Skip to contentSkip to conversation

leds: trigger: Add invert attribute to ledtrig-audio

PATCH In review Phil Sutter 2 participants 4 messages Aug 9, 2021 – Aug 10, 2021 LED

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 leds: trigger: Add invert attribute to ledtrig-audio

Inverting micmute LED used to be possible via a mixer setting, but conversion to LEDs class (probably) killed it. Re-establish the old functionality via sysfs attribute in audio LED triggers.

---
 drivers/leds/trigger/ledtrig-audio.c | 106 ++++++++++++++++++++++++++-
 1 file changed, 102 insertions(+), 4 deletions(-)
drivers/leds/trigger/ledtrig-audio.c +102 -4
index f76621e88482d..319ac4f514dfd 100644
@@ -6,9 +6,12 @@
 #include <linux/kernel.h>
 #include <linux/leds.h>
 #include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/sysfs.h>
 
 static struct led_trigger *ledtrig_audio[NUM_AUDIO_LEDS];
 static enum led_brightness audio_state[NUM_AUDIO_LEDS];
+static bool led_invert[NUM_AUDIO_LEDS];
 
 enum led_brightness ledtrig_audio_get(enum led_audio type)
 {
@@ -18,17 +21,112 @@ EXPORT_SYMBOL_GPL(ledtrig_audio_get);
 
 void ledtrig_audio_set(enum led_audio type, enum led_brightness state)
 {
+	if (led_invert[type])
+		state = !state;
+
 	audio_state[type] = state;
 	led_trigger_event(ledtrig_audio[type], state);
 }
 EXPORT_SYMBOL_GPL(ledtrig_audio_set);
 
+static ssize_t do_invert_show(enum led_audio type, char *buf)
+{
+	return sprintf(buf, "%u\n", led_invert[type]);
+}
+
+static ssize_t mute_invert_show(struct device *dev,
+				struct device_attribute *attr, char *buf)
+{
+	return do_invert_show(LED_AUDIO_MUTE, buf);
+}
+
+static ssize_t micmute_invert_show(struct device *dev,
+			       struct device_attribute *attr, char *buf)
+{
+	return do_invert_show(LED_AUDIO_MICMUTE, buf);
+}
+
+static ssize_t do_invert_store(enum led_audio type,
+			       const char *buf, size_t size)
+{
+	unsigned long state;
+	int ret;
+
+	ret = kstrtoul(buf, 0, &state);
+	if (ret)
+		return ret;
+
+	led_invert[type] = !!state;
+	ledtrig_audio_set(type, audio_state[type]);
+
+	return size;
+}
+
+static ssize_t mute_invert_store(struct device *dev,
+				 struct device_attribute *attr,
+				 const char *buf, size_t size)
+{
+	return do_invert_store(LED_AUDIO_MUTE, buf, size);
+}
+
+static ssize_t micmute_invert_store(struct device *dev,
+				    struct device_attribute *attr,
+				    const char *buf, size_t size)
+{
+	return do_invert_store(LED_AUDIO_MICMUTE, buf, size);
+}
+
+static struct device_attribute dev_attr_mute_invert =
+		__ATTR(invert, 0644, mute_invert_show, mute_invert_store);
+
+static struct attribute *audio_mute_trig_attrs[] = {
+	&dev_attr_mute_invert.attr,
+	NULL
+};
+ATTRIBUTE_GROUPS(audio_mute_trig);
+
+static struct device_attribute dev_attr_micmute_invert =
+		__ATTR(invert, 0644, micmute_invert_show, micmute_invert_store);
+
+static struct attribute *audio_micmute_trig_attrs[] = {
+	&dev_attr_micmute_invert.attr,
+	NULL
+};
+ATTRIBUTE_GROUPS(audio_micmute_trig);
+
+static void __init do_ledtrig_audio_init(const char *name,
+					 enum led_audio type,
+					 const struct attribute_group **groups)
+{
+	struct led_trigger *trigger;
+	int err;
+
+	trigger = kzalloc(sizeof(struct led_trigger), GFP_KERNEL);
+	if (!trigger) {
+		pr_warn("LED trigger %s failed to register (no memory)\n",
+			name);
+		goto out;
+	}
+
+	trigger->name = name;
+	trigger->groups = groups;
+
+	err = led_trigger_register(trigger);
+	if (err < 0) {
+		kfree(trigger);
+		trigger = NULL;
+		pr_warn("LED trigger %s failed to register (%d)\n", name, err);
+	}
+out:
+	ledtrig_audio[type] = trigger;
+}
+
 static int __init ledtrig_audio_init(void)
 {
-	led_trigger_register_simple("audio-mute",
-				    &ledtrig_audio[LED_AUDIO_MUTE]);
-	led_trigger_register_simple("audio-micmute",
-				    &ledtrig_audio[LED_AUDIO_MICMUTE]);
+	do_ledtrig_audio_init("audio-mute", LED_AUDIO_MUTE,
+			      audio_mute_trig_groups);
+	do_ledtrig_audio_init("audio-micmute", LED_AUDIO_MICMUTE,
+			      audio_micmute_trig_groups);
 	return 0;
 }
 module_init(ledtrig_audio_init);
Signature · 1 line
2.32.0
#1 lore raw

Hi!

