…which could cause double-destruction/undefined behavior.

…its to ensure proper error handling visibility.

- Migrate BVH_BuildQueue to std::mutex and std::lock_guard (RAII) and simplify Fetch/Enqueue implementations.
- Fix unique_lock construction in BRepCheck_Analyzer to use aFaceEdgeRes->GetMutex().
- Initialize TreatmentFunctor::Mutex properly (default construct instead of NULL).
- Add #include <unistd.h> on POSIX for mmap usage in Standard_MMgrOpt.
- Minor header fixes (Message_ProgressIndicator and others) and additional Standard_ErrorHandler includes in various modules.

Ensure signal symbols are available when compiling non-Windows sources by including <signal.h> earlier in OSD_signal.cxx.

…ders

- Make Standard_ErrorHandler::GetMutex() return a heap-allocated static std::mutex
  to avoid destruction-order issues during static deinitialization.
- Qualify GetMutex() definition in the .cxx and add its declaration & explanatory comment in the .hxx.
- Clean up Callback method declarations formatting in Standard_ErrorHandler.hxx.
- Add missing NCollection_Array1/NCollection_Shared includes to Select3D_SensitivePrimitiveArray.hxx.

CallHandler already acquires THE_SIGNAL_MUTEX, so remove the extra std::lock_guard
from TranslateSE to avoid double-locking and potential deadlock/reentrancy during
structured-exception translation.

…rTest_ContinuousRedrawer

… adopt shared locking

- Standard_ErrorHandler: remove GetMutex() and the old mutex pointer; introduce a namespace-local std::shared_mutex (THE_GLOBAL_MUTEX).
  Use std::shared_lock for read-only FindHandler path and std::unique_lock for modifying paths (constructor, Unlink, FindHandler with unlink). Add comments/TODO about possible thread_local optimization.

- Aspect_VKeySet (hxx/cxx): replace Standard_Mutex with std::shared_mutex; use std::shared_lock in const readers and std::lock_guard/std::unique_lock for writers.
  Add KeyDown_Unlocked / KeyUp_Unlocked helpers and thin locked wrappers to avoid double-locking.

- Graphic3d_MediaTexture / Graphic3d_MediaTextureSet (hxx/cxx): replace Handle(Standard_HMutex)/Standard_HMutex with std::mutex / std::mutex&; initialize mutex as std::mutex(); switch to std::unique_lock<std::mutex> for critical sections; tidy headers (include <mutex>, <Standard_MemoryUtils.hxx> as needed).

- AIS_ViewController: update lock to std::unique_lock<std::shared_mutex> when accessing myKeys.Mutex().

General: migrate away from custom mutex types to standard C++ mutexes, enable concurrent read access where beneficial, and simplify locking patterns.

- Replace Standard_Mutex member with std::mutex and adjust includes.
- Replace Standard_Mutex::Sentry usages with std::lock_guard<std::mutex>.
- Remove redundant lock in pushPlayEvent and add comment that caller must hold the mutex.

…Prs modules

Replace legacy Standard_Mutex usage with std::mutex across:
- StdPrs_BRepFont (header/member/Mutex() accessor and RenderGlyph lock)
- StdPrs_BRepTextBuilder (use std::lock_guard for font mutex)
- StdPrs_WFShape (functor mutex type and locking)

Also update includes to <mutex> and mark mutex mutable where required.

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Signed-off-by: Pasukhin Dmitry <pasuhinsvzn@gmail.com>

This reverts commit 1899df7.

- Add -D CMAKE_CXX_STANDARD=17 and -D BUILD_CPP_STANDARD="C++17" to both
  Windows and Linux configure steps in .github/actions/build-tinspector/action.yml.
- Remove unnecessary mutex locks in AIS_ViewController::handleViewOrientationKeys
  and StdPrs_BRepTextBuilder::Perform to avoid redundant locking / potential
  dead-locks.

…erbose make install

Replace std::unique_lock with std::lock_guard in several places to simplify locking:
- DE_PluginHolder.hxx: ctor/dtor
- Graphic3d_MediaTexture.cxx: GetImage
- Graphic3d_MediaTextureSet.cxx: LockFrame, ReleaseFrame, SwapFrames

Also replace manual myMutex.lock()/unlock() in Standard_MMgrOpt::Free with a scoped std::lock_guard and a narrowed block scope to improve exception safety and avoid potential deadlocks.

- Replace volatile boolean flags with std::atomic<bool> in BRepExtrema_DistShapeShape,
  use atomic load/store (with memory_order) inside TreatmentFunctor and set up local
  atomic copies for the parallel functor, copying results back after the parallel loop.
- Add missing <mutex> include to XSControl_WorkSession for proper use of std::lock_guard.
- Remove unused <Standard_MemoryUtils.hxx> include from Graphic3d_MediaTextureSet.

Move Pause() and Resume() implementations to the source file, add Standard_EXPORT to their declarations and protect pushPlayEvent with std::lock_guard<std::mutex> to ensure proper synchronization.

Add explicit <mutex> includes for headers that use std::mutex to avoid
relying on transitive includes and ensure correct compilation with
stricter include ordering.