* [PATCH v10 0/5] ZII RAVE platform driver
@ 2017-10-31 16:36 ` Andrey Smirnov
0 siblings, 0 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel-u79uwXL29TY76Z2rM5mHXA
Cc: Andrey Smirnov, Pavel Machek, Greg Kroah-Hartman,
cphealy-Re5JQEeQqe8AvxtiuMwx3w, Andy Shevchenko, Lee Jones,
Lucas Stach, Nikita Yushchenko, Guenter Roeck, Rob Herring,
Mark Rutland, devicetree-u79uwXL29TY76Z2rM5mHXA,
Sebastian Reichel
Hi everyone,
This patch series is v10 of the driver for supervisory processor found
on RAVE series of devices from ZII. Supervisory processor is a PIC
microcontroller connected to various electrical subsystems on RAVE
devices whose firmware implements protocol to command/qery them.
NOTE:
* This driver dependends on crc_ccitt_false(), added by
2da9378d531f8cc6670c7497f20d936b706ab80b in 'linux-next', the patch
was pulled in by Andrew Morton and is currently avaiting users, so
this series might have to go in through Andrew's tree
Changes since [v9]:
- Converted watchdog driver to use watchdog_active() instead of
watchdog_hw_running() and replaced WARN_ON with a regular error
message as per feedback from Guenter
- Changed rave_sp_wdt_start() to set WDOG_HW_RUNNING only if
communicating with hardware was sucessful
- Collected Reviewd-by from Sebastian (for serdev related patches)
- Collected Acked-by from Rob (for watchdog DT bindings)
Changes since [v8]:
- Driver moved from drivers/platform to drivers/mfd
- Collected Reviewed-by from Guenter (for patches 1, 2 and 3)
- Incorporated feedback from Guenter into watchdog driver
- Incorporated feedback from Rob into watchdog DT bindings
- Removed struct rave_sp_rsp_status, which was a leftover from v5
-> v6 code removal.
- Fixed minor problems reported by checkpatch
Changes since [v7]:
- Added watchdog driver to the patchset, so it would be easier to
understand how parent/children drivers are tied together
- Added serdev patches to implement devm_serdev_device_open() and make .remove optional
- "Added" missing serdev_device_close() by converting the driver
to use devm_serdev_device_open()
- Converted the driver to use devm_of_platform_populate()
- Removed needless dependency on MFD_CORE
- Removed dependency on SERIAL_DEV_CTRL_TTYPORT
Changes since [v6]:
- Patch 2/2 has been applied by Lee so it is no longer a part of the series
- Removed all sysfs and debugfs attribute to reduce the scope of
the driver propsed for inclusion. This is not a critical to have
feature and can be added/discussed later.
Changes since [v5]:
- Fixed a build break, introduced by a last minute change in [v5]
- Moved majority of attributes that were exposed over sysfs to debugfs
- Document remaining sysfs attributes in Documentation/ABI/testing/sysfs-platform-rave-sp
Changes since [v4]:
- Replaced usage of DEVICE_ATTR with DEVICE_ATTR_RW
- Fixed a number of warnings produces by sparse tool
- Incorporated event more feedback from Andy Shevchenko
- Collected Reviewed-by from Andy
Changes since [v3]:
- Re-collected lost Acked-by from Rob
- Incorporated further feedback from Andy Shevchenko
- Dropped useless change (stray newline) to drivers/mfd/Makefile
Changes since [v2]:
- Fixed swapped command codes in rave_sp_common_get_boot_source()
and rave_sp_common_set_boot_source() revealed by further testing
of the code
- Incorporated feedback from Andy Shevchenko
Changes since [v1]:
- Updated wording in DT-bindings as per Rob's request.
- Collected Rob's Acked-by for patch 2/2
Feedback is greatly appreciated!
Thanks,
Andrey Smirnov
[v9] lkml.kernel.org/r/20171025190421.18415-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v8] lkml.kernel.org/r/20171018170136.12347-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v7] lkml.kernel.org/r/20171013061321.31252-2-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v6] lkml.kernel.org/r/20170828163131.24815-2-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v5] lkml.kernel.org/r/20170728142704.11156-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v4] lkml.kernel.org/r/20170725184450.13171-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v3] lkml.kernel.org/r/20170724150915.4824-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v2] lkml.kernel.org/r/20170718175604.11735-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
[v1] lkml.kernel.org/r/20170710170449.4544-1-andrew.smirnov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
Andrey Smirnov (5):
serdev: Make .remove in struct serdev_device_driver optional
serdev: Introduce devm_serdev_device_open()
mfd: Add driver for RAVE Supervisory Processor
watchdog: Add RAVE SP watchdog driver
dt-bindings: watchdog: Add bindings for RAVE SP watchdog driver
.../bindings/watchdog/zii,rave-sp-wdt.txt | 39 ++
Documentation/driver-model/devres.txt | 3 +
drivers/mfd/Kconfig | 9 +
drivers/mfd/Makefile | 2 +
drivers/mfd/rave-sp.c | 659 +++++++++++++++++++++
drivers/platform/Kconfig | 1 +
drivers/tty/serdev/core.c | 31 +-
drivers/watchdog/Kconfig | 7 +
drivers/watchdog/Makefile | 1 +
drivers/watchdog/rave-sp-wdt.c | 349 +++++++++++
include/linux/rave-sp.h | 54 ++
include/linux/serdev.h | 1 +
12 files changed, 1154 insertions(+), 2 deletions(-)
create mode 100644 Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
create mode 100644 drivers/mfd/rave-sp.c
create mode 100644 drivers/watchdog/rave-sp-wdt.c
create mode 100644 include/linux/rave-sp.h
--
2.13.6
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 22+ messages in thread* [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional
2017-10-31 16:36 ` Andrey Smirnov
(?)
@ 2017-10-31 16:36 ` Andrey Smirnov
2017-11-01 23:48 ` Rob Herring
2017-11-04 11:24 ` Greg Kroah-Hartman
-1 siblings, 2 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel
Cc: Andrey Smirnov, linux-serial, Rob Herring, cphealy,
Guenter Roeck, Lucas Stach, Nikita Yushchenko, Lee Jones,
Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko, Johan Hovold,
Sebastian Reichel
Using devres infrastructure it is possible to write a serdev driver
that doesn't have any code that needs to be called as a part of
.remove. Add code to make .remove optional.
Cc: linux-kernel@vger.kernel.org
Cc: linux-serial@vger.kernel.org
Cc: Rob Herring <robh@kernel.org>
Cc: cphealy@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Lee Jones <lee.jones@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Johan Hovold <johan@kernel.org>
Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
drivers/tty/serdev/core.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/tty/serdev/core.c b/drivers/tty/serdev/core.c
index c68fb3a8ea1c..f500f6a2ca88 100644
--- a/drivers/tty/serdev/core.c
+++ b/drivers/tty/serdev/core.c
@@ -252,8 +252,8 @@ static int serdev_drv_probe(struct device *dev)
static int serdev_drv_remove(struct device *dev)
{
const struct serdev_device_driver *sdrv = to_serdev_device_driver(dev->driver);
-
- sdrv->remove(to_serdev_device(dev));
+ if (sdrv->remove)
+ sdrv->remove(to_serdev_device(dev));
return 0;
}
--
2.13.6
^ permalink raw reply related [flat|nested] 22+ messages in thread* Re: [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional
2017-10-31 16:36 ` [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional Andrey Smirnov
@ 2017-11-01 23:48 ` Rob Herring
2017-11-04 11:24 ` Greg Kroah-Hartman
1 sibling, 0 replies; 22+ messages in thread
From: Rob Herring @ 2017-11-01 23:48 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-serial, Chris Healy, Guenter Roeck,
Lucas Stach, Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman,
Pavel Machek, Andy Shevchenko, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 11:36 AM, Andrey Smirnov
<andrew.smirnov@gmail.com> wrote:
> Using devres infrastructure it is possible to write a serdev driver
> that doesn't have any code that needs to be called as a part of
> .remove. Add code to make .remove optional.
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-serial@vger.kernel.org
> Cc: Rob Herring <robh@kernel.org>
> Cc: cphealy@gmail.com
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
> drivers/tty/serdev/core.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Acked-by: Rob Herring <robh@kernel.org>
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional
2017-10-31 16:36 ` [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional Andrey Smirnov
2017-11-01 23:48 ` Rob Herring
@ 2017-11-04 11:24 ` Greg Kroah-Hartman
2017-11-04 17:05 ` Sebastian Reichel
1 sibling, 1 reply; 22+ messages in thread
From: Greg Kroah-Hartman @ 2017-11-04 11:24 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-serial, Rob Herring, cphealy, Guenter Roeck,
Lucas Stach, Nikita Yushchenko, Lee Jones, Pavel Machek,
Andy Shevchenko, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:52AM -0700, Andrey Smirnov wrote:
> Using devres infrastructure it is possible to write a serdev driver
> that doesn't have any code that needs to be called as a part of
> .remove. Add code to make .remove optional.
What about manual unbind from userspace through sysfs? You need to have
a remove function. All drivers need that, to not have it is pretty lazy :)
thanks,
greg k-h
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH v10 1/5] serdev: Make .remove in struct serdev_device_driver optional
2017-11-04 11:24 ` Greg Kroah-Hartman
@ 2017-11-04 17:05 ` Sebastian Reichel
0 siblings, 0 replies; 22+ messages in thread
From: Sebastian Reichel @ 2017-11-04 17:05 UTC (permalink / raw)
To: Greg Kroah-Hartman
Cc: Andrey Smirnov, linux-kernel, linux-serial, Rob Herring, cphealy,
Guenter Roeck, Lucas Stach, Nikita Yushchenko, Lee Jones,
Pavel Machek, Andy Shevchenko, Johan Hovold
[-- Attachment #1: Type: text/plain, Size: 906 bytes --]
Hi Greg,
On Sat, Nov 04, 2017 at 12:24:31PM +0100, Greg Kroah-Hartman wrote:
> On Tue, Oct 31, 2017 at 09:36:52AM -0700, Andrey Smirnov wrote:
> > Using devres infrastructure it is possible to write a serdev driver
> > that doesn't have any code that needs to be called as a part of
> > .remove. Add code to make .remove optional.
>
> What about manual unbind from userspace through sysfs? You need to have
> a remove function. All drivers need that, to not have it is pretty lazy :)
Resources,that have been requested via devres are free'd by the
devres framework after the driver specific remove function. This
also works for manual unbind.
If *all* driver resources are allocated using devres, you will end
up with an empty remove function. In that case it makes sense to
remove it completly. This is what quite a few (mainline) i2c and
spi drivers actually do.
-- Sebastian
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]
^ permalink raw reply [flat|nested] 22+ messages in thread
* [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open()
2017-10-31 16:36 ` Andrey Smirnov
(?)
(?)
@ 2017-10-31 16:36 ` Andrey Smirnov
2017-11-01 23:48 ` Rob Herring
2017-11-05 15:38 ` Johan Hovold
-1 siblings, 2 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel
Cc: Andrey Smirnov, linux-serial, Rob Herring, cphealy,
Guenter Roeck, Lucas Stach, Nikita Yushchenko, Lee Jones,
Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko, Johan Hovold,
Sebastian Reichel
Add code implementing managed version of serdev_device_open() for
serdev device drivers that "open" the device during driver's lifecycle
only once (e.g. opened in .probe() and closed in .remove()).
Cc: linux-kernel@vger.kernel.org
Cc: linux-serial@vger.kernel.org
Cc: Rob Herring <robh@kernel.org>
Cc: cphealy@gmail.com
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Lee Jones <lee.jones@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Johan Hovold <johan@kernel.org>
Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
Documentation/driver-model/devres.txt | 3 +++
drivers/tty/serdev/core.c | 27 +++++++++++++++++++++++++++
include/linux/serdev.h | 1 +
3 files changed, 31 insertions(+)
diff --git a/Documentation/driver-model/devres.txt b/Documentation/driver-model/devres.txt
index 69f08c0f23a8..e9c6b5cfeec1 100644
--- a/Documentation/driver-model/devres.txt
+++ b/Documentation/driver-model/devres.txt
@@ -383,6 +383,9 @@ RESET
devm_reset_control_get()
devm_reset_controller_register()
+SERDEV
+ devm_serdev_device_open()
+
SLAVE DMA ENGINE
devm_acpi_dma_controller_register()
diff --git a/drivers/tty/serdev/core.c b/drivers/tty/serdev/core.c
index f500f6a2ca88..b3a785665c6f 100644
--- a/drivers/tty/serdev/core.c
+++ b/drivers/tty/serdev/core.c
@@ -116,6 +116,33 @@ void serdev_device_close(struct serdev_device *serdev)
}
EXPORT_SYMBOL_GPL(serdev_device_close);
+static void devm_serdev_device_release(struct device *dev, void *dr)
+{
+ serdev_device_close(*(struct serdev_device **)dr);
+}
+
+int devm_serdev_device_open(struct device *dev, struct serdev_device *serdev)
+{
+ struct serdev_device **dr;
+ int ret;
+
+ dr = devres_alloc(devm_serdev_device_release, sizeof(*dr), GFP_KERNEL);
+ if (!dr)
+ return -ENOMEM;
+
+ ret = serdev_device_open(serdev);
+ if (ret) {
+ devres_free(dr);
+ return ret;
+ }
+
+ *dr = serdev;
+ devres_add(dev, dr);
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(devm_serdev_device_open);
+
void serdev_device_write_wakeup(struct serdev_device *serdev)
{
complete(&serdev->write_comp);
diff --git a/include/linux/serdev.h b/include/linux/serdev.h
index e69402d4a8ae..9929063bd45d 100644
--- a/include/linux/serdev.h
+++ b/include/linux/serdev.h
@@ -193,6 +193,7 @@ static inline int serdev_controller_receive_buf(struct serdev_controller *ctrl,
int serdev_device_open(struct serdev_device *);
void serdev_device_close(struct serdev_device *);
+int devm_serdev_device_open(struct device *, struct serdev_device *);
unsigned int serdev_device_set_baudrate(struct serdev_device *, unsigned int);
void serdev_device_set_flow_control(struct serdev_device *, bool);
int serdev_device_write_buf(struct serdev_device *, const unsigned char *, size_t);
--
2.13.6
^ permalink raw reply related [flat|nested] 22+ messages in thread* Re: [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open()
2017-10-31 16:36 ` [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open() Andrey Smirnov
@ 2017-11-01 23:48 ` Rob Herring
2017-11-05 15:38 ` Johan Hovold
1 sibling, 0 replies; 22+ messages in thread
From: Rob Herring @ 2017-11-01 23:48 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-serial, Chris Healy, Guenter Roeck,
Lucas Stach, Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman,
Pavel Machek, Andy Shevchenko, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 11:36 AM, Andrey Smirnov
<andrew.smirnov@gmail.com> wrote:
> Add code implementing managed version of serdev_device_open() for
> serdev device drivers that "open" the device during driver's lifecycle
> only once (e.g. opened in .probe() and closed in .remove()).
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-serial@vger.kernel.org
> Cc: Rob Herring <robh@kernel.org>
> Cc: cphealy@gmail.com
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
> Documentation/driver-model/devres.txt | 3 +++
> drivers/tty/serdev/core.c | 27 +++++++++++++++++++++++++++
> include/linux/serdev.h | 1 +
> 3 files changed, 31 insertions(+)
Acked-by: Rob Herring <robh@kernel.org>
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open()
2017-10-31 16:36 ` [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open() Andrey Smirnov
2017-11-01 23:48 ` Rob Herring
@ 2017-11-05 15:38 ` Johan Hovold
2017-11-05 22:19 ` Andrey Smirnov
1 sibling, 1 reply; 22+ messages in thread
From: Johan Hovold @ 2017-11-05 15:38 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-serial, Rob Herring, cphealy, Guenter Roeck,
Lucas Stach, Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman,
Pavel Machek, Andy Shevchenko, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:53AM -0700, Andrey Smirnov wrote:
> Add code implementing managed version of serdev_device_open() for
> serdev device drivers that "open" the device during driver's lifecycle
> only once (e.g. opened in .probe() and closed in .remove()).
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-serial@vger.kernel.org
> Cc: Rob Herring <robh@kernel.org>
> Cc: cphealy@gmail.com
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
> Documentation/driver-model/devres.txt | 3 +++
> drivers/tty/serdev/core.c | 27 +++++++++++++++++++++++++++
> include/linux/serdev.h | 1 +
> 3 files changed, 31 insertions(+)
> +int devm_serdev_device_open(struct device *dev, struct serdev_device *serdev)
> +{
> + struct serdev_device **dr;
> + int ret;
> +
> + dr = devres_alloc(devm_serdev_device_release, sizeof(*dr), GFP_KERNEL);
> + if (!dr)
> + return -ENOMEM;
> +
> + ret = serdev_device_open(serdev);
> + if (ret) {
> + devres_free(dr);
> + return ret;
> + }
> +
> + *dr = serdev;
> + devres_add(dev, dr);
> +
> + return ret;
This would be more readable as return 0.
> +}
> +EXPORT_SYMBOL_GPL(devm_serdev_device_open);
Johan
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 2/5] serdev: Introduce devm_serdev_device_open()
2017-11-05 15:38 ` Johan Hovold
@ 2017-11-05 22:19 ` Andrey Smirnov
0 siblings, 0 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-11-05 22:19 UTC (permalink / raw)
To: Johan Hovold
Cc: linux-kernel, linux-serial, Rob Herring, Chris Healy,
Guenter Roeck, Lucas Stach, Nikita Yushchenko, Lee Jones,
Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko,
Sebastian Reichel
On Sun, Nov 5, 2017 at 7:38 AM, Johan Hovold <johan@kernel.org> wrote:
> On Tue, Oct 31, 2017 at 09:36:53AM -0700, Andrey Smirnov wrote:
>> Add code implementing managed version of serdev_device_open() for
>> serdev device drivers that "open" the device during driver's lifecycle
>> only once (e.g. opened in .probe() and closed in .remove()).
>>
>> Cc: linux-kernel@vger.kernel.org
>> Cc: linux-serial@vger.kernel.org
>> Cc: Rob Herring <robh@kernel.org>
>> Cc: cphealy@gmail.com
>> Cc: Guenter Roeck <linux@roeck-us.net>
>> Cc: Lucas Stach <l.stach@pengutronix.de>
>> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Cc: Lee Jones <lee.jones@linaro.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>> ---
>> Documentation/driver-model/devres.txt | 3 +++
>> drivers/tty/serdev/core.c | 27 +++++++++++++++++++++++++++
>> include/linux/serdev.h | 1 +
>> 3 files changed, 31 insertions(+)
>
>> +int devm_serdev_device_open(struct device *dev, struct serdev_device *serdev)
>> +{
>> + struct serdev_device **dr;
>> + int ret;
>> +
>> + dr = devres_alloc(devm_serdev_device_release, sizeof(*dr), GFP_KERNEL);
>> + if (!dr)
>> + return -ENOMEM;
>> +
>> + ret = serdev_device_open(serdev);
>> + if (ret) {
>> + devres_free(dr);
>> + return ret;
>> + }
>> +
>> + *dr = serdev;
>> + devres_add(dev, dr);
>> +
>> + return ret;
>
> This would be more readable as return 0.
Sure, I'll change that in v11.
Thanks,
Andrey Smrinov
^ permalink raw reply [flat|nested] 22+ messages in thread
* [PATCH v10 3/5] mfd: Add driver for RAVE Supervisory Processor
2017-10-31 16:36 ` Andrey Smirnov
` (2 preceding siblings ...)
(?)
@ 2017-10-31 16:36 ` Andrey Smirnov
2017-11-05 15:38 ` Johan Hovold
-1 siblings, 1 reply; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel
Cc: Andrey Smirnov, cphealy, Lucas Stach, Nikita Yushchenko,
Lee Jones, Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko,
Guenter Roeck, Rob Herring, Johan Hovold, Sebastian Reichel
Add a driver for RAVE Supervisory Processor, an MCU implementing
various bits of housekeeping functionality (watchdoging, backlight
control, LED control, etc) on RAVE family of products by Zodiac
Inflight Innovations.
This driver implementes core MFD/serdev device as well as
communication subroutines necessary for commanding the device.
Cc: linux-kernel@vger.kernel.org
Cc: cphealy@gmail.com
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Lee Jones <lee.jones@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Rob Herring <robh@kernel.org>
Cc: Johan Hovold <johan@kernel.org>
Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Tested-by: Chris Healy <cphealy@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
drivers/mfd/Kconfig | 9 +
drivers/mfd/Makefile | 2 +
drivers/mfd/rave-sp.c | 659 +++++++++++++++++++++++++++++++++++++++++++++++
drivers/platform/Kconfig | 1 +
include/linux/rave-sp.h | 54 ++++
5 files changed, 725 insertions(+)
create mode 100644 drivers/mfd/rave-sp.c
create mode 100644 include/linux/rave-sp.h
diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig
index fc5e4fef89d2..8e37d4d4fad0 100644
--- a/drivers/mfd/Kconfig
+++ b/drivers/mfd/Kconfig
@@ -1818,4 +1818,13 @@ config MFD_VEXPRESS_SYSREG
on the ARM Ltd. Versatile Express board.
endmenu
+
+config RAVE_SP_CORE
+ tristate "RAVE SP MCU core driver"
+ depends on SERIAL_DEV_BUS
+ select CRC_CCITT
+ help
+ Select this to get support for the Supervisory Processor
+ device found on several devices in RAVE line of hardware.
+
endif
diff --git a/drivers/mfd/Makefile b/drivers/mfd/Makefile
index c3d0a1b39bb6..808ab024a796 100644
--- a/drivers/mfd/Makefile
+++ b/drivers/mfd/Makefile
@@ -226,3 +226,5 @@ obj-$(CONFIG_MFD_SUN4I_GPADC) += sun4i-gpadc.o
obj-$(CONFIG_MFD_STM32_LPTIMER) += stm32-lptimer.o
obj-$(CONFIG_MFD_STM32_TIMERS) += stm32-timers.o
obj-$(CONFIG_MFD_MXS_LRADC) += mxs-lradc.o
+
+obj-$(CONFIG_RAVE_SP_CORE) += rave-sp.o
diff --git a/drivers/mfd/rave-sp.c b/drivers/mfd/rave-sp.c
new file mode 100644
index 000000000000..42202eb63e75
--- /dev/null
+++ b/drivers/mfd/rave-sp.c
@@ -0,0 +1,659 @@
+/*
+ * Multifunction core driver for Zodiac Inflight Innovations
+ * SP MCU that is connected via dedicated UART port
+ *
+ * Copyright (C) 2017 Zodiac Inflight Innovations
+ *
+ * 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, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/atomic.h>
+#include <linux/crc-ccitt.h>
+#include <linux/delay.h>
+#include <linux/export.h>
+#include <linux/init.h>
+#include <linux/slab.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_device.h>
+#include <linux/rave-sp.h>
+#include <linux/sched.h>
+#include <linux/serdev.h>
+#include <asm/unaligned.h>
+
+/*
+ * UART protocol using following entities:
+ * - message to MCU => ACK response
+ * - event from MCU => event ACK
+ *
+ * Frame structure:
+ * <STX> <DATA> <CHECKSUM> <ETX>
+ * Where:
+ * - STX - is start of transmission character
+ * - ETX - end of transmission
+ * - DATA - payload
+ * - CHECKSUM - checksum calculated on <DATA>
+ *
+ * If <DATA> or <CHECKSUM> contain one of control characters, then it is
+ * escaped using <DLE> control code. Added <DLE> does not participate in
+ * checksum calculation.
+ */
+#define RAVE_SP_STX 0x02
+#define RAVE_SP_ETX 0x03
+#define RAVE_SP_DLE 0x10
+
+#define RAVE_SP_MAX_DATA_SIZE 64
+#define RAVE_SP_CHECKSUM_SIZE 2 /* Worst case scenario on RDU2 */
+/*
+ * We don't store STX, ETX and unescaped bytes, so Rx is only
+ * DATA + CSUM
+ */
+#define RAVE_SP_RX_BUFFER_SIZE \
+ (RAVE_SP_MAX_DATA_SIZE + RAVE_SP_CHECKSUM_SIZE)
+
+#define RAVE_SP_STX_ETX_SIZE 2
+/*
+ * For Tx we have to have space for everything, STX, EXT and
+ * potentially stuffed DATA + CSUM data + csum
+ */
+#define RAVE_SP_TX_BUFFER_SIZE \
+ (RAVE_SP_STX_ETX_SIZE + 2 * RAVE_SP_RX_BUFFER_SIZE)
+
+#define RAVE_SP_BOOT_SOURCE_GET 0
+#define RAVE_SP_BOOT_SOURCE_SET 1
+
+#define RAVE_SP_RDU2_BOARD_TYPE_RMB 0
+#define RAVE_SP_RDU2_BOARD_TYPE_DEB 1
+
+#define RAVE_SP_BOOT_SOURCE_SD 0
+#define RAVE_SP_BOOT_SOURCE_EMMC 1
+#define RAVE_SP_BOOT_SOURCE_NOR 2
+
+/**
+ * enum rave_sp_deframer_state - Possible state for de-framer
+ *
+ * @RAVE_SP_EXPECT_SOF: Scanning input for start-of-frame marker
+ * @RAVE_SP_EXPECT_DATA: Got start of frame marker, collecting frame
+ * @RAVE_SP_EXPECT_ESCAPED_DATA: Got escape character, collecting escaped byte
+ */
+enum rave_sp_deframer_state {
+ RAVE_SP_EXPECT_SOF,
+ RAVE_SP_EXPECT_DATA,
+ RAVE_SP_EXPECT_ESCAPED_DATA,
+};
+
+/**
+ * struct rave_sp - Device protocol deframer
+ *
+ * @state: Current state of the deframer
+ * @data: Buffer used to collect deframed data
+ * @length: Number of bytes de-framed so far
+ */
+struct rave_sp_deframer {
+ enum rave_sp_deframer_state state;
+ unsigned char data[RAVE_SP_RX_BUFFER_SIZE];
+ size_t length;
+};
+
+/**
+ * struct rave_sp_reply - reply as per RAVE device protocol
+ *
+ * @length: Expected reply length
+ * @data: Buffer to store reply payload in
+ * @code: Expected reply code
+ * @ackid: Expected reply ACK ID
+ * @completion: Successful reply reception completion
+ */
+struct rave_sp_reply {
+ size_t length;
+ void *data;
+ u8 code;
+ u8 ackid;
+ struct completion received;
+};
+
+/**
+ * struct rave_sp_checksum - Variant specific checksum implementation details
+ *
+ * @length: Caculated checksum length
+ * @subroutine: Utilized checksum algorithm implementation
+ */
+struct rave_sp_checksum {
+ size_t length;
+ void (*subroutine)(const u8 *, size_t, u8 *);
+};
+
+/**
+ * struct rave_sp_variant_cmds - Variant specific command routines
+ *
+ * @translate: Generic to variant specific command mapping routine
+ *
+ */
+struct rave_sp_variant_cmds {
+ int (*translate)(enum rave_sp_command);
+};
+
+/**
+ * struct rave_sp_variant - RAVE supervisory processor core variant
+ *
+ * @checksum: Variant specific checksum implementation
+ * @cmd: Variant specific command pointer table
+ *
+ */
+struct rave_sp_variant {
+ const struct rave_sp_checksum *checksum;
+ struct rave_sp_variant_cmds cmd;
+};
+
+/**
+ * struct rave_sp - RAVE supervisory processor core
+ *
+ * @serdev: Pointer to underlying serdev
+ * @deframer: Stored state of the protocol deframer
+ * @ackid: ACK ID used in last reply sent to the device
+ * @bus_lock: Lock to serialize access to the device
+ * @reply_lock: Lock protecting @reply
+ * @reply: Pointer to memory to store reply payload
+ *
+ * @variant: Device variant specific information
+ * @event_notifier_list: Input event notification chain
+ *
+ */
+struct rave_sp {
+ struct serdev_device *serdev;
+ struct rave_sp_deframer deframer;
+ atomic_t ackid;
+ struct mutex bus_lock;
+ struct mutex reply_lock;
+ struct rave_sp_reply *reply;
+
+ const struct rave_sp_variant *variant;
+ struct blocking_notifier_head event_notifier_list;
+};
+
+static bool rave_sp_id_is_event(u8 code)
+{
+ return (code & 0xF0) == RAVE_SP_EVNT_BASE;
+}
+
+static void rave_sp_unregister_event_notifier(struct device *dev, void *res)
+{
+ struct rave_sp *sp = dev_get_drvdata(dev->parent);
+ struct notifier_block *nb = *(struct notifier_block **)res;
+ struct blocking_notifier_head *bnh = &sp->event_notifier_list;
+
+ WARN_ON(blocking_notifier_chain_unregister(bnh, nb));
+}
+
+int devm_rave_sp_register_event_notifier(struct device *dev,
+ struct notifier_block *nb)
+{
+ struct rave_sp *sp = dev_get_drvdata(dev->parent);
+ struct notifier_block **rcnb;
+ int ret;
+
+ rcnb = devres_alloc(rave_sp_unregister_event_notifier,
+ sizeof(*rcnb), GFP_KERNEL);
+ if (!rcnb)
+ return -ENOMEM;
+
+ ret = blocking_notifier_chain_register(&sp->event_notifier_list, nb);
+ if (!ret) {
+ *rcnb = nb;
+ devres_add(dev, rcnb);
+ } else {
+ devres_free(rcnb);
+ }
+
+ return ret;
+}
+EXPORT_SYMBOL_GPL(devm_rave_sp_register_event_notifier);
+
+static void csum_8b2c(const u8 *buf, size_t size, u8 *crc)
+{
+ *crc = *buf++;
+ size--;
+
+ while (size--)
+ *crc += *buf++;
+
+ *crc = 1 + ~(*crc);
+}
+
+static void csum_ccitt(const u8 *buf, size_t size, u8 *crc)
+{
+ const u16 calculated = crc_ccitt_false(0xffff, buf, size);
+
+ /*
+ * While the rest of the wire protocol is little-endian,
+ * CCITT-16 CRC in RDU2 device is sent out in big-endian order.
+ */
+ put_unaligned_be16(calculated, crc);
+}
+
+static void *stuff(unsigned char *dest, const unsigned char *src, size_t n)
+{
+ while (n--) {
+ const unsigned char byte = *src++;
+
+ switch (byte) {
+ case RAVE_SP_STX:
+ case RAVE_SP_ETX:
+ case RAVE_SP_DLE:
+ *dest++ = RAVE_SP_DLE;
+ /* FALLTHROUGH */
+ default:
+ *dest++ = byte;
+ }
+ }
+
+ return dest;
+}
+
+static int rave_sp_write(struct rave_sp *sp, const u8 *data, u8 data_size)
+{
+ const size_t checksum_length = sp->variant->checksum->length;
+ unsigned char frame[RAVE_SP_TX_BUFFER_SIZE];
+ unsigned char crc[RAVE_SP_CHECKSUM_SIZE];
+ unsigned char *dest = frame;
+ size_t length;
+
+ if (WARN_ON(checksum_length > sizeof(crc)))
+ return -ENOMEM;
+
+ if (WARN_ON(data_size > sizeof(frame)))
+ return -ENOMEM;
+
+ sp->variant->checksum->subroutine(data, data_size, crc);
+
+ *dest++ = RAVE_SP_STX;
+ dest = stuff(dest, data, data_size);
+ dest = stuff(dest, crc, checksum_length);
+ *dest++ = RAVE_SP_ETX;
+
+ length = dest - frame;
+
+ print_hex_dump(KERN_DEBUG, "rave-sp tx: ", DUMP_PREFIX_NONE,
+ 16, 1, frame, length, false);
+
+ return serdev_device_write(sp->serdev, frame, length, HZ);
+}
+
+static u8 rave_sp_reply_code(u8 command)
+{
+ /*
+ * There isn't a single rule that describes command code ->
+ * ACK code transformation, but, going through various
+ * versions of ICDs, there appear to be three distinct groups
+ * that can be described by simple transformation.
