gnulib:test-stdalign failing on powerpc-darwin8, gcc-4.0

[David Fang]

Hi,
	I initially reported a this as a GNUTLS-3.3.9 test failure, but 
was then redirected to the gnulib project, the origin of this test.

http://savannah.gnu.org/support/?108705

I'm getting 1 test failure on powerpc-darwin8 with test-stdalign.
Compiled with the system compiler, Apple's gcc-4.0.

% ./test-stdalign
test-stdalign.c:88: assertion '(uintptr_t) &(static_char_alignas) % 
TEST_ALIGNMENT == 0' failed
Abort

I locally changed the CHECK_ALIGNED macro to print, and I see the 
following:

#define CHECK_ALIGNED(var) printf( #var " %d\n", (uintptr_t) &(var) % 
TEST_ALIGNMENT )

static_char_alignas 5
static_char_Alignas 4
static_short_alignas 2
static_short_Alignas 0
static_int_alignas 4
static_int_Alignas 0
static_long_alignas 4
static_long_Alignas 0
static_int64_t_alignas 0
static_int64_t_Alignas 0
static_float_alignas 4
static_float_Alignas 0
static_double_alignas 0
static_double_Alignas 0
static_struct1_alignas 5
static_struct1_Alignas 4
static_struct2_alignas 0
static_struct2_Alignas 6
static_struct3_alignas 0
static_struct3_Alignas 4
static_struct4_alignas 0
static_struct4_Alignas 4

powerpc-darwin is known to have some unconventional data structure 
alignment rules in its ABI.  How should these results be interpreted?
Is there a problem in the compiler, or rather some incorrect assumption 
about the ABI in the test itself?  This is the system compiler, after all.

Thanks for any insights.

Fang


--
David Fang
http://www.csl.cornell.edu/~fang/