From: Andre Przywara <andre.przywara@arm.com>
To: Arnaud Ferraris <arnaud.ferraris@collabora.com>
Cc: u-boot@lists.denx.de, Jagan Teki <jagan@amarulasolutions.com>,
Samuel Holland <samuel@sholland.org>,
Maxime Ripard <mripard@kernel.org>
Subject: Re: [PATCH 1/2] board: sunxi: enable status LED early
Date: Tue, 7 Sep 2021 23:42:16 +0100 [thread overview]
Message-ID: <20210907234216.2446b437@slackpad.fritz.box> (raw)
In-Reply-To: <e529fcca-cfd5-9d13-059e-88a067d30a42@collabora.com>
On Tue, 7 Sep 2021 13:41:11 +0200
Arnaud Ferraris <arnaud.ferraris@collabora.com> wrote:
Hi Arnaud,
> Thanks for your feedback!
>
> Le 07/09/2021 à 01:46, Andre Przywara a écrit :
> > On Mon, 6 Sep 2021 22:57:52 +0200
> > Arnaud Ferraris <arnaud.ferraris@collabora.com> wrote:
> >
> > Hi Arnaud,
> >
> >> diff --git a/board/sunxi/board.c b/board/sunxi/board.c
> >> index 1a46100e40..6e0bf5fbf9 100644
> >> --- a/board/sunxi/board.c
> >> +++ b/board/sunxi/board.c
> >> @@ -46,6 +46,9 @@
> >> #include <spl.h>
> >> #include <sy8106a.h>
> >> #include <asm/setup.h>
> >> +#if defined(CONFIG_LED_STATUS) && defined(CONFIG_SPL_DRIVERS_MISC)
> >> +#include <status_led.h>
> >
> > status_led.h already guards for CONFIG_LED_STATUS, so you can just
> > unconditionally include this here, no #ifdefs needed.
>
> Understood, will do.
>
> >
> >> +#endif
> >>
> >> #if defined CONFIG_VIDEO_LCD_PANEL_I2C && !(defined CONFIG_SPL_BUILD)
> >> /* So that we can use pin names in Kconfig and sunxi_name_to_gpio() */
> >> @@ -672,6 +675,10 @@ void sunxi_board_init(void)
> >> {
> >> int power_failed = 0;
> >>
> >> +#if defined(CONFIG_LED_STATUS) && defined(CONFIG_SPL_DRIVERS_MISC)
> >
> > Unfortunately status_led.h doesn't define a dummy prototype for this,
> > so we need the #ifdef CONFIG_LED_STATUS. But I don't think you need
> > CONFIG_SPL_DRIVERS_MISC here. If that should only trigger in the SPL, use:
> >
> > if (IS_ENABLED(CONFIG_SPL_BUILD))
> > status_led_init();
> >
>
> Actually the status_led driver isn't compiled into SPL unless
> CONFIG_SPL_DRIVERS_MISC is set, resulting in a link error if that's not
> the case. We should therefore check for both CONFIG_LED_STATUS and
> CONFIG_SPL_DRIVERS_MISC to prevent build errors both at compile time and
> link time.
>
> The alternative would be:
>
> #ifdef CONFIG_LED_STATUS
> if (IS_ENABLED(CONFIG_SPL_DRIVERS_MISC))
> status_led_init();
> #endif
Right, the nasty bit here is that you could just have LED_STATUS for
U-Boot proper, but there is no separate CONFIG_SPL_LED_STATUS, so we
need both symbols, although it looks weird.
> Which is more or less the same, except that it relies on the compiler
> optimizing out this function call if CONFIG_SPL_DRIVERS_MISC isn't
> defined. That assumption feels less safe to me, but overall I'm fine
> with both implementations.
Yes, but I still prefer to have as little #ifdef as possible,
because ...
> Please also note the whole sunxi_board_init() function itself is already
> inside a #ifdef CONFIG_SPL_BUILD block.
You are right, I missed that. That whole file is an #ifdef
nightmare, and especially nested #ifdefs are the worst. So I would like
to not *add* more of them, hence asking for IS_ENABLED() wherever
possible.
Thanks,
Andre
> Cheers,
> Arnaud
>
> > Cheers,
> > Andre
> >
> >> + status_led_init();
> >> +#endif
> >> +
> >> #ifdef CONFIG_SY8106A_POWER
> >> power_failed = sy8106a_set_vout1(CONFIG_SY8106A_VOUT1_VOLT);
> >> #endif
next prev parent reply other threads:[~2021-09-07 22:42 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-06 20:57 [PATCH] PinePhone: enable LED on boot for improved visual feedback Arnaud Ferraris
2021-09-06 20:57 ` [PATCH 1/2] board: sunxi: enable status LED early Arnaud Ferraris
2021-09-06 23:46 ` Andre Przywara
2021-09-07 11:41 ` Arnaud Ferraris
2021-09-07 22:42 ` Andre Przywara [this message]
2021-09-06 20:57 ` [PATCH 2/2] pinephone_defconfig: add support for early-boot status LED Arnaud Ferraris
2021-09-06 23:46 ` Andre Przywara
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210907234216.2446b437@slackpad.fritz.box \
--to=andre.przywara@arm.com \
--cc=arnaud.ferraris@collabora.com \
--cc=jagan@amarulasolutions.com \
--cc=mripard@kernel.org \
--cc=samuel@sholland.org \
--cc=u-boot@lists.denx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox