Nested counted loops¶

Generated workload catalogue · Execution contract

Purpose¶

Exercise two nested branch sites with repeated inner-loop exits.

Setup¶

Property	Value
Workload ID	`nested`
Selection	`--workload nested`
Generator	`legacyProgram() in RV32IWorkloads.cpp`
End condition	EBREAK after the loop

x1 = iterations; x3 starts at zero. Each outer iteration initializes x2=8.

N means the requested iteration count. Registers start from reset; the loop's x2 is not an ABI stack pointer.

Implementation¶

Run eight increments of x3, using x2 as the inner countdown; then decrement the outer x1 counter.

for (unsigned j = 0; j < 8; ++j) ++x3;
--x1;

This is a behavioral summary of the generated assembly, not the host program being timed.

Expected result¶

x3=8*N, x1=x2=0. For N=12, x3=96. Each inner BNE is taken seven times and then not taken.

The runner validates the executed architectural state with the instruction oracle. This legacy generator does not publish a separate expected_result checksum; the equations above describe its loop.

Measurement and running¶

Useful for predictor loop behavior; also includes arithmetic dependencies and outer-loop control. No call stack is involved.

build/rv32i_experiment --core advanced --workload nested \
  --iterations 2  \
  --fidelity behavioral --period 4000 \
  --allocation-limit-mib 512 --max-cycles 4096

The explicit period is for a bounded functional check, not a hardware-speed result. For comparisons, use the serial campaign driver, identical case inputs, and a certified per-target clock.

Source and tests¶

src/analysis/RV32IWorkloads.cpp: emitted instructions.
tests/checks/RV32IExperimentCLITest.py: legacy loops across all three cores.
tests/checks/RV32IBenchmarkCLITest.py and tests/checks/BenchmarkCampaignCLITest.py: shared execution/campaign checks.