diff mbox series

[RFC,v2,2/2] watchdog: pseries-wdt: initial support for PAPR virtual watchdog timers

Message ID 20220509174357.5448-3-cheloha@linux.ibm.com (mailing list archive)
State Changes Requested
Headers show
Series Add driver for PAPR watchdog timers | expand

Commit Message

Scott Cheloha May 9, 2022, 5:43 p.m. UTC
PAPR v2.12 defines a new hypercall, H_WATCHDOG.  It permits guest
control of one or more virtual watchdog timers.  The timers have
millisecond granularity and terminate the guest upon expiration.

This patch adds a new driver for these timers, "pseries-wdt".

Unfortunately, these timers defy PowerPC device conventions.  They are
not affixed to any bus, nor do they have full representation in the
device tree.  As a workaround, we represent them as platform devices.
A platform device corresponding to a given timer is registered with
the platform bus via a machine_subsys_initcall() if firmware support
for the H_WATCHDOG hypercall is detected early in boot.

For now, only a single timer is registered with the platform bus and
exposed to userspace.  In the future we could expose additional timers
if there is good reason to do so.

Signed-off-by: Scott Cheloha <cheloha@linux.ibm.com>
---
 arch/powerpc/include/asm/firmware.h       |   3 +-
 arch/powerpc/platforms/pseries/firmware.c |   1 +
 arch/powerpc/platforms/pseries/setup.c    |  17 ++
 drivers/watchdog/Kconfig                  |   8 +
 drivers/watchdog/Makefile                 |   1 +
 drivers/watchdog/pseries-wdt.c            | 300 ++++++++++++++++++++++
 6 files changed, 329 insertions(+), 1 deletion(-)
 create mode 100644 drivers/watchdog/pseries-wdt.c

Comments

Tzung-Bi Shih May 10, 2022, 2:35 a.m. UTC | #1
On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
> +#define SETFIELD(_v, _b, _e)	\
> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
> +#define GETFIELD(_v, _b, _e)	\
> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))

From `./scripts/checkpatch.pl --strict`:
WARNING: please, no spaces at the start of a line

> +#define PSERIES_WDTQL_MUST_STOP       	1

From `./scripts/checkpatch.pl --strict`:
WARNING: please, no space before tabs

> +static const struct kernel_param_ops action_ops = { .set = action_set };
> +module_param_cb(action, &action_ops, NULL, S_IRUGO);

From `./scripts/checkpatch.pl --strict`:
WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
octal permissions '0444'.

> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");

The line exceeds 100 columns.

> +static bool nowayout = WATCHDOG_NOWAYOUT;
> +module_param(nowayout, bool, S_IRUGO);

From `./scripts/checkpatch.pl --strict`:
WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
octal permissions '0444'.

> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");

From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.

> +#define WATCHDOG_TIMEOUT 60
> +static unsigned int timeout = WATCHDOG_TIMEOUT;
> +module_param(timeout, uint, S_IRUGO);

From `./scripts/checkpatch.pl --strict`:
WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
octal permissions '0444'.

> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");

From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.

> +struct pseries_wdt {
> +	struct watchdog_device wd;
> +	unsigned long num;		/* NB: Watchdog numbers are 1-based */

What does NB stand for?  Could it be removed from the comment?

Does `timer_id` or some other equivalent names make more sense for the
variable?

> +static int pseries_wdt_start(struct watchdog_device *wdd)
> +{
[...]
> +	rc = plpar_hcall_norets(H_WATCHDOG, flags, pw->num, msecs);
> +	if (rc != H_SUCCESS) {
> +		dev_crit(dev, "H_WATCHDOG: %ld: failed to start timer %lu",
> +			 rc, pw->num);
> +	       	return -EIO;

From `./scripts/checkpatch.pl --strict`:
ERROR: code indent should use tabs where possible
WARNING: please, no space before tabs

> +static struct watchdog_info pseries_wdt_info = {
> +	.identity = DRV_NAME,
> +	.options = WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE | WDIOF_SETTIMEOUT \
> +	    | WDIOF_PRETIMEOUT,

From `./scripts/checkpatch.pl --strict`:
WARNING: Avoid unnecessary line continuations

> +static const struct watchdog_ops pseries_wdt_ops = {
> +	.owner = THIS_MODULE,
> +	.ping = pseries_wdt_start,

Does this mean: it needs hard restart for every ping?

> +static int pseries_wdt_probe(struct platform_device *pdev)
> +{
[...]
> +	rc = plpar_hcall(H_WATCHDOG, ret, PSERIES_WDTF_OP_QUERY);
> +	if (rc != H_SUCCESS)
> +		return (rc == H_FUNCTION) ? -ENODEV : -EIO;

The parentheses can be dropped.

> +	pw = devm_kzalloc(&pdev->dev, sizeof *pw, GFP_KERNEL);
> +	if (pw == NULL)

From `./scripts/checkpatch.pl --strict`:
CHECK: Comparison to NULL could be written "!pw"

> +	pw->num = pdev->id + 1;		/* 0-based -> 1-based */

Didn't see where the platform device was registered but using the pdev->id as
the timer id could be unreliable (e.g. from auto increment).
Guenter Roeck May 10, 2022, 3:34 a.m. UTC | #2
On 5/9/22 19:35, Tzung-Bi Shih wrote:
> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>> +#define SETFIELD(_v, _b, _e)	\
>> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>> +#define GETFIELD(_v, _b, _e)	\
>> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: please, no spaces at the start of a line
> 
>> +#define PSERIES_WDTQL_MUST_STOP       	1
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: please, no space before tabs
> 
>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
> 
> The line exceeds 100 columns.
> 
>> +static bool nowayout = WATCHDOG_NOWAYOUT;
>> +module_param(nowayout, bool, S_IRUGO);
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
> 
>>From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
> 
>> +#define WATCHDOG_TIMEOUT 60
>> +static unsigned int timeout = WATCHDOG_TIMEOUT;
>> +module_param(timeout, uint, S_IRUGO);
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
> 
>>From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
> 
>> +struct pseries_wdt {
>> +	struct watchdog_device wd;
>> +	unsigned long num;		/* NB: Watchdog numbers are 1-based */
> 
> What does NB stand for?  Could it be removed from the comment?
> 

Latin "Nota Bene", for "This is important".
All comments should be important, so I agree, this has little
if any value.

