[Clang] Fix Microsoft ABI inheritance model when member pointer is used in a base specifier #91990

MitalAshok · 2024-05-13T16:19:00Z

Fix CXXRecordDecl::isParsingBaseSpecifiers so that it is true while parsing base specifiers instead of directly after they have been parsed.

-fcomplete-member-pointers now issues a diagnostic when a member pointer is used in a base specifier.

llvmbot · 2024-05-13T16:19:31Z

@llvm/pr-subscribers-clang

Author: Mital Ashok (MitalAshok)

Changes

Fix CXXRecordDecl::isParsingBaseSpecifiers so that it is true while parsing base specifiers instead of directly after they have been parsed.

-fcomplete-member-pointers now issues a diagnostic when a member pointer is used in a base specifier.

-fcomplete-member-pointers has also been relaxed to not issue a diagnostic for incomplete classes with an explicit __{single|multiple|virtual}_inheritance attribute, whose completeness would not affect the representation of pointer-to-member objects.

Full diff: https://github.com/llvm/llvm-project/pull/91990.diff

9 Files Affected:

(modified) clang/docs/ReleaseNotes.rst (+4)
(modified) clang/include/clang/AST/DeclCXX.h (+3-2)
(modified) clang/include/clang/Basic/DiagnosticSemaKinds.td (+2)
(modified) clang/lib/AST/DeclCXX.cpp (+1-5)
(modified) clang/lib/Parse/ParseDeclCXX.cpp (+22)
(modified) clang/lib/Sema/SemaDeclCXX.cpp (-3)
(modified) clang/lib/Sema/SemaType.cpp (+14-9)
(modified) clang/test/SemaCXX/complete-member-pointers.cpp (+23-1)
(modified) clang/test/SemaCXX/member-pointer-ms.cpp (+8)

diff --git a/clang/docs/ReleaseNotes.rst b/clang/docs/ReleaseNotes.rst
index 4702b8c10cdbb..054332c2cee4e 100644
--- a/clang/docs/ReleaseNotes.rst
+++ b/clang/docs/ReleaseNotes.rst
@@ -79,6 +79,10 @@ ABI Changes in This Version
 - Fixed Microsoft calling convention when returning classes that have a deleted
   copy assignment operator. Such a class should be returned indirectly.
 
+- Fixed Microsoft layout of pointer-to-members of classes when the layout is needed
+  directly or indirectly by the base classes of a class. These should use the most
+  general unspecified inheritance layout. Also affects -fcomplete-member-pointers.
+
 AST Dumping Potentially Breaking Changes
 ----------------------------------------
 
diff --git a/clang/include/clang/AST/DeclCXX.h b/clang/include/clang/AST/DeclCXX.h
index fb52ac804849d..45b3b47ed51c5 100644
--- a/clang/include/clang/AST/DeclCXX.h
+++ b/clang/include/clang/AST/DeclCXX.h
@@ -297,7 +297,8 @@ class CXXRecordDecl : public RecordDecl {
     LLVM_PREFERRED_TYPE(bool)
     unsigned IsLambda : 1;
 
-    /// Whether we are currently parsing base specifiers.
+    /// Whether we are currently parsing base specifiers; the
+    /// colon has been consumed but the beginning left brace hasn't.
     LLVM_PREFERRED_TYPE(bool)
     unsigned IsParsingBaseSpecifiers : 1;
 
@@ -598,7 +599,7 @@ class CXXRecordDecl : public RecordDecl {
     return !hasDefinition() || !isDynamicClass() || hasAnyDependentBases();
   }
 
-  void setIsParsingBaseSpecifiers() { data().IsParsingBaseSpecifiers = true; }
+  void setIsParsingBaseSpecifiers(bool to = true) { data().IsParsingBaseSpecifiers = to; }
 
