* [PATCH v2 00/10] Add PWM framework and device-tree support.
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
This patch series adds very rudimentary device-tree support for PWM devices.
I fully realize that this is early work and it should not be merged as is.
However I wanted to post these patches for review to make sure I'm not on a
wild-goose chase.
With all of these patches applied (plus one board-specific patch that is not
included), I'm able to control the backlight on the device I'm working on
using the sysfs interface provided by the pwm-backlight driver and the
backlight class.
This series is based on Sascha Hauer's series of patches[0] to add a generic
PWM framework. The first patch in this series is taken from Sascha's branch,
while the second patch enables each PWM chip to provide multiple PWM devices
(none of the drivers in Sascha's branch are converted yet). When this series
is ready I think it would be best to merge patches 1 and 2.
Patch 3 adds some code to lookup a PWM chip given its device-tree handle. This
code will be used later on by the pwm-backlight driver to find the PWM device
that it should be using. Device tree binding documentation is also provided.
Patch 4 was taken from the Chromium tree and is required to provide proper
clocking of the Tegra2 PWFM controller. All Chromium-specific tags have been
removed from the commit message.
Patch 5 cleans up the clock registration for Tegra2 because patch 6 will only
instantiate one device for the PWFM controller instead of four.
Patch 6 adds a generic PWM framework driver for the Tegra2 PWFM controller.
The code is taken from the Chromium tree with some adjustments to integrate it
with the PWM framework.
Patch 7 implements device tree based probing for the Tegra2 PWFM controller.
Patches 8 and 9 are ports of the Blackfin PWM and the PXA PWM drivers to the
PWM framework. These are only compile-tested as I do not have any hardware to
test them on. They are meant as test-bed for the framework. I intend to port
other drivers over to the framework for the next spin.
Patch 10 implements DT-based probing in the pwm-backlight driver. Note that
this code only handles the "pwm" property (by looking up the PWM device via
the new PWM DT binding) and the "default-" and "max-brightness" properties.
Switching power to the backlight via GPIOs is not supported yet.
The whole series is based on the linux-next tree from 20120202. I think I've
addressed all of the concerns raised in the first version, and those I
haven't yet were still open for discussion (or I've explicitly marked them as
TODO).
Thierry
[0]: http://git.pengutronix.de/?p=imx/linux-2.6.git;a=shortlog;h=refs/heads/pwmlib
Sascha Hauer (1):
PWM: add pwm framework support
Simon Que (1):
arm/tegra: Fix PWM clock programming
Thierry Reding (8):
pwm: Allow chips to support multiple PWMs.
of: Add PWM support.
arm/tegra: Provide clock for only one PWM controller
pwm: Add NVIDIA Tegra SoC support
arm/tegra: Add PWFM controller device tree probing
pwm: Add Blackfin support
pwm: Add PXA support
pwm-backlight: Add rudimentary device tree support
Documentation/devicetree/bindings/pwm/pwm.txt | 50 +++
.../bindings/video/backlight/pwm-backlight | 16 +
Documentation/pwm.txt | 56 +++
MAINTAINERS | 6 +
arch/arm/boot/dts/tegra20.dtsi | 6 +
arch/arm/boot/dts/tegra30.dtsi | 6 +
arch/arm/mach-tegra/board-dt-tegra20.c | 1 +
arch/arm/mach-tegra/board-dt-tegra30.c | 1 +
arch/arm/mach-tegra/clock.h | 1 +
arch/arm/mach-tegra/tegra2_clocks.c | 33 ++-
arch/arm/plat-pxa/Makefile | 1 -
arch/arm/plat-pxa/pwm.c | 304 ----------------
arch/blackfin/Kconfig | 10 -
arch/blackfin/kernel/Makefile | 1 -
arch/blackfin/kernel/pwm.c | 100 ------
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/of/Kconfig | 6 +
drivers/of/Makefile | 1 +
drivers/of/pwm.c | 130 +++++++
drivers/pwm/Kconfig | 40 +++
drivers/pwm/Makefile | 4 +
drivers/pwm/core.c | 367 ++++++++++++++++++++
drivers/pwm/pwm-bfin.c | 164 +++++++++
drivers/pwm/pwm-pxa.c | 243 +++++++++++++
drivers/pwm/pwm-tegra.c | 266 ++++++++++++++
drivers/video/backlight/Kconfig | 2 +-
drivers/video/backlight/pwm_bl.c | 81 ++++-
include/linux/of_pwm.h | 51 +++
include/linux/pwm.h | 67 ++++
30 files changed, 1584 insertions(+), 433 deletions(-)
create mode 100644 Documentation/devicetree/bindings/pwm/pwm.txt
create mode 100644 Documentation/devicetree/bindings/video/backlight/pwm-backlight
create mode 100644 Documentation/pwm.txt
delete mode 100644 arch/arm/plat-pxa/pwm.c
delete mode 100644 arch/blackfin/kernel/pwm.c
create mode 100644 drivers/of/pwm.c
create mode 100644 drivers/pwm/Kconfig
create mode 100644 drivers/pwm/Makefile
create mode 100644 drivers/pwm/core.c
create mode 100644 drivers/pwm/pwm-bfin.c
create mode 100644 drivers/pwm/pwm-pxa.c
create mode 100644 drivers/pwm/pwm-tegra.c
create mode 100644 include/linux/of_pwm.h
--
1.7.9
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 00/10] Add PWM framework and device-tree support.
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
This patch series adds very rudimentary device-tree support for PWM devices.
I fully realize that this is early work and it should not be merged as is.
However I wanted to post these patches for review to make sure I'm not on a
wild-goose chase.
With all of these patches applied (plus one board-specific patch that is not
included), I'm able to control the backlight on the device I'm working on
using the sysfs interface provided by the pwm-backlight driver and the
backlight class.
This series is based on Sascha Hauer's series of patches[0] to add a generic
PWM framework. The first patch in this series is taken from Sascha's branch,
while the second patch enables each PWM chip to provide multiple PWM devices
(none of the drivers in Sascha's branch are converted yet). When this series
is ready I think it would be best to merge patches 1 and 2.
Patch 3 adds some code to lookup a PWM chip given its device-tree handle. This
code will be used later on by the pwm-backlight driver to find the PWM device
that it should be using. Device tree binding documentation is also provided.
Patch 4 was taken from the Chromium tree and is required to provide proper
clocking of the Tegra2 PWFM controller. All Chromium-specific tags have been
removed from the commit message.
Patch 5 cleans up the clock registration for Tegra2 because patch 6 will only
instantiate one device for the PWFM controller instead of four.
Patch 6 adds a generic PWM framework driver for the Tegra2 PWFM controller.
The code is taken from the Chromium tree with some adjustments to integrate it
with the PWM framework.
Patch 7 implements device tree based probing for the Tegra2 PWFM controller.
Patches 8 and 9 are ports of the Blackfin PWM and the PXA PWM drivers to the
PWM framework. These are only compile-tested as I do not have any hardware to
test them on. They are meant as test-bed for the framework. I intend to port
other drivers over to the framework for the next spin.
Patch 10 implements DT-based probing in the pwm-backlight driver. Note that
this code only handles the "pwm" property (by looking up the PWM device via
the new PWM DT binding) and the "default-" and "max-brightness" properties.
Switching power to the backlight via GPIOs is not supported yet.
The whole series is based on the linux-next tree from 20120202. I think I've
addressed all of the concerns raised in the first version, and those I
haven't yet were still open for discussion (or I've explicitly marked them as
TODO).
Thierry
[0]: http://git.pengutronix.de/?p=imx/linux-2.6.git;a=shortlog;h=refs/heads/pwmlib
Sascha Hauer (1):
PWM: add pwm framework support
Simon Que (1):
arm/tegra: Fix PWM clock programming
Thierry Reding (8):
pwm: Allow chips to support multiple PWMs.
of: Add PWM support.
arm/tegra: Provide clock for only one PWM controller
pwm: Add NVIDIA Tegra SoC support
arm/tegra: Add PWFM controller device tree probing
pwm: Add Blackfin support
pwm: Add PXA support
pwm-backlight: Add rudimentary device tree support
Documentation/devicetree/bindings/pwm/pwm.txt | 50 +++
.../bindings/video/backlight/pwm-backlight | 16 +
Documentation/pwm.txt | 56 +++
MAINTAINERS | 6 +
arch/arm/boot/dts/tegra20.dtsi | 6 +
arch/arm/boot/dts/tegra30.dtsi | 6 +
arch/arm/mach-tegra/board-dt-tegra20.c | 1 +
arch/arm/mach-tegra/board-dt-tegra30.c | 1 +
arch/arm/mach-tegra/clock.h | 1 +
arch/arm/mach-tegra/tegra2_clocks.c | 33 ++-
arch/arm/plat-pxa/Makefile | 1 -
arch/arm/plat-pxa/pwm.c | 304 ----------------
arch/blackfin/Kconfig | 10 -
arch/blackfin/kernel/Makefile | 1 -
arch/blackfin/kernel/pwm.c | 100 ------
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/of/Kconfig | 6 +
drivers/of/Makefile | 1 +
drivers/of/pwm.c | 130 +++++++
drivers/pwm/Kconfig | 40 +++
drivers/pwm/Makefile | 4 +
drivers/pwm/core.c | 367 ++++++++++++++++++++
drivers/pwm/pwm-bfin.c | 164 +++++++++
drivers/pwm/pwm-pxa.c | 243 +++++++++++++
drivers/pwm/pwm-tegra.c | 266 ++++++++++++++
drivers/video/backlight/Kconfig | 2 +-
drivers/video/backlight/pwm_bl.c | 81 ++++-
include/linux/of_pwm.h | 51 +++
include/linux/pwm.h | 67 ++++
30 files changed, 1584 insertions(+), 433 deletions(-)
create mode 100644 Documentation/devicetree/bindings/pwm/pwm.txt
create mode 100644 Documentation/devicetree/bindings/video/backlight/pwm-backlight
create mode 100644 Documentation/pwm.txt
delete mode 100644 arch/arm/plat-pxa/pwm.c
delete mode 100644 arch/blackfin/kernel/pwm.c
create mode 100644 drivers/of/pwm.c
create mode 100644 drivers/pwm/Kconfig
create mode 100644 drivers/pwm/Makefile
create mode 100644 drivers/pwm/core.c
create mode 100644 drivers/pwm/pwm-bfin.c
create mode 100644 drivers/pwm/pwm-pxa.c
create mode 100644 drivers/pwm/pwm-tegra.c
create mode 100644 include/linux/of_pwm.h
--
1.7.9
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 01/10] PWM: add pwm framework support
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: Sascha Hauer, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
From: Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
This patch adds framework support for PWM (pulse width modulation) devices.
The is a barebone PWM API already in the kernel under include/linux/pwm.h,
but it does not allow for multiple drivers as each of them implements the
pwm_*() functions.
There are other PWM framework patches around from Bill Gatliff. Unlike
his framework this one does not change the existing API for PWMs so that
this framework can act as a drop in replacement for the existing API.
Why another framework?
Several people argue that there should not be another framework for PWMs
but they should be integrated into one of the existing frameworks like led
or hwmon. Unlike these frameworks the PWM framework is agnostic to the
purpose of the PWM. In fact, a PWM can drive a LED, but this makes the
LED framework a user of a PWM, like already done in leds-pwm.c. The gpio
framework also is not suitable for PWMs. Every gpio could be turned into
a PWM using timer based toggling, but on the other hand not every PWM hardware
device can be turned into a gpio due to the lack of hardware capabilities.
This patch does not try to improve the PWM API yet, this could be done in
subsequent patches.
Signed-off-by: Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
Acked-by: Kurt Van Dijck <kurt.van.dijck-/BeEPy95v10@public.gmane.org>
Reviewed-by: Arnd Bergmann <arnd-r2nGTMty4D4@public.gmane.org>
Reviewed-by: Matthias Kaehlcke <matthias-RprLehDfhQ3k1uMJSBkQmQ@public.gmane.org>
---
Documentation/pwm.txt | 56 +++++++++++++
MAINTAINERS | 6 ++
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/pwm/Kconfig | 12 +++
drivers/pwm/Makefile | 1 +
drivers/pwm/core.c | 220 +++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/pwm.h | 37 ++++++++
8 files changed, 335 insertions(+), 0 deletions(-)
create mode 100644 Documentation/pwm.txt
create mode 100644 drivers/pwm/Kconfig
create mode 100644 drivers/pwm/Makefile
create mode 100644 drivers/pwm/core.c
diff --git a/Documentation/pwm.txt b/Documentation/pwm.txt
new file mode 100644
index 0000000..c7c5cb1
--- /dev/null
+++ b/Documentation/pwm.txt
@@ -0,0 +1,56 @@
+Pulse Width Modulation (PWM) interface
+
+This provides an overview about the Linux PWM interface
+
+PWMs are commonly used for controlling LEDs, fans or vibrators in
+cell phones. PWMs with a fixed purpose have no need implementing
+the Linux PWM API (although they could). However, PWMs are often
+found as discrete devices on SoCs which have no fixed purpose. It's
+up to the board designer to connect them to LEDs or fans. To provide
+this kind of flexibility the generic PWM API exists.
+
+Identifying PWMs
+----------------
+
+PWMs are identified by unique ids throughout the system. A platform
+should call pwmchip_reserve() during init time to reserve the id range
+for internal PWMs so that users have a fixed id to refer to specific
+PWMs.
+
+Using PWMs
+----------
+
+A PWM can be requested using pwm_request() and freed after usage with
+pwm_free(). After being requested a PWM has to be configured using
+
+int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns);
+
+To start/stop toggling the PWM output use pwm_enable()/pwm_disable().
+
+Implementing a PWM driver
+-------------------------
+
+Currently there are two ways to implement pwm drivers. Traditionally
+there only has been the barebone API meaning that each driver has
+to implement the pwm_*() functions itself. This means that it's impossible
+to have multiple PWM drivers in the system. For this reason it's mandatory
+for new drivers to use the generic PWM framework.
+A new PWM device can be added using pwmchip_add() and removed again with
+pwmchip_remove(). pwmchip_add() takes a filled in struct pwm_chip as
+argument which provides the ops and the pwm id to the framework.
+
+Locking
+-------
+
+The PWM core list manipulations are protected by a mutex, so pwm_request()
+and pwm_free() may not be called from an atomic context. Currently the
+PWM core does not enforce any locking to pwm_enable(), pwm_disable() and
+pwm_config(), so the calling context is currently driver specific. This
+is an issue derived from the former barebone API and should be fixed soon.
+
+Helpers
+-------
+
+Currently a PWM can only be configured with period_ns and duty_ns. For several
+use cases freq_hz and duty_percent might be better. Instead of calculating
+this in your driver please consider adding appropriate helpers to the framework.
diff --git a/MAINTAINERS b/MAINTAINERS
index eb8a518..a422dd5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5356,6 +5356,12 @@ S: Maintained
F: Documentation/video4linux/README.pvrusb2
F: drivers/media/video/pvrusb2/
+PPWM core
+M: Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
+L: linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
+S: Maintained
+F: drivers/pwm/
+
PXA2xx/PXA3xx SUPPORT
M: Eric Miao <eric.y.miao-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
M: Russell King <linux-lFZ/pmaqli7XmaaqVzeoHQ@public.gmane.org>
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 03ac866..a88882a 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -144,4 +144,6 @@ source "drivers/devfreq/Kconfig"
source "drivers/modem_shm/Kconfig"
+source "drivers/pwm/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index 2571dff..447ef50 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -8,6 +8,7 @@
# GPIO must come after pinctrl as gpios may need to mux pins etc
obj-y += pinctrl/
obj-y += gpio/
+obj-y += pwm/
obj-$(CONFIG_PCI) += pci/
obj-$(CONFIG_PARISC) += parisc/
obj-$(CONFIG_RAPIDIO) += rapidio/
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
new file mode 100644
index 0000000..93c1052
--- /dev/null
+++ b/drivers/pwm/Kconfig
@@ -0,0 +1,12 @@
+menuconfig PWM
+ bool "PWM Support"
+ help
+ This enables PWM support through the generic PWM framework.
+ You only need to enable this, if you also want to enable
+ one or more of the PWM drivers below.
+
+ If unsure, say N.
+
+if PWM
+
+endif
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
new file mode 100644
index 0000000..3469c3d
--- /dev/null
+++ b/drivers/pwm/Makefile
@@ -0,0 +1 @@
+obj-$(CONFIG_PWM) += core.o
diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
new file mode 100644
index 0000000..71de479
--- /dev/null
+++ b/drivers/pwm/core.c
@@ -0,0 +1,220 @@
+/*
+ * Generic pwmlib implementation
+ *
+ * Copyright (C) 2011 Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
+ *
+ * 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, 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; see the file COPYING. If not, write to
+ * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+#include <linux/module.h>
+#include <linux/pwm.h>
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/device.h>
+
+struct pwm_device {
+ struct pwm_chip *chip;
+ const char *label;
+ unsigned long flags;
+#define FLAG_REQUESTED 0
+#define FLAG_ENABLED 1
+ struct list_head node;
+};
+
+static LIST_HEAD(pwm_list);
+
+static DEFINE_MUTEX(pwm_lock);
+
+static struct pwm_device *_find_pwm(int pwm_id)
+{
+ struct pwm_device *pwm;
+
+ list_for_each_entry(pwm, &pwm_list, node) {
+ if (pwm->chip->pwm_id == pwm_id)
+ return pwm;
+ }
+
+ return NULL;
+}
+
+/**
+ * pwmchip_add() - register a new pwm
+ * @chip: the pwm
+ *
+ * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
+ * a dynamically assigned id will be used, otherwise the id specified,
+ */
+int pwmchip_add(struct pwm_chip *chip)
+{
+ struct pwm_device *pwm;
+ int ret = 0;
+
+ pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
+ if (!pwm)
+ return -ENOMEM;
+
+ pwm->chip = chip;
+
+ mutex_lock(&pwm_lock);
+
+ if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
+ ret = -EBUSY;
+ goto out;
+ }
+
+ list_add_tail(&pwm->node, &pwm_list);
+out:
+ mutex_unlock(&pwm_lock);
+
+ if (ret)
+ kfree(pwm);
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(pwmchip_add);
+
+/**
+ * pwmchip_remove() - remove a pwm
+ * @chip: the pwm
+ *
+ * remove a pwm. This function may return busy if the pwm is still requested.
+ */
+int pwmchip_remove(struct pwm_chip *chip)
+{
+ struct pwm_device *pwm;
+ int ret = 0;
+
+ mutex_lock(&pwm_lock);
+
+ pwm = _find_pwm(chip->pwm_id);
+ if (!pwm) {
+ ret = -ENOENT;
+ goto out;
+ }
+
+ if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
+ ret = -EBUSY;
+ goto out;
+ }
+
+ list_del(&pwm->node);
+
+ kfree(pwm);
+out:
+ mutex_unlock(&pwm_lock);
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(pwmchip_remove);
+
+/*
+ * pwm_request - request a PWM device
+ */
+struct pwm_device *pwm_request(int pwm_id, const char *label)
+{
+ struct pwm_device *pwm;
+ int ret;
+
+ mutex_lock(&pwm_lock);
+
+ pwm = _find_pwm(pwm_id);
+ if (!pwm) {
+ pwm = ERR_PTR(-ENOENT);
+ goto out;
+ }
+
+ if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
+ pwm = ERR_PTR(-EBUSY);
+ goto out;
+ }
+
+ if (!try_module_get(pwm->chip->ops->owner)) {
+ pwm = ERR_PTR(-ENODEV);
+ goto out;
+ }
+
+ if (pwm->chip->ops->request) {
+ ret = pwm->chip->ops->request(pwm->chip);
+ if (ret) {
+ pwm = ERR_PTR(ret);
+ goto out_put;
+ }
+ }
+
+ pwm->label = label;
+ set_bit(FLAG_REQUESTED, &pwm->flags);
+
+ goto out;
+
+out_put:
+ module_put(pwm->chip->ops->owner);
+out:
+ mutex_unlock(&pwm_lock);
+
+ return pwm;
+}
+EXPORT_SYMBOL_GPL(pwm_request);
+
+/*
+ * pwm_free - free a PWM device
+ */
+void pwm_free(struct pwm_device *pwm)
+{
+ mutex_lock(&pwm_lock);
+
+ if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
+ pr_warning("PWM device already freed\n");
+ goto out;
+ }
+
+ pwm->label = NULL;
+
+ module_put(pwm->chip->ops->owner);
+out:
+ mutex_unlock(&pwm_lock);
+}
+EXPORT_SYMBOL_GPL(pwm_free);
+
+/*
+ * pwm_config - change a PWM device configuration
+ */
+int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
+{
+ return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
+}
+EXPORT_SYMBOL_GPL(pwm_config);
+
+/*
+ * pwm_enable - start a PWM output toggling
+ */
+int pwm_enable(struct pwm_device *pwm)
+{
+ if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
+ return pwm->chip->ops->enable(pwm->chip);
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(pwm_enable);
+
+/*
+ * pwm_disable - stop a PWM output toggling
+ */
+void pwm_disable(struct pwm_device *pwm)
+{
+ if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
+ pwm->chip->ops->disable(pwm->chip);
+}
+EXPORT_SYMBOL_GPL(pwm_disable);
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index 7c77575..df9681b 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -28,4 +28,41 @@ int pwm_enable(struct pwm_device *pwm);
*/
void pwm_disable(struct pwm_device *pwm);
+#ifdef CONFIG_PWM
+struct pwm_chip;
+
+/**
+ * struct pwm_ops - PWM operations
+ * @request: optional hook for requesting a PWM
+ * @free: optional hook for freeing a PWM
+ * @config: configure duty cycles and period length for this PWM
+ * @enable: enable PWM output toggling
+ * @disable: disable PWM output toggling
+ */
+struct pwm_ops {
+ int (*request)(struct pwm_chip *chip);
+ void (*free)(struct pwm_chip *chip);
+ int (*config)(struct pwm_chip *chip, int duty_ns,
+ int period_ns);
+ int (*enable)(struct pwm_chip *chip);
+ void (*disable)(struct pwm_chip *chip);
+ struct module *owner;
+};
+
+/**
+ * struct pwm_chip - abstract a PWM
+ * @label: for diagnostics
+ * @owner: helps prevent removal of modules exporting active PWMs
+ * @ops: The callbacks for this PWM
+ */
+struct pwm_chip {
+ int pwm_id;
+ const char *label;
+ struct pwm_ops *ops;
+};
+
+int pwmchip_add(struct pwm_chip *chip);
+int pwmchip_remove(struct pwm_chip *chip);
+#endif
+
#endif /* __LINUX_PWM_H */
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 01/10] PWM: add pwm framework support
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
From: Sascha Hauer <s.hauer@pengutronix.de>
This patch adds framework support for PWM (pulse width modulation) devices.
The is a barebone PWM API already in the kernel under include/linux/pwm.h,
but it does not allow for multiple drivers as each of them implements the
pwm_*() functions.
There are other PWM framework patches around from Bill Gatliff. Unlike
his framework this one does not change the existing API for PWMs so that
this framework can act as a drop in replacement for the existing API.
Why another framework?
Several people argue that there should not be another framework for PWMs
but they should be integrated into one of the existing frameworks like led
or hwmon. Unlike these frameworks the PWM framework is agnostic to the
purpose of the PWM. In fact, a PWM can drive a LED, but this makes the
LED framework a user of a PWM, like already done in leds-pwm.c. The gpio
framework also is not suitable for PWMs. Every gpio could be turned into
a PWM using timer based toggling, but on the other hand not every PWM hardware
device can be turned into a gpio due to the lack of hardware capabilities.
This patch does not try to improve the PWM API yet, this could be done in
subsequent patches.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Acked-by: Kurt Van Dijck <kurt.van.dijck@eia.be>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Matthias Kaehlcke <matthias@kaehlcke.net>
---
Documentation/pwm.txt | 56 +++++++++++++
MAINTAINERS | 6 ++
drivers/Kconfig | 2 +
drivers/Makefile | 1 +
drivers/pwm/Kconfig | 12 +++
drivers/pwm/Makefile | 1 +
drivers/pwm/core.c | 220 +++++++++++++++++++++++++++++++++++++++++++++++++
include/linux/pwm.h | 37 ++++++++
8 files changed, 335 insertions(+), 0 deletions(-)
create mode 100644 Documentation/pwm.txt
create mode 100644 drivers/pwm/Kconfig
create mode 100644 drivers/pwm/Makefile
create mode 100644 drivers/pwm/core.c
diff --git a/Documentation/pwm.txt b/Documentation/pwm.txt
new file mode 100644
index 0000000..c7c5cb1
--- /dev/null
+++ b/Documentation/pwm.txt
@@ -0,0 +1,56 @@
+Pulse Width Modulation (PWM) interface
+
+This provides an overview about the Linux PWM interface
+
+PWMs are commonly used for controlling LEDs, fans or vibrators in
+cell phones. PWMs with a fixed purpose have no need implementing
+the Linux PWM API (although they could). However, PWMs are often
+found as discrete devices on SoCs which have no fixed purpose. It's
+up to the board designer to connect them to LEDs or fans. To provide
+this kind of flexibility the generic PWM API exists.
+
+Identifying PWMs
+----------------
+
+PWMs are identified by unique ids throughout the system. A platform
+should call pwmchip_reserve() during init time to reserve the id range
+for internal PWMs so that users have a fixed id to refer to specific
+PWMs.
+
+Using PWMs
+----------
+
+A PWM can be requested using pwm_request() and freed after usage with
+pwm_free(). After being requested a PWM has to be configured using
+
+int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns);
+
+To start/stop toggling the PWM output use pwm_enable()/pwm_disable().
+
+Implementing a PWM driver
+-------------------------
+
+Currently there are two ways to implement pwm drivers. Traditionally
+there only has been the barebone API meaning that each driver has
+to implement the pwm_*() functions itself. This means that it's impossible
+to have multiple PWM drivers in the system. For this reason it's mandatory
+for new drivers to use the generic PWM framework.
+A new PWM device can be added using pwmchip_add() and removed again with
+pwmchip_remove(). pwmchip_add() takes a filled in struct pwm_chip as
+argument which provides the ops and the pwm id to the framework.
+
+Locking
+-------
+
+The PWM core list manipulations are protected by a mutex, so pwm_request()
+and pwm_free() may not be called from an atomic context. Currently the
+PWM core does not enforce any locking to pwm_enable(), pwm_disable() and
+pwm_config(), so the calling context is currently driver specific. This
+is an issue derived from the former barebone API and should be fixed soon.
+
+Helpers
+-------
+
+Currently a PWM can only be configured with period_ns and duty_ns. For several
+use cases freq_hz and duty_percent might be better. Instead of calculating
+this in your driver please consider adding appropriate helpers to the framework.
diff --git a/MAINTAINERS b/MAINTAINERS
index eb8a518..a422dd5 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5356,6 +5356,12 @@ S: Maintained
F: Documentation/video4linux/README.pvrusb2
F: drivers/media/video/pvrusb2/
+PPWM core
+M: Sascha Hauer <s.hauer@pengutronix.de>
+L: linux-kernel at vger.kernel.org
+S: Maintained
+F: drivers/pwm/
+
PXA2xx/PXA3xx SUPPORT
M: Eric Miao <eric.y.miao@gmail.com>
M: Russell King <linux@arm.linux.org.uk>
diff --git a/drivers/Kconfig b/drivers/Kconfig
index 03ac866..a88882a 100644
--- a/drivers/Kconfig
+++ b/drivers/Kconfig
@@ -144,4 +144,6 @@ source "drivers/devfreq/Kconfig"
source "drivers/modem_shm/Kconfig"
+source "drivers/pwm/Kconfig"
+
endmenu
diff --git a/drivers/Makefile b/drivers/Makefile
index 2571dff..447ef50 100644
--- a/drivers/Makefile
+++ b/drivers/Makefile
@@ -8,6 +8,7 @@
# GPIO must come after pinctrl as gpios may need to mux pins etc
obj-y += pinctrl/
obj-y += gpio/
+obj-y += pwm/
obj-$(CONFIG_PCI) += pci/
obj-$(CONFIG_PARISC) += parisc/
obj-$(CONFIG_RAPIDIO) += rapidio/
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
new file mode 100644
index 0000000..93c1052
--- /dev/null
+++ b/drivers/pwm/Kconfig
@@ -0,0 +1,12 @@
+menuconfig PWM
+ bool "PWM Support"
+ help
+ This enables PWM support through the generic PWM framework.
+ You only need to enable this, if you also want to enable
+ one or more of the PWM drivers below.
+
+ If unsure, say N.
+
+if PWM
+
+endif
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
new file mode 100644
index 0000000..3469c3d
--- /dev/null
+++ b/drivers/pwm/Makefile
@@ -0,0 +1 @@
+obj-$(CONFIG_PWM) += core.o
diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
new file mode 100644
index 0000000..71de479
--- /dev/null
+++ b/drivers/pwm/core.c
@@ -0,0 +1,220 @@
+/*
+ * Generic pwmlib implementation
+ *
+ * Copyright (C) 2011 Sascha Hauer <s.hauer@pengutronix.de>
+ *
+ * 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, 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; see the file COPYING. If not, write to
+ * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ */
+#include <linux/module.h>
+#include <linux/pwm.h>
+#include <linux/list.h>
+#include <linux/mutex.h>
+#include <linux/err.h>
+#include <linux/slab.h>
+#include <linux/device.h>
+
+struct pwm_device {
+ struct pwm_chip *chip;
+ const char *label;
+ unsigned long flags;
+#define FLAG_REQUESTED 0
+#define FLAG_ENABLED 1
+ struct list_head node;
+};
+
+static LIST_HEAD(pwm_list);
+
+static DEFINE_MUTEX(pwm_lock);
+
+static struct pwm_device *_find_pwm(int pwm_id)
+{
+ struct pwm_device *pwm;
+
+ list_for_each_entry(pwm, &pwm_list, node) {
+ if (pwm->chip->pwm_id == pwm_id)
+ return pwm;
+ }
+
+ return NULL;
+}
+
+/**
+ * pwmchip_add() - register a new pwm
+ * @chip: the pwm
+ *
+ * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
+ * a dynamically assigned id will be used, otherwise the id specified,
+ */
+int pwmchip_add(struct pwm_chip *chip)
+{
+ struct pwm_device *pwm;
+ int ret = 0;
+
+ pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
+ if (!pwm)
+ return -ENOMEM;
+
+ pwm->chip = chip;
+
+ mutex_lock(&pwm_lock);
+
+ if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
+ ret = -EBUSY;
+ goto out;
+ }
+
+ list_add_tail(&pwm->node, &pwm_list);
+out:
+ mutex_unlock(&pwm_lock);
+
+ if (ret)
+ kfree(pwm);
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(pwmchip_add);
+
+/**
+ * pwmchip_remove() - remove a pwm
+ * @chip: the pwm
+ *
+ * remove a pwm. This function may return busy if the pwm is still requested.
+ */
+int pwmchip_remove(struct pwm_chip *chip)
+{
+ struct pwm_device *pwm;
+ int ret = 0;
+
+ mutex_lock(&pwm_lock);
+
+ pwm = _find_pwm(chip->pwm_id);
+ if (!pwm) {
+ ret = -ENOENT;
+ goto out;
+ }
+
+ if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
+ ret = -EBUSY;
+ goto out;
+ }
+
+ list_del(&pwm->node);
+
+ kfree(pwm);
+out:
+ mutex_unlock(&pwm_lock);
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(pwmchip_remove);
+
+/*
+ * pwm_request - request a PWM device
+ */
+struct pwm_device *pwm_request(int pwm_id, const char *label)
+{
+ struct pwm_device *pwm;
+ int ret;
+
+ mutex_lock(&pwm_lock);
+
+ pwm = _find_pwm(pwm_id);
+ if (!pwm) {
+ pwm = ERR_PTR(-ENOENT);
+ goto out;
+ }
+
+ if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
+ pwm = ERR_PTR(-EBUSY);
+ goto out;
+ }
+
+ if (!try_module_get(pwm->chip->ops->owner)) {
+ pwm = ERR_PTR(-ENODEV);
+ goto out;
+ }
+
+ if (pwm->chip->ops->request) {
+ ret = pwm->chip->ops->request(pwm->chip);
+ if (ret) {
+ pwm = ERR_PTR(ret);
+ goto out_put;
+ }
+ }
+
+ pwm->label = label;
+ set_bit(FLAG_REQUESTED, &pwm->flags);
+
+ goto out;
+
+out_put:
+ module_put(pwm->chip->ops->owner);
+out:
+ mutex_unlock(&pwm_lock);
+
+ return pwm;
+}
+EXPORT_SYMBOL_GPL(pwm_request);
+
+/*
+ * pwm_free - free a PWM device
+ */
+void pwm_free(struct pwm_device *pwm)
+{
+ mutex_lock(&pwm_lock);
+
+ if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
+ pr_warning("PWM device already freed\n");
+ goto out;
+ }
+
+ pwm->label = NULL;
+
+ module_put(pwm->chip->ops->owner);
+out:
+ mutex_unlock(&pwm_lock);
+}
+EXPORT_SYMBOL_GPL(pwm_free);
+
+/*
+ * pwm_config - change a PWM device configuration
+ */
+int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
+{
+ return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
+}
+EXPORT_SYMBOL_GPL(pwm_config);
+
+/*
+ * pwm_enable - start a PWM output toggling
+ */
+int pwm_enable(struct pwm_device *pwm)
+{
+ if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
+ return pwm->chip->ops->enable(pwm->chip);
+
+ return 0;
+}
+EXPORT_SYMBOL_GPL(pwm_enable);
+
+/*
+ * pwm_disable - stop a PWM output toggling
+ */
+void pwm_disable(struct pwm_device *pwm)
+{
+ if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
+ pwm->chip->ops->disable(pwm->chip);
+}
+EXPORT_SYMBOL_GPL(pwm_disable);
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index 7c77575..df9681b 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -28,4 +28,41 @@ int pwm_enable(struct pwm_device *pwm);
*/
void pwm_disable(struct pwm_device *pwm);
+#ifdef CONFIG_PWM
+struct pwm_chip;
+
+/**
+ * struct pwm_ops - PWM operations
+ * @request: optional hook for requesting a PWM
+ * @free: optional hook for freeing a PWM
+ * @config: configure duty cycles and period length for this PWM
+ * @enable: enable PWM output toggling
+ * @disable: disable PWM output toggling
+ */
+struct pwm_ops {
+ int (*request)(struct pwm_chip *chip);
+ void (*free)(struct pwm_chip *chip);
+ int (*config)(struct pwm_chip *chip, int duty_ns,
+ int period_ns);
+ int (*enable)(struct pwm_chip *chip);
+ void (*disable)(struct pwm_chip *chip);
+ struct module *owner;
+};
+
+/**
+ * struct pwm_chip - abstract a PWM
+ * @label: for diagnostics
+ * @owner: helps prevent removal of modules exporting active PWMs
+ * @ops: The callbacks for this PWM
+ */
+struct pwm_chip {
+ int pwm_id;
+ const char *label;
+ struct pwm_ops *ops;
+};
+
+int pwmchip_add(struct pwm_chip *chip);
+int pwmchip_remove(struct pwm_chip *chip);
+#endif
+
#endif /* __LINUX_PWM_H */
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
This commit modifies the PWM core to support multiple PWMs per struct
pwm_chip. It achieves this in a similar way to how gpiolib works, by
allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
starting at a given offset (pwm_chip.base >= 0). A chip specifies how
many PWMs it controls using the npwm member. Each of the functions in
the pwm_ops structure gets an additional argument that specified the PWM
number (it can be converted to a per-chip index by subtracting the
chip's base).