> Does `timer_id` or some other equivalent names make more sense for the
> variable?
> 
>> +static int pseries_wdt_start(struct watchdog_device *wdd)
>> +{
> [...]
>> +	rc = plpar_hcall_norets(H_WATCHDOG, flags, pw->num, msecs);
>> +	if (rc != H_SUCCESS) {
>> +		dev_crit(dev, "H_WATCHDOG: %ld: failed to start timer %lu",
>> +			 rc, pw->num);
>> +	       	return -EIO;
> 
>>From `./scripts/checkpatch.pl --strict`:
> ERROR: code indent should use tabs where possible
> WARNING: please, no space before tabs
> 
>> +static struct watchdog_info pseries_wdt_info = {
>> +	.identity = DRV_NAME,
>> +	.options = WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE | WDIOF_SETTIMEOUT \
>> +	    | WDIOF_PRETIMEOUT,
> 
>>From `./scripts/checkpatch.pl --strict`:
> WARNING: Avoid unnecessary line continuations
> 
>> +static const struct watchdog_ops pseries_wdt_ops = {
>> +	.owner = THIS_MODULE,
>> +	.ping = pseries_wdt_start,
> 
> Does this mean: it needs hard restart for every ping?
> 

If there is no separate ping function, there is no need to point
the ping function to the start function. The watchdog core uses
the start function automatically in this case.

>> +static int pseries_wdt_probe(struct platform_device *pdev)
>> +{
> [...]
>> +	rc = plpar_hcall(H_WATCHDOG, ret, PSERIES_WDTF_OP_QUERY);
>> +	if (rc != H_SUCCESS)
>> +		return (rc == H_FUNCTION) ? -ENODEV : -EIO;
> 
> The parentheses can be dropped.
> 
>> +	pw = devm_kzalloc(&pdev->dev, sizeof *pw, GFP_KERNEL);
>> +	if (pw == NULL)
> 
>>From `./scripts/checkpatch.pl --strict`:
> CHECK: Comparison to NULL could be written "!pw"
> 
>> +	pw->num = pdev->id + 1;		/* 0-based -> 1-based */
> 
> Didn't see where the platform device was registered but using the pdev->id as
> the timer id could be unreliable (e.g. from auto increment).

That is at the beginning of the patch, in arch/powerpc/platforms/pseries/setup.c,
which calls platform_device_register_simple() with an explicit device ID.

I agree, though, that it is fragile: The code in the probe function
explicitly checks for a negative ID, which would only be provided if
platform_device_register() explicitly declares that (ie if it provides
PLATFORM_DEVID_NONE as device ID). But if that is a concern, there
might as well be code registering a platform device with
PLATFORM_DEVID_AUTO as device ID, and then th device id would be automatic
and more or less random. I think the instantiation code should be more
explicit: Either assume that PLATFORM_DEVID_NONE or PLATFORM_DEVID_AUTO
will never be used to register the watchdog devices, or provide other
information such as platform data to make it explicit.

Also, the changes in arch code need to be made in separate patches.

Thanks,
Guenter
Scott Cheloha May 10, 2022, 3:48 p.m. UTC | #3
> On May 9, 2022, at 9:35 PM, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
> 
> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>> +#define SETFIELD(_v, _b, _e)	\
>> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>> +#define GETFIELD(_v, _b, _e)	\
>> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
> 
> From `./scripts/checkpatch.pl --strict`:
> WARNING: please, no spaces at the start of a line
> 
>> +#define PSERIES_WDTQL_MUST_STOP       	1
> 
> From `./scripts/checkpatch.pl --strict`:
> WARNING: please, no space before tabs
> 
>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
> 
> From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
> 
> The line exceeds 100 columns.

I was under the impression that strings were an exception to the
line-length rule.  Is that not the case?

`scripts/checkpatch.pl --strict` complains if I break the string up:

WARNING: quoted string split across lines
#162: FILE: drivers/watchdog/pseries-wdt.c:162:
+MODULE_PARM_DESC(action, "Action taken when watchdog expires: "
+		 "\"hard-poweroff\", \"hard-restart\", or \"dump-restart\" "

WARNING: quoted string split across lines
#163: FILE: drivers/watchdog/pseries-wdt.c:163:
+		 "\"hard-poweroff\", \"hard-restart\", or \"dump-restart\" "
+		 "(default=\"hard-restart\")");

>> +static bool nowayout = WATCHDOG_NOWAYOUT;
>> +module_param(nowayout, bool, S_IRUGO);
> 
> From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
> 
> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.

> 
>> +#define WATCHDOG_TIMEOUT 60
>> +static unsigned int timeout = WATCHDOG_TIMEOUT;
>> +module_param(timeout, uint, S_IRUGO);
> 
> From `./scripts/checkpatch.pl --strict`:
> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
> octal permissions '0444'.
> 
>> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
> 
> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
> 
>> +struct pseries_wdt {
>> +	struct watchdog_device wd;
>> +	unsigned long num;		/* NB: Watchdog numbers are 1-based */
> 
> What does NB stand for?  Could it be removed from the comment?

Sure, removed.

> Does `timer_id` or some other equivalent names make more sense for the
> variable?

The hardware documentation calls the parameter "watchdogNumber", so
I think "num" is better than "id" in this case.
Guenter Roeck May 10, 2022, 4:04 p.m. UTC | #4
On 5/10/22 08:48, Scott Cheloha wrote:
>> On May 9, 2022, at 9:35 PM, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>>
>> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>>> +#define SETFIELD(_v, _b, _e)	\
>>> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>>> +#define GETFIELD(_v, _b, _e)	\
>>> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
>>
>>  From `./scripts/checkpatch.pl --strict`:
>> WARNING: please, no spaces at the start of a line
>>
>>> +#define PSERIES_WDTQL_MUST_STOP       	1
>>
>>  From `./scripts/checkpatch.pl --strict`:
>> WARNING: please, no space before tabs
>>
>>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
>>
>>  From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
>>
>> The line exceeds 100 columns.
> 
> I was under the impression that strings were an exception to the
> line-length rule.  Is that not the case?
> 
You can use

MODULE_PARM_DESC(action,
		 "Some text");

> `scripts/checkpatch.pl --strict` complains if I break the string up:
> 
> WARNING: quoted string split across lines
> #162: FILE: drivers/watchdog/pseries-wdt.c:162:
> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: "
> +		 "\"hard-poweroff\", \"hard-restart\", or \"dump-restart\" "
> 
> WARNING: quoted string split across lines
> #163: FILE: drivers/watchdog/pseries-wdt.c:163:
> +		 "\"hard-poweroff\", \"hard-restart\", or \"dump-restart\" "
> +		 "(default=\"hard-restart\")");
> 
>>> +static bool nowayout = WATCHDOG_NOWAYOUT;
>>> +module_param(nowayout, bool, S_IRUGO);
>>
>>  From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
>>
>>  From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
> 
>>
>>> +#define WATCHDOG_TIMEOUT 60
>>> +static unsigned int timeout = WATCHDOG_TIMEOUT;
>>> +module_param(timeout, uint, S_IRUGO);
>>
>>  From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
>>
>>  From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
>>
>>> +struct pseries_wdt {
>>> +	struct watchdog_device wd;
>>> +	unsigned long num;		/* NB: Watchdog numbers are 1-based */
>>
>> What does NB stand for?  Could it be removed from the comment?
> 
> Sure, removed.
> 
>> Does `timer_id` or some other equivalent names make more sense for the
>> variable?
> 
> The hardware documentation calls the parameter "watchdogNumber", so
> I think "num" is better than "id" in this case.
>
Scott Cheloha May 10, 2022, 4:15 p.m. UTC | #5
> On May 10, 2022, at 11:04 AM, Guenter Roeck <linux@roeck-us.net> wrote:
> 
> On 5/10/22 08:48, Scott Cheloha wrote:
>>> On May 9, 2022, at 9:35 PM, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>>> 
>>> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>>>> +#define SETFIELD(_v, _b, _e)	\
>>>> + (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>>>> +#define GETFIELD(_v, _b, _e)	\
>>>> + (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
>>> 
>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: please, no spaces at the start of a line
>>> 
>>>> +#define PSERIES_WDTQL_MUST_STOP 	1
>>> 
>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: please, no space before tabs
>>> 
>>>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>>>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
>>> 
>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>>> octal permissions '0444'.
>>> 
>>>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
>>> 
>>> The line exceeds 100 columns.
>> I was under the impression that strings were an exception to the
>> line-length rule. Is that not the case?
> You can use
> 
> MODULE_PARM_DESC(action,
> 		 "Some text");

The line is still over 100 columns if I do this.

I can shrink the line by removing the valid inputs from the string if 100 columns
is a hard rule.

If so, where should I document the valid inputs instead?

Is Documentation/watchdog/watchdog-parameters.rst a better place for them?
Guenter Roeck May 10, 2022, 7 p.m. UTC | #6
On 5/10/22 09:15, Scott Cheloha wrote:
> 
> 
>> On May 10, 2022, at 11:04 AM, Guenter Roeck <linux@roeck-us.net> wrote:
>>
>> On 5/10/22 08:48, Scott Cheloha wrote:
>>>> On May 9, 2022, at 9:35 PM, Tzung-Bi Shih <tzungbi@kernel.org> wrote:
>>>>
>>>> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>>>>> +#define SETFIELD(_v, _b, _e)	\
>>>>> + (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>>>>> +#define GETFIELD(_v, _b, _e)	\
>>>>> + (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
>>>>
>>>>  From `./scripts/checkpatch.pl --strict`:
>>>> WARNING: please, no spaces at the start of a line
>>>>
>>>>> +#define PSERIES_WDTQL_MUST_STOP 	1
>>>>
>>>>  From `./scripts/checkpatch.pl --strict`:
>>>> WARNING: please, no space before tabs
>>>>
>>>>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>>>>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
>>>>
>>>>  From `./scripts/checkpatch.pl --strict`:
>>>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>>>> octal permissions '0444'.
>>>>
>>>>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
>>>>
>>>> The line exceeds 100 columns.
>>> I was under the impression that strings were an exception to the
>>> line-length rule. Is that not the case?
>> You can use
>>
>> MODULE_PARM_DESC(action,
>> 		 "Some text");
> 
> The line is still over 100 columns if I do this.
> 

You might consider using shorter strings, such as "poweroff", "restart",
"dump". The meaning needs to be documented anyway. You could also drop
"or" because it doesn't add value in this context, and the '"' don't really
add value either.

On a side note, personally I don't like textual module parameters because
they make the command line extremely long and add complexity to decoding
it, but that is your call and responsibility (and you are not the only one
using strings).

