* [PATCH v2 0/4] JZ4740 watchdog cleanups
@ 2019-06-07 16:24 Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 1/4] watchdog: jz4740: Use register names from <linux/mfd/ingenic-tcu.h> Paul Cercueil
` (3 more replies)
0 siblings, 4 replies; 10+ messages in thread
From: Paul Cercueil @ 2019-06-07 16:24 UTC (permalink / raw)
To: Guenter Roeck, Wim Van Sebroeck; +Cc: od, linux-watchdog, linux-kernel
V1 patchset didn't apply anymore on v5.2-rc3 so I rebased on top of it.
The previous patch 4/4 (which added SPDX license notifier) has been
dropped since somebody else did the job in -rc3.
So I added another cleanup as patch 4/4, and I tweaked patch 2/4 to
adjust to Guenter's feedback on V1.
Cheers
-Paul
^ permalink raw reply [flat|nested] 10+ messages in thread
* [PATCH v2 1/4] watchdog: jz4740: Use register names from <linux/mfd/ingenic-tcu.h>
2019-06-07 16:24 [PATCH v2 0/4] JZ4740 watchdog cleanups Paul Cercueil
@ 2019-06-07 16:24 ` Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 2/4] watchdog: jz4740: Avoid starting watchdog in set_timeout Paul Cercueil
` (2 subsequent siblings)
3 siblings, 0 replies; 10+ messages in thread
From: Paul Cercueil @ 2019-06-07 16:24 UTC (permalink / raw)
To: Guenter Roeck, Wim Van Sebroeck
Cc: od, linux-watchdog, linux-kernel, Paul Cercueil
Use the macros from <linux/mfd/ingenic-tcu.h> instead of declaring our
own.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
---
Notes:
v2: Rebased on v5.2-rc3
drivers/watchdog/jz4740_wdt.c | 39 ++++++++++++++---------------------
1 file changed, 16 insertions(+), 23 deletions(-)
diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
index 313358b2e0b1..a2592c4fff02 100644
--- a/drivers/watchdog/jz4740_wdt.c
+++ b/drivers/watchdog/jz4740_wdt.c
@@ -4,6 +4,7 @@
* JZ4740 Watchdog driver
*/
+#include <linux/mfd/ingenic-tcu.h>
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/types.h>
@@ -19,23 +20,16 @@
#include <asm/mach-jz4740/timer.h>
-#define JZ_REG_WDT_TIMER_DATA 0x0
-#define JZ_REG_WDT_COUNTER_ENABLE 0x4
-#define JZ_REG_WDT_TIMER_COUNTER 0x8
-#define JZ_REG_WDT_TIMER_CONTROL 0xC
-
#define JZ_WDT_CLOCK_PCLK 0x1
#define JZ_WDT_CLOCK_RTC 0x2
#define JZ_WDT_CLOCK_EXT 0x4
-#define JZ_WDT_CLOCK_DIV_SHIFT 3
-
-#define JZ_WDT_CLOCK_DIV_1 (0 << JZ_WDT_CLOCK_DIV_SHIFT)
-#define JZ_WDT_CLOCK_DIV_4 (1 << JZ_WDT_CLOCK_DIV_SHIFT)
-#define JZ_WDT_CLOCK_DIV_16 (2 << JZ_WDT_CLOCK_DIV_SHIFT)
-#define JZ_WDT_CLOCK_DIV_64 (3 << JZ_WDT_CLOCK_DIV_SHIFT)
-#define JZ_WDT_CLOCK_DIV_256 (4 << JZ_WDT_CLOCK_DIV_SHIFT)
-#define JZ_WDT_CLOCK_DIV_1024 (5 << JZ_WDT_CLOCK_DIV_SHIFT)
+#define JZ_WDT_CLOCK_DIV_1 (0 << TCU_TCSR_PRESCALE_LSB)
+#define JZ_WDT_CLOCK_DIV_4 (1 << TCU_TCSR_PRESCALE_LSB)
+#define JZ_WDT_CLOCK_DIV_16 (2 << TCU_TCSR_PRESCALE_LSB)
+#define JZ_WDT_CLOCK_DIV_64 (3 << TCU_TCSR_PRESCALE_LSB)
+#define JZ_WDT_CLOCK_DIV_256 (4 << TCU_TCSR_PRESCALE_LSB)
+#define JZ_WDT_CLOCK_DIV_1024 (5 << TCU_TCSR_PRESCALE_LSB)
#define DEFAULT_HEARTBEAT 5
#define MAX_HEARTBEAT 2048
@@ -63,7 +57,7 @@ static int jz4740_wdt_ping(struct watchdog_device *wdt_dev)
{
struct jz4740_wdt_drvdata *drvdata = watchdog_get_drvdata(wdt_dev);
- writew(0x0, drvdata->base + JZ_REG_WDT_TIMER_COUNTER);
+ writew(0x0, drvdata->base + TCU_REG_WDT_TCNT);
return 0;
}
@@ -86,18 +80,17 @@ static int jz4740_wdt_set_timeout(struct watchdog_device *wdt_dev,
break;
}
timeout_value >>= 2;
- clock_div += (1 << JZ_WDT_CLOCK_DIV_SHIFT);
+ clock_div += (1 << TCU_TCSR_PRESCALE_LSB);
}
- writeb(0x0, drvdata->base + JZ_REG_WDT_COUNTER_ENABLE);
- writew(clock_div, drvdata->base + JZ_REG_WDT_TIMER_CONTROL);
+ writeb(0x0, drvdata->base + TCU_REG_WDT_TCER);
+ writew(clock_div, drvdata->base + TCU_REG_WDT_TCSR);
- writew((u16)timeout_value, drvdata->base + JZ_REG_WDT_TIMER_DATA);
- writew(0x0, drvdata->base + JZ_REG_WDT_TIMER_COUNTER);
- writew(clock_div | JZ_WDT_CLOCK_RTC,
- drvdata->base + JZ_REG_WDT_TIMER_CONTROL);
+ writew((u16)timeout_value, drvdata->base + TCU_REG_WDT_TDR);
+ writew(0x0, drvdata->base + TCU_REG_WDT_TCNT);
+ writew(clock_div | JZ_WDT_CLOCK_RTC, drvdata->base + TCU_REG_WDT_TCSR);
- writeb(0x1, drvdata->base + JZ_REG_WDT_COUNTER_ENABLE);
+ writeb(0x1, drvdata->base + TCU_REG_WDT_TCER);
wdt_dev->timeout = new_timeout;
return 0;
@@ -115,7 +108,7 @@ static int jz4740_wdt_stop(struct watchdog_device *wdt_dev)
{
struct jz4740_wdt_drvdata *drvdata = watchdog_get_drvdata(wdt_dev);
- writeb(0x0, drvdata->base + JZ_REG_WDT_COUNTER_ENABLE);
+ writeb(0x0, drvdata->base + TCU_REG_WDT_TCER);
jz4740_timer_disable_watchdog();
return 0;
--
2.21.0.593.g511ec345e18
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH v2 2/4] watchdog: jz4740: Avoid starting watchdog in set_timeout
2019-06-07 16:24 [PATCH v2 0/4] JZ4740 watchdog cleanups Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 1/4] watchdog: jz4740: Use register names from <linux/mfd/ingenic-tcu.h> Paul Cercueil
@ 2019-06-07 16:24 ` Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 3/4] watchdog: jz4740: Drop dependency on MACH_JZ47xx Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module Paul Cercueil
3 siblings, 0 replies; 10+ messages in thread
From: Paul Cercueil @ 2019-06-07 16:24 UTC (permalink / raw)
To: Guenter Roeck, Wim Van Sebroeck
Cc: od, linux-watchdog, linux-kernel, Paul Cercueil
Previously the jz4740_wdt_set_timeout() function was starting the timer
unconditionally, even if it was stopped when that function was entered.
