Article 71308 of comp.sys.cbm: Path: news.acns.nwu.edu!newsfeed.acns.nwu.edu!news.ece.nwu.edu!news.cse.psu.edu!rutgers!cam-news-feed2.bbnplanet.com!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!europa.clark.net!newsfeeds.sol.net!sol.net!spool.mu.edu!newsspool.sol.net!munnari.OZ.AU!metro!metro!seagoon.newcastle.edu.au!cc.newcastle.edu.au!ecbm From: "Bruce R. McFarling" Newsgroups: comp.sys.sinclair,comp.sys.cbm Subject: Re: Spectrum Emulator for C64 Date: Fri, 11 Jul 1997 17:23:40 +1000 Organization: The University of Newcastle Lines: 52 Message-ID: References: <339d8a2a.3224107@news.demon.co.uk> <11408.imc@comlab.ox.ac.uk> <11440.imc@comlab.ox.ac.uk> NNTP-Posting-Host: cc.newcastle.edu.au Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <11440.imc@comlab.ox.ac.uk> Xref: news.acns.nwu.edu comp.sys.sinclair:41201 comp.sys.cbm:71308 On 10 Jul 1997, Ian Collier wrote: > In article , > "Bruce R. McFarling" wrote: > > BTW, a fully unrolled loop (obviously) doesn't care about byte > >order. However, if we restricted the 8-byte BCD number to be page aligned, > >the following can be put into RAM somewhere (you'll notice that it aint > >ROMMable) > > >op2 = *+11 > > Surely 11 is not the right number if you are referring to the $FF00 below? Why not? Actually, I meant that literally, rather than rhetorically, since I don't have it in front of me at this office. Remember that all instructions are one byte, all zero-page address modes add one byte, and all absolute address modes add two bytes, and OP1 is on the zero page. > It would be easier and clearer if you coded it as x+1 where x is the label > which you have called "-". The "-" is a technique used in some 6502 assemblers to handle the many short loops that the 6502 encourages you to write: it is a purely relative reference that means "one back" -- where BCC -- would skip past the first previous "-" to the next previous "-" (similarly for "BCC +" and "BCC ++"). And soitanly it would have been easier and clearer to write OP2 = OP2_LOAD+1 ; ... OP2_LOAD: - LDA $FF00,X > ... > Making assumptions about page alignment allows you to write INC L (4) > instead of INC HL (6) to advance the pointer on each step. Other than > that, "misbehaving" can't really help you as everything is done in registers > so there's no point in modifying a constant in the code (since there aren't > any, and putting one in would make it slower). OK, I'll subtract however many 2 clock cycles that makes it and refigure. Then someone can post a 65816 routine that shows what an upgraded C64 can do. Virtually, Bruce R. McFarling, Newcastle, NSW ecbm@cc.newcastle.edu.au Article 71309 of comp.sys.cbm: Path: news.acns.nwu.edu!newsfeed.acns.nwu.edu!news.ece.nwu.edu!news.cse.psu.edu!rutgers!cam-news-feed2.bbnplanet.com!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!paladin.american.edu!news.ecn.uoknor.edu!munnari.OZ.AU!metro!metro!seagoon.newcastle.edu.au!cc.newcastle.edu.au!ecbm From: "Bruce R. McFarling" Newsgroups: comp.sys.sinclair,comp.sys.cbm Subject: Re: Spectrum Emulator for C64 Date: Fri, 11 Jul 1997 17:26:03 +1000 Organization: The University of Newcastle Lines: 14 Message-ID: References: <339d8a2a.3224107@news.demon.co.uk> <11408.imc@comlab.ox.ac.uk> <11441.imc@comlab.ox.ac.uk> NNTP-Posting-Host: cc.newcastle.edu.au Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <11441.imc@comlab.ox.ac.uk> Xref: news.acns.nwu.edu comp.sys.sinclair:41202 comp.sys.cbm:71309 On 10 Jul 1997, Ian Collier wrote: > PS I notice that you didn't seem to set decimal mode in that example. Did someone erase a SED between Australia and c.s.