Article View: comp.sys.sun.hardware
Article #2598Re: SS20 too slow for 2D or hw malfunction?
From: Daniel J. Pavon-Moreno
Date: Sat, 21 Jun 2003 20:11
Date: Sat, 21 Jun 2003 20:11
38 lines
3296 bytes
3296 bytes
ultrasparc3@hotmail.com wrote: > In <bctcuf$fj2$1@jedi.apana.org.au> cd@lios.apana.org.au (Kralizec Craig) writes: > >>Make absolutely sure you configure the Xservers file correctly too - to get >>the best performance you *must* run Xsun in Truecolor mode. This means you >>need to specify: > > > and if you have SX-optimized software (XIL pipelines, etc) you should > reserve some of your memory for the SX to use. it requires a chunk > of physically-contiguous memory for best operation since it operates > on physical, not virtual, addresses. the sx_cmem kernel module did > that and you really only got excellent performance using the graphics > plug-ins (XIL pipelines, etc.) that made use of sx_cmem. there was a > sxconfig program that would set it up for you. refer to the man page > for details. > Oh, I reserved 32MB of contiguous memory, but both imagetool & showmetv (which are linked against the xil libraries) run really slow. In particular, window scrolling in imagetool is a joke. And image operations using the pallete (rotations in particular) often lead to image corruption in my system. > >>Using a single SM71 CPU doesn't help either. You *need* as much CPU >>brute-force as possible when using any 24-bit video platform, and since the >>SX/cg14 is tied directly into the local buses on the SS20 system board, >>better CPU performance translates directly into better video performance. > > > SuperSPARC-only modules also did a better (at least until processors > faster than 60MHz showed up) since having the SuperCACHE would add a > cycle or two of latency to the MBus accesses which is how you talk > to the cg14. Interesting. So you traded cpu performance for gfx performance. Great design.
Technical Details
Message-ID:
<3EF4BBDC.5080103@neoholistic.org>Path:
arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!newsfeed.stanford.edu!cyclone.bc.net!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news-out.superfeed.net!propagator2-maxim!news-in.superfeed.net!nsnmpen1-lo.nuria.telefonica-data.net!news.ya.com!yacom!not-for-mailReferences:
<3EEFB4DA.53AF5BAC@neoholistic.org> <bctcuf$fj2$1@jedi.apana.org.au> <2003520.02731.14904@cable.prodigy.com>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: f4671,8b1b6e084b2535a8 X-Google-Attributes: gidf4671,public X-Google-Thread: 113606,8e82577aad67e10b X-Google-Attributes: gid113606,public X-Google-Thread: 10d053,8e82577aad67e10b X-Google-Attributes: gid10d053,public X-Google-ArrivalTime: 2003-06-21 12:58:01 PST Path: arc1-ca.pugleaf.net!archiver1.google.com!news1.google.com!newsfeed.stanford.edu!cyclone.bc.net!sjc70.webusenet.com!news.webusenet.com!pd2nf1so.cg.shawcable.net!residential.shaw.ca!news-out.superfeed.net!propagator2-maxim!news-in.superfeed.net!nsnmpen1-lo.nuria.telefonica-data.net!news.ya.com!yacom!not-for-mail From: "Daniel J. Pavon-Moreno" <djpavon@neoholistic.org> Newsgroups: comp.sys.sun.hardware,aus.computers.sun,uk.comp.sys.sun Subject: Re: SS20 too slow for 2D or hw malfunction? Date: Sat, 21 Jun 2003 22:11:08 +0200 Organization: ya.com internet factory Lines: 39 Message-ID: <3EF4BBDC.5080103@neoholistic.org> References: <3EEFB4DA.53AF5BAC@neoholistic.org> <bctcuf$fj2$1@jedi.apana.org.au> <2003520.02731.14904@cable.prodigy.com> NNTP-Posting-Host: 80-36-7-132.adsl.ya.com Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Trace: news.ya.com 1056225476 22231 80.36.7.132 (21 Jun 2003 19:57:56 GMT) X-Complaints-To: newsmaster@ya.com NNTP-Posting-Date: Sat, 21 Jun 2003 19:57:56 +0000 (UTC) User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.0) Gecko/20020623 Debian/1.0.0-0.woody.1 X-Accept-Language: es, en