[libcxx-commits] [libcxx] [libc++][hardening] Categorize assertions related to strict weak ordering (PR #77405)

[Konstantin Varlamov via libcxx-commits]

================
@@ -341,23 +354,31 @@ _LIBCPP_HARDENING_MODE_DEBUG
 #    define _LIBCPP_ASSERT_VALID_ELEMENT_ACCESS(expression, message)     _LIBCPP_ASSERT(expression, message)
 #    define _LIBCPP_ASSERT_NON_NULL(expression, message)                 _LIBCPP_ASSERT(expression, message)
 #    define _LIBCPP_ASSERT_NON_OVERLAPPING_RANGES(expression, message)   _LIBCPP_ASSERT(expression, message)
+#    define _LIBCPP_ASSERT_ARGUMENT_WITHIN_DOMAIN(expression, message)   _LIBCPP_ASSERT(expression, message)
 #    define _LIBCPP_ASSERT_COMPATIBLE_ALLOCATOR(expression, message)     _LIBCPP_ASSERT(expression, message)
 #    define _LIBCPP_ASSERT_UNCATEGORIZED(expression, message)            _LIBCPP_ASSERT(expression, message)
 #    define _LIBCPP_ASSERT_PEDANTIC(expression, message)                 _LIBCPP_ASSERT(expression, message)
 // Disabled checks.
+#    define _LIBCPP_ASSERT_INTRUSIVE(expression, message)                _LIBCPP_ASSUME(expression)
 #    define _LIBCPP_ASSERT_INTERNAL(expression, message)                 _LIBCPP_ASSUME(expression)
 
 // Debug hardening mode checks.
 
 #  elif _LIBCPP_HARDENING_MODE == _LIBCPP_HARDENING_MODE_DEBUG
 
+#ifndef _LIBCPP_DEBUG_STRICT_WEAK_ORDERING_CHECK
+#define _LIBCPP_DEBUG_STRICT_WEAK_ORDERING_CHECK
----------------
var-const wrote:

This restores the approach from the original patch (https://reviews.llvm.org/D150264). I think it's a very interesting question: do we want separate knobs for these checks, i.e., are they so expensive that users might not want to enable them even in the debug mode? Or should we just unconditionally enable these checks in the debug mode and remove/phase out the macro?

https://github.com/llvm/llvm-project/pull/77405