Quoted 3 lines

Inverting micmute LED used to be possible via a mixer setting, but conversion to LEDs class (probably) killed it. Re-establish the old functionality via sysfs attribute in audio LED triggers.

So we have both invert and inverted attributes. Fun :-).

See sysfs-class-led and sysfs-class-led-trigger-oneshot.

We definitely want this documented. We probably want this for most triggers, maybe it should get one implementation in library somewhere?

Otherwise it makes sense.

drivers/leds/trigger/ledtrig-audio.c EXPORT_SYMBOL_GPL(ledtrig_audio_get); · line 60

Quoted 12 lines
drivers/leds/trigger/ledtrig-audio.c +12 -0
+static ssize_t do_invert_store(enum led_audio type,
+			       const char *buf, size_t size)
+{
+	unsigned long state;
+	int ret;
+
+	ret = kstrtoul(buf, 0, &state);
+	if (ret)
+		return ret;
+
+	led_invert[type] = !!state;
+	ledtrig_audio_set(type, audio_state[type]);

Accepting 42 as valid value sounds wrong. Anyway, this should do what oneshot trigger does.

Best regards,
Pavel

Signature · 1 line
http://www.livejournal.com/~pavelmachek
#2 lore raw

Hi Pavel,

Quoted 5 lines · Pavel Machek
Quoted 3 lines

Inverting micmute LED used to be possible via a mixer setting, but conversion to LEDs class (probably) killed it. Re-establish the old functionality via sysfs attribute in audio LED triggers.

So we have both invert and inverted attributes. Fun :-).

Hmm! :)

Are you talking about LED_BLINK_INVERT flag? I see a few triggers allow inversion but didn't find LED drivers exporting such a property.

Quoted 1 line

See sysfs-class-led and sysfs-class-led-trigger-oneshot.

I think I "copied" from oneshot trigger when writing this patch.

Quoted 2 lines

We definitely want this documented. We probably want this for most triggers, maybe it should get one implementation in library somewhere?

Should this be an implicit attribute of simple triggers only or all? In the latter case (which could simplify some triggers) I guess the value inversion has to take place in led_set_brightness_nopm(), the lowest level function triggers may use.

How does inversion work, actually? LED_OFF <-> LED_ON is trivial, but what about LED_HALF and LED_FULL? Leaving LED_HALF as-is seems logical, but the opposite of LED_OFF might be LED_ON or LED_FULL. Does max_brightness determine that?

Quoted 18 lines

Otherwise it makes sense.

drivers/leds/trigger/ledtrig-audio.c EXPORT_SYMBOL_GPL(ledtrig_audio_get); · line 60

Quoted 12 lines
drivers/leds/trigger/ledtrig-audio.c +12 -0
+static ssize_t do_invert_store(enum led_audio type,
+			       const char *buf, size_t size)
+{
+	unsigned long state;
+	int ret;
+
+	ret = kstrtoul(buf, 0, &state);
+	if (ret)
+		return ret;
+
+	led_invert[type] = !!state;
+	ledtrig_audio_set(type, audio_state[type]);

Accepting 42 as valid value sounds wrong. Anyway, this should do what oneshot trigger does.

Similarities to oneshot are not a coincidence. :)

Cheers, Phil

#3 lore raw

Hi!

Quoted 17 lines · Pavel Machek
Quoted 5 lines
Quoted 3 lines

Inverting micmute LED used to be possible via a mixer setting, but conversion to LEDs class (probably) killed it. Re-establish the old functionality via sysfs attribute in audio LED triggers.

So we have both invert and inverted attributes. Fun :-).

Hmm! :)

Are you talking about LED_BLINK_INVERT flag? I see a few triggers allow inversion but didn't find LED drivers exporting such a property.

Quoted 1 line

See sysfs-class-led and sysfs-class-led-trigger-oneshot.

I think I "copied" from oneshot trigger when writing this patch.

Quoted 7 lines
Quoted 2 lines

We definitely want this documented. We probably want this for most triggers, maybe it should get one implementation in library somewhere?

Should this be an implicit attribute of simple triggers only or all? In the latter case (which could simplify some triggers) I guess the value inversion has to take place in led_set_brightness_nopm(), the lowest level function triggers may use.

Actual inversion is trivial (so it may not make sense to share), but sysfs interface is not. Sharing that would be good.

Quoted 4 lines

How does inversion work, actually? LED_OFF <-> LED_ON is trivial, but what about LED_HALF and LED_FULL? Leaving LED_HALF as-is seems logical, but the opposite of LED_OFF might be LED_ON or LED_FULL. Does max_brightness determine that?

max_brightness determines that, and it would be simply

if (->inverted)
b = ->max_brighntess-b;

Best regards,
Pavel

Signature · 1 line
http://www.livejournal.com/~pavelmachek