EDG tracking issue

Reported

EDG rejects requires-clause in a definition within a function template edgcpp/compiler#264
EDG's `__is_convertible_to` and `default_initializable` mishandle various scenarios edgcpp/compiler#266
EDG's `std::nullptr_t` has relational operators, but it should not edgcpp/compiler#267
EDG `__builtin_is_implicit_lifetime` isn't checking for trivial constructors edgcpp/compiler#268
EDG can incorrectly allow pointers to base and derived classes to be subtracted edgcpp/compiler#269
EDG's `__is_constructible` mishandles an explicit ctor being overloaded with a deleted ctor edgcpp/compiler#270
EDG hasn't implemented CWG-2256 "Lifetime of trivially-destructible objects" edgcpp/compiler#271
Variable template is instantiated too early edgcpp/compiler#272
EDG doesn't detect narrowing conversions from a user-defined type to a pointer type to `bool` edgcpp/compiler#273
EDG's `__is_trivially_assignable` mishandles a case that affects `std::ranges::single_view` edgcpp/compiler#274
P2590R2 Explicit Lifetime Management edgcpp/compiler#275
EDG rejects `std::expected` trying to propagate triviality of assignment operations edgcpp/compiler#276
Layout for `[[no_unique_address]]` members is not optimal under Cfront-like layout edgcpp/compiler#277
Catastrophic error involving `constexpr` `std::string` in a `union` edgcpp/compiler#278
EDG constexpr evaluation improperly handles integral conversions and narrow types edgcpp/compiler#283
EDG's overload resolution incorrectly considers some ambiguous conversions to be unambiguous edgcpp/compiler#284
EDG's `__is_aggregate` builtin should be updated to match LWG-3823's preconditions edgcpp/compiler#285
EDG mishandles the element type in constexpr allocation edgcpp/compiler#286

Need to report

We have workarounds for the following bugs, which we should reduce to be library-free and report to the edgcpp/compiler repo.

VSO-1951821 EDGcpfe/27063 EDG instantiates the wrong overload of `promise_type::operator new` for `generator` machinery

STL/tests/std/tests/P2502R2_generator/test.cpp

Lines 273 to 286 in da52dc0

    
           #ifndef __EDG__ // TRANSITION, VSO-1951821 
        
               { 
        
                   auto g = [](allocator_arg_t, StatefulAlloc<int>, const int hi) -> generator<int, int, StatefulAlloc<char>> { 
        
                       constexpr size_t n = 64; 
        
                       int some_ints[n]; 
        
                       for (int i = 0; i < hi; ++i) { 
        
                           co_yield some_ints[i % n] = i; 
        
                       } 
        
                   }; 
        
                   test_one<gen_traits<int, int, StatefulAlloc<char>>, int, int, int>( 
        
                       g(allocator_arg, StatefulAlloc<int>{42}, 1024), views::iota(0, 1024)); 
        
               } 
        
           #endif // ^^^ no workaround ^^^

Modules

There are an unknown number of blocking issues for EDG consuming header units and named modules. Currently, EDG coverage is disabled in the header units test, and hasn't been added to the named modules test.

Fixed, need to remove workarounds

VSO-2931699 EDG should support if consteval downlevel in C++20 mode for parity with C1XX/Clang
- Fixed by VS-PR-737109 on 2026-05-12. We probably can't remove workarounds until CUDA gets this.

Navigation Menu

Repository navigation

EDG tracking issue #1621

Reported

Need to report

VSO-1951821 EDGcpfe/27063 EDG instantiates the wrong overload of `promise_type::operator new` for `generator` machinery

Modules

Fixed, need to remove workarounds

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

	#ifndef __EDG__ // TRANSITION, VSO-1951821
	{
	auto g = [](allocator_arg_t, StatefulAlloc<int>, const int hi) -> generator<int, int, StatefulAlloc<char>> {
	constexpr size_t n = 64;
	int some_ints[n];
	for (int i = 0; i < hi; ++i) {
	co_yield some_ints[i % n] = i;
	}
	};

	test_one<gen_traits<int, int, StatefulAlloc<char>>, int, int, int>(
	g(allocator_arg, StatefulAlloc<int>{42}, 1024), views::iota(0, 1024));
	}
	#endif // ^^^ no workaround ^^^

EDG tracking issue #1621

Description

Reported

Need to report

VSO-1951821 EDGcpfe/27063 EDG instantiates the wrong overload of promise_type::operator new for generator machinery

Modules

Fixed, need to remove workarounds

Activity

StephanTLavavej commented on Feb 17, 2022

cpplearner commented on Oct 8, 2023

cpplearner commented on Oct 8, 2023

cpplearner commented on Oct 8, 2023

cpplearner commented on Oct 8, 2023

cpplearner commented on Oct 8, 2023

cpplearner commented on Oct 9, 2023

cpplearner commented on Oct 9, 2023

StephanTLavavej commented on Oct 10, 2023

cpplearner commented on Oct 10, 2023

cpplearner commented on Oct 10, 2023

cpplearner commented on Oct 11, 2023

StephanTLavavej commented on Oct 12, 2023

cpplearner commented on Oct 12, 2023

cpplearner commented on Jan 5, 2024

cpplearner commented on Jan 5, 2024

CaseyCarter commented on Jan 18, 2024

CaseyCarter commented on Jan 23, 2024

cpplearner commented on Feb 3, 2024

StephanTLavavej commented on Feb 5, 2024

cpplearner commented on Oct 1, 2026

cpplearner commented on Oct 11, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions

VSO-1951821 EDGcpfe/27063 EDG instantiates the wrong overload of `promise_type::operator new` for `generator` machinery