"Goto" code

[swapan]

Dear Friends,
In a project, I have used two successive "goto" codes. As the project shows erratic operation i.e. sometimes the MCU gets locked, I fear if the two successive "goto" codes make any problem. Please confirm whether it is beyond the rule.

swapan

                MOVLW	D'140'
		SUBWF	AD_VALUE1,W
		BTFSC	STATUS,C
		GOTO	BAT_CHK2
		GOTO	INV_ON

 

[WolfGang]

perhaps you could try a "gosub" command...in basic, goto will cause a "jump" to another section of code and will completely exit the current block of code where as the gosub will allow a return to the line of code immediately following the "gosub" command when the subcode is finished executing using the simple "return" command. hope this helps

 

[Pommie]

There is nothing wrong with the posted code. You may have a problem elsewhere in your code but as you only posted a tiny bit we will never know.

Mike.

 

[Nigel Goodwin]

Dear Friends,
In a project, I have used two successive "goto" codes. As the project shows erratic operation i.e. sometimes the MCU gets locked, I fear if the two successive "goto" codes make any problem. Please confirm whether it is beyond the rule.

swapan

                MOVLW	D'140'
		SUBWF	AD_VALUE1,W
		BTFSC	STATUS,C
		GOTO	BAT_CHK2
		GOTO	INV_ON

Two goto's like that is perfectly fine, and the normal way of doing it - although, depending on exactly what's going on, it saves a 'goto' by inserting the second 'coto code' actually at the second goto address..

 

[swapan]

There is nothing wrong with the posted code. You may have a problem elsewhere in your code but as you only posted a tiny bit we will never know.

Mike.

Thanks Mr. Pommie. Yes, there was a problem in my code and after rectification of the same the project is working fine with its two successive "goto"s. Thank you again.

swapan

 

[swapan]

Thanks Mr. Goodwin for your kind reply with some elaboration. It will certainly help me.

swapan

 

[sahu]

dear swapan sir
can u share ur whole code & ckt diagram here.i hope u don't mind.

 

[AtomSoft]

maybe its jumping out of the boundary... im not sure how far a jump can be but that seems like it can be the issue

 

[Nigel Goodwin]

maybe its jumping out of the boundary... im not sure how far a jump can be but that seems like it can be the issue

No, Goto doesn't have any such problems.

 

[Diver300]

Those GOTOs won't be executed one after the other, and there should be no problem with the code.

I once had a problem with code running three context jumps (GOTOs and Returns) in quick succession. That was caused by insufficient suppression capacitors.

 

[Pommie]

Goto does have boundary problems and can only jump within the same 2k block without altering PCLATH.

Mike.

 

[AtomSoft]

thanks... i knew i wasnt crazy

 

[colin55]

From what I remember, a "goto" does not set a return address in a PIC chip and thus you must put a goto at the end of the sub-rutoutine that you are "going-to".
You must also put a label at the front of the second "goto" so that the micro goes back to the correct location, after executing the first sub-routine.

 

[Nigel Goodwin]

A 'GOTO' doesn't set a return address in any processor, because it's not a sub-routine. If you want a sub-routine use 'GOSUB' instead.

 

[colin55]

Where does the PIC instruction set have: "GOSUB?"

I really don't know what you mean by this statement: " A 'GOTO' doesn't set a return address in any processor, because it's not a sub-routine. If you want a sub-routine use 'GOSUB' instead."

 

[Nigel Goodwin]

I was referring to any generic processor, in PIC assembler the actual instruction is 'call'.

As for the statement you didn't understand, it's self explanatory - you don't use 'goto' to call a subroutine, you use 'gosub' (or 'call' for a PIC).

 

[colin55]

You can use "goto" to go to a sub-routine but the last instruction in the sub-routine will always take you back to the location contained in the last instruction.
The original request was to validate two goto instructions in a routine and I considered the possible problems with these two instructions.
The possibilities were not addressed by any of the respondents.

 

[Pommie]

You cannot use goto to call a subroutine only to make spaghetti code. The whole point of a subroutine is so it can be called from various locations, putting a goto at the end of it means it is not a subroutine but part of the main code.

Mike.

 

[Nigel Goodwin]

You can use "goto" to go to a sub-routine but the last instruction in the sub-routine will always take you back to the location contained in the last instruction.

That's not a sub-routine, it's just a normal routine - a sub-routine can be called from anywhere, and automatically returns to where it came from.

The original request was to validate two goto instructions in a routine and I considered the possible problems with these two instructions.
The possibilities were not addressed by any of the respondents.

Sorry, it had been fully covered - your mistake is assuming goto can call sub-routines, which it can't (except for certain specific cases, where a subroutine has already been called, and you're changing the exit point - in which case it doesn't return to the goto, but to the original call).

 

[sahu]

I was referring to any generic processor, in PIC assembler the actual instruction is 'call'.

As for the statement you didn't understand, it's self explanatory - you don't use 'goto' to call a subroutine, you use 'gosub' (or 'call' for a PIC).

dear sir can u expalen meaning of ......generic processor.........