Guenter

> I can shrink the line by removing the valid inputs from the string if 100 columns
> is a hard rule.
> 
> If so, where should I document the valid inputs instead?
> 
> Is Documentation/watchdog/watchdog-parameters.rst a better place for them?
>
Alexey Kardashevskiy May 11, 2022, 5:08 a.m. UTC | #7
On 5/10/22 13:34, Guenter Roeck wrote:
> On 5/9/22 19:35, Tzung-Bi Shih wrote:
>> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>>> +#define SETFIELD(_v, _b, _e)    \
>>> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), 
>>> (_e)))
>>> +#define GETFIELD(_v, _b, _e)    \
>>> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> 
>>> PPC_BITLSHIFT(_e))
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: please, no spaces at the start of a line
>>
>>> +#define PSERIES_WDTQL_MUST_STOP           1
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: please, no space before tabs
>>
>>> +static const struct kernel_param_ops action_ops = { .set = 
>>> action_set };
>>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: 
>>> \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" 
>>> (default=\"hard-restart\")");
>>
>> The line exceeds 100 columns.
>>
>>> +static bool nowayout = WATCHDOG_NOWAYOUT;
>>> +module_param(nowayout, bool, S_IRUGO);
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started 
>>> (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
>>
>>> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
>>
>>> +#define WATCHDOG_TIMEOUT 60
>>> +static unsigned int timeout = WATCHDOG_TIMEOUT;
>>> +module_param(timeout, uint, S_IRUGO);
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>> octal permissions '0444'.
>>
>>> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds 
>>> (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
>>
>>> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
>>
>>> +struct pseries_wdt {
>>> +    struct watchdog_device wd;
>>> +    unsigned long num;        /* NB: Watchdog numbers are 1-based */
>>
>> What does NB stand for?  Could it be removed from the comment?
>>
> 
> Latin "Nota Bene", for "This is important".
> All comments should be important, so I agree, this has little
> if any value.
> 
>> Does `timer_id` or some other equivalent names make more sense for the
>> variable?
>>
>>> +static int pseries_wdt_start(struct watchdog_device *wdd)
>>> +{
>> [...]
>>> +    rc = plpar_hcall_norets(H_WATCHDOG, flags, pw->num, msecs);
>>> +    if (rc != H_SUCCESS) {
>>> +        dev_crit(dev, "H_WATCHDOG: %ld: failed to start timer %lu",
>>> +             rc, pw->num);
>>> +               return -EIO;
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> ERROR: code indent should use tabs where possible
>> WARNING: please, no space before tabs
>>
>>> +static struct watchdog_info pseries_wdt_info = {
>>> +    .identity = DRV_NAME,
>>> +    .options = WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE | 
>>> WDIOF_SETTIMEOUT \
>>> +        | WDIOF_PRETIMEOUT,
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> WARNING: Avoid unnecessary line continuations
>>
>>> +static const struct watchdog_ops pseries_wdt_ops = {
>>> +    .owner = THIS_MODULE,
>>> +    .ping = pseries_wdt_start,
>>
>> Does this mean: it needs hard restart for every ping?
>>
> 
> If there is no separate ping function, there is no need to point
> the ping function to the start function. The watchdog core uses
> the start function automatically in this case.
> 
>>> +static int pseries_wdt_probe(struct platform_device *pdev)
>>> +{
>> [...]
>>> +    rc = plpar_hcall(H_WATCHDOG, ret, PSERIES_WDTF_OP_QUERY);
>>> +    if (rc != H_SUCCESS)
>>> +        return (rc == H_FUNCTION) ? -ENODEV : -EIO;
>>
>> The parentheses can be dropped.
>>
>>> +    pw = devm_kzalloc(&pdev->dev, sizeof *pw, GFP_KERNEL);
>>> +    if (pw == NULL)
>>
>>> From `./scripts/checkpatch.pl --strict`:
>> CHECK: Comparison to NULL could be written "!pw"
>>
>>> +    pw->num = pdev->id + 1;        /* 0-based -> 1-based */
>>
>> Didn't see where the platform device was registered but using the 
>> pdev->id as
>> the timer id could be unreliable (e.g. from auto increment).
> 
> That is at the beginning of the patch, in 
> arch/powerpc/platforms/pseries/setup.c,
> which calls platform_device_register_simple() with an explicit device ID.
> 
> I agree, though, that it is fragile: The code in the probe function
> explicitly checks for a negative ID, which would only be provided if
> platform_device_register() explicitly declares that (ie if it provides
> PLATFORM_DEVID_NONE as device ID). But if that is a concern, there
> might as well be code registering a platform device with
> PLATFORM_DEVID_AUTO as device ID, and then th device id would be automatic
> and more or less random. I think the instantiation code should be more
> explicit: Either assume that PLATFORM_DEVID_NONE or PLATFORM_DEVID_AUTO
> will never be used to register the watchdog devices, or provide other
> information such as platform data to make it explicit.
> 
> Also, the changes in arch code need to be made in separate patches.


Often this is true but in case like this - the driver itself is dead 
code until the platform enables it and if there is a problem with the 
driver - bisect will point to the device enablement commit. Backporting 
becomes a problem as it is going to be 3 apart patches vs. one:
1) define hypercall
2) implement the driver
3) enable the driver + device FW feature flag.

Instead, either maintainer (powerpc or watchdog) says "ack" and the 
other one puts it in the tree, I saw this in practice. Thanks,



> Thanks,
> Guenter
Guenter Roeck May 11, 2022, 5:27 a.m. UTC | #8
On 5/10/22 22:08, Alexey Kardashevskiy wrote:
> 
> 
> On 5/10/22 13:34, Guenter Roeck wrote:
>> On 5/9/22 19:35, Tzung-Bi Shih wrote:
>>> On Mon, May 09, 2022 at 12:43:57PM -0500, Scott Cheloha wrote:
>>>> +#define SETFIELD(_v, _b, _e)    \
>>>> +    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
>>>> +#define GETFIELD(_v, _b, _e)    \
>>>> +    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: please, no spaces at the start of a line
>>>
>>>> +#define PSERIES_WDTQL_MUST_STOP           1
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: please, no space before tabs
>>>
>>>> +static const struct kernel_param_ops action_ops = { .set = action_set };
>>>> +module_param_cb(action, &action_ops, NULL, S_IRUGO);
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>>> octal permissions '0444'.
>>>
>>>> +MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
>>>
>>> The line exceeds 100 columns.
>>>
>>>> +static bool nowayout = WATCHDOG_NOWAYOUT;
>>>> +module_param(nowayout, bool, S_IRUGO);
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>>> octal permissions '0444'.
>>>
>>>> +MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
>>>
>>>> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
>>>
>>>> +#define WATCHDOG_TIMEOUT 60
>>>> +static unsigned int timeout = WATCHDOG_TIMEOUT;
>>>> +module_param(timeout, uint, S_IRUGO);
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: Symbolic permissions 'S_IRUGO' are not preferred. Consider using
>>> octal permissions '0444'.
>>>
>>>> +MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
>>>
>>>> From `./scripts/checkpatch.pl --strict`, the line exceeds 100 columns.
>>>
>>>> +struct pseries_wdt {
>>>> +    struct watchdog_device wd;
>>>> +    unsigned long num;        /* NB: Watchdog numbers are 1-based */
>>>
>>> What does NB stand for?  Could it be removed from the comment?
>>>
>>
>> Latin "Nota Bene", for "This is important".
>> All comments should be important, so I agree, this has little
>> if any value.
>>
>>> Does `timer_id` or some other equivalent names make more sense for the
>>> variable?
>>>
>>>> +static int pseries_wdt_start(struct watchdog_device *wdd)
>>>> +{
>>> [...]
>>>> +    rc = plpar_hcall_norets(H_WATCHDOG, flags, pw->num, msecs);
>>>> +    if (rc != H_SUCCESS) {
>>>> +        dev_crit(dev, "H_WATCHDOG: %ld: failed to start timer %lu",
>>>> +             rc, pw->num);
>>>> +               return -EIO;
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> ERROR: code indent should use tabs where possible
>>> WARNING: please, no space before tabs
>>>
>>>> +static struct watchdog_info pseries_wdt_info = {
>>>> +    .identity = DRV_NAME,
>>>> +    .options = WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE | WDIOF_SETTIMEOUT \
>>>> +        | WDIOF_PRETIMEOUT,
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> WARNING: Avoid unnecessary line continuations
>>>
>>>> +static const struct watchdog_ops pseries_wdt_ops = {
>>>> +    .owner = THIS_MODULE,
>>>> +    .ping = pseries_wdt_start,
>>>
>>> Does this mean: it needs hard restart for every ping?
>>>
>>
>> If there is no separate ping function, there is no need to point
>> the ping function to the start function. The watchdog core uses
>> the start function automatically in this case.
>>
>>>> +static int pseries_wdt_probe(struct platform_device *pdev)
>>>> +{
>>> [...]
>>>> +    rc = plpar_hcall(H_WATCHDOG, ret, PSERIES_WDTF_OP_QUERY);
>>>> +    if (rc != H_SUCCESS)
>>>> +        return (rc == H_FUNCTION) ? -ENODEV : -EIO;
>>>
>>> The parentheses can be dropped.
>>>
>>>> +    pw = devm_kzalloc(&pdev->dev, sizeof *pw, GFP_KERNEL);
>>>> +    if (pw == NULL)
>>>
>>>> From `./scripts/checkpatch.pl --strict`:
>>> CHECK: Comparison to NULL could be written "!pw"
>>>
>>>> +    pw->num = pdev->id + 1;        /* 0-based -> 1-based */
>>>
>>> Didn't see where the platform device was registered but using the pdev->id as
>>> the timer id could be unreliable (e.g. from auto increment).
>>
>> That is at the beginning of the patch, in arch/powerpc/platforms/pseries/setup.c,
>> which calls platform_device_register_simple() with an explicit device ID.
>>
>> I agree, though, that it is fragile: The code in the probe function
>> explicitly checks for a negative ID, which would only be provided if
>> platform_device_register() explicitly declares that (ie if it provides
>> PLATFORM_DEVID_NONE as device ID). But if that is a concern, there
>> might as well be code registering a platform device with
>> PLATFORM_DEVID_AUTO as device ID, and then th device id would be automatic
>> and more or less random. I think the instantiation code should be more
>> explicit: Either assume that PLATFORM_DEVID_NONE or PLATFORM_DEVID_AUTO
>> will never be used to register the watchdog devices, or provide other
>> information such as platform data to make it explicit.
>>
>> Also, the changes in arch code need to be made in separate patches.
> 
> 
> Often this is true but in case like this - the driver itself is dead code until the platform enables it and if there is a problem with the driver - bisect will point to the device enablement commit. Backporting becomes a problem as it is going to be 3 apart patches vs. one:
> 1) define hypercall
> 2) implement the driver
> 3) enable the driver + device FW feature flag.
> 
> Instead, either maintainer (powerpc or watchdog) says "ack" and the other one puts it in the tree, I saw this in practice. Thanks,
> 

I won't accept this. I don't mind giving an Ack to a patch applied elsewhere,
and I don't mind applying patches from another subsystem if Acked by affected
maintainers, but I strongly believe that there should be no cross-subsystem
patches. Also, I don't consider 2) and 3) to be a single logical change,
and the rule for patches is "one logical change per patch".

Other maintainers may handle this differently, but that is their call and
responsibility.

