Is this possible?

GC4130

I have some code in SX/B that works very well for a particular sensor. I need to address multiple sensors of the same type and I am trying to figure out a way to make the original code more dynamic, all I need to do is change the I/O pin definition. Any help would be greatly appreciated. Thanks

'example

A var rb.1
B·var rb.2
temp var·byte

Test sub

Test:
...............
......
return

main:
··· temp=A
··· Test
··· temp=B
··· Test
goto main


·

Bean

George,
SX/B does not have anything like PIN variables. That is because the compiler "hard codes" the pin into the assembly instructions. A different pin requires different instructions.
Depending on what you are doing in your TEST routine you might be able to pass a bit mask to the routine.
Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012


Product web site: www.sxvm.com

"It's not getting what you want, it's wanting what you've got."
·

GC4130

Bean,

Could you explain a "bit mask rountine". My "Test" rountine is not complicated at all, it just allows me to measure the pulse width from an accelerometer. Problem is I have 4 of them and I want to·save code space by dynamically changing·the pin through the temp variable. If this is not possible, I guess I just have to have·4 individual "Test" rountines for·my 4 sensors with the specific pin designations in each.

thanks

George

Bean

George,
Can you post your program ? Or at least the "test" subroutine. It's easier to show you the code, than to explain how to do it.
Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012


Product web site: www.sxvm.com

"It's not getting what you want, it's wanting what you've got."
·

KenM

George,
Bit masking is essentially performing an AND function to determine the state of one bit.

check  ds 1         ;define a one byte space for status
checkm ds 1         ;define a one byte space for memory of check
 
 
 
 
 
;later in your code
 
and    check,#%00000001
 
;the value of check now indicates status of bit0
 
mov    check, #checkm
and    check,#%00000010
;the value of check now indicates status of bit1

There are other ways to address a single bit of a byte

Could you explain a "bit mask rountine".

GC4130

The code I have has alot of irrelavant information for this particular problem so I will try to explain with an abreviated version.

"This is what I would like to have"

A var rb.1
B·var rb.2
temp var·byte

Test sub

Test:
···· if temp=1 then check1
···· if temp=0 then check2
·····check1:
········· ...........
···· return
···· check2:
··········...........
···· return

main:
··· temp=A
··· Test
··· temp=B
··· Test
goto main






"This is what I Think I will be forced to do"

A var rb.1
B·var rb.2


Test1 sub
Test2 sub

Test1:
···· if A=1 then check1
···· if A=0 then check2
·····check1:
········· ...........
···· return
···· check2:
··········...........
···· return

Test2:
···· if B=1 then check3
···· if B=0 then check4
·····check3:
········· ...........
···· return
···· check4:
··········...........
·····return

main:
··· Test1
··· Test2
goto main

Bean

George,
I need to know what is inside the "check" routine.
Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012


Product web site: www.sxvm.com

"It's not getting what you want, it's wanting what you've got."
·

GC4130

Does it really matter what is inside check?, my problem is defined outside of this function Do you think there is some way I can rewrite my code to be more generalized like what I first posted.

"This is what I would like to have"

A var rb.1
B·var rb.2
temp var·byte

Test sub

Test:
···· if temp=1 then check1
···· if temp=0 then check2
·····check1:
········· ...........
···· return
···· check2:
··········...........
···· return

main:
··· temp=A
··· Test
··· temp=B
··· Test
goto main

Bean

George,
To answer your original question with what you have provided...

Yes it is possible.

Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012


Product web site: www.sxvm.com

"It's not getting what you want, it's wanting what you've got."
·

Paul Baker

Here is some code I banged out to show it is possible, I haven't tested it so there maybe bugs or syntax errors. It consists of two subroutines Fetch2bit and Store2bit, the first takes a single argument whose value is 0-3 (4 two bit values can be stored in each byte) and returns the value located at the specified position (into __Param1), the second takes two arguments the first is the storage position (0-3) and the second is the value to store (0-3). These routines do not perform any check on the bounds of the arguments so if you·use values outside of 0-3·you will get·unpredicitable·behaviour. You could also modify this to perfom on arrays if you need more than 4 two bit variables, but I leave that to you.

TwoBitMask CON %00000011
 
ByteVar  VAR byte
TempVar  VAR byte
TwoBit  VAR byte
ShiftMask VAR byte
 
Fetch2bit sub 1
Store2bit sub 2
 
Fetch2bit:
  TempVar = __Param1              'store 2 bit var position
  TwoBit = ByteVar                'make a copy of data
  DO WHILE TempVar > 0
    TwoBit >> 2                   'shift data down
    DEC TempVar
  LOOP
  __Param1 = TempVar & TwoBitMask 'store two bits into return value
  RETURN
 
Store2bit:
  TempVar = __Param1              'store 2 bit var position
  TwoBit = __Param2               'store 2 bit value
  ShiftMask = !TwoBitMask         'create masker (to clear both bits)
  DO WHILE TempVar > 0
    TwoBit << 2                   'shift value and mask to next position
    ShiftMask << 2
    DEC TempVar
  LOOP
  ByteVar = ByteVar & ShiftMask   'zero out both bits of variable
  ByteVar = ByteVar | TwoBit      'store bits
  RETURN

Post Edited (Paul Baker) : 5/31/2005 5:53:15 PM GMT

Bean

Paul,
You can use variables for the shift count.
Instead of your loop in fetch you can do:

TempVar = __PARAM1
TempVar = TempVar·* 2
TwoBit = ByteVar >> TempVar

Bean.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"SX-Video·Module" Now available from Parallax for only $28.95 http://www.parallax.com/detail.asp?product_id=30012


Product web site: www.sxvm.com

"It's not getting what you want, it's wanting what you've got."


Post Edited (Bean (Hitt Consulting)) : 5/31/2005 6:47:01 PM GMT

Paul Baker

Thanks Bean, my mind still thinks in assembly, plus the docs only being in the distro and me being at work, I was winging the code.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

GC4130

Maybe this is because I am a newbie to the SX but I can't see how I could implement this bit masking routine to the Original code I posted?? How does this relate to changing the pin designations? ·

Paul Baker

My code was for illustration purposes, I couched it in general subroutines that it could be used generically with a program. To distill it even further you do the following steps:

To read:
Figure out which position you need (0-3) (we'll call it P)
make a copy of the variable you want the two bits from (this can be a port register)
Right shift the copy 2*P times.
AND the result with the constant 3
what's left is the two bits you wanted in the lowest two bits of your copy of the data.

To write:
Figure out which position you need (0-3) (we'll call it P)
Make a copy of the bits that you want store (or use the original if you do not care if its destroyed)
Create a variable and store $FC into it (this is your mask which is the inverse of the above mask)
Left shift the data to be store and the mask 2*P times.
AND the mask with the location the data is to be stored
OR the data to be stored with the location it is to be stored.
Now those two bits are inserted into the location without messing with the other bits.

Either create a subroutine as I did above or just embed it into your code.

GC4130

Paul,

Thanks for the explaination I think I get it now. I'll try to implement it in my code.

George



·