[libcxx-commits] [libcxx] [libcxx][algorithm] Optimize std::stable_sort via radix sort algorithm (PR #104683)

[Nikolas Klauser via libcxx-commits]

=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?=,
=?utf-8?b?0JTQvNC40YLRgNC40Lkg0JjQtw=?Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/104683 at github.com>


================
@@ -45,7 +47,19 @@ struct __stable_sort {
     auto __last_iter = ranges::next(__first, __last);
 
     auto&& __projected_comp = std::__make_projected(__comp, __proj);
-    std::__stable_sort_impl<_RangeAlgPolicy>(std::move(__first), __last_iter, __projected_comp);
+    constexpr auto __default_comp =
+        __desugars_to_v<__totally_ordered_less_tag, _Comp, iter_value_t<_Iter>, iter_value_t<_Iter> >;
+    constexpr auto __default_proj        = __is_identity<_Proj>::value;
+    constexpr auto __integral_value      = is_integral_v<iter_value_t<_Iter>>;
+    constexpr auto __integral_projection = __default_proj && __integral_value;
----------------
philnik777 wrote:

If we do the dispatching in `__stable_sort_impl` instead we don't need to do the dispatching twice.

https://github.com/llvm/llvm-project/pull/104683