Re: [PATCH 04/12] i3c: master: Support ACPI enumeration

From: Akhil R <akhilrajeev@nvidia.com>
To: <thierry.reding@kernel.org>
Cc: <Frank.Li@nxp.com>, <acpica-devel@lists.linux.dev>,
	<akhilrajeev@nvidia.com>, <alexandre.belloni@bootlin.com>,
	<conor+dt@kernel.org>, <devicetree@vger.kernel.org>,
	<ebiggers@kernel.org>, <fredrik.markstrom@est.tech>,
	<jonathanh@nvidia.com>, <krzk@kernel.org>, <lenb@kernel.org>,
	<linux-acpi@vger.kernel.org>, <linux-hwmon@vger.kernel.org>,
	<linux-i3c@lists.infradead.org>, <linux-kernel@vger.kernel.org>,
	<linux-tegra@vger.kernel.org>, <linux@roeck-us.net>,
	<lkp@intel.com>, <llvm@lists.linux.dev>,
	<miquel.raynal@bootlin.com>, <oe-kbuild-all@lists.linux.dev>,
	<p.zabel@pengutronix.de>, <rafael@kernel.org>,
	<robert.moore@intel.com>, <robh@kernel.org>
Subject: Re: [PATCH 04/12] i3c: master: Support ACPI enumeration
Date: Tue, 31 Mar 2026 15:39:04 +0530	[thread overview]
Message-ID: <20260331100904.30932-1-akhilrajeev@nvidia.com> (raw)
In-Reply-To: <acO_PUjS_VG07qcS@orome>

On Wed, 25 Mar 2026 11:59:12 +0100, Thierry Reding wrote:
> On Tue, Mar 24, 2026 at 10:52:15PM +0530, Akhil R wrote:
>> On Tue, 24 Mar 2026 09:43:27 +0100, Alexandre Belloni wrote:
>> 
>> ...
>> 
>> >> #include <linux/acpi.h> is added in PATCH 03/12. The functions' prototypes
>> >> are present in acpi.h. I think the bot checked this patch individually,
>> >> or did I miss something?
>> >> 
>> > 
>> > #include <acpi/acpi_bus.h> is behind an #ifdef in acpi.h and your code
>> > is not.
>> 
>> Thanks for pointing Alexandre and Guenter. I also noticed that we do not
>> have stub functions for a few of the acpi_* functions in #else.
>> 
>> Looks like I will have to guard calls to these functions under
>> #ifdef CONFIG_ACPI.
> 
> Alternatively it might make sense to add the stubs in a separate patch.
> I don't know if they were purposefully left out or nobody's ever run
> into the lack of these before.

I looked into this and it turns out to be more involved than expected,
and requires adding stubs at multiple layers. We may end up either
sneaking in only what we require or a non-trivial change which may
involve too many parameters.

If you would agree, I will guard the ACPI calls used in this patchset with
#ifdef CONFIG_ACPI to keep things self-contained. If you think adding the
stubs is worthwhile, we can take them up as a separate series.

Best Regards,
Akhil

-- 
linux-i3c mailing list
linux-i3c@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-i3c