Re: [Qemu-arm] [Qemu-devel] [PATCH 00/74] pc: acpi: convert DSDT to AML API and drop ASL templates support

[Shannon Zhao <zhaoshenglong@huawei.com>]

On 2015/12/21 21:00, Igor Mammedov wrote:
> On Sat, 19 Dec 2015 22:38:41 +0200
> "Michael S. Tsirkin" <mst@redhat.com> wrote:
> 
>> > On Thu, Dec 10, 2015 at 12:40:54AM +0100, Igor Mammedov wrote:
>>> > > Due to huge size, CCing only cover letter instead of individual patches.
>>> > > 
>>> > > Series consist of 2 parts the 1st part prefixed 'acpi:' adds necessary
>>> > > AML API functions and the second part converts DSDT using existing and
>>> > > new AML API.
>>> > > 
>>> > > Series does exact byte by byte conversion and passes ACPI tables
>>> > > 'make check' tests.
>>> > > The conversion first moves common for PIIX4/Q35 parts, getting rid of *.dsl
>>> > > includes and then converts PIIX4 and Q35 parts of DSDT.
>>> > > 
>>> > > Diff-stat looks nice but actual code base is reduced by ~2000LOC
>>> > > while the rest of 10000 removals is dropping precompiled AML
>>> > > templates from tree.
>>> > > 
>>> > > There are some AML parts that could be optimized/simplified and shared
>>> > > between PIIX4/Q35/ARM but doing it will break exact match with original
>>> > > tests, hence it's left out of the scope of this series.
>>> > > 
>>> > > CC: "Michael S. Tsirkin" <mst@redhat.com> (supporter:ACPI/SMBIOS)
>>> > > CC: Shannon Zhao <zhaoshenglong@huawei.com> (maintainer:ARM ACPI Subsystem)
>>> > > CC: Peter Maydell <peter.maydell@linaro.org> (maintainer:ARM)
>>> > > CC: Paolo Bonzini <pbonzini@redhat.com> (maintainer:X86)
>>> > > CC: Richard Henderson <rth@twiddle.net> (maintainer:X86)
>>> > > CC: Eduardo Habkost <ehabkost@redhat.com> (maintainer:X86)
>>> > > CC: qemu-arm@nongnu.org (open list:ARM ACPI Subsystem)
>> > 
>> > I think this is a nice series, and I think we can live with the fact
>> > make check warns in the middle.
>> > 
>> > I sent some comments I'd like to see addressed.
>> > 
>> > Main points
>> > 	- one v2 patch seems garbled
>> > 	- please don't create functions with tons of parameters
>> > 	  which just do if (a) on each of these internally.
>> > 	  instead, move common code into a small function that
>> > 	  can be reused without so many conditionals.
>> > 	- I really dislike local variables starting with a_ for
>> > 	  no good reason.
> Michael,
> 
> Could you apply already reviewed AML API patches 1-24
> and push it to master early, pls?
> 
+1

BTW, Below patches are already in master.

[PATCH 15/74] acpi: support serialized method
[PATCH 24/74] acpi: extend aml_interrupt() to support multiple irqs

> That will help to reduce collisions with other AML
> patches on list and provide more complete API for others to use.
> 

-- 
Shannon