Re: don't error for duplicate identical type definitions

[Antoine Jacoutot]

 Any objections?

Considering I am the one who requested it, obviously no objection :-)
I've successfully tested this with ports that used to need ports/lang/gcc/4.9 .

So as far as I am concerned, OK for me.


 Index: c-decl.c
 ===
 RCS file: /OpenBSD/src/gnu/gcc/gcc/c-decl.c,v
 retrieving revision 1.2
 diff -u -p -r1.2 c-decl.c
 --- c-decl.c  29 Apr 2010 18:37:37 -  1.2
 +++ c-decl.c  2 May 2015 14:35:28 -
 @@ -1285,9 +1285,17 @@ diagnose_mismatched_decls (tree newdecl,
if (DECL_IN_SYSTEM_HEADER (newdecl) || DECL_IN_SYSTEM_HEADER (olddecl))
   return true;  /* Allow OLDDECL to continue in use.  */
  
 -  error (redefinition of typedef %q+D, newdecl);
 -  locate_old_decl (olddecl, error);
 -  return false;
 +  if (pedantic)
 + {
 +   pedwarn (redefinition of typedef %q+D, newdecl);
 +   if (flag_pedantic_errors)
 + {
 +   locate_old_decl (olddecl, error);
 +   return false;
 + }
 + }
 +
 +  return true;
  }
  
/* Function declarations can either be 'static' or 'extern' (no
 

-- 
Antoine

don't error for duplicate identical type definitions

[Miod Vallat]

Our current base compiler will abort with an error if it finds a typedef
being redefined, even if the new definition is no different from the one
it knows about.

Allowing this kind of type redefinition is a new C11 feature, which some
third-party software has already started using. Moreover, recent gcc
compilers will indeed accept this, even in non-c11 mode, with an
optional warning.

Consider the following testcase:
$ cat type.c
typedef unsigned char uint8_t;
typedef unsigned char char_t;

/* the following should not cause errors */
typedef unsigned char uint8_t;
typedef char_t uint8_t;

/* the following should cause errors */
typedef unsigned int uint8_t;
$

With the current in-tree compiler, it will produce these errors:

type.c:5: error: redefinition of typedef 'uint8_t'
type.c:1: error: previous declaration of 'uint8_t' was here
type.c:6: error: redefinition of typedef 'uint8_t'
type.c:5: error: previous declaration of 'uint8_t' was here
type.c:9: error: conflicting types for 'uint8_t'
type.c:6: error: previous declaration of 'uint8_t' was here

with the diff I am proposing below, it will only produce these errors:

type.c:9: error: conflicting types for 'uint8_t'
type.c:6: error: previous declaration of 'uint8_t' was here

and adding -pedantic will produce these warnings:

type.c:5: warning: redefinition of typedef 'uint8_t'
type.c:6: warning: redefinition of typedef 'uint8_t'

Any objections?

Index: c-decl.c
===
RCS file: /OpenBSD/src/gnu/gcc/gcc/c-decl.c,v
retrieving revision 1.2
diff -u -p -r1.2 c-decl.c
--- c-decl.c29 Apr 2010 18:37:37 -  1.2
+++ c-decl.c2 May 2015 14:35:28 -
@@ -1285,9 +1285,17 @@ diagnose_mismatched_decls (tree newdecl,
   if (DECL_IN_SYSTEM_HEADER (newdecl) || DECL_IN_SYSTEM_HEADER (olddecl))
return true;  /* Allow OLDDECL to continue in use.  */
 
-  error (redefinition of typedef %q+D, newdecl);
-  locate_old_decl (olddecl, error);
-  return false;
+  if (pedantic)
+   {
+ pedwarn (redefinition of typedef %q+D, newdecl);
+ if (flag_pedantic_errors)
+   {
+ locate_old_decl (olddecl, error);
+ return false;
+   }
+   }
+
+  return true;
 }
 
   /* Function declarations can either be 'static' or 'extern' (no