X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f From: Rugxulo Newsgroups: comp.os.msdos.djgpp Subject: Re: CWSDPMI r5 test update status? Date: Mon, 22 Dec 2008 23:03:41 -0800 (PST) Organization: http://groups.google.com Lines: 32 Message-ID: <3c9a01aa-72ac-40c0-b9cf-e4007cae17fd@x14g2000yqk.googlegroups.com> References: NNTP-Posting-Host: 65.13.115.246 Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 X-Trace: posting.google.com 1230015821 2158 127.0.0.1 (23 Dec 2008 07:03:41 GMT) X-Complaints-To: groups-abuse AT google DOT com NNTP-Posting-Date: Tue, 23 Dec 2008 07:03:41 +0000 (UTC) Complaints-To: groups-abuse AT google DOT com Injection-Info: x14g2000yqk.googlegroups.com; posting-host=65.13.115.246; posting-account=p5rsXQoAAAB8KPnVlgg9E_vlm2dvVhfO User-Agent: G2/1.0 X-HTTP-UserAgent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.0.5) Gecko/2008120122 Firefox/3.0.5,gzip(gfe),gzip(gfe) To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id mBN7F4w9022083 Reply-To: djgpp AT delorie DOT com Hi, On Dec 22, 6:30 pm, "Charles Sandmann" wrote: > > Just wondering if the CWSDPMI r5 test update will be promoted to release > > anytime soon. > > There have been no reported problems with the r5 update, but I have been > too busy to deal with it recently. If you need it, grab the test release > and start using it. Trivial updates to 8 year old software usually aren't > urgent. The SSE thing is only an issue for pure DOS since modern OSes (Win98SE, Linux 2.4) do it for you. The advantage of having CWSDPMI do it for us is that a). we don't have to manually do it ourselves in every single app, b). we can use the standard "ring 3" CWSDSTUB without worrying about crashing the machine, and c). virtual memory!! (Previously I was manually turning it on myself and using D3X, which is stub only and only ring 0 without virtual memory.) I'm not a math guru at all, and GCC / DJGPP doesn't really generate SSE automatically so far, so really you have to hand-code your SSE (like Matt Mahoney did for PAQ8). It can bring noticeable speedups, just slightly hard to understand, IMHO. (But bigger registers, independent of FPU, etc. vs. MMX. Besides, don't new Core 2s / Phenoms have bigger SSE bandwidth, e.g. 256 bits now?? So SSE should be faster now. And they say MMX is deprecated, no float doubles like SSE2, only integers. So it's considered less useful.) I'm trying to package up a new paq8o8z update, but as always, it's such a wimpy port, so caveat emptor. Besides, nobody but me seems to care anyways. (What, recompressing .JPGs in FreeDOS isn't fun?) ;-)