Technical data

Organization Blocks
System Software for S7-300/400 System and Standard Functions - Volume 1/2
A5E00709327-01
1-19
1.10 Multicomputing Interrupt Organization Block (OB60)
Description
Using the multicomputing interrupt, you can make sure that the reaction of the
CPUs is synchronized to an event during multicomputing. In contrast to hardware
interrupts triggered by signal modules, the multicomputing interrupt can only be
output by CPUs.
Understanding the Operation of Multicomputing Interrupt OBs
A multicomputing interrupt is triggered by calling SFC35 "MP_ALM." During
multicomputing, this brings about a synchronized OB60 start on all CPUs of the
bus segment unless you have disabled OB60 (with SFC39 "DIS_IRT") or delayed it
(with SFC41 "DIS_AIRT"). If you have not loaded OB60 on a CPU, the CPU
returns to the last priority class before the interrupt and continues program
execution there. In single processor operation and when using segmented racks,
OB60 is only started on the CPU on which you called SFC35 "MP_ALM."
When your program calls SFC35 "MP_ALM," you supply a job ID. This ID is
transferred to all CPUs. This allows you to react to a specific event. If you program
OB60 differently on the various CPUs, this may result in different execution times
for the OB. In this case, the CPUs return to the interrupted priority class at different
times. If the next multicomputing interrupt is output by a CPU while another CPU is
still busy executing the OB60 of the previous multicomputing interrupt, then OB60
is not started either on the requesting or on any other CPU belonging to the bus
segment. This is illustrated in the following figure taking the example of two CPUs.
You are informed of the outcome by the function value of the called SFC35.
CPU 1:
CPU 2:
Cycle
OB60 execution
OB60 execution OB60 execution
Cycle
OB60 execution
Cycle
Cycle
"CALL SFC35"
"CALL SFC35" "CALL SFC35"