The total maximum number of PWM devices is currently fixed to 64, but
can easily be made configurable via Kconfig.
The patch is incomplete in that it doesn't convert any existing drivers
that are now broken.
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
Changes in v2:
- add 'struct device *dev' field to pwm_chip, drop label field
- use radix tree for PWM descriptors
- add pwm_set_chip_data() and pwm_get_chip_data() functions to allow
storing and retrieving chip-specific per-PWM data
TODO:
- pass chip-indexed PWM number (PWM descriptor?) to drivers
- merge with Sascha's patch
drivers/pwm/core.c | 287 ++++++++++++++++++++++++++++++++++++++-------------
include/linux/pwm.h | 37 +++++--
2 files changed, 242 insertions(+), 82 deletions(-)
diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
index 71de479..a223bd6 100644
--- a/drivers/pwm/core.c
+++ b/drivers/pwm/core.c
@@ -17,154 +17,292 @@
* along with this program; see the file COPYING. If not, write to
* the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
*/
+
#include <linux/module.h>
+#include <linux/of_pwm.h>
#include <linux/pwm.h>
+#include <linux/radix-tree.h>
#include <linux/list.h>
#include <linux/mutex.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/device.h>
+#define MAX_PWMS 1024
+
struct pwm_device {
- struct pwm_chip *chip;
- const char *label;
+ const char *label;
+ unsigned int pwm;
+};
+
+struct pwm_desc {
+ struct pwm_chip *chip;
+ void *chip_data;
unsigned long flags;
#define FLAG_REQUESTED 0
#define FLAG_ENABLED 1
- struct list_head node;
+ struct pwm_device pwm;
};
-static LIST_HEAD(pwm_list);
-
static DEFINE_MUTEX(pwm_lock);
+static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
+static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
+
+static struct pwm_desc *pwm_to_desc(unsigned int pwm)
+{
+ return radix_tree_lookup(&pwm_desc_tree, pwm);
+}
+
+static struct pwm_desc *alloc_desc(unsigned int pwm)
+{
+ struct pwm_desc *desc;
+
+ desc = kzalloc(sizeof(*desc), GFP_KERNEL);
+ if (!desc)
+ return NULL;
-static struct pwm_device *_find_pwm(int pwm_id)
+ return desc;
+}
+
+static void free_desc(unsigned int pwm)
+{
+ struct pwm_desc *desc = pwm_to_desc(pwm);
+
+ radix_tree_delete(&pwm_desc_tree, pwm);
+ kfree(desc);
+}
+
+static int alloc_descs(int pwm, unsigned int from, unsigned int count)
{
- struct pwm_device *pwm;
+ unsigned int start;
+ unsigned int i;
+ int ret = 0;
+
+ if (pwm >= 0) {
+ if (from > pwm)
+ return -EINVAL;
+
+ from = pwm;
+ }
+
+ start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
+ count, 0);
- list_for_each_entry(pwm, &pwm_list, node) {
- if (pwm->chip->pwm_id == pwm_id)
- return pwm;
+ if ((pwm >= 0) && (start != pwm))
+ return -EEXIST;
+
+ if (start + count > MAX_PWMS)
+ return -ENOSPC;
+
+ for (i = start; i < start + count; i++) {
+ struct pwm_desc *desc = alloc_desc(i);
+ if (!desc) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ radix_tree_insert(&pwm_desc_tree, i, desc);
}
- return NULL;
+ bitmap_set(allocated_pwms, start, count);
+
+ return start;
+
+err:
+ for (i--; i >= 0; i--)
+ free_desc(i);
+
+ return ret;
+}
+
+static void free_descs(unsigned int from, unsigned int count)
+{
+ unsigned int i;
+
+ for (i = from; i < from + count; i++)
+ free_desc(i);
+
+ bitmap_clear(allocated_pwms, from, count);
}
/**
- * pwmchip_add() - register a new pwm
- * @chip: the pwm
- *
- * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
- * a dynamically assigned id will be used, otherwise the id specified,
+ * pwm_set_chip_data - set private chip data for a PWM
+ * @pwm: PWM number
+ * @data: pointer to chip-specific data
*/
-int pwmchip_add(struct pwm_chip *chip)
+int pwm_set_chip_data(unsigned int pwm, void *data)
{
- struct pwm_device *pwm;
- int ret = 0;
+ struct pwm_desc *desc = pwm_to_desc(pwm);
- pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
- if (!pwm)
- return -ENOMEM;
+ if (!desc)
+ return -EINVAL;
- pwm->chip = chip;
+ desc->chip_data = data;
+ return 0;
+}
+
+/**
+ * pwm_get_chip_data - get private chip data for a PWM
+ * @pwm: PWM number
+ */
+void *pwm_get_chip_data(unsigned int pwm)
+{
+ struct pwm_desc *desc = pwm_to_desc(pwm);
+
+ return desc ? desc->chip_data : NULL;
+}
+
+/**
+ * pwmchip_add() - register a new PWM chip
+ * @chip: the PWM chip to add
+ *
+ * Register a new PWM chip. If pwm->base < 0 then a dynamically assigned base
+ * will be used.
+ */
+int pwmchip_add(struct pwm_chip *chip)
+{
+ unsigned int i;
+ int ret;
mutex_lock(&pwm_lock);
- if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
- ret = -EBUSY;
+ ret = alloc_descs(chip->base, 0, chip->npwm);
+ if (ret < 0)
goto out;
+
+ chip->base = ret;
+ ret = 0;
+
+ /* initialize descriptors */
+ for (i = chip->base; i < chip->base + chip->npwm; i++) {
+ struct pwm_desc *desc = pwm_to_desc(i);
+
+ if (!desc) {
+ pr_debug("pwm: descriptor %u not initialized\n", i);
+ continue;
+ }
+
+ desc->chip = chip;
}
- list_add_tail(&pwm->node, &pwm_list);
+ of_pwmchip_add(chip);
+
out:
mutex_unlock(&pwm_lock);
-
- if (ret)
- kfree(pwm);
-
return ret;
}
EXPORT_SYMBOL_GPL(pwmchip_add);
/**
- * pwmchip_remove() - remove a pwm
- * @chip: the pwm
+ * pwmchip_remove() - remove a PWM chip
+ * @chip: the PWM chip to remove
*
- * remove a pwm. This function may return busy if the pwm is still requested.
+ * Removes a PWM chip. This function may return busy if the PWM chip provides
+ * a PWM device that is still requested.
*/
int pwmchip_remove(struct pwm_chip *chip)
{
- struct pwm_device *pwm;
+ unsigned int i;
int ret = 0;
mutex_lock(&pwm_lock);
- pwm = _find_pwm(chip->pwm_id);
- if (!pwm) {
- ret = -ENOENT;
- goto out;
- }
+ for (i = chip->base; i < chip->base + chip->npwm; i++) {
+ struct pwm_desc *desc = pwm_to_desc(i);
- if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
- ret = -EBUSY;
- goto out;
+ if (test_bit(FLAG_REQUESTED, &desc->flags)) {
+ ret = -EBUSY;
+ goto out;
+ }
}
- list_del(&pwm->node);
+ free_descs(chip->base, chip->npwm);
+ of_pwmchip_remove(chip);
- kfree(pwm);
out:
mutex_unlock(&pwm_lock);
-
return ret;
}
EXPORT_SYMBOL_GPL(pwmchip_remove);
+/**
+ * pwmchip_find() - iterator for locating a specific pwm_chip
+ * @data: data to pass to match function
+ * @match: callback function to check pwm_chip
+ */
+struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
+ void *data))
+{
+ struct pwm_chip *chip = NULL;
+ unsigned long i;
+
+ mutex_lock(&pwm_lock);
+
+ for_each_set_bit(i, allocated_pwms, MAX_PWMS) {
+ struct pwm_desc *desc = pwm_to_desc(i);
+
+ if (!desc || !desc->chip)
+ continue;
+
+ if (match(desc->chip, data)) {
+ chip = desc->chip;
+ break;
+ }
+ }
+
+ mutex_unlock(&pwm_lock);
+
+ return chip;
+}
+EXPORT_SYMBOL_GPL(pwmchip_find);
+
/*
* pwm_request - request a PWM device
*/
-struct pwm_device *pwm_request(int pwm_id, const char *label)
+struct pwm_device *pwm_request(int pwm, const char *label)
{
- struct pwm_device *pwm;
+ struct pwm_device *dev;
+ struct pwm_desc *desc;
int ret;
+ if ((pwm < 0) || (pwm >= MAX_PWMS))
+ return ERR_PTR(-ENOENT);
+
mutex_lock(&pwm_lock);
- pwm = _find_pwm(pwm_id);
- if (!pwm) {
- pwm = ERR_PTR(-ENOENT);
- goto out;
- }
+ desc = pwm_to_desc(pwm);
+ dev = &desc->pwm;
- if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
- pwm = ERR_PTR(-EBUSY);
+ if (test_bit(FLAG_REQUESTED, &desc->flags)) {
+ dev = ERR_PTR(-EBUSY);
goto out;
}
- if (!try_module_get(pwm->chip->ops->owner)) {
- pwm = ERR_PTR(-ENODEV);
+ if (!try_module_get(desc->chip->ops->owner)) {
+ dev = ERR_PTR(-ENODEV);
goto out;
}
- if (pwm->chip->ops->request) {
- ret = pwm->chip->ops->request(pwm->chip);
+ if (desc->chip->ops->request) {
+ ret = desc->chip->ops->request(desc->chip, pwm);
if (ret) {
- pwm = ERR_PTR(ret);
+ dev = ERR_PTR(ret);
goto out_put;
}
}
- pwm->label = label;
- set_bit(FLAG_REQUESTED, &pwm->flags);
+ dev->label = label;
+ dev->pwm = pwm;
+ set_bit(FLAG_REQUESTED, &desc->flags);
goto out;
out_put:
- module_put(pwm->chip->ops->owner);
+ module_put(desc->chip->ops->owner);
out:
mutex_unlock(&pwm_lock);
- return pwm;
+ return dev;
}
EXPORT_SYMBOL_GPL(pwm_request);
@@ -173,16 +311,19 @@ EXPORT_SYMBOL_GPL(pwm_request);
*/
void pwm_free(struct pwm_device *pwm)
{
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
mutex_lock(&pwm_lock);
- if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
+ if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
pr_warning("PWM device already freed\n");
goto out;
}
pwm->label = NULL;
+ pwm->pwm = 0;
- module_put(pwm->chip->ops->owner);
+ module_put(desc->chip->ops->owner);
out:
mutex_unlock(&pwm_lock);
}
@@ -193,7 +334,9 @@ EXPORT_SYMBOL_GPL(pwm_free);
*/
int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
{
- return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+ return desc->chip->ops->config(desc->chip, pwm->pwm, duty_ns,
+ period_ns);
}
EXPORT_SYMBOL_GPL(pwm_config);
@@ -202,8 +345,10 @@ EXPORT_SYMBOL_GPL(pwm_config);
*/
int pwm_enable(struct pwm_device *pwm)
{
- if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
- return pwm->chip->ops->enable(pwm->chip);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
+ if (!test_and_set_bit(FLAG_ENABLED, &desc->flags))
+ return desc->chip->ops->enable(desc->chip, pwm->pwm);
return 0;
}
@@ -214,7 +359,9 @@ EXPORT_SYMBOL_GPL(pwm_enable);
*/
void pwm_disable(struct pwm_device *pwm)
{
- if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
- pwm->chip->ops->disable(pwm->chip);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
+ if (test_and_clear_bit(FLAG_ENABLED, &desc->flags))
+ desc->chip->ops->disable(desc->chip, pwm->pwm);
}
EXPORT_SYMBOL_GPL(pwm_disable);
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index df9681b..0f8d105 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -32,37 +32,50 @@ void pwm_disable(struct pwm_device *pwm);
struct pwm_chip;
/**
- * struct pwm_ops - PWM operations
+ * struct pwm_ops - PWM controller operations
* @request: optional hook for requesting a PWM
* @free: optional hook for freeing a PWM
* @config: configure duty cycles and period length for this PWM
* @enable: enable PWM output toggling
* @disable: disable PWM output toggling
+ * @owner: helps prevent removal of modules exporting active PWMs
*/
struct pwm_ops {
- int (*request)(struct pwm_chip *chip);
- void (*free)(struct pwm_chip *chip);
- int (*config)(struct pwm_chip *chip, int duty_ns,
+ int (*request)(struct pwm_chip *chip,
+ unsigned int pwm);
+ void (*free)(struct pwm_chip *chip,
+ unsigned int pwm);
+ int (*config)(struct pwm_chip *chip,
+ unsigned int pwm, int duty_ns,
int period_ns);
- int (*enable)(struct pwm_chip *chip);
- void (*disable)(struct pwm_chip *chip);
+ int (*enable)(struct pwm_chip *chip,
+ unsigned int pwm);
+ void (*disable)(struct pwm_chip *chip,
+ unsigned int pwm);
struct module *owner;
};
/**
- * struct pwm_chip - abstract a PWM
- * @label: for diagnostics
- * @owner: helps prevent removal of modules exporting active PWMs
- * @ops: The callbacks for this PWM
+ * struct pwm_chip - abstract a PWM controller
+ * @dev: device providing the PWMs
+ * @ops: callbacks for this PWM controller
+ * @base: number of first PWM controlled by this chip
+ * @npwm: number of PWMs controlled by this chip
*/
struct pwm_chip {
- int pwm_id;
- const char *label;
+ struct device *dev;
struct pwm_ops *ops;
+ int base;
+ unsigned int npwm;
};
+int pwm_set_chip_data(unsigned int pwm, void *data);
+void *pwm_get_chip_data(unsigned int pwm);
+
int pwmchip_add(struct pwm_chip *chip);
int pwmchip_remove(struct pwm_chip *chip);
+struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
+ void *data));
#endif
#endif /* __LINUX_PWM_H */
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
This commit modifies the PWM core to support multiple PWMs per struct
pwm_chip. It achieves this in a similar way to how gpiolib works, by
allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
starting at a given offset (pwm_chip.base >= 0). A chip specifies how
many PWMs it controls using the npwm member. Each of the functions in
the pwm_ops structure gets an additional argument that specified the PWM
number (it can be converted to a per-chip index by subtracting the
chip's base).
The total maximum number of PWM devices is currently fixed to 64, but
can easily be made configurable via Kconfig.
The patch is incomplete in that it doesn't convert any existing drivers
that are now broken.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
Changes in v2:
- add 'struct device *dev' field to pwm_chip, drop label field
- use radix tree for PWM descriptors
- add pwm_set_chip_data() and pwm_get_chip_data() functions to allow
storing and retrieving chip-specific per-PWM data
TODO:
- pass chip-indexed PWM number (PWM descriptor?) to drivers
- merge with Sascha's patch
drivers/pwm/core.c | 287 ++++++++++++++++++++++++++++++++++++++-------------
include/linux/pwm.h | 37 +++++--
2 files changed, 242 insertions(+), 82 deletions(-)
diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
index 71de479..a223bd6 100644
--- a/drivers/pwm/core.c
+++ b/drivers/pwm/core.c
@@ -17,154 +17,292 @@
* along with this program; see the file COPYING. If not, write to
* the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
*/
+
#include <linux/module.h>
+#include <linux/of_pwm.h>
#include <linux/pwm.h>
+#include <linux/radix-tree.h>
#include <linux/list.h>
#include <linux/mutex.h>
#include <linux/err.h>
#include <linux/slab.h>
#include <linux/device.h>
+#define MAX_PWMS 1024
+
struct pwm_device {
- struct pwm_chip *chip;
- const char *label;
+ const char *label;
+ unsigned int pwm;
+};
+
+struct pwm_desc {
+ struct pwm_chip *chip;
+ void *chip_data;
unsigned long flags;
#define FLAG_REQUESTED 0
#define FLAG_ENABLED 1
- struct list_head node;
+ struct pwm_device pwm;
};
-static LIST_HEAD(pwm_list);
-
static DEFINE_MUTEX(pwm_lock);
+static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
+static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
+
+static struct pwm_desc *pwm_to_desc(unsigned int pwm)
+{
+ return radix_tree_lookup(&pwm_desc_tree, pwm);
+}
+
+static struct pwm_desc *alloc_desc(unsigned int pwm)
+{
+ struct pwm_desc *desc;
+
+ desc = kzalloc(sizeof(*desc), GFP_KERNEL);
+ if (!desc)
+ return NULL;
-static struct pwm_device *_find_pwm(int pwm_id)
+ return desc;
+}
+
+static void free_desc(unsigned int pwm)
+{
+ struct pwm_desc *desc = pwm_to_desc(pwm);
+
+ radix_tree_delete(&pwm_desc_tree, pwm);
+ kfree(desc);
+}
+
+static int alloc_descs(int pwm, unsigned int from, unsigned int count)
{
- struct pwm_device *pwm;
+ unsigned int start;
+ unsigned int i;
+ int ret = 0;
+
+ if (pwm >= 0) {
+ if (from > pwm)
+ return -EINVAL;
+
+ from = pwm;
+ }
+
+ start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
+ count, 0);
- list_for_each_entry(pwm, &pwm_list, node) {
- if (pwm->chip->pwm_id == pwm_id)
- return pwm;
+ if ((pwm >= 0) && (start != pwm))
+ return -EEXIST;
+
+ if (start + count > MAX_PWMS)
+ return -ENOSPC;
+
+ for (i = start; i < start + count; i++) {
+ struct pwm_desc *desc = alloc_desc(i);
+ if (!desc) {
+ ret = -ENOMEM;
+ goto err;
+ }
+
+ radix_tree_insert(&pwm_desc_tree, i, desc);
}
- return NULL;
+ bitmap_set(allocated_pwms, start, count);
+
+ return start;
+
+err:
+ for (i--; i >= 0; i--)
+ free_desc(i);
+
+ return ret;
+}
+
+static void free_descs(unsigned int from, unsigned int count)
+{
+ unsigned int i;
+
+ for (i = from; i < from + count; i++)
+ free_desc(i);
+
+ bitmap_clear(allocated_pwms, from, count);
}
/**
- * pwmchip_add() - register a new pwm
- * @chip: the pwm
- *
- * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
- * a dynamically assigned id will be used, otherwise the id specified,
+ * pwm_set_chip_data - set private chip data for a PWM
+ * @pwm: PWM number
+ * @data: pointer to chip-specific data
*/
-int pwmchip_add(struct pwm_chip *chip)
+int pwm_set_chip_data(unsigned int pwm, void *data)
{
- struct pwm_device *pwm;
- int ret = 0;
+ struct pwm_desc *desc = pwm_to_desc(pwm);
- pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
- if (!pwm)
- return -ENOMEM;
+ if (!desc)
+ return -EINVAL;
- pwm->chip = chip;
+ desc->chip_data = data;
+ return 0;
+}
+
+/**
+ * pwm_get_chip_data - get private chip data for a PWM
+ * @pwm: PWM number
+ */
+void *pwm_get_chip_data(unsigned int pwm)
+{
+ struct pwm_desc *desc = pwm_to_desc(pwm);
+
+ return desc ? desc->chip_data : NULL;
+}
+
+/**
+ * pwmchip_add() - register a new PWM chip
+ * @chip: the PWM chip to add
+ *
+ * Register a new PWM chip. If pwm->base < 0 then a dynamically assigned base
+ * will be used.
+ */
+int pwmchip_add(struct pwm_chip *chip)
+{
+ unsigned int i;
+ int ret;
mutex_lock(&pwm_lock);
- if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
- ret = -EBUSY;
+ ret = alloc_descs(chip->base, 0, chip->npwm);
+ if (ret < 0)
goto out;
+
+ chip->base = ret;
+ ret = 0;
+
+ /* initialize descriptors */
+ for (i = chip->base; i < chip->base + chip->npwm; i++) {
+ struct pwm_desc *desc = pwm_to_desc(i);
+
+ if (!desc) {
+ pr_debug("pwm: descriptor %u not initialized\n", i);
+ continue;
+ }
+
+ desc->chip = chip;
}
- list_add_tail(&pwm->node, &pwm_list);
+ of_pwmchip_add(chip);
+
out:
mutex_unlock(&pwm_lock);
-
- if (ret)
- kfree(pwm);
-
return ret;
}
EXPORT_SYMBOL_GPL(pwmchip_add);
/**
- * pwmchip_remove() - remove a pwm
- * @chip: the pwm
+ * pwmchip_remove() - remove a PWM chip
+ * @chip: the PWM chip to remove
*
- * remove a pwm. This function may return busy if the pwm is still requested.
+ * Removes a PWM chip. This function may return busy if the PWM chip provides
+ * a PWM device that is still requested.
*/
int pwmchip_remove(struct pwm_chip *chip)
{
- struct pwm_device *pwm;
+ unsigned int i;
int ret = 0;
mutex_lock(&pwm_lock);
- pwm = _find_pwm(chip->pwm_id);
- if (!pwm) {
- ret = -ENOENT;
- goto out;
- }
+ for (i = chip->base; i < chip->base + chip->npwm; i++) {
+ struct pwm_desc *desc = pwm_to_desc(i);
- if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
- ret = -EBUSY;
- goto out;
+ if (test_bit(FLAG_REQUESTED, &desc->flags)) {
+ ret = -EBUSY;
+ goto out;
+ }
}
- list_del(&pwm->node);
+ free_descs(chip->base, chip->npwm);
+ of_pwmchip_remove(chip);
- kfree(pwm);
out:
mutex_unlock(&pwm_lock);
-
return ret;
}
EXPORT_SYMBOL_GPL(pwmchip_remove);
+/**
+ * pwmchip_find() - iterator for locating a specific pwm_chip
+ * @data: data to pass to match function
+ * @match: callback function to check pwm_chip
+ */
+struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
+ void *data))
+{
+ struct pwm_chip *chip = NULL;
+ unsigned long i;
+
+ mutex_lock(&pwm_lock);
+
+ for_each_set_bit(i, allocated_pwms, MAX_PWMS) {
+ struct pwm_desc *desc = pwm_to_desc(i);
+
+ if (!desc || !desc->chip)
+ continue;
+
+ if (match(desc->chip, data)) {
+ chip = desc->chip;
+ break;
+ }
+ }
+
+ mutex_unlock(&pwm_lock);
+
+ return chip;
+}
+EXPORT_SYMBOL_GPL(pwmchip_find);
+
/*
* pwm_request - request a PWM device
*/
-struct pwm_device *pwm_request(int pwm_id, const char *label)
+struct pwm_device *pwm_request(int pwm, const char *label)
{
- struct pwm_device *pwm;
+ struct pwm_device *dev;
+ struct pwm_desc *desc;
int ret;
+ if ((pwm < 0) || (pwm >= MAX_PWMS))
+ return ERR_PTR(-ENOENT);
+
mutex_lock(&pwm_lock);
- pwm = _find_pwm(pwm_id);
- if (!pwm) {
- pwm = ERR_PTR(-ENOENT);
- goto out;
- }
+ desc = pwm_to_desc(pwm);
+ dev = &desc->pwm;
- if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
- pwm = ERR_PTR(-EBUSY);
+ if (test_bit(FLAG_REQUESTED, &desc->flags)) {
+ dev = ERR_PTR(-EBUSY);
goto out;
}
- if (!try_module_get(pwm->chip->ops->owner)) {
- pwm = ERR_PTR(-ENODEV);
+ if (!try_module_get(desc->chip->ops->owner)) {
+ dev = ERR_PTR(-ENODEV);
goto out;
}
- if (pwm->chip->ops->request) {
- ret = pwm->chip->ops->request(pwm->chip);
+ if (desc->chip->ops->request) {
+ ret = desc->chip->ops->request(desc->chip, pwm);
if (ret) {
- pwm = ERR_PTR(ret);
+ dev = ERR_PTR(ret);
goto out_put;
}
}
- pwm->label = label;
- set_bit(FLAG_REQUESTED, &pwm->flags);
+ dev->label = label;
+ dev->pwm = pwm;
+ set_bit(FLAG_REQUESTED, &desc->flags);
goto out;
out_put:
- module_put(pwm->chip->ops->owner);
+ module_put(desc->chip->ops->owner);
out:
mutex_unlock(&pwm_lock);
- return pwm;
+ return dev;
}
EXPORT_SYMBOL_GPL(pwm_request);
@@ -173,16 +311,19 @@ EXPORT_SYMBOL_GPL(pwm_request);
*/
void pwm_free(struct pwm_device *pwm)
{
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
mutex_lock(&pwm_lock);
- if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
+ if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
pr_warning("PWM device already freed\n");
goto out;
}
pwm->label = NULL;
+ pwm->pwm = 0;
- module_put(pwm->chip->ops->owner);
+ module_put(desc->chip->ops->owner);
out:
mutex_unlock(&pwm_lock);
}
@@ -193,7 +334,9 @@ EXPORT_SYMBOL_GPL(pwm_free);
*/
int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
{
- return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+ return desc->chip->ops->config(desc->chip, pwm->pwm, duty_ns,
+ period_ns);
}
EXPORT_SYMBOL_GPL(pwm_config);
@@ -202,8 +345,10 @@ EXPORT_SYMBOL_GPL(pwm_config);
*/
int pwm_enable(struct pwm_device *pwm)
{
- if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
- return pwm->chip->ops->enable(pwm->chip);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
+ if (!test_and_set_bit(FLAG_ENABLED, &desc->flags))
+ return desc->chip->ops->enable(desc->chip, pwm->pwm);
return 0;
}
@@ -214,7 +359,9 @@ EXPORT_SYMBOL_GPL(pwm_enable);
*/
void pwm_disable(struct pwm_device *pwm)
{
- if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
- pwm->chip->ops->disable(pwm->chip);
+ struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
+
+ if (test_and_clear_bit(FLAG_ENABLED, &desc->flags))
+ desc->chip->ops->disable(desc->chip, pwm->pwm);
}
EXPORT_SYMBOL_GPL(pwm_disable);
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index df9681b..0f8d105 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -32,37 +32,50 @@ void pwm_disable(struct pwm_device *pwm);
struct pwm_chip;
/**
- * struct pwm_ops - PWM operations
+ * struct pwm_ops - PWM controller operations
* @request: optional hook for requesting a PWM
* @free: optional hook for freeing a PWM
* @config: configure duty cycles and period length for this PWM
* @enable: enable PWM output toggling
* @disable: disable PWM output toggling
+ * @owner: helps prevent removal of modules exporting active PWMs
*/
struct pwm_ops {
- int (*request)(struct pwm_chip *chip);
- void (*free)(struct pwm_chip *chip);
- int (*config)(struct pwm_chip *chip, int duty_ns,
+ int (*request)(struct pwm_chip *chip,
+ unsigned int pwm);
+ void (*free)(struct pwm_chip *chip,
+ unsigned int pwm);
+ int (*config)(struct pwm_chip *chip,
+ unsigned int pwm, int duty_ns,
int period_ns);
- int (*enable)(struct pwm_chip *chip);
- void (*disable)(struct pwm_chip *chip);
+ int (*enable)(struct pwm_chip *chip,
+ unsigned int pwm);
+ void (*disable)(struct pwm_chip *chip,
+ unsigned int pwm);
struct module *owner;
};
/**
- * struct pwm_chip - abstract a PWM
- * @label: for diagnostics
- * @owner: helps prevent removal of modules exporting active PWMs
- * @ops: The callbacks for this PWM
+ * struct pwm_chip - abstract a PWM controller
+ * @dev: device providing the PWMs
+ * @ops: callbacks for this PWM controller
+ * @base: number of first PWM controlled by this chip
+ * @npwm: number of PWMs controlled by this chip
*/
struct pwm_chip {
- int pwm_id;
- const char *label;
+ struct device *dev;
struct pwm_ops *ops;
+ int base;
+ unsigned int npwm;
};
+int pwm_set_chip_data(unsigned int pwm, void *data);
+void *pwm_get_chip_data(unsigned int pwm);
+
int pwmchip_add(struct pwm_chip *chip);
int pwmchip_remove(struct pwm_chip *chip);
+struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
+ void *data));
#endif
#endif /* __LINUX_PWM_H */
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 03/10] of: Add PWM support.
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: Mark Brown, Sascha Hauer, Colin Cross, Rob Herring,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
This patch adds helpers to support device tree bindings for the generic
PWM API. Device tree binding documentation for PWM controllers is also
provided.
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
Changes in v2:
- add device tree binding documentation
- add of_xlate to parse controller-specific PWM-specifier
Documentation/devicetree/bindings/pwm/pwm.txt | 50 ++++++++++
drivers/of/Kconfig | 6 +
drivers/of/Makefile | 1 +
drivers/of/pwm.c | 130 +++++++++++++++++++++++++
include/linux/of_pwm.h | 51 ++++++++++
include/linux/pwm.h | 17 +++
6 files changed, 255 insertions(+), 0 deletions(-)
create mode 100644 Documentation/devicetree/bindings/pwm/pwm.txt
create mode 100644 drivers/of/pwm.c
create mode 100644 include/linux/of_pwm.h
diff --git a/Documentation/devicetree/bindings/pwm/pwm.txt b/Documentation/devicetree/bindings/pwm/pwm.txt
new file mode 100644
index 0000000..1c34e86
--- /dev/null
+++ b/Documentation/devicetree/bindings/pwm/pwm.txt
@@ -0,0 +1,50 @@
+Specifying PWM information for devices
+======================================
+
+1) PWM user nodes
+-----------------
+
+PWM users should specify a list of PWM devices that they want to use
+with a property containing a 'pwm-list':
+
+ pwm-list ::= <single-pwm> [pwm-list]
+ single-pwm ::= <pwm-phandle> <pwm-specifier>
+ pwm-phandle : phandle to PWM controller node
+ pwm-specifier : array of #pwm-cells specifying the given PWM
+ (controller specific)
+
+PWM properties should be named "[<name>-]pwms". Exact meaning of each
+pwms property must be documented in the device tree binding for each
+device.
+
+The following example could be used to describe a PWM-based backlight
+device:
+
+ pwm: pwm {
+ #pwm-cells = <2>;
+ };
+
+ [...]
+
+ bl: backlight {
+ pwms = <&pwm 0 5000000>;
+ };
+
+pwm-specifier typically encodes the chip-relative PWM number and the PWM
+period in nanoseconds.
+
+2) PWM controller nodes
+-----------------------
+
+PWM controller nodes must specify the number of cells used for the
+specifier using the '#pwm-cells' property.
+
+An example PWM controller might look like this:
+
+ pwm: pwm@7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
+TODO: do we need an empty 'pwm-controller' property?
diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
index 268163d..9635468 100644
--- a/drivers/of/Kconfig
+++ b/drivers/of/Kconfig
@@ -56,6 +56,12 @@ config OF_GPIO
help
OpenFirmware GPIO accessors
+config OF_PWM
+ def_bool y
+ depends on PWM
+ help
+ OpenFirmware PWM accessors
+
config OF_I2C
def_tristate I2C
depends on I2C && !SPARC
diff --git a/drivers/of/Makefile b/drivers/of/Makefile
index a73f5a5..3dd13e3 100644
--- a/drivers/of/Makefile
+++ b/drivers/of/Makefile
@@ -5,6 +5,7 @@ obj-$(CONFIG_OF_ADDRESS) += address.o
obj-$(CONFIG_OF_IRQ) += irq.o
obj-$(CONFIG_OF_DEVICE) += device.o platform.o
obj-$(CONFIG_OF_GPIO) += gpio.o
+obj-$(CONFIG_OF_PWM) += pwm.o
obj-$(CONFIG_OF_I2C) += of_i2c.o
obj-$(CONFIG_OF_NET) += of_net.o
obj-$(CONFIG_OF_SPI) += of_spi.o
diff --git a/drivers/of/pwm.c b/drivers/of/pwm.c
new file mode 100644
index 0000000..d6f7f33
--- /dev/null
+++ b/drivers/of/pwm.c
@@ -0,0 +1,130 @@
+/*
+ * OF helpers for the PWM API
+ *
+ * Copyright (c) 2011 Avionic Design GmbH
+ *
+ * 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.
+ */
+
+#include <linux/errno.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pwm.h>
+
+static int of_pwmchip_is_match(struct pwm_chip *chip, void *data)
+{
+ return chip->dev ? chip->dev->of_node == data : 0;
+}
+
+/**
+ * of_node_to_pwmchip() - finds the PWM chip associated with a device node
+ * @np: device node of the PWM chip
+ *
+ * Returns a pointer to the PWM chip associated with the specified device
+ * node or NULL if the device node doesn't represent a PWM chip.
+ */
+struct pwm_chip *of_node_to_pwmchip(struct device_node *np)
+{
+ return pwmchip_find(np, of_pwmchip_is_match);
+}
+
+int of_pwm_simple_xlate(struct pwm_chip *pc, const struct of_phandle_args *args,
+ struct pwm_spec *spec)
+{
+ if (pc->of_pwm_n_cells < 2)
+ return -EINVAL;
+
+ if (args->args_count < pc->of_pwm_n_cells)
+ return -EINVAL;
+
+ if (args->args[0] >= pc->npwm)
+ return -EINVAL;
+
+ if (spec)
+ spec->period = args->args[1];
+
+ return args->args[0];
+}
+
+void of_pwmchip_add(struct pwm_chip *chip)
+{
+ if (!chip->dev || !chip->dev->of_node)
+ return;
+
+ if (!chip->of_xlate) {
+ chip->of_xlate = of_pwm_simple_xlate;
+ chip->of_pwm_n_cells = 2;
+ }
+
+ of_node_get(chip->dev->of_node);
+}
+
+void of_pwmchip_remove(struct pwm_chip *chip)
+{
+ if (chip->dev && chip->dev->of_node)
+ of_node_put(chip->dev->of_node);
+}
+
+/**
+ * of_get_named_pwm() - get a PWM number and period to use with the PWM API
+ * @np: device node to get the PWM from
+ * @propname: property name containing PWM specifier(s)
+ * @index: index of the PWM
+ * @spec: a pointer to a struct pwm_spec to fill in
+ *
+ * Returns PWM number to use with the Linux generic PWM API or a negative
+ * error code on failure. If @spec is not NULL the function fills in the
+ * values parsed from the device tree.
+ */
+int of_get_named_pwm(struct device_node *np, const char *propname,
+ int index, struct pwm_spec *spec)
+{
+ struct of_phandle_args args;
+ struct pwm_chip *pc;
+ int ret;
+
+ ret = of_parse_phandle_with_args(np, propname, "#pwm-cells", index,
+ &args);
+ if (ret) {
+ pr_debug("%s(): can't parse pwms property\n", __func__);
+ goto out;
+ }
+
+ pc = of_node_to_pwmchip(args.np);
+ if (!pc) {
+ ret = -ENODEV;
+ goto put;
+ }
+
+ if (args.args_count != pc->of_pwm_n_cells) {
+ pr_debug("%s: wrong #pwm-cells for %s\n", np->full_name,
+ args.np->full_name);
+ ret = -EINVAL;
+ goto put;
+ }
+
+ /*
+ * reset the specifier structure since .of_xlate might decide not to
+ * fill it in
+ */
+ if (spec)
+ spec->period = 0;
+
+ ret = pc->of_xlate(pc, &args, spec);
+ if (ret < 0)
+ goto put;
+
+ ret += pc->base;
+
+put:
+ of_node_put(args.np);
+out:
+ pr_debug("%s() exited with status %d\n", __func__, ret);
+ return ret;
+}
+EXPORT_SYMBOL(of_get_named_pwm);
diff --git a/include/linux/of_pwm.h b/include/linux/of_pwm.h
new file mode 100644
index 0000000..a6af951
--- /dev/null
+++ b/include/linux/of_pwm.h
@@ -0,0 +1,51 @@
+/*
+ * OF helpers for the PWM API
+ *
+ * Copyright (c) 2011 Avionic Design GmbH
+ *
+ * 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.
+ */
+
+#ifndef __LINUX_OF_PWM_H
+#define __LINUX_OF_PWM_H
+
+#include <linux/pwm.h>
+
+struct device_node;
+
+#ifdef CONFIG_OF_PWM
+
+struct pwm_chip *of_node_to_pwmchip(struct device_node *np);
+int of_get_named_pwm(struct device_node *np, const char *propname,
+ int index, struct pwm_spec *spec);
+void of_pwmchip_add(struct pwm_chip *pc);
+void of_pwmchip_remove(struct pwm_chip *pc);
+
+#else
+
+static inline struct pwm_chip *of_node_to_pwmchip(struct device_node *np)
+{
+ return NULL;
+}
+
+static inline int of_get_named_pwm(struct device_node *np,
+ const char *propname, int index,
+ unsigned int *period_ns)
+{
+ return -ENOSYS;
+}
+
+static inline void of_pwmchip_add(struct pwm_chip *pc)
+{
+}
+
+static inline void of_pwmchip_remove(struct pwm_chip *pc)
+{
+}
+
+#endif /* CONFIG_OF_PWM */
+
+#endif /* __LINUX_OF_PWM_H */
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index 0f8d105..8595be4 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -1,6 +1,8 @@
#ifndef __LINUX_PWM_H
#define __LINUX_PWM_H
+#include <linux/of.h>
+
struct pwm_device;
/*
@@ -56,6 +58,14 @@ struct pwm_ops {
};
/**
+ * struct pwm_spec - device tree PWM specifier
+ * @period: PWM period (in nanoseconds)
+ */
+struct pwm_spec {
+ unsigned int period;
+};
+
+/**
* struct pwm_chip - abstract a PWM controller
* @dev: device providing the PWMs
* @ops: callbacks for this PWM controller
@@ -67,6 +77,13 @@ struct pwm_chip {
struct pwm_ops *ops;
int base;
unsigned int npwm;
+
+#ifdef CONFIG_OF_PWM
+ int (*of_xlate)(struct pwm_chip *pc,
+ const struct of_phandle_args *args,
+ struct pwm_spec *spec);
+ unsigned int of_pwm_n_cells;
+#endif
};
int pwm_set_chip_data(unsigned int pwm, void *data);
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 03/10] of: Add PWM support.
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
This patch adds helpers to support device tree bindings for the generic
PWM API. Device tree binding documentation for PWM controllers is also
provided.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
Changes in v2:
- add device tree binding documentation
- add of_xlate to parse controller-specific PWM-specifier
Documentation/devicetree/bindings/pwm/pwm.txt | 50 ++++++++++
drivers/of/Kconfig | 6 +
drivers/of/Makefile | 1 +
drivers/of/pwm.c | 130 +++++++++++++++++++++++++
include/linux/of_pwm.h | 51 ++++++++++
include/linux/pwm.h | 17 +++
6 files changed, 255 insertions(+), 0 deletions(-)
create mode 100644 Documentation/devicetree/bindings/pwm/pwm.txt
create mode 100644 drivers/of/pwm.c
create mode 100644 include/linux/of_pwm.h
diff --git a/Documentation/devicetree/bindings/pwm/pwm.txt b/Documentation/devicetree/bindings/pwm/pwm.txt
new file mode 100644
index 0000000..1c34e86
--- /dev/null
+++ b/Documentation/devicetree/bindings/pwm/pwm.txt
@@ -0,0 +1,50 @@
+Specifying PWM information for devices
+======================================
+
+1) PWM user nodes
+-----------------
+
+PWM users should specify a list of PWM devices that they want to use
+with a property containing a 'pwm-list':
+
+ pwm-list ::= <single-pwm> [pwm-list]
+ single-pwm ::= <pwm-phandle> <pwm-specifier>
+ pwm-phandle : phandle to PWM controller node
+ pwm-specifier : array of #pwm-cells specifying the given PWM
+ (controller specific)
+
+PWM properties should be named "[<name>-]pwms". Exact meaning of each
+pwms property must be documented in the device tree binding for each
+device.
+
+The following example could be used to describe a PWM-based backlight
+device:
+
+ pwm: pwm {
+ #pwm-cells = <2>;
+ };
+
+ [...]
+
+ bl: backlight {
+ pwms = <&pwm 0 5000000>;
+ };
+
+pwm-specifier typically encodes the chip-relative PWM number and the PWM
+period in nanoseconds.
+
+2) PWM controller nodes
+-----------------------
+
+PWM controller nodes must specify the number of cells used for the
+specifier using the '#pwm-cells' property.
+
+An example PWM controller might look like this:
+
+ pwm: pwm at 7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
+TODO: do we need an empty 'pwm-controller' property?
diff --git a/drivers/of/Kconfig b/drivers/of/Kconfig
index 268163d..9635468 100644
--- a/drivers/of/Kconfig
+++ b/drivers/of/Kconfig
@@ -56,6 +56,12 @@ config OF_GPIO
help
OpenFirmware GPIO accessors
+config OF_PWM
+ def_bool y
+ depends on PWM
+ help
+ OpenFirmware PWM accessors
+
config OF_I2C
def_tristate I2C
depends on I2C && !SPARC
diff --git a/drivers/of/Makefile b/drivers/of/Makefile
index a73f5a5..3dd13e3 100644
--- a/drivers/of/Makefile
+++ b/drivers/of/Makefile
@@ -5,6 +5,7 @@ obj-$(CONFIG_OF_ADDRESS) += address.o
obj-$(CONFIG_OF_IRQ) += irq.o
obj-$(CONFIG_OF_DEVICE) += device.o platform.o
obj-$(CONFIG_OF_GPIO) += gpio.o
+obj-$(CONFIG_OF_PWM) += pwm.o
obj-$(CONFIG_OF_I2C) += of_i2c.o
obj-$(CONFIG_OF_NET) += of_net.o
obj-$(CONFIG_OF_SPI) += of_spi.o
diff --git a/drivers/of/pwm.c b/drivers/of/pwm.c
new file mode 100644
index 0000000..d6f7f33
--- /dev/null
+++ b/drivers/of/pwm.c
@@ -0,0 +1,130 @@
+/*
+ * OF helpers for the PWM API
+ *
+ * Copyright (c) 2011 Avionic Design GmbH
+ *
+ * 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.
+ */
+
+#include <linux/errno.h>
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pwm.h>
+
+static int of_pwmchip_is_match(struct pwm_chip *chip, void *data)
+{
+ return chip->dev ? chip->dev->of_node == data : 0;
+}
+
+/**
+ * of_node_to_pwmchip() - finds the PWM chip associated with a device node
+ * @np: device node of the PWM chip
+ *
+ * Returns a pointer to the PWM chip associated with the specified device
+ * node or NULL if the device node doesn't represent a PWM chip.
+ */
+struct pwm_chip *of_node_to_pwmchip(struct device_node *np)
+{
+ return pwmchip_find(np, of_pwmchip_is_match);
+}
+
+int of_pwm_simple_xlate(struct pwm_chip *pc, const struct of_phandle_args *args,
+ struct pwm_spec *spec)
+{
+ if (pc->of_pwm_n_cells < 2)
+ return -EINVAL;
+
+ if (args->args_count < pc->of_pwm_n_cells)
+ return -EINVAL;
+
+ if (args->args[0] >= pc->npwm)
+ return -EINVAL;
+
+ if (spec)
+ spec->period = args->args[1];
+
+ return args->args[0];
+}
+
+void of_pwmchip_add(struct pwm_chip *chip)
+{
+ if (!chip->dev || !chip->dev->of_node)
+ return;
+
+ if (!chip->of_xlate) {
+ chip->of_xlate = of_pwm_simple_xlate;
+ chip->of_pwm_n_cells = 2;
+ }
+
+ of_node_get(chip->dev->of_node);
+}
+
+void of_pwmchip_remove(struct pwm_chip *chip)
+{
+ if (chip->dev && chip->dev->of_node)
+ of_node_put(chip->dev->of_node);
+}
+
+/**
+ * of_get_named_pwm() - get a PWM number and period to use with the PWM API
+ * @np: device node to get the PWM from
+ * @propname: property name containing PWM specifier(s)
+ * @index: index of the PWM
+ * @spec: a pointer to a struct pwm_spec to fill in
+ *
+ * Returns PWM number to use with the Linux generic PWM API or a negative
+ * error code on failure. If @spec is not NULL the function fills in the
+ * values parsed from the device tree.
+ */
+int of_get_named_pwm(struct device_node *np, const char *propname,
+ int index, struct pwm_spec *spec)
+{
+ struct of_phandle_args args;
+ struct pwm_chip *pc;
+ int ret;
+
+ ret = of_parse_phandle_with_args(np, propname, "#pwm-cells", index,
+ &args);
+ if (ret) {
+ pr_debug("%s(): can't parse pwms property\n", __func__);
+ goto out;
+ }
+
+ pc = of_node_to_pwmchip(args.np);
+ if (!pc) {
+ ret = -ENODEV;
+ goto put;
+ }
+
+ if (args.args_count != pc->of_pwm_n_cells) {
+ pr_debug("%s: wrong #pwm-cells for %s\n", np->full_name,
+ args.np->full_name);
+ ret = -EINVAL;
+ goto put;
+ }
+
+ /*
+ * reset the specifier structure since .of_xlate might decide not to
+ * fill it in
+ */
+ if (spec)
+ spec->period = 0;
+
+ ret = pc->of_xlate(pc, &args, spec);
+ if (ret < 0)
+ goto put;
+
+ ret += pc->base;
+
+put:
+ of_node_put(args.np);
+out:
+ pr_debug("%s() exited with status %d\n", __func__, ret);
+ return ret;
+}
+EXPORT_SYMBOL(of_get_named_pwm);
diff --git a/include/linux/of_pwm.h b/include/linux/of_pwm.h
new file mode 100644
index 0000000..a6af951
--- /dev/null
+++ b/include/linux/of_pwm.h
@@ -0,0 +1,51 @@
+/*
+ * OF helpers for the PWM API
+ *
+ * Copyright (c) 2011 Avionic Design GmbH
+ *
+ * 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.
+ */
+
+#ifndef __LINUX_OF_PWM_H
+#define __LINUX_OF_PWM_H
+
+#include <linux/pwm.h>
+
+struct device_node;
+
+#ifdef CONFIG_OF_PWM
+
+struct pwm_chip *of_node_to_pwmchip(struct device_node *np);
+int of_get_named_pwm(struct device_node *np, const char *propname,
+ int index, struct pwm_spec *spec);
+void of_pwmchip_add(struct pwm_chip *pc);
+void of_pwmchip_remove(struct pwm_chip *pc);
+
+#else
+
+static inline struct pwm_chip *of_node_to_pwmchip(struct device_node *np)
+{
+ return NULL;
+}
+
+static inline int of_get_named_pwm(struct device_node *np,
+ const char *propname, int index,
+ unsigned int *period_ns)
+{
+ return -ENOSYS;
+}
+
+static inline void of_pwmchip_add(struct pwm_chip *pc)
+{
+}
+
+static inline void of_pwmchip_remove(struct pwm_chip *pc)
+{
+}
+
+#endif /* CONFIG_OF_PWM */
+
+#endif /* __LINUX_OF_PWM_H */
diff --git a/include/linux/pwm.h b/include/linux/pwm.h
index 0f8d105..8595be4 100644
--- a/include/linux/pwm.h
+++ b/include/linux/pwm.h
@@ -1,6 +1,8 @@
#ifndef __LINUX_PWM_H
#define __LINUX_PWM_H
+#include <linux/of.h>
+
struct pwm_device;
/*
@@ -56,6 +58,14 @@ struct pwm_ops {
};
/**
+ * struct pwm_spec - device tree PWM specifier
+ * @period: PWM period (in nanoseconds)
+ */
+struct pwm_spec {
+ unsigned int period;
+};
+
+/**
* struct pwm_chip - abstract a PWM controller
* @dev: device providing the PWMs
* @ops: callbacks for this PWM controller
@@ -67,6 +77,13 @@ struct pwm_chip {
struct pwm_ops *ops;
int base;
unsigned int npwm;
+
+#ifdef CONFIG_OF_PWM
+ int (*of_xlate)(struct pwm_chip *pc,
+ const struct of_phandle_args *args,
+ struct pwm_spec *spec);
+ unsigned int of_pwm_n_cells;
+#endif
};
int pwm_set_chip_data(unsigned int pwm, void *data);
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 04/10] arm/tegra: Fix PWM clock programming
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: Simon Que, Bill Huang,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
From: Simon Que <sque-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
PWM clock source registers in Tegra 2 have different clock source selection bit
fields than other registers. PWM clock source bits in CLK_SOURCE_PWM_0 register
are located at bit field bit[30:28] while others are at bit field bit[31:30] in
their respective clock source register.
This patch updates the clock programming to correctly reflect that, by adding a
flag to indicate the alternate bit field format and checking for it when
selecting a clock source (parent clock).
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
Signed-off-by: Bill Huang <bilhuang-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
Signed-off-by: Simon Que <sque-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
---
Changes in v2:
- update commit message to account for changes
arch/arm/mach-tegra/clock.h | 1 +
arch/arm/mach-tegra/tegra2_clocks.c | 28 ++++++++++++++++++++++++----
2 files changed, 25 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-tegra/clock.h b/arch/arm/mach-tegra/clock.h
index bc30065..e18d77e 100644
--- a/arch/arm/mach-tegra/clock.h
+++ b/arch/arm/mach-tegra/clock.h
@@ -49,6 +49,7 @@
#define PLLM (1 << 20)
#define DIV_U71_INT (1 << 21)
#define DIV_U71_IDLE (1 << 22)
+#define PERIPH_SOURCE_CLK_4BIT (1 << 23)
#define ENABLE_ON_INIT (1 << 28)
#define PERIPH_ON_APB (1 << 29)
diff --git a/arch/arm/mach-tegra/tegra2_clocks.c b/arch/arm/mach-tegra/tegra2_clocks.c
index 74d314f..379f201 100644
--- a/arch/arm/mach-tegra/tegra2_clocks.c
+++ b/arch/arm/mach-tegra/tegra2_clocks.c
@@ -69,6 +69,8 @@
#define PERIPH_CLK_SOURCE_MASK (3<<30)
#define PERIPH_CLK_SOURCE_SHIFT 30
+#define PERIPH_CLK_SOURCE_4BIT_MASK (7<<28)
+#define PERIPH_CLK_SOURCE_4BIT_SHIFT 28
#define PERIPH_CLK_SOURCE_ENABLE (1<<28)
#define PERIPH_CLK_SOURCE_DIVU71_MASK 0xFF
#define PERIPH_CLK_SOURCE_DIVU16_MASK 0xFFFF
@@ -908,9 +910,16 @@ static void tegra2_periph_clk_init(struct clk *c)
u32 val = clk_readl(c->reg);
const struct clk_mux_sel *mux = NULL;
const struct clk_mux_sel *sel;
+ u32 shift;
+
+ if (c->flags & PERIPH_SOURCE_CLK_4BIT)
+ shift = PERIPH_CLK_SOURCE_4BIT_SHIFT;
+ else
+ shift = PERIPH_CLK_SOURCE_SHIFT;
+
if (c->flags & MUX) {
for (sel = c->inputs; sel->input != NULL; sel++) {
- if (val >> PERIPH_CLK_SOURCE_SHIFT == sel->value)
+ if (val >> shift == sel->value)
mux = sel;
}
BUG_ON(!mux);
@@ -1023,12 +1032,23 @@ static int tegra2_periph_clk_set_parent(struct clk *c, struct clk *p)
{
u32 val;
const struct clk_mux_sel *sel;
+ u32 mask, shift;
+
pr_debug("%s: %s %s\n", __func__, c->name, p->name);
+
+ if (c->flags & PERIPH_SOURCE_CLK_4BIT) {
+ mask = PERIPH_CLK_SOURCE_4BIT_MASK;
+ shift = PERIPH_CLK_SOURCE_4BIT_SHIFT;
+ } else {
+ mask = PERIPH_CLK_SOURCE_MASK;
+ shift = PERIPH_CLK_SOURCE_SHIFT;
+ }
+
for (sel = c->inputs; sel->input != NULL; sel++) {
if (sel->input == p) {
val = clk_readl(c->reg);
- val &= ~PERIPH_CLK_SOURCE_MASK;
- val |= (sel->value) << PERIPH_CLK_SOURCE_SHIFT;
+ val &= ~mask;
+ val |= (sel->value) << shift;
if (c->refcnt)
clk_enable(p);
@@ -2126,7 +2146,7 @@ static struct clk tegra_list_clks[] = {
PERIPH_CLK("i2s2", "tegra-i2s.1", NULL, 18, 0x104, 26000000, mux_pllaout0_audio2x_pllp_clkm, MUX | DIV_U71),
PERIPH_CLK("spdif_out", "spdif_out", NULL, 10, 0x108, 100000000, mux_pllaout0_audio2x_pllp_clkm, MUX | DIV_U71),
PERIPH_CLK("spdif_in", "spdif_in", NULL, 10, 0x10c, 100000000, mux_pllp_pllc_pllm, MUX | DIV_U71),
- PERIPH_CLK("pwm", "pwm", NULL, 17, 0x110, 432000000, mux_pllp_pllc_audio_clkm_clk32, MUX | DIV_U71),
+ PERIPH_CLK("pwm", "pwm", NULL, 17, 0x110, 432000000, mux_pllp_pllc_audio_clkm_clk32, MUX | DIV_U71 | PERIPH_SOURCE_CLK_4BIT),
PERIPH_CLK("spi", "spi", NULL, 43, 0x114, 40000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
PERIPH_CLK("xio", "xio", NULL, 45, 0x120, 150000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
PERIPH_CLK("twc", "twc", NULL, 16, 0x12c, 150000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 04/10] arm/tegra: Fix PWM clock programming
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
From: Simon Que <sque@chromium.org>
PWM clock source registers in Tegra 2 have different clock source selection bit
fields than other registers. PWM clock source bits in CLK_SOURCE_PWM_0 register
are located at bit field bit[30:28] while others are at bit field bit[31:30] in
their respective clock source register.
This patch updates the clock programming to correctly reflect that, by adding a
flag to indicate the alternate bit field format and checking for it when
selecting a clock source (parent clock).
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Signed-off-by: Bill Huang <bilhuang@nvidia.com>
Signed-off-by: Simon Que <sque@chromium.org>
---
Changes in v2:
- update commit message to account for changes
arch/arm/mach-tegra/clock.h | 1 +
arch/arm/mach-tegra/tegra2_clocks.c | 28 ++++++++++++++++++++++++----
2 files changed, 25 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-tegra/clock.h b/arch/arm/mach-tegra/clock.h
index bc30065..e18d77e 100644
--- a/arch/arm/mach-tegra/clock.h
+++ b/arch/arm/mach-tegra/clock.h
@@ -49,6 +49,7 @@
#define PLLM (1 << 20)
#define DIV_U71_INT (1 << 21)
#define DIV_U71_IDLE (1 << 22)
+#define PERIPH_SOURCE_CLK_4BIT (1 << 23)
#define ENABLE_ON_INIT (1 << 28)
#define PERIPH_ON_APB (1 << 29)
diff --git a/arch/arm/mach-tegra/tegra2_clocks.c b/arch/arm/mach-tegra/tegra2_clocks.c
index 74d314f..379f201 100644
--- a/arch/arm/mach-tegra/tegra2_clocks.c
+++ b/arch/arm/mach-tegra/tegra2_clocks.c
@@ -69,6 +69,8 @@
#define PERIPH_CLK_SOURCE_MASK (3<<30)
#define PERIPH_CLK_SOURCE_SHIFT 30
+#define PERIPH_CLK_SOURCE_4BIT_MASK (7<<28)
+#define PERIPH_CLK_SOURCE_4BIT_SHIFT 28
#define PERIPH_CLK_SOURCE_ENABLE (1<<28)
#define PERIPH_CLK_SOURCE_DIVU71_MASK 0xFF
#define PERIPH_CLK_SOURCE_DIVU16_MASK 0xFFFF
@@ -908,9 +910,16 @@ static void tegra2_periph_clk_init(struct clk *c)
u32 val = clk_readl(c->reg);
const struct clk_mux_sel *mux = NULL;
const struct clk_mux_sel *sel;
+ u32 shift;
+
+ if (c->flags & PERIPH_SOURCE_CLK_4BIT)
+ shift = PERIPH_CLK_SOURCE_4BIT_SHIFT;
+ else
+ shift = PERIPH_CLK_SOURCE_SHIFT;
+
if (c->flags & MUX) {
for (sel = c->inputs; sel->input != NULL; sel++) {
- if (val >> PERIPH_CLK_SOURCE_SHIFT == sel->value)
+ if (val >> shift == sel->value)
mux = sel;
}
BUG_ON(!mux);
@@ -1023,12 +1032,23 @@ static int tegra2_periph_clk_set_parent(struct clk *c, struct clk *p)
{
u32 val;
const struct clk_mux_sel *sel;
+ u32 mask, shift;
+
pr_debug("%s: %s %s\n", __func__, c->name, p->name);
+
+ if (c->flags & PERIPH_SOURCE_CLK_4BIT) {
+ mask = PERIPH_CLK_SOURCE_4BIT_MASK;
+ shift = PERIPH_CLK_SOURCE_4BIT_SHIFT;
+ } else {
+ mask = PERIPH_CLK_SOURCE_MASK;
+ shift = PERIPH_CLK_SOURCE_SHIFT;
+ }
+
for (sel = c->inputs; sel->input != NULL; sel++) {
if (sel->input == p) {
val = clk_readl(c->reg);
- val &= ~PERIPH_CLK_SOURCE_MASK;
- val |= (sel->value) << PERIPH_CLK_SOURCE_SHIFT;
+ val &= ~mask;
+ val |= (sel->value) << shift;
if (c->refcnt)
clk_enable(p);
@@ -2126,7 +2146,7 @@ static struct clk tegra_list_clks[] = {
PERIPH_CLK("i2s2", "tegra-i2s.1", NULL, 18, 0x104, 26000000, mux_pllaout0_audio2x_pllp_clkm, MUX | DIV_U71),
PERIPH_CLK("spdif_out", "spdif_out", NULL, 10, 0x108, 100000000, mux_pllaout0_audio2x_pllp_clkm, MUX | DIV_U71),
PERIPH_CLK("spdif_in", "spdif_in", NULL, 10, 0x10c, 100000000, mux_pllp_pllc_pllm, MUX | DIV_U71),
- PERIPH_CLK("pwm", "pwm", NULL, 17, 0x110, 432000000, mux_pllp_pllc_audio_clkm_clk32, MUX | DIV_U71),
+ PERIPH_CLK("pwm", "pwm", NULL, 17, 0x110, 432000000, mux_pllp_pllc_audio_clkm_clk32, MUX | DIV_U71 | PERIPH_SOURCE_CLK_4BIT),
PERIPH_CLK("spi", "spi", NULL, 43, 0x114, 40000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
PERIPH_CLK("xio", "xio", NULL, 45, 0x120, 150000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
PERIPH_CLK("twc", "twc", NULL, 16, 0x12c, 150000000, mux_pllp_pllc_pllm_clkm, MUX | DIV_U71),
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 05/10] arm/tegra: Provide clock for only one PWM controller
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
A subsequent patch will add a generic PWM API driver for the Tegra2 PWFM
controller, supporting all four PWM devices with a single PWM chip. The
device will be named tegra-pwm and only one clock needs to be provided.
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
Acked-by: Stephen Warren <swarren-DDmLM1+adcrQT0dZR+AlfA@public.gmane.org>
---
arch/arm/mach-tegra/tegra2_clocks.c | 5 +----
1 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-tegra/tegra2_clocks.c b/arch/arm/mach-tegra/tegra2_clocks.c
index 379f201..6eec82d 100644
--- a/arch/arm/mach-tegra/tegra2_clocks.c
+++ b/arch/arm/mach-tegra/tegra2_clocks.c
@@ -2245,10 +2245,7 @@ static struct clk_duplicate tegra_clk_duplicates[] = {
CLK_DUPLICATE("usbd", "tegra-otg", NULL),
CLK_DUPLICATE("hdmi", "tegradc.0", "hdmi"),
CLK_DUPLICATE("hdmi", "tegradc.1", "hdmi"),
- CLK_DUPLICATE("pwm", "tegra_pwm.0", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.1", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.2", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.3", NULL),
+ CLK_DUPLICATE("pwm", "tegra-pwm", NULL),
CLK_DUPLICATE("host1x", "tegra_grhost", "host1x"),
CLK_DUPLICATE("2d", "tegra_grhost", "gr2d"),
CLK_DUPLICATE("3d", "tegra_grhost", "gr3d"),
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 05/10] arm/tegra: Provide clock for only one PWM controller
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
A subsequent patch will add a generic PWM API driver for the Tegra2 PWFM
controller, supporting all four PWM devices with a single PWM chip. The
device will be named tegra-pwm and only one clock needs to be provided.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
---
arch/arm/mach-tegra/tegra2_clocks.c | 5 +----
1 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/arch/arm/mach-tegra/tegra2_clocks.c b/arch/arm/mach-tegra/tegra2_clocks.c
index 379f201..6eec82d 100644
--- a/arch/arm/mach-tegra/tegra2_clocks.c
+++ b/arch/arm/mach-tegra/tegra2_clocks.c
@@ -2245,10 +2245,7 @@ static struct clk_duplicate tegra_clk_duplicates[] = {
CLK_DUPLICATE("usbd", "tegra-otg", NULL),
CLK_DUPLICATE("hdmi", "tegradc.0", "hdmi"),
CLK_DUPLICATE("hdmi", "tegradc.1", "hdmi"),
- CLK_DUPLICATE("pwm", "tegra_pwm.0", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.1", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.2", NULL),
- CLK_DUPLICATE("pwm", "tegra_pwm.3", NULL),
+ CLK_DUPLICATE("pwm", "tegra-pwm", NULL),
CLK_DUPLICATE("host1x", "tegra_grhost", "host1x"),
CLK_DUPLICATE("2d", "tegra_grhost", "gr2d"),
CLK_DUPLICATE("3d", "tegra_grhost", "gr3d"),
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 06/10] pwm: Add NVIDIA Tegra SoC support
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
This commit adds a generic PWM framework driver for the PWFM controller
found on NVIDIA Tegra SoCs. The driver is based on code from the
Chromium kernel tree and was originally written by Gary King (NVIDIA)
and later modified by Simon Que (Chromium).
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
Changes in v2:
- set pwm_chip.dev field
- rename clk_enb to enable
- introduce NUM_PWM macro instead of hard-coding the number of PWM
devices
- update comment in tegra_pwm_config
- fix coding-style for multi-line comments
- use module_platform_driver
- change license to GPL
drivers/pwm/Kconfig | 10 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-tegra.c | 266 +++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 277 insertions(+), 0 deletions(-)
create mode 100644 drivers/pwm/pwm-tegra.c
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index 93c1052..bda6f23 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -9,4 +9,14 @@ menuconfig PWM
if PWM
+config PWM_TEGRA
+ tristate "NVIDIA Tegra PWM support"
+ depends on ARCH_TEGRA
+ help
+ Generic PWM framework driver for the PWFM controller found on NVIDIA
+ Tegra SoCs.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-tegra.
+
endif
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 3469c3d..12300f5 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1 +1,2 @@
obj-$(CONFIG_PWM) += core.o
+obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-tegra.c b/drivers/pwm/pwm-tegra.c
new file mode 100644
index 0000000..87ee350
--- /dev/null
+++ b/drivers/pwm/pwm-tegra.c
@@ -0,0 +1,266 @@
+/*
+ * drivers/pwm/pwm-tegra.c
+ *
+ * Tegra pulse-width-modulation controller driver
+ *
+ * Copyright (c) 2010, NVIDIA Corporation.
+ * Based on arch/arm/plat-mxc/pwm.c by Sascha Hauer <s.hauer-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
+ *
+ * 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.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ */
+
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pwm.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#define PWM_ENABLE (1 << 31)
+#define PWM_DUTY_WIDTH 8
+#define PWM_DUTY_SHIFT 16
+#define PWM_SCALE_WIDTH 13
+#define PWM_SCALE_SHIFT 0
+
+#define NUM_PWM 4
+
+struct tegra_pwm_chip {
+ struct pwm_chip chip;
+ struct device *dev;
+
+ struct clk *clk;
+
+ int enable[NUM_PWM];
+ void __iomem *mmio_base;
+};
+
+static inline struct tegra_pwm_chip *to_tegra_pwm_chip(struct pwm_chip *chip)
+{
+ return container_of(chip, struct tegra_pwm_chip, chip);
+}
+
+static inline int pwm_writel(struct tegra_pwm_chip *chip, unsigned int pwm,
+ unsigned long val)
+{
+ unsigned long offset = pwm << 4;
+ int rc;
+
+ rc = clk_enable(chip->clk);
+ if (WARN_ON(rc))
+ return rc;
+
+ writel(val, chip->mmio_base + offset);
+ clk_disable(chip->clk);
+
+ return 0;
+}
+
+static int tegra_pwm_config(struct pwm_chip *chip, unsigned int pwm,
+ int duty_ns, int period_ns)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+ unsigned long long c;
+ unsigned long rate, hz;
+ u32 val = 0;
+
+ /*
+ * Convert from duty_ns / period_ns to a fixed number of duty
+ * ticks per (1 << PWM_DUTY_WIDTH) cycles.
+ */
+ c = duty_ns * ((1 << PWM_DUTY_WIDTH) - 1);
+ do_div(c, period_ns);
+
+ val = (u32)c << PWM_DUTY_SHIFT;
+
+ /*
+ * Compute the prescaler value for which (1 << PWM_DUTY_WIDTH)
+ * cycles at the PWM clock rate will take period_ns nanoseconds.
+ */
+ rate = clk_get_rate(pc->clk) >> PWM_DUTY_WIDTH;
+ hz = 1000000000ul / period_ns;
+
+ rate = (rate + (hz / 2)) / hz;
+
+ /*
+ * Since the actual PWM divider is the register's frequency divider
+ * field minus 1, we need to decrement to get the correct value to
+ * write to the register.
+ */
+ if (rate > 0)
+ rate--;
+
+ /*
+ * Make sure that the rate will fit in the register's frequency
+ * divider field.
+ */
+ if (rate >> PWM_SCALE_WIDTH)
+ return -EINVAL;
+
+ val |= (rate << PWM_SCALE_SHIFT);
+
+ /* If the PWM channel is enabled, keep it enabled */
+ if (pc->enable[num])
+ val |= PWM_ENABLE;
+
+ return pwm_writel(pc, num, val);
+}
+
+static int tegra_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+ int rc = 0;
+
+ if (!pc->enable[num]) {
+ rc = clk_enable(pc->clk);
+ if (!rc) {
+ unsigned long offset = num << 4;
+ u32 val;
+
+ val = readl(pc->mmio_base + offset);
+ val |= PWM_ENABLE;
+ writel(val, pc->mmio_base + offset);
+
+ pc->enable[num] = 1;
+ }
+ }
+
+ return 0;
+}
+
+static void tegra_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+
+ if (pc->enable[num]) {
+ unsigned long offset = num << 4;
+ u32 val;
+
+ val = readl(pc->mmio_base + offset);
+ val &= ~PWM_ENABLE;
+ writel(val, pc->mmio_base + offset);
+
+ clk_disable(pc->clk);
+ pc->enable[num] = 0;
+ }
+}
+
+static struct pwm_ops tegra_pwm_ops = {
+ .config = tegra_pwm_config,
+ .enable = tegra_pwm_enable,
+ .disable = tegra_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int tegra_pwm_probe(struct platform_device *pdev)
+{
+ struct tegra_pwm_chip *pwm;
+ struct resource *r;
+ int ret;
+
+ pwm = devm_kzalloc(&pdev->dev, sizeof(*pwm), GFP_KERNEL);
+ if (!pwm) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ pwm->dev = &pdev->dev;
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!r) {
+ dev_err(&pdev->dev, "no memory resources defined\n");
+ return -ENODEV;
+ }
+
+ r = devm_request_mem_region(&pdev->dev, r->start, resource_size(r),
+ pdev->name);
+ if (!r) {
+ dev_err(&pdev->dev, "failed to request memory\n");
+ return -EBUSY;
+ }
+
+ pwm->mmio_base = devm_ioremap(&pdev->dev, r->start, resource_size(r));
+ if (!pwm->mmio_base) {
+ dev_err(&pdev->dev, "failed to ioremap() region\n");
+ return -ENODEV;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+
+ pwm->clk = clk_get(&pdev->dev, NULL);
+ if (IS_ERR(pwm->clk))
+ return PTR_ERR(pwm->clk);
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &tegra_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = NUM_PWM;
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ clk_put(pwm->clk);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __devexit tegra_pwm_remove(struct platform_device *pdev)
+{
+ struct tegra_pwm_chip *pwm = platform_get_drvdata(pdev);
+ int i;
+
+ if (WARN_ON(!pwm))
+ return -ENODEV;
+
+ pwmchip_remove(&pwm->chip);
+
+ for (i = 0; i < NUM_PWM; i++) {
+ pwm_writel(pwm, i, 0);
+
+ if (pwm->enable[i])
+ clk_disable(pwm->clk);
+ }
+
+ clk_put(pwm->clk);
+
+ return 0;
+}
+
+#ifdef CONFIG_OF
+static struct of_device_id tegra_pwm_of_match[] = {
+ { .compatible = "nvidia,tegra20-pwm" },
+ { }
+};
+#endif
+
+static struct platform_driver tegra_pwm_driver = {
+ .driver = {
+ .name = "tegra-pwm",
+ .of_match_table = of_match_ptr(tegra_pwm_of_match),
+ },
+ .probe = tegra_pwm_probe,
+ .remove = __devexit_p(tegra_pwm_remove),
+};
+
+module_platform_driver(tegra_pwm_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("NVIDIA Corporation");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 06/10] pwm: Add NVIDIA Tegra SoC support
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
This commit adds a generic PWM framework driver for the PWFM controller
found on NVIDIA Tegra SoCs. The driver is based on code from the
Chromium kernel tree and was originally written by Gary King (NVIDIA)
and later modified by Simon Que (Chromium).
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
Changes in v2:
- set pwm_chip.dev field
- rename clk_enb to enable
- introduce NUM_PWM macro instead of hard-coding the number of PWM
devices
- update comment in tegra_pwm_config
- fix coding-style for multi-line comments
- use module_platform_driver
- change license to GPL
drivers/pwm/Kconfig | 10 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-tegra.c | 266 +++++++++++++++++++++++++++++++++++++++++++++++
3 files changed, 277 insertions(+), 0 deletions(-)
create mode 100644 drivers/pwm/pwm-tegra.c
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index 93c1052..bda6f23 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -9,4 +9,14 @@ menuconfig PWM
if PWM
+config PWM_TEGRA
+ tristate "NVIDIA Tegra PWM support"
+ depends on ARCH_TEGRA
+ help
+ Generic PWM framework driver for the PWFM controller found on NVIDIA
+ Tegra SoCs.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-tegra.
+
endif
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 3469c3d..12300f5 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1 +1,2 @@
obj-$(CONFIG_PWM) += core.o
+obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-tegra.c b/drivers/pwm/pwm-tegra.c
new file mode 100644
index 0000000..87ee350
--- /dev/null
+++ b/drivers/pwm/pwm-tegra.c
@@ -0,0 +1,266 @@
+/*
+ * drivers/pwm/pwm-tegra.c
+ *
+ * Tegra pulse-width-modulation controller driver
+ *
+ * Copyright (c) 2010, NVIDIA Corporation.
+ * Based on arch/arm/plat-mxc/pwm.c by Sascha Hauer <s.hauer@pengutronix.de>
+ *
+ * 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.,
+ * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ */
+
+#include <linux/clk.h>
+#include <linux/err.h>
+#include <linux/io.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/pwm.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+
+#define PWM_ENABLE (1 << 31)
+#define PWM_DUTY_WIDTH 8
+#define PWM_DUTY_SHIFT 16
+#define PWM_SCALE_WIDTH 13
+#define PWM_SCALE_SHIFT 0
+
+#define NUM_PWM 4
+
+struct tegra_pwm_chip {
+ struct pwm_chip chip;
+ struct device *dev;
+
+ struct clk *clk;
+
+ int enable[NUM_PWM];
+ void __iomem *mmio_base;
+};
+
+static inline struct tegra_pwm_chip *to_tegra_pwm_chip(struct pwm_chip *chip)
+{
+ return container_of(chip, struct tegra_pwm_chip, chip);
+}
+
+static inline int pwm_writel(struct tegra_pwm_chip *chip, unsigned int pwm,
+ unsigned long val)
+{
+ unsigned long offset = pwm << 4;
+ int rc;
+
+ rc = clk_enable(chip->clk);
+ if (WARN_ON(rc))
+ return rc;
+
+ writel(val, chip->mmio_base + offset);
+ clk_disable(chip->clk);
+
+ return 0;
+}
+
+static int tegra_pwm_config(struct pwm_chip *chip, unsigned int pwm,
+ int duty_ns, int period_ns)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+ unsigned long long c;
+ unsigned long rate, hz;
+ u32 val = 0;
+
+ /*
+ * Convert from duty_ns / period_ns to a fixed number of duty
+ * ticks per (1 << PWM_DUTY_WIDTH) cycles.
+ */
+ c = duty_ns * ((1 << PWM_DUTY_WIDTH) - 1);
+ do_div(c, period_ns);
+
+ val = (u32)c << PWM_DUTY_SHIFT;
+
+ /*
+ * Compute the prescaler value for which (1 << PWM_DUTY_WIDTH)
+ * cycles@the PWM clock rate will take period_ns nanoseconds.
+ */
+ rate = clk_get_rate(pc->clk) >> PWM_DUTY_WIDTH;
+ hz = 1000000000ul / period_ns;
+
+ rate = (rate + (hz / 2)) / hz;
+
+ /*
+ * Since the actual PWM divider is the register's frequency divider
+ * field minus 1, we need to decrement to get the correct value to
+ * write to the register.
+ */
+ if (rate > 0)
+ rate--;
+
+ /*
+ * Make sure that the rate will fit in the register's frequency
+ * divider field.
+ */
+ if (rate >> PWM_SCALE_WIDTH)
+ return -EINVAL;
+
+ val |= (rate << PWM_SCALE_SHIFT);
+
+ /* If the PWM channel is enabled, keep it enabled */
+ if (pc->enable[num])
+ val |= PWM_ENABLE;
+
+ return pwm_writel(pc, num, val);
+}
+
+static int tegra_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+ int rc = 0;
+
+ if (!pc->enable[num]) {
+ rc = clk_enable(pc->clk);
+ if (!rc) {
+ unsigned long offset = num << 4;
+ u32 val;
+
+ val = readl(pc->mmio_base + offset);
+ val |= PWM_ENABLE;
+ writel(val, pc->mmio_base + offset);
+
+ pc->enable[num] = 1;
+ }
+ }
+
+ return 0;
+}
+
+static void tegra_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct tegra_pwm_chip *pc = to_tegra_pwm_chip(chip);
+ unsigned int num = pwm - chip->base;
+
+ if (pc->enable[num]) {
+ unsigned long offset = num << 4;
+ u32 val;
+
+ val = readl(pc->mmio_base + offset);
+ val &= ~PWM_ENABLE;
+ writel(val, pc->mmio_base + offset);
+
+ clk_disable(pc->clk);
+ pc->enable[num] = 0;
+ }
+}
+
+static struct pwm_ops tegra_pwm_ops = {
+ .config = tegra_pwm_config,
+ .enable = tegra_pwm_enable,
+ .disable = tegra_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int tegra_pwm_probe(struct platform_device *pdev)
+{
+ struct tegra_pwm_chip *pwm;
+ struct resource *r;
+ int ret;
+
+ pwm = devm_kzalloc(&pdev->dev, sizeof(*pwm), GFP_KERNEL);
+ if (!pwm) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ pwm->dev = &pdev->dev;
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (!r) {
+ dev_err(&pdev->dev, "no memory resources defined\n");
+ return -ENODEV;
+ }
+
+ r = devm_request_mem_region(&pdev->dev, r->start, resource_size(r),
+ pdev->name);
+ if (!r) {
+ dev_err(&pdev->dev, "failed to request memory\n");
+ return -EBUSY;
+ }
+
+ pwm->mmio_base = devm_ioremap(&pdev->dev, r->start, resource_size(r));
+ if (!pwm->mmio_base) {
+ dev_err(&pdev->dev, "failed to ioremap() region\n");
+ return -ENODEV;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+
+ pwm->clk = clk_get(&pdev->dev, NULL);
+ if (IS_ERR(pwm->clk))
+ return PTR_ERR(pwm->clk);
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &tegra_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = NUM_PWM;
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ clk_put(pwm->clk);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __devexit tegra_pwm_remove(struct platform_device *pdev)
+{
+ struct tegra_pwm_chip *pwm = platform_get_drvdata(pdev);
+ int i;
+
+ if (WARN_ON(!pwm))
+ return -ENODEV;
+
+ pwmchip_remove(&pwm->chip);
+
+ for (i = 0; i < NUM_PWM; i++) {
+ pwm_writel(pwm, i, 0);
+
+ if (pwm->enable[i])
+ clk_disable(pwm->clk);
+ }
+
+ clk_put(pwm->clk);
+
+ return 0;
+}
+
+#ifdef CONFIG_OF
+static struct of_device_id tegra_pwm_of_match[] = {
+ { .compatible = "nvidia,tegra20-pwm" },
+ { }
+};
+#endif
+
+static struct platform_driver tegra_pwm_driver = {
+ .driver = {
+ .name = "tegra-pwm",
+ .of_match_table = of_match_ptr(tegra_pwm_of_match),
+ },
+ .probe = tegra_pwm_probe,
+ .remove = __devexit_p(tegra_pwm_remove),
+};
+
+module_platform_driver(tegra_pwm_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("NVIDIA Corporation");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 07/10] arm/tegra: Add PWFM controller device tree probing
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: Mark Brown, Sascha Hauer, Colin Cross, Rob Herring,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
Add auxdata to instantiate a device tree for the PWFM controller and
include a corresponding node in the device tree.
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
arch/arm/boot/dts/tegra20.dtsi | 6 ++++++
arch/arm/boot/dts/tegra30.dtsi | 6 ++++++
arch/arm/mach-tegra/board-dt-tegra20.c | 1 +
arch/arm/mach-tegra/board-dt-tegra30.c | 1 +
4 files changed, 14 insertions(+), 0 deletions(-)
diff --git a/arch/arm/boot/dts/tegra20.dtsi b/arch/arm/boot/dts/tegra20.dtsi
index ec1f010..c34e6c3 100644
--- a/arch/arm/boot/dts/tegra20.dtsi
+++ b/arch/arm/boot/dts/tegra20.dtsi
@@ -148,6 +148,12 @@
interrupts = < 0 91 0x04 >;
};
+ pwm: pwm@7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
emc@7000f400 {
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm/boot/dts/tegra30.dtsi b/arch/arm/boot/dts/tegra30.dtsi
index ac4b75c..77bd985 100644
--- a/arch/arm/boot/dts/tegra30.dtsi
+++ b/arch/arm/boot/dts/tegra30.dtsi
@@ -146,6 +146,12 @@
interrupts = < 0 91 0x04 >;
};
+ pwm: pwm@7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
sdhci@78000000 {
compatible = "nvidia,tegra30-sdhci", "nvidia,tegra20-sdhci";
reg = <0x78000000 0x200>;
diff --git a/arch/arm/mach-tegra/board-dt-tegra20.c b/arch/arm/mach-tegra/board-dt-tegra20.c
index 7a95e0b..bad0dd1 100644
--- a/arch/arm/mach-tegra/board-dt-tegra20.c
+++ b/arch/arm/mach-tegra/board-dt-tegra20.c
@@ -73,6 +73,7 @@ struct of_dev_auxdata tegra20_auxdata_lookup[] __initdata = {
&tegra_ehci2_device.dev.platform_data),
OF_DEV_AUXDATA("nvidia,tegra20-ehci", TEGRA_USB3_BASE, "tegra-ehci.2",
&tegra_ehci3_device.dev.platform_data),
+ OF_DEV_AUXDATA("nvidia,tegra20-pwm", TEGRA_PWFM_BASE, "tegra-pwm", NULL),
{}
};
diff --git a/arch/arm/mach-tegra/board-dt-tegra30.c b/arch/arm/mach-tegra/board-dt-tegra30.c
index b4124b1..dc7324c 100644
--- a/arch/arm/mach-tegra/board-dt-tegra30.c
+++ b/arch/arm/mach-tegra/board-dt-tegra30.c
@@ -51,6 +51,7 @@ struct of_dev_auxdata tegra30_auxdata_lookup[] __initdata = {
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000C500, "tegra-i2c.2", NULL),
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000C700, "tegra-i2c.3", NULL),
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000D000, "tegra-i2c.4", NULL),
+ OF_DEV_AUXDATA("nvidia,tegra20-pwm", TEGRA_PWFM_BASE, "tegra-pwm", NULL),
{}
};
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 07/10] arm/tegra: Add PWFM controller device tree probing
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
Add auxdata to instantiate a device tree for the PWFM controller and
include a corresponding node in the device tree.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
arch/arm/boot/dts/tegra20.dtsi | 6 ++++++
arch/arm/boot/dts/tegra30.dtsi | 6 ++++++
arch/arm/mach-tegra/board-dt-tegra20.c | 1 +
arch/arm/mach-tegra/board-dt-tegra30.c | 1 +
4 files changed, 14 insertions(+), 0 deletions(-)
diff --git a/arch/arm/boot/dts/tegra20.dtsi b/arch/arm/boot/dts/tegra20.dtsi
index ec1f010..c34e6c3 100644
--- a/arch/arm/boot/dts/tegra20.dtsi
+++ b/arch/arm/boot/dts/tegra20.dtsi
@@ -148,6 +148,12 @@
interrupts = < 0 91 0x04 >;
};
+ pwm: pwm at 7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
emc at 7000f400 {
#address-cells = <1>;
#size-cells = <0>;
diff --git a/arch/arm/boot/dts/tegra30.dtsi b/arch/arm/boot/dts/tegra30.dtsi
index ac4b75c..77bd985 100644
--- a/arch/arm/boot/dts/tegra30.dtsi
+++ b/arch/arm/boot/dts/tegra30.dtsi
@@ -146,6 +146,12 @@
interrupts = < 0 91 0x04 >;
};
+ pwm: pwm at 7000a000 {
+ compatible = "nvidia,tegra20-pwm";
+ reg = <0x7000a000 0x100>;
+ #pwm-cells = <2>;
+ };
+
sdhci at 78000000 {
compatible = "nvidia,tegra30-sdhci", "nvidia,tegra20-sdhci";
reg = <0x78000000 0x200>;
diff --git a/arch/arm/mach-tegra/board-dt-tegra20.c b/arch/arm/mach-tegra/board-dt-tegra20.c
index 7a95e0b..bad0dd1 100644
--- a/arch/arm/mach-tegra/board-dt-tegra20.c
+++ b/arch/arm/mach-tegra/board-dt-tegra20.c
@@ -73,6 +73,7 @@ struct of_dev_auxdata tegra20_auxdata_lookup[] __initdata = {
&tegra_ehci2_device.dev.platform_data),
OF_DEV_AUXDATA("nvidia,tegra20-ehci", TEGRA_USB3_BASE, "tegra-ehci.2",
&tegra_ehci3_device.dev.platform_data),
+ OF_DEV_AUXDATA("nvidia,tegra20-pwm", TEGRA_PWFM_BASE, "tegra-pwm", NULL),
{}
};
diff --git a/arch/arm/mach-tegra/board-dt-tegra30.c b/arch/arm/mach-tegra/board-dt-tegra30.c
index b4124b1..dc7324c 100644
--- a/arch/arm/mach-tegra/board-dt-tegra30.c
+++ b/arch/arm/mach-tegra/board-dt-tegra30.c
@@ -51,6 +51,7 @@ struct of_dev_auxdata tegra30_auxdata_lookup[] __initdata = {
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000C500, "tegra-i2c.2", NULL),
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000C700, "tegra-i2c.3", NULL),
OF_DEV_AUXDATA("nvidia,tegra20-i2c", 0x7000D000, "tegra-i2c.4", NULL),
+ OF_DEV_AUXDATA("nvidia,tegra20-pwm", TEGRA_PWFM_BASE, "tegra-pwm", NULL),
{}
};
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 08/10] pwm: Add Blackfin support
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
arch/blackfin/Kconfig | 10 ---
arch/blackfin/kernel/Makefile | 1 -
arch/blackfin/kernel/pwm.c | 100 -------------------------
drivers/pwm/Kconfig | 9 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-bfin.c | 164 +++++++++++++++++++++++++++++++++++++++++
6 files changed, 174 insertions(+), 111 deletions(-)
delete mode 100644 arch/blackfin/kernel/pwm.c
create mode 100644 drivers/pwm/pwm-bfin.c
diff --git a/arch/blackfin/Kconfig b/arch/blackfin/Kconfig
index abe5a9e..3f23fd3 100644
--- a/arch/blackfin/Kconfig
+++ b/arch/blackfin/Kconfig
@@ -949,16 +949,6 @@ config BFIN_GPTIMERS
To compile this driver as a module, choose M here: the module
will be called gptimers.
-config HAVE_PWM
- tristate "Enable PWM API support"
- depends on BFIN_GPTIMERS
- help
- Enable support for the Pulse Width Modulation framework (as
- found in linux/pwm.h).
-
- To compile this driver as a module, choose M here: the module
- will be called pwm.
-
choice
prompt "Uncached DMA region"
default DMA_UNCACHED_1M
diff --git a/arch/blackfin/kernel/Makefile b/arch/blackfin/kernel/Makefile
index 1f88edd..03f60a6 100644
--- a/arch/blackfin/kernel/Makefile
+++ b/arch/blackfin/kernel/Makefile
@@ -21,7 +21,6 @@ obj-$(CONFIG_FUNCTION_TRACER) += ftrace-entry.o
obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
CFLAGS_REMOVE_ftrace.o = -pg
-obj-$(CONFIG_HAVE_PWM) += pwm.o
obj-$(CONFIG_IPIPE) += ipipe.o
obj-$(CONFIG_BFIN_GPTIMERS) += gptimers.o
obj-$(CONFIG_CPLB_INFO) += cplbinfo.o
diff --git a/arch/blackfin/kernel/pwm.c b/arch/blackfin/kernel/pwm.c
deleted file mode 100644
index 33f5942..0000000
--- a/arch/blackfin/kernel/pwm.c
+++ /dev/null
@@ -1,100 +0,0 @@
-/*
- * Blackfin Pulse Width Modulation (PWM) core
- *
- * Copyright (c) 2011 Analog Devices Inc.
- *
- * Licensed under the GPL-2 or later.
- */
-
-#include <linux/module.h>
-#include <linux/pwm.h>
-#include <linux/slab.h>
-
-#include <asm/gptimers.h>
-#include <asm/portmux.h>
-
-struct pwm_device {
- unsigned id;
- unsigned short pin;
-};
-
-static const unsigned short pwm_to_gptimer_per[] = {
- P_TMR0, P_TMR1, P_TMR2, P_TMR3, P_TMR4, P_TMR5,
- P_TMR6, P_TMR7, P_TMR8, P_TMR9, P_TMR10, P_TMR11,
-};
-
-struct pwm_device *pwm_request(int pwm_id, const char *label)
-{
- struct pwm_device *pwm;
- int ret;
-
- /* XXX: pwm_id really should be unsigned */
- if (pwm_id < 0)
- return NULL;
-
- pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
- if (!pwm)
- return pwm;
-
- pwm->id = pwm_id;
- if (pwm->id >= ARRAY_SIZE(pwm_to_gptimer_per))
- goto err;
-
- pwm->pin = pwm_to_gptimer_per[pwm->id];
- ret = peripheral_request(pwm->pin, label);
- if (ret)
- goto err;
-
- return pwm;
- err:
- kfree(pwm);
- return NULL;
-}
-EXPORT_SYMBOL(pwm_request);
-
-void pwm_free(struct pwm_device *pwm)
-{
- peripheral_free(pwm->pin);
- kfree(pwm);
-}
-EXPORT_SYMBOL(pwm_free);
-
-int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
-{
- unsigned long period, duty;
- unsigned long long val;
-
- if (duty_ns < 0 || duty_ns > period_ns)
- return -EINVAL;
-
- val = (unsigned long long)get_sclk() * period_ns;
- do_div(val, NSEC_PER_SEC);
- period = val;
-
- val = (unsigned long long)period * duty_ns;
- do_div(val, period_ns);
- duty = period - val;
-
- if (duty >= period)
- duty = period - 1;
-
- set_gptimer_config(pwm->id, TIMER_MODE_PWM | TIMER_PERIOD_CNT);
- set_gptimer_pwidth(pwm->id, duty);
- set_gptimer_period(pwm->id, period);
-
- return 0;
-}
-EXPORT_SYMBOL(pwm_config);
-
-int pwm_enable(struct pwm_device *pwm)
-{
- enable_gptimer(pwm->id);
- return 0;
-}
-EXPORT_SYMBOL(pwm_enable);
-
-void pwm_disable(struct pwm_device *pwm)
-{
- disable_gptimer(pwm->id);
-}
-EXPORT_SYMBOL(pwm_disable);
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index bda6f23..eb54042 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -9,6 +9,15 @@ menuconfig PWM
if PWM
+config PWM_BFIN
+ tristate "Blackfin PWM support"
+ depends on BFIN_GPTIMERS
+ help
+ Generic PWM framework driver for Blackfin.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-bfin.
+
config PWM_TEGRA
tristate "NVIDIA Tegra PWM support"
depends on ARCH_TEGRA
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 12300f5..251b8d2 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1,2 +1,3 @@
obj-$(CONFIG_PWM) += core.o
+obj-$(CONFIG_PWM_BFIN) += pwm-bfin.o
obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-bfin.c b/drivers/pwm/pwm-bfin.c
new file mode 100644
index 0000000..accfa86
--- /dev/null
+++ b/drivers/pwm/pwm-bfin.c
@@ -0,0 +1,164 @@
+/*
+ * Blackfin Pulse Width Modulation (PWM) core
+ *
+ * Copyright (c) 2011 Analog Devices Inc.
+ *
+ * Licensed under the GPL-2 or later.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pwm.h>
+#include <linux/slab.h>
+
+#include <asm/gptimers.h>
+#include <asm/portmux.h>
+
+struct bfin_pwm_chip {
+ struct pwm_chip chip;
+};
+
+struct bfin_pwm {
+ unsigned short pin;
+};
+
+static const unsigned short pwm_to_gptimer_per[] = {
+ P_TMR0, P_TMR1, P_TMR2, P_TMR3, P_TMR4, P_TMR5,
+ P_TMR6, P_TMR7, P_TMR8, P_TMR9, P_TMR10, P_TMR11,
+};
+
+static int bfin_pwm_request(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct bfin_pwm *priv;
+ int ret;
+
+ if (pwm >= ARRAY_SIZE(pwm_to_gptimer_per))
+ return -EINVAL;
+
+ priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+ if (!priv)
+ return -ENOMEM;
+
+ priv->pin = pwm_to_gptimer_per[pwm];
+
+ ret = peripheral_request(priv->pin, NULL);
+ if (ret) {
+ kfree(priv);
+ return ret;
+ }
+
+ pwm_set_chip_data(pwm, priv);
+
+ return 0;
+}
+
+static void bfin_pwm_free(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct bfin_pwm *priv = pwm_get_chip_data(pwm);
+
+ if (priv) {
+ peripheral_free(priv->pin);
+ kfree(priv);
+ }
+}
+
+static int bfin_pwm_config(struct pwm_chip *chip, unsigned int pwm,
+ int duty_ns, int period_ns)
+{
+ unsigned int num = pwm - chip->base;
+ unsigned long period, duty;
+ unsigned long long val;
+
+ if (duty_ns < 0 || duty_ns > period_ns)
+ return -EINVAL;
+
+ val = (unsigned long long)get_sclk() * period_ns;
+ do_div(val, NSEC_PER_SEC);
+ period = val;
+
+ val = (unsigned long long)period * duty_ns;
+ do_div(val, period_ns);
+ duty = period - val;
+
+ if (duty >= period)
+ duty = period - 1;
+
+ set_gptimer_config(num, TIMER_MODE_PWM | TIMER_PERIOD_CNT);
+ set_gptimer_pwidth(num, duty);
+ set_gptimer_period(num, period);
+
+ return 0;
+}
+
+static int bfin_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ unsigned int num = pwm - chip->base;
+
+ enable_gptimer(num);
+
+ return 0;
+}
+
+static void bfin_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ unsigned int num = pwm - chip->base;
+
+ disable_gptimer(num);
+}
+
+static struct pwm_ops bfin_pwm_ops = {
+ .request = bfin_pwm_request,
+ .free = bfin_pwm_free,
+ .config = bfin_pwm_config,
+ .enable = bfin_pwm_enable,
+ .disable = bfin_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int bfin_pwm_probe(struct platform_device *pdev)
+{
+ struct bfin_pwm_chip *pwm;
+ int ret;
+
+ pwm = devm_kzalloc(&pdev->dev, sizeof(*pwm), GFP_KERNEL);
+ if (!pwm) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &bfin_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = 12;
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __devexit bfin_pwm_remove(struct platform_device *pdev)
+{
+ struct bfin_pwm_chip *pwm = platform_get_drvdata(pdev);
+
+ pwmchip_remove(&pwm->chip);
+
+ return 0;
+}
+
+static struct platform_driver bfin_pwm_driver = {
+ .driver = {
+ .name = "bfin-pwm",
+ },
+ .probe = bfin_pwm_probe,
+ .remove = __devexit_p(bfin_pwm_remove),
+};
+
+module_platform_driver(bfin_pwm_driver);
+
+MODULE_LICENSE("GPL");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 08/10] pwm: Add Blackfin support
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
arch/blackfin/Kconfig | 10 ---
arch/blackfin/kernel/Makefile | 1 -
arch/blackfin/kernel/pwm.c | 100 -------------------------
drivers/pwm/Kconfig | 9 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-bfin.c | 164 +++++++++++++++++++++++++++++++++++++++++
6 files changed, 174 insertions(+), 111 deletions(-)
delete mode 100644 arch/blackfin/kernel/pwm.c
create mode 100644 drivers/pwm/pwm-bfin.c
diff --git a/arch/blackfin/Kconfig b/arch/blackfin/Kconfig
index abe5a9e..3f23fd3 100644
--- a/arch/blackfin/Kconfig
+++ b/arch/blackfin/Kconfig
@@ -949,16 +949,6 @@ config BFIN_GPTIMERS
To compile this driver as a module, choose M here: the module
will be called gptimers.
-config HAVE_PWM
- tristate "Enable PWM API support"
- depends on BFIN_GPTIMERS
- help
- Enable support for the Pulse Width Modulation framework (as
- found in linux/pwm.h).
-
- To compile this driver as a module, choose M here: the module
- will be called pwm.
-
choice
prompt "Uncached DMA region"
default DMA_UNCACHED_1M
diff --git a/arch/blackfin/kernel/Makefile b/arch/blackfin/kernel/Makefile
index 1f88edd..03f60a6 100644
--- a/arch/blackfin/kernel/Makefile
+++ b/arch/blackfin/kernel/Makefile
@@ -21,7 +21,6 @@ obj-$(CONFIG_FUNCTION_TRACER) += ftrace-entry.o
obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
CFLAGS_REMOVE_ftrace.o = -pg
-obj-$(CONFIG_HAVE_PWM) += pwm.o
obj-$(CONFIG_IPIPE) += ipipe.o
obj-$(CONFIG_BFIN_GPTIMERS) += gptimers.o
obj-$(CONFIG_CPLB_INFO) += cplbinfo.o
diff --git a/arch/blackfin/kernel/pwm.c b/arch/blackfin/kernel/pwm.c
deleted file mode 100644
index 33f5942..0000000
--- a/arch/blackfin/kernel/pwm.c
+++ /dev/null
@@ -1,100 +0,0 @@
-/*
- * Blackfin Pulse Width Modulation (PWM) core
- *
- * Copyright (c) 2011 Analog Devices Inc.
- *
- * Licensed under the GPL-2 or later.
- */
-
-#include <linux/module.h>
-#include <linux/pwm.h>
-#include <linux/slab.h>
-
-#include <asm/gptimers.h>
-#include <asm/portmux.h>
-
-struct pwm_device {
- unsigned id;
- unsigned short pin;
-};
-
-static const unsigned short pwm_to_gptimer_per[] = {
- P_TMR0, P_TMR1, P_TMR2, P_TMR3, P_TMR4, P_TMR5,
- P_TMR6, P_TMR7, P_TMR8, P_TMR9, P_TMR10, P_TMR11,
-};
-
-struct pwm_device *pwm_request(int pwm_id, const char *label)
-{
- struct pwm_device *pwm;
- int ret;
-
- /* XXX: pwm_id really should be unsigned */
- if (pwm_id < 0)
- return NULL;
-
- pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
- if (!pwm)
- return pwm;
-
- pwm->id = pwm_id;
- if (pwm->id >= ARRAY_SIZE(pwm_to_gptimer_per))
- goto err;
-
- pwm->pin = pwm_to_gptimer_per[pwm->id];
- ret = peripheral_request(pwm->pin, label);
- if (ret)
- goto err;
-
- return pwm;
- err:
- kfree(pwm);
- return NULL;
-}
-EXPORT_SYMBOL(pwm_request);
-
-void pwm_free(struct pwm_device *pwm)
-{
- peripheral_free(pwm->pin);
- kfree(pwm);
-}
-EXPORT_SYMBOL(pwm_free);
-
-int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
-{
- unsigned long period, duty;
- unsigned long long val;
-
- if (duty_ns < 0 || duty_ns > period_ns)
- return -EINVAL;
-
- val = (unsigned long long)get_sclk() * period_ns;
- do_div(val, NSEC_PER_SEC);
- period = val;
-
- val = (unsigned long long)period * duty_ns;
- do_div(val, period_ns);
- duty = period - val;
-
- if (duty >= period)
- duty = period - 1;
-
- set_gptimer_config(pwm->id, TIMER_MODE_PWM | TIMER_PERIOD_CNT);
- set_gptimer_pwidth(pwm->id, duty);
- set_gptimer_period(pwm->id, period);
-
- return 0;
-}
-EXPORT_SYMBOL(pwm_config);
-
-int pwm_enable(struct pwm_device *pwm)
-{
- enable_gptimer(pwm->id);
- return 0;
-}
-EXPORT_SYMBOL(pwm_enable);
-
-void pwm_disable(struct pwm_device *pwm)
-{
- disable_gptimer(pwm->id);
-}
-EXPORT_SYMBOL(pwm_disable);
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index bda6f23..eb54042 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -9,6 +9,15 @@ menuconfig PWM
if PWM
+config PWM_BFIN
+ tristate "Blackfin PWM support"
+ depends on BFIN_GPTIMERS
+ help
+ Generic PWM framework driver for Blackfin.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-bfin.
+
config PWM_TEGRA
tristate "NVIDIA Tegra PWM support"
depends on ARCH_TEGRA
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 12300f5..251b8d2 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1,2 +1,3 @@
obj-$(CONFIG_PWM) += core.o
+obj-$(CONFIG_PWM_BFIN) += pwm-bfin.o
obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-bfin.c b/drivers/pwm/pwm-bfin.c
new file mode 100644
index 0000000..accfa86
--- /dev/null
+++ b/drivers/pwm/pwm-bfin.c
@@ -0,0 +1,164 @@
+/*
+ * Blackfin Pulse Width Modulation (PWM) core
+ *
+ * Copyright (c) 2011 Analog Devices Inc.
+ *
+ * Licensed under the GPL-2 or later.
+ */
+
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/pwm.h>
+#include <linux/slab.h>
+
+#include <asm/gptimers.h>
+#include <asm/portmux.h>
+
+struct bfin_pwm_chip {
+ struct pwm_chip chip;
+};
+
+struct bfin_pwm {
+ unsigned short pin;
+};
+
+static const unsigned short pwm_to_gptimer_per[] = {
+ P_TMR0, P_TMR1, P_TMR2, P_TMR3, P_TMR4, P_TMR5,
+ P_TMR6, P_TMR7, P_TMR8, P_TMR9, P_TMR10, P_TMR11,
+};
+
+static int bfin_pwm_request(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct bfin_pwm *priv;
+ int ret;
+
+ if (pwm >= ARRAY_SIZE(pwm_to_gptimer_per))
+ return -EINVAL;
+
+ priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+ if (!priv)
+ return -ENOMEM;
+
+ priv->pin = pwm_to_gptimer_per[pwm];
+
+ ret = peripheral_request(priv->pin, NULL);
+ if (ret) {
+ kfree(priv);
+ return ret;
+ }
+
+ pwm_set_chip_data(pwm, priv);
+
+ return 0;
+}
+
+static void bfin_pwm_free(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct bfin_pwm *priv = pwm_get_chip_data(pwm);
+
+ if (priv) {
+ peripheral_free(priv->pin);
+ kfree(priv);
+ }
+}
+
+static int bfin_pwm_config(struct pwm_chip *chip, unsigned int pwm,
+ int duty_ns, int period_ns)
+{
+ unsigned int num = pwm - chip->base;
+ unsigned long period, duty;
+ unsigned long long val;
+
+ if (duty_ns < 0 || duty_ns > period_ns)
+ return -EINVAL;
+
+ val = (unsigned long long)get_sclk() * period_ns;
+ do_div(val, NSEC_PER_SEC);
+ period = val;
+
+ val = (unsigned long long)period * duty_ns;
+ do_div(val, period_ns);
+ duty = period - val;
+
+ if (duty >= period)
+ duty = period - 1;
+
+ set_gptimer_config(num, TIMER_MODE_PWM | TIMER_PERIOD_CNT);
+ set_gptimer_pwidth(num, duty);
+ set_gptimer_period(num, period);
+
+ return 0;
+}
+
+static int bfin_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ unsigned int num = pwm - chip->base;
+
+ enable_gptimer(num);
+
+ return 0;
+}
+
+static void bfin_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ unsigned int num = pwm - chip->base;
+
+ disable_gptimer(num);
+}
+
+static struct pwm_ops bfin_pwm_ops = {
+ .request = bfin_pwm_request,
+ .free = bfin_pwm_free,
+ .config = bfin_pwm_config,
+ .enable = bfin_pwm_enable,
+ .disable = bfin_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int bfin_pwm_probe(struct platform_device *pdev)
+{
+ struct bfin_pwm_chip *pwm;
+ int ret;
+
+ pwm = devm_kzalloc(&pdev->dev, sizeof(*pwm), GFP_KERNEL);
+ if (!pwm) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &bfin_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = 12;
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int __devexit bfin_pwm_remove(struct platform_device *pdev)
+{
+ struct bfin_pwm_chip *pwm = platform_get_drvdata(pdev);
+
+ pwmchip_remove(&pwm->chip);
+
+ return 0;
+}
+
+static struct platform_driver bfin_pwm_driver = {
+ .driver = {
+ .name = "bfin-pwm",
+ },
+ .probe = bfin_pwm_probe,
+ .remove = __devexit_p(bfin_pwm_remove),
+};
+
+module_platform_driver(bfin_pwm_driver);
+
+MODULE_LICENSE("GPL");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 09/10] pwm: Add PXA support
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
arch/arm/plat-pxa/Makefile | 1 -
arch/arm/plat-pxa/pwm.c | 304 --------------------------------------------
drivers/pwm/Kconfig | 9 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-pxa.c | 243 +++++++++++++++++++++++++++++++++++
5 files changed, 253 insertions(+), 305 deletions(-)
delete mode 100644 arch/arm/plat-pxa/pwm.c
create mode 100644 drivers/pwm/pwm-pxa.c
diff --git a/arch/arm/plat-pxa/Makefile b/arch/arm/plat-pxa/Makefile
index f302d04..af8e484 100644
--- a/arch/arm/plat-pxa/Makefile
+++ b/arch/arm/plat-pxa/Makefile
@@ -8,5 +8,4 @@ obj-$(CONFIG_PXA3xx) += mfp.o
obj-$(CONFIG_PXA95x) += mfp.o
obj-$(CONFIG_ARCH_MMP) += mfp.o
-obj-$(CONFIG_HAVE_PWM) += pwm.o
obj-$(CONFIG_PXA_SSP) += ssp.o
diff --git a/arch/arm/plat-pxa/pwm.c b/arch/arm/plat-pxa/pwm.c
deleted file mode 100644
index ef32686..0000000
--- a/arch/arm/plat-pxa/pwm.c
+++ /dev/null
@@ -1,304 +0,0 @@
-/*
- * linux/arch/arm/mach-pxa/pwm.c
- *
- * simple driver for PWM (Pulse Width Modulator) controller
- *
- * 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.
- *
- * 2008-02-13 initial version
- * eric miao <eric.miao-eYqpPyKDWXRBDgjK7y7TUQ@public.gmane.org>
- */
-
-#include <linux/module.h>
-#include <linux/kernel.h>
-#include <linux/platform_device.h>
-#include <linux/slab.h>
-#include <linux/err.h>
-#include <linux/clk.h>
-#include <linux/io.h>
-#include <linux/pwm.h>
-
-#include <asm/div64.h>
-
-#define HAS_SECONDARY_PWM 0x10
-#define PWM_ID_BASE(d) ((d) & 0xf)
-
-static const struct platform_device_id pwm_id_table[] = {
- /* PWM has_secondary_pwm? */
- { "pxa25x-pwm", 0 },
- { "pxa27x-pwm", 0 | HAS_SECONDARY_PWM },
- { "pxa168-pwm", 1 },
- { "pxa910-pwm", 1 },
- { },
-};
-MODULE_DEVICE_TABLE(platform, pwm_id_table);
-
-/* PWM registers and bits definitions */
-#define PWMCR (0x00)
-#define PWMDCR (0x04)
-#define PWMPCR (0x08)
-
-#define PWMCR_SD (1 << 6)
-#define PWMDCR_FD (1 << 10)
-
-struct pwm_device {
- struct list_head node;
- struct pwm_device *secondary;
- struct platform_device *pdev;
-
- const char *label;
- struct clk *clk;
- int clk_enabled;
- void __iomem *mmio_base;
-
- unsigned int use_count;
- unsigned int pwm_id;
-};
-
-/*
- * period_ns = 10^9 * (PRESCALE + 1) * (PV + 1) / PWM_CLK_RATE
- * duty_ns = 10^9 * (PRESCALE + 1) * DC / PWM_CLK_RATE
- */
-int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
-{
- unsigned long long c;
- unsigned long period_cycles, prescale, pv, dc;
-
- if (pwm == NULL || period_ns == 0 || duty_ns > period_ns)
- return -EINVAL;
-
- c = clk_get_rate(pwm->clk);
- c = c * period_ns;
- do_div(c, 1000000000);
- period_cycles = c;
-
- if (period_cycles < 1)
- period_cycles = 1;
- prescale = (period_cycles - 1) / 1024;
- pv = period_cycles / (prescale + 1) - 1;
-
- if (prescale > 63)
- return -EINVAL;
-
- if (duty_ns == period_ns)
- dc = PWMDCR_FD;
- else
- dc = (pv + 1) * duty_ns / period_ns;
-
- /* NOTE: the clock to PWM has to be enabled first
- * before writing to the registers
- */
- clk_enable(pwm->clk);
- __raw_writel(prescale, pwm->mmio_base + PWMCR);
- __raw_writel(dc, pwm->mmio_base + PWMDCR);
- __raw_writel(pv, pwm->mmio_base + PWMPCR);
- clk_disable(pwm->clk);
-
- return 0;
-}
-EXPORT_SYMBOL(pwm_config);
-
-int pwm_enable(struct pwm_device *pwm)
-{
- int rc = 0;
-
- if (!pwm->clk_enabled) {
- rc = clk_enable(pwm->clk);
- if (!rc)
- pwm->clk_enabled = 1;
- }
- return rc;
-}
-EXPORT_SYMBOL(pwm_enable);
-
-void pwm_disable(struct pwm_device *pwm)
-{
- if (pwm->clk_enabled) {
- clk_disable(pwm->clk);
- pwm->clk_enabled = 0;
- }
-}
-EXPORT_SYMBOL(pwm_disable);
-
-static DEFINE_MUTEX(pwm_lock);
-static LIST_HEAD(pwm_list);
-
-struct pwm_device *pwm_request(int pwm_id, const char *label)
-{
- struct pwm_device *pwm;
- int found = 0;
-
- mutex_lock(&pwm_lock);
-
- list_for_each_entry(pwm, &pwm_list, node) {
- if (pwm->pwm_id == pwm_id) {
- found = 1;
- break;
- }
- }
-
- if (found) {
- if (pwm->use_count == 0) {
- pwm->use_count++;
- pwm->label = label;
- } else
- pwm = ERR_PTR(-EBUSY);
- } else
- pwm = ERR_PTR(-ENOENT);
-
- mutex_unlock(&pwm_lock);
- return pwm;
-}
-EXPORT_SYMBOL(pwm_request);
-
-void pwm_free(struct pwm_device *pwm)
-{
- mutex_lock(&pwm_lock);
-
- if (pwm->use_count) {
- pwm->use_count--;
- pwm->label = NULL;
- } else
- pr_warning("PWM device already freed\n");
-
- mutex_unlock(&pwm_lock);
-}
-EXPORT_SYMBOL(pwm_free);
-
-static inline void __add_pwm(struct pwm_device *pwm)
-{
- mutex_lock(&pwm_lock);
- list_add_tail(&pwm->node, &pwm_list);
- mutex_unlock(&pwm_lock);
-}
-
-static int __devinit pwm_probe(struct platform_device *pdev)
-{
- const struct platform_device_id *id = platform_get_device_id(pdev);
- struct pwm_device *pwm, *secondary = NULL;
- struct resource *r;
- int ret = 0;
-
- pwm = kzalloc(sizeof(struct pwm_device), GFP_KERNEL);
- if (pwm == NULL) {
- dev_err(&pdev->dev, "failed to allocate memory\n");
- return -ENOMEM;
- }
-
- pwm->clk = clk_get(&pdev->dev, NULL);
- if (IS_ERR(pwm->clk)) {
- ret = PTR_ERR(pwm->clk);
- goto err_free;
- }
- pwm->clk_enabled = 0;
-
- pwm->use_count = 0;
- pwm->pwm_id = PWM_ID_BASE(id->driver_data) + pdev->id;
- pwm->pdev = pdev;
-
- r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
- if (r == NULL) {
- dev_err(&pdev->dev, "no memory resource defined\n");
- ret = -ENODEV;
- goto err_free_clk;
- }
-
- r = request_mem_region(r->start, resource_size(r), pdev->name);
- if (r == NULL) {
- dev_err(&pdev->dev, "failed to request memory resource\n");
- ret = -EBUSY;
- goto err_free_clk;
- }
-
- pwm->mmio_base = ioremap(r->start, resource_size(r));
- if (pwm->mmio_base == NULL) {
- dev_err(&pdev->dev, "failed to ioremap() registers\n");
- ret = -ENODEV;
- goto err_free_mem;
- }
-
- if (id->driver_data & HAS_SECONDARY_PWM) {
- secondary = kzalloc(sizeof(struct pwm_device), GFP_KERNEL);
- if (secondary == NULL) {
- ret = -ENOMEM;
- goto err_free_mem;
- }
-
- *secondary = *pwm;
- pwm->secondary = secondary;
-
- /* registers for the second PWM has offset of 0x10 */
- secondary->mmio_base = pwm->mmio_base + 0x10;
- secondary->pwm_id = pdev->id + 2;
- }
-
- __add_pwm(pwm);
- if (secondary)
- __add_pwm(secondary);
-
- platform_set_drvdata(pdev, pwm);
- return 0;
-
-err_free_mem:
- release_mem_region(r->start, resource_size(r));
-err_free_clk:
- clk_put(pwm->clk);
-err_free:
- kfree(pwm);
- return ret;
-}
-
-static int __devexit pwm_remove(struct platform_device *pdev)
-{
- struct pwm_device *pwm;
- struct resource *r;
-
- pwm = platform_get_drvdata(pdev);
- if (pwm == NULL)
- return -ENODEV;
-
- mutex_lock(&pwm_lock);
-
- if (pwm->secondary) {
- list_del(&pwm->secondary->node);
- kfree(pwm->secondary);
- }
-
- list_del(&pwm->node);
- mutex_unlock(&pwm_lock);
-
- iounmap(pwm->mmio_base);
-
- r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
- release_mem_region(r->start, resource_size(r));
-
- clk_put(pwm->clk);
- kfree(pwm);
- return 0;
-}
-
-static struct platform_driver pwm_driver = {
- .driver = {
- .name = "pxa25x-pwm",
- .owner = THIS_MODULE,
- },
- .probe = pwm_probe,
- .remove = __devexit_p(pwm_remove),
- .id_table = pwm_id_table,
-};
-
-static int __init pwm_init(void)
-{
- return platform_driver_register(&pwm_driver);
-}
-arch_initcall(pwm_init);
-
-static void __exit pwm_exit(void)
-{
- platform_driver_unregister(&pwm_driver);
-}
-module_exit(pwm_exit);
-
-MODULE_LICENSE("GPL v2");
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index eb54042..0ef4f30 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -18,6 +18,15 @@ config PWM_BFIN
To compile this driver as a module, choose M here: the module
will be called pwm-bfin.
+config PWM_PXA
+ tristate "PXA PWM support"
+ depends on ARCH_PXA
+ help
+ Generic PWM framework driver for PXA.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-pxa.
+
config PWM_TEGRA
tristate "NVIDIA Tegra PWM support"
depends on ARCH_TEGRA
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 251b8d2..e859c51 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1,3 +1,4 @@
obj-$(CONFIG_PWM) += core.o
obj-$(CONFIG_PWM_BFIN) += pwm-bfin.o
+obj-$(CONFIG_PWM_PXA) += pwm-pxa.o
obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-pxa.c b/drivers/pwm/pwm-pxa.c
new file mode 100644
index 0000000..6434b18
--- /dev/null
+++ b/drivers/pwm/pwm-pxa.c
@@ -0,0 +1,243 @@
+/*
+ * linux/arch/arm/mach-pxa/pwm.c
+ *
+ * simple driver for PWM (Pulse Width Modulator) controller
+ *
+ * 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.
+ *
+ * 2008-02-13 initial version
+ * eric miao <eric.miao-eYqpPyKDWXRBDgjK7y7TUQ@public.gmane.org>
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/pwm.h>
+
+#include <asm/div64.h>
+
+#define HAS_SECONDARY_PWM 0x10
+#define PWM_ID_BASE(d) ((d) & 0xf)
+
+static const struct platform_device_id pwm_id_table[] = {
+ /* PWM has_secondary_pwm? */
+ { "pxa25x-pwm", 0 },
+ { "pxa27x-pwm", 0 | HAS_SECONDARY_PWM },
+ { "pxa168-pwm", 1 },
+ { "pxa910-pwm", 1 },
+ { },
+};
+MODULE_DEVICE_TABLE(platform, pwm_id_table);
+
+/* PWM registers and bits definitions */
+#define PWMCR (0x00)
+#define PWMDCR (0x04)
+#define PWMPCR (0x08)
+
+#define PWMCR_SD (1 << 6)
+#define PWMDCR_FD (1 << 10)
+
+struct pxa_pwm_chip {
+ struct pwm_chip chip;
+ struct device *dev;
+
+ struct clk *clk;
+ int clk_enabled;
+ void __iomem *mmio_base;
+};
+
+static inline struct pxa_pwm_chip *to_pxa_pwm_chip(struct pwm_chip *chip)
+{
+ return container_of(chip, struct pxa_pwm_chip, chip);
+}
+
+/*
+ * period_ns = 10^9 * (PRESCALE + 1) * (PV + 1) / PWM_CLK_RATE
+ * duty_ns = 10^9 * (PRESCALE + 1) * DC / PWM_CLK_RATE
+ */
+static int pxa_pwm_config(struct pwm_chip *chip, unsigned int pwm, int duty_ns, int period_ns)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+ unsigned long long c;
+ unsigned long period_cycles, prescale, pv, dc;
+ unsigned long offset;
+
+ if (period_ns == 0 || duty_ns > period_ns)
+ return -EINVAL;
+
+ offset = (pwm - chip->base) ? 0x10 : 0;
+
+ c = clk_get_rate(pc->clk);
+ c = c * period_ns;
+ do_div(c, 1000000000);
+ period_cycles = c;
+
+ if (period_cycles < 1)
+ period_cycles = 1;
+ prescale = (period_cycles - 1) / 1024;
+ pv = period_cycles / (prescale + 1) - 1;
+
+ if (prescale > 63)
+ return -EINVAL;
+
+ if (duty_ns == period_ns)
+ dc = PWMDCR_FD;
+ else
+ dc = (pv + 1) * duty_ns / period_ns;
+
+ /* NOTE: the clock to PWM has to be enabled first
+ * before writing to the registers
+ */
+ clk_enable(pc->clk);
+ __raw_writel(prescale, pc->mmio_base + offset + PWMCR);
+ __raw_writel(dc, pc->mmio_base + offset + PWMDCR);
+ __raw_writel(pv, pc->mmio_base + offset + PWMPCR);
+ clk_disable(pc->clk);
+
+ return 0;
+}
+
+static int pxa_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+ int rc = 0;
+
+ if (!pc->clk_enabled) {
+ rc = clk_enable(pc->clk);
+ if (!rc)
+ pc->clk_enabled = 1;
+ }
+ return rc;
+}
+
+static void pxa_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+
+ if (pc->clk_enabled) {
+ clk_disable(pc->clk);
+ pc->clk_enabled = 0;
+ }
+}
+
+static struct pwm_ops pxa_pwm_ops = {
+ .config = pxa_pwm_config,
+ .enable = pxa_pwm_enable,
+ .disable = pxa_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int __devinit pwm_probe(struct platform_device *pdev)
+{
+ const struct platform_device_id *id = platform_get_device_id(pdev);
+ struct pxa_pwm_chip *pwm;
+ struct resource *r;
+ int ret = 0;
+
+ pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
+ if (pwm == NULL) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ pwm->clk = clk_get(&pdev->dev, NULL);
+ if (IS_ERR(pwm->clk)) {
+ ret = PTR_ERR(pwm->clk);
+ goto err_free;
+ }
+ pwm->clk_enabled = 0;
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &pxa_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = (id->driver_data & HAS_SECONDARY_PWM) ? 2 : 1;
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (r == NULL) {
+ dev_err(&pdev->dev, "no memory resource defined\n");
+ ret = -ENODEV;
+ goto err_free_clk;
+ }
+
+ r = request_mem_region(r->start, resource_size(r), pdev->name);
+ if (r == NULL) {
+ dev_err(&pdev->dev, "failed to request memory resource\n");
+ ret = -EBUSY;
+ goto err_free_clk;
+ }
+
+ pwm->mmio_base = ioremap(r->start, resource_size(r));
+ if (pwm->mmio_base == NULL) {
+ dev_err(&pdev->dev, "failed to ioremap() registers\n");
+ ret = -ENODEV;
+ goto err_free_mem;
+ }
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ return ret;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+ return 0;
+
+err_free_mem:
+ release_mem_region(r->start, resource_size(r));
+err_free_clk:
+ clk_put(pwm->clk);
+err_free:
+ kfree(pwm);
+ return ret;
+}
+
+static int __devexit pwm_remove(struct platform_device *pdev)
+{
+ struct pxa_pwm_chip *chip;
+ struct resource *r;
+
+ chip = platform_get_drvdata(pdev);
+ if (chip == NULL)
+ return -ENODEV;
+
+ pwmchip_remove(&chip->chip);
+
+ iounmap(chip->mmio_base);
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ release_mem_region(r->start, resource_size(r));
+
+ clk_put(chip->clk);
+ kfree(chip);
+ return 0;
+}
+
+static struct platform_driver pwm_driver = {
+ .driver = {
+ .name = "pxa25x-pwm",
+ .owner = THIS_MODULE,
+ },
+ .probe = pwm_probe,
+ .remove = __devexit_p(pwm_remove),
+ .id_table = pwm_id_table,
+};
+
+static int __init pwm_init(void)
+{
+ return platform_driver_register(&pwm_driver);
+}
+arch_initcall(pwm_init);
+
+static void __exit pwm_exit(void)
+{
+ platform_driver_unregister(&pwm_driver);
+}
+module_exit(pwm_exit);
+
+MODULE_LICENSE("GPL v2");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 09/10] pwm: Add PXA support
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
arch/arm/plat-pxa/Makefile | 1 -
arch/arm/plat-pxa/pwm.c | 304 --------------------------------------------
drivers/pwm/Kconfig | 9 ++
drivers/pwm/Makefile | 1 +
drivers/pwm/pwm-pxa.c | 243 +++++++++++++++++++++++++++++++++++
5 files changed, 253 insertions(+), 305 deletions(-)
delete mode 100644 arch/arm/plat-pxa/pwm.c
create mode 100644 drivers/pwm/pwm-pxa.c
diff --git a/arch/arm/plat-pxa/Makefile b/arch/arm/plat-pxa/Makefile
index f302d04..af8e484 100644
--- a/arch/arm/plat-pxa/Makefile
+++ b/arch/arm/plat-pxa/Makefile
@@ -8,5 +8,4 @@ obj-$(CONFIG_PXA3xx) += mfp.o
obj-$(CONFIG_PXA95x) += mfp.o
obj-$(CONFIG_ARCH_MMP) += mfp.o
-obj-$(CONFIG_HAVE_PWM) += pwm.o
obj-$(CONFIG_PXA_SSP) += ssp.o
diff --git a/arch/arm/plat-pxa/pwm.c b/arch/arm/plat-pxa/pwm.c
deleted file mode 100644
index ef32686..0000000
--- a/arch/arm/plat-pxa/pwm.c
+++ /dev/null
@@ -1,304 +0,0 @@
-/*
- * linux/arch/arm/mach-pxa/pwm.c
- *
- * simple driver for PWM (Pulse Width Modulator) controller
- *
- * 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.
- *
- * 2008-02-13 initial version
- * eric miao <eric.miao@marvell.com>
- */
-
-#include <linux/module.h>
-#include <linux/kernel.h>
-#include <linux/platform_device.h>
-#include <linux/slab.h>
-#include <linux/err.h>
-#include <linux/clk.h>
-#include <linux/io.h>
-#include <linux/pwm.h>
-
-#include <asm/div64.h>
-
-#define HAS_SECONDARY_PWM 0x10
-#define PWM_ID_BASE(d) ((d) & 0xf)
-
-static const struct platform_device_id pwm_id_table[] = {
- /* PWM has_secondary_pwm? */
- { "pxa25x-pwm", 0 },
- { "pxa27x-pwm", 0 | HAS_SECONDARY_PWM },
- { "pxa168-pwm", 1 },
- { "pxa910-pwm", 1 },
- { },
-};
-MODULE_DEVICE_TABLE(platform, pwm_id_table);
-
-/* PWM registers and bits definitions */
-#define PWMCR (0x00)
-#define PWMDCR (0x04)
-#define PWMPCR (0x08)
-
-#define PWMCR_SD (1 << 6)
-#define PWMDCR_FD (1 << 10)
-
-struct pwm_device {
- struct list_head node;
- struct pwm_device *secondary;
- struct platform_device *pdev;
-
- const char *label;
- struct clk *clk;
- int clk_enabled;
- void __iomem *mmio_base;
-
- unsigned int use_count;
- unsigned int pwm_id;
-};
-
-/*
- * period_ns = 10^9 * (PRESCALE + 1) * (PV + 1) / PWM_CLK_RATE
- * duty_ns = 10^9 * (PRESCALE + 1) * DC / PWM_CLK_RATE
- */
-int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
-{
- unsigned long long c;
- unsigned long period_cycles, prescale, pv, dc;
-
- if (pwm == NULL || period_ns == 0 || duty_ns > period_ns)
- return -EINVAL;
-
- c = clk_get_rate(pwm->clk);
- c = c * period_ns;
- do_div(c, 1000000000);
- period_cycles = c;
-
- if (period_cycles < 1)
- period_cycles = 1;
- prescale = (period_cycles - 1) / 1024;
- pv = period_cycles / (prescale + 1) - 1;
-
- if (prescale > 63)
- return -EINVAL;
-
- if (duty_ns == period_ns)
- dc = PWMDCR_FD;
- else
- dc = (pv + 1) * duty_ns / period_ns;
-
- /* NOTE: the clock to PWM has to be enabled first
- * before writing to the registers
- */
- clk_enable(pwm->clk);
- __raw_writel(prescale, pwm->mmio_base + PWMCR);
- __raw_writel(dc, pwm->mmio_base + PWMDCR);
- __raw_writel(pv, pwm->mmio_base + PWMPCR);
- clk_disable(pwm->clk);
-
- return 0;
-}
-EXPORT_SYMBOL(pwm_config);
-
-int pwm_enable(struct pwm_device *pwm)
-{
- int rc = 0;
-
- if (!pwm->clk_enabled) {
- rc = clk_enable(pwm->clk);
- if (!rc)
- pwm->clk_enabled = 1;
- }
- return rc;
-}
-EXPORT_SYMBOL(pwm_enable);
-
-void pwm_disable(struct pwm_device *pwm)
-{
- if (pwm->clk_enabled) {
- clk_disable(pwm->clk);
- pwm->clk_enabled = 0;
- }
-}
-EXPORT_SYMBOL(pwm_disable);
-
-static DEFINE_MUTEX(pwm_lock);
-static LIST_HEAD(pwm_list);
-
-struct pwm_device *pwm_request(int pwm_id, const char *label)
-{
- struct pwm_device *pwm;
- int found = 0;
-
- mutex_lock(&pwm_lock);
-
- list_for_each_entry(pwm, &pwm_list, node) {
- if (pwm->pwm_id == pwm_id) {
- found = 1;
- break;
- }
- }
-
- if (found) {
- if (pwm->use_count == 0) {
- pwm->use_count++;
- pwm->label = label;
- } else
- pwm = ERR_PTR(-EBUSY);
- } else
- pwm = ERR_PTR(-ENOENT);
-
- mutex_unlock(&pwm_lock);
- return pwm;
-}
-EXPORT_SYMBOL(pwm_request);
-
-void pwm_free(struct pwm_device *pwm)
-{
- mutex_lock(&pwm_lock);
-
- if (pwm->use_count) {
- pwm->use_count--;
- pwm->label = NULL;
- } else
- pr_warning("PWM device already freed\n");
-
- mutex_unlock(&pwm_lock);
-}
-EXPORT_SYMBOL(pwm_free);
-
-static inline void __add_pwm(struct pwm_device *pwm)
-{
- mutex_lock(&pwm_lock);
- list_add_tail(&pwm->node, &pwm_list);
- mutex_unlock(&pwm_lock);
-}
-
-static int __devinit pwm_probe(struct platform_device *pdev)
-{
- const struct platform_device_id *id = platform_get_device_id(pdev);
- struct pwm_device *pwm, *secondary = NULL;
- struct resource *r;
- int ret = 0;
-
- pwm = kzalloc(sizeof(struct pwm_device), GFP_KERNEL);
- if (pwm == NULL) {
- dev_err(&pdev->dev, "failed to allocate memory\n");
- return -ENOMEM;
- }
-
- pwm->clk = clk_get(&pdev->dev, NULL);
- if (IS_ERR(pwm->clk)) {
- ret = PTR_ERR(pwm->clk);
- goto err_free;
- }
- pwm->clk_enabled = 0;
-
- pwm->use_count = 0;
- pwm->pwm_id = PWM_ID_BASE(id->driver_data) + pdev->id;
- pwm->pdev = pdev;
-
- r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
- if (r == NULL) {
- dev_err(&pdev->dev, "no memory resource defined\n");
- ret = -ENODEV;
- goto err_free_clk;
- }
-
- r = request_mem_region(r->start, resource_size(r), pdev->name);
- if (r == NULL) {
- dev_err(&pdev->dev, "failed to request memory resource\n");
- ret = -EBUSY;
- goto err_free_clk;
- }
-
- pwm->mmio_base = ioremap(r->start, resource_size(r));
- if (pwm->mmio_base == NULL) {
- dev_err(&pdev->dev, "failed to ioremap() registers\n");
- ret = -ENODEV;
- goto err_free_mem;
- }
-
- if (id->driver_data & HAS_SECONDARY_PWM) {
- secondary = kzalloc(sizeof(struct pwm_device), GFP_KERNEL);
- if (secondary == NULL) {
- ret = -ENOMEM;
- goto err_free_mem;
- }
-
- *secondary = *pwm;
- pwm->secondary = secondary;
-
- /* registers for the second PWM has offset of 0x10 */
- secondary->mmio_base = pwm->mmio_base + 0x10;
- secondary->pwm_id = pdev->id + 2;
- }
-
- __add_pwm(pwm);
- if (secondary)
- __add_pwm(secondary);
-
- platform_set_drvdata(pdev, pwm);
- return 0;
-
-err_free_mem:
- release_mem_region(r->start, resource_size(r));
-err_free_clk:
- clk_put(pwm->clk);
-err_free:
- kfree(pwm);
- return ret;
-}
-
-static int __devexit pwm_remove(struct platform_device *pdev)
-{
- struct pwm_device *pwm;
- struct resource *r;
-
- pwm = platform_get_drvdata(pdev);
- if (pwm == NULL)
- return -ENODEV;
-
- mutex_lock(&pwm_lock);
-
- if (pwm->secondary) {
- list_del(&pwm->secondary->node);
- kfree(pwm->secondary);
- }
-
- list_del(&pwm->node);
- mutex_unlock(&pwm_lock);
-
- iounmap(pwm->mmio_base);
-
- r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
- release_mem_region(r->start, resource_size(r));
-
- clk_put(pwm->clk);
- kfree(pwm);
- return 0;
-}
-
-static struct platform_driver pwm_driver = {
- .driver = {
- .name = "pxa25x-pwm",
- .owner = THIS_MODULE,
- },
- .probe = pwm_probe,
- .remove = __devexit_p(pwm_remove),
- .id_table = pwm_id_table,
-};
-
-static int __init pwm_init(void)
-{
- return platform_driver_register(&pwm_driver);
-}
-arch_initcall(pwm_init);
-
-static void __exit pwm_exit(void)
-{
- platform_driver_unregister(&pwm_driver);
-}
-module_exit(pwm_exit);
-
-MODULE_LICENSE("GPL v2");
diff --git a/drivers/pwm/Kconfig b/drivers/pwm/Kconfig
index eb54042..0ef4f30 100644
--- a/drivers/pwm/Kconfig
+++ b/drivers/pwm/Kconfig
@@ -18,6 +18,15 @@ config PWM_BFIN
To compile this driver as a module, choose M here: the module
will be called pwm-bfin.
+config PWM_PXA
+ tristate "PXA PWM support"
+ depends on ARCH_PXA
+ help
+ Generic PWM framework driver for PXA.
+
+ To compile this driver as a module, choose M here: the module
+ will be called pwm-pxa.
+
config PWM_TEGRA
tristate "NVIDIA Tegra PWM support"
depends on ARCH_TEGRA
diff --git a/drivers/pwm/Makefile b/drivers/pwm/Makefile
index 251b8d2..e859c51 100644
--- a/drivers/pwm/Makefile
+++ b/drivers/pwm/Makefile
@@ -1,3 +1,4 @@
obj-$(CONFIG_PWM) += core.o
obj-$(CONFIG_PWM_BFIN) += pwm-bfin.o
+obj-$(CONFIG_PWM_PXA) += pwm-pxa.o
obj-$(CONFIG_PWM_TEGRA) += pwm-tegra.o
diff --git a/drivers/pwm/pwm-pxa.c b/drivers/pwm/pwm-pxa.c
new file mode 100644
index 0000000..6434b18
--- /dev/null
+++ b/drivers/pwm/pwm-pxa.c
@@ -0,0 +1,243 @@
+/*
+ * linux/arch/arm/mach-pxa/pwm.c
+ *
+ * simple driver for PWM (Pulse Width Modulator) controller
+ *
+ * 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.