   bool isParsingBaseSpecifiers() const {
     return data().IsParsingBaseSpecifiers;
diff --git a/clang/include/clang/Basic/DiagnosticSemaKinds.td b/clang/include/clang/Basic/DiagnosticSemaKinds.td
index 9e82130c93609..a9f9f02651cff 100644
--- a/clang/include/clang/Basic/DiagnosticSemaKinds.td
+++ b/clang/include/clang/Basic/DiagnosticSemaKinds.td
@@ -8003,6 +8003,8 @@ def err_bad_memptr_rhs : Error<
 def err_bad_memptr_lhs : Error<
   "left hand operand to %0 must be a %select{|pointer to }1class "
   "compatible with the right hand operand, but is %2">;
+def note_memptr_incomplete_until_bases : Note<
+  "this will affect the ABI of the member pointer until the bases have been specified">;
 def err_memptr_incomplete : Error<
   "member pointer has incomplete base type %0">;
 def warn_exception_caught_by_earlier_handler : Warning<
diff --git a/clang/lib/AST/DeclCXX.cpp b/clang/lib/AST/DeclCXX.cpp
index 75c441293d62e..7f20a47e6a054 100644
--- a/clang/lib/AST/DeclCXX.cpp
+++ b/clang/lib/AST/DeclCXX.cpp
@@ -474,10 +474,8 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
   if (data().IsStandardLayout && NumBases > 1 && hasRepeatedBaseClass(this))
     data().IsStandardLayout = false;
 
-  if (VBases.empty()) {
-    data().IsParsingBaseSpecifiers = false;
+  if (VBases.empty())
     return;
-  }
 
   // Create base specifier for any direct or indirect virtual bases.
   data().VBases = new (C) CXXBaseSpecifier[VBases.size()];
@@ -488,8 +486,6 @@ CXXRecordDecl::setBases(CXXBaseSpecifier const * const *Bases,
       addedClassSubobject(Type->getAsCXXRecordDecl());
     data().getVBases()[I] = *VBases[I];
   }
-
-  data().IsParsingBaseSpecifiers = false;
 }
 
 unsigned CXXRecordDecl::getODRHash() const {
diff --git a/clang/lib/Parse/ParseDeclCXX.cpp b/clang/lib/Parse/ParseDeclCXX.cpp
index 65ddebca49bc6..681b9803142c2 100644
--- a/clang/lib/Parse/ParseDeclCXX.cpp
+++ b/clang/lib/Parse/ParseDeclCXX.cpp
@@ -2334,6 +2334,26 @@ void Parser::ParseClassSpecifier(tok::TokenKind TagTokKind,
   }
 }
 
+namespace {
+  struct ParsingBaseSpecifiersGuard {
+    CXXRecordDecl *RD = nullptr;
+    ParsingBaseSpecifiersGuard(Sema& S, Decl *D) {
+      if (D) {
+        S.AdjustDeclIfTemplate(D);
+        RD = cast<CXXRecordDecl>(D);
+        assert(!RD->isParsingBaseSpecifiers() && "Recursively parsing base specifiers of the same class?");
+        RD->setIsParsingBaseSpecifiers(true);
+      }
+    }
+    ~ParsingBaseSpecifiersGuard() {
+      if (RD) {
+        assert(RD->isParsingBaseSpecifiers() && "Stopped parsing base specifiers before exiting ParseBaseClause?");
+        RD->setIsParsingBaseSpecifiers(false);
+      }
+    }
+  };
+}
+
 /// ParseBaseClause - Parse the base-clause of a C++ class [C++ class.derived].
 ///
 ///       base-clause : [C++ class.derived]
@@ -2345,6 +2365,8 @@ void Parser::ParseBaseClause(Decl *ClassDecl) {
   assert(Tok.is(tok::colon) && "Not a base clause");
   ConsumeToken();
 
+  ParsingBaseSpecifiersGuard Guard(Actions, ClassDecl);
+
   // Build up an array of parsed base specifiers.
   SmallVector<CXXBaseSpecifier *, 8> BaseInfo;
 
diff --git a/clang/lib/Sema/SemaDeclCXX.cpp b/clang/lib/Sema/SemaDeclCXX.cpp
index 53238d355ea09..0bf7208605213 100644
--- a/clang/lib/Sema/SemaDeclCXX.cpp
+++ b/clang/lib/Sema/SemaDeclCXX.cpp
@@ -2862,9 +2862,6 @@ BaseResult Sema::ActOnBaseSpecifier(Decl *classdecl, SourceRange SpecifierRange,
   if (!Class)
     return true;
 