Now, the timer will be restarted only if it was already running before
this function is called.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
---
Notes:
v2: - Don't disable TCU timer if it was already disabled
- Don't enable TCU timer if it was already enabled
drivers/watchdog/jz4740_wdt.c | 14 +++++++++++++-
1 file changed, 13 insertions(+), 1 deletion(-)
diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
index a2592c4fff02..7519d80c5d05 100644
--- a/drivers/watchdog/jz4740_wdt.c
+++ b/drivers/watchdog/jz4740_wdt.c
@@ -68,6 +68,7 @@ static int jz4740_wdt_set_timeout(struct watchdog_device *wdt_dev,
unsigned int rtc_clk_rate;
unsigned int timeout_value;
unsigned short clock_div = JZ_WDT_CLOCK_DIV_1;
+ u8 tcer;
rtc_clk_rate = clk_get_rate(drvdata->rtc_clk);
@@ -83,6 +84,7 @@ static int jz4740_wdt_set_timeout(struct watchdog_device *wdt_dev,
clock_div += (1 << TCU_TCSR_PRESCALE_LSB);
}
+ tcer = readb(drvdata->base + TCU_REG_WDT_TCER);
writeb(0x0, drvdata->base + TCU_REG_WDT_TCER);
writew(clock_div, drvdata->base + TCU_REG_WDT_TCSR);
@@ -90,7 +92,8 @@ static int jz4740_wdt_set_timeout(struct watchdog_device *wdt_dev,
writew(0x0, drvdata->base + TCU_REG_WDT_TCNT);
writew(clock_div | JZ_WDT_CLOCK_RTC, drvdata->base + TCU_REG_WDT_TCSR);
- writeb(0x1, drvdata->base + TCU_REG_WDT_TCER);
+ if (tcer & TCU_WDT_TCER_TCEN)
+ writeb(TCU_WDT_TCER_TCEN, drvdata->base + TCU_REG_WDT_TCER);
wdt_dev->timeout = new_timeout;
return 0;
@@ -98,9 +101,18 @@ static int jz4740_wdt_set_timeout(struct watchdog_device *wdt_dev,
static int jz4740_wdt_start(struct watchdog_device *wdt_dev)
{
+ struct jz4740_wdt_drvdata *drvdata = watchdog_get_drvdata(wdt_dev);
+ u8 tcer;
+
+ tcer = readb(drvdata->base + TCU_REG_WDT_TCER);
+
jz4740_timer_enable_watchdog();
jz4740_wdt_set_timeout(wdt_dev, wdt_dev->timeout);
+ /* Start watchdog if it wasn't started already */
+ if (!(tcer & TCU_WDT_TCER_TCEN))
+ writeb(TCU_WDT_TCER_TCEN, drvdata->base + TCU_REG_WDT_TCER);
+
return 0;
}
--
2.21.0.593.g511ec345e18
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH v2 3/4] watchdog: jz4740: Drop dependency on MACH_JZ47xx
2019-06-07 16:24 [PATCH v2 0/4] JZ4740 watchdog cleanups Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 1/4] watchdog: jz4740: Use register names from <linux/mfd/ingenic-tcu.h> Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 2/4] watchdog: jz4740: Avoid starting watchdog in set_timeout Paul Cercueil
@ 2019-06-07 16:24 ` Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module Paul Cercueil
3 siblings, 0 replies; 10+ messages in thread
From: Paul Cercueil @ 2019-06-07 16:24 UTC (permalink / raw)
To: Guenter Roeck, Wim Van Sebroeck
Cc: od, linux-watchdog, linux-kernel, Paul Cercueil
Depending on MACH_JZ47xx prevent us from creating a generic kernel that
works on more than one MIPS board. Instead, we just depend on MIPS being
set.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
---
Notes:
v2: Rebased on v5.2-rc3
drivers/watchdog/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index ffe754539f5a..20dd9efc9416 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig
@@ -1634,7 +1634,7 @@ config INDYDOG
config JZ4740_WDT
tristate "Ingenic jz4740 SoC hardware watchdog"
- depends on MACH_JZ4740 || MACH_JZ4780
+ depends on MIPS
select WATCHDOG_CORE
help
Hardware driver for the built-in watchdog timer on Ingenic jz4740 SoCs.
