http://gcc.gnu.org/bugzilla/show_bug.cgi?id=55768
Bug #: 55768
Summary: Use of unknown __attribute__ in function definition is
flagged with an error
Classification: Unclassified
Product: gcc
Version: 4.8.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassig...@gcc.gnu.org
ReportedBy: dnovi...@gcc.gnu.org
CC: ja...@gcc.gnu.org
I am not sure whether this is a bug, but it is a behaviour change since 4.7.
The compiler usually rejects attributes applied to function definitions:
int foo(int x, int y) __attribute__((const))
{
return x * y;
}
$ gcc47 -c a.cc
a.cc:1:23: error: attributes are not allowed on a function-definition
$
But if the compiler does not recognize the attribute, then it says nothing:
int foo(int x, int y) __attribute__((no_thread_safety_analysis))
{
return x * y;
}
$ gcc47 -c a.cc
$
This has changed in 4.8. We now emit an error even if we do not understand the
attribute:
a.cc:1:23: error: attributes are not allowed on a function-definition
int foo(int x, int y) __attribute__((no_thread_safety_analysis))
^
Is this a case where the user code should be changed to accomodate the rule of
not allowing attributes on function definitions?
In fact, shouldn't the compiler just allow attributes on function definitions?
