linux-kernel - Re: [PATCH 4/4] power_supply: bq24261 charger driver

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20140204113621.GB2450@amd.pavel.ucw.cz>
Date:	Tue, 4 Feb 2014 12:36:21 +0100
From:	Pavel Machek <pavel@....cz>
To:	Jenny TC <jenny.tc@...el.com>
Cc:	linux-kernel@...r.kernel.org,
	Dmitry Eremin-Solenikov <dbaryshkov@...il.com>,
	Anton Vorontsov <cbouatmailru@...il.com>,
	Anton Vorontsov <anton.vorontsov@...aro.org>,
	Kim Milo <Milo.Kim@...com>, Lee Jones <lee.jones@...aro.org>,
	Jingoo Han <jg1.han@...sung.com>,
	Chanwoo Choi <cw00.choi@...sung.com>,
	Sachin Kamat <sachin.kamat@...aro.org>,
	Lars-Peter Clausen <lars@...afoo.de>,
	Pali Rohár <pali.rohar@...il.com>,
	Rhyland Klein <rklein@...dia.com>,
	"Rafael J. Wysocki" <rjw@...ysocki.net>,
	David Woodhouse <dwmw2@...radead.org>,
	Tony Lindgren <tony@...mide.com>,
	Russell King <linux@....linux.org.uk>,
	Sebastian Reichel <sre@...g0.de>, aaro.koskinen@....fi,
	Pallala Ramakrishna <ramakrishna.pallala@...el.com>,
	freemangordon@....bg, linux-omap@...r.kernel.org
Subject: Re: [PATCH 4/4] power_supply: bq24261 charger driver

Hi!


> +#define DEV_MANUFACTURER "TI"
> +#define DEV_MANUFACTURER_NAME_SIZE 4

This is unneccessarily complicated for no reason. You copy "TI" to
struct, just so that ou can return pointer to the field on
get_property.

What about simply returning "TI" from get_property, without defines
and copying?

> +#define BQ24261_MIN_CV 3500
> +#define BQ24261_MAX_CV 4440

Other defines use uV as an unit :-(.

> +static void lookup_regval(u16 tbl[][2], size_t size, u16 in_val, u8 *out_val)
> +{
> +	int i;
> +
> +	for (i = 1; i < size; ++i)
> +		if (in_val < tbl[i][0])
> +			break;
> +
> +	*out_val = (u8) tbl[i - 1][1];
> +}

Umm. Could we simply return the value?

> +static void bq24261_cc_to_reg(int cc, u8 *reg_val)
> +{
> +
> +	cc = cc < BQ24261_MAX_CC ? cc : BQ24261_MAX_CC;
> +	cc = cc - BQ24261_MIN_CC;

clamp_t?

> +	*reg_val = cc > 0 ? ((cc/100) << 3) & 0xFF : 0;
> +}

Just return the value?

> +static void bq24261_cv_to_reg(int cv, u8 *reg_val)
> +{
> +	int val;
> +
> +	val = clamp_t(int, cv, BQ24261_MIN_CV, BQ24261_MAX_CV);
> +	*reg_val =
> +		(((val - BQ24261_MIN_CV) / BQ24261_CV_DIV)
> +			<< BQ24261_CV_BIT_POS);
> +}

Not sure if the defines really make it more readable. It should be
consistent with the above/below functions...

> +static inline void bq24261_iterm_to_reg(int iterm, u8 *regval)
> +{
> +	iterm = iterm < BQ24261_MAX_ITERM ? iterm : BQ24261_MAX_ITERM;
> +	iterm = iterm - BQ24261_MIN_ITERM;

clamp_t?

> +	*regval = iterm > 0 ? (iterm/50) & 0xFF : 0;
> +}

Just return the value.

> +static inline void bq24261_sfty_tmr_to_reg(int tmr, u8 *regval)
> +{
> +	return lookup_regval(bq24261_sfty_tmr, ARRAY_SIZE(bq24261_sfty_tmr),
> +			     tmr, regval);
> +}

Just return the value... returning void values with explicit return is
"interesting".

> +	/* If status is fault, wait for READY before enabling the charging */
> +
> +	if (!is_ready) {
> +		ret = wait_event_timeout(chip->wait_ready,
> +			(chip->chrgr_stat != BQ24261_CHRGR_STAT_READY),
> +				HZ);
> +		dev_info(&chip->client->dev,
> +			"chrgr_stat=%x\n", chip->chrgr_stat);
> +		if (ret == 0) {
> +			dev_err(&chip->client->dev,
> +				"Waiting for Charger Ready Failed.Enabling charging anyway\n");
> +		}
> +	}

So charger has a problem, and we force it on, anyway? Also put space
after ".".

> +static inline int bq24261_set_cv(struct bq24261_charger *chip, int cv)
> +{
> +	int bat_volt;
> +	int ret;
> +	u8 reg_val;
> +	u8 vindpm_val = 0x0;
> +
> +	/*
> +	* Setting VINDPM value as per the battery voltage
> +	*  VBatt           Vindpm     Register Setting
> +	*  < 3.7v           4.2v       0x0 (default)
> +	*  3.71v - 3.96v    4.36v      0x2
> +	*  > 3.96v          4.6v       0x5
> +	*/
> +	ret = get_battery_voltage(&bat_volt);
> +	if (ret) {
> +		dev_err(&chip->client->dev,
> +			"Error getting battery voltage!!\n");
> +	} else {

You forget the error value and continue anyway.

> +static inline void resume_charging(struct bq24261_charger *chip)
> +{
> +
> +	if (chip->is_charger_enabled)
> +		bq24261_enable_charger(chip, true);
> +	if (chip->inlmt)
> +		bq24261_set_inlmt(chip, chip->inlmt);
> +	if (chip->cc)
> +		bq24261_set_cc(chip, chip->cc);
> +	if (chip->cv)
> +		bq24261_set_cv(chip, chip->cv);
> +	if (chip->is_charging_enabled)
> +		bq24261_enable_charging(chip, true);

What about some error checking?

Is it wise to enable charging when setting voltage failed?

> +static inline bool is_bq24261_enabled(struct bq24261_charger *chip)
> +{
> +	if (chip->cable_type == PSY_CHARGER_CABLE_TYPE_NONE)
> +		return false;
> +	else if (!chip->is_charger_enabled)
> +		return false;

Kill the else.

> +static inline int get_battery_voltage(int *volt)
> +{
> +	struct power_supply *psy;
> +	union power_supply_propval val;
> +	int ret;
> +
> +	psy = get_psy_battery();
> +	if (!psy)
> +		return -EINVAL;

Hmm. Does this assume just one battery in the system?

Is it good idea? Older machines contain main and memory backup
batteries. Newer machines contain keyboard and display battery....


									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/