linux-kernel - Re: [PATCH] [media] s5p-fimc: Only build suspend/resume for PM

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Thu, 2 Oct 2014 15:50:35 +0200
From:	Thierry Reding <thierry.reding@...il.com>
To:	Mauro Carvalho Chehab <mchehab@....samsung.com>
Cc:	linux-media@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] [media] s5p-fimc: Only build suspend/resume for PM

On Thu, Oct 02, 2014 at 10:43:21AM -0300, Mauro Carvalho Chehab wrote:
> Em Thu, 02 Oct 2014 10:48:11 +0200
> Thierry Reding <thierry.reding@...il.com> escreveu:
> 
> > From: Thierry Reding <treding@...dia.com>
> > 
> > If power management is disabled these functions become unused, so there
> > is no reason to build them. This fixes a couple of build warnings when
> > PM(_SLEEP,_RUNTIME) is not enabled.
> > 
> > Signed-off-by: Thierry Reding <treding@...dia.com>
> > ---
> >  drivers/media/platform/exynos4-is/fimc-core.c | 2 ++
> >  1 file changed, 2 insertions(+)
> > 
> > diff --git a/drivers/media/platform/exynos4-is/fimc-core.c b/drivers/media/platform/exynos4-is/fimc-core.c
> > index b70fd996d794..8e7435bfa1f9 100644
> > --- a/drivers/media/platform/exynos4-is/fimc-core.c
> > +++ b/drivers/media/platform/exynos4-is/fimc-core.c
> > @@ -832,6 +832,7 @@ err:
> >  	return -ENXIO;
> >  }
> >  
> > +#if defined(CONFIG_PM_SLEEP) || defined(CONFIG_PM_RUNTIME)
> >  static int fimc_m2m_suspend(struct fimc_dev *fimc)
> >  {
> >  	unsigned long flags;
> > @@ -870,6 +871,7 @@ static int fimc_m2m_resume(struct fimc_dev *fimc)
> >  
> >  	return 0;
> >  }
> > +#endif
> 
> The patch obviously is correct, but I'm wandering if aren't there a way
> to avoid the if/endif.
> 
> Not tested here, but perhaps if we mark those functions as inline, the
> C compiler would do the right thing without generating any warnings.
> 
> If not, maybe we could use some macro like:
> 
> #if defined(CONFIG_PM_SLEEP) || defined(CONFIG_PM_RUNTIME)
> 	#define PM_SLEEP_FUNC
> #else
> 	#define PM_SLEEP_FUNC	inline
> #endif
> 
> And put it at pm.h.
> 
> That should be enough to shut up to warning without adding any footprint
> if PM is disabled.

I think you could use __maybe_unused to that effect, but it has the
disadvantage of hiding such messages forever. For instance if the
suspend/resume code was ever to be removed, then you wouldn't get a
warning at all.

And there's a corresponding #ifdef for the fimc_runtime_{suspend,resume}
and fimc_{suspend,resume} already anyway, so I don't see much point in
trying to avoid this particular #ifdef.

Thierry

Content of type "application/pgp-signature" skipped