Re: arch/m68k/include/asm/raw_io.h:91:13: warning: array subscript 0 is outside array bounds of 'volatile u16[0]' {aka 'volatile short unsigned int[]'}

[Rob Herring]

On Tue, Sep 19, 2023 at 7:09 AM kernel test robot <lkp@xxxxxxxxx> wrote:
>
> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> head:   2cf0f715623872823a72e451243bbf555d10d032
> commit: f1a43aadb5a690e141a3b6700e2a40c1d4dbe088 watchdog: Enable COMPILE_TEST for more drivers
> date:   5 weeks ago
> config: m68k-allyesconfig (https://download.01.org/0day-ci/archive/20230919/202309192013.vI4DKHmw-lkp@xxxxxxxxx/config)
> compiler: m68k-linux-gcc (GCC) 13.2.0
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20230919/202309192013.vI4DKHmw-lkp@xxxxxxxxx/reproduce)
>
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@xxxxxxxxx>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202309192013.vI4DKHmw-lkp@xxxxxxxxx/
>
> All warnings (new ones prefixed by >>):
>
>    In file included from arch/m68k/include/asm/io_mm.h:25,
>                     from arch/m68k/include/asm/io.h:8,
>                     from include/linux/io.h:13,
>                     from drivers/watchdog/machzwd.c:39:
>    In function 'zf_set_timer',
>        inlined from 'zf_timer_on' at drivers/watchdog/machzwd.c:218:2:
> >> arch/m68k/include/asm/raw_io.h:91:13: warning: array subscript 0 is outside array bounds of 'volatile u16[0]' {aka 'volatile short unsigned int[]'} [-Warray-bounds=]
>       91 |         __w = ((*(__force volatile u16 *) ((_addr & 0xFFFF0000UL) + ((__v >> 8)<<1)))); \
>          |         ~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>    arch/m68k/include/asm/io_mm.h:228:20: note: in expansion of macro 'rom_out_le16'
>      228 |                  : rom_out_le16(isa_itw(port), (val)))
>          |                    ^~~~~~~~~~~~
>    arch/m68k/include/asm/io_mm.h:356:42: note: in expansion of macro 'isa_rom_outw'
>      356 | #define outw(val, port) ((port) < 1024 ? isa_rom_outw((val), (port)) : out_le16((port), (val)))
>          |                                          ^~~~~~~~~~~~
>    drivers/watchdog/machzwd.c:74:53: note: in expansion of macro 'outw'
>       74 | #define zf_writew(port, data)  { outb(port, INDEX); outw(data, DATA_W); }
>          |                                                     ^~~~
>    drivers/watchdog/machzwd.c:173:17: note: in expansion of macro 'zf_writew'
>      173 |                 zf_writew(COUNTER_1, new);
>          |                 ^~~~~~~~~
>    In function 'zf_timer_on':
>    cc1: note: source object is likely at address zero

This seems to be some newish check in gcc which looks for fixed
pointers below 4KB[1]. The linked issue says more was planned for
gcc-13, but I haven't found what that is. AFAICT, that shouldn't
happen with this config because isa_itw() should be variable and the
compiler shouldn't be able to determine the value of _addr. However, a
config with CONFIG_Q40=n, CONFIG_AMIGA_PCMCIA=n, and
CONFIG_ATARI_ROM_ISA=n would have a fixed NULL value and could trigger
the warning. This should also have warnings everywhere outw() (and
others) are used with a constant port value.

Rob

[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99578