Skip to contentSkip to conversation

i2c: parport: Switch to use module_parport_driver()

PATCH Applied Andy Shevchenko 3 participants 5 messages Jul 12, 2021 – Aug 10, 2021

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 i2c: parport: Switch to use module_parport_driver()

Switch to use module_parport_driver() to reduce boilerplate code.

---
v2: fixed compilation error (Jean, LKP)

 drivers/i2c/busses/i2c-parport.c | 36 ++++++++++----------------------
 1 file changed, 11 insertions(+), 25 deletions(-)
drivers/i2c/busses/i2c-parport.c +11 -25
index a535889acca6..231145c48728 100644
@@ -267,6 +267,16 @@ static void i2c_parport_attach(struct parport *port)
 	int i;
 	struct pardev_cb i2c_parport_cb;
 
+	if (type < 0) {
+		pr_warn("adapter type unspecified\n");
+		return;
+	}
+
+	if (type >= ARRAY_SIZE(adapter_parm)) {
+		pr_warn("invalid type (%d)\n", type);
+		return;
+	}
+
 	for (i = 0; i < MAX_DEVICE; i++) {
 		if (parport[i] == -1)
 			continue;
@@ -392,32 +402,8 @@ static struct parport_driver i2c_parport_driver = {
 	.detach = i2c_parport_detach,
 	.devmodel = true,
 };
-
-/* ----- Module loading, unloading and information ------------------------ */
-
-static int __init i2c_parport_init(void)
-{
-	if (type < 0) {
-		pr_warn("adapter type unspecified\n");
-		return -ENODEV;
-	}
-
-	if (type >= ARRAY_SIZE(adapter_parm)) {
-		pr_warn("invalid type (%d)\n", type);
-		return -ENODEV;
-	}
-
-	return parport_register_driver(&i2c_parport_driver);
-}
-
-static void __exit i2c_parport_exit(void)
-{
-	parport_unregister_driver(&i2c_parport_driver);
-}
+module_parport_driver(i2c_parport_driver);
 
 MODULE_AUTHOR("Jean Delvare <[email protected]>");
 MODULE_DESCRIPTION("I2C bus over parallel port");
 MODULE_LICENSE("GPL");
-
-module_init(i2c_parport_init);
-module_exit(i2c_parport_exit);
Signature · 1 line
2.30.2
Reviewed-by on v2 1/1
#1 lore raw

Hi Andy,

Quoted 1 line · Andy Shevchenko

Switch to use module_parport_driver() to reduce boilerplate code.

This has the downside of moving the sanity check of the type parameter to run time, instead of driver load time. In particular this means that loading the i2c-parport driver without specifying the type will no longer fail. The driver will load successfully, but won't do anything.

While I prefer user errors to be reported as soon as possible, I don't really mind here, as parallel port drivers are not something worth debating over at this point in time. As a matter of fact, I can't possibly test this change as I no longer have a parallel port on any of my systems.

So if that's the direction we want to take then so be it.

Quoted 61 lines

---
v2: fixed compilation error (Jean, LKP)

 drivers/i2c/busses/i2c-parport.c | 36 ++++++++++----------------------
 1 file changed, 11 insertions(+), 25 deletions(-)
drivers/i2c/busses/i2c-parport.c +11 -25
index a535889acca6..231145c48728 100644
@@ -267,6 +267,16 @@ static void i2c_parport_attach(struct parport *port)
 	int i;
 	struct pardev_cb i2c_parport_cb;
 
+	if (type < 0) {
+		pr_warn("adapter type unspecified\n");
+		return;
+	}
+
+	if (type >= ARRAY_SIZE(adapter_parm)) {
+		pr_warn("invalid type (%d)\n", type);
+		return;
+	}
+
 	for (i = 0; i < MAX_DEVICE; i++) {
 		if (parport[i] == -1)
 			continue;
@@ -392,32 +402,8 @@ static struct parport_driver i2c_parport_driver = {
 	.detach = i2c_parport_detach,
 	.devmodel = true,
 };
-
-/* ----- Module loading, unloading and information ------------------------ */
-
-static int __init i2c_parport_init(void)
-{
-	if (type < 0) {
-		pr_warn("adapter type unspecified\n");
-		return -ENODEV;
-	}
-
-	if (type >= ARRAY_SIZE(adapter_parm)) {
-		pr_warn("invalid type (%d)\n", type);
-		return -ENODEV;
-	}
-
-	return parport_register_driver(&i2c_parport_driver);
-}
-
-static void __exit i2c_parport_exit(void)
-{
-	parport_unregister_driver(&i2c_parport_driver);
-}
+module_parport_driver(i2c_parport_driver);
 
