[PATCH 3/5] irqchip: armada-370-xp: re-enable per-CPU interrupts at resume time

[Marcin Wojtas]

Thomas,

2015-10-26 6:09 GMT+01:00 Thomas Petazzoni
<thomas.petazzoni at free-electrons.com>:
> Marcin,
>
> On Mon, 26 Oct 2015 05:35:46 +0100, Marcin Wojtas wrote:
>
>> Thanks for the explanation - now it's clear.
>
> Good :-) Hopefully the explanation in PATCH 5/5 is also clear enough.

The Ascii-art is beutiful, indeed:)

>
>> Btw, I checked the patches with mvneta in both 'standby' and 'mem'
>> modes on A38x (with not-yet-submitted support for PM in mvneta and
>> pinctrl) and everything works properly. Hence:
>
> Thanks for the testing. However, I wonder why you think those changes
> are need to get mvneta to work fine with the 'standby' mode ? While I
> do agree that they are need for the 'mem' mode, they shouldn't be
> needed for the 'standby' mode. For now, the standby mode only puts the
> CPU into deep-idle, and that's all: all devices remain powered on, and
> they don't lose their state.
>

Yes, you are right - without any pm_ops the driver works well after
suspend/resume in standby. However in the linux mem and standby is
treated exactly the same as pm sleep, so the same routines are
executed in both modes. Hence the s2ram support cannot spoil standby.

Best regards,
Marcin