linux-kernel - Re: [PATCH 2/2] pm: allow drivers to drop #ifdef and __maybe

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

Date:   Fri, 9 Apr 2021 12:17:04 +0900
From:   Masahiro Yamada <masahiroy@...nel.org>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Linux Kbuild mailing list <linux-kbuild@...r.kernel.org>,
        "open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
        Paul Cercueil <paul@...pouillou.net>,
        Len Brown <len.brown@...el.com>, Pavel Machek <pavel@....cz>,
        "Rafael J. Wysocki" <rjw@...ysocki.net>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Linux PM list <linux-pm@...r.kernel.org>
Subject: Re: [PATCH 2/2] pm: allow drivers to drop #ifdef and __maybe_unused
 from pm callbacks

On Fri, Apr 9, 2021 at 6:30 AM Arnd Bergmann <arnd@...db.de> wrote:
>
> On Thu, Apr 8, 2021 at 11:00 PM Masahiro Yamada <masahiroy@...nel.org> wrote:
> >
> > Drivers typically surround suspend and resume callbacks with #ifdef
> > CONFIG_PM(_SLEEP) or mark them as __maybe_unused in order to avoid
> > -Wunused-const-variable warnings.
> >
> > With this commit, drivers will be able to remove #ifdef CONFIG_PM(_SLEEP)
> > and __maybe_unsed because unused functions are dropped by the compiler
> > instead of the preprocessor.
> >
> > Signed-off-by: Masahiro Yamada <masahiroy@...nel.org>
>
> I tried this before and could not get it to work right.
>
> >
> > -#ifdef CONFIG_PM_SLEEP
> > +#define pm_ptr(_ptr)           PTR_IF(IS_ENABLED(CONFIG_PM), _ptr)
> > +#define pm_sleep_ptr(_ptr)     PTR_IF(IS_ENABLED(CONFIG_PM_SLEEP), _ptr)
> > +
> >  #define SET_SYSTEM_SLEEP_PM_OPS(suspend_fn, resume_fn) \
> > -       .suspend = suspend_fn, \
> > -       .resume = resume_fn, \
> > -       .freeze = suspend_fn, \
> > -       .thaw = resume_fn, \
> > -       .poweroff = suspend_fn, \
> > -       .restore = resume_fn,
> > -#else
> > -#define SET_SYSTEM_SLEEP_PM_OPS(suspend_fn, resume_fn)
> > -#endif
> > +       .suspend  = pm_sleep_ptr(suspend_fn), \
> > +       .resume   = pm_sleep_ptr(resume_fn), \
> > +       .freeze   = pm_sleep_ptr(suspend_fn), \
> > +       .thaw     = pm_sleep_ptr(resume_fn), \
> > +       .poweroff = pm_sleep_ptr(suspend_fn), \
> > +       .restore  = pm_sleep_ptr(resume_fn),
>
> The problem that I think you inevitably hit is that you run into a missing
> declaration for any driver that still uses an #ifdef around a static
> function.
>
> The only way I can see us doing this is to create a new set of
> macros that behave like the version you propose here but leave
> the old macros in place until the last such #ifdef has been removed.
>
>        Arnd

Agh, you are right.
We cannot change all the callsites atomically due to the huge amount of users.

-- 
Best Regards
Masahiro Yamada