Re: rtc regression [463a86304cae ("char/genrtc: x86: remove remnants of asm/rtc.h")]

From: Ville Syrjälä
Date: Tue Aug 09 2016 - 12:55:21 EST


On Tue, Aug 09, 2016 at 06:09:27PM +0200, Arnd Bergmann wrote:
> On Tuesday, August 9, 2016 6:56:49 PM CEST Ville Syrjälä wrote:
> > Hi,
> >
> > commit 463a86304cae ("char/genrtc: x86: remove remnants of asm/rtc.h")
> > broke rtc for me. Neither hwclock or rtcwake work anymore. This is just
> > a very standard x86-64 IVB box, and it was reported that machines in
> > our i915 test farm are having rtc related problems as well.
> >
> > The first time I run hwclock after rebooting I get this:
> > open("/dev/rtc", O_RDONLY) = 3
> > ioctl(3, PHN_SET_REGS or RTC_UIE_ON, 0) = 0
> > select(4, [3], NULL, NULL, {10, 0}) = 0 (Timeout)
> > ioctl(3, PHN_NOT_OH or RTC_UIE_OFF, 0) = 0
> > close(3) = 0
> >
> > On all subsequent runs I get this:
> > open("/dev/rtc", O_RDONLY) = 3
> > ioctl(3, PHN_SET_REGS or RTC_UIE_ON, 0) = -1 EINVAL (Invalid argument)
> > ioctl(3, RTC_RD_TIME, 0x7ffd76b3ae70) = -1 EINVAL (Invalid argument)
> > close(3) = 0
> >
> > 463a86304cae^ gets me back to working condition:
> > open("/dev/rtc", O_RDONLY) = 3
> > ioctl(3, PHN_SET_REGS or RTC_UIE_ON, 0) = 0
> > select(4, [3], NULL, NULL, {10, 0}) = 1 (in [3], left {9, 530998})
> > ioctl(3, PHN_NOT_OH or RTC_UIE_OFF, 0) = 0
> > ioctl(3, RTC_RD_TIME, {tm_sec=20, tm_min=17, tm_hour=15, tm_mday=9, tm_mon=7, tm_year=116, ...}) = 0
> > close(3) = 0
>
> That commit was supposed to have no effect on the binary, but it
> seems I had a typo in one file. Can you try this patch on top
> of mainline?
>
> Arnd
>
> diff --git a/arch/x86/kernel/hpet.c b/arch/x86/kernel/hpet.c
> index ed16e58658a4..c6dfd801df97 100644
> --- a/arch/x86/kernel/hpet.c
> +++ b/arch/x86/kernel/hpet.c
> @@ -1242,7 +1242,7 @@ irqreturn_t hpet_rtc_interrupt(int irq, void *dev_id)
> memset(&curr_time, 0, sizeof(struct rtc_time));
>
> if (hpet_rtc_flags & (RTC_UIE | RTC_AIE))
> - mc146818_set_time(&curr_time);
> + mc146818_get_time(&curr_time);

Heh. Pretty obvious once you see it :)

Works for me.
Tested-by: Ville Syrjälä <ville.syrjala@xxxxxxxxxxxxxxx>

>
> if (hpet_rtc_flags & RTC_UIE &&
> curr_time.tm_sec != hpet_prev_update_sec) {

--
Ville Syrjälä
Intel OTC