+ *
+ * 2008-02-13 initial version
+ * eric miao <eric.miao@marvell.com>
+ */
+
+#include <linux/module.h>
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <linux/clk.h>
+#include <linux/io.h>
+#include <linux/pwm.h>
+
+#include <asm/div64.h>
+
+#define HAS_SECONDARY_PWM 0x10
+#define PWM_ID_BASE(d) ((d) & 0xf)
+
+static const struct platform_device_id pwm_id_table[] = {
+ /* PWM has_secondary_pwm? */
+ { "pxa25x-pwm", 0 },
+ { "pxa27x-pwm", 0 | HAS_SECONDARY_PWM },
+ { "pxa168-pwm", 1 },
+ { "pxa910-pwm", 1 },
+ { },
+};
+MODULE_DEVICE_TABLE(platform, pwm_id_table);
+
+/* PWM registers and bits definitions */
+#define PWMCR (0x00)
+#define PWMDCR (0x04)
+#define PWMPCR (0x08)
+
+#define PWMCR_SD (1 << 6)
+#define PWMDCR_FD (1 << 10)
+
+struct pxa_pwm_chip {
+ struct pwm_chip chip;
+ struct device *dev;
+
+ struct clk *clk;
+ int clk_enabled;
+ void __iomem *mmio_base;
+};
+
+static inline struct pxa_pwm_chip *to_pxa_pwm_chip(struct pwm_chip *chip)
+{
+ return container_of(chip, struct pxa_pwm_chip, chip);
+}
+
+/*
+ * period_ns = 10^9 * (PRESCALE + 1) * (PV + 1) / PWM_CLK_RATE
+ * duty_ns = 10^9 * (PRESCALE + 1) * DC / PWM_CLK_RATE
+ */
+static int pxa_pwm_config(struct pwm_chip *chip, unsigned int pwm, int duty_ns, int period_ns)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+ unsigned long long c;
+ unsigned long period_cycles, prescale, pv, dc;
+ unsigned long offset;
+
+ if (period_ns == 0 || duty_ns > period_ns)
+ return -EINVAL;
+
+ offset = (pwm - chip->base) ? 0x10 : 0;
+
+ c = clk_get_rate(pc->clk);
+ c = c * period_ns;
+ do_div(c, 1000000000);
+ period_cycles = c;
+
+ if (period_cycles < 1)
+ period_cycles = 1;
+ prescale = (period_cycles - 1) / 1024;
+ pv = period_cycles / (prescale + 1) - 1;
+
+ if (prescale > 63)
+ return -EINVAL;
+
+ if (duty_ns == period_ns)
+ dc = PWMDCR_FD;
+ else
+ dc = (pv + 1) * duty_ns / period_ns;
+
+ /* NOTE: the clock to PWM has to be enabled first
+ * before writing to the registers
+ */
+ clk_enable(pc->clk);
+ __raw_writel(prescale, pc->mmio_base + offset + PWMCR);
+ __raw_writel(dc, pc->mmio_base + offset + PWMDCR);
+ __raw_writel(pv, pc->mmio_base + offset + PWMPCR);
+ clk_disable(pc->clk);
+
+ return 0;
+}
+
+static int pxa_pwm_enable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+ int rc = 0;
+
+ if (!pc->clk_enabled) {
+ rc = clk_enable(pc->clk);
+ if (!rc)
+ pc->clk_enabled = 1;
+ }
+ return rc;
+}
+
+static void pxa_pwm_disable(struct pwm_chip *chip, unsigned int pwm)
+{
+ struct pxa_pwm_chip *pc = to_pxa_pwm_chip(chip);
+
+ if (pc->clk_enabled) {
+ clk_disable(pc->clk);
+ pc->clk_enabled = 0;
+ }
+}
+
+static struct pwm_ops pxa_pwm_ops = {
+ .config = pxa_pwm_config,
+ .enable = pxa_pwm_enable,
+ .disable = pxa_pwm_disable,
+ .owner = THIS_MODULE,
+};
+
+static int __devinit pwm_probe(struct platform_device *pdev)
+{
+ const struct platform_device_id *id = platform_get_device_id(pdev);
+ struct pxa_pwm_chip *pwm;
+ struct resource *r;
+ int ret = 0;
+
+ pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
+ if (pwm == NULL) {
+ dev_err(&pdev->dev, "failed to allocate memory\n");
+ return -ENOMEM;
+ }
+
+ pwm->clk = clk_get(&pdev->dev, NULL);
+ if (IS_ERR(pwm->clk)) {
+ ret = PTR_ERR(pwm->clk);
+ goto err_free;
+ }
+ pwm->clk_enabled = 0;
+
+ pwm->chip.dev = &pdev->dev;
+ pwm->chip.ops = &pxa_pwm_ops;
+ pwm->chip.base = -1;
+ pwm->chip.npwm = (id->driver_data & HAS_SECONDARY_PWM) ? 2 : 1;
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ if (r == NULL) {
+ dev_err(&pdev->dev, "no memory resource defined\n");
+ ret = -ENODEV;
+ goto err_free_clk;
+ }
+
+ r = request_mem_region(r->start, resource_size(r), pdev->name);
+ if (r == NULL) {
+ dev_err(&pdev->dev, "failed to request memory resource\n");
+ ret = -EBUSY;
+ goto err_free_clk;
+ }
+
+ pwm->mmio_base = ioremap(r->start, resource_size(r));
+ if (pwm->mmio_base == NULL) {
+ dev_err(&pdev->dev, "failed to ioremap() registers\n");
+ ret = -ENODEV;
+ goto err_free_mem;
+ }
+
+ ret = pwmchip_add(&pwm->chip);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "pwmchip_add() failed: %d\n", ret);
+ return ret;
+ }
+
+ platform_set_drvdata(pdev, pwm);
+ return 0;
+
+err_free_mem:
+ release_mem_region(r->start, resource_size(r));
+err_free_clk:
+ clk_put(pwm->clk);
+err_free:
+ kfree(pwm);
+ return ret;
+}
+
+static int __devexit pwm_remove(struct platform_device *pdev)
+{
+ struct pxa_pwm_chip *chip;
+ struct resource *r;
+
+ chip = platform_get_drvdata(pdev);
+ if (chip == NULL)
+ return -ENODEV;
+
+ pwmchip_remove(&chip->chip);
+
+ iounmap(chip->mmio_base);
+
+ r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+ release_mem_region(r->start, resource_size(r));
+
+ clk_put(chip->clk);
+ kfree(chip);
+ return 0;
+}
+
+static struct platform_driver pwm_driver = {
+ .driver = {
+ .name = "pxa25x-pwm",
+ .owner = THIS_MODULE,
+ },
+ .probe = pwm_probe,
+ .remove = __devexit_p(pwm_remove),
+ .id_table = pwm_id_table,
+};
+
+static int __init pwm_init(void)
+{
+ return platform_driver_register(&pwm_driver);
+}
+arch_initcall(pwm_init);
+
+static void __exit pwm_exit(void)
+{
+ platform_driver_unregister(&pwm_driver);
+}
+module_exit(pwm_exit);
+
+MODULE_LICENSE("GPL v2");
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 10/10] pwm-backlight: Add rudimentary device tree support
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 15:19 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ
Cc: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Sascha Hauer, Arnd Bergmann,
Matthias Kaehlcke, Kurt Van Dijck, Rob Herring, Grant Likely,
Colin Cross, Olof Johansson, Richard Purdie, Mark Brown,
Mitch Bradley, Mike Frysinger, Eric Miao
This commit adds very basic support for device tree probing. Currently,
only a PWM and maximum and default brightness values can be specified.
Enabling or disabling backlight power via GPIOs is not yet supported.
A pointer to the exit() callback is stored in the driver data to keep it
around until the driver is unloaded.
Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
---
Changes in v2:
- avoid oops by keeping a reference to the platform-specific exit()
callback
.../bindings/video/backlight/pwm-backlight | 16 ++++
drivers/video/backlight/Kconfig | 2 +-
drivers/video/backlight/pwm_bl.c | 81 ++++++++++++++++++--
3 files changed, 90 insertions(+), 9 deletions(-)
create mode 100644 Documentation/devicetree/bindings/video/backlight/pwm-backlight
diff --git a/Documentation/devicetree/bindings/video/backlight/pwm-backlight b/Documentation/devicetree/bindings/video/backlight/pwm-backlight
new file mode 100644
index 0000000..ce65280
--- /dev/null
+++ b/Documentation/devicetree/bindings/video/backlight/pwm-backlight
@@ -0,0 +1,16 @@
+pwm-backlight bindings
+
+Required properties:
+ - compatible: "pwm-backlight"
+ - default-brightness: the default brightness setting
+ - max-brightness: the maximum brightness setting
+ - pwm: OF device-tree PWM specification
+
+Example:
+
+ backlight {
+ compatible = "pwm-backlight";
+ default-brightness = <224>;
+ max-brightness = <255>;
+ pwm = <&pwm 0 5000000>;
+ };
diff --git a/drivers/video/backlight/Kconfig b/drivers/video/backlight/Kconfig
index 49e7d83..37982fa 100644
--- a/drivers/video/backlight/Kconfig
+++ b/drivers/video/backlight/Kconfig
@@ -233,7 +233,7 @@ config BACKLIGHT_CARILLO_RANCH
config BACKLIGHT_PWM
tristate "Generic PWM based Backlight Driver"
- depends on HAVE_PWM
+ depends on HAVE_PWM || PWM
help
If you have a LCD backlight adjustable by PWM, say Y to enable
this driver.
diff --git a/drivers/video/backlight/pwm_bl.c b/drivers/video/backlight/pwm_bl.c
index 342b7d7..5c81397 100644
--- a/drivers/video/backlight/pwm_bl.c
+++ b/drivers/video/backlight/pwm_bl.c
@@ -17,6 +17,7 @@
#include <linux/fb.h>
#include <linux/backlight.h>
#include <linux/err.h>
+#include <linux/of_pwm.h>
#include <linux/pwm.h>
#include <linux/pwm_backlight.h>
#include <linux/slab.h>
@@ -31,6 +32,7 @@ struct pwm_bl_data {
void (*notify_after)(struct device *,
int brightness);
int (*check_fb)(struct device *, struct fb_info *);
+ void (*exit)(struct device *);
};
static int pwm_backlight_update_status(struct backlight_device *bl)
@@ -83,17 +85,79 @@ static const struct backlight_ops pwm_backlight_ops = {
.check_fb = pwm_backlight_check_fb,
};
+#ifdef CONFIG_OF
+static int pwm_backlight_parse_dt(struct device *dev,
+ struct platform_pwm_backlight_data *data)
+{
+ struct device_node *node = dev->of_node;
+ struct pwm_spec spec;
+ u32 value;
+ int ret;
+
+ if (!node)
+ return -ENODEV;
+
+ memset(data, 0, sizeof(*data));
+
+ ret = of_get_named_pwm(node, "pwm", 0, &spec);
+ if (ret < 0)
+ return ret;
+
+ data->pwm_period_ns = spec.period;
+ data->pwm_id = ret;
+
+ ret = of_property_read_u32(node, "default-brightness", &value);
+ if (ret < 0)
+ return ret;
+
+ data->dft_brightness = value;
+
+ ret = of_property_read_u32(node, "max-brightness", &value);
+ if (ret < 0)
+ return ret;
+
+ data->max_brightness = value;
+
+ /*
+ * TODO: Most users of this driver use a number of GPIOs to control
+ * backlight power. Support for specifying these needs to be
+ * added.
+ */
+
+ return 0;
+}
+
+static struct of_device_id pwm_backlight_of_match[] = {
+ { .compatible = "pwm-backlight" },
+ { }
+};
+
+MODULE_DEVICE_TABLE(of, pwm_backlight_of_match);
+#else
+static int pwm_backlight_parse_dt(struct device *dev,
+ struct platform_pwm_backlight_data *data)
+{
+ return -ENODEV;
+}
+#endif
+
static int pwm_backlight_probe(struct platform_device *pdev)
{
struct backlight_properties props;
struct platform_pwm_backlight_data *data = pdev->dev.platform_data;
+ struct platform_pwm_backlight_data defdata;
struct backlight_device *bl;
struct pwm_bl_data *pb;
int ret;
if (!data) {
- dev_err(&pdev->dev, "failed to find platform data\n");
- return -EINVAL;
+ ret = pwm_backlight_parse_dt(&pdev->dev, &defdata);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "failed to find platform data\n");
+ return ret;
+ }
+
+ data = &defdata;
}
if (data->init) {
@@ -113,6 +177,7 @@ static int pwm_backlight_probe(struct platform_device *pdev)
pb->notify = data->notify;
pb->notify_after = data->notify_after;
pb->check_fb = data->check_fb;
+ pb->exit = data->exit;
pb->lth_brightness = data->lth_brightness *
(data->pwm_period_ns / data->max_brightness);
pb->dev = &pdev->dev;
@@ -152,7 +217,6 @@ err_alloc:
static int pwm_backlight_remove(struct platform_device *pdev)
{
- struct platform_pwm_backlight_data *data = pdev->dev.platform_data;
struct backlight_device *bl = platform_get_drvdata(pdev);
struct pwm_bl_data *pb = dev_get_drvdata(&bl->dev);
@@ -160,8 +224,8 @@ static int pwm_backlight_remove(struct platform_device *pdev)
pwm_config(pb->pwm, 0, pb->period);
pwm_disable(pb->pwm);
pwm_free(pb->pwm);
- if (data->exit)
- data->exit(&pdev->dev);
+ if (pb->exit)
+ pb->exit(&pdev->dev);
return 0;
}
@@ -195,11 +259,12 @@ static SIMPLE_DEV_PM_OPS(pwm_backlight_pm_ops, pwm_backlight_suspend,
static struct platform_driver pwm_backlight_driver = {
.driver = {
- .name = "pwm-backlight",
- .owner = THIS_MODULE,
+ .name = "pwm-backlight",
+ .owner = THIS_MODULE,
#ifdef CONFIG_PM
- .pm = &pwm_backlight_pm_ops,
+ .pm = &pwm_backlight_pm_ops,
#endif
+ .of_match_table = of_match_ptr(pwm_backlight_of_match),
},
.probe = pwm_backlight_probe,
.remove = pwm_backlight_remove,
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* [PATCH v2 10/10] pwm-backlight: Add rudimentary device tree support
@ 2012-02-06 15:19 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-06 15:19 UTC (permalink / raw)
To: linux-arm-kernel
This commit adds very basic support for device tree probing. Currently,
only a PWM and maximum and default brightness values can be specified.
Enabling or disabling backlight power via GPIOs is not yet supported.
A pointer to the exit() callback is stored in the driver data to keep it
around until the driver is unloaded.
Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
---
Changes in v2:
- avoid oops by keeping a reference to the platform-specific exit()
callback
.../bindings/video/backlight/pwm-backlight | 16 ++++
drivers/video/backlight/Kconfig | 2 +-
drivers/video/backlight/pwm_bl.c | 81 ++++++++++++++++++--
3 files changed, 90 insertions(+), 9 deletions(-)
create mode 100644 Documentation/devicetree/bindings/video/backlight/pwm-backlight
diff --git a/Documentation/devicetree/bindings/video/backlight/pwm-backlight b/Documentation/devicetree/bindings/video/backlight/pwm-backlight
new file mode 100644
index 0000000..ce65280
--- /dev/null
+++ b/Documentation/devicetree/bindings/video/backlight/pwm-backlight
@@ -0,0 +1,16 @@
+pwm-backlight bindings
+
+Required properties:
+ - compatible: "pwm-backlight"
+ - default-brightness: the default brightness setting
+ - max-brightness: the maximum brightness setting
+ - pwm: OF device-tree PWM specification
+
+Example:
+
+ backlight {
+ compatible = "pwm-backlight";
+ default-brightness = <224>;
+ max-brightness = <255>;
+ pwm = <&pwm 0 5000000>;
+ };
diff --git a/drivers/video/backlight/Kconfig b/drivers/video/backlight/Kconfig
index 49e7d83..37982fa 100644
--- a/drivers/video/backlight/Kconfig
+++ b/drivers/video/backlight/Kconfig
@@ -233,7 +233,7 @@ config BACKLIGHT_CARILLO_RANCH
config BACKLIGHT_PWM
tristate "Generic PWM based Backlight Driver"
- depends on HAVE_PWM
+ depends on HAVE_PWM || PWM
help
If you have a LCD backlight adjustable by PWM, say Y to enable
this driver.
diff --git a/drivers/video/backlight/pwm_bl.c b/drivers/video/backlight/pwm_bl.c
index 342b7d7..5c81397 100644
--- a/drivers/video/backlight/pwm_bl.c
+++ b/drivers/video/backlight/pwm_bl.c
@@ -17,6 +17,7 @@
#include <linux/fb.h>
#include <linux/backlight.h>
#include <linux/err.h>
+#include <linux/of_pwm.h>
#include <linux/pwm.h>
#include <linux/pwm_backlight.h>
#include <linux/slab.h>
@@ -31,6 +32,7 @@ struct pwm_bl_data {
void (*notify_after)(struct device *,
int brightness);
int (*check_fb)(struct device *, struct fb_info *);
+ void (*exit)(struct device *);
};
static int pwm_backlight_update_status(struct backlight_device *bl)
@@ -83,17 +85,79 @@ static const struct backlight_ops pwm_backlight_ops = {
.check_fb = pwm_backlight_check_fb,
};
+#ifdef CONFIG_OF
+static int pwm_backlight_parse_dt(struct device *dev,
+ struct platform_pwm_backlight_data *data)
+{
+ struct device_node *node = dev->of_node;
+ struct pwm_spec spec;
+ u32 value;
+ int ret;
+
+ if (!node)
+ return -ENODEV;
+
+ memset(data, 0, sizeof(*data));
+
+ ret = of_get_named_pwm(node, "pwm", 0, &spec);
+ if (ret < 0)
+ return ret;
+
+ data->pwm_period_ns = spec.period;
+ data->pwm_id = ret;
+
+ ret = of_property_read_u32(node, "default-brightness", &value);
+ if (ret < 0)
+ return ret;
+
+ data->dft_brightness = value;
+
+ ret = of_property_read_u32(node, "max-brightness", &value);
+ if (ret < 0)
+ return ret;
+
+ data->max_brightness = value;
+
+ /*
+ * TODO: Most users of this driver use a number of GPIOs to control
+ * backlight power. Support for specifying these needs to be
+ * added.
+ */
+
+ return 0;
+}
+
+static struct of_device_id pwm_backlight_of_match[] = {
+ { .compatible = "pwm-backlight" },
+ { }
+};
+
+MODULE_DEVICE_TABLE(of, pwm_backlight_of_match);
+#else
+static int pwm_backlight_parse_dt(struct device *dev,
+ struct platform_pwm_backlight_data *data)
+{
+ return -ENODEV;
+}
+#endif
+
static int pwm_backlight_probe(struct platform_device *pdev)
{
struct backlight_properties props;
struct platform_pwm_backlight_data *data = pdev->dev.platform_data;
+ struct platform_pwm_backlight_data defdata;
struct backlight_device *bl;
struct pwm_bl_data *pb;
int ret;
if (!data) {
- dev_err(&pdev->dev, "failed to find platform data\n");
- return -EINVAL;
+ ret = pwm_backlight_parse_dt(&pdev->dev, &defdata);
+ if (ret < 0) {
+ dev_err(&pdev->dev, "failed to find platform data\n");
+ return ret;
+ }
+
+ data = &defdata;
}
if (data->init) {
@@ -113,6 +177,7 @@ static int pwm_backlight_probe(struct platform_device *pdev)
pb->notify = data->notify;
pb->notify_after = data->notify_after;
pb->check_fb = data->check_fb;
+ pb->exit = data->exit;
pb->lth_brightness = data->lth_brightness *
(data->pwm_period_ns / data->max_brightness);
pb->dev = &pdev->dev;
@@ -152,7 +217,6 @@ err_alloc:
static int pwm_backlight_remove(struct platform_device *pdev)
{
- struct platform_pwm_backlight_data *data = pdev->dev.platform_data;
struct backlight_device *bl = platform_get_drvdata(pdev);
struct pwm_bl_data *pb = dev_get_drvdata(&bl->dev);
@@ -160,8 +224,8 @@ static int pwm_backlight_remove(struct platform_device *pdev)
pwm_config(pb->pwm, 0, pb->period);
pwm_disable(pb->pwm);
pwm_free(pb->pwm);
- if (data->exit)
- data->exit(&pdev->dev);
+ if (pb->exit)
+ pb->exit(&pdev->dev);
return 0;
}
@@ -195,11 +259,12 @@ static SIMPLE_DEV_PM_OPS(pwm_backlight_pm_ops, pwm_backlight_suspend,
static struct platform_driver pwm_backlight_driver = {
.driver = {
- .name = "pwm-backlight",
- .owner = THIS_MODULE,
+ .name = "pwm-backlight",
+ .owner = THIS_MODULE,
#ifdef CONFIG_PM
- .pm = &pwm_backlight_pm_ops,
+ .pm = &pwm_backlight_pm_ops,
#endif
+ .of_match_table = of_match_ptr(pwm_backlight_of_match),
},
.probe = pwm_backlight_probe,
.remove = pwm_backlight_remove,
--
1.7.9
^ permalink raw reply related [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-06 21:22 ` Lars-Peter Clausen
-1 siblings, 0 replies; 44+ messages in thread
From: Lars-Peter Clausen @ 2012-02-06 21:22 UTC (permalink / raw)
To: Thierry Reding
Cc: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mitch Bradley,
Mark Brown, Mike Frysinger, Arnd Bergmann, Sascha Hauer,
Colin Cross, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
On 02/06/2012 04:19 PM, Thierry Reding wrote:
> This commit modifies the PWM core to support multiple PWMs per struct
> pwm_chip.
I think you should mention what motivates this change.
> It achieves this in a similar way to how gpiolib works, by
> allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> starting at a given offset (pwm_chip.base >= 0).
If we've learned one thing from gpiolib, I think it is that using a global
index to identify a resource was a bad idea.
> A chip specifies how
> many PWMs it controls using the npwm member. Each of the functions in
> the pwm_ops structure gets an additional argument that specified the PWM
> number (it can be converted to a per-chip index by subtracting the
> chip's base).
>
> The total maximum number of PWM devices is currently fixed to 64, but
> can easily be made configurable via Kconfig.
The code says 1024.
>
> The patch is incomplete in that it doesn't convert any existing drivers
> that are now broken.
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-06 21:22 ` Lars-Peter Clausen
0 siblings, 0 replies; 44+ messages in thread
From: Lars-Peter Clausen @ 2012-02-06 21:22 UTC (permalink / raw)
To: linux-arm-kernel
On 02/06/2012 04:19 PM, Thierry Reding wrote:
> This commit modifies the PWM core to support multiple PWMs per struct
> pwm_chip.
I think you should mention what motivates this change.
> It achieves this in a similar way to how gpiolib works, by
> allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> starting at a given offset (pwm_chip.base >= 0).
If we've learned one thing from gpiolib, I think it is that using a global
index to identify a resource was a bad idea.
> A chip specifies how
> many PWMs it controls using the npwm member. Each of the functions in
> the pwm_ops structure gets an additional argument that specified the PWM
> number (it can be converted to a per-chip index by subtracting the
> chip's base).
>
> The total maximum number of PWM devices is currently fixed to 64, but
> can easily be made configurable via Kconfig.
The code says 1024.
>
> The patch is incomplete in that it doesn't convert any existing drivers
> that are now broken.
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-06 21:22 ` Lars-Peter Clausen
@ 2012-02-07 7:04 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-07 7:04 UTC (permalink / raw)
To: Lars-Peter Clausen
Cc: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mitch Bradley,
Mark Brown, Mike Frysinger, Arnd Bergmann, Sascha Hauer,
Colin Cross, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1: Type: text/plain, Size: 2226 bytes --]
* Lars-Peter Clausen wrote:
> On 02/06/2012 04:19 PM, Thierry Reding wrote:
> > This commit modifies the PWM core to support multiple PWMs per struct
> > pwm_chip.
>
> I think you should mention what motivates this change.
Okay, I can add that.
> > It achieves this in a similar way to how gpiolib works, by
> > allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> > starting at a given offset (pwm_chip.base >= 0).
>
> If we've learned one thing from gpiolib, I think it is that using a global
> index to identify a resource was a bad idea.
Yes, that concern has been raised several times already. The reason for it
being this way is that the current API requires it. So either we keep it as
is for now and change it over to indexing on a per-chip basis later (which
has the advantage of not breaking anything currently in-tree) or we make that
change now, in which case this patch series will probably double in size. I'm
also not sure if I can reasonably keep the series bisectible that way.
One of the major problems when converting to a non-global namespace is how to
represent the relationship in code. For device tree this should be easy to do
because it has all the infrastructure in place. For non-device-tree devices I
have no idea yet how this could be done. Perhaps by using something like the
clock API and using names for lookup?
> > A chip specifies how
> > many PWMs it controls using the npwm member. Each of the functions in
> > the pwm_ops structure gets an additional argument that specified the PWM
> > number (it can be converted to a per-chip index by subtracting the
> > chip's base).
> >
> > The total maximum number of PWM devices is currently fixed to 64, but
> > can easily be made configurable via Kconfig.
>
> The code says 1024.
Right, I missed that. Thanks.
> > The patch is incomplete in that it doesn't convert any existing drivers
> > that are now broken.
I just noticed that this isn't completely valid anymore either since I've
ported two of the drivers. However, The remaining drivers won't work properly
when built against a kernel with PWM framework support because of the
duplicate symbols.
Thierry
[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-07 7:04 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-07 7:04 UTC (permalink / raw)
To: linux-arm-kernel
* Lars-Peter Clausen wrote:
> On 02/06/2012 04:19 PM, Thierry Reding wrote:
> > This commit modifies the PWM core to support multiple PWMs per struct
> > pwm_chip.
>
> I think you should mention what motivates this change.
Okay, I can add that.
> > It achieves this in a similar way to how gpiolib works, by
> > allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> > starting at a given offset (pwm_chip.base >= 0).
>
> If we've learned one thing from gpiolib, I think it is that using a global
> index to identify a resource was a bad idea.
Yes, that concern has been raised several times already. The reason for it
being this way is that the current API requires it. So either we keep it as
is for now and change it over to indexing on a per-chip basis later (which
has the advantage of not breaking anything currently in-tree) or we make that
change now, in which case this patch series will probably double in size. I'm
also not sure if I can reasonably keep the series bisectible that way.
One of the major problems when converting to a non-global namespace is how to
represent the relationship in code. For device tree this should be easy to do
because it has all the infrastructure in place. For non-device-tree devices I
have no idea yet how this could be done. Perhaps by using something like the
clock API and using names for lookup?
> > A chip specifies how
> > many PWMs it controls using the npwm member. Each of the functions in
> > the pwm_ops structure gets an additional argument that specified the PWM
> > number (it can be converted to a per-chip index by subtracting the
> > chip's base).
> >
> > The total maximum number of PWM devices is currently fixed to 64, but
> > can easily be made configurable via Kconfig.
>
> The code says 1024.
Right, I missed that. Thanks.
> > The patch is incomplete in that it doesn't convert any existing drivers
> > that are now broken.
I just noticed that this isn't completely valid anymore either since I've
ported two of the drivers. However, The remaining drivers won't work properly
when built against a kernel with PWM framework support because of the
duplicate symbols.
Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120207/f0c17d08/attachment.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-07 7:04 ` Thierry Reding
@ 2012-02-07 11:38 ` Mark Brown
-1 siblings, 0 replies; 44+ messages in thread
From: Mark Brown @ 2012-02-07 11:38 UTC (permalink / raw)
To: Thierry Reding
Cc: Lars-Peter Clausen, devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
Mitch Bradley, Mike Frysinger, Arnd Bergmann, Sascha Hauer,
Colin Cross, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1: Type: text/plain, Size: 561 bytes --]
On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> * Lars-Peter Clausen wrote:
> > If we've learned one thing from gpiolib, I think it is that using a global
> > index to identify a resource was a bad idea.
> Yes, that concern has been raised several times already. The reason for it
> being this way is that the current API requires it. So either we keep it as
Might be worth calling this out in the changelog to preempt further
discussion - it's a totally sensible approach to take but it's not
going to be immediately obvious to reviewers.
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 836 bytes --]
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-07 11:38 ` Mark Brown
0 siblings, 0 replies; 44+ messages in thread
From: Mark Brown @ 2012-02-07 11:38 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> * Lars-Peter Clausen wrote:
> > If we've learned one thing from gpiolib, I think it is that using a global
> > index to identify a resource was a bad idea.
> Yes, that concern has been raised several times already. The reason for it
> being this way is that the current API requires it. So either we keep it as
Might be worth calling this out in the changelog to preempt further
discussion - it's a totally sensible approach to take but it's not
going to be immediately obvious to reviewers.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120207/75c22e28/attachment.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-06 15:19 ` Thierry Reding
@ 2012-02-07 22:53 ` Ryan Mallon
-1 siblings, 0 replies; 44+ messages in thread
From: Ryan Mallon @ 2012-02-07 22:53 UTC (permalink / raw)
To: Thierry Reding
Cc: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mitch Bradley,
Mark Brown, Mike Frysinger, Arnd Bergmann, Sascha Hauer,
Colin Cross, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
On 07/02/12 02:19, Thierry Reding wrote:
Hi Thierry,
A few comments below,
~Ryan
> This commit modifies the PWM core to support multiple PWMs per struct
> pwm_chip. It achieves this in a similar way to how gpiolib works, by
> allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> starting at a given offset (pwm_chip.base >= 0). A chip specifies how
> many PWMs it controls using the npwm member. Each of the functions in
> the pwm_ops structure gets an additional argument that specified the PWM
> number (it can be converted to a per-chip index by subtracting the
> chip's base).
>
> The total maximum number of PWM devices is currently fixed to 64, but
> can easily be made configurable via Kconfig.
It would be better to make the code handle arbitrary numbers of PWMs. A
Kconfig knob becomes annoying when you have more than one platform
configured into the kernel.
> The patch is incomplete in that it doesn't convert any existing drivers
> that are now broken.
Does this patch actually break the drivers in terms of not building or
running? If so, can this patch series be reworked a bit to allow the old
PWM framework to be used until all of the drivers are converted?
>
> Signed-off-by: Thierry Reding <thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
> ---
> Changes in v2:
> - add 'struct device *dev' field to pwm_chip, drop label field
> - use radix tree for PWM descriptors
> - add pwm_set_chip_data() and pwm_get_chip_data() functions to allow
> storing and retrieving chip-specific per-PWM data
>
> TODO:
> - pass chip-indexed PWM number (PWM descriptor?) to drivers
> - merge with Sascha's patch
>
> drivers/pwm/core.c | 287 ++++++++++++++++++++++++++++++++++++++-------------
> include/linux/pwm.h | 37 +++++--
> 2 files changed, 242 insertions(+), 82 deletions(-)
>
> diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
> index 71de479..a223bd6 100644
> --- a/drivers/pwm/core.c
> +++ b/drivers/pwm/core.c
> @@ -17,154 +17,292 @@
> * along with this program; see the file COPYING. If not, write to
> * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
> */
> +
> #include <linux/module.h>
> +#include <linux/of_pwm.h>
> #include <linux/pwm.h>
> +#include <linux/radix-tree.h>
> #include <linux/list.h>
> #include <linux/mutex.h>
> #include <linux/err.h>
> #include <linux/slab.h>
> #include <linux/device.h>
>
> +#define MAX_PWMS 1024
> +
> struct pwm_device {
> - struct pwm_chip *chip;
> - const char *label;
> + const char *label;
> + unsigned int pwm;
> +};
> +
> +struct pwm_desc {
> + struct pwm_chip *chip;
> + void *chip_data;
> unsigned long flags;
> #define FLAG_REQUESTED 0
> #define FLAG_ENABLED 1
> - struct list_head node;
> + struct pwm_device pwm;
> };
Do pwm_desc and pwm_device really need to be separate structs?
pwm_device only has two fields, which could easily be added to pwm_desc
(and rename it to pmw_device if you like). They are both opaque
structures, so it makes no difference to the users of the framework.
> -static LIST_HEAD(pwm_list);
> -
> static DEFINE_MUTEX(pwm_lock);
> +static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
> +static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
I missed any discussion of this from v1, but why was this approach
chosen? The bitmap arbitrarily limits the maximum number of PWMs and is
also a little bit (nitpicky) wasteful when a platform doesn't need
anywhere near 1024 PWMs.
In most cases I would expect that platforms only have a handful of PWMs
(say less than 32), in which case a list lookup isn't too bad.
As someone else mentioned, it might be best to drop the global numbering
scheme and have each pwm_chip have its own numbering for its PWMs. So
requesting a PWM would first require getting a handle to the PWM chip it
is on. If a chip has a fixed number of PWMs (e.g. set a registration
time) then the PWMs within a chip can just be an array with O(1) lookup.
> +static struct pwm_desc *pwm_to_desc(unsigned int pwm)
> +{
> + return radix_tree_lookup(&pwm_desc_tree, pwm);
> +}
> +
> +static struct pwm_desc *alloc_desc(unsigned int pwm)
> +{
> + struct pwm_desc *desc;
> +
> + desc = kzalloc(sizeof(*desc), GFP_KERNEL);
> + if (!desc)
> + return NULL;
>
> -static struct pwm_device *_find_pwm(int pwm_id)
> + return desc;
> +}
> +
> +static void free_desc(unsigned int pwm)
> +{
> + struct pwm_desc *desc = pwm_to_desc(pwm);
> +
> + radix_tree_delete(&pwm_desc_tree, pwm);
> + kfree(desc);
> +}
> +
> +static int alloc_descs(int pwm, unsigned int from, unsigned int count)
> {
You don't really need the from argument. There is only one caller for
alloc_descs and it passes 0 for from. So you could re-write this as:
static int alloc_descs(int pwm, unsigned int count)
{
unsigned int from = 0;
...
if (pwm > MAX_PWMS)
return -EINVAL;
if (pwm >= 0)
from = pwm;
> - struct pwm_device *pwm;
> + unsigned int start;
> + unsigned int i;
> + int ret = 0;
> +
> + if (pwm >= 0) {
> + if (from > pwm)
> + return -EINVAL;
> +
> + from = pwm;
> + }
This doesn't catch some caller errors:
if (pwm > MAX_PWMS || from > MAX_PWMS)
return -EINVAL;
> +
> + start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
> + count, 0);
Do the PWM indexes need to be contiguous for a given PWM chip? You could
probably grab each bit individually. That said, I think a better
solution is to have per-chip numbering.
> - list_for_each_entry(pwm, &pwm_list, node) {
> - if (pwm->chip->pwm_id == pwm_id)
> - return pwm;
> + if ((pwm >= 0) && (start != pwm))
Don't need the inner parens.
> + return -EEXIST;
> +
> + if (start + count > MAX_PWMS)
> + return -ENOSPC;
Is this possible? Can bitmap_find_next_zero_area return a start position
where the count would exceed the maximum?
> +
> + for (i = start; i < start + count; i++) {
> + struct pwm_desc *desc = alloc_desc(i);
> + if (!desc) {
> + ret = -ENOMEM;
> + goto err;
> + }
> +
> + radix_tree_insert(&pwm_desc_tree, i, desc);
> }
>
> - return NULL;
> + bitmap_set(allocated_pwms, start, count);
> +
> + return start;
> +
> +err:
> + for (i--; i >= 0; i--)
Nitpick:
while (--i >= 0)
is a bit simpler.
> + free_desc(i);
> +
> + return ret;
> +}
> +
> +static void free_descs(unsigned int from, unsigned int count)
> +{
> + unsigned int i;
> +
> + for (i = from; i < from + count; i++)
> + free_desc(i);
> +
> + bitmap_clear(allocated_pwms, from, count);
> }
>
> /**
> - * pwmchip_add() - register a new pwm
> - * @chip: the pwm
> - *
> - * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
> - * a dynamically assigned id will be used, otherwise the id specified,
> + * pwm_set_chip_data - set private chip data for a PWM
> + * @pwm: PWM number
> + * @data: pointer to chip-specific data
> */
> -int pwmchip_add(struct pwm_chip *chip)
> +int pwm_set_chip_data(unsigned int pwm, void *data)
This interface is a bit ugly. Shouldn't the user need to have a handle
to the pwm_device in order to access the chip_data? Why should callers
be able to set/grab chip data from random PWMs?
> {
> - struct pwm_device *pwm;
> - int ret = 0;
> + struct pwm_desc *desc = pwm_to_desc(pwm);
>
> - pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
> - if (!pwm)
> - return -ENOMEM;
> + if (!desc)
> + return -EINVAL;
>
> - pwm->chip = chip;
> + desc->chip_data = data;
> + return 0;
> +}
> +
> +/**
> + * pwm_get_chip_data - get private chip data for a PWM
> + * @pwm: PWM number
> + */
> +void *pwm_get_chip_data(unsigned int pwm)
> +{
> + struct pwm_desc *desc = pwm_to_desc(pwm);
> +
> + return desc ? desc->chip_data : NULL;
> +}
> +
> +/**
> + * pwmchip_add() - register a new PWM chip
> + * @chip: the PWM chip to add
> + *
> + * Register a new PWM chip. If pwm->base < 0 then a dynamically assigned base
> + * will be used.
