Post by thread:Increment/Decrement "w"

Hi folks, Yesterday, I needed snippets of code which which will increment/decrement the "w" register without affecting any of the other registers in a 12C508. For Increment, I came up with: xorlw 0x0FF ; Get 1s Complement of Number addwf Reg, w ; w = Reg + (w^0x0FF) subwf Reg, w ; w = Reg + ((Reg + (w^0x0FF))^0x0FF) + 1 ; w = w + 1 and for Decrement: subwf Reg, w ; w = Reg + (2^0x0FF) + 1 xorlw 0x0FF ; Get 1s Complement of Result addwf Reg, w ; w = w - 1 "Reg" can be any Register in the PICMicro which cannot change during the course of the three instructions (in the low-end PICMicros, this is any port other than the timer and I/O ports). I've added these to my list of PICMicro "snippets" at: http://www.myke.com/PICMicro/snippet.htm and I will see if I can find a general case that can be used for "addlw" and "sublw" in the low-end devices. myke

Hi Myke. Nice to see ya again ;-) Do remember some discussion about tricks with ZERO_REG ? (with permanent zero value. John, Scott and other guys were involved...) It is some physically unimplemented location in memory and any operation read zero from it. This will allow you to reduce you code down to 2 instruction for every operation (I still not tested it yet but it looks workable) WBR Dmitry. Myke Predko wrote: {Quote hidden}> > Hi folks, > > Yesterday, I needed snippets of code which which will increment/decrement > the "w" register without affecting any of the other registers in a 12C508. > > For Increment, I came up with: > > xorlw 0x0FF ; Get 1s Complement of Number > addwf Reg, w ; w = Reg + (w^0x0FF) > subwf Reg, w ; w = Reg + ((Reg + (w^0x0FF))^0x0FF) + 1 subwf zero_reg,W ; -W xorlw 0x0FF ;-(-W)-1 = W-1 > and for Decrement: > > subwf Reg, w ; w = Reg + (2^0x0FF) + 1 > xorlw 0x0FF ; Get 1s Complement of Result > addwf Reg, w ; w = w - 1 xorlw 0x0FF ;-W -1 subwf zero_reg,W ;-(-W -1) = W+1 {Quote hidden}> "Reg" can be any Register in the PICMicro which cannot change during the > course of the three instructions (in the low-end PICMicros, this is any port > other than the timer and I/O ports). > > I've added these to my list of PICMicro "snippets" at: > > http://www.myke.com/PICMicro/snippet.htm > > and I will see if I can find a general case that can be used for "addlw" and > "sublw" in the low-end devices. > > myke

At 10:53 23/06/99 +1000, you wrote: >This is probably old news but I thought it was a good code snippet for >decrementing 16 bits. > >decf Lo >incfsz Lo,w >decf Hi Is it just me or is the above incorrect? Jim ________________________________________ Email: .....newfoundKILLspam@spam@pipeline.com.au http://www.new-elect.com WARP-13 SALE now on. $48USD with world delivery. MPLAB compatible PIC programmers and firmware upgrades for many programmers. ________________________________________

> At 10:53 23/06/99 +1000, you wrote: > >This is probably old news but I thought it was a good code snippet for > >decrementing 16 bits. > > > >decf Lo > >incfsz Lo,w > >decf Hi > > Is it just me or is the above incorrect? Yeah... incfsz really needs to be incfsnz (if only there were such an instruction). Better to use: movlw 1 subwf Lo btfsc STATUS,C decf Hi Unless anyone has a better idea? Cheers, Ben

At 08:59 PM 6/22/99 -0700, you wrote: >Yeah... incfsz really needs to be incfsnz (if only there were such an >instruction). >Better to use: > > movlw 1 > subwf Lo > btfsc STATUS,C > decf Hi > >Unless anyone has a better idea? Its not better, but an alternative: decf Lo decf Hi incfsz Lo,w incf Hi > >Cheers, >Ben > Sean | | Sean Breheny | Amateur Radio Callsign: KA3YXM | Electrical Engineering Student \--------------=---------------- Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 shb7KILLspamcornell.edu ICQ #: 3329174 ________________________________________________________ NetZero - We believe in a FREE Internet. Shouldn't you? Get your FREE Internet Access and Email at http://www.netzero.net/download/index.html

Hi, Ben Stragnell wrote: > movlw 1 > subwf Lo > btfsc STATUS,C > decf Hi > > Unless anyone has a better idea? dec16: movf lo,f skpnz decf hi,f decf lo,f Same number of cycles but does not touch W. regards, Reggie -- e-mail: .....rberdinKILLspam.....bigfoot.com ICQ#: 31651436 URL: http://www.bigfoot.com/~rberdin

