Path: ...!feeds.phibee-telecom.net!news.mixmin.net!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Terje Mathisen <terje.mathisen@tmsw.no>
Newsgroups: comp.arch
Subject: Re: Computer architects leaving Intel...
Date: Mon, 16 Sep 2024 10:37:47 +0200
Organization: A noiseless patient Spider
Lines: 55
Message-ID: <vc8qor$2od9v$1@dont-email.me>
References: <2024Aug30.161204@mips.complang.tuwien.ac.at>
 <vbcob9$dvp4$1@dont-email.me> <vbd6ia$e0ld$2@dont-email.me>
 <UxpCO.174965$Hld5.7714@fx15.iad> <vc41rl$1fhjd$1@dont-email.me>
 <2024Sep14.152652@mips.complang.tuwien.ac.at>
 <d93c1dc0455692767c89ea9f7bd47ed1@www.novabbs.org>
 <vc4o0l$1kuqf$1@dont-email.me> <vc6vno$285g2$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Injection-Date: Mon, 16 Sep 2024 10:37:47 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="72cf2766236da00f6658f17f99cdc0aa";
	logging-data="2897215"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/E1HkrglKhhecJp0J6u5F6Bcyl6cdaGkCJdu5UAcMNXg=="
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:91.0) Gecko/20100101
 Firefox/91.0 SeaMonkey/2.53.19
Cancel-Lock: sha1:oY5A6VdegEJnSNuTeCr/rh3znP4=
In-Reply-To: <vc6vno$285g2$1@dont-email.me>
Bytes: 3222

David Brown wrote:
> On 14/09/2024 21:26, Thomas Koenig wrote:
>> MitchAlsup1 <mitchalsup@aol.com> schrieb:
>>
>>> In many cases int is slower now than long -- which violates the notio=
n
>>> of int from K&R days.
>>
>> That's a designers's choice, I think.=C2=A0 It is possible to add 32-b=
it
>> instructions which should be as fast (or possibly faster) than
>> 64-bit instructions, as AMD64 and ARM have shown.
>>
>=20
> For some kinds of instructions, that's true - for others, it's not so=20
> easy without either making rather complicated instructions or having=20
> assembly instructions with undefined behaviour (imagine the terror that=
=20
> would bring to some people!).
>=20
> A classic example would be for "y =3D p[x++];" in a loop.=C2=A0 For a 6=
4-bit=20
> type x, you would set up one register once with "p + x", and then have =
a=20
> load with post-increment instruction in the loop.=C2=A0 You can also do=
 that=20
> with x as a 32-bit int, unless you are of the opinion that enough apple=
s=20
> added to a pile should give a negative number of apples.=C2=A0 But with=
 a=20
> wrapping type for x - such as unsigned int in C or modulo types in Ada,=
=20
> you have little choice but to hold "p" and "x" separately in registers,=
=20
> add them for every load, and do the increment and modulo operation.=C2=A0=
 I=20
> really can't see this all being handled by a single instruction.

This becomes much simpler in Rust where usize is the only legal index typ=
e:

Yeah, you have to actually write it as

   y =3D p[x];
   x +=3D 1;

instead of a single line, but this makes zero difference to the=20
compiler, right?

Terje

--=20
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"