Re: [PATCH v7 8/8] thermal/drivers/loongson2: Add Loongson-2K0300 support

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] thermal: loongson2: missing delay before reading ADC after initialization
--

commit 97ab3025b749644e2a07310bc78907ed2d3b519c
Author: Binbin Zhou <zhoubinbin@loongson.cn>
Subject: thermal/drivers/loongson2: Add Loongson-2K0300 support

This commit adds support for the Loongson-2K0300 thermal sensor, configuring
it with calibration data read from the syscon CHIP ID registers. It also
introduces a fallback mechanism to use an old fuse formula if the initial
calibrated temperature reading is outside of a valid range.

> diff --git a/drivers/thermal/loongson2_thermal.c b/drivers/thermal/loongson2_thermal.c
> index dbfb03aa08338..1db61509dbf37 100644
> --- a/drivers/thermal/loongson2_thermal.c
> +++ b/drivers/thermal/loongson2_thermal.c
[ ... ]
> @@ -83,6 +118,101 @@ static int loongson2_thermal_set_trips(struct thermal_zone_device *tz, int low,
[ ... ]
> +static int ls2k0300_thermal_hw_init(struct platform_device *pdev,
> +				    struct loongson2_thermal_data *data)
> +{
> +	struct device *dev = &pdev->dev;
> +	int ret;
> +
> +	data->regmap_cfg = syscon_regmap_lookup_by_phandle(dev->of_node, "loongson,chipid");
> +	if (IS_ERR(data->regmap_cfg))
> +		return PTR_ERR(data->regmap_cfg);
> +
> +	ret = ls2k0300_thermal_read_calib(data);
> +	if (ret)
> +		return dev_err_probe(dev, ret, "failed to read calibration data");
> +
> +	writel(LS2K0300_THSENS_INIT_MASK, data->ctrl_reg + LS2K0300_THSENS_CFG_REG);
> +
> +	ls2k0300_thermal_detect_old_fuse(dev, data);

[Severity: Medium]
Will this immediate ADC read return uninitialized or stale data?

The sensor is initialized by writing LS2K0300_THSENS_INIT_MASK, and on the
very next instruction ls2k0300_thermal_detect_old_fuse() reads the ADC
output without polling a readiness bit or applying a small delay.

If the hardware requires conversion latency, reading a 0 or stale value here
would produce a deeply negative temperature that spuriously satisfies the
temp < LOONGSON2_LOWEST_VALID_TEMP condition. Does this risk incorrectly
discarding valid calibration data and permanently enabling the uncalibrated
fallback logic for the boot session?

> +
> +	return 0;
> +}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1791512067.git.zhoubinbin@loongson.cn?part=8