linux-kernel - Re: [PATCH] rtc: bd70528: fix link error

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:   Tue, 9 Jul 2019 05:25:50 +0000
From:   "Vaittinen, Matti" <Matti.Vaittinen@...rohmeurope.com>
To:     "arnd@...db.de" <arnd@...db.de>
CC:     "linux-rtc@...r.kernel.org" <linux-rtc@...r.kernel.org>,
        "a.zummo@...ertech.it" <a.zummo@...ertech.it>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "alexandre.belloni@...tlin.com" <alexandre.belloni@...tlin.com>,
        "lee.jones@...aro.org" <lee.jones@...aro.org>,
        "rdunlap@...radead.org" <rdunlap@...radead.org>
Subject: Re: [PATCH] rtc: bd70528: fix link error

Hello Arnd,

(Added Randy who also sent a patch to fix this)

On Mon, 2019-07-08 at 18:10 +0200, Arnd Bergmann wrote:
> On Mon, Jul 8, 2019 at 3:24 PM Vaittinen, Matti
> <Matti.Vaittinen@...rohmeurope.com> wrote:
> 
> > On Mon, 2019-07-08 at 14:41 +0200, Arnd Bergmann wrote:
> > > With CONFIG_BD70528_WATCHDOG=m, a built-in rtc driver cannot call
> > > into the low-level functions that are part of the watchdog
> > > module:
> > > 
> > > drivers/rtc/rtc-bd70528.o: In function `bd70528_set_time':
> > > rtc-bd70528.c:(.text+0x22c): undefined reference to
> > > `bd70528_wdt_lock'
> > > rtc-bd70528.c:(.text+0x2a8): undefined reference to
> > > `bd70528_wdt_unlock'
> > > drivers/rtc/rtc-bd70528.o: In function
> > > `bd70528_set_rtc_based_timers':
> > > rtc-bd70528.c:(.text+0x50c): undefined reference to
> > > `bd70528_wdt_set'
> > > 
> > > Add a Kconfig dependency on this driver, but still allow compile-
> > > testing
> > > without it.
> > > 
> > > Fixes: 32a4a4ebf768 ("rtc: bd70528: Initial support for ROHM
> > > bd70528
> > > RTC")
> > > Signed-off-by: Arnd Bergmann <arnd@...db.de>
> > > ---
> > >  drivers/rtc/Kconfig | 2 ++
> > >  1 file changed, 2 insertions(+)
> > > 
> > > diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig
> > > index 3bfc04a86529..7b071cc74422 100644
> > > --- a/drivers/rtc/Kconfig
> > > +++ b/drivers/rtc/Kconfig
> > > @@ -498,8 +498,10 @@ config RTC_DRV_M41T80_WDT
> > >       help
> > >         If you say Y here you will get support for the
> > >         watchdog timer in the ST M41T60 and M41T80 RTC chips
> > > series.
> > > +
> > >  config RTC_DRV_BD70528
> > >       tristate "ROHM BD70528 PMIC RTC"
> > > +     depends on BD70528_WATCHDOG || (COMPILE_TEST &&
> > > !BD70528_WATCHDOG)
> > 
> > I am not fan of this. There may well be use-cases where it is
> > desirable
> > to leave the watchdog out but still compile in the RTC. This is why
> > we
> > have static inline stubs in the header for cases where WDG is not
> > compiled in. (RTC does not need to stop WDG if WDG driver is not
> > included)
> > 
> > Adding dependency from RTC to MFD for BD70528 should be done - this
> > will avoid most of the issues (And there has been few patches sent
> > for
> > this already). But that's still not complete solution because
> > configuring RTC and MFD to be built in-kernel and WDG as a module
> > will
> > cause errors again.
> > 
> > Is there a way to force WDG in-kernel if RTC is in-kernel? (Or
> > disallow configuring RTC in-kernel if WDG is a module - while still
> > allow RTC to be built without WDG?
> 
> We could make this
> 
>         depends on BD70528_WATCHDOG || !BD70528_WATCHDOG
> 
> which would allow building with or without watchdog, even when not
> compile-testing, but still disallow the combination of
> .

Thanks for teaching me Arnd! That is clever :) We need something like

depends on MFD_ROHM_BD70528 && (BD70528_WATCHDOG || !BD70528_WATCHDOG)

(I'm not sure if parenthesis are Ok and respected in Kconfig). I would
never have thought of BD70528_WATCHDOG || !BD70528_WATCHDOG - it looks
awkward at first sight but indeed - depends on BD70528_WATCHDOG
disallows BD70528_WATCHDOG=m with RTC_DRV_BD70528=y while
!BD70528_WATCHDOG allows BD70528_WATCHDOG=n. Brilliant and exactly what
we need :) Thanks a bunch!

> 
>        Arnd