-  // We haven't yet attached the base specifiers.
-  Class->setIsParsingBaseSpecifiers();
-
   // We do not support any C++11 attributes on base-specifiers yet.
   // Diagnose any attributes we see.
   for (const ParsedAttr &AL : Attributes) {
diff --git a/clang/lib/Sema/SemaType.cpp b/clang/lib/Sema/SemaType.cpp
index fddc3545ecb61..0a801f1797d23 100644
--- a/clang/lib/Sema/SemaType.cpp
+++ b/clang/lib/Sema/SemaType.cpp
@@ -9463,17 +9463,22 @@ bool Sema::RequireCompleteTypeImpl(SourceLocation Loc, QualType T,
 
   if (const MemberPointerType *MPTy = T->getAs<MemberPointerType>()) {
     if (!MPTy->getClass()->isDependentType()) {
-      if (getLangOpts().CompleteMemberPointers &&
-          !MPTy->getClass()->getAsCXXRecordDecl()->isBeingDefined() &&
-          RequireCompleteType(Loc, QualType(MPTy->getClass(), 0), Kind,
-                              diag::err_memptr_incomplete))
-        return true;
-
       // We lock in the inheritance model once somebody has asked us to ensure
       // that a pointer-to-member type is complete.
-      if (Context.getTargetInfo().getCXXABI().isMicrosoft()) {
-        (void)isCompleteType(Loc, QualType(MPTy->getClass(), 0));
-        assignInheritanceModel(*this, MPTy->getMostRecentCXXRecordDecl());
+      if (Context.getTargetInfo().getCXXABI().isMicrosoft() || getLangOpts().CompleteMemberPointers) {
+        QualType Class = QualType(MPTy->getClass(), 0);
+        (void)isCompleteType(Loc, Class);
+        CXXRecordDecl *RD = MPTy->getMostRecentCXXRecordDecl();
+        assignInheritanceModel(*this, RD);
+
+        if (getLangOpts().CompleteMemberPointers && RD->getMSInheritanceModel() == MSInheritanceModel::Unspecified) {
+          if (RD->hasDefinition() && RD->isParsingBaseSpecifiers()) {
+            Diag(Loc, diag::err_memptr_incomplete) << Class;
+            Diag(RD->getDefinition()->getLocation(), diag::note_memptr_incomplete_until_bases);
+          } else if (!RequireCompleteType(Loc, Class, diag::err_memptr_incomplete))
+            Diag(Loc, diag::err_memptr_incomplete) << Class;
+          return true;
+        }
       }
     }
   }
diff --git a/clang/test/SemaCXX/complete-member-pointers.cpp b/clang/test/SemaCXX/complete-member-pointers.cpp
index 942bb703a9223..41c9b98438c1a 100644
--- a/clang/test/SemaCXX/complete-member-pointers.cpp
+++ b/clang/test/SemaCXX/complete-member-pointers.cpp
@@ -1,4 +1,6 @@
-// RUN: %clang_cc1 -verify -fsyntax-only -fcomplete-member-pointers %s
+// RUN: %clang_cc1 -verify -fsyntax-only -fc++-abi=itanium -fms-extensions -fcomplete-member-pointers %s
+// RUN: %clang_cc1 -verify -fsyntax-only -triple=x86_64-unknown-win32 -fc++-abi=microsoft -fms-compatibility -fcomplete-member-pointers %s
+// RUN: %clang_cc1 -verify -fsyntax-only -triple=x86_64-unknown-win32 -fc++-abi=itanium -fms-compatibility -fcomplete-member-pointers %s
 
 struct S; // expected-note {{forward declaration of 'S'}}
 typedef int S::*t;
@@ -13,3 +15,23 @@ template <typename T>
 struct S3 {
   int T::*foo;
 };
+
+struct __single_inheritance S4;
+int S4::* baz;
+
+template<int I> struct Base {};
+struct __single_inheritance S5 : Base<sizeof(int S5::*)> {};
+struct
+S6  // #S6
+:
+Base<sizeof(int S6::*)>
+// expected-error@-1 {{member pointer has incomplete base type 'S6'}}
+//   expected-note@#S6 {{this will affect the ABI of the member pointer until the bases have been specified}}
+{
+};
+
+template<typename T> struct S7 {};
+int S7<void>::* qux;  // #qux
+template<> struct S7<void> {};
+// expected-error@-1 {{explicit specialization of 'S7<void>' after instantiation}}
+//   expected-note@#qux {{implicit instantiation first required here}}
diff --git a/clang/test/SemaCXX/member-pointer-ms.cpp b/clang/test/SemaCXX/member-pointer-ms.cpp
index 3271ff0c623a2..efbab8b28498f 100644
--- a/clang/test/SemaCXX/member-pointer-ms.cpp
+++ b/clang/test/SemaCXX/member-pointer-ms.cpp
@@ -215,6 +215,14 @@ struct NewUnspecified;
 SingleTemplate<void (IncSingle::*)()> tmpl_single;
 UnspecTemplate<void (NewUnspecified::*)()> tmpl_unspec;
 
