Sean T Allen

I've realized that this is a subset of #4174. The issue isn't really that we are working with an initialized array, it's that "we are reading past the end". If...

@stefandd there's no passing tests that address everything noted that should be tested per #4173 to verify that such a fix is complete. Full tests of known problems are required...

@stefandd if we moved forward with this PR for address #4174 then release notes etc would need to be updated as well. But again, I think it is reasonable to...

This is lacking test coverage to be accepted for 4174. The release notes should be updated, and it still has an extra release notes file that needs to be removed....

Some happy path tests I would want to see: - [ ] copy into destination so that it doesn't all fit within the destination where you want to copy something...

@ergl i'm intrigued by how this happens. do you know what pass that error is coming from? I assume it is coming from a pass after the pass where we...

I don't think silently not doing this is a good API here. Most of Array if you overshoot an index are partial... see - read_U8 - read_U16 - etc for...

Note, it appears that `copy_from` has similar issues.

@stefandd would you be interested in implementing the above? You already did a decent amount of work towards the tests and implementation.

I think this is problematic as I could disallow FFI to a 3rd party module and it could still use cpointer with an offset to get access to arbitrary memory...