Re: [PATCH v6 4/4] thermal: Add Tegra SOCTHERM thermal management driver

[Mikko Perttunen <mikko.perttunen-/1wQRMveznE@public.gmane.org>]

On 09/26/2014 02:45 PM, Thierry Reding wrote:
> On Fri, Sep 26, 2014 at 12:43:13PM +0300, Mikko Perttunen wrote:
>> From: Mikko Perttunen <mperttunen-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
>>
>> This adds support for the Tegra SOCTHERM thermal sensing and management
>> system found in the Tegra124 system-on-chip. This initial driver supports
>> temperature polling for four thermal zones.
>>
>> Signed-off-by: Mikko Perttunen <mperttunen-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
>> ---
>> v6: fixed sparse warning for wrong order of "__iomem *"
>
> Sorry for jumping in so late. This looks generally good, but I have a
> couple of comments of a stylistic nature. I haven't closely followed
> earlier rounds of the series, so please let me know if I'm bringing up
> issues that have already been discussed.
>
>> diff --git a/drivers/thermal/tegra_soctherm.c b/drivers/thermal/tegra_soctherm.c
> [...]
>> @@ -0,0 +1,471 @@
>> +/*
>> + * drivers/thermal/tegra_soctherm.c
>
> This line is not really necessary and likely to become stale if files
> are moved around.

Removed.

>
>> +#include <linux/clk.h>
>> +#include <linux/err.h>
>> +#include <linux/io.h>
>> +#include <linux/module.h>
>> +#include <linux/of.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/reset.h>
>> +#include <linux/thermal.h>
>> +#include <linux/delay.h>
>> +#include <linux/interrupt.h>
>> +#include <linux/bitops.h>
>> +#include <soc/tegra/fuse.h>
>
> These should be sorted alphabetically and I prefer separating them into
> sections, so that the last of the linux/* includes is separated from the
> soc/* includes by a blank line.

Fixed.

>
>> +
>> +#define SENSOR_CONFIG0				0
>> +#define		SENSOR_CONFIG0_STOP		BIT(0)
>> +#define		SENSOR_CONFIG0_TALL_SHIFT	8
>> +#define		SENSOR_CONFIG0_TCALC_OVER	BIT(4)
>> +#define		SENSOR_CONFIG0_OVER		BIT(3)
>> +#define		SENSOR_CONFIG0_CPTR_OVER	BIT(2)
>> +#define SENSOR_CONFIG1				4
>> +#define		SENSOR_CONFIG1_TSAMPLE_SHIFT	0
>> +#define		SENSOR_CONFIG1_TIDDQ_EN_SHIFT	15
>> +#define		SENSOR_CONFIG1_TEN_COUNT_SHIFT	24
>> +#define		SENSOR_CONFIG1_TEMP_ENABLE	BIT(31)
>> +#define SENSOR_CONFIG2				8
>> +#define		SENSOR_CONFIG2_THERMA_SHIFT	16
>> +#define		SENSOR_CONFIG2_THERMB_SHIFT	0
>> +
>> +#define SENSOR_PDIV				0x1c0
>> +#define		SENSOR_PDIV_T124		0x8888
>> +#define SENSOR_HOTSPOT_OFF			0x1c4
>> +#define		SENSOR_HOTSPOT_OFF_T124		0x00060600
>> +#define SENSOR_TEMP1				0x1c8
>> +#define SENSOR_TEMP2				0x1cc
>> +
>> +#define SENSOR_TEMP_MASK			0xffff
>> +#define READBACK_VALUE_MASK			0xff00
>> +#define READBACK_VALUE_SHIFT			8
>> +#define READBACK_ADD_HALF			BIT(7)
>> +#define READBACK_NEGATE				BIT(1)
>
> These don't seem to be aligned in the same way as the field definitions
> for the other registers. Was that on purpose? I also think using two
> tabs to indent field definitions is somewhat excessive. Another common
> pattern I've seen used is:
>
> 	#define REGISTER_1_NAME 0x000
> 	#define  REGISTER_1_FIELD_MASK (0xf << 4)
>
> 	#define REGISTER_2_NAME 0x004
> 	#define  REGISTER_2_FIELD_MASK (0xf << 8)

I agree, the indentation was a bit excessive. The readback defines 
weren't indented since they weren't really nested under anything. I 
removed the indentation and added some spacing; I think it looks a bit 
better now.

