[libc++] Refactor how basic_string and vector hoist exception-throwing functions

[ldionne]

Summary

In basic_string and vector, we've been encapsulating all exception
throwing code paths in helper functions of a base class, which are defined
in the compiled library. For example, vector_base_common defines two
methods, throw_length_error() and __throw_out_of_range(), and the class
is externally instantiated in the library. This was done a long time ago,
but after investigating, I believe the goal of the current design was to:

1. Encapsulate the code to throw an exception (which is non-trivial) in an externally-defined function so that the important code paths that call it (e.g. vector::at) are free from that code. Basically, the intent is for the "hot" code path to contain a single conditional jump (based on checking the error condition) to an externally-defined function, which handles all the exception-throwing business.

2. Avoid defining this exception-throwing function once per instantiation of the class template. In other words, we want a single copy of __throw_length_error even if we have vector<int>, vector<char>, etc.

3. Encapsulate the passing of the container-specific string (i.e. "vector" and "basic_string") to the underlying exception-throwing function so that object files don't contain those duplicated string literals. For example, we'd like to have a single "vector" string literal for passing to std::__throw_length_error in the library, instead of having one per translation unit.

However, the way this is achieved right now has two problems:

• Using a base class and exporting it is really weird - I've been confused about this ever since I first saw it. It's just a really unusual way of achieving the above goals. Also, it's made even worse by the fact that the definitions of throw_length_error and throw_out_of_range appear in the headers despite always being intended to be defined in the compiled library (via the extern template instantiation).

• We end up exporting those functions as weak symbols, which isn't great for load times. Instead, it would be better to export those as strong symbols from the library.

This patch fixes those issues while retaining ABI compatibility (e.g. we
still export the exact same symbols as before). Note that we need to
keep the base classes as-is to avoid breaking the ABI of someone who
might inherit from std::basic_string or std::vector.

[ldionne]

Don't touch the base classes since that's an ABI break.

[ldionne]

Add comments about ABI compat of base classes

[Mordante]

libcxx/include/string

677

Do we want to remove this is ABI v2?

libcxx/include/string

 
 template<class _CharT, class _Traits, class _Allocator>
 basic_string<_CharT, _Traits, _Allocator>
 operator+(const basic_string<_CharT, _Traits, _Allocator>& __x, _CharT __y);
 
 _LIBCPP_EXTERN_TEMPLATE(_LIBCPP_FUNC_VIS string operator+<char, char_traits<char>, allocator<char> >(char const*, string const&))
 
 template <bool>
-class _LIBCPP_TEMPLATE_VIS __basic_string_common
-{
-protected:
-    _LIBCPP_NORETURN void __throw_length_error() const;
-    _LIBCPP_NORETURN void __throw_out_of_range() const;
-};
-
-template <bool __b>
-void
-__basic_string_common<__b>::__throw_length_error() const
-{
-    _VSTD::__throw_length_error("basic_string");
-}
+struct __basic_string_common;
 
-template <bool __b>
-void
-__basic_string_common<__b>::__throw_out_of_range() const
-{
-    _VSTD::__throw_out_of_range("basic_string");
-}
-
-_LIBCPP_EXTERN_TEMPLATE(class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS __basic_string_common<true>)
+template <>
+struct __basic_string_common<true> {
+    // Both are defined in string.cpp
+    _LIBCPP_NORETURN _LIBCPP_EXPORTED_FROM_ABI void __throw_length_error() const;
+    _LIBCPP_NORETURN _LIBCPP_EXPORTED_FROM_ABI void __throw_out_of_range() const;
+};
 
 template <class _Iter>
 struct __string_is_trivial_iterator : public false_type {};
 
 template <class _Tp>
 struct __string_is_trivial_iterator<_Tp*>
     : public is_arithmetic<_Tp> {};
 

