[libcxx-commits] [libcxx] [libc++] Add some _LIBCPP_ASSUMEs for bounded iterators (PR #109033)

[Louis Dionne via libcxx-commits]

https://github.com/ldionne commented:

When I configure my local tree with

```
-DLIBCXX_HARDENING_MODE=fast
-DLIBCXX_ABI_DEFINES="_LIBCPP_ABI_BOUNDED_ITERATORS;_LIBCPP_ABI_BOUNDED_ITERATORS_IN_STRING;_LIBCPP_ABI_BOUNDED_ITERATORS_IN_VECTOR"
```

and then run this benchmark with and without the patch:

```
$ libcxx-lit <build> -sv --param optimization=speed --time-tests --show-all --param enable_benchmarks=run -j1 libcxx/test/benchmarks/containers/sequence/vector.bench.cpp
```

I get those results:

```
$ libcxx/utils/libcxx-compare-benchmarks build/{default,candidate} libcxx/test/benchmarks/containers/sequence/vector.bench.cpp
Comparing build/default/libcxx/test/benchmarks/containers/sequence/Output/vector.bench.cpp.dir/benchmark-result.json to build/candidate/libcxx/test/benchmarks/containers/sequence/Output/vector.bench.cpp.dir/benchmark-result.json
Benchmark                                                                Time             CPU      Time Old      Time New       CPU Old       CPU New
-----------------------------------------------------------------------------------------------------------------------------------------------------
std::vector<int>::iterate-whole-container/32                          -0.0273         -0.0271            11            11            11            11
std::vector<int>::iterate-whole-container/1024                        -0.0257         -0.0256           336           328           336           328
std::vector<int>::iterate-whole-container/8192                        -0.0375         -0.0374          2657          2558          2657          2558
std::vector<std::string>::iterate-whole-container/32                  +0.3467         +0.3473            11            15            11            15
std::vector<std::string>::iterate-whole-container/1024                -0.0394         -0.0387           341           328           341           328
std::vector<std::string>::iterate-whole-container/8192                -0.0041         -0.0040          2571          2560          2570          2560
OVERALL_GEOMEAN                                                       +0.0272         +0.0276             0             0             0             0
```

I think the result for `std::vector<std::string>::iterate-whole-container/32` is a fluke due to the small size of the container. But I also don't see any improvement, which is a bit surprising to me. Any ideas?

https://github.com/llvm/llvm-project/pull/109033