https://sourceware.org/bugzilla/show_bug.cgi?id=34758
--- Comment #4 from Martin Mirchev <martin.mirchev at u dot nus.edu> --- Hi Jens, Thanks for looking at it. To answer directly: the patch was drafted with an LLM-based coding assistant, not written by me from scratch. I reviewed it, reproduced the internal error, and tested the change (the reproducer then gives the CFI error plus a warning, and a matched remember/restore pair assembles to an identical object). But under the binutils policy, I don't think it can be treated as my own work. So please treat the attachment as an illustration of the problem rather than a patch submission. If it helps, I'm happy to write a fix myself, manually and without LLM-generated code, and send it to the list under the usual DCO, if you and the maintainers agree with that approach. Otherwise, you're of course welcome to write your own fix with "Reported-by: Martin Mirchev". Either way, I can test patches against the reproducer. The same applies to the proposed fixes I attached to bugs 34755-34760, 34770, 34872 and 34873. Please treat those as illustrations too. The bug reports and reproducers themselves stand, and the same offer to write a manual fix holds for them. Thanks, Martin -- You are receiving this mail because: You are on the CC list for the bug.
