--------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00008Date: 04/29/97 From: DIMITRI SMITS Time: 11:00pm \/To: BARTON PAUL LEVENSON (Read 2 times) Subj: Transcendental Functions > DM> Any chance you could post the qbasic? I'm rather > DM> interested in the algorithms. Taylor's theorem was > I don't think I'm allowed to here -- you'll have to > ask the moderator. If he allows it, I'll post them. if the moderator doesn't allow it, could you NETMAIL it to me then? either to 2:292/8013.12 or o01s96@zorro.ruca.ua.ac.be thnx Greetz, Da Gongo --- FMail 1.02 * Origin: Only The Lonely ... (2:292/8013.12) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00009Date: 05/05/97 From: REJEAN LAMBERT Time: 07:34pm \/To: ALL (Read 2 times) Subj: Pure asm help needed! I'm fairly new to assembler (not quite a year), and up to now, i've only used it with the turbo languages (inline asm). Now I have written a program entirely in assembler, but within Turbo Pascal. It goes like this: program Doofus; begin asm; ... end; end. What do I have to change to turn my tp-inline program into a standalone ASM file that would compile with TASM or MASM ? BTW the program is a speedy version of the Munching Squares ( Y = X xor T ). Uses 2 bytes of stack space (push 0a000h, pop es) that could easily be removed. No variables, no memory allocation. It's as plain as you can get. --- GEcho 1.00 * Origin: Glow in the Dark BBS ... flickering in!! (1:163/115) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00010Date: 05/04/97 From: ANDREW FRANK Time: 11:52pm \/To: DENIS BOYLES (Read 2 times) Subj: Re: Accessing memory on As Denis Boyles had made known to Andrew Frank, on 01 May 97 14:18:02, I quote. DB> Probably a good idea, I did the same thing when I tried inline with DB> QC to. When in doubt save it to the stack. :) I was "curious" thought DB> of what QC was doing "behind the scences" and examined the code. Which DB> is where I found SI/DI being save for me already. I've looked at what needs to be done for my problem, and decided that I'll live with C, atleast until I make some more progress learning the basics of assembly. DB> Hmm, there is, but it would depend on how you mean by "loading" more DB> than one byte. Do you mean int a register itself? Well, you can DB> depending on the CPU and code. DB> As well as the BYTE, you can load a WORD size value. DB> ie: DB> lodsw I suppose that this only works if the word is at an even address? I've heard that it's more complicated, and more time consuming if it's at an odd address. ... 355/113 A pretty good recipe for a Pi. --- Blue Wave/386 v2.30 [NR] * Origin: fks Online! * Ontario, Canada * (905)820-7273 * (1:259/423) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00011Date: 05/05/97 From: JOE KOSS Time: 04:51pm \/To: DORAN MOPPERT (Read 2 times) Subj: Re: scrolling > It does not complicate it any more than your own approach (which > I accept as valid - I'm just adding my own two cents =P) if the > music player source code is available. Otherwise it's a quite > simple matter to store and call the original int vector. Self > modifying code at its easiest =). Hmm, I think it does. Now your music sequencer must run at a specific frequency .. enhancing the granularity of the sequencing would break the rest of the program (unless you hack in new code to keep a "compatible" counter around .. ) It is generally poor programming practice for one library to require that another library "do something" that isn't part of the overall intent of said library. (Its a music replayer, not a timer) --- GEcho 1.02+ * Origin: Midnight Madness <-> Hartford, CT (1:142/8076) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00012Date: 05/05/97 From: JOE KOSS Time: 04:53pm \/To: SAM IZZO (Read 2 times) Subj: Re: scrolling > > What is LSB? > > JK> Least Significant Byte > Least Significant Bit No, I surely meant Least Significant Byte and Most Significant Byte, as the source code I posted suggests: > > JK> xchg ah, al > > JK> in al, 50h ; MSB of timer 0 > > JK> xchg ah, al and woah, that should be 60h not 50h --- GEcho 1.02+ * Origin: Midnight Madness <-> Hartford, CT (1:142/8076) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00013Date: 05/06/97 From: N.FERRI@AGORA.STM.IT Time: 01:03am \/To: ALL (Read 2 times) Subj: Re: External Programs From: N.Ferri@agora.stm.it Subject: Re: External Programs Scritto il 5-Mag-97 alle 17:01:54 da Denis.Boyles@f290.n153.z1.fidonet.org: &>& So, for your example you would need to change the initialized data &>& of CMDTAIL above to: &>& cmdtail db 6," /T /P",0Dh &>& So, the tail itself would be 6 bytes in length, and would end with &>& the 0D character. The first "space" is to make M/T/P as M /T /P, &>& though for a / DOS will seperate that anyway. (ie. M/T or M /T &>& should work in DOS) Thanks for your explanations, but it doesn't work for me. I'm trying to launch sys.com with the command tail a: (sys a:). I always get the message "important parameter missing" :( It seems it cannot see the "a:" switch. I wrote it the following way: cmdtail db 3," a:",0dh cmdtail db 2,"a:",0dh None of the above works. Why ? What's wrong with it ? Thanks Nicola --- MMMR v4.70reg | The home of Net 346 Services --- * Origin: (1:346/100) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00014Date: 05/05/97 From: JOHN GARDENIERS Time: 07:25pm \/To: DENIS BOYLES (Read 2 times) Subj: DOS return code (Errorle Hi Denis, -=> 01 May 97 15:08, Denis Boyles wrote to John Gardeniers <=- JG>> According to RB's interrupt list INT 21h, function 4Dh will retrieve JG>> the return code from the previous program. DB> So they say. I suspect they've simply left out a couple of important points, such as exactly when the routine will or won't work. JG>> small program which exits with a known errorlevel I have no problem JG>> using that number from within a batch file. DB> I tried that too and had the same results. The ERRORLEVEL part would DB> work, but I would always get 0,0 for when using func. 4D. As you can imagine, there's a lot to be said for a simple way to save the errorlevel from one program, and use it at a later stage. Right now I do it with environment variables, but that's a far from ideal solution. DB> It would appear that the function is only valid directly after you DB> execute the program. On the commandline, the program is being DB> executed from the shell. Which in turn would get the exit code itself DB> and store it away for the ERRORLEVEL. That's about the way I see it. Command.com (or alternative) swallows the last return code. Just to make matters even worse, Command.com doesn't appear to return anything other than a zero when you use it to execute it's built-in commands (copy, dir, etc.), even if there are problems. DB> However if you run your own program which EXECs another, then the DB> function does work. I just "checked" with my EXEC.ASM demo program, DB> which I just re-posted to someone else. True, but that isn't always an option. Looks like I'll just have to find a better way in a different way. Any idea of where the errorlevel is stored while executing a batch file? ttyl, >>> Fuse >>> ... Psst! Your .ZIP file is open. --- GoldED/386 3.00.Alpha4+ * Origin: The Cubby House, assembled by hand (3:632/360.70) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00015Date: 05/03/97 From: PHIL QUINTON Time: 02:07pm \/To: CAREY BLOODWORTH (Read 2 times) Subj: Serial Chip / Modem contr CB> A _WINMODEM_..! That's a whole different thing entirely. CB> All you had ever said was a HSP modem. Sorry, I was under the impression that they were the same! PQ>> So I want to write some. I need to know how to intercept Comm port PQ>>intercepts CB> Sorry, it's not that simple. Not even close to being that CB> simple. It's not just a matter of taking the signals and CB> converting them to/from analog/digital. There is a lot of CB> very delicate signal processing to deal with echo CB> cancellation, terrible phone lines, data compression, error CB> correction (both MNP and v42bis type). What a Git. I forgot about echo cancellation, and terrible phone lines. CB> I've heard there is some code that you can run to turn some CB> sound cards into a 2400 baud modem, and that is not trivial. CB> Going from 2400 to 14.4k is a massive leap in programming CB> requirements, signal processing, licensing requirements (the CB> specs are copyright....), and so on. CB> A lot of people have disliked the RPI and Winmodems because CB> they don't come with DOS drivers, but nobody is writing them CB> because there are simply too many details to deal with. Probably because it is easier to intercept IO calls via windows better, using VxD style files and all. CB> Incidentally, software versions (like Winmodem) of 33.6k and CB> 56k modems should appear in the not too distant future. I know, mine is a 33.6k. My 133 cannot connect at 33.6 because it isn't powerful enough. Perhaps now I will give up thinking of writting a Dos driver for it, I would need all the algorithms to apply to the numbers coming from the DSP to sort out noise etc.. Thanks for your help, and patience.. -=> Phil <=- --- * Origin: Watch the Shadows, Win95 is following.. (2:250/607.96) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00016Date: 05/04/97 From: MARIUS BENDIKSEN Time: 12:37pm \/To: PAUL CHAMBERLAIN (Read 2 times) Subj: need info-protected mode > Questions: > ---------- > 1a. How does the AVL bit (bit 52) in the GDT descriptor affect the > operation of the CPU when a selector is loaded into a segment > register ? If a selector is loaded into a segment register, and the AVL bit is clear (0), the processor will signal an exception. > 2. TYPE is a 4 bit field which affects the segment operation. > Bit 0 of TYPE states if the selector has been accessed. > The CPU sets this bit in the descriptor. > True or False ? True. > And explain why true or false ? It could be useful to know wether or not the selector has been accessed, primarily for reasons of swap optimalization and debugging. --- BBBS/NT v3.33 How-D * Origin: Circle of Protection - +47 55961259 (2:211/37) --------------- FIDO MESSAGE AREA==> TOPIC: 145 ASSEMBLY LANG. Ref: E5B00017Date: 05/03/97 From: DAVID KIRSCHBAUM Time: 10:37am \/To: BARRY BLOCK (Read 2 times) Subj: Bosskey.Asm - sysreq.asm > Hello Flemming, DB>>> push CS DB>>> pop DS DB>>> mov SS,[_SS] ;restore them back again DB>>> mov SP,[_SP] DB>>> ... FS>> Are you sure this is safe? You are using the stack here. Not FS>> good when SS and SP are messed up, eh? :) FS>> There's nothing wrong with: FS>> mov ss,cs:[_ss] FS>> mov sp,cs:[_sp] > I don't see the problem. push,then pop. Stack is not altered, right? > However, I tried the tsr with Denis's change and it still crashes :) > More needs to be done. > Kind regards, Barry I'd disable interrupts before I messed with the stack pointers though. --- * Origin: Toad Hall (1:3634/2.4)