Guenter
diff mbox series

Patch

diff --git a/arch/powerpc/include/asm/firmware.h b/arch/powerpc/include/asm/firmware.h
index 8dddd34b8ecf..398e0b5e485f 100644
--- a/arch/powerpc/include/asm/firmware.h
+++ b/arch/powerpc/include/asm/firmware.h
@@ -55,6 +55,7 @@ 
 #define FW_FEATURE_RPT_INVALIDATE ASM_CONST(0x0000010000000000)
 #define FW_FEATURE_FORM2_AFFINITY ASM_CONST(0x0000020000000000)
 #define FW_FEATURE_ENERGY_SCALE_INFO ASM_CONST(0x0000040000000000)
+#define FW_FEATURE_WATCHDOG	ASM_CONST(0x0000080000000000)
 
 #ifndef __ASSEMBLY__
 
@@ -76,7 +77,7 @@  enum {
 		FW_FEATURE_DRC_INFO | FW_FEATURE_BLOCK_REMOVE |
 		FW_FEATURE_PAPR_SCM | FW_FEATURE_ULTRAVISOR |
 		FW_FEATURE_RPT_INVALIDATE | FW_FEATURE_FORM2_AFFINITY |
-		FW_FEATURE_ENERGY_SCALE_INFO,
+		FW_FEATURE_ENERGY_SCALE_INFO | FW_FEATURE_WATCHDOG,
 	FW_FEATURE_PSERIES_ALWAYS = 0,
 	FW_FEATURE_POWERNV_POSSIBLE = FW_FEATURE_OPAL | FW_FEATURE_ULTRAVISOR,
 	FW_FEATURE_POWERNV_ALWAYS = 0,
diff --git a/arch/powerpc/platforms/pseries/firmware.c b/arch/powerpc/platforms/pseries/firmware.c
index 09c119b2f623..080108d129ed 100644
--- a/arch/powerpc/platforms/pseries/firmware.c
+++ b/arch/powerpc/platforms/pseries/firmware.c
@@ -67,6 +67,7 @@  hypertas_fw_features_table[] = {
 	{FW_FEATURE_PAPR_SCM,		"hcall-scm"},
 	{FW_FEATURE_RPT_INVALIDATE,	"hcall-rpt-invalidate"},
 	{FW_FEATURE_ENERGY_SCALE_INFO,	"hcall-energy-scale-info"},
+	{FW_FEATURE_WATCHDOG,		"hcall-watchdog"},
 };
 
 /* Build up the firmware features bitmask using the contents of
diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c
index 955ff8aa1644..6e8b67ea2a33 100644
--- a/arch/powerpc/platforms/pseries/setup.c
+++ b/arch/powerpc/platforms/pseries/setup.c
@@ -14,6 +14,7 @@ 
 
 #include <linux/cpu.h>
 #include <linux/errno.h>
+#include <linux/platform_device.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
@@ -169,6 +170,22 @@  static void __init fwnmi_init(void)
 #endif
 }
 
+/*
+ * Affix a device for the first timer to the platform bus if
+ * we have firmware support for the H_WATCHDOG hypercall.
+ */
+static struct platform_device *pseries_wdt_pdev;
+
+static __init int pseries_wdt_init(void)
+{
+	if (!firmware_has_feature(FW_FEATURE_WATCHDOG))
+		return 0;
+	pseries_wdt_pdev = platform_device_register_simple("pseries-wdt",
+							   0, NULL, 0);
+	return 0;
+}
+machine_subsys_initcall(pseries, pseries_wdt_init);
+
 static void pseries_8259_cascade(struct irq_desc *desc)
 {
 	struct irq_chip *chip = irq_desc_get_chip(desc);
diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index c4e82a8d863f..c5755f1e4540 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig
@@ -1932,6 +1932,14 @@  config MEN_A21_WDT
 
 # PPC64 Architecture
 
+config PSERIES_WDT
+	tristate "POWER Architecture Platform Watchdog Timer"
+	depends on PPC_PSERIES
+	select WATCHDOG_CORE
+	help
+	  Driver for virtual watchdog timers provided by PAPR
+	  hypervisors (e.g. pHyp, KVM).
+
 config WATCHDOG_RTAS
 	tristate "RTAS watchdog"
 	depends on PPC_RTAS
diff --git a/drivers/watchdog/Makefile b/drivers/watchdog/Makefile
index f7da867e8782..f35660409f17 100644
--- a/drivers/watchdog/Makefile
+++ b/drivers/watchdog/Makefile
@@ -184,6 +184,7 @@  obj-$(CONFIG_BOOKE_WDT) += booke_wdt.o
 obj-$(CONFIG_MEN_A21_WDT) += mena21_wdt.o
 