Was it April fools day yesterday? Sorry about my little snippet, it obviously doesn't work. Here's some more to make up. Rotate a value without using carry. rlf RAM,w rlf RAM Test if a value is in a certain range W = value addlw 255 - Hi addlw (Hi - Lo) + 1 Carry is set if W is in range Lo - Hi -- Best regards Tony 'The Engine' - Design your own programmer. http://www.picnpoke.com Email EraseMEpicnpokespam_OUTTakeThisOuTcdi.com.au

At 09:06 AM 6/24/99 +1000, you wrote: >Was it April fools day yesterday? > >Sorry about my little snippet, it obviously doesn't work. > >Here's some more to make up. > >Rotate a value without using carry. > >rlf RAM,w >rlf RAM I'm afraid I don't get this one. Why isn't carry affected? In fact, if it weren't affected,the first line would do nothing to affect the second. Sean | | Sean Breheny | Amateur Radio Callsign: KA3YXM | Electrical Engineering Student \--------------=---------------- Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 shb7spam_OUTcornell.edu ICQ #: 3329174 ________________________________________________________ NetZero - We believe in a FREE Internet. Shouldn't you? Get your FREE Internet Access and Email at http://www.netzero.net/download/index.html

Sean Breheny wrote: >Rotate a value without using carry. > > >rlf RAM,w > >rlf RAM > > I'm afraid I don't get this one. Why isn't carry affected? In fact, if it > weren't affected,the first line would do nothing to affect the second. It does 'affect' the carry, but it doesn't 'use' it. In other words rotate the byte left once. -- Best regards Tony 'The Engine' - Design your own programmer. http://www.picnpoke.com Email @spam@picnpokeKILLspamcdi.com.au

Ok, I get it, I was a bit slow on the uptake.<G> The first line places the MSBit of RAM into Carry and then the second actually performs the rotate. Thanks, Sean At 01:24 PM 6/24/99 +1000, you wrote: {Quote hidden}>Sean Breheny wrote: > >>Rotate a value without using carry. >> >> >rlf RAM,w >> >rlf RAM >> >> I'm afraid I don't get this one. Why isn't carry affected? In fact, if it >> weren't affected,the first line would do nothing to affect the second. > >It does 'affect' the carry, but it doesn't 'use' it. >In other words rotate the byte left once. > >-- >Best regards > >Tony > >'The Engine' - Design your own programmer. > >http://www.picnpoke.com >Email KILLspampicnpokeKILLspamcdi.com.au > | | Sean Breheny | Amateur Radio Callsign: KA3YXM | Electrical Engineering Student \--------------=---------------- Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 RemoveMEshb7TakeThisOuTcornell.edu ICQ #: 3329174 ________________________________________________________ NetZero - We believe in a FREE Internet. Shouldn't you? Get your FREE Internet Access and Email at http://www.netzero.net/download/index.html

> Van: Sean Breheny <spamBeGoneshb7spamBeGoneCORNELL.EDU> > Aan: TakeThisOuTPICLISTEraseMEspam_OUTMITVMA.MIT.EDU > Onderwerp: Re: Increment/Decrement "w" > Datum: donderdag 24 juni 1999 4:34 > > At 09:06 AM 6/24/99 +1000, you wrote: > >Was it April fools day yesterday? > > > >Sorry about my little snippet, it obviously doesn't work. > > > >Here's some more to make up. > > > >Rotate a value without using carry. > > > >rlf RAM,w > >rlf RAM > > I'm afraid I don't get this one. Why isn't carry affected? In fact, if it > weren't affected,the first line would do nothing to affect the second. Sean, the normal Rotate-instructions are 9-bit (Carry into lowest bit, highest bit into Carry). This modification creates a 8-bit Rotate. (Highest bit into Lowest bit). Greetz, Rudy Wieser

Tony Nixon wrote: > > Sean Breheny wrote: > > >Rotate a value without using carry. > > > > >rlf RAM,w > > >rlf RAM > > > > I'm afraid I don't get this one. Why isn't carry affected? In fact, if it > > weren't affected,the first line would do nothing to affect the second. > > It does 'affect' the carry, but it doesn't 'use' it. > In other words rotate the byte left once. If you would like to rotate without affection the carry you can do it like that: ;right cycle shift rrf RAM,W rrf RAM,F addlw 0x80 ;restore carry ;left cycle shift rlf RAM,W rlf RAM,F movwf temp_cy rrf temp_cy,f ;restore carry Any ideas how to make second one shorter ? ;-) WBR Dmitry.

