[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAHp75VdrEGDixMjPS3kXHqgcLNOSJ3eqYAfz9dX1kkU8NxF0mQ@mail.gmail.com>
Date: Mon, 9 Oct 2017 02:10:15 +0300
From: Andy Shevchenko <andy.shevchenko@...il.com>
To: Mario Limonciello <Mario.Limonciello@...l.com>
Cc: "dvhart@...radead.org" <dvhart@...radead.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Platform Driver <platform-driver-x86@...r.kernel.org>,
Andy Lutomirski <luto@...nel.org>, quasisec@...gle.com,
Pali Rohár <pali.rohar@...il.com>,
"Rafael J. Wysocki" <rjw@...ysocki.net>, mjg59@...gle.com,
Christoph Hellwig <hch@....de>, Greg KH <greg@...ah.com>
Subject: Re: [PATCH v5 09/14] platform/x86: dell-smbios: Introduce dispatcher
for SMM calls
On Mon, Oct 9, 2017 at 12:45 AM, <Mario.Limonciello@...l.com> wrote:
>> >> config DELL_SMBIOS
>> >> tristate
>> >
>> >> + depends on DELL_SMBIOS_SMM
>> >> +
>> >
>> > If the above dependency is true (which I'm quite sure not) this split
>> > should never happen.
>>
>> Hmm... One more thought here.
>>
>> Which is library and which is not? IOW the question is "can
>> DELL_SMBIOS be a standalone working module"?
>>
>
> It can technically be compiled as a standalone module, but it won't do
> anything as a standalone module as of this patch.
>
> After the sysfs tokens patch it could be used to provide token information
> without the SMM or WMI drivers.
> (although this interface will have to change based upon Greg's most recent
> feedback).
>
> In terms of transitioning people moving from older kernels to new, it makes
> most sense to me that it's automatically selecting DELL_SMBIOS_SMM if you
> had it selected previously, but I'll defer to you judgement if the dependency
> should be dropped since it can be a standalone module that just doesn't do
> anything.
Okay, if we can call it "a library" then remove the reverse dependency
as I mentioned in first comment.
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists