Path: csiph.com!usenet.pasdenom.info!gegeweb.org!de-l.enfer-du-nord.net!feeder1.enfer-du-nord.net!feeds.phibee-telecom.net!newsfeed.xs4all.nl!newsfeed2.news.xs4all.nl!xs4all!news2.euro.net!postnews2.euro.net!news.wanadoo.nl!not-for-mail
From: mhx@iae.nl (Marcel Hendrix)
Subject: Re: Offline compilation of Forth
Newsgroups: comp.lang.forth
Message-ID: <97821397018434@frunobulax.edu>
Date: Wed, 6 Feb 2013 22:53:06 +0200
References: <1540968.teCMtnv6FK@sunwukong.fritz.box>
X-Newsreader: iForth 2.0 console (October 21, 2006)
Lines: 18
Organization: Wanadoo
NNTP-Posting-Date: 06 Feb 2013 21:52:56 GMT
NNTP-Posting-Host: s529d937f.adsl.online.nl
X-Trace: 1360187576 dr7.euro.net 240 82.157.147.127:58431
X-Complaints-To: abuse@wanadoo.nl
Xref: csiph.com comp.lang.forth:19518

Bernd Paysan <bernd.paysan@gmx.de> writes Re: Offline compilation of Forth
[..]
> For x64, having SSE/SSE2 is indeed a must.  David Kühling added SSE/SSE2 
> to the x64 version of my assembler, and it's in total now 730 lines.  
> What's now missing are only recent additions like AVX and AES-NI.

Can you give a hint where to find it? The bigForth distribution
(sourceforge/subversion) shows assem486.fb (a screen file) which only seems to have a 
handful of 3Dnow and mmx words (no SSE/SSE2).

I am very curious how David got a full SSE/SSE2 assembler in 100 lines,
given that most instructions have tiny abnormalities and never quite the
same addressing possibilities. For example, the cvt(t)xx2yyy instruction 
needs 50 lines just to enter all the different opcode name varieties
(the SSE/SSE2 part, covering all instructions, is 510 lines in total).

-marcel