Pattern Generation

Patterns are Python sources evaluated in an Origen application context. They operate on the active DUT and tester and are rendered by the tester selected in the target set.

Basic Pattern

Use a Pattern context to delimit generation:

with Pattern(pin_header="all") as pat:
    origen.tester.set_timeset("functional")
    origen.tester.cc("Reset sequence")
    origen.dut.pin("reset").drive(0)
    origen.tester.repeat(5)
    origen.dut.pin("reset").drive(1)
    origen.tester.cycle()

Applications may use the injected dut() and tester() helpers or the explicit origen.dut and origen.tester objects. Prefer one style consistently within a project.

Pin State And Cycles

Driving or verifying a pin changes modeled state. cycle commits one cycle; repeat commits repeated cycles. Set a timeset before cycling whenever the target does not establish a suitable default.

Register And Protocol Operations

Register writes and verifies call block controllers, which translate them into protocol pin operations. This keeps pattern source at the register-intent level:

reg = origen.dut.reg("control")
reg.set_data(0x5)
reg.write()

Captures And Overlays

Capture marks cycles whose pin data should be returned by the tester. Overlay marks data that a downstream tool may substitute. Support and syntax vary by backend; keep approved output for patterns using either feature.

Generating

Invoke generation with the application command and active targets:

$ origen generate example/patterns/reset.py \
    --target dut/eagle --target tester/smt7

Run origen generate --help for output and reference-directory options.

Regression Strategy

Test pattern source at two levels: inspect model/AST behavior where practical, and compare representative generated files against approved references. Avoid approving large diffs without identifying the model or renderer change that caused them.

• Pattern API
  • Common Operations
  • Comments