+ */
+ switch (command) {
+ case 0xA0 ... 0xBE:
+ /*
+ * Commands implemented by firmware found in RDU1 and
+ * older devices all seem to obey the following rule
+ */
+ return command + 0x20;
+ case 0xE0 ... 0xEF:
+ /*
+ * Events emitted by all versions of the firmare use
+ * least significant bit to get an ACK code
+ */
+ return command | 0x01;
+ default:
+ /*
+ * Commands implemented by firmware found in RDU2 are
+ * similar to "old" commands, but they use slightly
+ * different offset
+ */
+ return command + 0x40;
+ }
+}
+
+int rave_sp_exec(struct rave_sp *sp,
+ void *__data, size_t data_size,
+ void *reply_data, size_t reply_data_size)
+{
+ struct rave_sp_reply reply = {
+ .data = reply_data,
+ .length = reply_data_size,
+ .received = COMPLETION_INITIALIZER_ONSTACK(reply.received),
+ };
+ unsigned char *data = __data;
+ int command, ret = 0;
+ u8 ackid;
+
+ command = sp->variant->cmd.translate(data[0]);
+ if (command < 0)
+ return command;
+
+ ackid = atomic_inc_return(&sp->ackid);
+ reply.ackid = ackid;
+ reply.code = rave_sp_reply_code((u8)command),
+
+ mutex_lock(&sp->bus_lock);
+
+ mutex_lock(&sp->reply_lock);
+ sp->reply = &reply;
+ mutex_unlock(&sp->reply_lock);
+
+ data[0] = command;
+ data[1] = ackid;
+
+ rave_sp_write(sp, data, data_size);
+
+ if (!wait_for_completion_timeout(&reply.received, HZ)) {
+ dev_err(&sp->serdev->dev, "Command timeout\n");
+ ret = -ETIMEDOUT;
+
+ mutex_lock(&sp->reply_lock);
+ sp->reply = NULL;
+ mutex_unlock(&sp->reply_lock);
+ }
+
+ mutex_unlock(&sp->bus_lock);
+ return ret;
+}
+EXPORT_SYMBOL(rave_sp_exec);
+
+static void rave_sp_receive_event(struct rave_sp *sp,
+ const unsigned char *data, size_t length)
+{
+ u8 cmd[] = {
+ [0] = rave_sp_reply_code(data[0]),
+ [1] = data[1],
+ };
+
+ rave_sp_write(sp, cmd, sizeof(cmd));
+
+ blocking_notifier_call_chain(&sp->event_notifier_list,
+ rave_sp_action_pack(data[0], data[2]),
+ NULL);
+}
+
+static void rave_sp_receive_reply(struct rave_sp *sp,
+ const unsigned char *data, size_t length)
+{
+ struct device *dev = &sp->serdev->dev;
+ struct rave_sp_reply *reply;
+ const size_t payload_length = length - 2;
+
+ mutex_lock(&sp->reply_lock);
+ reply = sp->reply;
+
+ if (reply) {
+ if (reply->code == data[0] && reply->ackid == data[1] &&
+ payload_length >= reply->length) {
+ /*
+ * We are relying on memcpy(dst, src, 0) to be a no-op
+ * when handling commands that have a no-payload reply
+ */
+ memcpy(reply->data, &data[2], reply->length);
+ complete(&reply->received);
+ sp->reply = NULL;
+ } else {
+ dev_err(dev, "Ignoring incorrect reply\n");
+ dev_dbg(dev, "Code: expected = 0x%08x received = 0x%08x\n",
+ reply->code, data[0]);
+ dev_dbg(dev, "ACK ID: expected = 0x%08x received = 0x%08x\n",
+ reply->ackid, data[1]);
+ dev_dbg(dev, "Length: expected = %zu received = %zu\n",
+ reply->length, payload_length);
+ }
+ }
+
+ mutex_unlock(&sp->reply_lock);
+}
+
+static void rave_sp_receive_frame(struct rave_sp *sp,
+ const unsigned char *data,
+ size_t length)
+{
+ const size_t checksum_length = sp->variant->checksum->length;
+ const size_t payload_length = length - checksum_length;
+ const u8 *crc_reported = &data[payload_length];
+ struct device *dev = &sp->serdev->dev;
+ u8 crc_calculated[checksum_length];
+
+ print_hex_dump(KERN_DEBUG, "rave-sp rx: ", DUMP_PREFIX_NONE,
+ 16, 1, data, length, false);
+
+ if (unlikely(length <= checksum_length)) {
+ dev_warn(dev, "Dropping short frame\n");
+ return;
+ }
+
+ sp->variant->checksum->subroutine(data, payload_length,
+ crc_calculated);
+
+ if (memcmp(crc_calculated, crc_reported, checksum_length)) {
+ dev_warn(dev, "Dropping bad frame\n");
+ return;
+ }
+
+ if (rave_sp_id_is_event(data[0]))
+ rave_sp_receive_event(sp, data, length);
+ else
+ rave_sp_receive_reply(sp, data, length);
+}
+
+static int rave_sp_receive_buf(struct serdev_device *serdev,
+ const unsigned char *buf, size_t size)
+{
+ struct device *dev = &serdev->dev;
+ struct rave_sp *sp = dev_get_drvdata(dev);
+ struct rave_sp_deframer *deframer = &sp->deframer;
+ const unsigned char *src = buf;
+ const unsigned char *end = buf + size;
+ bool reset_framer = false;
+
+ while (src < end) {
+ const unsigned char byte = *src++;
+
+ switch (deframer->state) {
+ case RAVE_SP_EXPECT_SOF:
+ if (byte == RAVE_SP_STX)
+ deframer->state = RAVE_SP_EXPECT_DATA;
+ continue;
+
+ case RAVE_SP_EXPECT_DATA:
+ switch (byte) {
+ case RAVE_SP_ETX:
+ rave_sp_receive_frame(sp,
+ deframer->data,
+ deframer->length);
+ reset_framer = true;
+ break;
+ case RAVE_SP_STX:
+ dev_warn(dev, "Bad frame: STX before ETX\n");
+ reset_framer = true;
+ break;
+ case RAVE_SP_DLE:
+ deframer->state = RAVE_SP_EXPECT_ESCAPED_DATA;
+ continue;
+ }
+
+ case RAVE_SP_EXPECT_ESCAPED_DATA: /* FALLTHROUGH */
+ deframer->data[deframer->length++] = byte;
+
+ if (deframer->length == sizeof(deframer->data)) {
+ dev_warn(dev, "Bad frame: Too long\n");
+ reset_framer = true;
+ break;
+ }
+
+ deframer->state = RAVE_SP_EXPECT_DATA;
+ break;
+ }
+ }
+
+ if (reset_framer) {
+ deframer->state = RAVE_SP_EXPECT_SOF;
+ deframer->length = 0;
+ }
+
+ return src - buf;
+}
+
+static int rave_sp_rdu1_cmd_translate(enum rave_sp_command command)
+{
+ if (command >= RAVE_SP_CMD_STATUS &&
+ command <= RAVE_SP_CMD_CONTROL_EVENTS)
+ return command;
+
+ return -EINVAL;
+}
+
+static int rave_sp_rdu2_cmd_translate(enum rave_sp_command command)
+{
+ if (command >= RAVE_SP_CMD_GET_FIRMWARE_VERSION &&
+ command <= RAVE_SP_CMD_GET_GPIO_STATE)
+ return command;
+
+ if (command == RAVE_SP_CMD_REQ_COPPER_REV) {
+ /*
+ * As per RDU2 ICD 3.4.47 CMD_GET_COPPER_REV code is
+ * different from that for RDU1 and it is set to 0x28.
+ */
+ return 0x28;
+ }
+
+ return rave_sp_rdu1_cmd_translate(command);
+}
+
+static int rave_sp_default_cmd_translate(enum rave_sp_command command)
+{
+ /*
+ * All of the following command codes were taken from "Table :
+ * Communications Protocol Message Types" in section 3.3
+ * "MESSAGE TYPES" of Rave PIC24 ICD.
+ */
+ switch (command) {
+ case RAVE_SP_CMD_GET_FIRMWARE_VERSION:
+ return 0x11;
+ case RAVE_SP_CMD_GET_BOOTLOADER_VERSION:
+ return 0x12;
+ case RAVE_SP_CMD_BOOT_SOURCE:
+ return 0x14;
+ case RAVE_SP_CMD_SW_WDT:
+ return 0x1C;
+ case RAVE_SP_CMD_RESET:
+ return 0x1E;
+ case RAVE_SP_CMD_RESET_REASON:
+ return 0x1F;
+ default:
+ return -EINVAL;
+ }
+}
+
+static const struct rave_sp_checksum rave_sp_checksum_8b2c = {
+ .length = 1,
+ .subroutine = csum_8b2c,
+};
+
+static const struct rave_sp_checksum rave_sp_checksum_ccitt = {
+ .length = 2,
+ .subroutine = csum_ccitt,
+};
+
+static const struct rave_sp_variant rave_sp_legacy = {
+ .checksum = &rave_sp_checksum_8b2c,
+ .cmd = {
+ .translate = rave_sp_default_cmd_translate,
+ },
+};
+
+static const struct rave_sp_variant rave_sp_rdu1 = {
+ .checksum = &rave_sp_checksum_8b2c,
+ .cmd = {
+ .translate = rave_sp_rdu1_cmd_translate,
+ },
+};
+
+static const struct rave_sp_variant rave_sp_rdu2 = {
+ .checksum = &rave_sp_checksum_ccitt,
+ .cmd = {
+ .translate = rave_sp_rdu2_cmd_translate,
+ },
+};
+
+static const struct of_device_id rave_sp_dt_ids[] = {
+ { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_rdu1 },
+ { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_rdu2 },
+ { /* sentinel */ }
+};
+
+static const struct serdev_device_ops rave_sp_serdev_device_ops = {
+ .receive_buf = rave_sp_receive_buf,
+ .write_wakeup = serdev_device_write_wakeup,
+};
+
+static int rave_sp_probe(struct serdev_device *serdev)
+{
+ struct device *dev = &serdev->dev;
+ struct rave_sp *sp;
+ u32 baud;
+ int ret;
+
+ if (of_property_read_u32(dev->of_node, "current-speed", &baud)) {
+ dev_err(dev,
+ "'current-speed' is not specified in device node\n");
+ return -EINVAL;
+ }
+
+ sp = devm_kzalloc(dev, sizeof(*sp), GFP_KERNEL);
+ if (!sp)
+ return -ENOMEM;
+
+ sp->serdev = serdev;
+ dev_set_drvdata(dev, sp);
+
+ sp->variant = of_device_get_match_data(dev);
+ if (!sp->variant)
+ return -ENODEV;
+
+ mutex_init(&sp->bus_lock);
+ mutex_init(&sp->reply_lock);
+ BLOCKING_INIT_NOTIFIER_HEAD(&sp->event_notifier_list);
+
+ serdev_device_set_client_ops(serdev, &rave_sp_serdev_device_ops);
+ ret = devm_serdev_device_open(dev, serdev);
+ if (ret)
+ return ret;
+
+ serdev_device_set_baudrate(serdev, baud);
+
+ return devm_of_platform_populate(dev);
+}
+
+MODULE_DEVICE_TABLE(of, rave_sp_dt_ids);
+
+static struct serdev_device_driver rave_sp_drv = {
+ .probe = rave_sp_probe,
+ .driver = {
+ .name = "rave-sp",
+ .of_match_table = rave_sp_dt_ids,
+ },
+};
+module_serdev_device_driver(rave_sp_drv);
+
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR("Andrey Vostrikov <andrey.vostrikov@cogentembedded.com>");
+MODULE_AUTHOR("Nikita Yushchenko <nikita.yoush@cogentembedded.com>");
+MODULE_AUTHOR("Andrey Smirnov <andrew.smirnov@gmail.com>");
+MODULE_DESCRIPTION("RAVE SP core driver");
diff --git a/drivers/platform/Kconfig b/drivers/platform/Kconfig
index c11db8bceea1..989d38b587bf 100644
--- a/drivers/platform/Kconfig
+++ b/drivers/platform/Kconfig
@@ -8,3 +8,4 @@ endif
source "drivers/platform/goldfish/Kconfig"
source "drivers/platform/chrome/Kconfig"
+
diff --git a/include/linux/rave-sp.h b/include/linux/rave-sp.h
new file mode 100644
index 000000000000..811929afc1a9
--- /dev/null
+++ b/include/linux/rave-sp.h
@@ -0,0 +1,54 @@
+#ifndef _LINUX_RAVE_SP_H_
+#define _LINUX_RAVE_SP_H_
+
+enum rave_sp_command {
+ RAVE_SP_CMD_GET_FIRMWARE_VERSION = 0x20,
+ RAVE_SP_CMD_GET_BOOTLOADER_VERSION = 0x21,
+ RAVE_SP_CMD_BOOT_SOURCE = 0x26,
+ RAVE_SP_CMD_GET_BOARD_COPPER_REV = 0x2B,
+ RAVE_SP_CMD_GET_GPIO_STATE = 0x2F,
+
+ RAVE_SP_CMD_STATUS = 0xA0,
+ RAVE_SP_CMD_SW_WDT = 0xA1,
+ RAVE_SP_CMD_PET_WDT = 0xA2,
+ RAVE_SP_CMD_RESET = 0xA7,
+ RAVE_SP_CMD_RESET_REASON = 0xA8,
+
+ RAVE_SP_CMD_REQ_COPPER_REV = 0xB6,
+ RAVE_SP_CMD_GET_I2C_DEVICE_STATUS = 0xBA,
+ RAVE_SP_CMD_GET_SP_SILICON_REV = 0xB9,
+ RAVE_SP_CMD_CONTROL_EVENTS = 0xBB,
+
+ RAVE_SP_EVNT_BASE = 0xE0,
+};
+
+struct rave_sp;
+
+static inline unsigned long rave_sp_action_pack(u8 event, u8 value)
+{
+ return ((unsigned long)value << 8) | event;
+}
+
+static inline u8 rave_sp_action_unpack_event(unsigned long action)
+{
+ return action;
+}
+
+static inline u8 rave_sp_action_unpack_value(unsigned long action)
+{
+ return action >> 8;
+}
+
+int rave_sp_exec(struct rave_sp *sp,
+ void *__data, size_t data_size,
+ void *reply_data, size_t reply_data_size);
+int devm_rave_sp_register_event_notifier(struct device *dev,
+ struct notifier_block *nb);
+
+#define COMPATIBLE_RAVE_SP_NIU "zii,rave-sp-niu"
+#define COMPATIBLE_RAVE_SP_MEZZ "zii,rave-sp-mezz"
+#define COMPATIBLE_RAVE_SP_ESB "zii,rave-sp-esb"
+#define COMPATIBLE_RAVE_SP_RDU1 "zii,rave-sp-rdu1"
+#define COMPATIBLE_RAVE_SP_RDU2 "zii,rave-sp-rdu2"
+
+#endif /* _LINUX_RAVE_SP_H_ */
--
2.13.6
^ permalink raw reply related [flat|nested] 22+ messages in thread* Re: [PATCH v10 3/5] mfd: Add driver for RAVE Supervisory Processor
2017-10-31 16:36 ` [PATCH v10 3/5] mfd: Add driver for RAVE Supervisory Processor Andrey Smirnov
@ 2017-11-05 15:38 ` Johan Hovold
2017-11-05 22:02 ` Andrey Smirnov
0 siblings, 1 reply; 22+ messages in thread
From: Johan Hovold @ 2017-11-05 15:38 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, cphealy, Lucas Stach, Nikita Yushchenko, Lee Jones,
Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko, Guenter Roeck,
Rob Herring, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:54AM -0700, Andrey Smirnov wrote:
> Add a driver for RAVE Supervisory Processor, an MCU implementing
> various bits of housekeeping functionality (watchdoging, backlight
> control, LED control, etc) on RAVE family of products by Zodiac
> Inflight Innovations.
>
> This driver implementes core MFD/serdev device as well as
> communication subroutines necessary for commanding the device.
>
> Cc: linux-kernel@vger.kernel.org
> Cc: cphealy@gmail.com
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Rob Herring <robh@kernel.org>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Tested-by: Chris Healy <cphealy@gmail.com>
> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
> +int rave_sp_exec(struct rave_sp *sp,
> + void *__data, size_t data_size,
> + void *reply_data, size_t reply_data_size)
> +{
> + struct rave_sp_reply reply = {
> + .data = reply_data,
> + .length = reply_data_size,
> + .received = COMPLETION_INITIALIZER_ONSTACK(reply.received),
> + };
> + unsigned char *data = __data;
> + int command, ret = 0;
> + u8 ackid;
> +
> + command = sp->variant->cmd.translate(data[0]);
> + if (command < 0)
> + return command;
> +
> + ackid = atomic_inc_return(&sp->ackid);
> + reply.ackid = ackid;
> + reply.code = rave_sp_reply_code((u8)command),
> +
> + mutex_lock(&sp->bus_lock);
> +
> + mutex_lock(&sp->reply_lock);
> + sp->reply = &reply;
> + mutex_unlock(&sp->reply_lock);
> +
> + data[0] = command;
> + data[1] = ackid;
> +
> + rave_sp_write(sp, data, data_size);
> +
> + if (!wait_for_completion_timeout(&reply.received, HZ)) {
> + dev_err(&sp->serdev->dev, "Command timeout\n");
> + ret = -ETIMEDOUT;
> +
> + mutex_lock(&sp->reply_lock);
> + sp->reply = NULL;
> + mutex_unlock(&sp->reply_lock);
> + }
> +
> + mutex_unlock(&sp->bus_lock);
> + return ret;
> +}
> +EXPORT_SYMBOL(rave_sp_exec);
EXPORT_SYMBOL_GPL?
> +static int rave_sp_receive_buf(struct serdev_device *serdev,
> + const unsigned char *buf, size_t size)
> +{
> + struct device *dev = &serdev->dev;
> + struct rave_sp *sp = dev_get_drvdata(dev);
> + struct rave_sp_deframer *deframer = &sp->deframer;
> + const unsigned char *src = buf;
> + const unsigned char *end = buf + size;
> + bool reset_framer = false;
> +
> + while (src < end) {
> + const unsigned char byte = *src++;
> +
> + switch (deframer->state) {
> + case RAVE_SP_EXPECT_SOF:
> + if (byte == RAVE_SP_STX)
> + deframer->state = RAVE_SP_EXPECT_DATA;
> + continue;
> +
> + case RAVE_SP_EXPECT_DATA:
> + switch (byte) {
> + case RAVE_SP_ETX:
> + rave_sp_receive_frame(sp,
> + deframer->data,
> + deframer->length);
> + reset_framer = true;
> + break;
> + case RAVE_SP_STX:
> + dev_warn(dev, "Bad frame: STX before ETX\n");
> + reset_framer = true;
> + break;
> + case RAVE_SP_DLE:
> + deframer->state = RAVE_SP_EXPECT_ESCAPED_DATA;
> + continue;
> + }
> +
> + case RAVE_SP_EXPECT_ESCAPED_DATA: /* FALLTHROUGH */
The fallthrough comment should precede the case statement (or you will
soon receive a patch from the people working enabling
-Wimplicit-fallthrough).
> + deframer->data[deframer->length++] = byte;
> +
> + if (deframer->length == sizeof(deframer->data)) {
> + dev_warn(dev, "Bad frame: Too long\n");
> + reset_framer = true;
> + break;
> + }
> +
> + deframer->state = RAVE_SP_EXPECT_DATA;
> + break;
> + }
> + }
> +
> + if (reset_framer) {
> + deframer->state = RAVE_SP_EXPECT_SOF;
> + deframer->length = 0;
> + }
> +
> + return src - buf;
> +}
> +MODULE_LICENSE("GPL v2");
This doesn't match the license text (GPL-2.0+).
> +MODULE_AUTHOR("Andrey Vostrikov <andrey.vostrikov@cogentembedded.com>");
> +MODULE_AUTHOR("Nikita Yushchenko <nikita.yoush@cogentembedded.com>");
> +MODULE_AUTHOR("Andrey Smirnov <andrew.smirnov@gmail.com>");
> +MODULE_DESCRIPTION("RAVE SP core driver");
> diff --git a/drivers/platform/Kconfig b/drivers/platform/Kconfig
> index c11db8bceea1..989d38b587bf 100644
> --- a/drivers/platform/Kconfig
> +++ b/drivers/platform/Kconfig
> @@ -8,3 +8,4 @@ endif
> source "drivers/platform/goldfish/Kconfig"
>
> source "drivers/platform/chrome/Kconfig"
> +
This chunk should not be here.
> diff --git a/include/linux/rave-sp.h b/include/linux/rave-sp.h
> new file mode 100644
> index 000000000000..811929afc1a9
> --- /dev/null
> +++ b/include/linux/rave-sp.h
I think this one should live in include/linux/mfd/
> @@ -0,0 +1,54 @@
> +#ifndef _LINUX_RAVE_SP_H_
> +#define _LINUX_RAVE_SP_H_
> +
> +enum rave_sp_command {
> + RAVE_SP_CMD_GET_FIRMWARE_VERSION = 0x20,
> + RAVE_SP_CMD_GET_BOOTLOADER_VERSION = 0x21,
> + RAVE_SP_CMD_BOOT_SOURCE = 0x26,
> + RAVE_SP_CMD_GET_BOARD_COPPER_REV = 0x2B,
> + RAVE_SP_CMD_GET_GPIO_STATE = 0x2F,
> +
> + RAVE_SP_CMD_STATUS = 0xA0,
> + RAVE_SP_CMD_SW_WDT = 0xA1,
> + RAVE_SP_CMD_PET_WDT = 0xA2,
> + RAVE_SP_CMD_RESET = 0xA7,
> + RAVE_SP_CMD_RESET_REASON = 0xA8,
> +
> + RAVE_SP_CMD_REQ_COPPER_REV = 0xB6,
> + RAVE_SP_CMD_GET_I2C_DEVICE_STATUS = 0xBA,
> + RAVE_SP_CMD_GET_SP_SILICON_REV = 0xB9,
> + RAVE_SP_CMD_CONTROL_EVENTS = 0xBB,
> +
> + RAVE_SP_EVNT_BASE = 0xE0,
> +};
Johan
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 3/5] mfd: Add driver for RAVE Supervisory Processor
2017-11-05 15:38 ` Johan Hovold
@ 2017-11-05 22:02 ` Andrey Smirnov
0 siblings, 0 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-11-05 22:02 UTC (permalink / raw)
To: Johan Hovold
Cc: linux-kernel, Chris Healy, Lucas Stach, Nikita Yushchenko,
Lee Jones, Greg Kroah-Hartman, Pavel Machek, Andy Shevchenko,
Guenter Roeck, Rob Herring, Sebastian Reichel
On Sun, Nov 5, 2017 at 7:38 AM, Johan Hovold <johan@kernel.org> wrote:
> On Tue, Oct 31, 2017 at 09:36:54AM -0700, Andrey Smirnov wrote:
>> Add a driver for RAVE Supervisory Processor, an MCU implementing
>> various bits of housekeeping functionality (watchdoging, backlight
>> control, LED control, etc) on RAVE family of products by Zodiac
>> Inflight Innovations.
>>
>> This driver implementes core MFD/serdev device as well as
>> communication subroutines necessary for commanding the device.
>>
>> Cc: linux-kernel@vger.kernel.org
>> Cc: cphealy@gmail.com
>> Cc: Lucas Stach <l.stach@pengutronix.de>
>> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Cc: Lee Jones <lee.jones@linaro.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Cc: Guenter Roeck <linux@roeck-us.net>
>> Cc: Rob Herring <robh@kernel.org>
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Tested-by: Chris Healy <cphealy@gmail.com>
>> Reviewed-by: Guenter Roeck <linux@roeck-us.net>
>> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>> ---
>
>> +int rave_sp_exec(struct rave_sp *sp,
>> + void *__data, size_t data_size,
>> + void *reply_data, size_t reply_data_size)
>> +{
>> + struct rave_sp_reply reply = {
>> + .data = reply_data,
>> + .length = reply_data_size,
>> + .received = COMPLETION_INITIALIZER_ONSTACK(reply.received),
>> + };
>> + unsigned char *data = __data;
>> + int command, ret = 0;
>> + u8 ackid;
>> +
>> + command = sp->variant->cmd.translate(data[0]);
>> + if (command < 0)
>> + return command;
>> +
>> + ackid = atomic_inc_return(&sp->ackid);
>> + reply.ackid = ackid;
>> + reply.code = rave_sp_reply_code((u8)command),
>> +
>> + mutex_lock(&sp->bus_lock);
>> +
>> + mutex_lock(&sp->reply_lock);
>> + sp->reply = &reply;
>> + mutex_unlock(&sp->reply_lock);
>> +
>> + data[0] = command;
>> + data[1] = ackid;
>> +
>> + rave_sp_write(sp, data, data_size);
>> +
>> + if (!wait_for_completion_timeout(&reply.received, HZ)) {
>> + dev_err(&sp->serdev->dev, "Command timeout\n");
>> + ret = -ETIMEDOUT;
>> +
>> + mutex_lock(&sp->reply_lock);
>> + sp->reply = NULL;
>> + mutex_unlock(&sp->reply_lock);
>> + }
>> +
>> + mutex_unlock(&sp->bus_lock);
>> + return ret;
>> +}
>> +EXPORT_SYMBOL(rave_sp_exec);
>
> EXPORT_SYMBOL_GPL?
>
Yep, will change in v11.
>> +static int rave_sp_receive_buf(struct serdev_device *serdev,
>> + const unsigned char *buf, size_t size)
>> +{
>> + struct device *dev = &serdev->dev;
>> + struct rave_sp *sp = dev_get_drvdata(dev);
>> + struct rave_sp_deframer *deframer = &sp->deframer;
>> + const unsigned char *src = buf;
>> + const unsigned char *end = buf + size;
>> + bool reset_framer = false;
>> +
>> + while (src < end) {
>> + const unsigned char byte = *src++;
>> +
>> + switch (deframer->state) {
>> + case RAVE_SP_EXPECT_SOF:
>> + if (byte == RAVE_SP_STX)
>> + deframer->state = RAVE_SP_EXPECT_DATA;
>> + continue;
>> +
>> + case RAVE_SP_EXPECT_DATA:
>> + switch (byte) {
>> + case RAVE_SP_ETX:
>> + rave_sp_receive_frame(sp,
>> + deframer->data,
>> + deframer->length);
>> + reset_framer = true;
>> + break;
>> + case RAVE_SP_STX:
>> + dev_warn(dev, "Bad frame: STX before ETX\n");
>> + reset_framer = true;
>> + break;
>> + case RAVE_SP_DLE:
>> + deframer->state = RAVE_SP_EXPECT_ESCAPED_DATA;
>> + continue;
>> + }
>> +
>> + case RAVE_SP_EXPECT_ESCAPED_DATA: /* FALLTHROUGH */
>
> The fallthrough comment should precede the case statement (or you will
> soon receive a patch from the people working enabling
> -Wimplicit-fallthrough).
>
Sure, will change in v11.
>> + deframer->data[deframer->length++] = byte;
>> +
>> + if (deframer->length == sizeof(deframer->data)) {
>> + dev_warn(dev, "Bad frame: Too long\n");
>> + reset_framer = true;
>> + break;
>> + }
>> +
>> + deframer->state = RAVE_SP_EXPECT_DATA;
>> + break;
>> + }
>> + }
>> +
>> + if (reset_framer) {
>> + deframer->state = RAVE_SP_EXPECT_SOF;
>> + deframer->length = 0;
>> + }
>> +
>> + return src - buf;
>> +}
>
>> +MODULE_LICENSE("GPL v2");
>
> This doesn't match the license text (GPL-2.0+).
>
Will update in v11.
>> +MODULE_AUTHOR("Andrey Vostrikov <andrey.vostrikov@cogentembedded.com>");
>> +MODULE_AUTHOR("Nikita Yushchenko <nikita.yoush@cogentembedded.com>");
>> +MODULE_AUTHOR("Andrey Smirnov <andrew.smirnov@gmail.com>");
>> +MODULE_DESCRIPTION("RAVE SP core driver");
>> diff --git a/drivers/platform/Kconfig b/drivers/platform/Kconfig
>> index c11db8bceea1..989d38b587bf 100644
>> --- a/drivers/platform/Kconfig
>> +++ b/drivers/platform/Kconfig
>> @@ -8,3 +8,4 @@ endif
>> source "drivers/platform/goldfish/Kconfig"
>>
>> source "drivers/platform/chrome/Kconfig"
>> +
>
> This chunk should not be here.
>
Missed this one. Will remove in v11.
>> diff --git a/include/linux/rave-sp.h b/include/linux/rave-sp.h
>> new file mode 100644
>> index 000000000000..811929afc1a9
>> --- /dev/null
>> +++ b/include/linux/rave-sp.h
>
> I think this one should live in include/linux/mfd/
My bad, forgot to move that file. Will do in v11.
Thanks,
Andrey Smirnov
^ permalink raw reply [flat|nested] 22+ messages in thread
* [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-10-31 16:36 ` Andrey Smirnov
` (3 preceding siblings ...)
(?)
@ 2017-10-31 16:36 ` Andrey Smirnov
2017-10-31 17:21 ` Guenter Roeck
` (2 more replies)
-1 siblings, 3 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel
Cc: Andrey Smirnov, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Guenter Roeck, Rob Herring, Johan Hovold,
Sebastian Reichel
This driver provides access to RAVE SP watchdog functionality.
Cc: linux-kernel@vger.kernel.org
Cc: linux-watchdog@vger.kernel.org
Cc: cphealy@gmail.com
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Lee Jones <lee.jones@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Rob Herring <robh@kernel.org>
Cc: Johan Hovold <johan@kernel.org>
Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
drivers/watchdog/Kconfig | 7 +
drivers/watchdog/Makefile | 1 +
drivers/watchdog/rave-sp-wdt.c | 349 +++++++++++++++++++++++++++++++++++++++++
3 files changed, 357 insertions(+)
create mode 100644 drivers/watchdog/rave-sp-wdt.c
diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index c722cbfdc7e6..533a72248cd1 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig
@@ -223,6 +223,13 @@ config ZIIRAVE_WATCHDOG
To compile this driver as a module, choose M here: the
module will be called ziirave_wdt.
+config RAVE_SP_WATCHDOG
+ tristate "RAVE SP Watchdog timer"
+ depends on RAVE_SP_CORE
+ select WATCHDOG_CORE
+ help
+ Support for the watchdog on RAVE SP device.
+
# ALPHA Architecture
# ARM Architecture
diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
index 56adf9fa67d0..5c9556c09f6e 100644
--- a/drivers/watchdog/Makefile
+++ b/drivers/watchdog/Makefile
@@ -223,3 +223,4 @@ obj-$(CONFIG_MAX77620_WATCHDOG) += max77620_wdt.o
obj-$(CONFIG_ZIIRAVE_WATCHDOG) += ziirave_wdt.o
obj-$(CONFIG_SOFT_WATCHDOG) += softdog.o
obj-$(CONFIG_MENF21BMC_WATCHDOG) += menf21bmc_wdt.o
+obj-$(CONFIG_RAVE_SP_WATCHDOG) += rave-sp-wdt.o
diff --git a/drivers/watchdog/rave-sp-wdt.c b/drivers/watchdog/rave-sp-wdt.c
new file mode 100644
index 000000000000..8a2943f7c123
--- /dev/null
+++ b/drivers/watchdog/rave-sp-wdt.c
@@ -0,0 +1,349 @@
+/*
+ * rave-sp-wdt.c - Watchdog driver present in RAVE SP
+ *
+ * Copyright (C) 2017 Zodiac Inflight Innovation
+ *
+ * Driver for parent device can be found in drivers/mfd/rave-sp.c
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of
+ * the License, or (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/delay.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/nvmem-consumer.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/rave-sp.h>
+#include <linux/reboot.h>
+#include <linux/slab.h>
+#include <linux/watchdog.h>
+
+enum {
+ RAVE_SP_RESET_BYTE = 1,
+ RAVE_SP_RESET_REASON_NORMAL = 0,
+ RAVE_SP_RESET_DELAY_MS = 500,
+};
+
+/**
+ * struct rave_sp_wdt_variant - RAVE SP watchdog variant
+ *
+ * @max_timeout: Largest possible watchdog timeout setting
+ * @min_timeout: Smallest possible watchdog timeout setting
+ *
+ * @configure: Function to send configuration command
+ * @restart: Function to send "restart" command
+ */
+struct rave_sp_wdt_variant {
+ unsigned int max_timeout;
+ unsigned int min_timeout;
+
+ int (*configure)(struct watchdog_device *, bool);
+ int (*restart)(struct watchdog_device *);
+};
+
+/**
+ * struct rave_sp_wdt - RAVE SP watchdog
+ *
+ * @wdd: Underlying watchdog device
+ * @sp: Pointer to parent RAVE SP device
+ * @variant: Device specific variant information
+ * @reboot_notifier: Reboot notifier implementing machine reset
+ */
+struct rave_sp_wdt {
+ struct watchdog_device wdd;
+ struct rave_sp *sp;
+ const struct rave_sp_wdt_variant *variant;
+ struct notifier_block reboot_notifier;
+};
+
+static struct rave_sp_wdt *to_rave_sp_wdt(struct watchdog_device *wdd)
+{
+ return container_of(wdd, struct rave_sp_wdt, wdd);
+}
+
+static int rave_sp_wdt_exec(struct watchdog_device *wdd, void *data,
+ size_t data_size)
+{
+ return rave_sp_exec(to_rave_sp_wdt(wdd)->sp,
+ data, data_size, NULL, 0);
+}
+
+static int rave_sp_wdt_legacy_configure(struct watchdog_device *wdd, bool on)
+{
+ u8 cmd[] = {
+ [0] = RAVE_SP_CMD_SW_WDT,
+ [1] = 0,
+ [2] = 0,
+ [3] = on,
+ [4] = on ? wdd->timeout : 0,
+ };
+
+ return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
+}
+
+static int rave_sp_wdt_rdu_configure(struct watchdog_device *wdd, bool on)
+{
+ u8 cmd[] = {
+ [0] = RAVE_SP_CMD_SW_WDT,
+ [1] = 0,
+ [2] = on,
+ [3] = (u8)wdd->timeout,
+ [4] = (u8)(wdd->timeout >> 8),
+ };
+
+ return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
+}
+
+/**
+ * rave_sp_wdt_configure - Configure watchdog device
+ *
+ * @wdd: Device to configure
+ * @on: Desired state of the watchdog timer (ON/OFF)
+ *
+ * This function configures two aspects of the watchdog timer:
+ *
+ * - Wheither it is ON or OFF
+ * - Its timeout duration
+ *
+ * with first aspect specified via function argument and second via
+ * the value of 'wdd->timeout'.
