On Wed, Jul 29, 2015 at 8:02 AM, Leo Yan [off-list ref] wrote:

When use rtc-pl031 for suspend test on Hisilicon's SoC Hi6220, Usually
the data register (DR) will read back as value zero. So the suspend
test code will set the match register (MR) for 10 seconds' timeout; But
there have chance later will read back some random values from DR
register; So finally miss with match value and will not trigger
waken up event anymore.

This issue can be dismissed by reset registers in initialization flow;
And this code have no harm for ST's variant.

Signed-off-by: Leo Yan <redacted>

I don't understand this...

+       /* Init registers */
+       writel(0x0, ldata->base + RTC_LR);

This will reset the clock to jan 1st 1970 on every reboot.
The idea is that the RTC should *preserve* the system time
if you reboot the system, so NACK.

Usually userspace has a script using hwclock to read the
system time from the rtc to system time with hwclock -s
after userspace comes up. Likewise it writes it back with
hwclock -w before rebooting.

+       writel(0x0, ldata->base + RTC_DR);

This is a read-only register in the PL031 clean variant.
What do you want to achieve here? Is this register writeable
on the HiSilicon?

+       writel(0x0, ldata->base + RTC_IMSC);

OK

+       writel(RTC_BIT_AI, ldata->base + RTC_ICR);

So why do we want to have the alarm enabled by
default, before the kernel nor userspace has requested
it?

If your problem is with suspend/resume I suggest you work
on the [runtime]_suspend/resume hooks instead of probe().
Possibly you need to save/restore state across suspend/resume.

Yours,
Linus Walleij