[RFC PATCH 0/3] Introduce pm_ptr() / pm_sleep_ptr()

From: Paul Cercueil <paul@crapouillou.net>
To: "Rafael J . Wysocki" <rjw@rjwysocki.net>,
	Len Brown <len.brown@intel.com>, Pavel Machek <pavel@ucw.cz>
Cc: Ulf Hansson <ulf.hansson@linaro.org>,
	od@zcrc.me, linux-pm@vger.kernel.org, linux-mmc@vger.kernel.org,
	linux-kernel@vger.kernel.org,
	Paul Cercueil <paul@crapouillou.net>
Subject: [RFC PATCH 0/3] Introduce pm_ptr() / pm_sleep_ptr()
Date: Tue, 11 Feb 2020 13:03:18 -0300	[thread overview]
Message-ID: <20200211160321.22124-1-paul@crapouillou.net> (raw)

Hi,

I've seen many times things like:

#ifdef CONFIG_PM_SLEEP
static SIMPLE_DEV_PM_OPS(foo_pm_ops, foo_suspend, foo_resume);
#define FOO_PM_OPS (&foo_pm_ops)
#else
#define FOO_PM_OPS NULL
#endif
static struct platform_driver foo_driver = {
		.driver.pm = FOO_PM_OPS,
};

And always wondered why there was no of-match-ptr-like macro to make
things cleaner.

So this RFC adds two macros, pm_ptr() and pm_sleep_ptr(), which resolve
to their argument when CONFIG_PM or CONFIG_PM_SLEEP (respectively) are
enabled, or NULL otherwise.

Patch 3/3 is an example of what it would look like when used in a
driver.

Comments welcome.

Cheers,
-Paul

Paul Cercueil (3):
  PM: introduce pm_ptr() and pm_sleep_ptr()
  PM: Make *_DEV_PM_OPS macros use __maybe_unused
  mmc: jz4740: Use pm_sleep_ptr() macro

 drivers/mmc/host/jz4740_mmc.c | 12 +++---------
 include/linux/pm.h            | 16 ++++++++++++++--
 2 files changed, 17 insertions(+), 11 deletions(-)

-- 
2.25.0