[PATCH v10 00/21] Introduce ACPI for ARM64 based on ACPI 5.1

[Lorenzo Pieralisi]

On Tue, Mar 24, 2015 at 10:02:53PM +0000, Grant Likely wrote:
> On Thu, 19 Mar 2015 19:39:27 +0000
> , Will Deacon <will.deacon at arm.com>
>  wrote:
> > On Thu, Mar 19, 2015 at 10:17:27AM +0000, Lorenzo Pieralisi wrote:
> > > On Thu, Mar 19, 2015 at 04:09:33AM +0000, Hanjun Guo wrote:
> > > > On 2015/3/19 3:05, Will Deacon wrote:
> > > > > On Wed, Mar 11, 2015 at 12:39:26PM +0000, Hanjun Guo wrote:
> > > > >> This patch set already tested on multi platforms:
> > > > >>  - AMD Seattle board;
> > > > >>  - Cavium Thunder board;
> > > > >>  - Huawei D02 board;
> > > > >>  - Qualcomm ARM64 platform
> > > > >>
> > > > >> This version 10 patch set address some minor comments and collect ACKs and
> > > > >> Reviewed-bys for v9:
> > > > >>
> > > > >>  - new Acks from Rafael, Olof, Grant, Lorenzo
> > > > >>  - new way to handle typdef phys_cpuid_t which suggested by Rafael,
> > > > >>    but no functional change
> > > > >>  - Remove if(!phys) for early ioremappings
> > > > >>  - Rework sleep function for ARM64
> > > > >>  - Introduce linux/acpi_irq.h to hold acpi_irq_init()
> > > > >>  - Disable ACPI if not HW_REDUCED_ACPI compliant
> > > > >>  - Remove the doc of why ACPI on ARM
> > > > > So I've had a look at the current state of this series and I think there
> > > > > are a few immediate things left to do:
> > > > >
> > > > >   (1) Resolve the acpi=force cmdline issue highlighted by Lorenzo and
> > > > >       Catalin
> > > > 
> > > > Sure, it will be done after the confirmation with Ard.
> > > > 
> > > > >
> > > > >   (2) I believe Sudeep and Lorenzo have concerns about patch 13 (SMP init),
> > > > >       so I'm assuming there will be additional patches from them that are
> > > > >       required.
> > > > 
> > > > Sorry, I assume that it is about the print information for PSCI absent for SMP init, right?
> > > 
> > > Not only that, Sudeep has a patch to consolidate DT and ACPI SMP code,
> > > I am working on it, I do not think it should be a blocking point, patch
> > > coming asap on top of your series.
> > 
> > Well, I don't really want to merge the series without those patches so I
> > do think it blocks the code from getting into mainline.
> 
> Really? It's a pretty minor duplication problem and it's been identified
> as something requiring refactoring to both the ACPI and DT code. It
> isn't at all dangerous. Why is this a blocking point?

The SMP init ACPI/DT consolidation, in particular in relation to cpu_ops may
not be a blocking point, but it is not a whim either and it deserves some
thought.

I will post a patch asap and the ACPI parking protocol support patches
strictly depend on this clean-up to be completed.

Lorenzo