Hi Dmitry, Dmitry Kiryashov wrote: > ;left cycle shift > > rlf RAM,W > rlf RAM,F > movwf temp_cy > rrf temp_cy,f ;restore carry > > Any ideas how to make second one shorter ? ;-) rlf RAM,w rlf RAM,f movwf STATUS ; ;-) very BAD!!! This is _NOT_ recommended. This is just to give a solution to the problem to restore the carry using 3 cycles. This will affect other flags and bank settings depending on RAM contents. regards, Reggie -- e-mail: RemoveMErberdinTakeThisOuTbigfoot.com ICQ#: 31651436 URL: http://www.bigfoot.com/~rberdin

Tony Nixon <Tony.NixonEraseME.....ENG.MONASH.EDU.AU> wrote: > Test if a value is in a certain range > > W = value > > addlw 255 - Hi > addlw (Hi - Lo) + 1 > > Carry is set if W is in range Lo - Hi That can't possibly work! There aren't enough instructions! And I refuse to try it out to see if it works, since I know I'm right! Sorry, I couldn't resist. That's the sort of feedback I got when I posted this method here years ago. I normally use a macro for this, so that I don't have to remember the details. On the 12-bit core, there is no addlw instruction, so doing the equivalent requires a temporary register and three more instructions: movwf temp movlw 255-Hi addwf temp movlw (hi-lo)+1 addwf temp,w Cheers, Eric

Eric Smith wrote: > > Test if a value is in a certain range > > > > W = value > > > > addlw 255 - Hi > > addlw (Hi - Lo) + 1 > > > > Carry is set if W is in range Lo - Hi > > That can't possibly work! There aren't enough instructions! And I refuse > to try it out to see if it works, since I know I'm right! > Let's see :-) Lower = 50 Upper = 60 addlw 255 - Hi equates to addlw 195 addlw (Hi - Lo) + 1 equates to addlw 11 W = 50 ; lower limit addlw 195 ; w = 245 addlw 11 ; w = 0, C = 1, result = OK W = 55 ; in range addlw 195 ; w = 250 addlw 11 ; w = 5, C = 1, result = OK W = 60 ; upper limit addlw 195 ; w = 255 addlw 11 ; w = 11, C = 1, result = OK W = 49 ; under minimum addlw 195 ; w = 244 addlw 11 ; w = 255, C = 0, result = Not OK W = 61 ; over maximum addlw 195 ; w = 0 addlw 11 ; w = 11, C = 0, result = Not OK -- Best regards Tony 'The Engine' - Design your own programmer. http://www.picnpoke.com Email EraseMEpicnpokecdi.com.au

Tony Nixon wrote: {Quote hidden}> > Eric Smith wrote: > > > > Test if a value is in a certain range > > > > > > W = value > > > > > > addlw 255 - Hi > > > addlw (Hi - Lo) + 1 > > > > > > Carry is set if W is in range Lo - Hi > > > > That can't possibly work! There aren't enough instructions! And I refuse > > to try it out to see if it works, since I know I'm right! > > > > Let's see :-) > > Lower = 50 > Upper = 60 > > addlw 255 - Hi equates to addlw 195 > addlw (Hi - Lo) + 1 equates to addlw 11 Hi, Of what I understand is that Eric was the source of this snippet. He just emphasized the previous response he got when he first posted it here. BTW, the snippet works. regards, Reggie -- e-mail: RemoveMErberdinEraseMEEraseMEbigfoot.com ICQ#: 31651436 URL: http://www.bigfoot.com/~rberdin

> I think most (if not all of the confusion came from people not knowing > wether Hi and Lo were in RAM or constants at assembly time. > > ...and yes, it does work. It works for all values of Hi and Lo except 255 and 0. It is not the only sequence of the same length to do range checking. Of the several sequences that we have in our compilers each of them trade destructive testing for code size and bias the tested variable against either 0 or 255 allowing a single conditional branch. Walter Banks http://www.bytecraft.com

Hi Dennis, > No, just a comment from the price in Black Adder. Just kidding :-) > "addlw 255-Hi" > is this not a high level bit of code or does the PIC suddenly do // > processing? because if Hi is RAM or variable as you indicate the compiler > will take the absolute value of it PORTB for example, and thus the result > will always be the same. This value is not a variable or high level code. It is a constant and is calculated only at assembly time. The programmer desides on the values. > I am just saying that I think some code is > missing! And that the original poster is correct not enoguh instructions. Nothing missing, except as Eric Smith mentioned, the assembler must be able to truncate values over 255 or do it yourself. Just a bit of trivia... My soldering iron tip broke the other day, and while waiting for a replacement, I had to use a small 3mm desoldering tip. It works a treat for soldering IC pins. -- Best regards Tony 'The Engine' - Write your own programmer software. http://www.picnpoke.com Email RemoveMEpicnpokespam_OUTKILLspamcdi.com.au