On Fri, May 28, 2010 at 1:12 PM, Howard Hinnant <[email protected]> wrote: > On May 28, 2010, at 4:10 PM, Eli Friedman wrote: > >> On Fri, May 28, 2010 at 11:04 AM, Howard Hinnant <[email protected]> wrote: >>> Author: hhinnant >>> Date: Fri May 28 13:04:31 2010 >>> New Revision: 104982 >>> >>> URL: http://llvm.org/viewvc/llvm-project?rev=104982&view=rev >>> Log: >>> Fixed [support.types]. <cstddef> wasn't definining NULL or offsetof. >>> >>> Modified: >>> libcxx/trunk/include/cstddef >>> >>> Modified: libcxx/trunk/include/cstddef >>> URL: >>> http://llvm.org/viewvc/llvm-project/libcxx/trunk/include/cstddef?rev=104982&r1=104981&r2=104982&view=diff >>> ============================================================================== >>> --- libcxx/trunk/include/cstddef (original) >>> +++ libcxx/trunk/include/cstddef Fri May 28 13:04:31 2010 >>> @@ -34,9 +34,11 @@ >>> */ >>> >>> #include <__config> >>> +#define __need_NULL >>> #define __need_ptrdiff_t >>> #define __need_size_t >>> #include <stddef.h> >>> +#include <stddef.h> >> >> Umm, what? > > <chuckle> I know, I know. On Mac OS the first trip through <stddef.h> > doesn't define offsetof. Or said differently, offsetof doesn't get defined > if any of __need_* are defined. The first trip creates the desired macros > and #undefines the request. The second trip creates offsetof. <shrug>
Why not get rid of the __need_ defines and just include stddef.h once? -Eli _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
