home *** CD-ROM | disk | FTP | other *** search
- Path: news.eunet.fi!fipnet!kone!jsaarinen
- Newsgroups: comp.sys.amiga.programmer
- X-NewsReader: IntuiNews 1.2b (31.7.94)
- References: <38232280@kone.fipnet.fi> <987.6609T1075T1727@ifi.uio.no>
- From: "Jyrki Saarinen" <jsaarinen@kone.fipnet.fi>
- Date: Tue, 6 Feb 96 17:15:08 UT
- Comments: Illegal date header - new date added by quicknews
- X-Original-Date: Tue, 06 Feb 96 18:59:54
- MIME-Version: 1.0
- Content-Type: text/plain; charset=iso-8859-1
- Content-Transfer-Encoding: binary
- Subject: Re: PPC compilers
- Message-ID: <38232333@kone.fipnet.fi>
-
-
- > That should fix the delay problem. For 1 addx and 2 addx'es. :)
- > Remember when you use 2 addx the fraction parts are swapped, so the
- > fract and x-flag is calculated on the addx above, so there will be no
- > delay problems on the second addx. Everything should be correct!
-
- Yep. Now the optimal polygon routine looks like this:
- (Gouraud/Texture with a Shading table)
-
- move.w d2,d4 ;u<<8
- move.l a0,d5 ;g<<8
- move.b d2,d4 ;+v
- move.b (a2,d4.w),d5 ;read texel
- move.b (a3,d5.l),(a4)+ ;read pixel from the shading table
- addx.l d3,d2 ;u+=ustep
- addx.l d1,d0 ;v+=vstep
- add.l a1,a0 ;g+=gstep
-
- That should be about 30 68020/68030 cycles per pixel.
-
- Now, how the hell this should be scheduled for maximum
- performance on the 68040 and on the 68060? Anyone?
-
- -- _
- a Stellar programmer _ //
- "Amiga - back for the future" \X/
-