
) -Python version (
python --version
) - python 3.10.4Python packages (in your virtualenv:
pip freeze
)Docker version (
docker --version
)Additional context Tried it on multiple different systems, and virtual machines and appears to always be the same issue.