--
2.21.0.593.g511ec345e18
^ permalink raw reply related [flat|nested] 10+ messages in thread
* [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-07 16:24 [PATCH v2 0/4] JZ4740 watchdog cleanups Paul Cercueil
` (2 preceding siblings ...)
2019-06-07 16:24 ` [PATCH v2 3/4] watchdog: jz4740: Drop dependency on MACH_JZ47xx Paul Cercueil
@ 2019-06-07 16:24 ` Paul Cercueil
2019-06-07 17:29 ` Guenter Roeck
2019-06-08 7:51 ` Christophe Leroy
3 siblings, 2 replies; 10+ messages in thread
From: Paul Cercueil @ 2019-06-07 16:24 UTC (permalink / raw)
To: Guenter Roeck, Wim Van Sebroeck
Cc: od, linux-watchdog, linux-kernel, Paul Cercueil
This allows the probe function to be dropped after the kernel finished
its initialization, in the case where the driver was not compiled as a
module.
Signed-off-by: Paul Cercueil <paul@crapouillou.net>
---
Notes:
v2: New patch
drivers/watchdog/jz4740_wdt.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
index 7519d80c5d05..2061788c1939 100644
--- a/drivers/watchdog/jz4740_wdt.c
+++ b/drivers/watchdog/jz4740_wdt.c
@@ -157,7 +157,7 @@ static const struct of_device_id jz4740_wdt_of_matches[] = {
MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
#endif
-static int jz4740_wdt_probe(struct platform_device *pdev)
+static int __init_or_module jz4740_wdt_probe(struct platform_device *pdev)
{
struct device *dev = &pdev->dev;
struct jz4740_wdt_drvdata *drvdata;
--
2.21.0.593.g511ec345e18
^ permalink raw reply related [flat|nested] 10+ messages in thread
* Re: [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-07 16:24 ` [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module Paul Cercueil
@ 2019-06-07 17:29 ` Guenter Roeck
2019-06-08 7:51 ` Christophe Leroy
1 sibling, 0 replies; 10+ messages in thread
From: Guenter Roeck @ 2019-06-07 17:29 UTC (permalink / raw)
To: Paul Cercueil; +Cc: Wim Van Sebroeck, od, linux-watchdog, linux-kernel
On Fri, Jun 07, 2019 at 06:24:29PM +0200, Paul Cercueil wrote:
> This allows the probe function to be dropped after the kernel finished
> its initialization, in the case where the driver was not compiled as a
> module.
>
> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
> ---
>
> Notes:
> v2: New patch
>
> drivers/watchdog/jz4740_wdt.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
> index 7519d80c5d05..2061788c1939 100644
> --- a/drivers/watchdog/jz4740_wdt.c
> +++ b/drivers/watchdog/jz4740_wdt.c
> @@ -157,7 +157,7 @@ static const struct of_device_id jz4740_wdt_of_matches[] = {
> MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
> #endif
>
> -static int jz4740_wdt_probe(struct platform_device *pdev)
> +static int __init_or_module jz4740_wdt_probe(struct platform_device *pdev)
> {
> struct device *dev = &pdev->dev;
> struct jz4740_wdt_drvdata *drvdata;
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-07 16:24 ` [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module Paul Cercueil
2019-06-07 17:29 ` Guenter Roeck
@ 2019-06-08 7:51 ` Christophe Leroy
2019-06-08 9:57 ` Paul Cercueil
1 sibling, 1 reply; 10+ messages in thread
From: Christophe Leroy @ 2019-06-08 7:51 UTC (permalink / raw)
To: Paul Cercueil, Guenter Roeck, Wim Van Sebroeck
Cc: od, linux-watchdog, linux-kernel
Hi Paul,
Le 07/06/2019 à 18:24, Paul Cercueil a écrit :
> This allows the probe function to be dropped after the kernel finished
> its initialization, in the case where the driver was not compiled as a
> module.
