david942j

Sorry for late response, I haven't checked here for days. I would need some time to recall what I was thinking about the `new_load_method`, so let me answer other easier...

> 2. Related to previous question, can i have an example on how this may go bad( sorry about being lazy and not figuring out myself) I think it just...

> 1. I see that mgap moves the offsets by extension size of all headers beyond the LOAD segment being patched. Will it be correct to assume, we can create...

Not sure if I understand your question.. > i can guess why this has to be done,we may have other content following it that may prevent us from inline patching...

> How about appending "libcrypto.so.1\x00" to the original "libc.so.6\x00xyz\x00" and moving all data / headers after strtab_string? It's still possible we don't have enough space even adding this. The `new_load_method`...

A note to myself: patchelf.rb fails with NotImplementedError when patching the system glibc on Ubuntu 20.04 https://gitlab.com/david942j/libcdb/-/blob/master/libc/libc6_2.31-0ubuntu9_amd64/lib/x86_64-linux-gnu/libc-2.31.so ``` ➔ bundle exec patchelf.rb --so libc.so.217 /lib/x86_64-linux-gnu/libc.so.6 libc.patch Traceback (most recent call...

Any update here :) ?

yes?

Hi @lem0nify , since I don't have Windows environment for testing, could you try using a debugger with the stack backtrace to see where the crash point is?

@furunkel just let you know I've pushed crabstone to rubygems.org