Still silent after the pinctrl dedup. Three diagnostic additions so the next boot tells us exactly where the failure actually is: - IKCONFIG=y + IKCONFIG_PROC=y so /proc/config.gz is queryable. init greps it for PINCTRL_JASPERLAKE, GPIOLIB, GPIO_ACPI, I2C_DESIGNWARE, SND_SOC_MAX98357A, SND_SOC_INTEL_SOF_RT5682 — that answers "is the driver actually compiled in" with zero ambiguity. - init dump expanded: kallsyms greps for jsl_pinctrl / max98357 / jasperlake_ prefixes; GPIO chip listing; scan of /sys/bus/platform for MX98360A and friends to see whether the driver bound and what resources came through; amixer contents; DAPM Speaker widget state. - ac-native stdout/stderr now redirect straight to /mnt/ac-native-{stdout,stderr}.log instead of /tmp. Previously the ALSA init trace (which device opened, mixer enumeration, volume settings) only reached USB on clean exit — useless when the user yanks the stick to reflash. Vfat buffers the writes but the file is there and readable after reboot. Still need to actually fix the audio — this commit just makes the next iteration one-shot-diagnosable instead of building + flashing + guessing. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>