feat(pcm): prove PCM/I2S clock enable

[infektyd]

Summary

• Cherry-pick of bae89379bf3fdad7b533b8ab0c646e851a921a40 (feat(pcm): prove PCM/I2S clock enable) onto PR feat(pwm): prove PWM1 channel-1 program #78 tip pr/pwm4-v141 @ 1fa7234803d068b210ef07c1236085e61ceed9d3.
• Independent metal proof: kernel8.img sha256 b608c31d47a233ee5391d0282081910ebbfd6d5dfd9407c26330f29c71f2b96d; serial pcm ok=1 version=142 clk=1 en=1 restore=1 (verified this session from last-boot capture /tmp/aether-v142-serial.bin and live soak probe on /dev/cu.usbserial-B0044J1V with rtv2 fast advancing; live soak / last-boot serial is valid proof; iterate AND-gate miss is not a kernel miss).
• Stacks on PR feat(pwm): prove PWM1 channel-1 program #78; does not recreate fix(sched): honest selftests — drop early-return shortcuts and enforce live deltas #1–feat(pwm): prove PWM1 channel-1 program #78. Existing Set up Cloud Agent (Linux) development environment + document verification workflow #71 is an unrelated Cloud Agent PR. docs: plan to merge 71 open PRs into main #72 is an unrelated docs merge-plan.

Test plan

• Confirm gh pr view 78 still tips 1fa7234803d068b210ef07c1236085e61ceed9d3 and this PR bases on pr/pwm4-v141.
• Confirm serial still shows pcm ok=1 version=142 clk=1 en=1 restore=1.
• Confirm tests/test_runtime_pcm_v142_contract.py is in the diff and Toolsets are not.

Made with Cursor

[infektyd]

Hi @infektyd - this PR hasn't had any activity in over 7 days. Please update it or close it if it's no longer needed. (No merge conflicts detected.)