>
>> +#define FUSE_TSENSOR8_CALIB			0x180
>> +#define FUSE_SPARE_REALIGNMENT_REG_0		0x1fc
>> +
>> +#define FUSE_TSENSOR_CALIB_CP_TS_BASE_MASK	0x1fff
>> +#define FUSE_TSENSOR_CALIB_FT_TS_BASE_MASK	(0x1fff << 13)
>> +#define FUSE_TSENSOR_CALIB_FT_TS_BASE_SHIFT	13
>> +
>> +#define FUSE_TSENSOR8_CALIB_CP_TS_BASE_MASK	0x3ff
>> +#define FUSE_TSENSOR8_CALIB_FT_TS_BASE_MASK	(0x7ff << 10)
>> +#define FUSE_TSENSOR8_CALIB_FT_TS_BASE_SHIFT	10
>> +
>> +#define FUSE_SPARE_REALIGNMENT_REG_SHIFT_CP_MASK 0x3f
>> +#define FUSE_SPARE_REALIGNMENT_REG_SHIFT_FT_MASK (0x1f << 21)
>> +#define FUSE_SPARE_REALIGNMENT_REG_SHIFT_FT_SHIFT 21
>
> These are also inconsistent with the above register definitions.

Fixed.

>
>> +
>> +#define NOMINAL_CALIB_FT_T124			105
>> +#define NOMINAL_CALIB_CP_T124			25
>
> What do FT and CP stand for? Given the _T124 suffix they would seem
> likely to change on future SoC generations. Perhaps they should be moved
> into a struct tegra_soctherm_soc or similar? The array of t124_tsensors
> would be another candidate to put into such a structure.

No idea. I haven't found any expansions for these in internal or 
external documentation/code. These are the names used by the downstream 
driver and some documentation, so I used them.

>
> That's something that could be deferred until support is added for a
> next generation.

I would prefer to do that.

>
>> +struct tegra_tsensor_configuration {
>> +	u32 tall, tsample, tiddq_en, ten_count;
>> +	u32 pdiv, tsample_ate, pdiv_ate;
>
> What's the significance of the "ate" suffix?

Again, no idea.

>
>> +};
>> +
>> +struct tegra_tsensor {
>> +	u32 base;
>> +	u32 calib_fuse_offset;
>> +	/* Correction values used to modify values read from calibration fuses */
>> +	s32 fuse_corr_alpha, fuse_corr_beta;
>> +};
>
> Both this structure and the one above use inconsistent grouping of
> fields (or at least I can't make out any grouping). Perhaps it would be
> more consistent to use one line per field, or group all fields of a data
> type in a single line?

Put fields of same type on same line.

>
>> +struct tegra_thermctl_zone {
>> +	void __iomem *temp_reg;
>> +	int temp_shift;
>
> Should this be unsigned? Also this is a thermal zone, so the temp_
> prefix doesn't add any context.

Fixed.

>
>> +static int calculate_shared_calibration(struct tsensor_shared_calibration *r)
>> +{
>> +	u32 val;
>> +	u32 shifted_cp, shifted_ft;
>
> The above two lines could be collapsed into one.

Fixed.

>
>> +static int calculate_tsensor_calibration(
>> +	const struct tegra_tsensor *sensor,
>> +	struct tsensor_shared_calibration shared,
>> +	u32 *calib
>> +)
>
> I think a more idiomatic way to write this would be:
>
> static int
> calculate_tsensor_calibration(const struct tegra_tsensor *sensor,
> 			      struct tsensor_shared_calibration shared,
> 			      u32 *calib)

If I do that, it will go over the 80 character limit by quite a few 
characters, which is why I didn't use that style. Personally I'm fine 
with either style.

>
> While at it, perhaps make shared a const * instead of passing it in by
> value?

That is possible, but I'm not sure what the difference would be. Is 
there a style rule forbidding by-value compound types? (Also if I change 
the style, it would go over 80 characters by even more.)

