Multiple JMP, with One LBL

wanher3 · Mar 28, 2019

Hello,

Could you have several JMP instructions pointing to one LBL ?
E.g
Rung 1 has: JMP (X1)
Rung 5 has: LBL (X1)
Rung 8 has: JMP (X1)

mk42 · Mar 28, 2019

wanher3 said:
Hello,

Could you have several JMP instructions pointing to one LBL ?
E.g
Rung 1 has: JMP (X1)
Rung 5 has: LBL (X1)
Rung 8 has: JMP (X1)

Not sure what platform you're using, but I've definitely done it before.

wanher3 · Mar 28, 2019

mk42 said:
Not sure what platform you're using, but I've definitely done it before.

I'm using Rockwell Studio 5000.

daba · Mar 29, 2019

The answer is definitely YES - multiple JMPs to one LBL is perfectly possible.

Obviously it doesn't work the other way, you can't have multiple LBLs with the same name/reference.

wanher3 · Mar 29, 2019

daba said:
The answer is definitely YES - multiple JMPs to one LBL is perfectly possible.

Obviously it doesn't work the other way, you can't have multiple LBLs with the same name/reference.

Does it cause any issues with the scan time or how does it affect the scan time ?

daba · Mar 29, 2019

wanher3 said:
Does it cause any issues with the scan time or how does it affect the scan time ?

If you are jumping forwards in the code, it shortens the scan time.

If you are jumping backwards, it lengthens it.

However, if you JMP backwards continuously, (i.e. without a "get-out" clause), you will fault the processor because the scan watchdog timer will time-out.

For repetitive "loops", I prefer to use the FOR instruction, the code executed resides in a separate routine file, and the FOR must terminate at some point

wanher3 · Mar 29, 2019

daba said:
If you are jumping forwards in the code, it shortens the scan time.

If you are jumping backwards, it lengthens it.

However, if you JMP backwards continuously, (i.e. without a "get-out" clause), you will fault the processor because the scan watchdog timer will time-out.

How can i prevent that fault to happen ? and how many backward jumps will actually cause the fault ?

mk42 · Mar 29, 2019

wanher3 said:
How can i prevent that fault to happen ? and how many backward jumps will actually cause the fault ?

It depends entirely on how much code you're jumping (and probably on a CPU setting for max cycle time).

The difference between jumping backwards 5 times and 6 times probably won't make a difference, unless you're jumping all the way from the end to the beginning.

In most situations where i've seen looping make sense, there is some calculation/algorithm/sort that needs to be done on an array of data. For an array of 10, that's probably OK. For an array of 1000, you might not like the affect on your scan time.

If the array is too big, it usually makes sense to just sort 1 item (or maybe several items) each scan, until it is all done a few seconds later.

wanher3 · Mar 29, 2019

mk42 said:
It depends entirely on how much code you're jumping (and probably on a CPU setting for max cycle time).

The difference between jumping backwards 5 times and 6 times probably won't make a difference, unless you're jumping all the way from the end to the beginning.

In most situations where i've seen looping make sense, there is some calculation/algorithm/sort that needs to be done on an array of data. For an array of 10, that's probably OK. For an array of 1000, you might not like the affect on your scan time.

If the array is too big, it usually makes sense to just sort 1 item (or maybe several items) each scan, until it is all done a few seconds later.

Thank you.

Ron Beaufort · Mar 29, 2019

If the array is too big, it usually makes sense to just sort 1 item (or maybe several items) each scan, until it is all done a few seconds later.

take a look at the FAL (File Arithmetic Logical) instruction - and pay particular attention to the MODE setting ...

daba · Mar 30, 2019

Ron Beaufort said:
take a look at the FAL (File Arithmetic Logical) instruction - and pay particular attention to the MODE setting ...

FAL is a "do" instruction...

If you are "looking", its sister is the FSC (File Search Compare) instruction ...

Multiple JMP, with One LBL

wanher3

Member

mk42

Lifetime Supporting Member

wanher3

Member

daba

Lifetime Supporting Member

wanher3

Member

daba

Lifetime Supporting Member

wanher3

Member

mk42

Lifetime Supporting Member

wanher3

Member

Ron Beaufort

Lifetime Supporting Member

daba

Lifetime Supporting Member

Similar Topics