+// Member pointers used in base specifiers force an unspecified inheritance model
+struct MemPtrInBase : UnspecTemplate<void (MemPtrInBase::*)()> {};
+#ifdef VMB
+struct __single_inheritance SpecifiedMemPtrInBase;
+struct SpecifiedMemPtrInBase : SingleTemplate<void (SpecifiedMemPtrInBase::*)()> {};
+struct __single_inheritance SpecifiedMemPtrInBase2 : SingleTemplate<void (SpecifiedMemPtrInBase2::*)()> {};
+#endif
+
 struct NewUnspecified { };
 
 static_assert(sizeof(void (NewUnspecified::*)()) == kUnspecifiedFunctionSize, "");

github-actions · 2024-05-13T16:21:56Z

✅ With the latest revision this PR passed the C/C++ code formatter.

MitalAshok · 2024-05-13T16:22:11Z

Example of the incompatibility: https://godbolt.org/z/Mn1T57WGb

struct unspecified_inheritance;
template<int I>
struct X {
    static_assert(I == sizeof(int unspecified_inheritance::*), "");
};

struct Y : X<sizeof(int Y::*)> {};

Currently, the sizeof(int Y::*) locks down the inheritance model of Y as single, when MSVC marks it as unspecified.

It was the intent that it should have been unspecified:

llvm-project/clang/lib/AST/MicrosoftCXXABI.cpp

Lines 223 to 224 in 89a080c

    
           if (!hasDefinition() || isParsingBaseSpecifiers()) 
        
             return MSInheritanceModel::Unspecified;

But isParsingBaseSpecifiers() was set after parsing instead of before.

…ed in a base specifier Fix CXXRecordDecl::isParsingBaseSpecifiers so that it is true while parsing base specifiers instead of directly after they have been parsed. -fcomplete-member-pointers now issues a diagnostic when a member pointer is used in a base specifier. -fcomplete-member-pointers has also been relaxed to not issue a diagnostic for incomplete classes with an explicit __{single|multiple|virtual}_inheritance attribute, whose completeness would not affect the representation of pointer-to-member objects.

…rosoft ABI

MaxEW707 · 2024-07-06T05:22:18Z

clang/include/clang/AST/DeclCXX.h

@@ -598,7 +599,9 @@ class CXXRecordDecl : public RecordDecl {
    return !hasDefinition() || !isDynamicClass() || hasAnyDependentBases();
  }

-  void setIsParsingBaseSpecifiers() { data().IsParsingBaseSpecifiers = true; }
+  void setIsParsingBaseSpecifiers(bool to = true) {


Suggested change

void setIsParsingBaseSpecifiers(bool to = true) {

void setIsParsingBaseSpecifiers(bool To = true) {

I am surprised clang-format didn't complain about starting with a lowercase for variables.
I can't think of a better name. I respond here if I do :).

… the Microsoft ABI" This reverts commit f343670. I'm planning on moving this to a separate pr which substantially changes -fcomplete-member-pointers

I'm planning on moving this to a separate pr which substantially changes -fcomplete-member-pointers

Endilll

Sema.h changes look good.

MaxEW707

LGTM.

@MitalAshok I noticed in your commit messages you mentioned reworking -fcomplete-member-pointers. Will that include fixing clang incorrectly warning when an inheritance model is explicitly specified? Godbolt for reference: https://godbolt.org/z/rYaWz6sra.

MitalAshok · 2024-07-08T11:15:29Z

Will that include fixing clang incorrectly warning when an inheritance model is explicitly specified?

Yes, I have a WIP of it here #98010 where that doesn't get a warning (https://github.com/llvm/llvm-project/pull/98010/files#diff-bec1ae40e5793adbd61e7a9dc41a4715a06e27888a384c0752bb6b4b09a1075fR20-R21)

MaxEW707 · 2024-07-16T03:08:57Z

@Endilll Wondering if this PR gets the green check from you.

I would like to get this merged before the clang-19 branch is created and just want to get the green light from a more senior member of the community before merging. Thanks :).

cor3ntin · 2024-07-17T14:58:04Z

@AaronBallman

AaronBallman

This generally LGTM but I'd love to hear from @rnk or @zmodem as they are quite knowledgeable about the MS ABI in general.

