Re: [PATCH 2/4] ASoC: tas2764: Deal with bogus initial temperature register value

[Mark Brown]

On Sat, Apr 25, 2026 at 10:44:04AM +1000, James Calligeros wrote:

> Return -ENODATA early if the temperature has not yet been sampled
> by the chip, and indicate a fault condition using HWMON_T_FAULT.

> +	case hwmon_temp_fault:
> +		*val = !!tas2764_read_die_temp(tas2764, val);
> +		return 0;

Should this pay more attention to the error code in case there was an
actual I/O error talking to the chip as opposed to a synthetic -ENODATA
caused by reading a 0 back?  If we can't talk to the chip we don't
specifically know if there was a temparature fault.

Attachment: signature.asc
Description: PGP signature