Rocksolid Light - comp.arch - Re: Concertina II... again?

Re: Concertina II... again?

<7f3f9eef-5a23-4997-ad32-037c4b93d747n@googlegroups.com>

https://news.novabbs.org/devel/article-flat.php?id=34565&group=comp.arch#34565

X-Received: by 2002:a05:620a:21c7:b0:777:357b:33b0 with SMTP id h7-20020a05620a21c700b00777357b33b0mr63914qka.5.1697563662947;
Tue, 17 Oct 2023 10:27:42 -0700 (PDT)
X-Received: by 2002:a05:6871:58d:b0:1ea:2dd6:6a86 with SMTP id
u13-20020a056871058d00b001ea2dd66a86mr1259926oan.9.1697563661270; Tue, 17 Oct
2023 10:27:41 -0700 (PDT)
Path: i2pn2.org!i2pn.org!weretis.net!feeder6.news.weretis.net!border-2.nntp.ord.giganews.com!nntp.giganews.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.arch
Date: Tue, 17 Oct 2023 10:27:41 -0700 (PDT)
In-Reply-To: <7991ac06-a03e-465b-8560-7538a7a8b611n@googlegroups.com>
Injection-Info: google-groups.googlegroups.com; posting-host=2001:56a:fa34:c000:e1fe:c53b:dea7:660f;
posting-account=1nOeKQkAAABD2jxp4Pzmx9Hx5g9miO8y
NNTP-Posting-Host: 2001:56a:fa34:c000:e1fe:c53b:dea7:660f
References: <7fb66351-9136-492e-9879-5b338eef1359n@googlegroups.com>
<677d90c6-fcac-43e0-a073-6165e59c8acbn@googlegroups.com> <a582b4da-eeb5-4347-ba65-f0b5d84589f7n@googlegroups.com>
<25ad0119-92d7-4dd2-8d47-397e78eed150n@googlegroups.com> <34cfc1b6-6ad1-4e3b-826c-e3758f0826e8n@googlegroups.com>
<a4cee9db-4b1f-431a-bc00-62d8db13124bn@googlegroups.com> <1c1a74af-29fa-4ea7-963a-55b438c96be0n@googlegroups.com>
<7991ac06-a03e-465b-8560-7538a7a8b611n@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <7f3f9eef-5a23-4997-ad32-037c4b93d747n@googlegroups.com>
Subject: Re: Concertina II... again?
From: jsavard@ecn.ab.ca (Quadibloc)
Injection-Date: Tue, 17 Oct 2023 17:27:42 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Lines: 26

by: Quadibloc - Tue, 17 Oct 2023 17:27 UTC

On Sunday, March 12, 2023 at 6:31:39 PM UTC-6, MitchAlsup wrote:

> Would I be rude in asserting that I think blocks and block headers are a putting you
> "up sh!t creak without a paddle".

This had started me thinking. Earlier, I did post a reply noting how I _could_ manage
to convert the ISA to one with conventional variable-length instructions.

Basically, the feature of having immediate values of all lengths available for instructions
was one thing that tended to push me in the direction of using a block structure - in
addition to the desire to include features normally associated with VLIW architectures,
like explicitly indicating instructions that can or can't execute in parallel.

So rather than converting to variable-length instructions, I've just been holding off on
doing any more work on Concertina II until I can come up with something new that will
help; either a better variable-length instruction format, or a much simpler block structure.

John Savard

Captain's Log, star date 21:34.5...

devel / comp.arch / Re: Concertina II... again?

Subject	Author
Re: Concertina II... again?	Quadibloc