Re: [PATCH v6 00/10] acpi, clocksource: add GTDT driver and GTDT support in arm_arch_timer

[Hanjun Guo]

On 2016/6/30 21:27, Rafael J. Wysocki wrote:
> On Thursday, June 30, 2016 10:10:02 AM Hanjun Guo wrote:
> > Hi Rafael,
> >
> > On 2016/6/30 9:37, Rafael J. Wysocki wrote:
> > > On Thursday, June 30, 2016 09:29:59 AM Fu Wei wrote:
> > > > Hi Rafael,
> > > >
> > > > On 30 June 2016 at 05:32, Rafael J. Wysocki <rafael@xxxxxxxxxx> wrote:
> > > > > On Wed, Jun 29, 2016 at 8:15 PM,  <fu.wei@xxxxxxxxxx> wrote:
> > > > > > From: Fu Wei <fu.wei@xxxxxxxxxx>
> > > > > >
> > > > > > This patchset:
> > > > > >     (1)Preparation for adding GTDT support in arm_arch_timer
> > > > > >         1. Move some enums and marcos to header file
> > > > > >         2. Add a new enum for spi type.
> > > > > >         3. Improve printk relevant code
> > > > > >
> > > > > >     (2)Introduce ACPI GTDT parser: drivers/acpi/gtdt.c
> > > > > >     Parse all kinds of timer in GTDT table of ACPI:arch timer,
> > > > > >     memory-mapped timer and SBSA Generic Watchdog timer.
> > > > > >     This driver can help to simplify all the relevant timer drivers,
> > > > > >     and separate all the ACPI GTDT knowledge from them.
> > > > > >
> > > > > >     (3)Simplify ACPI code for arm_arch_timer
> > > > > >
> > > > > >     (4)Add GTDT support for ARM memory-mapped timer
> > > > >
> > > > > GTDT is ARM-specific AFAICS.
> > > >
> > > > yes, you are right, it is.
> > > >
> > > > > If so, why do we need that code to reside in drivers/acpi/ ?
> > > >
> > > > Although  the GTDT is just for ARM64, but this driver is parsing one
> > > > of ACPI table,
> > > > I think that could be treated as ACPI driver.  Do I miss something? :-)
> > >
> > > Yes, you are.  Nobody except for ARM64 will ever need it.
> >
> > GTDT is part of ACPI spec, drivers/acpi/ is for driver code of
> > ACPI spec, I think it can stay in drivers/acpi/ from this point
> > of view, am I right?
>
> The question is not "Can it?", but "Does it need to?".
>
> It is in the spec, but still there's only one architecture needing it.
>
> There is no way to test it on any other architecture and no reason to build it
> for any other architecture, so why does it need to be located in drivers/acpi/ ?

I'm fine to move it to other places such as arch/arm64/kernel/, but I
would like to ask ARM64 maintainer's suggestion for this.

Will, Catalin, what's your opinion on this?

Thanks
Hanjun