[clang] Claim conformance for CWG466

[Endilll]

CWG466 "cv-qualifiers on pseudo-destructor type".
Richard claimed that we don't implement this DR because of one ill-formed example being accepted: a->CI::~VI();. This example is testing the behavior of calling a pseudo-destructor via a qualified name, where components of qualified name denote the same int type, but with different cv-qualifications.

Initially, the following wording from [expr.pseudo] quoted in CWG466 was left intact:

Furthermore, the two type-names in a pseudo-destructor-name of the form

:: (opt) nested-name-specifier (opt) type-name ::~ type-name

shall designate the same scalar type.

According to this wording, the example is indeed ill-formed.

P1131R2 merged wording for pseudo-destructors into regular destructor wording. Among other things, [expr.pseudo] was removed, and [expr.prim.id.qual]/2 was changed to read:

Where type-name ::~ type-name is used, the two type-names shall refer to the same type (ignoring cv-qualifications);

I believe P1131R2 made the example well-formed.

However, this wording didn't survive P1787R6 "Declarations and where to find them". In that paper I don't see the intent to make the example ill-formed again, so I assume it confirmed status-quo via other wording. My guess the new wording is http://eel.is/c++draft/basic.lookup#qual.general-4.6:

If a qualified name Q follows a ~:

• <...>
• The type-name that is or contains Q shall refer to its (original) lookup context (ignoring cv-qualification) under the interpretation established by at least one (successful) lookup performed.

[llvmbot]

@llvm/pr-subscribers-clang

Author: Vlad Serebrennikov (Endilll)

Changes

CWG466 "cv-qualifiers on pseudo-destructor type".
Richard claimed that we don't implement this DR because of one ill-formed example being accepted: a->CI::~VI();. This example is testing the behavior of calling a pseudo-destructor via a qualified name, where components of qualified name denote the same int type, but with different cv-qualifications.

Initially, the following wording from [expr.pseudo] quoted in CWG466 was left intact:
> Furthermore, the two type-names in a pseudo-destructor-name of the form
>
> :: (opt) nested-name-specifier (opt) type-name ::~ type-name
>
>shall designate the same scalar type.

According to this wording, the example is indeed ill-formed.

P1131R2 merged wording for pseudo-destructors into regular destructor wording. Among other things, [expr.pseudo] was removed, and [expr.prim.id.qual]/2 was changed to read:

> Where type-name ::~ type-name is used, the two type-names shall refer to the same type (ignoring cv-qualifications);

I believe P1131R2 made the example well-formed.

However, this wording didn't survive P1787R6 "Declarations and where to find them". In that paper I don't see the intent to make the example ill-formed again, so I assume it confirmed status-quo via other wording. My guess the new wording is http://eel.is/c++draft/basic.lookup#qual.general-4.6:
> If a qualified name Q follows a ~:
> - <...>
> - The type-name that is or contains Q shall refer to its (original) lookup context (ignoring cv-qualification) under the interpretation established by at least one (successful) lookup performed.

---

Full diff: https://github.com/llvm/llvm-project/pull/87748.diff

2 Files Affected:

• (modified) clang/test/CXX/drs/dr4xx.cpp (+2-2)
• (modified) clang/www/cxx_dr_status.html (+1-1)

diff --git a/clang/test/CXX/drs/dr4xx.cpp b/clang/test/CXX/drs/dr4xx.cpp
index 343c4ee6f3344e..34dd638c1d9b00 100644
--- a/clang/test/CXX/drs/dr4xx.cpp
+++ b/clang/test/CXX/drs/dr4xx.cpp
@@ -948,7 +948,7 @@ namespace dr460 { // dr460: yes
 // dr464: na
 // dr465: na
 
-namespace dr466 { // dr466: no
+namespace dr466 { // dr466: 2.8
   typedef int I;
   typedef const int CI;
   typedef volatile int VI;
@@ -960,7 +960,7 @@ namespace dr466 { // dr466: no
     a->CI::~CI();
     a->VI::~VI();
 
-    a->CI::~VI(); // FIXME: This is invalid; CI and VI are not the same scalar type.
+    a->CI::~VI(); // allowed by changes to [expr.id.prim.qual]/2 introduced in P1131R2
 
     b->~I();
     b->~CI();
diff --git a/clang/www/cxx_dr_status.html b/clang/www/cxx_dr_status.html
index 5e1e03dec1d484..887eaa1bdcd55f 100755
--- a/clang/www/cxx_dr_status.html
+++ b/clang/www/cxx_dr_status.html
@@ -2836,7 +2836,7 @@ <h2 id="cxxdr">C++ defect report implementation status</h2>
     <td><a href="https://cplusplus.github.io/CWG/issues/466.html">466</a></td>
     <td>CD1</td>
     <td>cv-qualifiers on pseudo-destructor type</td>
-    <td class="none" align="center">No</td>
+    <td class="full" align="center">Clang 2.8</td>
   </tr>
   <tr id="467">
     <td><a href="https://cplusplus.github.io/CWG/issues/467.html">467</a></td>

[cor3ntin]

I agree with that (the new wording in [basic.lookup.qual.general] ignore cv qualifications)