Re: [PATCH v3 1/5] leds: lm3532: Fix brightness control for i2c mode

From: Dan Murphy
Date: Tue Aug 27 2019 - 12:19:45 EST


On 8/27/19 11:04 AM, Tony Lindgren wrote:
* Dan Murphy <dmurphy@xxxxxx> [190827 13:02]:

On 8/27/19 7:44 AM, Dan Murphy wrote:

On 8/27/19 7:18 AM, Pavel Machek wrote:
On Mon 2019-08-26 15:44:37, Tony Lindgren wrote:
* Pavel Machek <pavel@xxxxxx> [190826 22:14]:
On Mon 2019-08-26 14:58:22, Tony Lindgren wrote:

* Dan Murphy <dmurphy@xxxxxx> [190820 19:53]:
Fix the brightness control for I2C mode. Instead of
changing the full scale current register update the ALS target
register for the appropriate banks.

In addition clean up some code errors and random misspellings found
during coding.

Tested on Droid4 as well as LM3532 EVM connected to
a BeagleBoneBlack

Fixes: e37a7f8d77e1 ("leds: lm3532: Introduce the
lm3532 LED driver")
Reported-by: Pavel Machek <pavel@xxxxxx>
Signed-off-by: Dan Murphy <dmurphy@xxxxxx>

v3 - Removed register define updates -
Looks like starting with this patch in Linux next the LCD on droid4
is so dim it's unreadable even with brightness set to 255. Setting
brightness to 0 does blank it completely though.

Did something maybe break with the various patch revisions or are
we now missing some dts patch?
Maybe missing dts patch. We should provide maximum current the LED can
Or i2c control is somehow broken and only als control now works?
With only setting CONFIG_LEDS_LM3532=m to the next branch I get full
brightness with 255.

I also see half brightness at 128 with the ramp down working.

I am not able to reproduce this issue on my device.

Just to make sure my data was right I did a clean rebuild on commit

from the led-next branch.

Just adding the above config flag. I still cannot reproduce the issue

See attached pic
OK thanks for checking. Probably you can reproduce the issue if you
reset things to commit c4b8354e5341 ("leds: lm3532: Fix brightness
control for i2c mode"). There might now be something uninitialized
with that commit depending on the hardware state on boot if you
care to check that. Or maybe there's some interaction with other
patches not yet at commit c4b8354e5341 level.

Ok the fix for that issue is in this patch

leds: lm3532: Fixes for the driver for stability

The fix for setting the brightness control register during the init forced the ALS mode to

be enabled. The Fixes for driver stability patch then set the led->mode to the correct register setting.

I confirmed again that things fail at commit c4b8354e5341. But
now testing with the same Linux next as yesterday things works again.
Not sure what's going on as failures with Linux next yestreday
made me start narrowing down what commit causes the issues.

Anyways, playing with loading and unloading the leds-lm3532.ko I
noticed we can also have unpaired regulator calls when using sysfs
brightness that the below patch attempts to fix. Not sure how I got
to the point of regulator warnings, but I was trying to enable
the brightness via sysfs. Maybe I had other patches too when
I got the regulator warnings.. But please check if the below
patch makes sense.

Makes sense did you want me to push a patch?




8< ---------------------------
diff --git a/drivers/leds/leds-lm3532.c b/drivers/leds/leds-lm3532.c
--- a/drivers/leds/leds-lm3532.c
+++ b/drivers/leds/leds-lm3532.c
@@ -127,6 +127,7 @@ struct lm3532_als_data {
* @num_leds - Number of LED strings are supported in this array
* @full_scale_current - The full-scale current setting for the current sink.
* @led_strings - The LED strings supported in this array
+ * @enabled - Enabled status
* @label - LED label
struct lm3532_led {
@@ -138,6 +139,7 @@ struct lm3532_led {
int ctrl_brt_pointer;
int num_leds;
int full_scale_current;
+ int enabled:1;
u32 led_strings[LM3532_MAX_CONTROL_BANKS];
char label[LED_MAX_NAME_SIZE];
@@ -292,11 +294,15 @@ static int lm3532_get_ramp_index(int ramp_time)
+/* Caller must take care of locking */
static int lm3532_led_enable(struct lm3532_led *led_data)
int ctrl_en_val = BIT(led_data->control_bank);
int ret;
+ if (led_data->enabled)
+ return 0;
ret = regmap_update_bits(led_data->priv->regmap, LM3532_REG_ENABLE,
ctrl_en_val, ctrl_en_val);
if (ret) {
@@ -304,14 +310,24 @@ static int lm3532_led_enable(struct lm3532_led *led_data)
return ret;
- return regulator_enable(led_data->priv->regulator);
+ ret = regulator_enable(led_data->priv->regulator);
+ if (ret < 0)
+ return ret;
+ led_data->enabled = 1;
+ return 0;
+/* Caller must take care of locking */
static int lm3532_led_disable(struct lm3532_led *led_data)
int ctrl_en_val = BIT(led_data->control_bank);
int ret;
+ if (!led_data->enabled)
+ return 0;
ret = regmap_update_bits(led_data->priv->regmap, LM3532_REG_ENABLE,
ctrl_en_val, 0);
if (ret) {
@@ -319,7 +335,13 @@ static int lm3532_led_disable(struct lm3532_led *led_data)
return ret;
- return regulator_disable(led_data->priv->regulator);
+ ret = regulator_disable(led_data->priv->regulator);
+ if (ret < 0)
+ return ret;
+ led_data->enabled = 0;
+ return 0;
static int lm3532_brightness_set(struct led_classdev *led_cdev,