Take the VHDL model of the arbiter provided and instantiate it in a SystemVerilog top module that also contains an interface, 4 instances of the provided request model, 4 test programs (one for each request device) and one test program that randomly generates resets for the system. Create random variables for the (1) number of cycles consumed before a device generates a request, (2) the number of cycles a request is kept high, (3) the number of cycles consumed before a reset is generated, and (4) the number of cycles the reset is asserted.
Use the following constraints:
(1) Range: 101 – 299 (2) Range: 1 – 100 (3) Range: 301 – 499 (4) Range: 4 – 9
Even though you are using different instantiations of a class, the method will give the same results. So, force the system to have different values by changing the number of calls to randomize in each test program.
test0 repeat(200)
p.randomize();
… test1 repeat(200)
p.randomize();
p.randomize();
…
test2 repeat(200)
p.randomize();
p.randomize();
p.randomize();
… test3 repeat(200)
p.randomize();
p.randomize();
p.randomize();
p.randomize();
…
Compile all your files with all forms of coverage enabled. Optimize using the command vopt top -o opttop +cover=sbecft from the transcript window, where top is your top-level module. Run 200 different test cases for each test program. Generate a coverage report for all types of coverage using Tools - Coverage Report - Text. Select Details and all kinds of Coverage.
Turn in your SystemVerilog source files and your coverage report file.
