Speaker route on the board, 2026-10-01

Source: docs/evidence/boot-verification/2026-10-01/boot-postcandidate.txt, the console transcript from the 2026-10-01 boot verification. The command date -u in it prints Thu Oct 1 02:51:01 PM UTC 2026. Evidence class: on-board console transcript (commands sent over the console, no physical observation). This file cites that capture for this change’s tasks and does not re-run anything.

Task 2.2b: the patched module binds (done)

The boot-time dmesg lines show the driver bound with the new default, and the control exists and reads off:

[    5.327187] canaan-k230-snd-inno sound: External I2S output disabled by default
numid=5,iface=MIXER,name='External I2S Output Switch'
  ; type=BOOLEAN,access=rw------,values=1
  : values=off

/proc/asound/cards and aplay -l list only K230I2SINNO (the Inno codec, 9140e000.inno_codec). The message is printed at 5.3 s into a boot, so the patched kernel was the booted one.

Task 4.2b: k230-speaker-test status (still open)

k230-speaker-test status ran and exited 0 (K230_OP_DONE_… 0). Two of its sections failed inside the script, though:

• amixer controls printed amixer: Control default open error: Host is down;
• the SDMODE line check printed gpioinfo: cannot find line 'gpiochip1'. The board’s libgpiod is 2.2.4, where gpioinfo takes line names and needs -c gpiochip1 to select a chip. This is the version mismatch that design.md flagged as UNVERIFIED.

So the status subcommand runs, but its GPIO and control checks do not work against the pinned libgpiod. 4.2b stays open until the script uses the 2.x syntax and a board run shows the SDMODE line state.