Changelog.md

@@ -1,6 +1,5 @@
 # Table of Contents
 - [VC Redist Lockdown](#vc-redist-lockdown)
-- [Expected in VS 2022 17.7 Preview 1](#expected-in-vs-2022-177-preview-1)
 - [Expected in VS 2022 17.6 Preview 3](#expected-in-vs-2022-176-preview-3)
 - [Expected in VS 2022 17.6 Preview 2](#expected-in-vs-2022-176-preview-2)
 - [Expected in VS 2022 17.6 Preview 1](#expected-in-vs-2022-176-preview-1)
@@ -25,7 +24,7 @@ To reduce the frequency of required reboots when updating VS (and VS Previews),
 
 * VS 2022 17.8 Preview 1
 
-# Expected in VS 2022 17.7 Preview 1
+# Expected in VS 2022 17.6 Preview 3
 - Merged C++23 features:
   * [P2505R5](https://wg21.link/P2505R5) [#3361](https://github.com/microsoft/STL/pull/3361) Monadic Functions For `expected`
   * [P2588R3](https://wg21.link/P2588R3) [#3455](https://github.com/microsoft/STL/pull/3455) `barrier`'s Phase Completion Guarantees
@@ -42,6 +41,7 @@ To reduce the frequency of required reboots when updating VS (and VS Previews),
   * Changed `vformat_to()` to instantiate its machinery in a lazier way. [#3403](https://github.com/microsoft/STL/pull/3403https://github.com/microsoft/STL/pull/3403)
     + When merely including `<format>` or `<chrono>`, this avoids instantiating dynamic initializers for facet IDs, and improves throughput slightly.
 - Improved performance:
+  * Optimized `<atomic>` for ARM64, massively improving load-acquire and store-release (measured 14.1x to 23.8x speedups - times not percent) and significantly improving sequentially consistent stores (measured 1.58x speedup). [#3399](https://github.com/microsoft/STL/pull/3399)
   * `ranges::find` now calls `memchr()` in more situations (when the vectorized implementation isn't available). [#3386](https://github.com/microsoft/STL/pull/3386)
 - Improved throughput:
   * Used `if constexpr` instead of overloads in internal machinery. [#3389](https://github.com/microsoft/STL/pull/3389)
@@ -62,10 +62,6 @@ To reduce the frequency of required reboots when updating VS (and VS Previews),
   * Fixed the build system to properly pass options when assembling the "alias objects". [#3402](https://github.com/microsoft/STL/pull/3402)
 - Updated `_MSVC_STL_UPDATE`. [#3395](https://github.com/microsoft/STL/pull/3395)
 
-# Expected in VS 2022 17.6 Preview 3
-- Improved performance:
-  * Optimized `<atomic>` for ARM64, massively improving load-acquire and store-release (measured 14.1x to 23.8x speedups - times not percent) and significantly improving sequentially consistent stores (measured 1.58x speedup). [#3399](https://github.com/microsoft/STL/pull/3399)
-
 # Expected in VS 2022 17.6 Preview 2
 - Merged C++23 features:
   * [P2404R3](https://wg21.link/P2404R3) [#3345](https://github.com/microsoft/STL/pull/3345) Move-Only Types For Comparison Concepts