>
>> +{
>> +	u32 val;
>> +	s32 actual_tsensor_ft, actual_tsensor_cp;
>> +	s32 delta_sens, delta_temp;
>> +	s32 mult, div;
>> +	s16 therma, thermb;
>> +	int err;
>> +
>> +	err = tegra_fuse_readl(sensor->calib_fuse_offset, &val);
>> +	if (err)
>> +		return err;
>> +
>> +	actual_tsensor_cp = (shared.base_cp * 64) + sign_extend32(val, 12);
>> +	val = (val & FUSE_TSENSOR_CALIB_FT_TS_BASE_MASK)
>> +		>> FUSE_TSENSOR_CALIB_FT_TS_BASE_SHIFT;
>
> I personally prefer the operator at the end of the previous line, but I
> guess that's mostly a matter of taste, so I'll leave it up to Eduardo.
>
>> +	actual_tsensor_ft = (shared.base_ft * 32) + sign_extend32(val, 12);
>> +
>> +	delta_sens = actual_tsensor_ft - actual_tsensor_cp;
>> +	delta_temp = shared.actual_temp_ft - shared.actual_temp_cp;
>> +
>> +	mult = t124_tsensor_config.pdiv * t124_tsensor_config.tsample_ate;
>> +	div = t124_tsensor_config.tsample * t124_tsensor_config.pdiv_ate;
>
> t124_tsensor_config is a pretty long name and repeated fairly often.
> It's also one of those things that will likely change in future
> generations (judging by the t124_ prefix), so perhaps it should be
> stored as a const * somewhere. Perhaps in struct tegra_tsensor, so the
> above becomes:

Added it as a .config field to tegra_tsensor.

>
> 	mult = sensor->config->pdiv * sensor->config->tsample_ate;
>
>> +
>> +	therma = div64_s64_precise((s64) delta_temp * (1LL << 13) * mult,
>> +		(s64) delta_sens * div);
>
> There should be no space between the (s64) and the variable in the above
> (and below, well, everywhere really). Also I find this difficult to read
> because the second line is strangely indented. Can you align it with the
> function arguments on the first line, please?

Fixed.

>
>> +	thermb = div64_s64_precise(
>> +		((s64) actual_tsensor_ft * shared.actual_temp_cp) -
>> +		((s64) actual_tsensor_cp * shared.actual_temp_ft),
>> +		(s64) delta_sens);
>
> The way you need to wrap this indicates that you should either make
> variable names shorter or split this computation up into several steps.

Split up.

>
>> +
>> +	therma = div64_s64_precise((s64) therma * sensor->fuse_corr_alpha,
>> +		(s64) 1000000LL);
>> +	thermb = div64_s64_precise((s64) thermb * sensor->fuse_corr_alpha +
>> +		sensor->fuse_corr_beta,
>> +		(s64) 1000000LL);
>> +
>> +	*calib = ((u16)(therma) << SENSOR_CONFIG2_THERMA_SHIFT) |
>
> Why the parentheses around "therma"?

Not sure. Removed.

>
>> +		((u16)thermb << SENSOR_CONFIG2_THERMB_SHIFT);
>> +
>> +	return 0;
>> +}
>> +
>> +static int enable_tsensor(struct tegra_soctherm *tegra,
>> +			  const struct tegra_tsensor *sensor,
>> +			  struct tsensor_shared_calibration shared)
>
> Again, shouldn't you pass in "shared" by reference?
>
>> +{
>> +	void __iomem *base = tegra->regs + sensor->base;
>> +	unsigned int val;
>> +	u32 calib;
>> +	int err;
>> +
>> +	err = calculate_tsensor_calibration(sensor, shared, &calib);
>> +	if (err)
>> +		return err;
>> +
>> +	val = 0;
>
> There's no need for this, if you...
>
>> +	val |= t124_tsensor_config.tall << SENSOR_CONFIG0_TALL_SHIFT;
>
> ... make this "val = ...;"
>
>> +	writel(val, base + SENSOR_CONFIG0);
>> +
>> +	val = 0;
>
> Same here.

Fixed.

