It seems that when the driver is built-in, the HID bus is
initialized after the driver is loaded, which whould cause
module_hid_driver() to fail.
Fix this by registering the driver after the HID bus using
late_initcall() in accordance with other hwmon HID drivers.
Compile-tested only.
Signed-off-by: Armin Wolf <[email protected]>
---
drivers/hwmon/corsair-psu.c | 18 +++++++++++++++++-
1 file changed, 17 insertions(+), 1 deletion(-)
diff --git a/drivers/hwmon/corsair-psu.c b/drivers/hwmon/corsair-psu.c
index 904890598c11..2c7c92272fe3 100644
--- a/drivers/hwmon/corsair-psu.c
+++ b/drivers/hwmon/corsair-psu.c
@@ -899,7 +899,23 @@ static struct hid_driver corsairpsu_driver = {
.reset_resume = corsairpsu_resume,
#endif
};
-module_hid_driver(corsairpsu_driver);
+
+static int __init corsair_init(void)
+{
+ return hid_register_driver(&corsairpsu_driver);
+}
+
+static void __exit corsair_exit(void)
+{
+ hid_unregister_driver(&corsairpsu_driver);
+}
+
+/*
+ * With module_init() the driver would load before the HID bus when
+ * built-in, so use late_initcall() instead.
+ */
+late_initcall(corsair_init);
+module_exit(corsair_exit);
MODULE_LICENSE("GPL");
MODULE_AUTHOR("Wilken Gottwalt <[email protected]>");
--
2.39.2
On Thu, 7 Dec 2023 22:07:23 +0100
Armin Wolf <[email protected]> wrote:
> It seems that when the driver is built-in, the HID bus is
> initialized after the driver is loaded, which whould cause
> module_hid_driver() to fail.
> Fix this by registering the driver after the HID bus using
> late_initcall() in accordance with other hwmon HID drivers.
>
> Compile-tested only.
So you did not test this? Well, I did.
[ 2.225831] Driver 'corsair-psu' was unable to register with bus_type 'hid' because the bus was not initialized.
[ 2.225835] amd_pstate: driver load is disabled, boot with specific mode to enable this
[ 2.226363] ledtrig-cpu: registered to indicate activity on CPUs
[ 2.226679] hid: raw HID events driver (C) Jiri Kosina
You are right, it is a timing issue and this can actually happen. I'm fine with
the fix.
Though, this could even be a bigger issue. There are currently 104 HID drivers
using the module_hid_driver macro. Maybe it would be a better idea to change the
module_hid_driver macro to use the lateinit calls instead of the plain init/exit
calls.
greetings,
Will
> Signed-off-by: Armin Wolf <[email protected]>
> ---
> drivers/hwmon/corsair-psu.c | 18 +++++++++++++++++-
> 1 file changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/hwmon/corsair-psu.c b/drivers/hwmon/corsair-psu.c
> index 904890598c11..2c7c92272fe3 100644
> --- a/drivers/hwmon/corsair-psu.c
> +++ b/drivers/hwmon/corsair-psu.c
> @@ -899,7 +899,23 @@ static struct hid_driver corsairpsu_driver = {
> .reset_resume = corsairpsu_resume,
> #endif
> };
> -module_hid_driver(corsairpsu_driver);
> +
> +static int __init corsair_init(void)
> +{
> + return hid_register_driver(&corsairpsu_driver);
> +}
> +
> +static void __exit corsair_exit(void)
> +{
> + hid_unregister_driver(&corsairpsu_driver);
> +}
> +
> +/*
> + * With module_init() the driver would load before the HID bus when
> + * built-in, so use late_initcall() instead.
> + */
> +late_initcall(corsair_init);
> +module_exit(corsair_exit);
>
> MODULE_LICENSE("GPL");
> MODULE_AUTHOR("Wilken Gottwalt <[email protected]>");
> --
> 2.39.2
>
On Thu, Dec 07, 2023 at 10:07:23PM +0100, Armin Wolf wrote:
> It seems that when the driver is built-in, the HID bus is
> initialized after the driver is loaded, which whould cause
> module_hid_driver() to fail.
> Fix this by registering the driver after the HID bus using
> late_initcall() in accordance with other hwmon HID drivers.
>
> Compile-tested only.
>
> Signed-off-by: Armin Wolf <[email protected]>
Applied. I dropped the "Compile-tested only" comment and added a note
that the patch was tested but the tested did not provide a Tested-by: tag.
Thanks,
Guenter
On Thu, 7 Dec 2023 22:07:23 +0100
Armin Wolf <[email protected]> wrote:
Just forgot. Did both, works just fine. Is it okay that way? This is my first
time.
Tested-by: Wilken Gottwalt <[email protected]>
Reviewed-by: Wilken Gottwalt <[email protected]>
> It seems that when the driver is built-in, the HID bus is
> initialized after the driver is loaded, which whould cause
> module_hid_driver() to fail.
> Fix this by registering the driver after the HID bus using
> late_initcall() in accordance with other hwmon HID drivers.
>
> Compile-tested only.
>
> Signed-off-by: Armin Wolf <[email protected]>
> ---
> drivers/hwmon/corsair-psu.c | 18 +++++++++++++++++-
> 1 file changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/hwmon/corsair-psu.c b/drivers/hwmon/corsair-psu.c
> index 904890598c11..2c7c92272fe3 100644
> --- a/drivers/hwmon/corsair-psu.c
> +++ b/drivers/hwmon/corsair-psu.c
> @@ -899,7 +899,23 @@ static struct hid_driver corsairpsu_driver = {
> .reset_resume = corsairpsu_resume,
> #endif
> };
> -module_hid_driver(corsairpsu_driver);
> +
> +static int __init corsair_init(void)
> +{
> + return hid_register_driver(&corsairpsu_driver);
> +}
> +
> +static void __exit corsair_exit(void)
> +{
> + hid_unregister_driver(&corsairpsu_driver);
> +}
> +
> +/*
> + * With module_init() the driver would load before the HID bus when
> + * built-in, so use late_initcall() instead.
> + */
> +late_initcall(corsair_init);
> +module_exit(corsair_exit);
>
> MODULE_LICENSE("GPL");
> MODULE_AUTHOR("Wilken Gottwalt <[email protected]>");
> --
> 2.39.2
>
On 12/8/23 10:44, Wilken Gottwalt wrote:
> On Thu, 7 Dec 2023 22:07:23 +0100
> Armin Wolf <[email protected]> wrote:
>
> Just forgot. Did both, works just fine. Is it okay that way? This is my first
> time.
>
> Tested-by: Wilken Gottwalt <[email protected]>
> Reviewed-by: Wilken Gottwalt <[email protected]>
>
Yes, it is perfect, just a bit late. I sent the pull request pretty much in parallel
(I didn't want to wait to make sure the patch makes it into v6.7-rc5).
Thanks,
Guenter