Re: [PATCH] leds: ledtrig-morse: send out morse code

From: Andreas Klinger
Date: Thu Jun 28 2018 - 16:30:33 EST


Hi Pavel,

Pavel Machek <pavel@xxxxxx> schrieb am Thu, 28. Jun 20:56:
> Hi!
>
> > Send out a morse code by using LEDs.
> >
> > This is useful especially on embedded systems without displays to tell the
> > user about error conditions and status information.
> >
> > The trigger will be called "morse"
> >
> > The string to be send is written into the file morse_string and sent out
> > with a workqueue. Supported are letters and digits.
> >
> > With the file dot_unit the minimal time unit can be adjusted in
> > milliseconds.
> >
> > Signed-off-by: Andreas Klinger <ak@xxxxxxxxxxxxx>
>
> Can we get more general "pattern" trigger? Some LEDs can do that in
> hardware, and it is more general than plain morse.
>
> Ouch and it already was implemented :-). Patch is in attachment.
>
> Pavel

The idea of the morse trigger is so be able so send out a short error
code or if needed also a complete sentence. The morse code don't need
extra explanation to the user. Decoding can be found everywhere.

With the pattern trigger one need a translation of the letters and
numbers into morse code. This is what the morse trigger is doing.

So from my perspective the pattern trigger is something different than
the morse one.

Another question:
The pattern trigger is not in mainline. Do you know why?

Andreas


>
> --
> (english) http://www.livejournal.com/~pavelmachek
> (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

> From rteysseyre@xxxxxxxxx Fri Mar 20 15:40:22 2015
> Return-Path: <rteysseyre@xxxxxxxxx>
> X-Original-To: pavel@xxxxxxxxxxxxxxxxxxxxxxxx
> Delivered-To: pavel@xxxxxxxxxxxxxxxxxxxxxxxx
> Received: from jabberwock.ucw.cz (jabberwock.ucw.cz [46.255.230.98])
> by atrey.karlin.mff.cuni.cz (Postfix) with ESMTP id 3275281ECB
> for <pavel@xxxxxxxxxxxxxxxxxxxxxxxx>; Fri, 20 Mar 2015 15:40:22 +0100 (CET)
> Received: by jabberwock.ucw.cz (Postfix)
> id 3053A1C00FF; Fri, 20 Mar 2015 15:40:22 +0100 (CET)
> Delivered-To: pavel@xxxxxxxxxxxxxxxxx
> Received: from mail-wg0-f46.google.com (mail-wg0-f46.google.com [74.125.82.46])
> (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits))
> (Client CN "smtp.gmail.com", Issuer "Google Internet Authority G2" (not verified))
> by jabberwock.ucw.cz (Postfix) with ESMTPS id 18EB61C00F9
> for <pavel@xxxxxx>; Fri, 20 Mar 2015 15:40:22 +0100 (CET)
> Received: by wgbcc7 with SMTP id cc7so91022009wgb.0
> for <pavel@xxxxxx>; Fri, 20 Mar 2015 07:40:21 -0700 (PDT)
> DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
> d=gmail.com; s=20120113;
> h=subject:from:to:cc:content-type:date:message-id:mime-version
> :content-transfer-encoding;
> bh=UdT+u6nHVkpFGm2AqZ0XFePjLX83mCQTkcoTCraWWvc=;
> b=b9VbjaS+0McMAtdeSJv1Lhb6RDg/noWghgIsbIZD/DpMnIXJ05Dyi160Jx28uS0YzT
> zyPmtmEv3IrnmkcCp7g6S6f04ATgi9q3tI+5qooaQS65V7bRtlpcTMoB8KEA38SsBtap
> 8Ng+65MHchSpmGyR/h0WxVg934ieNpySSLCQ/vSUu2lOxi0SLmyr93KqfZXQ6Yk5FCA3
> QOcCuiXpcLfmGbLwR2YT0uu3/60LOmWL6xt+J/bkWjo38qwqtH5phoa5ydjcUy5MZJSZ
> tn4qM8mQ8YH9EUJ+VjcxQX4nYdY+Mf3tsnSOieHNzUVt+2z1GmLcDi1Jfe8MtyPh8+9m
> dp+w==
> X-Received: by 10.194.108.137 with SMTP id hk9mr118158926wjb.112.1426862421666;