> + */
> +int pwmchip_add(struct pwm_chip *chip)
> +{
> + unsigned int i;
> + int ret;
>
> mutex_lock(&pwm_lock);
>
> - if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
> - ret = -EBUSY;
> + ret = alloc_descs(chip->base, 0, chip->npwm);
> + if (ret < 0)
> goto out;
> +
> + chip->base = ret;
> + ret = 0;
> +
> + /* initialize descriptors */
> + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> + struct pwm_desc *desc = pwm_to_desc(i);
> +
> + if (!desc) {
> + pr_debug("pwm: descriptor %u not initialized\n", i);
Should this be a WARN_ON?
> + continue;
> + }
> +
> + desc->chip = chip;
> }
>
> - list_add_tail(&pwm->node, &pwm_list);
> + of_pwmchip_add(chip);
> +
> out:
> mutex_unlock(&pwm_lock);
> -
> - if (ret)
> - kfree(pwm);
> -
> return ret;
> }
> EXPORT_SYMBOL_GPL(pwmchip_add);
>
> /**
> - * pwmchip_remove() - remove a pwm
> - * @chip: the pwm
> + * pwmchip_remove() - remove a PWM chip
> + * @chip: the PWM chip to remove
> *
> - * remove a pwm. This function may return busy if the pwm is still requested.
> + * Removes a PWM chip. This function may return busy if the PWM chip provides
> + * a PWM device that is still requested.
> */
> int pwmchip_remove(struct pwm_chip *chip)
> {
> - struct pwm_device *pwm;
> + unsigned int i;
> int ret = 0;
>
> mutex_lock(&pwm_lock);
>
> - pwm = _find_pwm(chip->pwm_id);
> - if (!pwm) {
> - ret = -ENOENT;
> - goto out;
> - }
> + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> + struct pwm_desc *desc = pwm_to_desc(i);
>
> - if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
> - ret = -EBUSY;
> - goto out;
> + if (test_bit(FLAG_REQUESTED, &desc->flags)) {
> + ret = -EBUSY;
> + goto out;
> + }
> }
>
> - list_del(&pwm->node);
> + free_descs(chip->base, chip->npwm);
> + of_pwmchip_remove(chip);
>
> - kfree(pwm);
> out:
> mutex_unlock(&pwm_lock);
> -
> return ret;
> }
> EXPORT_SYMBOL_GPL(pwmchip_remove);
>
> +/**
> + * pwmchip_find() - iterator for locating a specific pwm_chip
> + * @data: data to pass to match function
> + * @match: callback function to check pwm_chip
> + */
> +struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
> + void *data))
> +{
> + struct pwm_chip *chip = NULL;
> + unsigned long i;
> +
> + mutex_lock(&pwm_lock);
> +
> + for_each_set_bit(i, allocated_pwms, MAX_PWMS) {
> + struct pwm_desc *desc = pwm_to_desc(i);
> +
> + if (!desc || !desc->chip)
> + continue;
> +
> + if (match(desc->chip, data)) {
> + chip = desc->chip;
> + break;
> + }
> + }
> +
> + mutex_unlock(&pwm_lock);
> +
> + return chip;
> +}
> +EXPORT_SYMBOL_GPL(pwmchip_find);
So, propreitary modules are not allowed to use PWMs?
> +
> /*
> * pwm_request - request a PWM device
> */
> -struct pwm_device *pwm_request(int pwm_id, const char *label)
> +struct pwm_device *pwm_request(int pwm, const char *label)
> {
> - struct pwm_device *pwm;
> + struct pwm_device *dev;
> + struct pwm_desc *desc;
> int ret;
>
> + if ((pwm < 0) || (pwm >= MAX_PWMS))
Don't need the inner parens.
> + return ERR_PTR(-ENOENT);
-EINVAL maybe?
> +
> mutex_lock(&pwm_lock);
>
> - pwm = _find_pwm(pwm_id);
> - if (!pwm) {
> - pwm = ERR_PTR(-ENOENT);
> - goto out;
> - }
> + desc = pwm_to_desc(pwm);
> + dev = &desc->pwm;
This looks wrong. If the pwm_desc being requested doesn't exist, won't
this oops?
>
> - if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
> - pwm = ERR_PTR(-EBUSY);
> + if (test_bit(FLAG_REQUESTED, &desc->flags)) {
> + dev = ERR_PTR(-EBUSY);
> goto out;
> }
>
> - if (!try_module_get(pwm->chip->ops->owner)) {
> - pwm = ERR_PTR(-ENODEV);
> + if (!try_module_get(desc->chip->ops->owner)) {
> + dev = ERR_PTR(-ENODEV);
> goto out;
> }
>
> - if (pwm->chip->ops->request) {
> - ret = pwm->chip->ops->request(pwm->chip);
> + if (desc->chip->ops->request) {
> + ret = desc->chip->ops->request(desc->chip, pwm);
> if (ret) {
> - pwm = ERR_PTR(ret);
> + dev = ERR_PTR(ret);
> goto out_put;
> }
> }
>
> - pwm->label = label;
> - set_bit(FLAG_REQUESTED, &pwm->flags);
> + dev->label = label;
> + dev->pwm = pwm;
> + set_bit(FLAG_REQUESTED, &desc->flags);
>
> goto out;
>
> out_put:
> - module_put(pwm->chip->ops->owner);
> + module_put(desc->chip->ops->owner);
> out:
> mutex_unlock(&pwm_lock);
>
> - return pwm;
> + return dev;
> }
> EXPORT_SYMBOL_GPL(pwm_request);
>
> @@ -173,16 +311,19 @@ EXPORT_SYMBOL_GPL(pwm_request);
> */
> void pwm_free(struct pwm_device *pwm)
> {
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> mutex_lock(&pwm_lock);
pwm_lock is used to protect global addition and removal from the
radix/tree bitmap. Here you are also using it to protect the fields of a
specific pwm device? Maybe it would be better to have a per pwm device
lock for this?
> - if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
> + if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
> pr_warning("PWM device already freed\n");
> goto out;
> }
>
> pwm->label = NULL;
> + pwm->pwm = 0;
Why do this? Isn't index 0 a valid PWM index?a
>
> - module_put(pwm->chip->ops->owner);
> + module_put(desc->chip->ops->owner);
> out:
> mutex_unlock(&pwm_lock);
> }
> @@ -193,7 +334,9 @@ EXPORT_SYMBOL_GPL(pwm_free);
> */
> int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
> {
> - return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> + return desc->chip->ops->config(desc->chip, pwm->pwm, duty_ns,
> + period_ns);
> }
> EXPORT_SYMBOL_GPL(pwm_config);
>
> @@ -202,8 +345,10 @@ EXPORT_SYMBOL_GPL(pwm_config);
> */
> int pwm_enable(struct pwm_device *pwm)
> {
> - if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
> - return pwm->chip->ops->enable(pwm->chip);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> + if (!test_and_set_bit(FLAG_ENABLED, &desc->flags))
> + return desc->chip->ops->enable(desc->chip, pwm->pwm);
>
> return 0;
> }
> @@ -214,7 +359,9 @@ EXPORT_SYMBOL_GPL(pwm_enable);
> */
> void pwm_disable(struct pwm_device *pwm)
> {
> - if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
> - pwm->chip->ops->disable(pwm->chip);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> + if (test_and_clear_bit(FLAG_ENABLED, &desc->flags))
> + desc->chip->ops->disable(desc->chip, pwm->pwm);
> }
> EXPORT_SYMBOL_GPL(pwm_disable);
> diff --git a/include/linux/pwm.h b/include/linux/pwm.h
> index df9681b..0f8d105 100644
> --- a/include/linux/pwm.h
> +++ b/include/linux/pwm.h
> @@ -32,37 +32,50 @@ void pwm_disable(struct pwm_device *pwm);
> struct pwm_chip;
>
> /**
> - * struct pwm_ops - PWM operations
> + * struct pwm_ops - PWM controller operations
> * @request: optional hook for requesting a PWM
> * @free: optional hook for freeing a PWM
> * @config: configure duty cycles and period length for this PWM
> * @enable: enable PWM output toggling
> * @disable: disable PWM output toggling
> + * @owner: helps prevent removal of modules exporting active PWMs
> */
> struct pwm_ops {
> - int (*request)(struct pwm_chip *chip);
> - void (*free)(struct pwm_chip *chip);
> - int (*config)(struct pwm_chip *chip, int duty_ns,
> + int (*request)(struct pwm_chip *chip,
> + unsigned int pwm);
> + void (*free)(struct pwm_chip *chip,
> + unsigned int pwm);
> + int (*config)(struct pwm_chip *chip,
> + unsigned int pwm, int duty_ns,
> int period_ns);
> - int (*enable)(struct pwm_chip *chip);
> - void (*disable)(struct pwm_chip *chip);
> + int (*enable)(struct pwm_chip *chip,
> + unsigned int pwm);
> + void (*disable)(struct pwm_chip *chip,
> + unsigned int pwm);
> struct module *owner;
> };
>
> /**
> - * struct pwm_chip - abstract a PWM
> - * @label: for diagnostics
> - * @owner: helps prevent removal of modules exporting active PWMs
> - * @ops: The callbacks for this PWM
> + * struct pwm_chip - abstract a PWM controller
> + * @dev: device providing the PWMs
> + * @ops: callbacks for this PWM controller
> + * @base: number of first PWM controlled by this chip
> + * @npwm: number of PWMs controlled by this chip
> */
> struct pwm_chip {
> - int pwm_id;
> - const char *label;
> + struct device *dev;
> struct pwm_ops *ops;
> + int base;
> + unsigned int npwm;
> };
>
> +int pwm_set_chip_data(unsigned int pwm, void *data);
> +void *pwm_get_chip_data(unsigned int pwm);
> +
> int pwmchip_add(struct pwm_chip *chip);
> int pwmchip_remove(struct pwm_chip *chip);
> +struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
> + void *data));
> #endif
>
> #endif /* __LINUX_PWM_H */
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-07 22:53 ` Ryan Mallon
0 siblings, 0 replies; 44+ messages in thread
From: Ryan Mallon @ 2012-02-07 22:53 UTC (permalink / raw)
To: linux-arm-kernel
On 07/02/12 02:19, Thierry Reding wrote:
Hi Thierry,
A few comments below,
~Ryan
> This commit modifies the PWM core to support multiple PWMs per struct
> pwm_chip. It achieves this in a similar way to how gpiolib works, by
> allowing PWM ranges to be requested dynamically (pwm_chip.base == -1) or
> starting at a given offset (pwm_chip.base >= 0). A chip specifies how
> many PWMs it controls using the npwm member. Each of the functions in
> the pwm_ops structure gets an additional argument that specified the PWM
> number (it can be converted to a per-chip index by subtracting the
> chip's base).
>
> The total maximum number of PWM devices is currently fixed to 64, but
> can easily be made configurable via Kconfig.
It would be better to make the code handle arbitrary numbers of PWMs. A
Kconfig knob becomes annoying when you have more than one platform
configured into the kernel.
> The patch is incomplete in that it doesn't convert any existing drivers
> that are now broken.
Does this patch actually break the drivers in terms of not building or
running? If so, can this patch series be reworked a bit to allow the old
PWM framework to be used until all of the drivers are converted?
>
> Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
> ---
> Changes in v2:
> - add 'struct device *dev' field to pwm_chip, drop label field
> - use radix tree for PWM descriptors
> - add pwm_set_chip_data() and pwm_get_chip_data() functions to allow
> storing and retrieving chip-specific per-PWM data
>
> TODO:
> - pass chip-indexed PWM number (PWM descriptor?) to drivers
> - merge with Sascha's patch
>
> drivers/pwm/core.c | 287 ++++++++++++++++++++++++++++++++++++++-------------
> include/linux/pwm.h | 37 +++++--
> 2 files changed, 242 insertions(+), 82 deletions(-)
>
> diff --git a/drivers/pwm/core.c b/drivers/pwm/core.c
> index 71de479..a223bd6 100644
> --- a/drivers/pwm/core.c
> +++ b/drivers/pwm/core.c
> @@ -17,154 +17,292 @@
> * along with this program; see the file COPYING. If not, write to
> * the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
> */
> +
> #include <linux/module.h>
> +#include <linux/of_pwm.h>
> #include <linux/pwm.h>
> +#include <linux/radix-tree.h>
> #include <linux/list.h>
> #include <linux/mutex.h>
> #include <linux/err.h>
> #include <linux/slab.h>
> #include <linux/device.h>
>
> +#define MAX_PWMS 1024
> +
> struct pwm_device {
> - struct pwm_chip *chip;
> - const char *label;
> + const char *label;
> + unsigned int pwm;
> +};
> +
> +struct pwm_desc {
> + struct pwm_chip *chip;
> + void *chip_data;
> unsigned long flags;
> #define FLAG_REQUESTED 0
> #define FLAG_ENABLED 1
> - struct list_head node;
> + struct pwm_device pwm;
> };
Do pwm_desc and pwm_device really need to be separate structs?
pwm_device only has two fields, which could easily be added to pwm_desc
(and rename it to pmw_device if you like). They are both opaque
structures, so it makes no difference to the users of the framework.
> -static LIST_HEAD(pwm_list);
> -
> static DEFINE_MUTEX(pwm_lock);
> +static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
> +static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
I missed any discussion of this from v1, but why was this approach
chosen? The bitmap arbitrarily limits the maximum number of PWMs and is
also a little bit (nitpicky) wasteful when a platform doesn't need
anywhere near 1024 PWMs.
In most cases I would expect that platforms only have a handful of PWMs
(say less than 32), in which case a list lookup isn't too bad.
As someone else mentioned, it might be best to drop the global numbering
scheme and have each pwm_chip have its own numbering for its PWMs. So
requesting a PWM would first require getting a handle to the PWM chip it
is on. If a chip has a fixed number of PWMs (e.g. set a registration
time) then the PWMs within a chip can just be an array with O(1) lookup.
> +static struct pwm_desc *pwm_to_desc(unsigned int pwm)
> +{
> + return radix_tree_lookup(&pwm_desc_tree, pwm);
> +}
> +
> +static struct pwm_desc *alloc_desc(unsigned int pwm)
> +{
> + struct pwm_desc *desc;
> +
> + desc = kzalloc(sizeof(*desc), GFP_KERNEL);
> + if (!desc)
> + return NULL;
>
> -static struct pwm_device *_find_pwm(int pwm_id)
> + return desc;
> +}
> +
> +static void free_desc(unsigned int pwm)
> +{
> + struct pwm_desc *desc = pwm_to_desc(pwm);
> +
> + radix_tree_delete(&pwm_desc_tree, pwm);
> + kfree(desc);
> +}
> +
> +static int alloc_descs(int pwm, unsigned int from, unsigned int count)
> {
You don't really need the from argument. There is only one caller for
alloc_descs and it passes 0 for from. So you could re-write this as:
static int alloc_descs(int pwm, unsigned int count)
{
unsigned int from = 0;
...
if (pwm > MAX_PWMS)
return -EINVAL;
if (pwm >= 0)
from = pwm;
> - struct pwm_device *pwm;
> + unsigned int start;
> + unsigned int i;
> + int ret = 0;
> +
> + if (pwm >= 0) {
> + if (from > pwm)
> + return -EINVAL;
> +
> + from = pwm;
> + }
This doesn't catch some caller errors:
if (pwm > MAX_PWMS || from > MAX_PWMS)
return -EINVAL;
> +
> + start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
> + count, 0);
Do the PWM indexes need to be contiguous for a given PWM chip? You could
probably grab each bit individually. That said, I think a better
solution is to have per-chip numbering.
> - list_for_each_entry(pwm, &pwm_list, node) {
> - if (pwm->chip->pwm_id == pwm_id)
> - return pwm;
> + if ((pwm >= 0) && (start != pwm))
Don't need the inner parens.
> + return -EEXIST;
> +
> + if (start + count > MAX_PWMS)
> + return -ENOSPC;
Is this possible? Can bitmap_find_next_zero_area return a start position
where the count would exceed the maximum?
> +
> + for (i = start; i < start + count; i++) {
> + struct pwm_desc *desc = alloc_desc(i);
> + if (!desc) {
> + ret = -ENOMEM;
> + goto err;
> + }
> +
> + radix_tree_insert(&pwm_desc_tree, i, desc);
> }
>
> - return NULL;
> + bitmap_set(allocated_pwms, start, count);
> +
> + return start;
> +
> +err:
> + for (i--; i >= 0; i--)
Nitpick:
while (--i >= 0)
is a bit simpler.
> + free_desc(i);
> +
> + return ret;
> +}
> +
> +static void free_descs(unsigned int from, unsigned int count)
> +{
> + unsigned int i;
> +
> + for (i = from; i < from + count; i++)
> + free_desc(i);
> +
> + bitmap_clear(allocated_pwms, from, count);
> }
>
> /**
> - * pwmchip_add() - register a new pwm
> - * @chip: the pwm
> - *
> - * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
> - * a dynamically assigned id will be used, otherwise the id specified,
> + * pwm_set_chip_data - set private chip data for a PWM
> + * @pwm: PWM number
> + * @data: pointer to chip-specific data
> */
> -int pwmchip_add(struct pwm_chip *chip)
> +int pwm_set_chip_data(unsigned int pwm, void *data)
This interface is a bit ugly. Shouldn't the user need to have a handle
to the pwm_device in order to access the chip_data? Why should callers
be able to set/grab chip data from random PWMs?
> {
> - struct pwm_device *pwm;
> - int ret = 0;
> + struct pwm_desc *desc = pwm_to_desc(pwm);
>
> - pwm = kzalloc(sizeof(*pwm), GFP_KERNEL);
> - if (!pwm)
> - return -ENOMEM;
> + if (!desc)
> + return -EINVAL;
>
> - pwm->chip = chip;
> + desc->chip_data = data;
> + return 0;
> +}
> +
> +/**
> + * pwm_get_chip_data - get private chip data for a PWM
> + * @pwm: PWM number
> + */
> +void *pwm_get_chip_data(unsigned int pwm)
> +{
> + struct pwm_desc *desc = pwm_to_desc(pwm);
> +
> + return desc ? desc->chip_data : NULL;
> +}
> +
> +/**
> + * pwmchip_add() - register a new PWM chip
> + * @chip: the PWM chip to add
> + *
> + * Register a new PWM chip. If pwm->base < 0 then a dynamically assigned base
> + * will be used.
> + */
> +int pwmchip_add(struct pwm_chip *chip)
> +{
> + unsigned int i;
> + int ret;
>
> mutex_lock(&pwm_lock);
>
> - if (chip->pwm_id >= 0 && _find_pwm(chip->pwm_id)) {
> - ret = -EBUSY;
> + ret = alloc_descs(chip->base, 0, chip->npwm);
> + if (ret < 0)
> goto out;
> +
> + chip->base = ret;
> + ret = 0;
> +
> + /* initialize descriptors */
> + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> + struct pwm_desc *desc = pwm_to_desc(i);
> +
> + if (!desc) {
> + pr_debug("pwm: descriptor %u not initialized\n", i);
Should this be a WARN_ON?
> + continue;
> + }
> +
> + desc->chip = chip;
> }
>
> - list_add_tail(&pwm->node, &pwm_list);
> + of_pwmchip_add(chip);
> +
> out:
> mutex_unlock(&pwm_lock);
> -
> - if (ret)
> - kfree(pwm);
> -
> return ret;
> }
> EXPORT_SYMBOL_GPL(pwmchip_add);
>
> /**
> - * pwmchip_remove() - remove a pwm
> - * @chip: the pwm
> + * pwmchip_remove() - remove a PWM chip
> + * @chip: the PWM chip to remove
> *
> - * remove a pwm. This function may return busy if the pwm is still requested.
> + * Removes a PWM chip. This function may return busy if the PWM chip provides
> + * a PWM device that is still requested.
> */
> int pwmchip_remove(struct pwm_chip *chip)
> {
> - struct pwm_device *pwm;
> + unsigned int i;
> int ret = 0;
>
> mutex_lock(&pwm_lock);
>
> - pwm = _find_pwm(chip->pwm_id);
> - if (!pwm) {
> - ret = -ENOENT;
> - goto out;
> - }
> + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> + struct pwm_desc *desc = pwm_to_desc(i);
>
> - if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
> - ret = -EBUSY;
> - goto out;
> + if (test_bit(FLAG_REQUESTED, &desc->flags)) {
> + ret = -EBUSY;
> + goto out;
> + }
> }
>
> - list_del(&pwm->node);
> + free_descs(chip->base, chip->npwm);
> + of_pwmchip_remove(chip);
>
> - kfree(pwm);
> out:
> mutex_unlock(&pwm_lock);
> -
> return ret;
> }
> EXPORT_SYMBOL_GPL(pwmchip_remove);
>
> +/**
> + * pwmchip_find() - iterator for locating a specific pwm_chip
> + * @data: data to pass to match function
> + * @match: callback function to check pwm_chip
> + */
> +struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
> + void *data))
> +{
> + struct pwm_chip *chip = NULL;
> + unsigned long i;
> +
> + mutex_lock(&pwm_lock);
> +
> + for_each_set_bit(i, allocated_pwms, MAX_PWMS) {
> + struct pwm_desc *desc = pwm_to_desc(i);
> +
> + if (!desc || !desc->chip)
> + continue;
> +
> + if (match(desc->chip, data)) {
> + chip = desc->chip;
> + break;
> + }
> + }
> +
> + mutex_unlock(&pwm_lock);
> +
> + return chip;
> +}
> +EXPORT_SYMBOL_GPL(pwmchip_find);
So, propreitary modules are not allowed to use PWMs?
> +
> /*
> * pwm_request - request a PWM device
> */
> -struct pwm_device *pwm_request(int pwm_id, const char *label)
> +struct pwm_device *pwm_request(int pwm, const char *label)
> {
> - struct pwm_device *pwm;
> + struct pwm_device *dev;
> + struct pwm_desc *desc;
> int ret;
>
> + if ((pwm < 0) || (pwm >= MAX_PWMS))
Don't need the inner parens.
> + return ERR_PTR(-ENOENT);
-EINVAL maybe?
> +
> mutex_lock(&pwm_lock);
>
> - pwm = _find_pwm(pwm_id);
> - if (!pwm) {
> - pwm = ERR_PTR(-ENOENT);
> - goto out;
> - }
> + desc = pwm_to_desc(pwm);
> + dev = &desc->pwm;
This looks wrong. If the pwm_desc being requested doesn't exist, won't
this oops?
>
> - if (test_bit(FLAG_REQUESTED, &pwm->flags)) {
> - pwm = ERR_PTR(-EBUSY);
> + if (test_bit(FLAG_REQUESTED, &desc->flags)) {
> + dev = ERR_PTR(-EBUSY);
> goto out;
> }
>
> - if (!try_module_get(pwm->chip->ops->owner)) {
> - pwm = ERR_PTR(-ENODEV);
> + if (!try_module_get(desc->chip->ops->owner)) {
> + dev = ERR_PTR(-ENODEV);
> goto out;
> }
>
> - if (pwm->chip->ops->request) {
> - ret = pwm->chip->ops->request(pwm->chip);
> + if (desc->chip->ops->request) {
> + ret = desc->chip->ops->request(desc->chip, pwm);
> if (ret) {
> - pwm = ERR_PTR(ret);
> + dev = ERR_PTR(ret);
> goto out_put;
> }
> }
>
> - pwm->label = label;
> - set_bit(FLAG_REQUESTED, &pwm->flags);
> + dev->label = label;
> + dev->pwm = pwm;
> + set_bit(FLAG_REQUESTED, &desc->flags);
>
> goto out;
>
> out_put:
> - module_put(pwm->chip->ops->owner);
> + module_put(desc->chip->ops->owner);
> out:
> mutex_unlock(&pwm_lock);
>
> - return pwm;
> + return dev;
> }
> EXPORT_SYMBOL_GPL(pwm_request);
>
> @@ -173,16 +311,19 @@ EXPORT_SYMBOL_GPL(pwm_request);
> */
> void pwm_free(struct pwm_device *pwm)
> {
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> mutex_lock(&pwm_lock);
pwm_lock is used to protect global addition and removal from the
radix/tree bitmap. Here you are also using it to protect the fields of a
specific pwm device? Maybe it would be better to have a per pwm device
lock for this?
> - if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
> + if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
> pr_warning("PWM device already freed\n");
> goto out;
> }
>
> pwm->label = NULL;
> + pwm->pwm = 0;
Why do this? Isn't index 0 a valid PWM index?a
>
> - module_put(pwm->chip->ops->owner);
> + module_put(desc->chip->ops->owner);
> out:
> mutex_unlock(&pwm_lock);
> }
> @@ -193,7 +334,9 @@ EXPORT_SYMBOL_GPL(pwm_free);
> */
> int pwm_config(struct pwm_device *pwm, int duty_ns, int period_ns)
> {
> - return pwm->chip->ops->config(pwm->chip, duty_ns, period_ns);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> + return desc->chip->ops->config(desc->chip, pwm->pwm, duty_ns,
> + period_ns);
> }
> EXPORT_SYMBOL_GPL(pwm_config);
>
> @@ -202,8 +345,10 @@ EXPORT_SYMBOL_GPL(pwm_config);
> */
> int pwm_enable(struct pwm_device *pwm)
> {
> - if (!test_and_set_bit(FLAG_ENABLED, &pwm->flags))
> - return pwm->chip->ops->enable(pwm->chip);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> + if (!test_and_set_bit(FLAG_ENABLED, &desc->flags))
> + return desc->chip->ops->enable(desc->chip, pwm->pwm);
>
> return 0;
> }
> @@ -214,7 +359,9 @@ EXPORT_SYMBOL_GPL(pwm_enable);
> */
> void pwm_disable(struct pwm_device *pwm)
> {
> - if (test_and_clear_bit(FLAG_ENABLED, &pwm->flags))
> - pwm->chip->ops->disable(pwm->chip);
> + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> +
> + if (test_and_clear_bit(FLAG_ENABLED, &desc->flags))
> + desc->chip->ops->disable(desc->chip, pwm->pwm);
> }
> EXPORT_SYMBOL_GPL(pwm_disable);
> diff --git a/include/linux/pwm.h b/include/linux/pwm.h
> index df9681b..0f8d105 100644
> --- a/include/linux/pwm.h
> +++ b/include/linux/pwm.h
> @@ -32,37 +32,50 @@ void pwm_disable(struct pwm_device *pwm);
> struct pwm_chip;
>
> /**
> - * struct pwm_ops - PWM operations
> + * struct pwm_ops - PWM controller operations
> * @request: optional hook for requesting a PWM
> * @free: optional hook for freeing a PWM
> * @config: configure duty cycles and period length for this PWM
> * @enable: enable PWM output toggling
> * @disable: disable PWM output toggling
> + * @owner: helps prevent removal of modules exporting active PWMs
> */
> struct pwm_ops {
> - int (*request)(struct pwm_chip *chip);
> - void (*free)(struct pwm_chip *chip);
> - int (*config)(struct pwm_chip *chip, int duty_ns,
> + int (*request)(struct pwm_chip *chip,
> + unsigned int pwm);
> + void (*free)(struct pwm_chip *chip,
> + unsigned int pwm);
> + int (*config)(struct pwm_chip *chip,
> + unsigned int pwm, int duty_ns,
> int period_ns);
> - int (*enable)(struct pwm_chip *chip);
> - void (*disable)(struct pwm_chip *chip);
> + int (*enable)(struct pwm_chip *chip,
> + unsigned int pwm);
> + void (*disable)(struct pwm_chip *chip,
> + unsigned int pwm);
> struct module *owner;
> };
>
> /**
> - * struct pwm_chip - abstract a PWM
> - * @label: for diagnostics
> - * @owner: helps prevent removal of modules exporting active PWMs
> - * @ops: The callbacks for this PWM
> + * struct pwm_chip - abstract a PWM controller
> + * @dev: device providing the PWMs
> + * @ops: callbacks for this PWM controller
> + * @base: number of first PWM controlled by this chip
> + * @npwm: number of PWMs controlled by this chip
> */
> struct pwm_chip {
> - int pwm_id;
> - const char *label;
> + struct device *dev;
> struct pwm_ops *ops;
> + int base;
> + unsigned int npwm;
> };
>
> +int pwm_set_chip_data(unsigned int pwm, void *data);
> +void *pwm_get_chip_data(unsigned int pwm);
> +
> int pwmchip_add(struct pwm_chip *chip);
> int pwmchip_remove(struct pwm_chip *chip);
> +struct pwm_chip *pwmchip_find(void *data, int (*match)(struct pwm_chip *chip,
> + void *data));
> #endif
>
> #endif /* __LINUX_PWM_H */
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-07 22:53 ` Ryan Mallon
@ 2012-02-08 8:15 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 8:15 UTC (permalink / raw)
To: Ryan Mallon
Cc: devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mitch Bradley,
Mark Brown, Mike Frysinger, Arnd Bergmann, Sascha Hauer,
Colin Cross, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Matthias Kaehlcke,
linux-tegra-u79uwXL29TY76Z2rM5mHXA, Eric Miao,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1: Type: text/plain, Size: 10502 bytes --]
* Ryan Mallon wrote:
> On 07/02/12 02:19, Thierry Reding wrote:
> > The total maximum number of PWM devices is currently fixed to 64, but
> > can easily be made configurable via Kconfig.
>
> It would be better to make the code handle arbitrary numbers of PWMs. A
> Kconfig knob becomes annoying when you have more than one platform
> configured into the kernel.
AFAICT handling an arbitrary number of PWMs will only be possible once we get
rid of the global namespace and therefore should be postponed for later. I
may be wrong, though, so if anybody can point me in the right direction I'm
perfectly happy to change that in this series.
> > The patch is incomplete in that it doesn't convert any existing drivers
> > that are now broken.
>
> Does this patch actually break the drivers in terms of not building or
> running? If so, can this patch series be reworked a bit to allow the old
> PWM framework to be used until all of the drivers are converted?
That sentence is misleading. Since the new framework implements the exact
same API as the drivers, any unconverted drivers will still work unless they
are built within the same kernel as the framework.
The intent is to get the framework into an acceptable, backwards-compatible
shape and port existing drivers (i.e. PWM API providers) to the new framework
to ensure everything keeps working as usual.
After that, any issues with the current implementation can be addressed
(per-chip numbering, ...).
> > struct pwm_device {
> > - struct pwm_chip *chip;
> > - const char *label;
> > + const char *label;
> > + unsigned int pwm;
> > +};
> > +
> > +struct pwm_desc {
> > + struct pwm_chip *chip;
> > + void *chip_data;
> > unsigned long flags;
> > #define FLAG_REQUESTED 0
> > #define FLAG_ENABLED 1
> > - struct list_head node;
> > + struct pwm_device pwm;
> > };
>
>
> Do pwm_desc and pwm_device really need to be separate structs?
> pwm_device only has two fields, which could easily be added to pwm_desc
> (and rename it to pmw_device if you like). They are both opaque
> structures, so it makes no difference to the users of the framework.
This was meant to be a middle-ground between the PWM API and a central
framework. The current PWM API returns a pointer to struct pwm_device when
requesting a PWM but the plan was to keep this in line with gpiolib which
does a lot of the same managing.
Perhaps I should elaborate some more. One thing I find particularily nice
about gpiolib is how each GPIO can be addressed with just the GPIO number
which is initially requested and can only be used by the requester. So the
plan was to keep a set of pwm_desc structures which could be looked up via
the PWM number instead of using a pointer to struct pwm_device.
In retrospect I must say that this may not have been the best idea. If indeed
we want to end up with a per-chip numbering of PWM devices, a global number
won't do any good. Keeping the pwm_device structure and wiring it up with the
PWM chip would be a better fit.
Does that sound reasonable?
> > -static LIST_HEAD(pwm_list);
> > -
> > static DEFINE_MUTEX(pwm_lock);
> > +static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
> > +static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
>
> I missed any discussion of this from v1, but why was this approach
> chosen? The bitmap arbitrarily limits the maximum number of PWMs and is
> also a little bit (nitpicky) wasteful when a platform doesn't need
> anywhere near 1024 PWMs.
The initial implementation used a static array of MAX_PWMS pwm_desc
structures (with MAX_PWMS being 64), but in order not to waste memory
unnecessarily this was changed to a radix tree with the bitmap keeping track
of free ranges.
> In most cases I would expect that platforms only have a handful of PWMs
> (say less than 32), in which case a list lookup isn't too bad.
The difficulty with such a scheme is allocating a contiguous range of PWM IDs
for each chip.
> As someone else mentioned, it might be best to drop the global numbering
> scheme and have each pwm_chip have its own numbering for its PWMs. So
> requesting a PWM would first require getting a handle to the PWM chip it
> is on. If a chip has a fixed number of PWMs (e.g. set a registration
> time) then the PWMs within a chip can just be an array with O(1) lookup.
Yes, that would be the ultimate goal.
> > +static int alloc_descs(int pwm, unsigned int from, unsigned int count)
>
> > {
>
> You don't really need the from argument. There is only one caller for
> alloc_descs and it passes 0 for from. So you could re-write this as:
>
> static int alloc_descs(int pwm, unsigned int count)
> {
> unsigned int from = 0;
> ...
>
> if (pwm > MAX_PWMS)
> return -EINVAL;
> if (pwm >= 0)
> from = pwm;
You're right, I'll change that for the next series. Unless we decide to get
rid of the whole pwm_desc structure.
> > - struct pwm_device *pwm;
> > + unsigned int start;
> > + unsigned int i;
> > + int ret = 0;
> > +
> > + if (pwm >= 0) {
> > + if (from > pwm)
> > + return -EINVAL;
> > +
> > + from = pwm;
> > + }
>
>
> This doesn't catch some caller errors:
>
> if (pwm > MAX_PWMS || from > MAX_PWMS)
> return -EINVAL;
I'll fix that as well.
> > +
> > + start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
> > + count, 0);
>
> Do the PWM indexes need to be contiguous for a given PWM chip? You could
> probably grab each bit individually.
Now that you mention it, I don't think so. Since there can currently only be
a single provider of the PWM API there really aren't any restrictions at all
since each chip will simply get the first N indices anyway.
