https://github.com/boomanaiden154 created https://github.com/llvm/llvm-project/pull/244469
This patch adds a reasonably optimized clang_vector based memchr implementation. It is comparable in .text size to glibc and about 30% faster for inputs <64 bytes (90+% of inputs for server applications). We lose a little bit for larger inputs due to extra instructions in the main loop that LLVM isn't able to optimize away, but not by too much. The original implementatin was written by an agent as I prodded it to keep making the implementation faster. Then I heavily hand-edited mostly for style. Assisted-By: LLM _______________________________________________ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits
