Re: [v8,4/5] watchdog: Add RAVE SP watchdog driver

[Andrey Smirnov]

On Sat, Oct 21, 2017 at 9:47 AM, Guenter Roeck <linux@xxxxxxxxxxxx> wrote:
> On Wed, Oct 18, 2017 at 10:01:35AM -0700, Andrey Smirnov wrote:
>> This driver provides access to RAVE SP watchdog functionality.
>>
>> Cc: linux-kernel@xxxxxxxxxxxxxxx
>> Cc: linux-watchdog@xxxxxxxxxxxxxxx
>> Cc: cphealy@xxxxxxxxx
>> Cc: Lucas Stach <l.stach@xxxxxxxxxxxxxx>
>> Cc: Nikita Yushchenko <nikita.yoush@xxxxxxxxxxxxxxxxxx>
>> Cc: Lee Jones <lee.jones@xxxxxxxxxx>
>> Cc: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
>> Cc: Pavel Machek <pavel@xxxxxx>
>> Cc: Andy Shevchenko <andy.shevchenko@xxxxxxxxx>
>> Cc: Guenter Roeck <linux@xxxxxxxxxxxx>
>> Cc: Rob Herring <robh@xxxxxxxxxx>
>> Cc: Johan Hovold <johan@xxxxxxxxxx>
>> Signed-off-by: Nikita Yushchenko <nikita.yoush@xxxxxxxxxxxxxxxxxx>
>> Signed-off-by: Andrey Smirnov <andrew.smirnov@xxxxxxxxx>
>> ---
>>  drivers/watchdog/Kconfig       |   7 +
>>  drivers/watchdog/Makefile      |   1 +
>>  drivers/watchdog/rave-sp-wdt.c | 310 +++++++++++++++++++++++++++++++++++++++++
>>  3 files changed, 318 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..86fdc7fb1f92
>> --- /dev/null
>> +++ b/drivers/watchdog/rave-sp-wdt.c
>> @@ -0,0 +1,310 @@
>> +/*
>> + *  rave-sp-wdt.c - Watchdog driver preseing in RAVE SP
>
> present ?
>

Yup, my bad, will fix in v9.

>> + *
>> + * Copyright (C) 2017 Zodiac Inflight Innovation
>> + *
>> + * 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/of_device.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/rave-sp.h>
>> +#include <linux/watchdog.h>
>> +#include <linux/nvmem-consumer.h>
>> +#include <linux/reboot.h>
>> +#include <linux/slab.h>
>
> Alphabetic order please.
>

OK, will do.

>> +
>> +enum {
>> +     RAVE_SP_RESET_BYTE = 1,
>> +     RAVE_SP_RESET_REASON_NORMAL = 0,
>> +     RAVE_SP_RESET_DELAY_MS = 500,
>> +};
>> +
>> +struct rave_sp_wdt_variant {
>> +     unsigned int max_timeout;
>> +     unsigned int min_timeout;
>> +
>> +     int (*configure)(struct watchdog_device *);
>> +     int (*restart)(struct watchdog_device *);
>> +};
>> +
>> +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,
>> +                           void *reply, size_t reply_size)
>> +{
>> +     return rave_sp_exec(to_rave_sp_wdt(wdd)->sp,
>> +                         data, data_size, reply, reply_size);
>> +}
>
> This extra function is only called once and thus quite pointless.
>

True. Will remove in v9.

>> +
>> +static int rave_sp_wdt_exec(struct watchdog_device *wdd, void *data,
>> +                         size_t data_size)
>> +{
>> +     return __rave_sp_wdt_exec(wdd, data, data_size, NULL, 0);
>> +}
>> +
>> +
>
> Please run checkpatch --strict and fix what it reports.
>

Sure, will do.

>> +static int rave_sp_wdt_legacy_configure(struct watchdog_device *wdd)
>> +{
>> +     const bool enable = watchdog_hw_running(wdd);
>> +     u8 cmd[] = {
>> +             [0] = RAVE_SP_CMD_SW_WDT,
>> +             [1] = 0,
>> +             [2] = 0,
>> +             [3] = !!enable,
>
> Useless !!. It converts a boolean to a boolean. Besides, isn't it always
> true anyway ?

Good point, will remove the "!!". This function also is called as
.stop method (via rave_sp_wdt_configure) for corresponding
"watchdog_device" and I rely on code in watchdog_stop() to clear
WDOG_HW_RUNNING. Is this not a correct thing to do?

>
>> +             [4] = enable ? wdd->timeout : 0,
>> +     };
>
> Interesting; checkpatch doesn't require an empty line after variable
> declarations anymore ? Lets keep it anyway for consistency.
>

OK, will fix in v9.

>> +     return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static int rave_sp_wdt_rdu_configure(struct watchdog_device *wdd)
>> +{
>> +     u8 cmd[] = {
>> +             [0] = RAVE_SP_CMD_SW_WDT,
>> +             [1] = 0,
>> +             [2] = watchdog_hw_running(wdd),
>
> Nitpick, but isn't this always true ? There is no stop function,
> and HW_RUNNING is set before this function is called.
>

Same as above. Let me know if relying on HW_RUNNING being cleared by
watchdog_stop() is incorrect.

>> +             [3] = (u8) wdd->timeout,
>> +             [4] = (u8) (wdd->timeout >> 8),
>> +     };
>> +     return rave_sp_wdt_exec(wdd, cmd, sizeof(cmd));
>> +}
>> +
>> +static int rave_sp_wdt_configure(struct watchdog_device *wdd)
>> +{
>> +     return to_rave_sp_wdt(wdd)->variant->configure(wdd);
>> +}
>> +
>> +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 commuicate to SP via UART. Luckily for use SP will
>
> communicate
>

My bad, will fix in v9.

>> +      * 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 one second and hopefuly we'd never reach that
>
> RAVE_SP_RESET_DELAY_MS is 500(ms), not one second.
>

Sorry, meant to wait for twice that delay, hence one second, but wrote
incorrect code. Will fix in v9.

>> +      * return statement
>> +      */
>> +     mdelay(RAVE_SP_RESET_DELAY_MS);
>> +     return -EIO;
>> +}
>> +
>> +static int rave_sp_wdt_start(struct watchdog_device *wdd)
>> +{
>> +     set_bit(WDOG_HW_RUNNING, &wdd->status);
>> +     return rave_sp_wdt_configure(wdd);
>> +}
>> +
>> +static int rave_sp_wdt_set_timeout(struct watchdog_device *wdd,
>> +                                unsigned int timeout)
>> +{
>> +     wdd->timeout = timeout;
>> +     return rave_sp_wdt_configure(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_configure,
>> +     .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 (WARN_ON(!id))
>> +             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);
>> +
>> +     if (WARN_ON(!sp_wd->sp))
>> +             return -ENODEV;
>> +
> Not sure I understand the value of those tracebacks. Presumably the driver
> is instantiated through its parent. Is there precedent that other drivers
> in the same situation do this ?
>

No precedent. I was just being paranoid. Will remove in v9.

Thanks,
Andrey Smirnov