| fork | statement |
Dirty parts
| Syntax | A forkStatement is:
| ||||||||||||||||||
| Description | A fork activates (starts the concurrent execution of) a process declaration. If the process has parameters, a parenthesized list of expressions (expns) must follow the process' name (processId).
| ||||||||||||||||||
| Example | This program initiates (forks) two concurrent processes, one of which repeatedly outputs Hi and the other Ho. The resulting output is an unpredictable sequence of Hi's and Ho's, as greetings executes twice concurrently, one instance with its word set to Hi and the other with its word set to Ho.
process greetings ( word : string )
loop
put word
end loop
end greetings
fork greetings ( "Hi" )
fork greetings ( "Ho" )
| ||||||||||||||||||
| Details | See procedure declaration for details about parameters. The first optional reference in the fork statement must be a boolean variable reference. The fork sets this to true if the process is actually activated. If this fails to occur (probably because stack space could not be allocated), this reference is set to false. If the fork fails but this reference is omitted, an exception occurs. See exception handlers. The optional expn specifies the number of bytes for the process' stack; this overrides the optionally given stack size in the process declaration. The second optional reference must be a variable reference with the type addressint. See addressint. This variable is set to identify the process activation. This reference has the implementation-dependent meaning of locating the process' internal descriptor. In this explanation of the fork statement, we have up to this point ignored the possibility of processes exported from modules. If the process is being forked from outside of a module from which it has been exported, the syntax of the fork statement is:
fork moduleId . procedureId [ ( expn {, expn } ) ] …
In other words, the module's name and a dot must precede the process' name.
| ||||||||||||||||||