+ */
+static int rave_sp_wdt_configure(struct watchdog_device *wdd, bool on)
+{
+ return to_rave_sp_wdt(wdd)->variant->configure(wdd, on);
+}
+
+static int rave_sp_wdt_legacy_restart(struct watchdog_device *wdd)
+{
+ u8 cmd[] = {
+ [0] = RAVE_SP_CMD_RESET,
+ [1] = 0,
+ [2] = RAVE_SP_RESET_BYTE
+ };
+
+ return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
+}
+
+static int rave_sp_wdt_rdu_restart(struct watchdog_device *wdd)
+{
+ u8 cmd[] = {
+ [0] = RAVE_SP_CMD_RESET,
+ [1] = 0,
+ [2] = RAVE_SP_RESET_BYTE,
+ [3] = RAVE_SP_RESET_REASON_NORMAL
+ };
+
+ return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
+}
+
+static int rave_sp_wdt_reboot_notifier(struct notifier_block *nb,
+ unsigned long action, void *data)
+{
+ /*
+ * Restart handler is called in atomic context which means we
+ * can't communicate to SP via UART. Luckily for use SP will
+ * wait 500ms before actually resetting us, so we ask it to do
+ * so here and let the rest of the system go on wrapping
+ * things up.
+ */
+ if (action == SYS_DOWN || action == SYS_HALT) {
+ struct rave_sp_wdt *sp_wd =
+ container_of(nb, struct rave_sp_wdt, reboot_notifier);
+
+ const int ret = sp_wd->variant->restart(&sp_wd->wdd);
+
+ if (ret < 0)
+ dev_err(sp_wd->wdd.parent,
+ "Failed to issue restart command (%d)", ret);
+ return NOTIFY_OK;
+ }
+
+ return NOTIFY_DONE;
+}
+
+static int rave_sp_wdt_restart(struct watchdog_device *wdd,
+ unsigned long action, void *data)
+{
+ /*
+ * The actual work was done by reboot notifier above. SP
+ * firmware waits 500 ms before issuing reset, so let's hang
+ * here for twice that delay and hopefuly we'd never reach
+ * the return statement.
+ */
+ mdelay(2 * RAVE_SP_RESET_DELAY_MS);
+
+ return -EIO;
+}
+
+static int rave_sp_wdt_start(struct watchdog_device *wdd)
+{
+ int ret;
+
+ ret = rave_sp_wdt_configure(wdd, true);
+ if (!ret)
+ set_bit(WDOG_HW_RUNNING, &wdd->status);
+
+ return ret;
+}
+
+static int rave_sp_wdt_stop(struct watchdog_device *wdd)
+{
+ return rave_sp_wdt_configure(wdd, false);
+}
+
+static int rave_sp_wdt_set_timeout(struct watchdog_device *wdd,
+ unsigned int timeout)
+{
+ wdd->timeout = timeout;
+
+ return rave_sp_wdt_configure(wdd, watchdog_active(wdd));
+}
+
+static int rave_sp_wdt_ping(struct watchdog_device *wdd)
+{
+ u8 cmd[] = {
+ [0] = RAVE_SP_CMD_PET_WDT,
+ [1] = 0,
+ };
+
+ return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
+}
+
+static const struct watchdog_info rave_sp_wdt_info = {
+ .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE,
+ .identity = "RAVE SP Watchdog",
+};
+
+static const struct watchdog_ops rave_sp_wdt_ops = {
+ .owner = THIS_MODULE,
+ .start = rave_sp_wdt_start,
+ .stop = rave_sp_wdt_stop,
+ .ping = rave_sp_wdt_ping,
+ .set_timeout = rave_sp_wdt_set_timeout,
+ .restart = rave_sp_wdt_restart,
+};
+
+static const struct of_device_id rave_sp_wdt_of_match[] = {
+ { .compatible = "zii,rave-sp-watchdog" },
+ {}
+};
+
+static const struct rave_sp_wdt_variant rave_sp_wdt_legacy = {
+ .max_timeout = 255,
+ .min_timeout = 1,
+ .configure = rave_sp_wdt_legacy_configure,
+ .restart = rave_sp_wdt_legacy_restart,
+};
+
+static const struct rave_sp_wdt_variant rave_sp_wdt_rdu = {
+ .max_timeout = 180,
+ .min_timeout = 60,
+ .configure = rave_sp_wdt_rdu_configure,
+ .restart = rave_sp_wdt_rdu_restart,
+};
+
+static const struct of_device_id rave_sp_wdt_variants[] = {
+ { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
+ { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
+ { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
+ { /* sentinel */ }
+};
+
+static int rave_sp_wdt_probe(struct platform_device *pdev)
+{
+ struct device *dev = &pdev->dev;
+ const struct of_device_id *id;
+ struct watchdog_device *wdd;
+ struct rave_sp_wdt *sp_wd;
+ struct nvmem_cell *cell;
+ __le16 timeout = 0;
+ int ret;
+
+ id = of_match_device(rave_sp_wdt_variants, dev->parent);
+ if (!id) {
+ dev_err(dev, "Unknown parent device variant. Bailing out\n");
+ return -ENODEV;
+ }
+
+ sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
+ if (!sp_wd)
+ return -ENOMEM;
+
+ sp_wd->variant = id->data;
+ sp_wd->sp = dev_get_drvdata(dev->parent);
+
+ wdd = &sp_wd->wdd;
+ wdd->parent = dev;
+ wdd->info = &rave_sp_wdt_info;
+ wdd->ops = &rave_sp_wdt_ops;
+ wdd->min_timeout = sp_wd->variant->min_timeout;
+ wdd->max_timeout = sp_wd->variant->max_timeout;
+ wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
+ wdd->timeout = 60;
+
+ cell = nvmem_cell_get(dev, "wdt-timeout");
+ if (!IS_ERR(cell)) {
+ size_t len;
+ void *value = nvmem_cell_read(cell, &len);
+
+ if (!IS_ERR(value)) {
+ memcpy(&timeout, value, min(len, sizeof(timeout)));
+ kfree(value);
+ }
+ nvmem_cell_put(cell);
+ }
+ watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
+ watchdog_set_restart_priority(wdd, 255);
+
+ sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
+ ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
+ if (ret) {
+ dev_err(dev, "Failed to register reboot notifier\n");
+ return ret;
+ }
+
+ /*
+ * We don't know if watchdog is running now. To be sure, let's
+ * start it and depend on watchdog core to ping it
+ */
+ wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
+ ret = rave_sp_wdt_start(wdd);
+ if (ret) {
+ dev_err(dev, "Watchdog didn't start\n");
+ return ret;
+ }
+
+ return devm_watchdog_register_device(dev, wdd);
+}
+
+static struct platform_driver rave_sp_wdt_driver = {
+ .probe = rave_sp_wdt_probe,
+ .driver = {
+ .name = KBUILD_MODNAME,
+ .of_match_table = rave_sp_wdt_of_match,
+ },
+};
+
+module_platform_driver(rave_sp_wdt_driver);
+
+MODULE_DEVICE_TABLE(of, rave_sp_wdt_of_match);
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Andrey Vostrikov <andrey.vostrikov@cogentembedded.com>");
+MODULE_AUTHOR("Nikita Yushchenko <nikita.yoush@cogentembedded.com>");
+MODULE_AUTHOR("Andrey Smirnov <andrew.smirnov@gmail.com>");
+MODULE_DESCRIPTION("RAVE SP Watchdog driver");
+MODULE_ALIAS("platform:rave-sp-watchdog");
--
2.13.6
^ permalink raw reply related [flat|nested] 22+ messages in thread* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-10-31 16:36 ` [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver Andrey Smirnov
@ 2017-10-31 17:21 ` Guenter Roeck
2017-11-04 11:24 ` Greg Kroah-Hartman
2017-11-05 15:47 ` Johan Hovold
2 siblings, 0 replies; 22+ messages in thread
From: Guenter Roeck @ 2017-10-31 17:21 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Rob Herring, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
> This driver provides access to RAVE SP watchdog functionality.
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-watchdog@vger.kernel.org
> Cc: cphealy@gmail.com
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Rob Herring <robh@kernel.org>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> ---
> drivers/watchdog/Kconfig | 7 +
> drivers/watchdog/Makefile | 1 +
> drivers/watchdog/rave-sp-wdt.c | 349 +++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 357 insertions(+)
> create mode 100644 drivers/watchdog/rave-sp-wdt.c
>
> diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
> index c722cbfdc7e6..533a72248cd1 100644
> --- a/drivers/watchdog/Kconfig
> +++ b/drivers/watchdog/Kconfig
> @@ -223,6 +223,13 @@ config ZIIRAVE_WATCHDOG
> To compile this driver as a module, choose M here: the
> module will be called ziirave_wdt.
>
> +config RAVE_SP_WATCHDOG
> + tristate "RAVE SP Watchdog timer"
> + depends on RAVE_SP_CORE
> + select WATCHDOG_CORE
> + help
> + Support for the watchdog on RAVE SP device.
> +
> # ALPHA Architecture
>
> # ARM Architecture
> diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
> index 56adf9fa67d0..5c9556c09f6e 100644
> --- a/drivers/watchdog/Makefile
> +++ b/drivers/watchdog/Makefile
> @@ -223,3 +223,4 @@ obj-$(CONFIG_MAX77620_WATCHDOG) += max77620_wdt.o
> obj-$(CONFIG_ZIIRAVE_WATCHDOG) += ziirave_wdt.o
> obj-$(CONFIG_SOFT_WATCHDOG) += softdog.o
> obj-$(CONFIG_MENF21BMC_WATCHDOG) += menf21bmc_wdt.o
> +obj-$(CONFIG_RAVE_SP_WATCHDOG) += rave-sp-wdt.o
> diff --git a/drivers/watchdog/rave-sp-wdt.c b/drivers/watchdog/rave-sp-wdt.c
> new file mode 100644
> index 000000000000..8a2943f7c123
> --- /dev/null
> +++ b/drivers/watchdog/rave-sp-wdt.c
> @@ -0,0 +1,349 @@
> +/*
> + * rave-sp-wdt.c - Watchdog driver present in RAVE SP
> + *
> + * Copyright (C) 2017 Zodiac Inflight Innovation
> + *
> + * Driver for parent device can be found in drivers/mfd/rave-sp.c
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License as
> + * published by the Free Software Foundation; either version 2 of
> + * the License, or (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/delay.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/nvmem-consumer.h>
> +#include <linux/of_device.h>
> +#include <linux/platform_device.h>
> +#include <linux/rave-sp.h>
> +#include <linux/reboot.h>
> +#include <linux/slab.h>
> +#include <linux/watchdog.h>
> +
> +enum {
> + RAVE_SP_RESET_BYTE = 1,
> + RAVE_SP_RESET_REASON_NORMAL = 0,
> + RAVE_SP_RESET_DELAY_MS = 500,
> +};
> +
> +/**
> + * struct rave_sp_wdt_variant - RAVE SP watchdog variant
> + *
> + * @max_timeout: Largest possible watchdog timeout setting
> + * @min_timeout: Smallest possible watchdog timeout setting
> + *
> + * @configure: Function to send configuration command
> + * @restart: Function to send "restart" command
> + */
> +struct rave_sp_wdt_variant {
> + unsigned int max_timeout;
> + unsigned int min_timeout;
> +
> + int (*configure)(struct watchdog_device *, bool);
> + int (*restart)(struct watchdog_device *);
> +};
> +
> +/**
> + * struct rave_sp_wdt - RAVE SP watchdog
> + *
> + * @wdd: Underlying watchdog device
> + * @sp: Pointer to parent RAVE SP device
> + * @variant: Device specific variant information
> + * @reboot_notifier: Reboot notifier implementing machine reset
> + */
> +struct rave_sp_wdt {
> + struct watchdog_device wdd;
> + struct rave_sp *sp;
> + const struct rave_sp_wdt_variant *variant;
> + struct notifier_block reboot_notifier;
> +};
> +
> +static struct rave_sp_wdt *to_rave_sp_wdt(struct watchdog_device *wdd)
> +{
> + return container_of(wdd, struct rave_sp_wdt, wdd);
> +}
> +
> +static int rave_sp_wdt_exec(struct watchdog_device *wdd, void *data,
> + size_t data_size)
> +{
> + return rave_sp_exec(to_rave_sp_wdt(wdd)->sp,
> + data, data_size, NULL, 0);
> +}
> +
> +static int rave_sp_wdt_legacy_configure(struct watchdog_device *wdd, bool on)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_SW_WDT,
> + [1] = 0,
> + [2] = 0,
> + [3] = on,
> + [4] = on ? wdd->timeout : 0,
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_rdu_configure(struct watchdog_device *wdd, bool on)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_SW_WDT,
> + [1] = 0,
> + [2] = on,
> + [3] = (u8)wdd->timeout,
> + [4] = (u8)(wdd->timeout >> 8),
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +/**
> + * rave_sp_wdt_configure - Configure watchdog device
> + *
> + * @wdd: Device to configure
> + * @on: Desired state of the watchdog timer (ON/OFF)
> + *
> + * This function configures two aspects of the watchdog timer:
> + *
> + * - Wheither it is ON or OFF
> + * - Its timeout duration
> + *
> + * with first aspect specified via function argument and second via
> + * the value of 'wdd->timeout'.
> + */
> +static int rave_sp_wdt_configure(struct watchdog_device *wdd, bool on)
> +{
> + return to_rave_sp_wdt(wdd)->variant->configure(wdd, on);
> +}
> +
> +static int rave_sp_wdt_legacy_restart(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_RESET,
> + [1] = 0,
> + [2] = RAVE_SP_RESET_BYTE
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_rdu_restart(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_RESET,
> + [1] = 0,
> + [2] = RAVE_SP_RESET_BYTE,
> + [3] = RAVE_SP_RESET_REASON_NORMAL
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_reboot_notifier(struct notifier_block *nb,
> + unsigned long action, void *data)
> +{
> + /*
> + * Restart handler is called in atomic context which means we
> + * can't communicate to SP via UART. Luckily for use SP will
> + * wait 500ms before actually resetting us, so we ask it to do
> + * so here and let the rest of the system go on wrapping
> + * things up.
> + */
> + if (action == SYS_DOWN || action == SYS_HALT) {
> + struct rave_sp_wdt *sp_wd =
> + container_of(nb, struct rave_sp_wdt, reboot_notifier);
> +
> + const int ret = sp_wd->variant->restart(&sp_wd->wdd);
> +
> + if (ret < 0)
> + dev_err(sp_wd->wdd.parent,
> + "Failed to issue restart command (%d)", ret);
> + return NOTIFY_OK;
> + }
> +
> + return NOTIFY_DONE;
> +}
> +
> +static int rave_sp_wdt_restart(struct watchdog_device *wdd,
> + unsigned long action, void *data)
> +{
> + /*
> + * The actual work was done by reboot notifier above. SP
> + * firmware waits 500 ms before issuing reset, so let's hang
> + * here for twice that delay and hopefuly we'd never reach
> + * the return statement.
> + */
> + mdelay(2 * RAVE_SP_RESET_DELAY_MS);
> +
> + return -EIO;
> +}
> +
> +static int rave_sp_wdt_start(struct watchdog_device *wdd)
> +{
> + int ret;
> +
> + ret = rave_sp_wdt_configure(wdd, true);
> + if (!ret)
> + set_bit(WDOG_HW_RUNNING, &wdd->status);
> +
> + return ret;
> +}
> +
> +static int rave_sp_wdt_stop(struct watchdog_device *wdd)
> +{
> + return rave_sp_wdt_configure(wdd, false);
> +}
> +
> +static int rave_sp_wdt_set_timeout(struct watchdog_device *wdd,
> + unsigned int timeout)
> +{
> + wdd->timeout = timeout;
> +
> + return rave_sp_wdt_configure(wdd, watchdog_active(wdd));
> +}
> +
> +static int rave_sp_wdt_ping(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_PET_WDT,
> + [1] = 0,
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static const struct watchdog_info rave_sp_wdt_info = {
> + .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE,
> + .identity = "RAVE SP Watchdog",
> +};
> +
> +static const struct watchdog_ops rave_sp_wdt_ops = {
> + .owner = THIS_MODULE,
> + .start = rave_sp_wdt_start,
> + .stop = rave_sp_wdt_stop,
> + .ping = rave_sp_wdt_ping,
> + .set_timeout = rave_sp_wdt_set_timeout,
> + .restart = rave_sp_wdt_restart,
> +};
> +
> +static const struct of_device_id rave_sp_wdt_of_match[] = {
> + { .compatible = "zii,rave-sp-watchdog" },
> + {}
> +};
> +
> +static const struct rave_sp_wdt_variant rave_sp_wdt_legacy = {
> + .max_timeout = 255,
> + .min_timeout = 1,
> + .configure = rave_sp_wdt_legacy_configure,
> + .restart = rave_sp_wdt_legacy_restart,
> +};
> +
> +static const struct rave_sp_wdt_variant rave_sp_wdt_rdu = {
> + .max_timeout = 180,
> + .min_timeout = 60,
> + .configure = rave_sp_wdt_rdu_configure,
> + .restart = rave_sp_wdt_rdu_restart,
> +};
> +
> +static const struct of_device_id rave_sp_wdt_variants[] = {
> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
> + { /* sentinel */ }
> +};
> +
> +static int rave_sp_wdt_probe(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + const struct of_device_id *id;
> + struct watchdog_device *wdd;
> + struct rave_sp_wdt *sp_wd;
> + struct nvmem_cell *cell;
> + __le16 timeout = 0;
> + int ret;
> +
> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
> + if (!id) {
> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
> + return -ENODEV;
> + }
> +
> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
> + if (!sp_wd)
> + return -ENOMEM;
> +
> + sp_wd->variant = id->data;
> + sp_wd->sp = dev_get_drvdata(dev->parent);
> +
> + wdd = &sp_wd->wdd;
> + wdd->parent = dev;
> + wdd->info = &rave_sp_wdt_info;
> + wdd->ops = &rave_sp_wdt_ops;
> + wdd->min_timeout = sp_wd->variant->min_timeout;
> + wdd->max_timeout = sp_wd->variant->max_timeout;
> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
> + wdd->timeout = 60;
> +
> + cell = nvmem_cell_get(dev, "wdt-timeout");
> + if (!IS_ERR(cell)) {
> + size_t len;
> + void *value = nvmem_cell_read(cell, &len);
> +
> + if (!IS_ERR(value)) {
> + memcpy(&timeout, value, min(len, sizeof(timeout)));
> + kfree(value);
> + }
> + nvmem_cell_put(cell);
> + }
> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
> + watchdog_set_restart_priority(wdd, 255);
> +
> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
> + if (ret) {
> + dev_err(dev, "Failed to register reboot notifier\n");
> + return ret;
> + }
> +
> + /*
> + * We don't know if watchdog is running now. To be sure, let's
> + * start it and depend on watchdog core to ping it
> + */
> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
> + ret = rave_sp_wdt_start(wdd);
> + if (ret) {
> + dev_err(dev, "Watchdog didn't start\n");
> + return ret;
> + }
> +
> + return devm_watchdog_register_device(dev, wdd);
> +}
> +
> +static struct platform_driver rave_sp_wdt_driver = {
> + .probe = rave_sp_wdt_probe,
> + .driver = {
> + .name = KBUILD_MODNAME,
> + .of_match_table = rave_sp_wdt_of_match,
> + },
> +};
> +
> +module_platform_driver(rave_sp_wdt_driver);
> +
> +MODULE_DEVICE_TABLE(of, rave_sp_wdt_of_match);
> +MODULE_LICENSE("GPL");
> +MODULE_AUTHOR("Andrey Vostrikov <andrey.vostrikov@cogentembedded.com>");
> +MODULE_AUTHOR("Nikita Yushchenko <nikita.yoush@cogentembedded.com>");
> +MODULE_AUTHOR("Andrey Smirnov <andrew.smirnov@gmail.com>");
> +MODULE_DESCRIPTION("RAVE SP Watchdog driver");
> +MODULE_ALIAS("platform:rave-sp-watchdog");
> --
> 2.13.6
>
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-10-31 16:36 ` [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver Andrey Smirnov
2017-10-31 17:21 ` Guenter Roeck
@ 2017-11-04 11:24 ` Greg Kroah-Hartman
2017-11-05 15:55 ` Guenter Roeck
2017-11-05 15:47 ` Johan Hovold
2 siblings, 1 reply; 22+ messages in thread
From: Greg Kroah-Hartman @ 2017-11-04 11:24 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Pavel Machek, Andy Shevchenko,
Guenter Roeck, Rob Herring, Johan Hovold, Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
> This driver provides access to RAVE SP watchdog functionality.
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-watchdog@vger.kernel.org
> Cc: cphealy@gmail.com
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Rob Herring <robh@kernel.org>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
> drivers/watchdog/Kconfig | 7 +
> drivers/watchdog/Makefile | 1 +
> drivers/watchdog/rave-sp-wdt.c | 349 +++++++++++++++++++++++++++++++++++++++++
> 3 files changed, 357 insertions(+)
> create mode 100644 drivers/watchdog/rave-sp-wdt.c
>
> diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
> index c722cbfdc7e6..533a72248cd1 100644
> --- a/drivers/watchdog/Kconfig
> +++ b/drivers/watchdog/Kconfig
> @@ -223,6 +223,13 @@ config ZIIRAVE_WATCHDOG
> To compile this driver as a module, choose M here: the
> module will be called ziirave_wdt.
>
> +config RAVE_SP_WATCHDOG
> + tristate "RAVE SP Watchdog timer"
> + depends on RAVE_SP_CORE
> + select WATCHDOG_CORE
> + help
> + Support for the watchdog on RAVE SP device.
> +
> # ALPHA Architecture
>
> # ARM Architecture
> diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
> index 56adf9fa67d0..5c9556c09f6e 100644
> --- a/drivers/watchdog/Makefile
> +++ b/drivers/watchdog/Makefile
> @@ -223,3 +223,4 @@ obj-$(CONFIG_MAX77620_WATCHDOG) += max77620_wdt.o
> obj-$(CONFIG_ZIIRAVE_WATCHDOG) += ziirave_wdt.o
> obj-$(CONFIG_SOFT_WATCHDOG) += softdog.o
> obj-$(CONFIG_MENF21BMC_WATCHDOG) += menf21bmc_wdt.o
> +obj-$(CONFIG_RAVE_SP_WATCHDOG) += rave-sp-wdt.o
> diff --git a/drivers/watchdog/rave-sp-wdt.c b/drivers/watchdog/rave-sp-wdt.c
> new file mode 100644
> index 000000000000..8a2943f7c123
> --- /dev/null
> +++ b/drivers/watchdog/rave-sp-wdt.c
> @@ -0,0 +1,349 @@
> +/*
> + * rave-sp-wdt.c - Watchdog driver present in RAVE SP
> + *
> + * Copyright (C) 2017 Zodiac Inflight Innovation
> + *
> + * Driver for parent device can be found in drivers/mfd/rave-sp.c
> + *
> + * This program is free software; you can redistribute it and/or
> + * modify it under the terms of the GNU General Public License as
> + * published by the Free Software Foundation; either version 2 of
> + * the License, or (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/delay.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/nvmem-consumer.h>
> +#include <linux/of_device.h>
> +#include <linux/platform_device.h>
> +#include <linux/rave-sp.h>
> +#include <linux/reboot.h>
> +#include <linux/slab.h>
> +#include <linux/watchdog.h>
> +
> +enum {
> + RAVE_SP_RESET_BYTE = 1,
> + RAVE_SP_RESET_REASON_NORMAL = 0,
> + RAVE_SP_RESET_DELAY_MS = 500,
> +};
> +
> +/**
> + * struct rave_sp_wdt_variant - RAVE SP watchdog variant
> + *
> + * @max_timeout: Largest possible watchdog timeout setting
> + * @min_timeout: Smallest possible watchdog timeout setting
> + *
> + * @configure: Function to send configuration command
> + * @restart: Function to send "restart" command
> + */
> +struct rave_sp_wdt_variant {
> + unsigned int max_timeout;
> + unsigned int min_timeout;
> +
> + int (*configure)(struct watchdog_device *, bool);
> + int (*restart)(struct watchdog_device *);
> +};
> +
> +/**
> + * struct rave_sp_wdt - RAVE SP watchdog
> + *
> + * @wdd: Underlying watchdog device
> + * @sp: Pointer to parent RAVE SP device
> + * @variant: Device specific variant information
> + * @reboot_notifier: Reboot notifier implementing machine reset
> + */
> +struct rave_sp_wdt {
> + struct watchdog_device wdd;
> + struct rave_sp *sp;
> + const struct rave_sp_wdt_variant *variant;
> + struct notifier_block reboot_notifier;
> +};
> +
> +static struct rave_sp_wdt *to_rave_sp_wdt(struct watchdog_device *wdd)
> +{
> + return container_of(wdd, struct rave_sp_wdt, wdd);
> +}
> +
> +static int rave_sp_wdt_exec(struct watchdog_device *wdd, void *data,
> + size_t data_size)
> +{
> + return rave_sp_exec(to_rave_sp_wdt(wdd)->sp,
> + data, data_size, NULL, 0);
> +}
> +
> +static int rave_sp_wdt_legacy_configure(struct watchdog_device *wdd, bool on)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_SW_WDT,
> + [1] = 0,
> + [2] = 0,
> + [3] = on,
> + [4] = on ? wdd->timeout : 0,
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_rdu_configure(struct watchdog_device *wdd, bool on)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_SW_WDT,
> + [1] = 0,
> + [2] = on,
> + [3] = (u8)wdd->timeout,
> + [4] = (u8)(wdd->timeout >> 8),
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +/**
> + * rave_sp_wdt_configure - Configure watchdog device
> + *
> + * @wdd: Device to configure
> + * @on: Desired state of the watchdog timer (ON/OFF)
> + *
> + * This function configures two aspects of the watchdog timer:
> + *
> + * - Wheither it is ON or OFF
> + * - Its timeout duration
> + *
> + * with first aspect specified via function argument and second via
> + * the value of 'wdd->timeout'.
> + */
> +static int rave_sp_wdt_configure(struct watchdog_device *wdd, bool on)
> +{
> + return to_rave_sp_wdt(wdd)->variant->configure(wdd, on);
> +}
> +
> +static int rave_sp_wdt_legacy_restart(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_RESET,
> + [1] = 0,
> + [2] = RAVE_SP_RESET_BYTE
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_rdu_restart(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_RESET,
> + [1] = 0,
> + [2] = RAVE_SP_RESET_BYTE,
> + [3] = RAVE_SP_RESET_REASON_NORMAL
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static int rave_sp_wdt_reboot_notifier(struct notifier_block *nb,
> + unsigned long action, void *data)
> +{
> + /*
> + * Restart handler is called in atomic context which means we
> + * can't communicate to SP via UART. Luckily for use SP will
> + * wait 500ms before actually resetting us, so we ask it to do
> + * so here and let the rest of the system go on wrapping
> + * things up.
> + */
> + if (action == SYS_DOWN || action == SYS_HALT) {
> + struct rave_sp_wdt *sp_wd =
> + container_of(nb, struct rave_sp_wdt, reboot_notifier);
> +
> + const int ret = sp_wd->variant->restart(&sp_wd->wdd);
> +
> + if (ret < 0)
> + dev_err(sp_wd->wdd.parent,
> + "Failed to issue restart command (%d)", ret);
> + return NOTIFY_OK;
> + }
> +
> + return NOTIFY_DONE;
> +}
> +
> +static int rave_sp_wdt_restart(struct watchdog_device *wdd,
> + unsigned long action, void *data)
> +{
> + /*
> + * The actual work was done by reboot notifier above. SP
> + * firmware waits 500 ms before issuing reset, so let's hang
> + * here for twice that delay and hopefuly we'd never reach
> + * the return statement.
> + */
> + mdelay(2 * RAVE_SP_RESET_DELAY_MS);
> +
> + return -EIO;
> +}
> +
> +static int rave_sp_wdt_start(struct watchdog_device *wdd)
> +{
> + int ret;
> +
> + ret = rave_sp_wdt_configure(wdd, true);
> + if (!ret)
> + set_bit(WDOG_HW_RUNNING, &wdd->status);
> +
> + return ret;
> +}
> +
> +static int rave_sp_wdt_stop(struct watchdog_device *wdd)
> +{
> + return rave_sp_wdt_configure(wdd, false);
> +}
> +
> +static int rave_sp_wdt_set_timeout(struct watchdog_device *wdd,
> + unsigned int timeout)
> +{
> + wdd->timeout = timeout;
> +
> + return rave_sp_wdt_configure(wdd, watchdog_active(wdd));
> +}
> +
> +static int rave_sp_wdt_ping(struct watchdog_device *wdd)
> +{
> + u8 cmd[] = {
> + [0] = RAVE_SP_CMD_PET_WDT,
> + [1] = 0,
> + };
> +
> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
> +}
> +
> +static const struct watchdog_info rave_sp_wdt_info = {
> + .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE,
> + .identity = "RAVE SP Watchdog",
> +};
> +
> +static const struct watchdog_ops rave_sp_wdt_ops = {
> + .owner = THIS_MODULE,
> + .start = rave_sp_wdt_start,
> + .stop = rave_sp_wdt_stop,
> + .ping = rave_sp_wdt_ping,
> + .set_timeout = rave_sp_wdt_set_timeout,
> + .restart = rave_sp_wdt_restart,
> +};
> +
> +static const struct of_device_id rave_sp_wdt_of_match[] = {
> + { .compatible = "zii,rave-sp-watchdog" },
> + {}
> +};
> +
> +static const struct rave_sp_wdt_variant rave_sp_wdt_legacy = {
> + .max_timeout = 255,
> + .min_timeout = 1,
> + .configure = rave_sp_wdt_legacy_configure,
> + .restart = rave_sp_wdt_legacy_restart,
> +};
> +
> +static const struct rave_sp_wdt_variant rave_sp_wdt_rdu = {
> + .max_timeout = 180,
> + .min_timeout = 60,
> + .configure = rave_sp_wdt_rdu_configure,
> + .restart = rave_sp_wdt_rdu_restart,
> +};
> +
> +static const struct of_device_id rave_sp_wdt_variants[] = {
> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
> + { /* sentinel */ }
> +};
> +
> +static int rave_sp_wdt_probe(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + const struct of_device_id *id;
> + struct watchdog_device *wdd;
> + struct rave_sp_wdt *sp_wd;
> + struct nvmem_cell *cell;
> + __le16 timeout = 0;
> + int ret;
> +
> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
> + if (!id) {
> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
> + return -ENODEV;
> + }
> +
> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
> + if (!sp_wd)
> + return -ENOMEM;
> +
> + sp_wd->variant = id->data;
> + sp_wd->sp = dev_get_drvdata(dev->parent);
> +
> + wdd = &sp_wd->wdd;
> + wdd->parent = dev;
> + wdd->info = &rave_sp_wdt_info;
> + wdd->ops = &rave_sp_wdt_ops;
> + wdd->min_timeout = sp_wd->variant->min_timeout;
> + wdd->max_timeout = sp_wd->variant->max_timeout;
> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
> + wdd->timeout = 60;
> +
> + cell = nvmem_cell_get(dev, "wdt-timeout");
> + if (!IS_ERR(cell)) {
> + size_t len;
> + void *value = nvmem_cell_read(cell, &len);
> +
> + if (!IS_ERR(value)) {
> + memcpy(&timeout, value, min(len, sizeof(timeout)));
> + kfree(value);
> + }
> + nvmem_cell_put(cell);
> + }
> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
> + watchdog_set_restart_priority(wdd, 255);
> +
> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
> + if (ret) {
> + dev_err(dev, "Failed to register reboot notifier\n");
> + return ret;
> + }
> +
> + /*
> + * We don't know if watchdog is running now. To be sure, let's
> + * start it and depend on watchdog core to ping it
> + */
> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
> + ret = rave_sp_wdt_start(wdd);
> + if (ret) {
> + dev_err(dev, "Watchdog didn't start\n");
> + return ret;
> + }
> +
> + return devm_watchdog_register_device(dev, wdd);
> +}
> +
> +static struct platform_driver rave_sp_wdt_driver = {
> + .probe = rave_sp_wdt_probe,
No remove? :)
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-11-04 11:24 ` Greg Kroah-Hartman
@ 2017-11-05 15:55 ` Guenter Roeck
0 siblings, 0 replies; 22+ messages in thread
From: Guenter Roeck @ 2017-11-05 15:55 UTC (permalink / raw)
To: Greg Kroah-Hartman, Andrey Smirnov
Cc: linux-kernel, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Pavel Machek, Andy Shevchenko,
Rob Herring, Johan Hovold, Sebastian Reichel
On 11/04/2017 04:24 AM, Greg Kroah-Hartman wrote:
> On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
>> This driver provides access to RAVE SP watchdog functionality.
