On Mon, 8 Apr 2019 16:58:34 +0200
Alexandre Belloni <alexandre.belloni@xxxxxxxxxxx> wrote:
> On 08/04/2019 16:20:56+0200, Thomas Bogendoerfer wrote:
> > diff --git a/drivers/rtc/rtc-m48t35.c b/drivers/rtc/rtc-m48t35.c
> > index 0cf6507de3c7..05f0d91366af 100644
> > --- a/drivers/rtc/rtc-m48t35.c
> > +++ b/drivers/rtc/rtc-m48t35.c
> > @@ -24,6 +24,16 @@
> >
> > struct m48t35_rtc {
> > u8 pad[0x7ff8]; /* starts at 0x7ff8 */
> > +#ifdef CONFIG_SGI_IP27
> > + u8 hour;
> > + u8 min;
> > + u8 sec;
> > + u8 control;
> > + u8 year;
> > + u8 month;
> > + u8 date;
> > + u8 day;
> > +#else
>
> I'm not sure why the RTC driver has to know about that. Shouldn't your
> accessors be fixing that?
no, because the hardware is weird. RTC is connected to IOC3 byte bus and IOC3 is
connected to PCI. With a correct readb for PCI bus access to RTC behind IOC3 is byte
swapped.
Thomas.
--
SUSE Linux GmbH
GF: Felix Imendörffer, Mary Higgins, Sri Rasiah
HRB 21284 (AG Nürnberg)