> Fri, 20 Mar 2015 07:40:21 -0700 (PDT)
> Received: from [192.168.244.132] (fs-141-0-203-237.fullsave.info. [141.0.203.237])
> by mx.google.com with ESMTPSA id ha10sm6624401wjc.37.2015.03.20.07.40.20
> (version=SSLv3 cipher=RC4-SHA bits=128/128);
> Fri, 20 Mar 2015 07:40:20 -0700 (PDT)
> Subject: [PATCH v2] leds: Add arbitrary pattern trigger
> From: =?ISO-8859-1?Q?Rapha=EBl?= Teysseyre <rteysseyre@xxxxxxxxx>
> To: Pavel Machek <pavel@xxxxxx>
> Cc: Joe Xue <lgxue@xxxxxxxxxxx>, Bryan Wu <cooloney@xxxxxxxxx>,
> "rpurdie@xxxxxxxxx" <rpurdie@xxxxxxxxx>, Linux LED Subsystem
> <linux-leds@xxxxxxxxxxxxxxx>, lkml <linux-kernel@xxxxxxxxxxxxxxx>
> Content-Type: text/plain; charset="UTF-8"
> Date: Fri, 20 Mar 2015 15:42:02 +0100
> Message-ID: <1426862522.48232.3.camel@localhost>
> Mime-Version: 1.0
> X-Mailer: Evolution 2.32.3 (2.32.3-34.el6)
> Content-Transfer-Encoding: 8bit
> X-CRM114-Status: Good ( pR: 191.7191 )
> Status: RO
> Content-Length: 17146
>
> Hi all,
>
> Following your comments about [PATCH] leds: Add arbitrary pattern trigger,
> here is a revised version of this patch.
>
> This is intended for embedded systems without screen or network access
> to show a status (or error) code to a human.
>
> It's been tested on an ARM architecture (Xilinx Zynq 7010 SoC,
> which CPU is a dual ARM Cortex-A9), with a non-mainline
> kernel (xilinx-v2014.4, based of a 3.17.0 kernel).
> Unfortunately, I don't have other hardware to test it on.
> It compiles fine in a 3.19.0-rc1 source tree.
> Additional testing and comments would be appreciated.
>
> Changes since version 1:
> - Fixed typos in documentation and comments.
> - Fixed MODULE_LICENSE string.
> - No more mutex in atomic context.
> - Return EINVAL when invalid patterns are written to the "pattern" attribute.
>
>
>
> Add a new led trigger supporting arbitrary patterns.
> This is useful for embedded systems with no screen or network access.
>
> Export two sysfs attributes: "pattern", and "repeat".
>
> When "repeat"=-1, repeat the pattern indefinitely. Otherwise,
> repeat it the specified number of times. "pattern" specifies
> the pattern as comma separated couples of "brightness duration"
> values. See detailled documentation in patch.
>
> Signed-off-by: Raphaël Teysseyre <rteysseyre@xxxxxxxxx>
> ---
> Documentation/leds/ledtrig-pattern.txt | 86 +++++++
> drivers/leds/trigger/Kconfig | 10 +
> drivers/leds/trigger/Makefile | 1 +
> drivers/leds/trigger/ledtrig-pattern.c | 432 ++++++++++++++++++++++++++++++++
> 4 files changed, 529 insertions(+), 0 deletions(-)
> create mode 100644 Documentation/leds/ledtrig-pattern.txt
> create mode 100644 drivers/leds/trigger/ledtrig-pattern.c
>
> diff --git a/Documentation/leds/ledtrig-pattern.txt b/Documentation/leds/ledtrig-pattern.txt
> new file mode 100644
> index 0000000..579295e
> --- /dev/null
> +++ b/Documentation/leds/ledtrig-pattern.txt
> @@ -0,0 +1,86 @@
> +LED Pattern Trigger
> +===================
> +
> +This is a LED trigger allowing arbitrary pattern execution. It can do gradual
> +dimming. This trigger can be configured to repeat the pattern a number of
> +times or indefinitely. This is intended as a way of communication for embedded
> +systems with no screen.