I'm not sure that's what __init_or_module flag does.
As far as I understand, this flag makes the function being dropped only
when the kernel is built without modules support, ie without
CONFIG_MODULES. See
https://elixir.bootlin.com/linux/latest/source/include/linux/module.h#L145
In addition, I'm not sure you can simply define a probe function as
__init. What if someone tries to unbind and rebind the device through
sysfs for instance ?
It seems there is a special function called __platform_driver_probe()
for registering devices when the probe function is to be in __init, see
https://elixir.bootlin.com/linux/latest/source/drivers/base/platform.c#L684
Christophe
>
> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
> ---
>
> Notes:
> v2: New patch
>
> drivers/watchdog/jz4740_wdt.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
> index 7519d80c5d05..2061788c1939 100644
> --- a/drivers/watchdog/jz4740_wdt.c
> +++ b/drivers/watchdog/jz4740_wdt.c
> @@ -157,7 +157,7 @@ static const struct of_device_id jz4740_wdt_of_matches[] = {
> MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
> #endif
>
> -static int jz4740_wdt_probe(struct platform_device *pdev)
> +static int __init_or_module jz4740_wdt_probe(struct platform_device *pdev)
> {
> struct device *dev = &pdev->dev;
> struct jz4740_wdt_drvdata *drvdata;
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-08 7:51 ` Christophe Leroy
@ 2019-06-08 9:57 ` Paul Cercueil
2019-06-11 5:16 ` Christophe Leroy
0 siblings, 1 reply; 10+ messages in thread
From: Paul Cercueil @ 2019-06-08 9:57 UTC (permalink / raw)
To: Christophe Leroy
Cc: Guenter Roeck, Wim Van Sebroeck, od, linux-watchdog, linux-kernel
Hi Christophe,
Le sam. 8 juin 2019 à 9:51, Christophe Leroy <christophe.leroy@c-s.fr>
a écrit :
> Hi Paul,
>
> Le 07/06/2019 à 18:24, Paul Cercueil a écrit :
>> This allows the probe function to be dropped after the kernel
>> finished
>> its initialization, in the case where the driver was not compiled as
>> a
>> module.
>
> I'm not sure that's what __init_or_module flag does.
>
> As far as I understand, this flag makes the function being dropped
> only when the kernel is built without modules support, ie without
> CONFIG_MODULES. See
> https://elixir.bootlin.com/linux/latest/source/include/linux/module.h#L145
So it doesn't depend on the driver being built-in or compiled as a
module?
> In addition, I'm not sure you can simply define a probe function as
> __init. What if someone tries to unbind and rebind the device through
> sysfs for instance ?
Ouch. I feel stupid now.
> It seems there is a special function called __platform_driver_probe()
> for registering devices when the probe function is to be in __init,
> see
> https://elixir.bootlin.com/linux/latest/source/drivers/base/platform.c#L684
Yes, but only usable by drivers that won't defer probe, and it removes
the bind/unbind attributes from sysfs,
so it shouldn't be used for non-critical drivers, I think.
> Christophe
>
>>
>> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>> ---
>>
>> Notes:
>> v2: New patch
>>
>> drivers/watchdog/jz4740_wdt.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/watchdog/jz4740_wdt.c
>> b/drivers/watchdog/jz4740_wdt.c
>> index 7519d80c5d05..2061788c1939 100644
>> --- a/drivers/watchdog/jz4740_wdt.c
>> +++ b/drivers/watchdog/jz4740_wdt.c
>> @@ -157,7 +157,7 @@ static const struct of_device_id
>> jz4740_wdt_of_matches[] = {
>> MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
>> #endif
>> \x7f-static int jz4740_wdt_probe(struct platform_device *pdev)
>> +static int __init_or_module jz4740_wdt_probe(struct platform_device
>> *pdev)
>> {
>> struct device *dev = &pdev->dev;
>> struct jz4740_wdt_drvdata *drvdata;
>>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-08 9:57 ` Paul Cercueil
@ 2019-06-11 5:16 ` Christophe Leroy
2019-06-11 11:02 ` Guenter Roeck
0 siblings, 1 reply; 10+ messages in thread
From: Christophe Leroy @ 2019-06-11 5:16 UTC (permalink / raw)
To: Paul Cercueil
Cc: Guenter Roeck, Wim Van Sebroeck, od, linux-watchdog, linux-kernel
Hi Paul,
Le 08/06/2019 à 11:57, Paul Cercueil a écrit :
> Hi Christophe,
>
> Le sam. 8 juin 2019 à 9:51, Christophe Leroy <christophe.leroy@c-s.fr> a
> écrit :
>> Hi Paul,
>>
>> Le 07/06/2019 à 18:24, Paul Cercueil a écrit :
>>> This allows the probe function to be dropped after the kernel finished
>>> its initialization, in the case where the driver was not compiled as a
>>> module.
