MIME-Version: 1.0
References: <20220110025203.2545903-1-kuba@kernel.org> <CAHk-=wg-pW=bRuRUvhGmm0DgqZ45A0KaH85V5KkVoxGKX170Xg@mail.gmail.com>
 <20220110205603.6bc9b680@kicinski-fedora-PC1C0HJN.hsd1.ca.comcast.net>
In-Reply-To: <20220110205603.6bc9b680@kicinski-fedora-PC1C0HJN.hsd1.ca.comcast.net>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon, 10 Jan 2022 21:02:47 -0800
Message-ID: <CAHk-=wieO-N=XD73D=OKJ88Y_=GiF9OGm2yg3s_nvafPxAFYwg@mail.gmail.com>
Subject: Re: [GIT PULL] Networking for 5.17
To:     Jakub Kicinski <kuba@kernel.org>
Cc:     David Miller <davem@davemloft.net>,
        Netdev <netdev@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Jan 10, 2022 at 8:56 PM Jakub Kicinski <kuba@kernel.org> wrote:
>
>   The only explanation I can come up with it's that it was
> done so that people running olddefconfig right after the per-vendor
> split was introduced wouldn't lose all drivers.

Yeah, I think that was the reason. Don't hide existing drivers behind
vendors that got turned off.

I think it would probably be fine to have entirely new vendors (ie "no
existing drivers") default to 'n' when added, just to cut down on the
huge amount of driver questions.

But right now they all act the other way around, and are just a way to
explicitly turn off questions for a vendor you know you don't care
about.

It does perhaps get a bit confusing if some vendors are 'default y'
and others are 'default n'. It's easy enough to explain at the time
when that vendor is added, but then ten years passes, and somebody
asks why something like DEC or 3com is 'default y', while soem big new
vendor migth be 'default n'.

So I dunno.

            Linus