> +
> +The trigger can be activated from user space on LED class devices as shown
> +below:
> +
> + echo pattern > trigger
> +
> +This adds the following sysfs attributes to the LED:
> +
> + pattern - specifies the pattern. See syntax below.
> +
> + repeat - number of times the pattern must be repeated.
> + writing -1 to this file will make the pattern
> + repeat indefinitely.
> +
> +The pattern will be restarted each time a new value is written to
> +the pattern or repeat attribute. When dimming, the LED brightness
> +is set every 50 ms.
> +
> +pattern syntax:
> +The pattern is specified in the pattern attribute with an array of comma-
> +separated "brightness/length in miliseconds" values. The two components
> +of each value are to be separated by a space.
> +
> +For example, assuming the driven LED supports
> +intensity value from 0 to 255:
> +
> + echo 0 1000, 255 2000 > pattern
> +
> +Or:
> +
> + echo 0 1000, 255 2000, > pattern
> +
> +Will make the LED go gradually from zero-intensity to max (255) intensity
> +in 1000 milliseconds, then back to zero intensity in 2000 milliseconds:
> +
> +LED brightness
> + ^
> +255-| / \ / \ /
> + | / \ / \ /
> + | / \ / \ /
> + | / \ / \ /
> + 0-| / \/ \/
> + +---0----1----2----3----4----5----6------------> time (s)
> +
> +
> +
> +To make the LED go instantly from one brigntess value to another,
> +use zero-time lengths. For example:
> +
> + echo 0 1000, 0 0, 255 2000, 255 0 > pattern
> +
> +Will make the LED stay off for one second, then stay at max brightness
> +for two seconds:
> +
> +LED brightness
> + ^
> +255-| +---------+ +---------+
> + | | | | |
> + | | | | |
> + | | | | |
> + 0-| -----+ +----+ +----
> + +---0----1----2----3----4----5----6------------> time (s)
> +
> +
> +Notes:
> +
> +Patterns with invalid syntax are reported with EINVAL, the
> +resulting LED brightness is undefined. Reading the pattern
> +back returns an empty string.
> +
> +Patterns with less than two values, no value with time length > 50
> +milliseconds, or no two values with differing brightnesses
> +result in the LED being set at the brightness of the first value,
> +or zero if the pattern contains no value. EINVAL is returned,
> +and reading the pattern back returns an empty string.
> +
> +Because sysfs is used to define the pattern, patterns that need more than
> +PAGE_SIZE characters to describe aren't supported. PAGE_SIZE is system
> +dependent.
> diff --git a/drivers/leds/trigger/Kconfig b/drivers/leds/trigger/Kconfig
> index 49794b4..ce27bdb 100644
> --- a/drivers/leds/trigger/Kconfig
> +++ b/drivers/leds/trigger/Kconfig
> @@ -108,4 +108,14 @@ config LEDS_TRIGGER_CAMERA
> This enables direct flash/torch on/off by the driver, kernel space.
> If unsure, say Y.
>
> +config LEDS_TRIGGER_PATTERN
> + tristate "LED Pattern Trigger"
> + depends on LEDS_TRIGGERS
> + help
> + This allows LEDs blinking with an arbitrary pattern. Can be useful
> + on embedded systems with no screen to give out a status code to
> + a human.
