From: Hans Wennborg <hans@hanshq.net>
Date: Fri, 23 May 2014 20:16:41 +0000 (+0000)
Subject: Don't suppress warning about dllimport on typedefs etc. in MicrosoftExt mode
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=98a84f6794a8928fee755d3d9f9cf78949d34027;p=clang

Don't suppress warning about dllimport on typedefs etc. in MicrosoftExt mode

It's true the MSVC doesn't warn about dllimport when applied to e.g. a typedef,
but that applies to dllexport too. I'd like us to be consistent, and I think
the right thing to do is to warn.

The original test that came with implementing the old behaviour doesn't provide
a good motivation, and it said it was checking that we're not repoting an *error*,
which is still true since this is just a warning.

There are plenty of tests e.g. in Sema/dllimport.c to check that we do warn
about dllimport on non functions or variables.

Differential Revision: http://reviews.llvm.org/D3832

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@209546 91177308-0d34-0410-b5e6-96231b3b80d8
---

diff --git a/include/clang/Basic/Attr.td b/include/clang/Basic/Attr.td
index 1b011b0686..ce8ca1a568 100644
--- a/include/clang/Basic/Attr.td
+++ b/include/clang/Basic/Attr.td
@@ -1665,8 +1665,7 @@ def DLLExport : InheritableAttr, TargetSpecificAttr<TargetWindows> {
 
 def DLLImport : InheritableAttr, TargetSpecificAttr<TargetWindows> {
   let Spellings = [Declspec<"dllimport">, GCC<"dllimport">];
-  // Technically, the subjects for DllImport are Function and Var, but there is
-  // custom semantic handling required when MicrosoftExt is true.
+  let Subjects = SubjectList<[Function, Var]>;
   let Documentation = [Undocumented];
 }
 
diff --git a/lib/Sema/SemaDeclAttr.cpp b/lib/Sema/SemaDeclAttr.cpp
index 9775c58639..e579b11f1d 100644
--- a/lib/Sema/SemaDeclAttr.cpp
+++ b/lib/Sema/SemaDeclAttr.cpp
@@ -3836,18 +3836,6 @@ DLLImportAttr *Sema::mergeDLLImportAttr(Decl *D, SourceRange Range,
 }
 
 static void handleDLLImportAttr(Sema &S, Decl *D, const AttributeList &Attr) {
-  // Attribute can be applied only to functions or variables.
-  FunctionDecl *FD = dyn_cast<FunctionDecl>(D);
-  if (!FD && !isa<VarDecl>(D)) {
-    // Apparently Visual C++ thinks it is okay to not emit a warning
-    // in this case, so only emit a warning when -fms-extensions is not
-    // specified.
-    if (!S.getLangOpts().MicrosoftExt)
-      S.Diag(Attr.getLoc(), diag::warn_attribute_wrong_decl_type)
-        << Attr.getName() << ExpectedVariableOrFunction;
-    return;
-  }
-
   unsigned Index = Attr.getAttributeSpellingListIndex();
   DLLImportAttr *NewAttr = S.mergeDLLImportAttr(D, Attr.getRange(), Index);
   if (NewAttr)
diff --git a/test/Rewriter/dllimport-typedef.c b/test/Rewriter/dllimport-typedef.c
deleted file mode 100644
index b587f16f3e..0000000000
--- a/test/Rewriter/dllimport-typedef.c
+++ /dev/null
@@ -1,17 +0,0 @@
-// RUN: not %clang_cc1 -triple i686-pc-win32 -fms-extensions -fsyntax-only %s 2>&1 | FileCheck -check-prefix=CHECK-NEG %s
-// RUN: not %clang_cc1 -triple i686-pc-win32 -fsyntax-only %s 2>&1 | FileCheck -check-prefix=CHECK-POS %s
-
-// Do not report an error with including dllimport in the typedef when -fms-extensions is specified.
-// Addresses <rdar://problem/7653870>.
-typedef __declspec(dllimport) int CB(void);
-
-// This function is added just to trigger a diagnostic.  This way we can test how many
-// diagnostics we expect.
-void bar() { return 1; }
-
-// CHECK-NEG: error: void function 'bar' should not return a value
-// CHECK-NEG: {{^}}1 error generated
-// CHECK-POS: warning: 'dllimport' attribute only applies to variables and functions
-// CHECK-POS: error: void function 'bar' should not return a value
-// CHECK-POS: 1 warning and 1 error generated
-