Re: [PATCH v3 28/38] rtc: add HAS_IOPORT dependencies

[Arnd Bergmann]

On Mon, May 8, 2023, at 17:36, Niklas Schnelle wrote:
> On Tue, 2023-03-14 at 13:52 +0100, Alexandre Belloni wrote:
>> Hello,
>> 
>> On 14/03/2023 13:12:06+0100, Niklas Schnelle wrote:
>> > In a future patch HAS_IOPORT=n will result in inb()/outb() and friends
>> > not being declared. We thus need to add HAS_IOPORT as dependency for
>> > those drivers using them.
>> > 
>> > Co-developed-by: Arnd Bergmann <arnd@xxxxxxxxxx>
>> > Signed-off-by: Niklas Schnelle <schnelle@xxxxxxxxxxxxx>
>> > ---
>> >  drivers/rtc/Kconfig | 4 +++-
>> >  1 file changed, 3 insertions(+), 1 deletion(-)
>> > 
>> > diff --git a/drivers/rtc/Kconfig b/drivers/rtc/Kconfig
>> > index 5a71579af0a1..20aa77bf0a9f 100644
>> > --- a/drivers/rtc/Kconfig
>> > +++ b/drivers/rtc/Kconfig
>> > @@ -956,6 +956,7 @@ comment "Platform RTC drivers"
>> >  config RTC_DRV_CMOS
>> >  	tristate "PC-style 'CMOS'"
>> >  	depends on X86 || ARM || PPC || MIPS || SPARC64
>> > +	depends on HAS_IOPORT
>> 
>> Did you check that this will not break platforms that doesn't have RTC_PORT defined?
>> > 
>
> From what I can tell the CMOS_READ() macro this driver relies on uses
> some form of inb() style I/O port access in all its definitions. So my
> understanding is that this device is always accessed via I/O ports even
> if the variants differ slightly and would make no sense on a platform
> without any way of accessing I/O ports which is what lack of HAS_IOPORT
> means. From what I can see even without RTC_PORT being defined the
> CMOS_READ is still used. Hope that answers your question?

I think the m68k/atari and mips/dec variants don't necessarily
qualify as PIO, those are really just pointer dereferences, and
they don't use the actual inb/outb functions.

On atari, it looks like HAS_IOPORT may be set if ATARI_ROM_ISA
is, but on dec it's never enabled.

    Arnd