> +
> + If unsure, say N
> +
> endif # LEDS_TRIGGERS
> diff --git a/drivers/leds/trigger/Makefile b/drivers/leds/trigger/Makefile
> index 1abf48d..a739429 100644
> --- a/drivers/leds/trigger/Makefile
> +++ b/drivers/leds/trigger/Makefile
> @@ -8,3 +8,4 @@ obj-$(CONFIG_LEDS_TRIGGER_CPU) += ledtrig-cpu.o
> obj-$(CONFIG_LEDS_TRIGGER_DEFAULT_ON) += ledtrig-default-on.o
> obj-$(CONFIG_LEDS_TRIGGER_TRANSIENT) += ledtrig-transient.o
> obj-$(CONFIG_LEDS_TRIGGER_CAMERA) += ledtrig-camera.o
> +obj-$(CONFIG_LEDS_TRIGGER_PATTERN) += ledtrig-pattern.o
> diff --git a/drivers/leds/trigger/ledtrig-pattern.c b/drivers/leds/trigger/ledtrig-pattern.c
> new file mode 100644
> index 0000000..fdf231d
> --- /dev/null
> +++ b/drivers/leds/trigger/ledtrig-pattern.c
> @@ -0,0 +1,432 @@
> +/*
> + * Arbitrary pattern trigger
> + *
> + * Copyright 2015, Epsiline
> + *
> + * Author : Raphaël Teysseyre <rteysseyre@xxxxxxxxx>
> + *
> + * Idea discussed with Pavel Machek <pavel@xxxxxx> on
> + * <linux-leds@xxxxxxxxxxxxxxx> (march 2015, thread title
> + * [PATCH RFC] leds: Add status code trigger)
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + */
> +
> +#include <linux/types.h>
> +#include <linux/kernel.h>
> +#include <linux/slab.h>
> +#include <linux/leds.h>
> +#include <linux/module.h>
> +#include <linux/mutex.h>
> +#include "../leds.h"
> +
> +struct pattern_step {
> + int brightness;
> + int time_ms;
> +};
> +
> +struct pattern_trig_data {
> + struct pattern_step *steps; /* Array describing the pattern */
> + struct mutex lock;
> + char is_sane;
> + struct pattern_step *curr;
> + struct pattern_step *next;
> + int time_ms; /* Time in current step */
> + int nsteps; /* Number of steps */
> + int repeat; /* < 0 means repeat indefinitely */
> + struct workqueue_struct *queue;
> + struct delayed_work dwork;
> + struct led_classdev *led_cdev; /* Needed by pattern_trig_update() */
> +};
> +
> +#define UPDATE_INTERVAL 50
> +/* When doing gradual dimming, the led brightness
> + will be updated every UPDATE_INTERVAL milliseconds */
> +
> +#define PATTERN_SEPARATOR ","
> +
> +#define MAX_NSTEPS (PAGE_SIZE/4)
> +/* The "pattern" attribute contains at most PAGE_SIZE characters.
> + Each pattern step in this attribute needs at least 4 characters
> + (a 1-digit number for the led brighntess, a space,
> + a 1-digit number for the time, a PATTERN_SEPARATOR).
> + Therefore, there is at most PAGE_SIZE/4 steps. */
> +
> +static int pattern_trig_initialize_data(struct pattern_trig_data *data)
> +{
> + mutex_init(&data->lock);
> + mutex_lock(&data->lock);
> +
> + data->is_sane = 0;
> + data->steps = kzalloc(MAX_NSTEPS*sizeof(struct pattern_step),
> + GFP_KERNEL);
> + if (!data->steps)
> + return -ENOMEM;
> +
> + data->curr = NULL;
> + data->next = NULL;
> + data->time_ms = 0;
> + data->nsteps = 0;
> + data->repeat = -1;
> + mutex_unlock(&data->lock);
> + return 0;
> +}
> +
> +static void pattern_trig_clear_data(struct pattern_trig_data *data)
> +{
> + data->is_sane = 0;
> + kfree(data->steps);
> +}
> +
> +/*
> + * is_sane : pattern checking.
