Built-in operator!= reported ambiguous when the operand converts through different conversion functions

[danieljennings]

Issue Summary

EDG rejects h != p as ambiguous when h has both operator A*() and operator const void*() and p is a const A*. The two conversions go through different conversion functions, so per [over.ics.rank]/3.3 neither is better for h, and the exact match on p should make operator!=(const A*, const A*) the best candidate, as GCC, Clang and MSVC agree.

Reproducing Source Code

// EDG says h != p is ambiguous. GCC, Clang and MSVC accept it.
// EDG also accepts it if operator const void*() is removed.
//
// The built-in candidates [over.built]/15 include operator!=(const A*, const A*) and
// operator!=(const void*, const void*).
// - h: operator A*() vs operator const void*(). Two different conversion functions, so
//   neither conversion sequence is better [over.ics.rank]/3.3.
// - p: Exact Match for const A*, Conversion for const void*, so const A* is better
//   [over.ics.rank]/4.
// So operator!=(const A*, const A*) is the best viable function [over.match.best]/2.

struct A {};
struct H
{
    operator A *() const;
    operator const void *() const;
};

bool Test( const H &h, const A *p ) { return h != p; } // error: more than one operator "!=" matches these operands:

Command-line Options

cpfe --c++20 test.cpp

Type of Issue

[Front End] This code SHOULD compile with these options.

Additional Details

(--c++17 also repros it)

Demonstration that EDG disagrees with Clang/MSVC/GCC: https://godbolt.org/z/onesYfexc

I don't know the EDG configuration that applies to VS2026 Intellisense, but it also repros the problem.

Reproducing (Standard) Configuration(s)

linux-clang-release, windows-msvc-release

Reproducing (Non-standard) Configuration

No response

I acknowledge that:

• I have checked open issues and existing discussions before filing this report.
• The reproducing source code that I have provided does not depend on headers that I have not included with my submission.
• I have reduced the reproducing source code to the best of my ability.
• I have selected at least one standard configuration or uploaded a non-standard configuration dump.
• I understand code shared via this form is NOT confidential.
• I have permission to share the reproducing source code and grant permission for the project to use the source code (or derivatives of the source code) to create tests.
• I agree that any tests created using the reproducing source code (or deriatives of it) will be licensed under project license unless I included a LICENSE.txt file in my .zip or .tar.gz upload. If I included a LICENSE.txt file, any tests created using the reproducing source code will be governed by the terms of the license I provided instead.