[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8447d6730901281120k1153f682oa4e8f3e6249eee7b@mail.gmail.com>
Date: Wed, 28 Jan 2009 20:20:07 +0100
From: Davide Rizzo <elpa.rizzo@...il.com>
To: Greg KH <gregkh@...e.de>
Cc: "Hans J. Koch" <hjk@...utronix.de>, linux-kernel@...r.kernel.org,
ben-linux@...ff.org, linux-api@...r.kernel.org
Subject: Re: [PATCH 1/3] Driver for user access to internal clocks
> Same comments here as for your "timer.c" file:
> - too generic of a name
> - needs proper documentation
> - send any kernel api changes also to linux-api@...r.kernel.org
>
> thanks,
>
> greg k-h
>
Here is the modified user mode clock access driver.
I moved it in /drivers/misc and reorganized virtual files under /sys/class/clock
This driver is for user level programs to interact with system clocks.
It allows to read and modify rates and parents, using virtual files.
It requires the implementation of 2 additional functions in the clk interface:
clk_for_each() and clk_name().
Actually I implemented that functions only for Samsung S3C24xx platform.
Signed-off-by: Davide Rizzo <elpa.rizzo@...il.com>
---
diff -urpN linux-2.6.29-rc2/drivers/misc/Kconfig
linux-2.6.29-rc2.elpa/drivers/misc/Kconfig
--- linux-2.6.29-rc2/drivers/misc/Kconfig 2009-01-17 11:32:40.000000000 +0100
+++ linux-2.6.29-rc2.elpa/drivers/misc/Kconfig 2009-01-28
19:21:43.000000000 +0100
@@ -232,4 +232,19 @@ config DELL_LAPTOP
source "drivers/misc/c2port/Kconfig"
+config UM_CLOCK
+ bool "User mode clock driver"
+ depends on SYSFS
+ default n
+ ---help---
+ Allows user mode programs to configure and control internal clocks'
+ rates and parents through virtual files in sysfs.
+ This driver requires implementation of some additional functions
+ in architecture specific low-level drivers: clk_for_each() and
+ clk_get_name(). Moreover, it requires clk_get() to recognize the
+ "name.id" format
+ Currently they're implemented only on Samsung S3C24xx platforms.
+
+ If unsure, say N.
+
endif # MISC_DEVICES
diff -urpN linux-2.6.29-rc2/drivers/misc/Makefile
linux-2.6.29-rc2.elpa/drivers/misc/Makefile
--- linux-2.6.29-rc2/drivers/misc/Makefile 2009-01-17 11:32:40.000000000 +0100
+++ linux-2.6.29-rc2.elpa/drivers/misc/Makefile 2009-01-28
19:21:43.000000000 +0100
@@ -20,3 +20,4 @@ obj-$(CONFIG_SGI_XP) += sgi-xp/
obj-$(CONFIG_SGI_GRU) += sgi-gru/
obj-$(CONFIG_HP_ILO) += hpilo.o
obj-$(CONFIG_C2PORT) += c2port/
+obj-$(CONFIG_UM_CLOCK) += um_clock.o
diff -urpN linux-2.6.29-rc2/drivers/misc/um_clock.c
linux-2.6.29-rc2.elpa/drivers/misc/um_clock.c
--- linux-2.6.29-rc2/drivers/misc/um_clock.c 1970-01-01 01:00:00.000000000 +0100
+++ linux-2.6.29-rc2.elpa/drivers/misc/um_clock.c 2009-01-28
19:21:43.000000000 +0100
@@ -0,0 +1,174 @@
+/*
+ driver/misc/clock.c
+
+ Written Feb 2008 by Davide Rizzo <elpa.rizzo@...il.com>
+
+ This driver allows to read and modify internal clocks' rates using
+ virtual files. User can also read and modify parents.
+
+ This driver requires implementation of clk_name() and clk_enum() functions
+ in architecture specific clock.c
+
+ This program is free software; you can redistribute it and/or modify
+ it under the terms of the GNU General Public License as published by
+ the Free Software Foundation; either version 2 of the License, or
+ (at your option) any later version.
+
+ This program is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ GNU General Public License for more details.
+
+ You should have received a copy of the GNU General Public License
+ along with this program; if not, write to the Free Software
+ Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+*/
+
+#include <linux/device.h>
+#include <linux/clk.h>
+#include <linux/kdev_t.h>
+#include <linux/err.h>
+
+static DEFINE_MUTEX(mutex);
+
+static ssize_t rate_show(struct device *dev, struct device_attribute *attr,
+ char *buffer)
+{
+ unsigned long rate;
+ struct clk *clk;
+
+ clk = dev_get_drvdata(dev);
+ rate = 0;
+
+ mutex_lock(&mutex);
+
+ rate = clk_get_rate(clk);
+
+ mutex_unlock(&mutex);
+
+ return sprintf(buffer, "%ld\n", rate);
+}
+
+static ssize_t rate_store(struct device *dev, struct device_attribute *attr,
+ const char *buffer, size_t count)
+{
+ unsigned long rate;
+ struct clk *clk;
+ int err;
+
+ clk = dev_get_drvdata(dev);
+ err = strict_strtoul(buffer, 10, &rate);
+ if (err)
+ return err;
+
+ mutex_lock(&mutex);
+
+ if (rate != 0) {
+ clk_set_rate(clk, clk_round_rate(clk, rate));
+ clk_enable(clk);
+ } else
+ clk_disable(clk);
+
+ mutex_unlock(&mutex);
+
+ return count;
+}
+
+static ssize_t parent_show(struct device *dev,
+ struct device_attribute *attr, char *buffer)
+{
+ struct clk *parent, *clk;
+
+ clk = dev_get_drvdata(dev);
+
+ mutex_lock(&mutex);
+
+ parent = clk_get_parent(clk);
+ if (parent && !IS_ERR(parent)) {
+ const char *name = clk_get_name(parent);
+ if (IS_ERR(name)) {
+ mutex_unlock(&mutex);
+ return PTR_ERR(name);
+ }
+ strlcpy(buffer, name, PAGE_SIZE);
+ } else
+ buffer[0] = '\0';
+ strlcat(buffer, "\n", PAGE_SIZE);
+
+ mutex_unlock(&mutex);
+
+ return strlen(buffer);
+}
+
+static ssize_t parent_store(struct device *dev,
+ struct device_attribute *attr, const char *buffer, size_t count)
+{
+ struct clk *parent;
+
+ char *s = kstrdup(buffer, GFP_KERNEL);
+
+ if (!s)
+ return -ENOMEM;
+ if (s[strlen(s) - 1] == '\n')
+ s[strlen(s) - 1] = '\0';
+ parent = clk_get(dev, s);
+ kfree(s);
+ if (IS_ERR(parent))
+ return PTR_ERR(parent);
+
+ mutex_lock(&mutex);
+
+ clk_set_parent(dev_get_drvdata(dev), parent);
+
+ mutex_unlock(&mutex);
+
+ clk_put(parent);
+ return count;
+}
+
+static DEVICE_ATTR(rate, S_IRUGO | S_IWUSR, rate_show, rate_store);
+static DEVICE_ATTR(parent, S_IRUGO | S_IWUSR, parent_show, parent_store);
+
+static const struct attribute *clock_attrs[] = {
+ &dev_attr_rate.attr,
+ &dev_attr_parent.attr,
+ NULL,
+};
+
+static const struct attribute_group clock_attr_group = {
+ .attrs = (struct attribute **) clock_attrs,
+};
+
+static struct class gpclock_class = {
+ .name = "clock",
+ .owner = THIS_MODULE,
+};
+
+static int create_clock_attr(struct clk *clk, void *data)
+{
+ struct device *dev;
+ const char *name;
+
+ name = clk_get_name(clk);
+ if (name == NULL || IS_ERR(name)) {
+ pr_debug("Invalid clock's name\n");
+ return PTR_ERR(name);
+ }
+ dev = device_create(&gpclock_class, (struct device *)data, MKDEV(0, 0),
+ clk, name);
+ if (!dev)
+ return -ENODEV;
+ return sysfs_create_group(&dev->kobj, &clock_attr_group);
+}
+
+static int __init gpclock_init(void)
+{
+ int ret;
+
+ ret = class_register(&gpclock_class);
+ if (ret >= 0)
+ ret = clk_for_each(create_clock_attr, NULL);
+ return ret;
+}
+subsys_initcall(gpclock_init);
+
diff -urpN linux-2.6.29-rc2/Documentation/ABI/testing/sysfs-class-clock
linux-2.6.29-rc2.elpa/Documentation/ABI/testing/sysfs-class-clock
--- linux-2.6.29-rc2/Documentation/ABI/testing/sysfs-class-clock 1970-01-01
01:00:00.000000000 +0100
+++ linux-2.6.29-rc2.elpa/Documentation/ABI/testing/sysfs-class-clock 2009-01-28
19:21:43.000000000 +0100
@@ -0,0 +1,20 @@
+What: /sys/class/clock
+Date: January 2009
+Contact: Davide Rizzo <elpa.rizzo@...il.com>
+Description:
+ The /sys/class/clock directory will consist of a group of
+ subdirectories each one describing an internal clock in the kernel.
+ The name of the subdirectory is the clock's name.
+ If a clock is device specific, its name is in the form "name.id".
+ In each directory there are 2 virtual files: rate and parent.
+ Both these files are r/w, but there can be some architecture-dependent
+ limitations on write access on one or both of them.
+ For this interface to work, it needs some low-level implementations (look
+ at include/linux/clk.h):
+ - clk_for_each()
+ - clk_name()
+ - clk_get() should recognize also the "name.id" format
+ Actually I implemented these only for Samsung S3C SoC in
+ arch/arm/plat-s3c/clock.c
+
+Users:
--
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/
Powered by blists - more mailing lists