> + * A pattern satisfying these three conditions is reported as sane :
> + * - At least two steps
> + * - At least one step with time >= UPDATE_INTERVAL
> + * - At least two steps with differing brightnesses
> + * When @data->pattern isn't sane, a sensible brightness
> + * default is suggested in @brightness
> + *
> + * DO NOT call pattern_trig_update() on a not-sane pattern
> + * with is_sane = 1, you'll be punished with an infinite
> + * loop in the kernel.
> + */
> +static int is_sane(struct pattern_trig_data *data, int *brightness)
> +{
> + int i;
> + char stept_ok = 0;
> + char stepb_ok = 0;
> +
> + *brightness = 0;
> + if (data->nsteps < 1)
> + return 0;
> +
> + *brightness = data->steps[0].brightness;
> + if (data->nsteps < 2)
> + return 0;
> +
> + for (i = 0; i < data->nsteps; i++) {
> + if (data->steps[i].time_ms >= UPDATE_INTERVAL) {
> + if (stepb_ok)
> + return 1;
> + stept_ok = 1;
> + }
> + if (data->steps[i].brightness != data->steps[0].brightness) {
> + if (stept_ok)
> + return 1;
> + stepb_ok = 1;
> + }
> + }
> +
> + return 0;
> +}
> +
> +static int reset_pattern(struct pattern_trig_data *data,
> + struct led_classdev *led_cdev)
> +{
> + int brightness;
> +
> + mutex_lock(&data->lock);
> + data->is_sane = 0; /* Prevent pattern_trig_update()
> + from scheduling new work */
> + mutex_unlock(&data->lock);
> +
> + flush_workqueue(data->queue);
> +
> + mutex_lock(&data->lock);
> + if (is_sane(data, &brightness)) {
> + data->curr = data->steps;
> + data->next = data->steps + 1;
> + data->time_ms = 0;
> + data->is_sane = 1;
> + queue_delayed_work(data->queue, &data->dwork, 0);
> + mutex_unlock(&data->lock);
> + return 0;
> + }
> +
> + mutex_unlock(&data->lock);
> + led_set_brightness(led_cdev, brightness);
> + return -EINVAL;
> +}
> +
> +/* --- Sysfs handling --- */
> +
> +static ssize_t pattern_trig_show_repeat(
> + struct device *dev, struct device_attribute *attr, char *buf)
> +{
> + struct led_classdev *led_cdev = dev_get_drvdata(dev);
> + struct pattern_trig_data *data = led_cdev->trigger_data;
> +
> + return scnprintf(buf, PAGE_SIZE, "%d\n", data->repeat);
> +}
> +
> +static ssize_t pattern_trig_store_repeat(
> + struct device *dev, struct device_attribute *attr,
> + const char *buf, size_t count)
> +{
> + struct led_classdev *led_cdev = dev_get_drvdata(dev);
> + struct pattern_trig_data *data = led_cdev->trigger_data;
> + long res;
> + int err;
> +
> + err = kstrtol(buf, 10, &res);
> + if (err)
> + return err;
> +
> + mutex_lock(&data->lock);
> + data->repeat = res < 0 ? -1 : res;
> + mutex_unlock(&data->lock);
> +
> + reset_pattern(data, led_cdev);
> +
> + return count;
> +}
> +
> +DEVICE_ATTR(repeat, S_IRUGO | S_IWUSR,
> + pattern_trig_show_repeat, pattern_trig_store_repeat);
> +
> +static ssize_t pattern_trig_show_pattern(
> + struct device *dev, struct device_attribute *attr, char *buf)
> +{
> + struct led_classdev *led_cdev = dev_get_drvdata(dev);
> + struct pattern_trig_data *data = led_cdev->trigger_data;
> + ssize_t count = 0;
> + int i;
> +
> + if (!data->steps || !data->nsteps)
> + return 0;
> +
> + for (i = 0; i < data->nsteps; i++)