 # PPC64 Architecture
+obj-$(CONFIG_PSERIES_WDT) += pseries-wdt.o
 obj-$(CONFIG_WATCHDOG_RTAS) += wdrtas.o
 
 # S390 Architecture
diff --git a/drivers/watchdog/pseries-wdt.c b/drivers/watchdog/pseries-wdt.c
new file mode 100644
index 000000000000..501211bd3f26
--- /dev/null
+++ b/drivers/watchdog/pseries-wdt.c
@@ -0,0 +1,300 @@ 
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+/*
+ * Copyright (c) 2022 International Business Machines, Inc.
+ */
+
+#include <linux/bitops.h>
+#include <linux/kernel.h>
+#include <linux/limits.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/platform_device.h>
+#include <linux/watchdog.h>
+
+#define DRV_NAME "pseries-wdt"
+
+/*
+ * The PAPR's MSB->LSB bit ordering is is 0->63.  These macros
+ * simplify defining bitfields as described in the PAPR without
+ * needing to transpose values to the more C-like 63->0 ordering.
+ */
+#define SETFIELD(_v, _b, _e)	\
+    (((unsigned long)(_v) << PPC_BITLSHIFT(_e)) & PPC_BITMASK((_b), (_e)))
+#define GETFIELD(_v, _b, _e)	\
+    (((unsigned long)(_v) & PPC_BITMASK((_b), (_e))) >> PPC_BITLSHIFT(_e))
+
+/*
+ * H_WATCHDOG Hypercall Input
+ *
+ * R4: "flags":
+ *
+ *     A 64-bit value structured as follows:
+ *
+ *         Bits 0-46: Reserved (must be zero).
+ */
+#define PSERIES_WDTF_RESERVED	PPC_BITMASK(0, 46)
+
+/*
+ *         Bit 47: "leaveOtherWatchdogsRunningOnTimeout"
+ *
+ *             0  Stop outstanding watchdogs on timeout.
+ *             1  Leave outstanding watchdogs running on timeout.
+ */
+#define PSERIES_WDTF_LEAVE_OTHER	PPC_BIT(47)
+
+/*
+ *         Bits 48-55: "operation"
+ *
+ *             0x01  Start Watchdog
+ *             0x02  Stop Watchdog
+ *             0x03  Query Watchdog Capabilities
+ *             0x04  Query Watchdog LPM Requirement
+ */
+#define PSERIES_WDTF_OP(op)		SETFIELD((op), 48, 55)
+#define PSERIES_WDTF_OP_START		PSERIES_WDTF_OP(0x1)
+#define PSERIES_WDTF_OP_STOP		PSERIES_WDTF_OP(0x2)
+#define PSERIES_WDTF_OP_QUERY		PSERIES_WDTF_OP(0x3)
+#define PSERIES_WDTF_OP_QUERY_LPM	PSERIES_WDTF_OP(0x4)
+
+/*
+ *         Bits 56-63: "timeoutAction"
+ *
+ *             0x01  Hard poweroff
+ *             0x02  Hard restart
+ *             0x03  Dump restart
+ */
+#define PSERIES_WDTF_ACTION(ac)			SETFIELD(ac, 56, 63)
+#define PSERIES_WDTF_ACTION_HARD_POWEROFF	PSERIES_WDTF_ACTION(0x1)
+#define PSERIES_WDTF_ACTION_HARD_RESTART	PSERIES_WDTF_ACTION(0x2)
+#define PSERIES_WDTF_ACTION_DUMP_RESTART	PSERIES_WDTF_ACTION(0x3)
+
+/*
+ * R5: "watchdogNumber":
+ *
+ *     The target watchdog.  Watchdog numbers are 1-based.  The
+ *     maximum supported watchdog number may be obtained via the
+ *     "Query Watchdog Capabilities" operation.
+ *
+ *     This input is ignored for the "Query Watchdog Capabilities"
+ *     operation.
+ *
+ * R6: "timeoutInMs":
+ *
+ *     The timeout in milliseconds.  The minimum supported timeout may
+ *     be obtained via the "Query Watchdog Capabilities" operation.
+ *
+ *     This input is ignored for the "Stop Watchdog", "Query Watchdog
+ *     Capabilities", and "Query LPM Requirement" operations.
+ */
+
+/*
+ * H_WATCHDOG Hypercall Output
+ *
+ * R3: Return code
+ *
+ *     H_SUCCESS    The operation completed.
+ *
+ *     H_BUSY	    The hypervisor is too busy; retry the operation.
+ *
+ *     H_PARAMETER  The given "flags" are somehow invalid.  Either the
+ *                  "operation" or "timeoutAction" is invalid, or a
+ *                  reserved bit is set.
+ *
+ *     H_P2         The given "watchdogNumber" is zero or exceeds the
+ *                  supported maximum value.
+ *
+ *     H_P3         The given "timeoutInMs" is below the supported
+ *                  minimum value.
+ *
+ *     H_NOOP       The given "watchdogNumber" is already stopped.
+ *
+ *     H_HARDWARE   The operation failed for ineffable reasons.
+ *
+ *     H_FUNCTION   The H_WATCHDOG hypercall is not supported by this
+ *                  hypervisor.
+ *
+ * R4:
+ *
+ * - For the "Query Watchdog Capabilities" operation, a 64-bit
+ *   structure defined as:
+ *
+ *       Bits  0-15: The minimum supported timeout in milliseconds.
+ *       Bits 16-31: The number of watchdogs supported.
+ *       Bits 32-63: Reserved.
+ */
+#define PSERIES_WDTQ_MIN_TIMEOUT(cap)	GETFIELD((cap), 0, 15)
+#define PSERIES_WDTQ_MAX_NUMBER(cap)	GETFIELD((cap), 16, 31)
+#define PSERIES_WDTQ_RESERVED		PPC_BITMASK(32, 63)
+
+/*
+ * - For the "Query Watchdog LPM Requirement" operation:
+ *
+ *       1  The given "watchdogNumber" must be stopped prior to
+ *          suspending.
+ *
+ *       2  The given "watchdogNumber" does not have to be stopped
+ *          prior to suspending.
+ */
+#define PSERIES_WDTQL_MUST_STOP       	1
+#define PSERIES_WDTQL_NEED_NOT_STOP	2
+