>>
>> Cc: linux-kernel@vger.kernel.org
>> Cc: linux-watchdog@vger.kernel.org
>> Cc: cphealy@gmail.com
>> Cc: Lucas Stach <l.stach@pengutronix.de>
>> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Cc: Lee Jones <lee.jones@linaro.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Cc: Guenter Roeck <linux@roeck-us.net>
>> Cc: Rob Herring <robh@kernel.org>
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>> ---
>> drivers/watchdog/Kconfig | 7 +
>> drivers/watchdog/Makefile | 1 +
>> drivers/watchdog/rave-sp-wdt.c | 349 +++++++++++++++++++++++++++++++++++++++++
>> 3 files changed, 357 insertions(+)
>> create mode 100644 drivers/watchdog/rave-sp-wdt.c
>>
>> diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
>> index c722cbfdc7e6..533a72248cd1 100644
>> --- a/drivers/watchdog/Kconfig
>> +++ b/drivers/watchdog/Kconfig
>> @@ -223,6 +223,13 @@ config ZIIRAVE_WATCHDOG
>> To compile this driver as a module, choose M here: the
>> module will be called ziirave_wdt.
>>
>> +config RAVE_SP_WATCHDOG
>> + tristate "RAVE SP Watchdog timer"
>> + depends on RAVE_SP_CORE
>> + select WATCHDOG_CORE
>> + help
>> + Support for the watchdog on RAVE SP device.
>> +
>> # ALPHA Architecture
>>
>> # ARM Architecture
>> diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
>> index 56adf9fa67d0..5c9556c09f6e 100644
>> --- a/drivers/watchdog/Makefile
>> +++ b/drivers/watchdog/Makefile
>> @@ -223,3 +223,4 @@ obj-$(CONFIG_MAX77620_WATCHDOG) += max77620_wdt.o
>> obj-$(CONFIG_ZIIRAVE_WATCHDOG) += ziirave_wdt.o
>> obj-$(CONFIG_SOFT_WATCHDOG) += softdog.o
>> obj-$(CONFIG_MENF21BMC_WATCHDOG) += menf21bmc_wdt.o
>> +obj-$(CONFIG_RAVE_SP_WATCHDOG) += rave-sp-wdt.o
>> diff --git a/drivers/watchdog/rave-sp-wdt.c b/drivers/watchdog/rave-sp-wdt.c
>> new file mode 100644
>> index 000000000000..8a2943f7c123
>> --- /dev/null
>> +++ b/drivers/watchdog/rave-sp-wdt.c
>> @@ -0,0 +1,349 @@
>> +/*
>> + * rave-sp-wdt.c - Watchdog driver present in RAVE SP
>> + *
>> + * Copyright (C) 2017 Zodiac Inflight Innovation
>> + *
>> + * Driver for parent device can be found in drivers/mfd/rave-sp.c
>> + *
>> + * This program is free software; you can redistribute it and/or
>> + * modify it under the terms of the GNU General Public License as
>> + * published by the Free Software Foundation; either version 2 of
>> + * the License, or (at your option) any later version.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program; if not, see <http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include <linux/delay.h>
>> +#include <linux/kernel.h>
>> +#include <linux/module.h>
>> +#include <linux/nvmem-consumer.h>
>> +#include <linux/of_device.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/rave-sp.h>
>> +#include <linux/reboot.h>
>> +#include <linux/slab.h>
>> +#include <linux/watchdog.h>
>> +
>> +enum {
>> + RAVE_SP_RESET_BYTE = 1,
>> + RAVE_SP_RESET_REASON_NORMAL = 0,
>> + RAVE_SP_RESET_DELAY_MS = 500,
>> +};
>> +
>> +/**
>> + * struct rave_sp_wdt_variant - RAVE SP watchdog variant
>> + *
>> + * @max_timeout: Largest possible watchdog timeout setting
>> + * @min_timeout: Smallest possible watchdog timeout setting
>> + *
>> + * @configure: Function to send configuration command
>> + * @restart: Function to send "restart" command
>> + */
>> +struct rave_sp_wdt_variant {
>> + unsigned int max_timeout;
>> + unsigned int min_timeout;
>> +
>> + int (*configure)(struct watchdog_device *, bool);
>> + int (*restart)(struct watchdog_device *);
>> +};
>> +
>> +/**
>> + * struct rave_sp_wdt - RAVE SP watchdog
>> + *
>> + * @wdd: Underlying watchdog device
>> + * @sp: Pointer to parent RAVE SP device
>> + * @variant: Device specific variant information
>> + * @reboot_notifier: Reboot notifier implementing machine reset
>> + */
>> +struct rave_sp_wdt {
>> + struct watchdog_device wdd;
>> + struct rave_sp *sp;
>> + const struct rave_sp_wdt_variant *variant;
>> + struct notifier_block reboot_notifier;
>> +};
>> +
>> +static struct rave_sp_wdt *to_rave_sp_wdt(struct watchdog_device *wdd)
>> +{
>> + return container_of(wdd, struct rave_sp_wdt, wdd);
>> +}
>> +
>> +static int rave_sp_wdt_exec(struct watchdog_device *wdd, void *data,
>> + size_t data_size)
>> +{
>> + return rave_sp_exec(to_rave_sp_wdt(wdd)->sp,
>> + data, data_size, NULL, 0);
>> +}
>> +
>> +static int rave_sp_wdt_legacy_configure(struct watchdog_device *wdd, bool on)
>> +{
>> + u8 cmd[] = {
>> + [0] = RAVE_SP_CMD_SW_WDT,
>> + [1] = 0,
>> + [2] = 0,
>> + [3] = on,
>> + [4] = on ? wdd->timeout : 0,
>> + };
>> +
>> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static int rave_sp_wdt_rdu_configure(struct watchdog_device *wdd, bool on)
>> +{
>> + u8 cmd[] = {
>> + [0] = RAVE_SP_CMD_SW_WDT,
>> + [1] = 0,
>> + [2] = on,
>> + [3] = (u8)wdd->timeout,
>> + [4] = (u8)(wdd->timeout >> 8),
>> + };
>> +
>> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +/**
>> + * rave_sp_wdt_configure - Configure watchdog device
>> + *
>> + * @wdd: Device to configure
>> + * @on: Desired state of the watchdog timer (ON/OFF)
>> + *
>> + * This function configures two aspects of the watchdog timer:
>> + *
>> + * - Wheither it is ON or OFF
>> + * - Its timeout duration
>> + *
>> + * with first aspect specified via function argument and second via
>> + * the value of 'wdd->timeout'.
>> + */
>> +static int rave_sp_wdt_configure(struct watchdog_device *wdd, bool on)
>> +{
>> + return to_rave_sp_wdt(wdd)->variant->configure(wdd, on);
>> +}
>> +
>> +static int rave_sp_wdt_legacy_restart(struct watchdog_device *wdd)
>> +{
>> + u8 cmd[] = {
>> + [0] = RAVE_SP_CMD_RESET,
>> + [1] = 0,
>> + [2] = RAVE_SP_RESET_BYTE
>> + };
>> +
>> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static int rave_sp_wdt_rdu_restart(struct watchdog_device *wdd)
>> +{
>> + u8 cmd[] = {
>> + [0] = RAVE_SP_CMD_RESET,
>> + [1] = 0,
>> + [2] = RAVE_SP_RESET_BYTE,
>> + [3] = RAVE_SP_RESET_REASON_NORMAL
>> + };
>> +
>> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static int rave_sp_wdt_reboot_notifier(struct notifier_block *nb,
>> + unsigned long action, void *data)
>> +{
>> + /*
>> + * Restart handler is called in atomic context which means we
>> + * can't communicate to SP via UART. Luckily for use SP will
>> + * wait 500ms before actually resetting us, so we ask it to do
>> + * so here and let the rest of the system go on wrapping
>> + * things up.
>> + */
>> + if (action == SYS_DOWN || action == SYS_HALT) {
>> + struct rave_sp_wdt *sp_wd =
>> + container_of(nb, struct rave_sp_wdt, reboot_notifier);
>> +
>> + const int ret = sp_wd->variant->restart(&sp_wd->wdd);
>> +
>> + if (ret < 0)
>> + dev_err(sp_wd->wdd.parent,
>> + "Failed to issue restart command (%d)", ret);
>> + return NOTIFY_OK;
>> + }
>> +
>> + return NOTIFY_DONE;
>> +}
>> +
>> +static int rave_sp_wdt_restart(struct watchdog_device *wdd,
>> + unsigned long action, void *data)
>> +{
>> + /*
>> + * The actual work was done by reboot notifier above. SP
>> + * firmware waits 500 ms before issuing reset, so let's hang
>> + * here for twice that delay and hopefuly we'd never reach
>> + * the return statement.
>> + */
>> + mdelay(2 * RAVE_SP_RESET_DELAY_MS);
>> +
>> + return -EIO;
>> +}
>> +
>> +static int rave_sp_wdt_start(struct watchdog_device *wdd)
>> +{
>> + int ret;
>> +
>> + ret = rave_sp_wdt_configure(wdd, true);
>> + if (!ret)
>> + set_bit(WDOG_HW_RUNNING, &wdd->status);
>> +
>> + return ret;
>> +}
>> +
>> +static int rave_sp_wdt_stop(struct watchdog_device *wdd)
>> +{
>> + return rave_sp_wdt_configure(wdd, false);
>> +}
>> +
>> +static int rave_sp_wdt_set_timeout(struct watchdog_device *wdd,
>> + unsigned int timeout)
>> +{
>> + wdd->timeout = timeout;
>> +
>> + return rave_sp_wdt_configure(wdd, watchdog_active(wdd));
>> +}
>> +
>> +static int rave_sp_wdt_ping(struct watchdog_device *wdd)
>> +{
>> + u8 cmd[] = {
>> + [0] = RAVE_SP_CMD_PET_WDT,
>> + [1] = 0,
>> + };
>> +
>> + return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static const struct watchdog_info rave_sp_wdt_info = {
>> + .options = WDIOF_SETTIMEOUT | WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE,
>> + .identity = "RAVE SP Watchdog",
>> +};
>> +
>> +static const struct watchdog_ops rave_sp_wdt_ops = {
>> + .owner = THIS_MODULE,
>> + .start = rave_sp_wdt_start,
>> + .stop = rave_sp_wdt_stop,
>> + .ping = rave_sp_wdt_ping,
>> + .set_timeout = rave_sp_wdt_set_timeout,
>> + .restart = rave_sp_wdt_restart,
>> +};
>> +
>> +static const struct of_device_id rave_sp_wdt_of_match[] = {
>> + { .compatible = "zii,rave-sp-watchdog" },
>> + {}
>> +};
>> +
>> +static const struct rave_sp_wdt_variant rave_sp_wdt_legacy = {
>> + .max_timeout = 255,
>> + .min_timeout = 1,
>> + .configure = rave_sp_wdt_legacy_configure,
>> + .restart = rave_sp_wdt_legacy_restart,
>> +};
>> +
>> +static const struct rave_sp_wdt_variant rave_sp_wdt_rdu = {
>> + .max_timeout = 180,
>> + .min_timeout = 60,
>> + .configure = rave_sp_wdt_rdu_configure,
>> + .restart = rave_sp_wdt_rdu_restart,
>> +};
>> +
>> +static const struct of_device_id rave_sp_wdt_variants[] = {
>> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
>> + { /* sentinel */ }
>> +};
>> +
>> +static int rave_sp_wdt_probe(struct platform_device *pdev)
>> +{
>> + struct device *dev = &pdev->dev;
>> + const struct of_device_id *id;
>> + struct watchdog_device *wdd;
>> + struct rave_sp_wdt *sp_wd;
>> + struct nvmem_cell *cell;
>> + __le16 timeout = 0;
>> + int ret;
>> +
>> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
>> + if (!id) {
>> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
>> + return -ENODEV;
>> + }
>> +
>> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
>> + if (!sp_wd)
>> + return -ENOMEM;
>> +
>> + sp_wd->variant = id->data;
>> + sp_wd->sp = dev_get_drvdata(dev->parent);
>> +
>> + wdd = &sp_wd->wdd;
>> + wdd->parent = dev;
>> + wdd->info = &rave_sp_wdt_info;
>> + wdd->ops = &rave_sp_wdt_ops;
>> + wdd->min_timeout = sp_wd->variant->min_timeout;
>> + wdd->max_timeout = sp_wd->variant->max_timeout;
>> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
>> + wdd->timeout = 60;
>> +
>> + cell = nvmem_cell_get(dev, "wdt-timeout");
>> + if (!IS_ERR(cell)) {
>> + size_t len;
>> + void *value = nvmem_cell_read(cell, &len);
>> +
>> + if (!IS_ERR(value)) {
>> + memcpy(&timeout, value, min(len, sizeof(timeout)));
>> + kfree(value);
>> + }
>> + nvmem_cell_put(cell);
>> + }
>> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
>> + watchdog_set_restart_priority(wdd, 255);
>> +
>> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
>> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
>> + if (ret) {
>> + dev_err(dev, "Failed to register reboot notifier\n");
>> + return ret;
>> + }
>> +
>> + /*
>> + * We don't know if watchdog is running now. To be sure, let's
>> + * start it and depend on watchdog core to ping it
>> + */
>> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
>> + ret = rave_sp_wdt_start(wdd);
>> + if (ret) {
>> + dev_err(dev, "Watchdog didn't start\n");
>> + return ret;
>> + }
>> +
>> + return devm_watchdog_register_device(dev, wdd);
>> +}
>> +
>> +static struct platform_driver rave_sp_wdt_driver = {
>> + .probe = rave_sp_wdt_probe,
>
> No remove? :)
>
It would be an empty function.
Guenter
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-10-31 16:36 ` [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver Andrey Smirnov
2017-10-31 17:21 ` Guenter Roeck
2017-11-04 11:24 ` Greg Kroah-Hartman
@ 2017-11-05 15:47 ` Johan Hovold
2017-11-05 18:34 ` Guenter Roeck
2017-11-05 21:59 ` Andrey Smirnov
2 siblings, 2 replies; 22+ messages in thread
From: Johan Hovold @ 2017-11-05 15:47 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Guenter Roeck, Rob Herring, Johan Hovold,
Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
> This driver provides access to RAVE SP watchdog functionality.
>
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-watchdog@vger.kernel.org
> Cc: cphealy@gmail.com
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Rob Herring <robh@kernel.org>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> +static const struct of_device_id rave_sp_wdt_of_match[] = {
> + { .compatible = "zii,rave-sp-watchdog" },
> + {}
> +};
> +static const struct of_device_id rave_sp_wdt_variants[] = {
> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
> + { /* sentinel */ }
> +};
> +
> +static int rave_sp_wdt_probe(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + const struct of_device_id *id;
> + struct watchdog_device *wdd;
> + struct rave_sp_wdt *sp_wd;
> + struct nvmem_cell *cell;
> + __le16 timeout = 0;
> + int ret;
> +
> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
So as I already mentioned in a comment on an earlier version of the MFD
driver, I find this matching on the parent of_node to be an odd
pattern, which also does not seem to have any precedent in mainline.
It seems you really should be using two compatible strings for the
watchdog (for the legacy and rdu variants) rather than keep an entry for
every parent compatible-string in every child/cell driver (which all
need to be kept in sync).
But let's see what Rob and Lee says about this.
> + if (!id) {
> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
> + return -ENODEV;
> + }
> +
> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
> + if (!sp_wd)
> + return -ENOMEM;
> +
> + sp_wd->variant = id->data;
> + sp_wd->sp = dev_get_drvdata(dev->parent);
> +
> + wdd = &sp_wd->wdd;
> + wdd->parent = dev;
> + wdd->info = &rave_sp_wdt_info;
> + wdd->ops = &rave_sp_wdt_ops;
> + wdd->min_timeout = sp_wd->variant->min_timeout;
> + wdd->max_timeout = sp_wd->variant->max_timeout;
> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
> + wdd->timeout = 60;
> +
> + cell = nvmem_cell_get(dev, "wdt-timeout");
> + if (!IS_ERR(cell)) {
> + size_t len;
> + void *value = nvmem_cell_read(cell, &len);
> +
> + if (!IS_ERR(value)) {
> + memcpy(&timeout, value, min(len, sizeof(timeout)));
> + kfree(value);
> + }
> + nvmem_cell_put(cell);
> + }
> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
> + watchdog_set_restart_priority(wdd, 255);
> +
> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
> + if (ret) {
> + dev_err(dev, "Failed to register reboot notifier\n");
> + return ret;
> + }
> +
> + /*
> + * We don't know if watchdog is running now. To be sure, let's
> + * start it and depend on watchdog core to ping it
> + */
> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
> + ret = rave_sp_wdt_start(wdd);
> + if (ret) {
> + dev_err(dev, "Watchdog didn't start\n");
> + return ret;
> + }
> +
> + return devm_watchdog_register_device(dev, wdd);
What about stopping the watchdog on errors here?
And does watchdog core take care of calling stop() on unregister (i.e.
at unbind)?
