# 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.