>>
>> I'm not sure that's what __init_or_module flag does.
>>
>> As far as I understand, this flag makes the function being dropped
>> only when the kernel is built without modules support, ie without
>> CONFIG_MODULES. See
>> https://elixir.bootlin.com/linux/latest/source/include/linux/module.h#L145
>>
>
> So it doesn't depend on the driver being built-in or compiled as a module?
No it doesn't. This flag is for built-in functions that are needed by
init and modules init. If the kernel doesn't support modules, it can
drop the function after init. If the kernel supports modules, it has to
keep the function. That's what this flag is made for.
If your need is to mark a function so that it gets discarded after init
or module init, just mark it __init. If it is built in, it will be
dropped after init. If it is in a module, it will be dropped after the
module is initialised.
>
>> In addition, I'm not sure you can simply define a probe function as
>> __init. What if someone tries to unbind and rebind the device through
>> sysfs for instance ?
>
> Ouch. I feel stupid now.
>
>> It seems there is a special function called __platform_driver_probe()
>> for registering devices when the probe function is to be in __init,
>> see
>> https://elixir.bootlin.com/linux/latest/source/drivers/base/platform.c#L684
>>
>
> Yes, but only usable by drivers that won't defer probe, and it removes
> the bind/unbind attributes from sysfs,
> so it shouldn't be used for non-critical drivers, I think.
I guess it would make sense for watchdog drivers, we don't expect this
kind of driver to be unbinded, do we ?
Christophe
>
>> Christophe
>>
>>>
>>> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>>> ---
>>>
>>> Notes:
>>> v2: New patch
>>>
>>> drivers/watchdog/jz4740_wdt.c | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/drivers/watchdog/jz4740_wdt.c
>>> b/drivers/watchdog/jz4740_wdt.c
>>> index 7519d80c5d05..2061788c1939 100644
>>> --- a/drivers/watchdog/jz4740_wdt.c
>>> +++ b/drivers/watchdog/jz4740_wdt.c
>>> @@ -157,7 +157,7 @@ static const struct of_device_id
>>> jz4740_wdt_of_matches[] = {
>>> MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
>>> #endif
>>> \x7f-static int jz4740_wdt_probe(struct platform_device *pdev)
>>> +static int __init_or_module jz4740_wdt_probe(struct platform_device
>>> *pdev)
>>> {
>>> struct device *dev = &pdev->dev;
>>> struct jz4740_wdt_drvdata *drvdata;
>>>
>
^ permalink raw reply [flat|nested] 10+ messages in thread
* Re: [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module
2019-06-11 5:16 ` Christophe Leroy
@ 2019-06-11 11:02 ` Guenter Roeck
0 siblings, 0 replies; 10+ messages in thread
From: Guenter Roeck @ 2019-06-11 11:02 UTC (permalink / raw)
To: Christophe Leroy, Paul Cercueil
Cc: Wim Van Sebroeck, od, linux-watchdog, linux-kernel
On 6/10/19 10:16 PM, Christophe Leroy wrote:
> Hi Paul,
>
> Le 08/06/2019 à 11:57, Paul Cercueil a écrit :
>> Hi Christophe,
>>
>> Le sam. 8 juin 2019 à 9:51, Christophe Leroy <christophe.leroy@c-s.fr> a écrit :
>>> Hi Paul,
>>>
>>> Le 07/06/2019 à 18:24, Paul Cercueil a écrit :
>>>> This allows the probe function to be dropped after the kernel finished
>>>> its initialization, in the case where the driver was not compiled as a
>>>> module.
