Ryan Torok

Thank you to everyone for the quick responses and insightful comments! TL;DR: Issues with ASLR and the level of introspection into the runtime that would be required make stack snapshots...

As an aside, I think this question from my original comment: > is it possible that when we restore that stack in a new instantiation of the module that skips...

ECDSA test failures resolved by #22393

As you mostly pointed out already, using `/usr/bin/env` is a double-edged sword. On one hand, it's guaranteed to exist on all [POSIX-compliant ](https://discourse.nixos.org/t/add-bin-bash-to-avoid-unnecessary-pain/5673/2) systems, so it will be present on,...

It's really only the two `cc_wrapper` scripts that matter. Keeping the test scripts as `#!/bin/bash` is probably fine, unless you're planning on running CI on a non-FHS platform. The only...

> This is a case in which Copilot comments are highly relevant and much more thorough than what I could provide. Please take a look. Copilot was quick to remind...

Trying to rewrite `cc_wrapper.sh.tpl` to be POSIX-compliant caused too many bugs, and more importantly, required eliding a bunch of safety checks. The original script uses arrays and regular expressions everywhere,...

Can a maintainer please re-run CI on this PR?

I was able to replicate this failure locally in an Ubuntu container. I think the issue might be that the path of the `cc_wrapper_inner.sh` script is hard-coded relative to `cc_wrapper.sh`....

I believe I've fixed the issue. The `cc_wrapper_inner.sh` script was being found before, but the version of it I had before was inadvertently a combination of the original `cc_wrapper.sh` and...