>
>> +	val |= (t124_tsensor_config.tsample - 1) <<
>> +		SENSOR_CONFIG1_TSAMPLE_SHIFT;
>
> Now the << operator is at the end of the line, so I think whichever way
> you decide it should at least be consistent.

This now fits on one line.

>
>> +/* Translate from soctherm readback format to millicelsius.
>
> Block comments should have the first line empty, like so:
>
> 	/*
> 	 * bla...
> 	 */
>
>> + * The soctherm readback format in bits is as follows:
>> + *   TTTTTTTT H______N
>> + * where T's contain the temperature in Celsius,
>> + * H denotes an addition of 0.5 Celsius and N denotes negation
>> + * of the final value.
>> + */
>> +static inline long translate_temp(u16 val)
>
> I think it can be left to the compiler to decide whether or not to
> inline this particular function.

Removed 'inline'.

>
>> +static const int thermctl_temp_offsets[] = {
>> +	SENSOR_TEMP1, SENSOR_TEMP2, SENSOR_TEMP1, SENSOR_TEMP2
>> +};
>> +
>> +static const int thermctl_temp_shifts[] = {
>> +	16, 16, 0, 0
>> +};
>
> Perhaps these should be in a single array that describes the individual
> zones. Something like:
>
> 	struct thermctl_zone_desc {
> 		unsigned int offset;
> 		unsigned int shift;
> 	};
>
> 	static const struct thermctl_zone_desc zones[] = {
> 		{ SENSOR_TEMP1, 16 },
> 		...
> 	};
>
> And perhaps this array should have a t124_ prefix since it could vary
> per SoC generation?

Changed.

>
>> +static int tegra_soctherm_probe(struct platform_device *pdev)
>> +{
>> +	struct tegra_soctherm *tegra;
>> +	struct thermal_zone_device *tz;
>> +	struct tsensor_shared_calibration shared_calib;
>> +	int i;
>
> Should be unsigned.

Fixed.

>
>> +	int err = 0;
>
> I don't think this needs to be initialized.

Removed initialization.

>
>> +
>> +	const struct tegra_tsensor *tsensors = t124_tsensors;
>> +
>> +	tegra = devm_kzalloc(&pdev->dev, sizeof(*tegra), GFP_KERNEL);
>> +	if (!tegra)
>> +		return -ENOMEM;
>> +
>> +	tegra->regs = devm_ioremap_resource(&pdev->dev,
>> +		platform_get_resource(pdev, IORESOURCE_MEM, 0));
>
> Can this please be two separate statements for better readability?

Split up.

>
>> +	if (IS_ERR(tegra->regs)) {
>> +		dev_err(&pdev->dev, "can't get registers");
>
> No need for the message. devm_ioremap_resource() prints one for you on
> error.

Removed message.

>
>> +		return PTR_ERR(tegra->regs);
>> +	}
>> +
>> +	tegra->reset = devm_reset_control_get(&pdev->dev, "soctherm");
>> +	if (IS_ERR(tegra->reset)) {
>> +		dev_err(&pdev->dev, "can't get soctherm reset\n");
>> +		return PTR_ERR(tegra->reset);
>> +	}
>> +
>> +	tegra->clock_tsensor = devm_clk_get(&pdev->dev, "tsensor");
>> +	if (IS_ERR(tegra->clock_tsensor)) {
>> +		dev_err(&pdev->dev, "can't get clock tsensor\n");
>
> Nit: the wording of this is inconsistent with the reset error.

Fixed.

>
>> +		return PTR_ERR(tegra->clock_tsensor);
>> +	}
>> +
>> +	tegra->clock_soctherm = devm_clk_get(&pdev->dev, "soctherm");
>> +	if (IS_ERR(tegra->clock_soctherm)) {
>> +		dev_err(&pdev->dev, "can't get clock soctherm\n");
>
> Same here.

Fixed.

>
>> +		return PTR_ERR(tegra->clock_soctherm);
>> +	}
>> +
>> +	reset_control_assert(tegra->reset);
>> +
>> +	err = clk_prepare_enable(tegra->clock_soctherm);
>> +	if (err) {
>> +		reset_control_deassert(tegra->reset);
>
> Is it really useful to deassert in case of failure? After the assertion
> of the reset above, all hardware state will be gone and since the device
> won't be used, nobody will reinitialize it properly. Taking it out of
> reset is useless.

