Received-SPF: pass (google.com: domain of linux-kernel+bounces-32280-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Date: Sun, 21 Jan 2024 18:31:53 -0800
From: "H. Peter Anvin" <hpa@zytor.com>
To: Steven Rostedt <rostedt@goodmis.org>
CC: Thorsten Glaser <tg@debian.org>, Peter Zijlstra <peterz@infradead.org>,
        x86@kernel.org, torvalds@linuxfoundation.org,
        linux-kernel@vger.kernel.org, linux-toolchains@vger.kernel.org,
        jpoimboe@redhat.com, alexei.starovoitov@gmail.com, mhiramat@kernel.org
Subject: Re: [PATCH 1/2] x86: Remove dynamic NOP selection
User-Agent: K-9 Mail for Android
In-Reply-To: <20240121212236.02331f4c@rorschach.local.home>
References: <20210312113253.305040674@infradead.org> <20210312115749.065275711@infradead.org> <Pine.BSM.4.64L.2401200654480.19429@herc.mirbsd.org> <CE53F232-3D2D-4910-94B4-A4304F5990C7@zytor.com> <Pine.BSM.4.64L.2401201652490.29203@herc.mirbsd.org> <d23e1b70-7ac3-49cd-8d5f-1dedc0108603@zytor.com> <Pine.BSM.4.64L.2401212356390.999@herc.mirbsd.org> <20B57BDC-9A3F-4B02-9BA9-41477088A6CE@zytor.com> <20240121195649.7355e1d5@rorschach.local.home> <020a1803-25c4-4365-8c53-291becd94632@zytor.com> <20240121212236.02331f4c@rorschach.local.home>
Message-ID: <50F640B7-AB54-4571-9988-EAD5FD5933A4@zytor.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain;
 charset=utf-8
Content-Transfer-Encoding: quoted-printable

On January 21, 2024 6:22:36 PM PST, Steven Rostedt <rostedt@goodmis=2Eorg> =
wrote:
>On Sun, 21 Jan 2024 18:15:39 -0800
>"H=2E Peter Anvin" <hpa@zytor=2Ecom> wrote:
>
>> On 1/21/24 16:56, Steven Rostedt wrote:
>> >>
>> >> Yes, but it is a matter of where we optimize for performance as oppo=
sed to correctness=2E =20
>> >=20
>> > There is no such thing as "optimize for correctness", it is either
>> > correct or it is not=2E Correctness should always come before perform=
ance
>> > (at least that is what Thomas has pounded into me ;-)
>> >=20
>> > If a kernel use to work on a machine but a newer version no longer
>> > works, I call that a regression=2E
>> >  =20
>>=20
>> There absolutely is such a thing as "optimize for correctness=2E" It me=
ans=20
>> to keep the code clean, easily testable, and with a minimal number of=
=20
>> distinct code paths so that regressions and *especially* uncaught=20
>> regressions get caught quickly=2E
>
>I call that maintainability, not correctness=2E It is either correct and
>works, or is incorrect and does not work=2E
>
>You can change code to be more maintainable and still make it incorrect=
=2E
>
>-- Steve

Yes, of course=2E That's called failure :)