S6: Extend and reproduce
Question: What evidence makes a small extension reviewable?
Time: 30 minutes. Prerequisites: completed setup. See qualification and limitations before running.
Mechanism
- Choose an unused teaching opcode at the frozen source revision.
- Return four little-endian fields through the existing PRP helper.
- Reject invalid lengths and reserved fields before transferring data.
Predict
What should fail before the command exists, and what must still fail afterward?
Write down a predicted state or counter change before executing the commands.
Run
Host terminal, unless a command is explicitly labelled Guest. Source and
export the identity fields from connection.env. These labs overwrite the
explicit disposable namespace. Use a fresh VM for each independent arm.
# Guest, before and after rebuilding the separate checkout:
sudo python3 check-command.py /dev/nvme0 --mode 2
# Host, teaching checkout only:
git apply /path/to/solution.patch
ninja -C build qemu-system-x86_64
Check your evidence
- Preserve the failing baseline and passing patched receipt.
- Record both source diffs and binary hashes.
- Run the regression after restarting the lab VM with the new binary.
The host runner retrieves raw receipts and preserves the guest exit status. Keep unsuccessful attempts; do not replace them with a later successful log.
Interpret
Can another learner reproduce your claim from the bundle alone?
Explain the mechanism and the measurement boundary. A throughput number by itself is not evidence that the intended feature was exercised.
Recover
- Test unexpectedly passes before patch: inspect the actual binary and source.
- Unknown ABI: fail decoding instead of guessing the layout.
- Keep the emulator qtest and Linux guest regression as separate evidence.
Save the receipt before resetting the VM.
Materials and source
Slides, scripts and worksheet.
Videos are not yet recorded. The frozen public source is
1918eb9795ba.
The FEMU Manual explains the implementation and supported properties.