>>>
>>> I'm not sure that's what __init_or_module flag does.
>>>
>>> As far as I understand, this flag makes the function being dropped only when the kernel is built without modules support, ie without CONFIG_MODULES. See https://elixir.bootlin.com/linux/latest/source/include/linux/module.h#L145
>>
>> So it doesn't depend on the driver being built-in or compiled as a module?
>
> No it doesn't. This flag is for built-in functions that are needed by init and modules init. If the kernel doesn't support modules, it can drop the function after init. If the kernel supports modules, it has to keep the function. That's what this flag is made for.
>
> If your need is to mark a function so that it gets discarded after init or module init, just mark it __init. If it is built in, it will be dropped after init. If it is in a module, it will be dropped after the module is initialised.
>
>>
>>> In addition, I'm not sure you can simply define a probe function as __init. What if someone tries to unbind and rebind the device through sysfs for instance ?
>>
>> Ouch. I feel stupid now.
>>
>>> It seems there is a special function called __platform_driver_probe() for registering devices when the probe function is to be in __init, see https://elixir.bootlin.com/linux/latest/source/drivers/base/platform.c#L684
>>
>> Yes, but only usable by drivers that won't defer probe, and it removes the bind/unbind attributes from sysfs,
>> so it shouldn't be used for non-critical drivers, I think.
>
> I guess it would make sense for watchdog drivers, we don't expect this kind of driver to be unbinded, do we ?
>
Except for debugging, where it can be extremely useful to be able to do it.
Guenter
> Christophe
>
>>
>>> Christophe
>>>
>>>>
>>>> Signed-off-by: Paul Cercueil <paul@crapouillou.net>
>>>> ---
>>>>
>>>> Notes:
>>>> v2: New patch
>>>>
>>>> drivers/watchdog/jz4740_wdt.c | 2 +-
>>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>>
>>>> diff --git a/drivers/watchdog/jz4740_wdt.c b/drivers/watchdog/jz4740_wdt.c
>>>> index 7519d80c5d05..2061788c1939 100644
>>>> --- a/drivers/watchdog/jz4740_wdt.c
>>>> +++ b/drivers/watchdog/jz4740_wdt.c
>>>> @@ -157,7 +157,7 @@ static const struct of_device_id jz4740_wdt_of_matches[] = {
>>>> MODULE_DEVICE_TABLE(of, jz4740_wdt_of_matches);
>>>> #endif
>>>> \x7f-static int jz4740_wdt_probe(struct platform_device *pdev)
>>>> +static int __init_or_module jz4740_wdt_probe(struct platform_device *pdev)
>>>> {
>>>> struct device *dev = &pdev->dev;
>>>> struct jz4740_wdt_drvdata *drvdata;
>>>>
>>
>
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2019-06-11 11:02 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-06-07 16:24 [PATCH v2 0/4] JZ4740 watchdog cleanups Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 1/4] watchdog: jz4740: Use register names from <linux/mfd/ingenic-tcu.h> Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 2/4] watchdog: jz4740: Avoid starting watchdog in set_timeout Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 3/4] watchdog: jz4740: Drop dependency on MACH_JZ47xx Paul Cercueil
2019-06-07 16:24 ` [PATCH v2 4/4] watchdog: jz4740: Make probe function __init_or_module Paul Cercueil
2019-06-07 17:29 ` Guenter Roeck
2019-06-08 7:51 ` Christophe Leroy
2019-06-08 9:57 ` Paul Cercueil
2019-06-11 5:16 ` Christophe Leroy
2019-06-11 11:02 ` Guenter Roeck
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).