xroche wrote: @RKSimon -- thank you for taking the time to look at this. To set expectations: **this PR is not intended for code review.** It's a proof-of-concept demonstrating that `__int256` can be implemented as a builtin type following the `__int128` pattern, end-to-end across six LLVM subsystems.
The next step is to decide whether this is worth pursuing at all. If the approach has merit, a formal RFC goes to Discourse -- where I fully expect significant discussion, especially around AI assistance. Only if the RFC reaches consensus would we move to actual code review, with 15 incremental PRs (not this monolithic diff) as the review units. We're open to follow whatever direction the community decides. **Baseline test pre-patches**: regardless of the RFC outcome, these are useful standalone. I've extracted the 9 new `.ll` codegen test files into 3 draft PRs with baseline CHECK lines from upstream `llc`: - #183586 -- X86 (3 files) - #183587 -- AArch64 (5 files) - #183588 -- RISCV (1 file) **On AI assistance**: the trailers are required by policy, but concrete numbers are more informative. This project took ~43 hours of active pairing across 24 sessions over 14 days (~360 human prompts, ~15 per session on average). Only 30% of effort went to writing code -- the rest was testing, review, git hygiene, and communication. Every line has been reviewed and validated by me; the AI accelerates the mechanical parts (boilerplate, 291-file cross-referencing), but design decisions and bug catches are mine. Thank you again for your time and for the CTPOP work on #182830 -- happy to adjust the approach based on your feedback. https://github.com/llvm/llvm-project/pull/182733 _______________________________________________ lldb-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
