[cfe-commits] r147925 - in /cfe/trunk: include/clang/Basic/DiagnosticSemaKinds.td lib/Sema/SemaDecl.cpp test/Preprocessor/line-directive.c test/Sema/c11-typedef-redef.c

[Douglas Gregor]

On Jan 11, 2012, at 11:49 AM, Eli Friedman wrote:

> On Wed, Jan 11, 2012 at 11:44 AM, Matthieu Monrocq
> <matthieu.monrocq at gmail.com> wrote:
>> 
>> 
>> Le 11 janvier 2012 05:25, Douglas Gregor <dgregor at apple.com> a écrit :
>> 
>>> Author: dgregor
>>> Date: Tue Jan 10 22:25:01 2012
>>> New Revision: 147925
>>> 
>>> URL: http://llvm.org/viewvc/llvm-project?rev=147925&view=rev
>>> Log:
>>> C11 allows typedefs to be redefined. Implement this in C11 mode, and
>>> downgrade the default-error warning to an ExtWarn in
>>> C90/99. <rdar://problem/10668057>
>>> 
>>> Added:
>>>    cfe/trunk/test/Sema/c11-typedef-redef.c   (with props)
>>> Modified:
>>>    cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
>>>    cfe/trunk/lib/Sema/SemaDecl.cpp
>>>    cfe/trunk/test/Preprocessor/line-directive.c
>>> 
>>> Modified: cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td?rev=147925&r1=147924&r2=147925&view=diff
>>> 
>>> ==============================================================================
>>> --- cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td (original)
>>> +++ cfe/trunk/include/clang/Basic/DiagnosticSemaKinds.td Tue Jan 10
>>> 22:25:01 2012
>>> @@ -2751,9 +2751,9 @@
>>>   DiagGroup<"undefined-internal">;
>>>  def note_used_here : Note<"used here">;
>>> 
>>> -def warn_redefinition_of_typedef : Warning<
>>> -  "redefinition of typedef %0 is invalid in C">,
>>> -  InGroup<DiagGroup<"typedef-redefinition"> >, DefaultError;
>>> +def warn_redefinition_of_typedef : ExtWarn<
>>> +  "redefinition of typedef %0 is a C11 feature">,
>>> +  InGroup<DiagGroup<"typedef-redefinition"> >;
>>> 
>>>  def err_inline_declaration_block_scope : Error<
>>>   "inline declaration of %0 not allowed in block scope">;
>>> 
>>> Modified: cfe/trunk/lib/Sema/SemaDecl.cpp
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/Sema/SemaDecl.cpp?rev=147925&r1=147924&r2=147925&view=diff
>>> 
>>> ==============================================================================
>>> --- cfe/trunk/lib/Sema/SemaDecl.cpp (original)
>>> +++ cfe/trunk/lib/Sema/SemaDecl.cpp Tue Jan 10 22:25:01 2012
>>> @@ -1463,8 +1463,6 @@
>>> 
>>>   // The types match.  Link up the redeclaration chain if the old
>>>   // declaration was a typedef.
>>> -  // FIXME: this is a potential source of weirdness if the type
>>> -  // spellings don't match exactly.
>>>   if (TypedefNameDecl *Typedef = dyn_cast<TypedefNameDecl>(Old))
>>>     New->setPreviousDeclaration(Typedef);
>>> 
>>> @@ -1509,8 +1507,8 @@
>>>     return New->setInvalidDecl();
>>>   }
>>> 
>>> -  // Modules always permit redefinition of typedefs.
>>> -  if (getLangOptions().Modules)
>>> +  // Modules always permit redefinition of typedefs, as does C11.
>>> +  if (getLangOptions().Modules || getLangOptions().C11)
>>>     return;
>>> 
>>>   // If we have a redefinition of a typedef in C, emit a warning.  This
>>> warning
>>> 
>>> Modified: cfe/trunk/test/Preprocessor/line-directive.c
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Preprocessor/line-directive.c?rev=147925&r1=147924&r2=147925&view=diff
>>> 
>>> ==============================================================================
>>> --- cfe/trunk/test/Preprocessor/line-directive.c (original)
>>> +++ cfe/trunk/test/Preprocessor/line-directive.c Tue Jan 10 22:25:01 2012
>>> @@ -41,7 +41,7 @@
>>> 
>>>  # 192 "glomp.h" // not a system header.
>>>  typedef int x;  // expected-note {{previous definition is here}}
>>> -typedef int x;  // expected-error {{redefinition of typedef 'x' is
>>> invalid in C}}
>>> +typedef int x;  // expected-warning {{redefinition of typedef 'x' is a
>>> C11 feature}}
>>> 
>>>  # 192 "glomp.h" 3 // System header.
>>>  typedef int y;  // ok
>>> @@ -62,7 +62,7 @@
>>>  # 42 "blonk.h"  // DOES change system headerness.
>>> 
>>>  typedef int w;  // expected-note {{previous definition is here}}
>>> -typedef int w;  // expected-error {{redefinition of typedef 'w' is
>>> invalid in C}}
>>> +typedef int w;  // expected-warning {{redefinition of typedef 'w' is a
>>> C11 feature}}
>>> 
>>>  typedef int q;  // original definition in system header, should not
>>> diagnose.
>>> 
>>> 
>>> Added: cfe/trunk/test/Sema/c11-typedef-redef.c
>>> URL:
>>> http://llvm.org/viewvc/llvm-project/cfe/trunk/test/Sema/c11-typedef-redef.c?rev=147925&view=auto
>>> 
>>> ==============================================================================
>>> --- cfe/trunk/test/Sema/c11-typedef-redef.c (added)
>>> +++ cfe/trunk/test/Sema/c11-typedef-redef.c Tue Jan 10 22:25:01 2012
>>> @@ -0,0 +1,14 @@
>>> +// RUN: %clang_cc1 -std=c11 %s -verify
>>> +
>>> +typedef int type;
>>> +typedef type type;
>>> +typedef int type;
>>> +
>>> +void f(int N) {
>>> +  typedef int type2;
>>> +  typedef type type2;
>>> +  typedef int type2;
>>> +
>>> +  typedef int vla[N]; // expected-note{{previous definition is here}}
>>> +  typedef int vla[N]; // expected-error{{typedef redefinition with
>>> different types ('int [N]' vs 'int [N]')}}
>>> +}
>>> 
>> Hello Doug,
>> 
>> this last error is very weird:
>> - is the error off ?   (ie, it is acceptable to redefine the typedef for vla
>> but not handled yet)
>> - or is it simply that the diagnostic is unclear ?   (the FOO vs FOO part
>> looks strange, the types are identical as far as I can see).
> 
> It's just the diagnostic that's a bit unclear.

Yeah, I've cleaned it up in r147973.

	- Doug