[Git][wine/wine][master] 2 commits: Revert "kernel32/tests: Don't use _ReadWriteBarrier on ARM platforms."

[Alexandre Julliard (＠julliard)]

Alexandre Julliard pushed to branch master at wine / wine

Commits:
615afdc2 by Jinoh Kang at 2025-04-28T16:31:57+02:00
Revert "kernel32/tests: Don't use _ReadWriteBarrier on ARM platforms."

This reverts commit 504a305a60849704f2d83b55e9aabb26d89edf11.

The choice of weak barrier was by design. The stronger barrier broke the
positive half of the litmus test for FlushProcessWriteBuffers():
> virtual.c:4568: Test failed: expected write-read reordering with compiler barrier
>  only (got 0 reorderings)

FlushProcessWriteBuffers() is used in tandem with a compiler-only
barrier to implement fast synchronization, where only one side is
executed frequently and the other side only occassionally (e.g., garbage
collection).  Prominent applications include .NET and HotSpot JVM.

Although _ReadWriteBarrier() is "marked deprecated," it is in fact used
by Microsoft's own STL library (with explicit suppression of warning) to
implement a compiler-only memory barrier [1].  The deprecation notice
suggests C++11 atomics as the (only?) alternative.

[1]: https://github.com/microsoft/STL/blob/b5df16a98934319e2e6864d6036cbe9cd9c74f...

- - - - -
8db355e6 by Jinoh Kang at 2025-04-28T16:32:04+02:00
kernel32/tests: Don't use _ReadWriteBarrier() on clang.

Some clang versions do not support _ReadWriteBarrier() when targeting
armv7-windows (MSVC mode).

- - - - -

1 changed file:

- dlls/kernel32/tests/virtual.c

View it on GitLab: https://gitlab.winehq.org/wine/wine/-/compare/f46c44e0364080ef3c7aa726b9b6f8...