Re: [PATCH] of: Fix uninitialized child warning in for_each_child_of_node() if !OF

[Geert Uytterhoeven <geert-Td1EMuHUCqxL1ZNQvxDV9g@public.gmane.org>]

On Mon, Jan 27, 2014 at 5:21 PM, Rob Herring <robherring2-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
> On Mon, Jan 27, 2014 at 8:52 AM, Geert Uytterhoeven
> <geert-Td1EMuHUCqxL1ZNQvxDV9g@public.gmane.org> wrote:
>> On Mon, Jan 27, 2014 at 3:16 PM, Rob Herring <robherring2-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>>> On Sun, Jan 26, 2014 at 4:21 AM, Geert Uytterhoeven
>>> <geert-Td1EMuHUCqxL1ZNQvxDV9g@public.gmane.org> wrote:
>>>> If CONFIG_OF=n:
>>>>
>>>> drivers/i2c/i2c-mux.c: In function ‘i2c_add_mux_adapter’:
>>>> drivers/i2c/i2c-mux.c:158: warning: ‘child’ is used uninitialized in this function
>>>> drivers/leds/leds-lp55xx-common.c: In function ‘lp55xx_of_populate_pdata’:
>>>> drivers/leds/leds-lp55xx-common.c:560: warning: ‘child’ is used uninitialized in this function
>>>> drivers/leds/leds-pwm.c: In function ‘led_pwm_probe’:
>>>> drivers/leds/leds-pwm.c:89: warning: ‘child’ is used uninitialized in this function
>>>> drivers/mfd/stmpe.c: In function ‘stmpe_of_probe’:
>>>> drivers/mfd/stmpe.c:1112: warning: ‘child’ is used uninitialized in this function
>>>> drivers/mfd/tc3589x.c: In function ‘tc3589x_probe’:
>>>> drivers/mfd/tc3589x.c:324: warning: ‘child’ is used uninitialized in this function
>>>> drivers/power/charger-manager.c: In function ‘of_cm_parse_desc’:
>>>> drivers/power/charger-manager.c:1606: warning: ‘child’ is used uninitialized in this function
>>>>
>>>> Introduced by commit 00b2c76a6abbe082bb5afb89ee49ec325e9cd4d2
>>    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>>>> ("include/linux/of.h: make for_each_child_of_node() reference its args when
>>>> CONFIG_OF=n").
>>>>
>>>> Instead of dropping the "__of_use_dn(child)" again, explicitly set child
>>>> to NULL, to protect against code using child after the loop (child == NULL
>>>> is the loop termination condition if CONFIG_OF=y).
>>>>
>>>> Signed-off-by: Geert Uytterhoeven <geert-Td1EMuHUCqxL1ZNQvxDV9g@public.gmane.org>
>>>
>>> Just got fixed in mainline:
>>>
>>> commit 00b2c76a6abbe082bb5afb89ee49ec325e9cd4d2
>>
>> Nope, this is the one that introduced new warnings.
>
> Oh right, sorry. Read that too fast.
>
> Won't your change trigger warnings from -Wunused-but-set-variable in
> cases where child is not used? That would only be enabled for "W=1"
> though.
>
> I think the right fix here is all the for_each_X macros should not be
> conditional on CONFIG_OF and we should rely on the functions they call
> like of_get_next_child to do the right thing for !OF.

That's indeed an alternative fix. As long as of_get_next_child() and
friends return NULL, that'll work fine.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html