A rogue wakeup source not registered in wakeup_sources list is not visible
from wakeup_sources_stats_show. Check if the wakeup source is registered
properly by looking at the timer function.
Signed-off-by: Jin Qian <[email protected]>
---
drivers/base/power/wakeup.c | 19 +++++++++++++++++++
1 file changed, 19 insertions(+)
diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
index 7726200..f24c622 100644
--- a/drivers/base/power/wakeup.c
+++ b/drivers/base/power/wakeup.c
@@ -14,6 +14,7 @@
#include <linux/suspend.h>
#include <linux/seq_file.h>
#include <linux/debugfs.h>
+#include <linux/types.h>
#include <trace/events/power.h>
#include "power.h"
@@ -351,6 +352,19 @@ int device_set_wakeup_enable(struct device *dev, bool enable)
}
EXPORT_SYMBOL_GPL(device_set_wakeup_enable);
+/**
+ * validate_wakeup_source - validate the given wakeup source.
+ * @ws: Wakeup source to be validated.
+ */
+static bool validate_wakeup_source(struct wakeup_source *ws)
+{
+ /*
+ * Use timer function to check if the given source is initialized
+ * by wakeup_source_add.
+ */
+ return ws->timer.function == pm_wakeup_timer_fn;
+}
+
/*
* The functions below use the observation that each wakeup event starts a
* period in which the system should not be suspended. The moment this period
@@ -391,6 +405,11 @@ static void wakeup_source_activate(struct wakeup_source *ws)
{
unsigned int cec;
+ if (!validate_wakeup_source(ws)) {
+ pr_err("unregistered wakeup source: %s\n", ws->name);
+ return;
+ }
+
/*
* active wakeup source should bring the system
* out of PM_SUSPEND_FREEZE state
--
2.2.0.rc0.207.ga3a616c
On Wed 2015-04-22 17:50:10, Jin Qian wrote:
> A rogue wakeup source not registered in wakeup_sources list is not visible
> from wakeup_sources_stats_show. Check if the wakeup source is registered
> properly by looking at the timer function.
How often does that happen?
Will people on your systems actually notice that pr_err()?
Pavel
> Signed-off-by: Jin Qian <[email protected]>
> ---
> drivers/base/power/wakeup.c | 19 +++++++++++++++++++
> 1 file changed, 19 insertions(+)
>
> diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
> index 7726200..f24c622 100644
> --- a/drivers/base/power/wakeup.c
> +++ b/drivers/base/power/wakeup.c
> @@ -14,6 +14,7 @@
> #include <linux/suspend.h>
> #include <linux/seq_file.h>
> #include <linux/debugfs.h>
> +#include <linux/types.h>
> #include <trace/events/power.h>
>
> #include "power.h"
> @@ -351,6 +352,19 @@ int device_set_wakeup_enable(struct device *dev, bool enable)
> }
> EXPORT_SYMBOL_GPL(device_set_wakeup_enable);
>
> +/**
> + * validate_wakeup_source - validate the given wakeup source.
> + * @ws: Wakeup source to be validated.
> + */
> +static bool validate_wakeup_source(struct wakeup_source *ws)
> +{
> + /*
> + * Use timer function to check if the given source is initialized
> + * by wakeup_source_add.
> + */
> + return ws->timer.function == pm_wakeup_timer_fn;
> +}
> +
> /*
> * The functions below use the observation that each wakeup event starts a
> * period in which the system should not be suspended. The moment this period
> @@ -391,6 +405,11 @@ static void wakeup_source_activate(struct wakeup_source *ws)
> {
> unsigned int cec;
>
> + if (!validate_wakeup_source(ws)) {
> + pr_err("unregistered wakeup source: %s\n", ws->name);
> + return;
> + }
> +
> /*
> * active wakeup source should bring the system
> * out of PM_SUSPEND_FREEZE state
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
This is a rare case.
My thought is to prevent a rogue wakeup source from preventing system
suspend, which is very difficult to find out if the wakeup source is
not registered and not visible in /d/wakeup_sources. On the other
hand, if the wakeup source failed to activate, it's much easier to be
noticed considering we deal with particular wakeup source. I can make
the pr_err BUG_ON or WARN_ON if they're more noticeable.
Thanks,
jin
On Sat, Apr 25, 2015 at 1:12 PM, Pavel Machek <[email protected]> wrote:
> On Wed 2015-04-22 17:50:10, Jin Qian wrote:
>> A rogue wakeup source not registered in wakeup_sources list is not visible
>> from wakeup_sources_stats_show. Check if the wakeup source is registered
>> properly by looking at the timer function.
>
> How often does that happen?
>
> Will people on your systems actually notice that pr_err()?
> Pavel
>
>> Signed-off-by: Jin Qian <[email protected]>
>> ---
>> drivers/base/power/wakeup.c | 19 +++++++++++++++++++
>> 1 file changed, 19 insertions(+)
>>
>> diff --git a/drivers/base/power/wakeup.c b/drivers/base/power/wakeup.c
>> index 7726200..f24c622 100644
>> --- a/drivers/base/power/wakeup.c
>> +++ b/drivers/base/power/wakeup.c
>> @@ -14,6 +14,7 @@
>> #include <linux/suspend.h>
>> #include <linux/seq_file.h>
>> #include <linux/debugfs.h>
>> +#include <linux/types.h>
>> #include <trace/events/power.h>
>>
>> #include "power.h"
>> @@ -351,6 +352,19 @@ int device_set_wakeup_enable(struct device *dev, bool enable)
>> }
>> EXPORT_SYMBOL_GPL(device_set_wakeup_enable);
>>
>> +/**
>> + * validate_wakeup_source - validate the given wakeup source.
>> + * @ws: Wakeup source to be validated.
>> + */
>> +static bool validate_wakeup_source(struct wakeup_source *ws)
>> +{
>> + /*
>> + * Use timer function to check if the given source is initialized
>> + * by wakeup_source_add.
>> + */
>> + return ws->timer.function == pm_wakeup_timer_fn;
>> +}
>> +
>> /*
>> * The functions below use the observation that each wakeup event starts a
>> * period in which the system should not be suspended. The moment this period
>> @@ -391,6 +405,11 @@ static void wakeup_source_activate(struct wakeup_source *ws)
>> {
>> unsigned int cec;
>>
>> + if (!validate_wakeup_source(ws)) {
>> + pr_err("unregistered wakeup source: %s\n", ws->name);
>> + return;
>> + }
>> +
>> /*
>> * active wakeup source should bring the system
>> * out of PM_SUSPEND_FREEZE state
>
> --
> (english) http://www.livejournal.com/~pavelmachek
> (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html