> + count += scnprintf(buf + count, PAGE_SIZE - count,
> + "%d %d" PATTERN_SEPARATOR,
> + data->steps[i].brightness,
> + data->steps[i].time_ms);
> + buf[count - 1] = '\n';
> + buf[count] = '\0';
> +
> + return count + 1;
> +}
> +
> +static ssize_t pattern_trig_store_pattern(
> + struct device *dev, struct device_attribute *attr,
> + const char *buf, size_t count)
> +{
> + struct led_classdev *led_cdev = dev_get_drvdata(dev);
> + struct pattern_trig_data *data = led_cdev->trigger_data;
> + int cr = 0; /* Characters read after a successful sscanf call */
> + int ccr = 0; /* Characters read before looking for PATTERN_SEPARATOR */
> + int tcr = 0; /* Total characters read in buf */
> + int ccount; /* Number of successful conversions in a sscanf call */
> +
> + mutex_lock(&data->lock);
> + data->is_sane = 0;
> +
> + for (data->nsteps = 0; data->nsteps < MAX_NSTEPS; data->nsteps++) {
> + cr = 0;
> + ccount = sscanf(buf + tcr, " %d %d %n" PATTERN_SEPARATOR " %n",
> + &data->steps[data->nsteps].brightness,
> + &data->steps[data->nsteps].time_ms, &ccr, &cr);
> +
> + if (!cr) { /* PATTERN_SEPARATOR not reached */
> + if (ccount == 2) {
> + /* Successful conversion before
> + looking for PATTERN_SEPARATOR. */
> + data->nsteps++;
> + tcr += ccr;
> + }
> +
> + if (tcr != count) {
> + /* Invalid syntax */
> + data->nsteps = 0;
> + mutex_unlock(&data->lock);
> + return -EINVAL;
> + }
> +
> + mutex_unlock(&data->lock);
> +
> + if (reset_pattern(data, led_cdev)) {
> + /* Invalid pattern */
> + data->nsteps = 0;
> + return -EINVAL;
> + }
> +
> + return count;
> + }
> +
> + tcr += cr;
> + }
> +
> + /* Shouldn't reach that */
> + WARN(1, "MAX_NSTEP too small. Please report\n");
> + data->nsteps = 0;
> + mutex_unlock(&data->lock);
> + return -EINVAL;
> +}
> +
> +DEVICE_ATTR(pattern, S_IRUGO | S_IWUSR,
> + pattern_trig_show_pattern, pattern_trig_store_pattern);
> +
> +static int pattern_trig_create_sysfs_files(struct device *dev)
> +{
> + int err;
> +
> + err = device_create_file(dev, &dev_attr_repeat);
> + if (err)
> + return err;
> +
> + err = device_create_file(dev, &dev_attr_pattern);
> + if (err)
> + device_remove_file(dev, &dev_attr_repeat);
> +
> + return err;
> +}
> +
> +static void pattern_trig_remove_sysfs_files(struct device *dev)
> +{
> + device_remove_file(dev, &dev_attr_pattern);
> + device_remove_file(dev, &dev_attr_repeat);
> +}
> +
> +/* --- Led intensity updating --- */
> +
> +static int compute_brightness(struct pattern_trig_data *data)
> +{
> + if (data->time_ms == 0)
> + return data->curr->brightness;
> +
> + if (data->curr->time_ms == 0) /* Don't divide by zero */
> + return data->next->brightness;
> +
> + return data->curr->brightness + data->time_ms
> + * (data->next->brightness - data->curr->brightness)
> + / data->curr->time_ms;
> +}
> +
> +static void update_to_next_step(struct pattern_trig_data *data)
> +{
> + data->curr = data->next;
> + if (data->curr == data->steps)
> + data->repeat--;
> +
> + if (data->next == data->steps + data->nsteps - 1)
> + data->next = data->steps;
> + else
> + data->next++;
> +
> + data->time_ms = 0;
> +}
> +