AaronBallman · 2024-07-17T16:58:56Z

clang/lib/Sema/SemaType.cpp

+        if (RD->isBeingDefined()) {
+          if (RD->isParsingBaseSpecifiers()) {


Combine these into a single statement?

I really only want to call RequireCompleteTypein the else branch when !RD->isBeingDefined() (when called while RD->isParsingBaseSpecifiers(), there's an extra note 'definition of "[class name]" is incomplete until the closing "}"', which is misleading since for CompleteMemberPointers purposes it's complete after the bases are parsed)

I'm struggling to understand why this error diagnostic doesn't fire in your MS ABI test case example that forms a member pointer type in the base specifier list.

rnk · 2024-07-17T22:29:42Z

clang/test/SemaCXX/complete-member-pointers.cpp

+S5 // #S5
+:
+Base<sizeof(int S5::*)>
+// expected-error@-1 {{member pointer has incomplete base type 'S5'}}


Should this be an error in Microsoft mode? Shouldn't we silently use the unspecified model in that mode?

It's an error for -fcomplete-member-pointers which is supposed to tell you when you use a member pointer in a way that the MS ABI would have it be unspecified, regardless of if you are compiling with MS member pointers or Itanium member pointers

I agree on the desired behavior, but doesn't this test show that, in the MS ABI mode, clang emits a hard error about an incomplete type? That's my reading of what the code does, and what the test shows.

Yes, that's just how -fcomplete-member-pointers works currently. It's more like a -Werror=... diagnostic

rnk · 2024-07-17T22:30:37Z

clang/test/SemaCXX/member-pointer-ms.cpp

@@ -215,6 +215,14 @@ struct NewUnspecified;
 SingleTemplate<void (IncSingle::*)()> tmpl_single;
 UnspecTemplate<void (NewUnspecified::*)()> tmpl_unspec;

+// Member pointers used in base specifiers force an unspecified inheritance model
+struct MemPtrInBase : UnspecTemplate<void (MemPtrInBase::*)()> {};


I guess this memptr type is not required to be complete during the parsing of the base specifiers here.

It's not for simply naming the type for the template argument but UnspecTemplate's definition does complete it by taking its sizeof, forcing the inheritance model to be assigned

Yes, but it's not clear to me if that happens before or after we've finished parsing the base specifiers.

I see what you mean.
If there were two bases struct MemPtrInBase : UnspecTemplate<void (MemPtrInBase::*)()>, SomeOtherClass {, UnspecTemplate<void (MemPtrInBase::*)()> would be instantiated and completed before the token for SomeOtherClass is seen. So this is the case for Clang, but I don't see why a different compiler couldn't parse all the bases at once then instantiated them.

However, they need to be instantiated at least before the body of the struct is parsed (because the bases affect name lookup), and this is essentially the same time as when you are "parsing the base specifiers".
And of course it needs to match Microsoft behaviour and assign the unspecified inheritance model during this time.

rnk · 2024-07-17T22:30:49Z

clang/lib/Sema/SemaType.cpp

+        if (RD->isBeingDefined()) {
+          if (RD->isParsingBaseSpecifiers()) {


I'm struggling to understand why this error diagnostic doesn't fire in your MS ABI test case example that forms a member pointer type in the base specifier list.

MaxEW707 · 2024-09-06T06:52:06Z

@rnk @zmodem Sorry for the ping. Its been a bit since this got reviewed and I recently hit this ABI case when moving some parts of a larger project to clang.

rnk

Sorry, looks like I wrote pending comments and didn't publish them

rnk · 2024-07-19T17:55:07Z

clang/test/SemaCXX/complete-member-pointers.cpp

+S5 // #S5
+:
+Base<sizeof(int S5::*)>
+// expected-error@-1 {{member pointer has incomplete base type 'S5'}}


I agree on the desired behavior, but doesn't this test show that, in the MS ABI mode, clang emits a hard error about an incomplete type? That's my reading of what the code does, and what the test shows.

rnk · 2024-07-19T17:55:40Z

clang/test/SemaCXX/member-pointer-ms.cpp

@@ -215,6 +215,14 @@ struct NewUnspecified;
 SingleTemplate<void (IncSingle::*)()> tmpl_single;
 UnspecTemplate<void (NewUnspecified::*)()> tmpl_unspec;

+// Member pointers used in base specifiers force an unspecified inheritance model
+struct MemPtrInBase : UnspecTemplate<void (MemPtrInBase::*)()> {};