 #ifndef _LIBCPP_HAS_NO_CHAR8_T
     _LIBCPP_PREFERRED_NAME(u8string)
 #endif
 #ifndef _LIBCPP_HAS_NO_UNICODE_CHARS
     _LIBCPP_PREFERRED_NAME(u16string)
     _LIBCPP_PREFERRED_NAME(u32string)
 #endif
     basic_string
-    : private __basic_string_common<true>
+    : private __basic_string_common<true> // This base class is historical, but it needs to remain for ABI compatibility
 {
 public:
     typedef basic_string                                 __self;
     typedef basic_string_view<_CharT, _Traits>           __self_view;
     typedef _Traits                                      traits_type;
     typedef _CharT                                       value_type;
     typedef _Allocator                                   allocator_type;
     typedef allocator_traits<allocator_type>             __alloc_traits;

libcxx/include/vector

 
 _LIBCPP_PUSH_MACROS
 #include <__undef_macros>
 
 
 _LIBCPP_BEGIN_NAMESPACE_STD
 
 template <bool>
-class _LIBCPP_TEMPLATE_VIS __vector_base_common
-{
-protected:
-    _LIBCPP_INLINE_VISIBILITY __vector_base_common() {}
-    _LIBCPP_NORETURN void __throw_length_error() const;
-    _LIBCPP_NORETURN void __throw_out_of_range() const;
-};
-
-template <bool __b>
-void
-__vector_base_common<__b>::__throw_length_error() const
-{
-    _VSTD::__throw_length_error("vector");
-}
-
-template <bool __b>
-void
-__vector_base_common<__b>::__throw_out_of_range() const
-{
-    _VSTD::__throw_out_of_range("vector");
-}
+struct __vector_base_common;
 
-_LIBCPP_EXTERN_TEMPLATE(class _LIBCPP_EXTERN_TEMPLATE_TYPE_VIS __vector_base_common<true>)
+template <>
+struct __vector_base_common<true> {
+    // Both are defined in vector.cpp
+    _LIBCPP_NORETURN _LIBCPP_EXPORTED_FROM_ABI void __throw_length_error() const;
+    _LIBCPP_NORETURN _LIBCPP_EXPORTED_FROM_ABI void __throw_out_of_range() const;
+};
 
 template <class _Tp, class _Allocator>
 class __vector_base
-    : protected __vector_base_common<true>
+    : protected __vector_base_common<true> // This base class is historical, but it needs to remain for ABI compatibility
 {
 public:
     typedef _Allocator                               allocator_type;
     typedef allocator_traits<allocator_type>         __alloc_traits;
     typedef typename __alloc_traits::size_type       size_type;
 protected:
     typedef _Tp                                      value_type;
     typedef value_type&                              reference;

libcxx/src/string.cpp

 #include "cerrno"
 #include "limits"
 #include "stdexcept"
 #include <stdio.h>
 #include "__debug"
 
 _LIBCPP_BEGIN_NAMESPACE_STD
 
-template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __basic_string_common<true>;
+void __basic_string_common<true>::__throw_length_error() const {
+    _VSTD::__throw_length_error("basic_string");
+}
+
+void __basic_string_common<true>::__throw_out_of_range() const {
+    _VSTD::__throw_out_of_range("basic_string");
+}
 
 #define _LIBCPP_EXTERN_TEMPLATE_DEFINE(...) template __VA_ARGS__;
 #ifdef _LIBCPP_ABI_STRING_OPTIMIZED_EXTERNAL_INSTANTIATION
 _LIBCPP_STRING_UNSTABLE_EXTERN_TEMPLATE_LIST(_LIBCPP_EXTERN_TEMPLATE_DEFINE, char)
 _LIBCPP_STRING_UNSTABLE_EXTERN_TEMPLATE_LIST(_LIBCPP_EXTERN_TEMPLATE_DEFINE, wchar_t)
 #else
 _LIBCPP_STRING_V1_EXTERN_TEMPLATE_LIST(_LIBCPP_EXTERN_TEMPLATE_DEFINE, char)
 _LIBCPP_STRING_V1_EXTERN_TEMPLATE_LIST(_LIBCPP_EXTERN_TEMPLATE_DEFINE, wchar_t)

libcxx/src/vector.cpp

 //===------------------------- vector.cpp ---------------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 
 #include "vector"
 
 _LIBCPP_BEGIN_NAMESPACE_STD
 
-template class _LIBCPP_CLASS_TEMPLATE_INSTANTIATION_VIS __vector_base_common<true>;
+void __vector_base_common<true>::__throw_length_error() const {
+    _VSTD::__throw_length_error("vector");
+}
+
+void __vector_base_common<true>::__throw_out_of_range() const {
+    _VSTD::__throw_out_of_range("vector");
+}
 
 _LIBCPP_END_NAMESPACE_STD