> +static void pattern_trig_update(struct work_struct *work)
> +{
> + struct delayed_work *dwork = to_delayed_work(work);
> + struct pattern_trig_data *data =
> + container_of(dwork, struct pattern_trig_data, dwork);
> +
> + mutex_lock(&data->lock);
> +
> + if (!data->is_sane || !data->repeat) {
> + mutex_unlock(&data->lock);
> + return;
> + }
> +
> + if (data->time_ms > data->curr->time_ms)
> + update_to_next_step(data);
> +
> + /* is_sane() checked that there is at least
> + one step with time_ms >= UPDATE_INTERVAL
> + so we won't go in an infinite loop */
> + while (data->curr->time_ms < UPDATE_INTERVAL)
> + update_to_next_step(data);
> +
> + if (data->next->brightness == data->curr->brightness) {
> + /* Constant brightness for this step */
> + led_set_brightness(data->led_cdev, data->curr->brightness);
> + queue_delayed_work(data->queue, &data->dwork,
> + msecs_to_jiffies(data->curr->time_ms));
> + update_to_next_step(data);
> + } else {
> + /* Gradual dimming */
> + led_set_brightness(data->led_cdev, compute_brightness(data));
> + data->time_ms += UPDATE_INTERVAL;
> + queue_delayed_work(data->queue, &data->dwork,
> + msecs_to_jiffies(UPDATE_INTERVAL));
> + }
> +
> + mutex_unlock(&data->lock);
> +}
> +
> +/* --- Trigger activation --- */
> +
> +static void pattern_trig_activate(struct led_classdev *led_cdev)
> +{
> + struct pattern_trig_data *data = NULL;
> + int err;
> +
> + data = kzalloc(sizeof(*data), GFP_KERNEL);
> + if (!data)
> + return;
> +
> + err = pattern_trig_initialize_data(data);
> + if (err) {
> + kfree(data);
> + return;
> + }
> +
> + led_cdev->trigger_data = data;
> + data->led_cdev = led_cdev;
> + data->queue = alloc_ordered_workqueue("pattern trigger on %s", 0,
> + led_cdev->name ? led_cdev->name : "?");
> + INIT_DELAYED_WORK(&data->dwork, pattern_trig_update);
> + pattern_trig_create_sysfs_files(led_cdev->dev);
> +}
> +
> +static void pattern_trig_deactivate(struct led_classdev *led_cdev)
> +{
> + struct pattern_trig_data *data = led_cdev->trigger_data;
> +
> + if (data) {
> + pattern_trig_remove_sysfs_files(led_cdev->dev);
> +
> + mutex_lock(&data->lock);
> + data->is_sane = 0; /* Prevent pattern_trig_update()
> + from scheduling new work */
> + mutex_unlock(&data->lock);
> +
> + flush_workqueue(data->queue);
> + destroy_workqueue(data->queue);
> +
> + led_set_brightness(led_cdev, LED_OFF);
> + pattern_trig_clear_data(data);
> + kfree(data);
> +
> + led_cdev->trigger_data = NULL;
> + }
> +}
> +
> +static struct led_trigger pattern_led_trigger = {
> + .name = "pattern",
> + .activate = pattern_trig_activate,
> + .deactivate = pattern_trig_deactivate,
> +};
> +
> +/* --- Module loading/unloading --- */
> +
> +static int __init pattern_trig_init(void)
> +{
> + return led_trigger_register(&pattern_led_trigger);
> +}
> +
> +static void __exit pattern_trig_exit(void)
> +{
> + led_trigger_unregister(&pattern_led_trigger);
> +}
> +
> +module_init(pattern_trig_init);
> +module_exit(pattern_trig_exit);
> +
> +MODULE_AUTHOR("Raphael Teysseyre <rteysseyre@xxxxxxxxx");
> +MODULE_DESCRIPTION("Statuscode LED trigger");
> +MODULE_LICENSE("GPL v2");
> --
> 1.7.1
>
>
>