process-run

Name

process-run - schedule processes on a single CPU

Synopsis

process-run -l spec
process-run -l spec -c

Description

Process-run simulates scheduling of processes on a single CPU.

Arguments

• -l spec: a comma separated list of process specifications of the form N:prob, where N is the number of instructions, and prob is the probability of getting a CPU instruction else IO. The probability is an integer in the range[0, 100].

• -c: run simulation.

Example

Consider processes spec 3:50,5:100. It creates two processes:

• The first one has 3 instructions with 50% probability of getting a CPU instruction.
• The second one has 5 CPU instructions.

Simulate run on the CPU:

% process-run -l 3:50,5:100 -c
Process 0
  cpu
  io
  iod
  cpu
Process 1
  cpu
  cpu
  cpu
  cpu
  cpu
Clock   PID:0   PID:1
 1      run:cpu rdy
 2      blk:io  run:cpu
 3      blk:io  run:cpu
 4      blk:io  run:cpu
 5      rdy:iod run:cpu
 6      rdy:iod run:cpu
 7      run:cpu ok
 8      ok      ok

Note that the first process gets CPU,IO,CPU instructions. The iod means IO finished and the process is in ready state, waiting for the scheduler to pick it up.

The output sample reads:

• The first process runs for one cycle (clk 1) while the second process waits in the ready state.
• The first process blocks on the IO (clk 2) and remains in that state for the duration of IO (the cost is 3 cycles), while the second PID executes CPU instructions (clk 2-6).
• The first process is unblocked on the IO (clk 5) and remains in the ready state (clk 5-6) until the second process finishes.
• The second process completes and the remaining first process's CPU instruction runs (clk 7)
• All processes complete at the end (clk 8)