> +}
> +
> +static struct platform_driver rave_sp_wdt_driver = {
> + .probe = rave_sp_wdt_probe,
> + .driver = {
> + .name = KBUILD_MODNAME,
> + .of_match_table = rave_sp_wdt_of_match,
> + },
> +};
Johan
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-11-05 15:47 ` Johan Hovold
@ 2017-11-05 18:34 ` Guenter Roeck
2017-11-05 21:59 ` Andrey Smirnov
1 sibling, 0 replies; 22+ messages in thread
From: Guenter Roeck @ 2017-11-05 18:34 UTC (permalink / raw)
To: Johan Hovold, Andrey Smirnov
Cc: linux-kernel, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Rob Herring, Sebastian Reichel
On 11/05/2017 07:47 AM, Johan Hovold wrote:
> On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
>> This driver provides access to RAVE SP watchdog functionality.
>>
>> Cc: linux-kernel@vger.kernel.org
>> Cc: linux-watchdog@vger.kernel.org
>> Cc: cphealy@gmail.com
>> Cc: Lucas Stach <l.stach@pengutronix.de>
>> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Cc: Lee Jones <lee.jones@linaro.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Cc: Guenter Roeck <linux@roeck-us.net>
>> Cc: Rob Herring <robh@kernel.org>
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>
>> +static const struct of_device_id rave_sp_wdt_of_match[] = {
>> + { .compatible = "zii,rave-sp-watchdog" },
>> + {}
>> +};
>
>> +static const struct of_device_id rave_sp_wdt_variants[] = {
>> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
>> + { /* sentinel */ }
>> +};
>> +
>> +static int rave_sp_wdt_probe(struct platform_device *pdev)
>> +{
>> + struct device *dev = &pdev->dev;
>> + const struct of_device_id *id;
>> + struct watchdog_device *wdd;
>> + struct rave_sp_wdt *sp_wd;
>> + struct nvmem_cell *cell;
>> + __le16 timeout = 0;
>> + int ret;
>> +
>> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
>
> So as I already mentioned in a comment on an earlier version of the MFD
> driver, I find this matching on the parent of_node to be an odd
> pattern, which also does not seem to have any precedent in mainline.
>
> It seems you really should be using two compatible strings for the
> watchdog (for the legacy and rdu variants) rather than keep an entry for
> every parent compatible-string in every child/cell driver (which all
> need to be kept in sync).
>
> But let's see what Rob and Lee says about this.
>
>> + if (!id) {
>> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
>> + return -ENODEV;
>> + }
>> +
>> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
>> + if (!sp_wd)
>> + return -ENOMEM;
>> +
>> + sp_wd->variant = id->data;
>> + sp_wd->sp = dev_get_drvdata(dev->parent);
>> +
>> + wdd = &sp_wd->wdd;
>> + wdd->parent = dev;
>> + wdd->info = &rave_sp_wdt_info;
>> + wdd->ops = &rave_sp_wdt_ops;
>> + wdd->min_timeout = sp_wd->variant->min_timeout;
>> + wdd->max_timeout = sp_wd->variant->max_timeout;
>> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
>> + wdd->timeout = 60;
>> +
>> + cell = nvmem_cell_get(dev, "wdt-timeout");
>> + if (!IS_ERR(cell)) {
>> + size_t len;
>> + void *value = nvmem_cell_read(cell, &len);
>> +
>> + if (!IS_ERR(value)) {
>> + memcpy(&timeout, value, min(len, sizeof(timeout)));
>> + kfree(value);
>> + }
>> + nvmem_cell_put(cell);
>> + }
>> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
>> + watchdog_set_restart_priority(wdd, 255);
>> +
>> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
>> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
>> + if (ret) {
>> + dev_err(dev, "Failed to register reboot notifier\n");
>> + return ret;
>> + }
>> +
>> + /*
>> + * We don't know if watchdog is running now. To be sure, let's
>> + * start it and depend on watchdog core to ping it
>> + */
>> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
>> + ret = rave_sp_wdt_start(wdd);
>> + if (ret) {
>> + dev_err(dev, "Watchdog didn't start\n");
>> + return ret;
>> + }
>> +
>> + return devm_watchdog_register_device(dev, wdd);
>
> What about stopping the watchdog on errors here?
>
> And does watchdog core take care of calling stop() on unregister (i.e.
> at unbind)?
>
Both good points. This needs watchdog_stop_on_unregister(), and the
probe function needs to stop the watchdog if registration fails.
Guenter
>> +}
>> +
>> +static struct platform_driver rave_sp_wdt_driver = {
>> + .probe = rave_sp_wdt_probe,
>> + .driver = {
>> + .name = KBUILD_MODNAME,
>> + .of_match_table = rave_sp_wdt_of_match,
>> + },
>> +};
>
> Johan
> --
> To unsubscribe from this list: send the line "unsubscribe linux-watchdog" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
^ permalink raw reply [flat|nested] 22+ messages in thread* Re: [PATCH v10 4/5] watchdog: Add RAVE SP watchdog driver
2017-11-05 15:47 ` Johan Hovold
2017-11-05 18:34 ` Guenter Roeck
@ 2017-11-05 21:59 ` Andrey Smirnov
1 sibling, 0 replies; 22+ messages in thread
From: Andrey Smirnov @ 2017-11-05 21:59 UTC (permalink / raw)
To: Johan Hovold
Cc: linux-kernel, linux-watchdog, Chris Healy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Guenter Roeck, Rob Herring, Sebastian Reichel
On Sun, Nov 5, 2017 at 7:47 AM, Johan Hovold <johan@kernel.org> wrote:
> On Tue, Oct 31, 2017 at 09:36:55AM -0700, Andrey Smirnov wrote:
>> This driver provides access to RAVE SP watchdog functionality.
>>
>> Cc: linux-kernel@vger.kernel.org
>> Cc: linux-watchdog@vger.kernel.org
>> Cc: cphealy@gmail.com
>> Cc: Lucas Stach <l.stach@pengutronix.de>
>> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Cc: Lee Jones <lee.jones@linaro.org>
>> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
>> Cc: Pavel Machek <pavel@ucw.cz>
>> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
>> Cc: Guenter Roeck <linux@roeck-us.net>
>> Cc: Rob Herring <robh@kernel.org>
>> Cc: Johan Hovold <johan@kernel.org>
>> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
>> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
>
>> +static const struct of_device_id rave_sp_wdt_of_match[] = {
>> + { .compatible = "zii,rave-sp-watchdog" },
>> + {}
>> +};
>
>> +static const struct of_device_id rave_sp_wdt_variants[] = {
>> + { .compatible = COMPATIBLE_RAVE_SP_NIU, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_MEZZ, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_ESB, .data = &rave_sp_wdt_legacy },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU1, .data = &rave_sp_wdt_rdu },
>> + { .compatible = COMPATIBLE_RAVE_SP_RDU2, .data = &rave_sp_wdt_rdu },
>> + { /* sentinel */ }
>> +};
>> +
>> +static int rave_sp_wdt_probe(struct platform_device *pdev)
>> +{
>> + struct device *dev = &pdev->dev;
>> + const struct of_device_id *id;
>> + struct watchdog_device *wdd;
>> + struct rave_sp_wdt *sp_wd;
>> + struct nvmem_cell *cell;
>> + __le16 timeout = 0;
>> + int ret;
>> +
>> + id = of_match_device(rave_sp_wdt_variants, dev->parent);
>
> So as I already mentioned in a comment on an earlier version of the MFD
> driver, I find this matching on the parent of_node to be an odd
> pattern, which also does not seem to have any precedent in mainline.
>
> It seems you really should be using two compatible strings for the
> watchdog (for the legacy and rdu variants) rather than keep an entry for
> every parent compatible-string in every child/cell driver (which all
> need to be kept in sync).
>
That's doable, for sure, but the downside is that it opens the door
for nonsensical configurations as, for example, "zii,rave-sp-niu" as a
parent SP device and "zii,rave-watchdog-rdu" as child watchdog and it
forces majority of child device drivers to split their compatibility
strings into 2 or more "flavors". My preference was to avoid that and
instead deal with having to having to keep things in sync.
> But let's see what Rob and Lee says about this.
>
>> + if (!id) {
>> + dev_err(dev, "Unknown parent device variant. Bailing out\n");
>> + return -ENODEV;
>> + }
>> +
>> + sp_wd = devm_kzalloc(dev, sizeof(*sp_wd), GFP_KERNEL);
>> + if (!sp_wd)
>> + return -ENOMEM;
>> +
>> + sp_wd->variant = id->data;
>> + sp_wd->sp = dev_get_drvdata(dev->parent);
>> +
>> + wdd = &sp_wd->wdd;
>> + wdd->parent = dev;
>> + wdd->info = &rave_sp_wdt_info;
>> + wdd->ops = &rave_sp_wdt_ops;
>> + wdd->min_timeout = sp_wd->variant->min_timeout;
>> + wdd->max_timeout = sp_wd->variant->max_timeout;
>> + wdd->status = WATCHDOG_NOWAYOUT_INIT_STATUS;
>> + wdd->timeout = 60;
>> +
>> + cell = nvmem_cell_get(dev, "wdt-timeout");
>> + if (!IS_ERR(cell)) {
>> + size_t len;
>> + void *value = nvmem_cell_read(cell, &len);
>> +
>> + if (!IS_ERR(value)) {
>> + memcpy(&timeout, value, min(len, sizeof(timeout)));
>> + kfree(value);
>> + }
>> + nvmem_cell_put(cell);
>> + }
>> + watchdog_init_timeout(wdd, le16_to_cpu(timeout), dev);
>> + watchdog_set_restart_priority(wdd, 255);
>> +
>> + sp_wd->reboot_notifier.notifier_call = rave_sp_wdt_reboot_notifier;
>> + ret = devm_register_reboot_notifier(dev, &sp_wd->reboot_notifier);
>> + if (ret) {
>> + dev_err(dev, "Failed to register reboot notifier\n");
>> + return ret;
>> + }
>> +
>> + /*
>> + * We don't know if watchdog is running now. To be sure, let's
>> + * start it and depend on watchdog core to ping it
>> + */
>> + wdd->max_hw_heartbeat_ms = wdd->max_timeout * 1000;
>> + ret = rave_sp_wdt_start(wdd);
>> + if (ret) {
>> + dev_err(dev, "Watchdog didn't start\n");
>> + return ret;
>> + }
>> +
>> + return devm_watchdog_register_device(dev, wdd);
>
> What about stopping the watchdog on errors here?
>
> And does watchdog core take care of calling stop() on unregister (i.e.
> at unbind)?
>
I think Guenter already addressed those, so I'll update the code with
his suggestions in v11.
Thanks,
Andrey Smirnov
^ permalink raw reply [flat|nested] 22+ messages in thread
* [PATCH v10 5/5] dt-bindings: watchdog: Add bindings for RAVE SP watchdog driver
2017-10-31 16:36 ` Andrey Smirnov
` (4 preceding siblings ...)
(?)
@ 2017-10-31 16:36 ` Andrey Smirnov
2017-10-31 17:22 ` Guenter Roeck
-1 siblings, 1 reply; 22+ messages in thread
From: Andrey Smirnov @ 2017-10-31 16:36 UTC (permalink / raw)
To: linux-kernel
Cc: Andrey Smirnov, devicetree, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Guenter Roeck, Rob Herring, Johan Hovold,
Mark Rutland, Sebastian Reichel
Add Device Tree bindings for RAVE SP watchdog drvier - an MFD cell of
parent RAVE SP driver (documented in
Documentation/devicetree/bindings/mfd/zii,rave-sp.txt).
Cc: linux-kernel@vger.kernel.org
Cc: devicetree@vger.kernel.org
Cc: linux-watchdog@vger.kernel.org
Cc: cphealy@gmail.com
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Cc: Lee Jones <lee.jones@linaro.org>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Pavel Machek <pavel@ucw.cz>
Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Rob Herring <robh@kernel.org>
Cc: Johan Hovold <johan@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
.../bindings/watchdog/zii,rave-sp-wdt.txt | 39 ++++++++++++++++++++++
1 file changed, 39 insertions(+)
create mode 100644 Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
diff --git a/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt b/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
new file mode 100644
index 000000000000..ab16659e781d
--- /dev/null
+++ b/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
@@ -0,0 +1,39 @@
+Zodiac Inflight Innovations RAVE Supervisory Processor Watchdog Bindings
+
+RAVE SP watchdog device is a "MFD cell" device corresponding to
+watchdog functionality of RAVE Supervisory Processor. It is expected
+that its Device Tree node is specified as a child of the node
+corresponding to the parent RAVE SP device (as documented in
+Documentation/devicetree/bindings/mfd/zii,rave-sp.txt)
+
+Required properties:
+- compatible: Should be "zii,rave-sp-watchdog"
+
+Optional properties:
+
+- wdt-timeout: Two byte nvmem cell specified as per
+ Documentation/devicetree/bindings/nvmem/nvmem.txt
+
+Example:
+
+ rave-sp {
+ compatible = "zii,rave-sp-rdu1";
+ current-speed = <38400>;
+
+ eeprom {
+ compatible = "zii,pic-main-eeprom";
+ #address-cells = <1>;
+ #size-cells = <1>;
+
+ wdt_timeout: wdt-timeout@8E {
+ reg = <0x8E 2>;
+ };
+ };
+
+ watchdog {
+ compatible = "zii,rave-sp-watchdog";
+ nvmem-cells = <&wdt_timeout>;
+ nvmem-cell-names = "wdt-timeout";
+ };
+ }
+
--
2.13.6
^ permalink raw reply related [flat|nested] 22+ messages in thread* Re: [PATCH v10 5/5] dt-bindings: watchdog: Add bindings for RAVE SP watchdog driver
2017-10-31 16:36 ` [PATCH v10 5/5] dt-bindings: watchdog: Add bindings for " Andrey Smirnov
@ 2017-10-31 17:22 ` Guenter Roeck
0 siblings, 0 replies; 22+ messages in thread
From: Guenter Roeck @ 2017-10-31 17:22 UTC (permalink / raw)
To: Andrey Smirnov
Cc: linux-kernel, devicetree, linux-watchdog, cphealy, Lucas Stach,
Nikita Yushchenko, Lee Jones, Greg Kroah-Hartman, Pavel Machek,
Andy Shevchenko, Rob Herring, Johan Hovold, Mark Rutland,
Sebastian Reichel
On Tue, Oct 31, 2017 at 09:36:56AM -0700, Andrey Smirnov wrote:
> Add Device Tree bindings for RAVE SP watchdog drvier - an MFD cell of
> parent RAVE SP driver (documented in
> Documentation/devicetree/bindings/mfd/zii,rave-sp.txt).
>
> Cc: linux-kernel@vger.kernel.org
> Cc: devicetree@vger.kernel.org
> Cc: linux-watchdog@vger.kernel.org
> Cc: cphealy@gmail.com
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Cc: Lee Jones <lee.jones@linaro.org>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Pavel Machek <pavel@ucw.cz>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Guenter Roeck <linux@roeck-us.net>
> Cc: Rob Herring <robh@kernel.org>
> Cc: Johan Hovold <johan@kernel.org>
> Cc: Mark Rutland <mark.rutland@arm.com>
> Cc: Sebastian Reichel <sebastian.reichel@collabora.co.uk>
> Acked-by: Rob Herring <robh@kernel.org>
> Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com>
> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> ---
> .../bindings/watchdog/zii,rave-sp-wdt.txt | 39 ++++++++++++++++++++++
> 1 file changed, 39 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
>
> diff --git a/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt b/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
> new file mode 100644
> index 000000000000..ab16659e781d
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/watchdog/zii,rave-sp-wdt.txt
> @@ -0,0 +1,39 @@
> +Zodiac Inflight Innovations RAVE Supervisory Processor Watchdog Bindings
> +
> +RAVE SP watchdog device is a "MFD cell" device corresponding to
> +watchdog functionality of RAVE Supervisory Processor. It is expected
> +that its Device Tree node is specified as a child of the node
> +corresponding to the parent RAVE SP device (as documented in
> +Documentation/devicetree/bindings/mfd/zii,rave-sp.txt)
> +
> +Required properties:
> +- compatible: Should be "zii,rave-sp-watchdog"
> +
> +Optional properties:
> +
> +- wdt-timeout: Two byte nvmem cell specified as per
> + Documentation/devicetree/bindings/nvmem/nvmem.txt
> +
> +Example:
> +
> + rave-sp {
> + compatible = "zii,rave-sp-rdu1";
> + current-speed = <38400>;
> +
> + eeprom {
> + compatible = "zii,pic-main-eeprom";
> + #address-cells = <1>;
> + #size-cells = <1>;
> +
> + wdt_timeout: wdt-timeout@8E {
> + reg = <0x8E 2>;
> + };
> + };
> +
> + watchdog {
> + compatible = "zii,rave-sp-watchdog";
> + nvmem-cells = <&wdt_timeout>;
> + nvmem-cell-names = "wdt-timeout";
> + };
> + }
> +
> --
> 2.13.6
>
^ permalink raw reply [flat|nested] 22+ messages in thread