Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230509084007.24373-1-lukas.bulwahn@gmail.com>
 <7cbe80f0-dcc3-bfc3-3bd8-c9f96704618d@intel.com> <CAKXUXMxO4b-8wT-88fJ=++woH-p79_4R-vUr=aELNFYinGcEuw@mail.gmail.com>
 <ea00d762-b0df-ab72-988b-98c62ffc7949@intel.com>
In-Reply-To: <ea00d762-b0df-ab72-988b-98c62ffc7949@intel.com>
From:   Lukas Bulwahn <lukas.bulwahn@gmail.com>
Date:   Wed, 10 May 2023 09:46:40 +0200
Message-ID: <CAKXUXMw77cSBz6nBF4K5BT3AUU2Ac3JGTj6qXJ5mDnUFmRrWeg@mail.gmail.com>
Subject: Re: [PATCH] x86: make config X86_FEATURE_NAMES visible with EXPERT
To:     Dave Hansen <dave.hansen@intel.com>
Cc:     Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
        Dave Hansen <dave.hansen@linux.intel.com>, x86@kernel.org,
        "H . Peter Anvin" <hpa@zytor.com>, kernel-janitors@vger.kernel.org,
        linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, May 9, 2023 at 8:38=E2=80=AFPM Dave Hansen <dave.hansen@intel.com> =
wrote:
>
> On 5/9/23 11:20, Lukas Bulwahn wrote:
> > On Tue, May 9, 2023 at 4:07=E2=80=AFPM Dave Hansen <dave.hansen@intel.c=
om> wrote:
> >>
> >> On 5/9/23 01:40, Lukas Bulwahn wrote:
> >>> Commit 6a108a14fa35 ("kconfig: rename CONFIG_EMBEDDED to CONFIG_EXPER=
T")
> >>> introduces CONFIG_EXPERT to carry the previous intent of CONFIG_EMBED=
DED
> >>> and just gives that intent a much better name.
> >>
> >> That was quite the changelog, and I'm still not quite sure:
> >>
> >> What is the problem with the existing code?
> >>
> >> What is the end user visible impact of this problem and of this propos=
ed
> >> change?
> >>
> >
> > Thanks, Dave, for your feedback.
> >
> > The commit message is so lengthy, as I copied it from the original
> > cover letter into each commit, but I agree with your feedback.
>
> Uhh...  *What* cover letter?  A cover letter is usually the 0/NN message
> in a patch series.  It is not obvious at *all* that this plain
> non-numbered patch is part of a series.
>

Yeah, sending out the patch isolated that originated from a larger
patch series was a bad idea. I got that.

> > So maybe this commit message fits better:
> >
> > To simplify build configurations, the CONFIG_EMBEDDED is to be
> > removed. All configs that are only visible with CONFIG_EMBEDDED=3Dy
> > shall be now visible with CONFIG_EXPERT=3Dy.
> >
> > In the x86 architecture, the config X86_FEATURE_NAMES is visible when
> > CONFIG_EMBEDDED is enabled. Now, make the config X86_FEATURE_NAMES
> > visible when CONFIG_EXPERT is enabled.
> >
> > Dave, what do you think? If that is good enough for you, I will send
> > an updated patch with that commit message.
>
> CONFIG_EMBEDDED is being removed:
>
>         <INSERT LINK HERE>
>
> That means that everything in Kconfig that uses CONFIG_EMBEDDED needs to
> switch over to something else.
>
> Move X86_FEATURE_NAMES over to CONFIG_EXPERT instead of CONFIG_EMBEDDED.
>

Just for completeness of the mailing list archive, I will add the link:

CONFIG_EMBEDDED is being removed:

https://lore.kernel.org/all/20220908104337.11940-1-lukas.bulwahn@gmail.com/

Following Boris' suggestion, I have now created a different patch to
make the option non-visible in the prompt (see Link below). So, this
patch may not be required at all if the other patch is accepted. In
case there is a critical number of users out there that really want to
tinker with this option and we decide to make it visible with
CONFIG_EXPERT, I will propose this patch with the commit message you
provided. For now, this patch is superseded by the patch 1 of 2 below:

Link: https://lore.kernel.org/all/20230510065713.10996-1-lukas.bulwahn@gmai=
l.com/

Thanks for all the help.

Lukas

> --
>
> This actually isn't _great_ for x86.  We hid X86_FEATURE_NAMES behind
> EMBEDDED because we didn't want to see it 99% of the time.  But just
> about everyone uses EXPERT=3Dy, so the end result here is that everyone
> will now see X86_FEATURE_NAMES.
>
> Oh well.  It's just one Kconfig option.  Not a big deal.