> That said, I think a better solution is to have per-chip numbering.
Yes, I've heard that a number of times already. =)
> > - list_for_each_entry(pwm, &pwm_list, node) {
> > - if (pwm->chip->pwm_id == pwm_id)
> > - return pwm;
> > + if ((pwm >= 0) && (start != pwm))
>
>
> Don't need the inner parens.
I usually like to keep it explicit but I can change it, no problem.
> > + return -EEXIST;
> > +
> > + if (start + count > MAX_PWMS)
> > + return -ENOSPC;
>
>
> Is this possible? Can bitmap_find_next_zero_area return a start position
> where the count would exceed the maximum?
It can indeed. Whenever the area doesn't fit it will return a number past the
maximum (see lib/bitmap.c).
> > +
> > + for (i = start; i < start + count; i++) {
> > + struct pwm_desc *desc = alloc_desc(i);
> > + if (!desc) {
> > + ret = -ENOMEM;
> > + goto err;
> > + }
> > +
> > + radix_tree_insert(&pwm_desc_tree, i, desc);
> > }
> >
> > - return NULL;
> > + bitmap_set(allocated_pwms, start, count);
> > +
> > + return start;
> > +
> > +err:
> > + for (i--; i >= 0; i--)
>
>
> Nitpick:
>
> while (--i >= 0)
>
> is a bit simpler.
Okay.
> > + free_desc(i);
> > +
> > + return ret;
> > +}
> > +
> > +static void free_descs(unsigned int from, unsigned int count)
> > +{
> > + unsigned int i;
> > +
> > + for (i = from; i < from + count; i++)
> > + free_desc(i);
> > +
> > + bitmap_clear(allocated_pwms, from, count);
> > }
> >
> > /**
> > - * pwmchip_add() - register a new pwm
> > - * @chip: the pwm
> > - *
> > - * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
> > - * a dynamically assigned id will be used, otherwise the id specified,
> > + * pwm_set_chip_data - set private chip data for a PWM
> > + * @pwm: PWM number
> > + * @data: pointer to chip-specific data
> > */
> > -int pwmchip_add(struct pwm_chip *chip)
> > +int pwm_set_chip_data(unsigned int pwm, void *data)
>
>
> This interface is a bit ugly. Shouldn't the user need to have a handle
> to the pwm_device in order to access the chip_data? Why should callers
> be able to set/grab chip data from random PWMs?
That interface is meant to be used by drivers to set and retrieve driver-
private data for the PWM. Again, if we get rid of the radix tree and struct
pwm_desc this can be done via the pwm_device structure directly.
> > +int pwmchip_add(struct pwm_chip *chip)
[...]
> > + /* initialize descriptors */
> > + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> > + struct pwm_desc *desc = pwm_to_desc(i);
> > +
> > + if (!desc) {
> > + pr_debug("pwm: descriptor %u not initialized\n", i);
>
>
> Should this be a WARN_ON?
Yes. I just saw that WARN_ON can even be used within a condition, so I'll fix
that up as well. Unless, of course, pwm_desc is removed.
> > +EXPORT_SYMBOL_GPL(pwmchip_find);
>
> So, propreitary modules are not allowed to use PWMs?
I just copied that from gpiolib. I don't know, should they?
> > + if ((pwm < 0) || (pwm >= MAX_PWMS))
>
> Don't need the inner parens.
Okay.
> > + return ERR_PTR(-ENOENT);
>
> -EINVAL maybe?
Yes, that's better.
> > +
> > mutex_lock(&pwm_lock);
> >
> > - pwm = _find_pwm(pwm_id);
> > - if (!pwm) {
> > - pwm = ERR_PTR(-ENOENT);
> > - goto out;
> > - }
> > + desc = pwm_to_desc(pwm);
> > + dev = &desc->pwm;
>
> This looks wrong. If the pwm_desc being requested doesn't exist, won't
> this oops?
Yes, a check is missing here.
> > void pwm_free(struct pwm_device *pwm)
> > {
> > + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> > +
> > mutex_lock(&pwm_lock);
>
> pwm_lock is used to protect global addition and removal from the
> radix/tree bitmap. Here you are also using it to protect the fields of a
> specific pwm device? Maybe it would be better to have a per pwm device
> lock for this?
I'm not sure. pwm_free() would be the sole user currently, so maybe it would
be overkill at this point.
> > - if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
> > + if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
> > pr_warning("PWM device already freed\n");
> > goto out;
> > }
> >
> > pwm->label = NULL;
> > + pwm->pwm = 0;
>
>
> Why do this? Isn't index 0 a valid PWM index?
I added that just to zero out all fields in the structure. And 0 is a valid
index, yes. There is currently no such thing as an invalid PWM index. Perhaps
that needs to be addressed as well. Alternatively this could just be handled
by using the pwm_desc's flags field.
Thanks for the thorough review,
Thierry
[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 8:15 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 8:15 UTC (permalink / raw)
To: linux-arm-kernel
* Ryan Mallon wrote:
> On 07/02/12 02:19, Thierry Reding wrote:
> > The total maximum number of PWM devices is currently fixed to 64, but
> > can easily be made configurable via Kconfig.
>
> It would be better to make the code handle arbitrary numbers of PWMs. A
> Kconfig knob becomes annoying when you have more than one platform
> configured into the kernel.
AFAICT handling an arbitrary number of PWMs will only be possible once we get
rid of the global namespace and therefore should be postponed for later. I
may be wrong, though, so if anybody can point me in the right direction I'm
perfectly happy to change that in this series.
> > The patch is incomplete in that it doesn't convert any existing drivers
> > that are now broken.
>
> Does this patch actually break the drivers in terms of not building or
> running? If so, can this patch series be reworked a bit to allow the old
> PWM framework to be used until all of the drivers are converted?
That sentence is misleading. Since the new framework implements the exact
same API as the drivers, any unconverted drivers will still work unless they
are built within the same kernel as the framework.
The intent is to get the framework into an acceptable, backwards-compatible
shape and port existing drivers (i.e. PWM API providers) to the new framework
to ensure everything keeps working as usual.
After that, any issues with the current implementation can be addressed
(per-chip numbering, ...).
> > struct pwm_device {
> > - struct pwm_chip *chip;
> > - const char *label;
> > + const char *label;
> > + unsigned int pwm;
> > +};
> > +
> > +struct pwm_desc {
> > + struct pwm_chip *chip;
> > + void *chip_data;
> > unsigned long flags;
> > #define FLAG_REQUESTED 0
> > #define FLAG_ENABLED 1
> > - struct list_head node;
> > + struct pwm_device pwm;
> > };
>
>
> Do pwm_desc and pwm_device really need to be separate structs?
> pwm_device only has two fields, which could easily be added to pwm_desc
> (and rename it to pmw_device if you like). They are both opaque
> structures, so it makes no difference to the users of the framework.
This was meant to be a middle-ground between the PWM API and a central
framework. The current PWM API returns a pointer to struct pwm_device when
requesting a PWM but the plan was to keep this in line with gpiolib which
does a lot of the same managing.
Perhaps I should elaborate some more. One thing I find particularily nice
about gpiolib is how each GPIO can be addressed with just the GPIO number
which is initially requested and can only be used by the requester. So the
plan was to keep a set of pwm_desc structures which could be looked up via
the PWM number instead of using a pointer to struct pwm_device.
In retrospect I must say that this may not have been the best idea. If indeed
we want to end up with a per-chip numbering of PWM devices, a global number
won't do any good. Keeping the pwm_device structure and wiring it up with the
PWM chip would be a better fit.
Does that sound reasonable?
> > -static LIST_HEAD(pwm_list);
> > -
> > static DEFINE_MUTEX(pwm_lock);
> > +static DECLARE_BITMAP(allocated_pwms, MAX_PWMS);
> > +static RADIX_TREE(pwm_desc_tree, GFP_KERNEL);
>
> I missed any discussion of this from v1, but why was this approach
> chosen? The bitmap arbitrarily limits the maximum number of PWMs and is
> also a little bit (nitpicky) wasteful when a platform doesn't need
> anywhere near 1024 PWMs.
The initial implementation used a static array of MAX_PWMS pwm_desc
structures (with MAX_PWMS being 64), but in order not to waste memory
unnecessarily this was changed to a radix tree with the bitmap keeping track
of free ranges.
> In most cases I would expect that platforms only have a handful of PWMs
> (say less than 32), in which case a list lookup isn't too bad.
The difficulty with such a scheme is allocating a contiguous range of PWM IDs
for each chip.
> As someone else mentioned, it might be best to drop the global numbering
> scheme and have each pwm_chip have its own numbering for its PWMs. So
> requesting a PWM would first require getting a handle to the PWM chip it
> is on. If a chip has a fixed number of PWMs (e.g. set a registration
> time) then the PWMs within a chip can just be an array with O(1) lookup.
Yes, that would be the ultimate goal.
> > +static int alloc_descs(int pwm, unsigned int from, unsigned int count)
>
> > {
>
> You don't really need the from argument. There is only one caller for
> alloc_descs and it passes 0 for from. So you could re-write this as:
>
> static int alloc_descs(int pwm, unsigned int count)
> {
> unsigned int from = 0;
> ...
>
> if (pwm > MAX_PWMS)
> return -EINVAL;
> if (pwm >= 0)
> from = pwm;
You're right, I'll change that for the next series. Unless we decide to get
rid of the whole pwm_desc structure.
> > - struct pwm_device *pwm;
> > + unsigned int start;
> > + unsigned int i;
> > + int ret = 0;
> > +
> > + if (pwm >= 0) {
> > + if (from > pwm)
> > + return -EINVAL;
> > +
> > + from = pwm;
> > + }
>
>
> This doesn't catch some caller errors:
>
> if (pwm > MAX_PWMS || from > MAX_PWMS)
> return -EINVAL;
I'll fix that as well.
> > +
> > + start = bitmap_find_next_zero_area(allocated_pwms, MAX_PWMS, from,
> > + count, 0);
>
> Do the PWM indexes need to be contiguous for a given PWM chip? You could
> probably grab each bit individually.
Now that you mention it, I don't think so. Since there can currently only be
a single provider of the PWM API there really aren't any restrictions at all
since each chip will simply get the first N indices anyway.
> That said, I think a better solution is to have per-chip numbering.
Yes, I've heard that a number of times already. =)
> > - list_for_each_entry(pwm, &pwm_list, node) {
> > - if (pwm->chip->pwm_id == pwm_id)
> > - return pwm;
> > + if ((pwm >= 0) && (start != pwm))
>
>
> Don't need the inner parens.
I usually like to keep it explicit but I can change it, no problem.
> > + return -EEXIST;
> > +
> > + if (start + count > MAX_PWMS)
> > + return -ENOSPC;
>
>
> Is this possible? Can bitmap_find_next_zero_area return a start position
> where the count would exceed the maximum?
It can indeed. Whenever the area doesn't fit it will return a number past the
maximum (see lib/bitmap.c).
> > +
> > + for (i = start; i < start + count; i++) {
> > + struct pwm_desc *desc = alloc_desc(i);
> > + if (!desc) {
> > + ret = -ENOMEM;
> > + goto err;
> > + }
> > +
> > + radix_tree_insert(&pwm_desc_tree, i, desc);
> > }
> >
> > - return NULL;
> > + bitmap_set(allocated_pwms, start, count);
> > +
> > + return start;
> > +
> > +err:
> > + for (i--; i >= 0; i--)
>
>
> Nitpick:
>
> while (--i >= 0)
>
> is a bit simpler.
Okay.
> > + free_desc(i);
> > +
> > + return ret;
> > +}
> > +
> > +static void free_descs(unsigned int from, unsigned int count)
> > +{
> > + unsigned int i;
> > +
> > + for (i = from; i < from + count; i++)
> > + free_desc(i);
> > +
> > + bitmap_clear(allocated_pwms, from, count);
> > }
> >
> > /**
> > - * pwmchip_add() - register a new pwm
> > - * @chip: the pwm
> > - *
> > - * register a new pwm. pwm->pwm_id must be initialized. if pwm_id < 0 then
> > - * a dynamically assigned id will be used, otherwise the id specified,
> > + * pwm_set_chip_data - set private chip data for a PWM
> > + * @pwm: PWM number
> > + * @data: pointer to chip-specific data
> > */
> > -int pwmchip_add(struct pwm_chip *chip)
> > +int pwm_set_chip_data(unsigned int pwm, void *data)
>
>
> This interface is a bit ugly. Shouldn't the user need to have a handle
> to the pwm_device in order to access the chip_data? Why should callers
> be able to set/grab chip data from random PWMs?
That interface is meant to be used by drivers to set and retrieve driver-
private data for the PWM. Again, if we get rid of the radix tree and struct
pwm_desc this can be done via the pwm_device structure directly.
> > +int pwmchip_add(struct pwm_chip *chip)
[...]
> > + /* initialize descriptors */
> > + for (i = chip->base; i < chip->base + chip->npwm; i++) {
> > + struct pwm_desc *desc = pwm_to_desc(i);
> > +
> > + if (!desc) {
> > + pr_debug("pwm: descriptor %u not initialized\n", i);
>
>
> Should this be a WARN_ON?
Yes. I just saw that WARN_ON can even be used within a condition, so I'll fix
that up as well. Unless, of course, pwm_desc is removed.
> > +EXPORT_SYMBOL_GPL(pwmchip_find);
>
> So, propreitary modules are not allowed to use PWMs?
I just copied that from gpiolib. I don't know, should they?
> > + if ((pwm < 0) || (pwm >= MAX_PWMS))
>
> Don't need the inner parens.
Okay.
> > + return ERR_PTR(-ENOENT);
>
> -EINVAL maybe?
Yes, that's better.
> > +
> > mutex_lock(&pwm_lock);
> >
> > - pwm = _find_pwm(pwm_id);
> > - if (!pwm) {
> > - pwm = ERR_PTR(-ENOENT);
> > - goto out;
> > - }
> > + desc = pwm_to_desc(pwm);
> > + dev = &desc->pwm;
>
> This looks wrong. If the pwm_desc being requested doesn't exist, won't
> this oops?
Yes, a check is missing here.
> > void pwm_free(struct pwm_device *pwm)
> > {
> > + struct pwm_desc *desc = container_of(pwm, struct pwm_desc, pwm);
> > +
> > mutex_lock(&pwm_lock);
>
> pwm_lock is used to protect global addition and removal from the
> radix/tree bitmap. Here you are also using it to protect the fields of a
> specific pwm device? Maybe it would be better to have a per pwm device
> lock for this?
I'm not sure. pwm_free() would be the sole user currently, so maybe it would
be overkill at this point.
> > - if (!test_and_clear_bit(FLAG_REQUESTED, &pwm->flags)) {
> > + if (!test_and_clear_bit(FLAG_REQUESTED, &desc->flags)) {
> > pr_warning("PWM device already freed\n");
> > goto out;
> > }
> >
> > pwm->label = NULL;
> > + pwm->pwm = 0;
>
>
> Why do this? Isn't index 0 a valid PWM index?
I added that just to zero out all fields in the structure. And 0 is a valid
index, yes. There is currently no such thing as an invalid PWM index. Perhaps
that needs to be addressed as well. Alternatively this could just be handled
by using the pwm_desc's flags field.
Thanks for the thorough review,
Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120208/e70501a8/attachment-0001.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-08 8:15 ` Thierry Reding
@ 2012-02-08 9:00 ` Sascha Hauer
-1 siblings, 0 replies; 44+ messages in thread
From: Sascha Hauer @ 2012-02-08 9:00 UTC (permalink / raw)
To: Thierry Reding
Cc: linux-tegra-u79uwXL29TY76Z2rM5mHXA, Ryan Mallon,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mark Brown,
Matthias Kaehlcke, Rob Herring, Richard Purdie, Colin Cross,
Eric Miao, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> * Ryan Mallon wrote:
> > On 07/02/12 02:19, Thierry Reding wrote:
> > > The total maximum number of PWM devices is currently fixed to 64, but
> > > can easily be made configurable via Kconfig.
> >
> > It would be better to make the code handle arbitrary numbers of PWMs. A
> > Kconfig knob becomes annoying when you have more than one platform
> > configured into the kernel.
>
> AFAICT handling an arbitrary number of PWMs will only be possible once we get
> rid of the global namespace and therefore should be postponed for later. I
> may be wrong, though, so if anybody can point me in the right direction I'm
> perfectly happy to change that in this series.
>
> > > The patch is incomplete in that it doesn't convert any existing drivers
> > > that are now broken.
> >
> > Does this patch actually break the drivers in terms of not building or
> > running? If so, can this patch series be reworked a bit to allow the old
> > PWM framework to be used until all of the drivers are converted?
>
> That sentence is misleading. Since the new framework implements the exact
> same API as the drivers, any unconverted drivers will still work unless they
> are built within the same kernel as the framework.
I have a series porting all in kernel drivers to this framework. I can
follow up on this series once it's clear we want to go this way. With
this series the way will be free to move to a better pwm matching
without a global pwm index because the drivers won't be affected by
this change anymore.
Sascha
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 9:00 ` Sascha Hauer
0 siblings, 0 replies; 44+ messages in thread
From: Sascha Hauer @ 2012-02-08 9:00 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> * Ryan Mallon wrote:
> > On 07/02/12 02:19, Thierry Reding wrote:
> > > The total maximum number of PWM devices is currently fixed to 64, but
> > > can easily be made configurable via Kconfig.
> >
> > It would be better to make the code handle arbitrary numbers of PWMs. A
> > Kconfig knob becomes annoying when you have more than one platform
> > configured into the kernel.
>
> AFAICT handling an arbitrary number of PWMs will only be possible once we get
> rid of the global namespace and therefore should be postponed for later. I
> may be wrong, though, so if anybody can point me in the right direction I'm
> perfectly happy to change that in this series.
>
> > > The patch is incomplete in that it doesn't convert any existing drivers
> > > that are now broken.
> >
> > Does this patch actually break the drivers in terms of not building or
> > running? If so, can this patch series be reworked a bit to allow the old
> > PWM framework to be used until all of the drivers are converted?
>
> That sentence is misleading. Since the new framework implements the exact
> same API as the drivers, any unconverted drivers will still work unless they
> are built within the same kernel as the framework.
I have a series porting all in kernel drivers to this framework. I can
follow up on this series once it's clear we want to go this way. With
this series the way will be free to move to a better pwm matching
without a global pwm index because the drivers won't be affected by
this change anymore.
Sascha
--
Pengutronix e.K. | |
Industrial Linux Solutions | http://www.pengutronix.de/ |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 |
Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-07 7:04 ` Thierry Reding
@ 2012-02-08 9:13 ` Russell King - ARM Linux
-1 siblings, 0 replies; 44+ messages in thread
From: Russell King - ARM Linux @ 2012-02-08 9:13 UTC (permalink / raw)
To: Thierry Reding
Cc: Lars-Peter Clausen, linux-tegra-u79uwXL29TY76Z2rM5mHXA,
Mike Frysinger, Arnd Bergmann,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mark Brown,
Matthias Kaehlcke, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Colin Cross, Mitch Bradley, Eric Miao,
Sascha Hauer, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> One of the major problems when converting to a non-global namespace is how to
> represent the relationship in code. For device tree this should be easy to do
> because it has all the infrastructure in place. For non-device-tree devices I
> have no idea yet how this could be done. Perhaps by using something like the
> clock API and using names for lookup?
Well, the first step is to make sure that the pwm 'chip' drivers don't
know anything about the global numberspace.
That limits the extent of the global numberspace to the pwm core and
pwm users. You then don't have to worry about the pwm chip drivers
while you work out how to solve the other half of the problem.
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 9:13 ` Russell King - ARM Linux
0 siblings, 0 replies; 44+ messages in thread
From: Russell King - ARM Linux @ 2012-02-08 9:13 UTC (permalink / raw)
To: linux-arm-kernel
On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> One of the major problems when converting to a non-global namespace is how to
> represent the relationship in code. For device tree this should be easy to do
> because it has all the infrastructure in place. For non-device-tree devices I
> have no idea yet how this could be done. Perhaps by using something like the
> clock API and using names for lookup?
Well, the first step is to make sure that the pwm 'chip' drivers don't
know anything about the global numberspace.
That limits the extent of the global numberspace to the pwm core and
pwm users. You then don't have to worry about the pwm chip drivers
while you work out how to solve the other half of the problem.
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-08 8:15 ` Thierry Reding
@ 2012-02-08 9:17 ` Russell King - ARM Linux
-1 siblings, 0 replies; 44+ messages in thread
From: Russell King - ARM Linux @ 2012-02-08 9:17 UTC (permalink / raw)
To: Thierry Reding
Cc: Ryan Mallon, linux-tegra-u79uwXL29TY76Z2rM5mHXA, Mike Frysinger,
Arnd Bergmann, devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
Mark Brown, Matthias Kaehlcke, Rob Herring, Grant Likely,
Olof Johansson, Richard Purdie, Colin Cross, Mitch Bradley,
Eric Miao, Sascha Hauer,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> * Ryan Mallon wrote:
> > On 07/02/12 02:19, Thierry Reding wrote:
> > > The total maximum number of PWM devices is currently fixed to 64, but
> > > can easily be made configurable via Kconfig.
> >
> > It would be better to make the code handle arbitrary numbers of PWMs. A
> > Kconfig knob becomes annoying when you have more than one platform
> > configured into the kernel.
>
> AFAICT handling an arbitrary number of PWMs will only be possible once we get
> rid of the global namespace and therefore should be postponed for later. I
> may be wrong, though, so if anybody can point me in the right direction I'm
> perfectly happy to change that in this series.
One solution would be to look at how genirq handles this. Eg, it uses a
radix tree to store the irqdescs rather than an array, and a bitmap
showing which global numbers are allocated.
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 9:17 ` Russell King - ARM Linux
0 siblings, 0 replies; 44+ messages in thread
From: Russell King - ARM Linux @ 2012-02-08 9:17 UTC (permalink / raw)
To: linux-arm-kernel
On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> * Ryan Mallon wrote:
> > On 07/02/12 02:19, Thierry Reding wrote:
> > > The total maximum number of PWM devices is currently fixed to 64, but
> > > can easily be made configurable via Kconfig.
> >
> > It would be better to make the code handle arbitrary numbers of PWMs. A
> > Kconfig knob becomes annoying when you have more than one platform
> > configured into the kernel.
>
> AFAICT handling an arbitrary number of PWMs will only be possible once we get
> rid of the global namespace and therefore should be postponed for later. I
> may be wrong, though, so if anybody can point me in the right direction I'm
> perfectly happy to change that in this series.
One solution would be to look at how genirq handles this. Eg, it uses a
radix tree to store the irqdescs rather than an array, and a bitmap
showing which global numbers are allocated.
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-08 9:17 ` Russell King - ARM Linux
@ 2012-02-08 10:31 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 10:31 UTC (permalink / raw)
To: Russell King - ARM Linux
Cc: Ryan Mallon, linux-tegra-u79uwXL29TY76Z2rM5mHXA, Mike Frysinger,
Arnd Bergmann, devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ,
Mark Brown, Matthias Kaehlcke, Rob Herring, Grant Likely,
Olof Johansson, Richard Purdie, Colin Cross, Mitch Bradley,
Eric Miao, Sascha Hauer,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1: Type: text/plain, Size: 1465 bytes --]
* Russell King - ARM Linux wrote:
> On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> > * Ryan Mallon wrote:
> > > On 07/02/12 02:19, Thierry Reding wrote:
> > > > The total maximum number of PWM devices is currently fixed to 64, but
> > > > can easily be made configurable via Kconfig.
> > >
> > > It would be better to make the code handle arbitrary numbers of PWMs. A
> > > Kconfig knob becomes annoying when you have more than one platform
> > > configured into the kernel.
> >
> > AFAICT handling an arbitrary number of PWMs will only be possible once we get
> > rid of the global namespace and therefore should be postponed for later. I
> > may be wrong, though, so if anybody can point me in the right direction I'm
> > perfectly happy to change that in this series.
>
> One solution would be to look at how genirq handles this. Eg, it uses a
> radix tree to store the irqdescs rather than an array, and a bitmap
> showing which global numbers are allocated.
That's exactly the solution implemented by this second version of the series.
In fact I did turn to genirq for inspiration at the time I wrote the code.
My understanding was that Ryan proposed to get rid of the bitmap altogether
because it arbitrarily limits the number of PWMs. How that can be achieved
with a global namespace I don't know. Thus my proposal to keep it as-is for
now and get rid of it once per-chip indexing is implemented.
Thierry
[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 10:31 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 10:31 UTC (permalink / raw)
To: linux-arm-kernel
* Russell King - ARM Linux wrote:
> On Wed, Feb 08, 2012 at 09:15:08AM +0100, Thierry Reding wrote:
> > * Ryan Mallon wrote:
> > > On 07/02/12 02:19, Thierry Reding wrote:
> > > > The total maximum number of PWM devices is currently fixed to 64, but
> > > > can easily be made configurable via Kconfig.
> > >
> > > It would be better to make the code handle arbitrary numbers of PWMs. A
> > > Kconfig knob becomes annoying when you have more than one platform
> > > configured into the kernel.
> >
> > AFAICT handling an arbitrary number of PWMs will only be possible once we get
> > rid of the global namespace and therefore should be postponed for later. I
> > may be wrong, though, so if anybody can point me in the right direction I'm
> > perfectly happy to change that in this series.
>
> One solution would be to look at how genirq handles this. Eg, it uses a
> radix tree to store the irqdescs rather than an array, and a bitmap
> showing which global numbers are allocated.
That's exactly the solution implemented by this second version of the series.
In fact I did turn to genirq for inspiration at the time I wrote the code.
My understanding was that Ryan proposed to get rid of the bitmap altogether
because it arbitrarily limits the number of PWMs. How that can be achieved
with a global namespace I don't know. Thus my proposal to keep it as-is for
now and get rid of it once per-chip indexing is implemented.
Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120208/c61c04bf/attachment-0001.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-08 9:13 ` Russell King - ARM Linux
@ 2012-02-08 11:12 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 11:12 UTC (permalink / raw)
To: Russell King - ARM Linux
Cc: Lars-Peter Clausen, linux-tegra-u79uwXL29TY76Z2rM5mHXA,
Mike Frysinger, Arnd Bergmann,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mark Brown,
Matthias Kaehlcke, Rob Herring, Grant Likely, Olof Johansson,
Richard Purdie, Colin Cross, Mitch Bradley, Eric Miao,
Sascha Hauer, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1: Type: text/plain, Size: 1342 bytes --]
* Russell King - ARM Linux wrote:
> On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> > One of the major problems when converting to a non-global namespace is how to
> > represent the relationship in code. For device tree this should be easy to do
> > because it has all the infrastructure in place. For non-device-tree devices I
> > have no idea yet how this could be done. Perhaps by using something like the
> > clock API and using names for lookup?
>
> Well, the first step is to make sure that the pwm 'chip' drivers don't
> know anything about the global numberspace.
>
> That limits the extent of the global numberspace to the pwm core and
> pwm users. You then don't have to worry about the pwm chip drivers
> while you work out how to solve the other half of the problem.
I see. That would mean that pwm_request() gets a global PWM index, looks up
the corresponding PWM chip and device then pass the chip-relative index to
the driver's ops.
I think that can be simplified even further. For instance if the pwm_device
structure contains a pointer to its parent struct pwm_chip, then that can be
used directly to call into the driver's ops. And the ops could be modified to
take a pwm_device instead of the global PWM index. That would be similar to
how struct irq_data is used.
Thierry
[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 11:12 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 11:12 UTC (permalink / raw)
To: linux-arm-kernel
* Russell King - ARM Linux wrote:
> On Tue, Feb 07, 2012 at 08:04:00AM +0100, Thierry Reding wrote:
> > One of the major problems when converting to a non-global namespace is how to
> > represent the relationship in code. For device tree this should be easy to do
> > because it has all the infrastructure in place. For non-device-tree devices I
> > have no idea yet how this could be done. Perhaps by using something like the
> > clock API and using names for lookup?
>
> Well, the first step is to make sure that the pwm 'chip' drivers don't
> know anything about the global numberspace.
>
> That limits the extent of the global numberspace to the pwm core and
> pwm users. You then don't have to worry about the pwm chip drivers
> while you work out how to solve the other half of the problem.
I see. That would mean that pwm_request() gets a global PWM index, looks up
the corresponding PWM chip and device then pass the chip-relative index to
the driver's ops.
I think that can be simplified even further. For instance if the pwm_device
structure contains a pointer to its parent struct pwm_chip, then that can be
used directly to call into the driver's ops. And the ops could be modified to
take a pwm_device instead of the global PWM index. That would be similar to
how struct irq_data is used.
Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120208/57c89e5a/attachment.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
* Re: [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
2012-02-08 9:00 ` Sascha Hauer
@ 2012-02-08 11:16 ` Thierry Reding
-1 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 11:16 UTC (permalink / raw)
To: Sascha Hauer
Cc: linux-tegra-u79uwXL29TY76Z2rM5mHXA, Ryan Mallon,
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ, Mark Brown,
Matthias Kaehlcke, Rob Herring, Richard Purdie, Colin Cross,
Eric Miao, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
Kurt Van Dijck
[-- Attachment #1.1: Type: text/plain, Size: 622 bytes --]
* Sascha Hauer wrote:
> I have a series porting all in kernel drivers to this framework. I can
> follow up on this series once it's clear we want to go this way. With
> this series the way will be free to move to a better pwm matching
> without a global pwm index because the drivers won't be affected by
> this change anymore.
Nice. I already ported the PXA and Blackfin drivers in this series and was
planning to port more in subsequent versions but if you've already done most
of the work I can just skip. I should probably keep the Tegra driver in the
series because it doesn't exist in the kernel tree yet.
Thierry
[-- Attachment #1.2: Type: application/pgp-signature, Size: 198 bytes --]
[-- Attachment #2: Type: text/plain, Size: 192 bytes --]
_______________________________________________
devicetree-discuss mailing list
devicetree-discuss-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
https://lists.ozlabs.org/listinfo/devicetree-discuss
^ permalink raw reply [flat|nested] 44+ messages in thread
* [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs.
@ 2012-02-08 11:16 ` Thierry Reding
0 siblings, 0 replies; 44+ messages in thread
From: Thierry Reding @ 2012-02-08 11:16 UTC (permalink / raw)
To: linux-arm-kernel
* Sascha Hauer wrote:
> I have a series porting all in kernel drivers to this framework. I can
> follow up on this series once it's clear we want to go this way. With
> this series the way will be free to move to a better pwm matching
> without a global pwm index because the drivers won't be affected by
> this change anymore.
Nice. I already ported the PXA and Blackfin drivers in this series and was
planning to port more in subsequent versions but if you've already done most
of the work I can just skip. I should probably keep the Tegra driver in the
series because it doesn't exist in the kernel tree yet.
Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120208/e59e2c18/attachment.sig>
^ permalink raw reply [flat|nested] 44+ messages in thread
end of thread, other threads:[~2012-02-08 11:16 UTC | newest]
Thread overview: 44+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-02-06 15:19 [PATCH v2 00/10] Add PWM framework and device-tree support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
[not found] ` <1328541585-24642-1-git-send-email-thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
2012-02-06 15:19 ` [PATCH v2 01/10] PWM: add pwm framework support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 02/10] pwm: Allow chips to support multiple PWMs Thierry Reding
2012-02-06 15:19 ` Thierry Reding
[not found] ` <1328541585-24642-3-git-send-email-thierry.reding-RM9K5IK7kjKj5M59NBduVrNAH6kLmebB@public.gmane.org>
2012-02-06 21:22 ` Lars-Peter Clausen
2012-02-06 21:22 ` Lars-Peter Clausen
[not found] ` <4F3044A9.8000202-Qo5EllUWu/uELgA04lAiVw@public.gmane.org>
2012-02-07 7:04 ` Thierry Reding
2012-02-07 7:04 ` Thierry Reding
[not found] ` <20120207070400.GA29238-RM9K5IK7kjIQXX3q8xo1gnVAuStQJXxyR5q1nwbD4aMs9pC9oP6+/A@public.gmane.org>
2012-02-07 11:38 ` Mark Brown
2012-02-07 11:38 ` Mark Brown
2012-02-08 9:13 ` Russell King - ARM Linux
2012-02-08 9:13 ` Russell King - ARM Linux
[not found] ` <20120208091327.GH889-l+eeeJia6m9vn6HldHNs0ANdhmdF6hFW@public.gmane.org>
2012-02-08 11:12 ` Thierry Reding
2012-02-08 11:12 ` Thierry Reding
2012-02-07 22:53 ` Ryan Mallon
2012-02-07 22:53 ` Ryan Mallon
[not found] ` <4F31AB63.3020301-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
2012-02-08 8:15 ` Thierry Reding
2012-02-08 8:15 ` Thierry Reding
[not found] ` <20120208081508.GA6673-RM9K5IK7kjIQXX3q8xo1gnVAuStQJXxyR5q1nwbD4aMs9pC9oP6+/A@public.gmane.org>
2012-02-08 9:00 ` Sascha Hauer
2012-02-08 9:00 ` Sascha Hauer
[not found] ` <20120208090055.GP3852-bIcnvbaLZ9MEGnE8C9+IrQ@public.gmane.org>
2012-02-08 11:16 ` Thierry Reding
2012-02-08 11:16 ` Thierry Reding
2012-02-08 9:17 ` Russell King - ARM Linux
2012-02-08 9:17 ` Russell King - ARM Linux
[not found] ` <20120208091720.GI889-l+eeeJia6m9vn6HldHNs0ANdhmdF6hFW@public.gmane.org>
2012-02-08 10:31 ` Thierry Reding
2012-02-08 10:31 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 03/10] of: Add PWM support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 04/10] arm/tegra: Fix PWM clock programming Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 05/10] arm/tegra: Provide clock for only one PWM controller Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 06/10] pwm: Add NVIDIA Tegra SoC support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 07/10] arm/tegra: Add PWFM controller device tree probing Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 08/10] pwm: Add Blackfin support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 09/10] pwm: Add PXA support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
2012-02-06 15:19 ` [PATCH v2 10/10] pwm-backlight: Add rudimentary device tree support Thierry Reding
2012-02-06 15:19 ` Thierry Reding
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.