Removed deasserts.

>
>> +	/* Initialize raw sensors */
>> +
>> +	err = calculate_shared_calibration(&shared_calib);
>> +	if (err)
>> +		goto disable_clocks;
>> +
>> +	for (i = 0; i < ARRAY_SIZE(t124_tsensors); ++i) {
>> +		err = enable_tsensor(tegra, tsensors + i, shared_calib);
>> +		if (err)
>> +			goto disable_clocks;
>> +	}
>> +
>> +	writel(SENSOR_PDIV_T124, tegra->regs + SENSOR_PDIV);
>> +	writel(SENSOR_HOTSPOT_OFF_T124, tegra->regs + SENSOR_HOTSPOT_OFF);
>> +
>> +	/* Initialize thermctl sensors */
>> +
>> +	for (i = 0; i < ARRAY_SIZE(tegra->thermctl_tzs); ++i) {
>> +		struct tegra_thermctl_zone *zone =
>> +			devm_kzalloc(&pdev->dev, sizeof(*zone), GFP_KERNEL);
>> +		if (!zone) {
>> +			err = -ENOMEM;
>> +			--i;
>
> What's up with the i line here? Doesn't that trigger a compiler warning?
>
> Oh wait, that's whitespace highlighting screwing with the --i. One more
> reason to prefer i-- where passible. Also I think the more idiomatic way
> for this kind of cleanup is to not touch i here, but to do this with a
>
> 	while (i--)
>
> later on. See below.

Fixed.

>
>> +			goto unregister_tzs;
>> +		}
>> +
>> +		zone->temp_reg = tegra->regs + thermctl_temp_offsets[i];
>> +		zone->temp_shift = thermctl_temp_shifts[i];
>> +
>> +		tz = thermal_zone_of_sensor_register(
>> +			&pdev->dev, i, zone, tegra_thermctl_get_temp, NULL);
>
> This is weirdly wrapped. Better would be:
>
>> +		tz = thermal_zone_of_sensor_register(&pdev->dev, i, zone,
>> +						     tegra_thermctl_get_temp,
>> +						     NULL);
>

Fixed.

>> +unregister_tzs:
>> +	for (; i >= 0; i--)
>> +		thermal_zone_of_sensor_unregister(&pdev->dev,
>> +						  tegra->thermctl_tzs[i]);
>
> Like I said above, I think the more idiomatic way would be:
>
> 	while (i--)
> 		thermal_zone_of_sensor_unregister(...);
>

Fixed.

>> +static int tegra_soctherm_remove(struct platform_device *pdev)
>> +{
>> +	struct tegra_soctherm *tegra = platform_get_drvdata(pdev);
>> +	int i;
>
> unsigned again.

Fixed.

>
>> +
>> +	for (i = 0; i < ARRAY_SIZE(tegra->thermctl_tzs); ++i) {
>> +		thermal_zone_of_sensor_unregister(&pdev->dev,
>> +						  tegra->thermctl_tzs[i]);
>> +	}
>> +
>> +	clk_disable_unprepare(tegra->clock_tsensor);
>> +	clk_disable_unprepare(tegra->clock_soctherm);
>> +
>> +	return 0;
>> +}
>> +
>> +static struct platform_driver tegra_soctherm_driver = {
>> +	.probe = tegra_soctherm_probe,
>> +	.remove = tegra_soctherm_remove,
>> +	.driver = {
>> +		.name = "tegra_soctherm",
>
> I think we're primarily using tegra- as prefix for driver names.

Fixed.

>
>> +		.of_match_table = tegra_soctherm_of_match,
>> +	},
>> +};
>> +module_platform_driver(tegra_soctherm_driver);
>> +
>> +MODULE_AUTHOR("Mikko Perttunen <mperttunen-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>");
>> +MODULE_DESCRIPTION("Tegra SOCTHERM thermal management driver");
>
> Perhaps: "NVIDIA Tegra ..."?

Fixed.

>
> Thierry
>

Thanks,
Mikko