[all-commits] [llvm/llvm-project] 539117: Complete complete types early when importing types...

adrian-prantl via All-commits all-commits at lists.llvm.org
Fri Nov 22 10:00:43 PST 2019


  Branch: refs/heads/master
  Home:   https://github.com/llvm/llvm-project
  Commit: 539117616d7a4c5690f1b9284faf0d282cd79621
      https://github.com/llvm/llvm-project/commit/539117616d7a4c5690f1b9284faf0d282cd79621
  Author: Adrian Prantl <aprantl at apple.com>
  Date:   2019-11-22 (Fri, 22 Nov 2019)

  Changed paths:
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/Makefile
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/TestClangModulesAppUpdate.py
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/foo.m
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/main.m
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/module.modulemap
    A lldb/packages/Python/lldbsuite/test/lang/objc/modules-app-update/umbrella.h
    M lldb/source/Plugins/SymbolFile/DWARF/DWARFASTParserClang.cpp

  Log Message:
  -----------
  Complete complete types early when importing types from Clang module DWARF.

This affects -gmodules only.

Under normal operation pcm_type is a shallow forward declaration
that gets completed later. This is necessary to support cyclic
data structures. If, however, pcm_type is already complete (for
example, because it was loaded for a different target before),
the definition needs to be imported right away, too.
Type::ResolveClangType() effectively ignores the ResolveState
inside type_sp and only looks at IsDefined(), so it never calls
ClangASTImporter::ASTImporterDelegate::ImportDefinitionTo(),
which does extra work for Objective-C classes. This would result
in only the forward declaration to be visible.

An alternative implementation would be to sink this into Type::ResolveClangType ( https://github.com/llvm/llvm-project/blob/88235812a71d99c082e7aa2ef9356d43d1f83a80/lldb/source/Symbol/Type.cpp#L5809) though it isn't clear to me how to best do this from a layering perspective.

rdar://problem/52134074

Differential Revision: https://reviews.llvm.org/D70415




More information about the All-commits mailing list