RE: [PATCH v2 1/3] ACPI: Resolve objects on host-directed table loads

[Nikolaus Voss <nv@vosn.de>]

Bob,

On Fri, 14 Jun 2019, Moore, Robert wrote:
>
>
> -----Original Message-----
> From: Nikolaus Voss [mailto:nv@vosn.de]
> Sent: Friday, June 14, 2019 2:26 AM
> To: Rafael J. Wysocki <rafael@kernel.org>
> Cc: Rafael J. Wysocki <rjw@rjwysocki.net>; Len Brown <lenb@kernel.org>; Moore, Robert <robert.moore@intel.com>; Schmauss, Erik <erik.schmauss@intel.com>; Jacek Anaszewski <jacek.anaszewski@gmail.com>; Pavel Machek <pavel@ucw.cz>; Dan Murphy <dmurphy@ti.com>; Thierry Reding <thierry.reding@gmail.com>; ACPI Devel Maling List <linux-acpi@vger.kernel.org>; open list:ACPI COMPONENT ARCHITECTURE (ACPICA) <devel@acpica.org>; linux-leds@vger.kernel.org; Linux PWM List <linux-pwm@vger.kernel.org>; Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
> Subject: Re: [PATCH v2 1/3] ACPI: Resolve objects on host-directed table loads
>
> Hi Rafael,
>
> On Fri, 14 Jun 2019, Rafael J. Wysocki wrote:
>> On Wed, Jun 12, 2019 at 10:36 AM Nikolaus Voss
>> <nikolaus.voss@loewensteinmedical.de> wrote:
>>>
>>> If an ACPI SSDT overlay is loaded after built-in tables have been
>>> loaded e.g. via configfs or efivar_ssdt_load() it is necessary to
>>> rewalk the namespace to resolve references. Without this, relative
>>> and absolute paths like ^PCI0.SBUS or \_SB.PCI0.SBUS are not resolved
>>> correctly.
>>>
>>> Make configfs load use the same method as efivar_ssdt_load().
>>>
>>> Signed-off-by: Nikolaus Voss <nikolaus.voss@loewensteinmedical.de>
>>
>> This is fine by me, so
>>
>> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
>>
>> Or if you want me to take this patch (without the other two in the
>> series), please let me know.
>
> thanks. I think it would be the best if you take up this patch as it is 
> an independent topic. In retrospect it wasn't a good idea to put it into 
> this series.
>
> Kind regards,
> Niko
>
> I would have to ask, why is additional code needed for package 
> initialization/resolution? It already happens elsewhere in acpica. Bob

for built-in tables loaded via acpi_ex_load_table_op() everything is fine, 
because after acpi_tb_load_table() acpi_ns_walk_namespace() is called to 
resolve references.

My fix only affects tables loaded dynamically via either 
acpi_configfs driver (for debugging purposes) or efivar_ssdt_load() (to 
specify a table on the kernel's command line). They use acpi_load_table() 
to load the table from a caller-owned buffer. To resolve the references, 
it is again necessary to rewalk the namespace, which was simply missing in 
acpi_load_table().

Niko