Yes, but it's not clear to me if that happens before or after we've finished parsing the base specifiers.

llvmbot added clang Clang issues not falling into any other category clang:frontend Language frontend issues, e.g. anything involving "Sema" labels May 13, 2024

MitalAshok force-pushed the ms-member-pointer-in-base branch from a534708 to 5dc9193 Compare May 13, 2024 16:22

MitalAshok added 2 commits June 11, 2024 14:55

Merge branch 'main' into ms-member-pointer-in-base

85a823a

Add more places where -fcomplete-member-pointers should match the Mic…

f343670

…rosoft ABI

MaxEW707 reviewed Jul 6, 2024

View reviewed changes

MitalAshok added 4 commits July 6, 2024 09:48

Merge branch 'main' into ms-member-pointer-in-base

ff6a2b8

Revert "Add more places where -fcomplete-member-pointers should match…

66c8658

… the Microsoft ABI" This reverts commit f343670. I'm planning on moving this to a separate pr which substantially changes -fcomplete-member-pointers

Partially revert changes to -fcomplete-member-pointers

7058fb4

I'm planning on moving this to a separate pr which substantially changes -fcomplete-member-pointers

Manipulate the CXXRecordDecl in Sema instead of Parse

7c619ab

MitalAshok requested a review from Endilll as a code owner July 6, 2024 09:59

Endilll reviewed Jul 6, 2024

View reviewed changes

MaxEW707 approved these changes Jul 8, 2024

View reviewed changes

MaxEW707 requested a review from Sirraide July 9, 2024 02:17

Endilll requested a review from AaronBallman July 16, 2024 12:05

AaronBallman requested a review from rnk July 17, 2024 16:54

AaronBallman reviewed Jul 17, 2024

View reviewed changes

rnk reviewed Jul 17, 2024

View reviewed changes

rnk reviewed Sep 6, 2024

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Clang] Fix Microsoft ABI inheritance model when member pointer is used in a base specifier #91990

[Clang] Fix Microsoft ABI inheritance model when member pointer is used in a base specifier #91990

MitalAshok commented May 13, 2024 •

edited

Loading

llvmbot commented May 13, 2024

github-actions bot commented May 13, 2024 •

edited

Loading

MitalAshok commented May 13, 2024

MaxEW707 Jul 6, 2024

Endilll left a comment

MaxEW707 left a comment •

edited

Loading

MitalAshok commented Jul 8, 2024

MaxEW707 commented Jul 16, 2024

cor3ntin commented Jul 17, 2024

AaronBallman left a comment

AaronBallman Jul 17, 2024

MitalAshok Jul 17, 2024

rnk Jul 17, 2024

rnk Jul 17, 2024

MitalAshok Jul 18, 2024

rnk Jul 19, 2024

MitalAshok Sep 6, 2024

rnk Jul 17, 2024

MitalAshok Jul 18, 2024

rnk Jul 19, 2024

MitalAshok Sep 6, 2024

rnk Jul 17, 2024

MaxEW707 commented Sep 6, 2024

rnk left a comment

rnk Jul 19, 2024

rnk Jul 19, 2024

	void setIsParsingBaseSpecifiers(bool to = true) {
	void setIsParsingBaseSpecifiers(bool To = true) {

		if (RD->isBeingDefined()) {
		if (RD->isParsingBaseSpecifiers()) {

[Clang] Fix Microsoft ABI inheritance model when member pointer is used in a base specifier #91990

Are you sure you want to change the base?

[Clang] Fix Microsoft ABI inheritance model when member pointer is used in a base specifier #91990

Conversation

MitalAshok commented May 13, 2024 • edited Loading

llvmbot commented May 13, 2024

github-actions bot commented May 13, 2024 • edited Loading

MitalAshok commented May 13, 2024

Choose a reason for hiding this comment

Endilll left a comment

Choose a reason for hiding this comment

MaxEW707 left a comment • edited Loading

Choose a reason for hiding this comment

MitalAshok commented Jul 8, 2024

MaxEW707 commented Jul 16, 2024

cor3ntin commented Jul 17, 2024

AaronBallman left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

MaxEW707 commented Sep 6, 2024

rnk left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

MitalAshok commented May 13, 2024 •

edited

Loading

github-actions bot commented May 13, 2024 •

edited

Loading

MaxEW707 left a comment •

edited

Loading