*? Dagnab those nasty net gremlins. Did it also steal the PHP and PLP? Those are the cycle stealers SEx and CLx instructions are fairly cheap when the processor status doesn't have to be restored. Virtually, Bruce R. McFarling, Newcastle, NSW ecbm@cc.newcastle.edu.au Article 71307 of comp.sys.cbm: Path: news.acns.nwu.edu!newsfeed.acns.nwu.edu!news.ece.nwu.edu!news.cse.psu.edu!rutgers!cam-news-feed2.bbnplanet.com!cam-news-hub1.bbnplanet.com!cpk-news-hub1.bbnplanet.com!news.bbnplanet.com!paladin.american.edu!news.ecn.uoknor.edu!munnari.OZ.AU!metro!metro!seagoon.newcastle.edu.au!cc.newcastle.edu.au!ecbm From: "Bruce R. McFarling" Newsgroups: comp.sys.sinclair,comp.sys.cbm,comp.emulators.cbm Subject: Re: Spectrum Emulator for C64 Date: Fri, 11 Jul 1997 17:08:44 +1000 Organization: The University of Newcastle Lines: 25 Message-ID: References: <339d8a2a.3224107@news.demon.co.uk> <5ppnb4$1036@ds2.acs.ucalgary.ca> <5pv3b0$nkc@news.acns.nwu.edu> NNTP-Posting-Host: cc.newcastle.edu.au Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <5pv3b0$nkc@news.acns.nwu.edu> Xref: news.acns.nwu.edu comp.sys.sinclair:41200 comp.sys.cbm:71307 comp.emulators.cbm:22820 On 9 Jul 1997, Stephen Judd wrote: [Quoth I]: > > Rather 'many' than 'few' variables. Having few enough variables > >that they can be maintained in the Z80 register set without shuffling is > >an *advantage* for the 6502. > dis > or > Z80 > :) Well, put a bow tie on Beaker and color me red. #-8 Actually now that I read it again, it ought to have been: > >an *advantage* over the 6502. But I wasn't playing close enough attention and my fingers refused to type it that way. BTW, I take it that cycles in Z80 land are M cycles and not T cycles? Virtually, Bruce R. McFarling, Newcastle, NSW ecbm@cc.newcastle.edu.au Article 71465 of comp.sys.cbm: Path: news.acns.nwu.edu!newsfeed.acns.nwu.edu!news.ece.nwu.edu!news.cse.psu.edu!uwm.edu!spool.mu.edu!sol.net!spool.mu.edu!newsspool.sol.net!howland.erols.net!europa.clark.net!dispatch.news.demon.net!demon!delos.dra.hmg.gb!server1.netnews.ja.net!server5.netnews.ja.net!server6.netnews.ja.net!server4.netnews.ja.net!server2.netnews.ja.net!news.ox.ac.uk!news From: imc@ecs.ox.ac.uk (Ian Collier) Newsgroups: comp.sys.sinclair,comp.sys.cbm Subject: Re: Spectrum Emulator for C64 Date: 14 Jul 1997 17:21:55 GMT Organization: Oxford University Computing Laboratory, UK Message-ID: <11470.imc@comlab.ox.ac.uk> References: <339d8a2a.3224107@news.demon.co.uk> <11440.imc@comlab.ox.ac.uk> NNTP-Posting-Host: boothp2.ecs.ox.ac.uk X-Local-Date: Monday, 14th July 1997 at 6:21pm BST Lines: 38 Xref: news.acns.nwu.edu comp.sys.sinclair:41418 comp.sys.cbm:71465 In article , "Bruce R. McFarling" wrote: >On 10 Jul 1997, Ian Collier wrote: >> >op2 = *+11 >> > Surely 11 is not the right number if you are referring to the $FF00 below? > Why not? Actually, I meant that literally, rather than >rhetorically, since I don't have it in front of me at this office. >Remember that all instructions are one byte, all zero-page address modes >add one byte, and all absolute address modes add two bytes, and OP1 is on >the zero page. op2 = *+11 offset ADD16DEC: STA OP2+1 ; 4 0 CLC ; 2 3 TYA ; 2 4 ADC #8 ; 2 5 STA OP2 ; 4 7 CLC ; 2 10 LDX #(0-8) ; 2 11 - LDA $FF00,X ; 5 13 Am I wrong? > The "-" is a technique used in some 6502 assemblers to handle the >many short loops that the 6502 encourages you to write: it is a purely >relative reference that means "one back" -- where BCC -- would skip past >the first previous "-" to the next previous "-" (similarly for "BCC +" and >"BCC ++"). If it's _really_ short, I write the number of bytes directly. inc l:jr nz,1:inc h Otherwise I stick a label in. It's not _that_ much effort, surely? :-) -- ---- Ian Collier : imc@comlab.ox.ac.uk : WWW page (including Spectrum section): ------ http://www.comlab.ox.ac.uk/oucl/users/ian.collier/imc.html Article 71571 of comp.sys.cbm: Path: news.acns.nwu.edu!newsfeed.acns.nwu.edu!news.luc.edu!chi-news.cic.net!howland.erols.net!gatech!purdue!nick.arc.nasa.gov!news.Hawaii.Edu!munnari.OZ.AU!metro!metro!seagoon.newcastle.edu.au!cc.newcastle.edu.au!ecbm From: "Bruce R. McFarling" Newsgroups: comp.sys.sinclair,comp.sys.cbm Subject: Re: Spectrum Emulator for C64 Date: Tue, 15 Jul 1997 19:50:40 +1000 Organization: The University of Newcastle Lines: 75 Message-ID: References: <339d8a2a.3224107@news.demon.co.uk> <11408.imc@comlab.ox.ac.uk> <11440.imc@comlab.ox.ac.uk> Reply-To: "Bruce R. McFarling" NNTP-Posting-Host: cc.newcastle.edu.au Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In-Reply-To: <11440.imc@comlab.ox.ac.uk> Xref: news.acns.nwu.edu comp.sys.sinclair:41545 comp.sys.cbm:71571 On 10 Jul 1997, Ian Collier wrote: > >op2 = *+11 > > Surely 11 is not the right number if you are referring to the $FF00 below? > It would be easier and clearer if you coded it as x+1 where x is the label > which you have called "-". Surely it's not (the right number, that is). I tacked it on at the end. Here are some more. These are for a 16-digit value located at a given location in memory, targetted at a 16-digit span in the zero page passed in the X register. There are 8 of these, ADD_OP2, ADD_OP2_2, to ADD_OP2_9. ADD_OP2_x PHP SED LDY #7 TXA CLC ADC #7 TAX CLC - LDA OP2_x,Y ADC OP1,X STA OP1,X DEX DEY BNE - LDA OP2_x ADC OP1,X STA OP1,X PLP RTS This one fills a multiplication table with the products of the 15 digit quantity at OP2 ... OP2+7 (which also serves the role of the product of one) : TARGET = OP2+8 FINAL = 8*8 PHP SED LDX #7 - LDY #7 CLC - LDA OP2,Y ADC OP2,X STA TARGET,X DEX DEY BNE - - LDA OP2 ADC OP2,X STA TARGET,X TXA CLC ADC #15 TAX CMP #72 BMI -- PLP RTS Fixes, cycle counts and superior solutions for this particular problem welcome. And, surely it's clear what this particular problem is, since the quickest way I have to find out how fast a Z80 can do a multi-digit BCD multiplication is to post one for a 6502. Virtually, Bruce R. McFarling, Newcastle, NSW ecbm@cc.newcastle.edu.au