+static unsigned long action = PSERIES_WDTF_ACTION_HARD_RESTART;
+static int action_set(const char *val, const struct kernel_param *kp)
+{
+	if (strcmp(val, "hard-poweroff") == 0) {
+		action = PSERIES_WDTF_ACTION_HARD_POWEROFF;
+		return 0;
+	}
+	if (strcmp(val, "hard-restart") == 0) {
+		action = PSERIES_WDTF_ACTION_HARD_RESTART;
+		return 0;
+	}
+	if (strcmp(val, "dump-restart") == 0) {
+		action = PSERIES_WDTF_ACTION_DUMP_RESTART;
+		return 0;
+	}
+	return -EINVAL;
+}
+static const struct kernel_param_ops action_ops = { .set = action_set };
+module_param_cb(action, &action_ops, NULL, S_IRUGO);
+MODULE_PARM_DESC(action, "Action taken when watchdog expires: \"hard-poweroff\", \"hard-restart\", or \"dump-restart\" (default=\"hard-restart\")");
+
+static bool nowayout = WATCHDOG_NOWAYOUT;
+module_param(nowayout, bool, S_IRUGO);
+MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once started (default=" __MODULE_STRING(WATCHDOG_NOWAYOUT) ")");
+
+#define WATCHDOG_TIMEOUT 60
+static unsigned int timeout = WATCHDOG_TIMEOUT;
+module_param(timeout, uint, S_IRUGO);
+MODULE_PARM_DESC(timeout, "Initial watchdog timeout in seconds (default=" __MODULE_STRING(WATCHDOG_TIMEOUT) ")");
+
+struct pseries_wdt {
+	struct watchdog_device wd;
+	unsigned long num;		/* NB: Watchdog numbers are 1-based */
+};
+
+static int pseries_wdt_start(struct watchdog_device *wdd)
+{
+	struct device *dev = wdd->parent;
+	struct pseries_wdt *pw = watchdog_get_drvdata(wdd);
+	unsigned long flags, msecs;
+	long rc;
+
+	flags = action | PSERIES_WDTF_OP_START;
+	msecs = wdd->timeout * 1000UL;
+	rc = plpar_hcall_norets(H_WATCHDOG, flags, pw->num, msecs);
+	if (rc != H_SUCCESS) {
+		dev_crit(dev, "H_WATCHDOG: %ld: failed to start timer %lu",
+			 rc, pw->num);
+	       	return -EIO;
+	}
+	return 0;
+}
+
+static int pseries_wdt_stop(struct watchdog_device *wdd)
+{
+	struct device *dev = wdd->parent;
+	struct pseries_wdt *pw = watchdog_get_drvdata(wdd);
+	long rc;
+
+	rc = plpar_hcall_norets(H_WATCHDOG, PSERIES_WDTF_OP_STOP, pw->num);
+	if (rc != H_SUCCESS && rc != H_NOOP) {
+		dev_crit(dev, "H_WATCHDOG: %ld: failed to stop timer %lu",
+			 rc, pw->num);
+		return -EIO;
+	}
+	return 0;
+}
+
+static struct watchdog_info pseries_wdt_info = {
+	.identity = DRV_NAME,
+	.options = WDIOF_KEEPALIVEPING | WDIOF_MAGICCLOSE | WDIOF_SETTIMEOUT \
+	    | WDIOF_PRETIMEOUT,
+};
+
+static const struct watchdog_ops pseries_wdt_ops = {
+	.owner = THIS_MODULE,
+	.ping = pseries_wdt_start,
+	.start = pseries_wdt_start,
+	.stop = pseries_wdt_stop,
+};
+
+static int pseries_wdt_probe(struct platform_device *pdev)
+{
+	unsigned long ret[PLPAR_HCALL_BUFSIZE] = { 0 };
+	unsigned long cap, min_timeout_ms;
+	long rc;
+	struct pseries_wdt *pw;
+	int err;
+
+	rc = plpar_hcall(H_WATCHDOG, ret, PSERIES_WDTF_OP_QUERY);
+	if (rc != H_SUCCESS)
+		return (rc == H_FUNCTION) ? -ENODEV : -EIO;
+	cap = ret[0];
+
+	/*
+	 * If the underlying id exceeds the number of available timers
+	 * there is a bug in the platform device registration code.
+	 */
+	if (pdev->id < 0 || pdev->id >= PSERIES_WDTQ_MAX_NUMBER(cap))
+		return -ENODEV;
+
+	pw = devm_kzalloc(&pdev->dev, sizeof *pw, GFP_KERNEL);
+	if (pw == NULL)
+		return -ENOMEM;
+	pw->num = pdev->id + 1;		/* 0-based -> 1-based */
+	pw->wd.parent = &pdev->dev;
+	pw->wd.info = &pseries_wdt_info;
+	pw->wd.ops = &pseries_wdt_ops;
+	min_timeout_ms = PSERIES_WDTQ_MIN_TIMEOUT(cap);
+	pw->wd.min_timeout = roundup(min_timeout_ms, 1000) / 1000;
+	pw->wd.max_timeout = UINT_MAX;
+	watchdog_init_timeout(&pw->wd, timeout, NULL);
+	watchdog_set_nowayout(&pw->wd, nowayout);
+	watchdog_stop_on_reboot(&pw->wd);
+	watchdog_stop_on_unregister(&pw->wd);
+	watchdog_set_drvdata(&pw->wd, pw);
+
+	err = devm_watchdog_register_device(&pdev->dev, &pw->wd);
+	if (err)
+		return err;
+
+	platform_set_drvdata(pdev, &pw->wd);
+
+	return 0;
+}
+
+static int pseries_wdt_suspend(struct platform_device *pdev, pm_message_t state)
+{
+	struct watchdog_device *wd = platform_get_drvdata(pdev);
+
+	if (watchdog_active(wd))
+		return pseries_wdt_stop(wd);
+	return 0;
+}
+
+static int pseries_wdt_resume(struct platform_device *pdev)
+{
+	struct watchdog_device *wd = platform_get_drvdata(pdev);
+
+	if (watchdog_active(wd))
+		return pseries_wdt_start(wd);
+	return 0;
+}
+
+static struct platform_driver pseries_wdt_driver = {
+	.driver = {
+		.name = DRV_NAME,
+		.owner = THIS_MODULE,
+	},
+	.probe = pseries_wdt_probe,
+	.resume = pseries_wdt_resume,
+	.suspend = pseries_wdt_suspend,
+};
+module_platform_driver(pseries_wdt_driver);
+
+MODULE_AUTHOR("Alexey Kardashevskiy <aik@ozlabs.ru>");
+MODULE_AUTHOR("Scott Cheloha <cheloha@linux.ibm.com>");
+MODULE_DESCRIPTION("POWER Architecture Platform Watchdog Driver");
+MODULE_LICENSE("GPL");