[Bug middle-end/48124] [4.3/4.4/4.5/4.6/4.7 Regression] likely wrong code bug

["rguenth at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48124

--- Comment #3 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-15 10:21:04 UTC ---
Hm, e has alignment of 8 bytes and f of 4 but the assembler doesn't pad
e's size to 8 byte multiples appearantly

        .data
        .align 8
        .type   e, @object
        .size   e, 12
e:
        .byte   0
        .byte   0
        .byte   0
        .byte   0
        .value  0
        .byte   0
        .byte   0
        .byte   1
        .byte   0
        .zero   2
        .align 4
        .type   f, @object
        .size   f, 4
f:
        .long   1

this is probably a more general problem also with the vectorizer which
happily increases alignment of global variables independed on if their
size is a multiple of that alignment.

If the asm output machinery does not try to handle padding to multiples
of alignment size then that's the bug I guess.