[libc++] Optimize map::insert_or_assign (#155816)

`__emplace_unique` uses `__find_equal`, which can be significantly
faster than `lower_bound`. As a nice side-effect, this also changes the
implementation to the "naive" implementation of trying `insert` first,
and if that fails assign instead. This also matches the
`insert_or_assign` overloads with a hint.

```
Zen 2:
--------------------------------------------------------------------------------------------------------
Benchmark                                                                            old             new
--------------------------------------------------------------------------------------------------------
std::map<int, int>::insert_or_assign(key, value) (already present)/0             1.62 ns         1.53 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/32            5.78 ns         5.99 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/1024          21.5 ns         15.4 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/8192          26.2 ns         20.5 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/0                   22.5 ns         21.1 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/32                  42.9 ns         28.4 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/1024                 118 ns         92.0 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/8192                 227 ns          173 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/0     13.2 ns         18.9 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/32    65.6 ns         39.0 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/1024   127 ns         64.4 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/8192   134 ns         71.4 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/0           45.6 ns         37.3 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/32           142 ns         93.3 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/1024         288 ns          147 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/8192         368 ns          182 ns

Apple M4:
--------------------------------------------------------------------------------------------------------
Benchmark                                                                              old           new
--------------------------------------------------------------------------------------------------------
std::map<int, int>::insert_or_assign(key, value) (already present)/0              0.784 ns      0.740 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/32              2.52 ns       1.77 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/1024            8.72 ns       4.06 ns
std::map<int, int>::insert_or_assign(key, value) (already present)/8192            10.6 ns       3.98 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/0                     17.3 ns       17.2 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/32                    22.5 ns       19.3 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/1024                  56.8 ns       33.5 ns
std::map<int, int>::insert_or_assign(key, value) (new value)/8192                  88.2 ns       41.0 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/0       16.6 ns       11.8 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/32      13.7 ns       30.7 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/1024    46.7 ns       49.1 ns
std::map<std::string, int>::insert_or_assign(key, value) (already present)/8192    41.9 ns       76.9 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/0             40.0 ns       40.5 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/32            38.9 ns       40.0 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/1024          84.9 ns       96.9 ns
std::map<std::string, int>::insert_or_assign(key, value) (new value)/8192           166 ns        149 ns
```
3 files changed
tree: f1b2c0038b734a5eac0829ec1f1d3e29a2f8bfc8
  1. .ci/
  2. .github/
  3. bolt/
  4. clang/
  5. clang-tools-extra/
  6. cmake/
  7. compiler-rt/
  8. cross-project-tests/
  9. flang/
  10. flang-rt/
  11. libc/
  12. libclc/
  13. libcxx/
  14. libcxxabi/
  15. libsycl/
  16. libunwind/
  17. lld/
  18. lldb/
  19. llvm/
  20. llvm-libgcc/
  21. mlir/
  22. offload/
  23. openmp/
  24. orc-rt/
  25. polly/
  26. runtimes/
  27. third-party/
  28. utils/
  29. .clang-format
  30. .clang-format-ignore
  31. .clang-tidy
  32. .git-blame-ignore-revs
  33. .gitattributes
  34. .gitignore
  35. .mailmap
  36. CODE_OF_CONDUCT.md
  37. CONTRIBUTING.md
  38. LICENSE.TXT
  39. pyproject.toml
  40. README.md
  41. SECURITY.md
README.md

The LLVM Compiler Infrastructure

OpenSSF Scorecard OpenSSF Best Practices libc++

Welcome to the LLVM project!

This repository contains the source code for LLVM, a toolkit for the construction of highly optimized compilers, optimizers, and run-time environments.

The LLVM project has multiple components. The core of the project is itself called “LLVM”. This contains all of the tools, libraries, and header files needed to process intermediate representations and convert them into object files. Tools include an assembler, disassembler, bitcode analyzer, and bitcode optimizer.

C-like languages use the Clang frontend. This component compiles C, C++, Objective-C, and Objective-C++ code into LLVM bitcode -- and from there into object files, using LLVM.

Other components include: the libc++ C++ standard library, the LLD linker, and more.

Getting the Source Code and Building LLVM

Consult the Getting Started with LLVM page for information on building and running LLVM.

For information on how to contribute to the LLVM project, please take a look at the Contributing to LLVM guide.

Getting in touch

Join the LLVM Discourse forums, Discord chat, LLVM Office Hours or Regular sync-ups.

The LLVM project has adopted a code of conduct for participants to all modes of communication within the project.