 MODULE_AUTHOR("Jean Delvare <[email protected]>");
 MODULE_DESCRIPTION("I2C bus over parallel port");
 MODULE_LICENSE("GPL");
-
-module_init(i2c_parport_init);
-module_exit(i2c_parport_exit);
Signature · 2 lines
Jean Delvare
SUSE L3 Support
#2 lore raw
Quoted 9 lines · Jean Delvare

Hi Andy,

Quoted 1 line · Andy Shevchenko

Switch to use module_parport_driver() to reduce boilerplate code.

This has the downside of moving the sanity check of the type parameter to run time, instead of driver load time. In particular this means that loading the i2c-parport driver without specifying the type will no longer fail.

And this is actually an advantage of the change if you think about it. Now we can have a module that won't fail at boot time and give the user a chance to amend parameter at run time if it was wrong at boot time.

Quoted 11 lines

The driver will load successfully, but won't do anything.

While I prefer user errors to be reported as soon as possible, I don't really mind here, as parallel port drivers are not something worth debating over at this point in time. As a matter of fact, I can't possibly test this change as I no longer have a parallel port on any of my systems.

So if that's the direction we want to take then so be it.

Thanks!

Quoted 66 lines
Quoted 61 lines

---
v2: fixed compilation error (Jean, LKP)

 drivers/i2c/busses/i2c-parport.c | 36 ++++++++++----------------------
 1 file changed, 11 insertions(+), 25 deletions(-)
drivers/i2c/busses/i2c-parport.c +11 -25
index a535889acca6..231145c48728 100644
@@ -267,6 +267,16 @@ static void i2c_parport_attach(struct parport *port)
 	int i;
 	struct pardev_cb i2c_parport_cb;
 
+	if (type < 0) {
+		pr_warn("adapter type unspecified\n");
+		return;
+	}
+
+	if (type >= ARRAY_SIZE(adapter_parm)) {
+		pr_warn("invalid type (%d)\n", type);
+		return;
+	}
+
 	for (i = 0; i < MAX_DEVICE; i++) {
 		if (parport[i] == -1)
 			continue;
@@ -392,32 +402,8 @@ static struct parport_driver i2c_parport_driver = {
 	.detach = i2c_parport_detach,
 	.devmodel = true,
 };
-
-/* ----- Module loading, unloading and information ------------------------ */
-
-static int __init i2c_parport_init(void)
-{
-	if (type < 0) {
-		pr_warn("adapter type unspecified\n");
-		return -ENODEV;
-	}
-
-	if (type >= ARRAY_SIZE(adapter_parm)) {
-		pr_warn("invalid type (%d)\n", type);
-		return -ENODEV;
-	}
-
-	return parport_register_driver(&i2c_parport_driver);
-}
-
-static void __exit i2c_parport_exit(void)
-{
-	parport_unregister_driver(&i2c_parport_driver);
-}
+module_parport_driver(i2c_parport_driver);
 
 MODULE_AUTHOR("Jean Delvare <[email protected]>");
 MODULE_DESCRIPTION("I2C bus over parallel port");
 MODULE_LICENSE("GPL");
-
-module_init(i2c_parport_init);
-module_exit(i2c_parport_exit);
Signature · 2 lines
Jean Delvare
SUSE L3 Support
Signature · 2 lines
With Best Regards,
Andy Shevchenko
#3 lore raw
Quoted 4 lines · Andy Shevchenko
Quoted 3 lines · Jean Delvare
Quoted 1 line · Andy Shevchenko

Switch to use module_parport_driver() to reduce boilerplate code.

Quoted 8 lines
Quoted 4 lines

This has the downside of moving the sanity check of the type parameter to run time, instead of driver load time. In particular this means that loading the i2c-parport driver without specifying the type will no longer fail.

And this is actually an advantage of the change if you think about it. Now we can have a module that won't fail at boot time and give the user a chance to amend parameter at run time if it was wrong at boot time.

Okay, for built-in driver it might require to check and fix permissions of the parameter sysfs node (currently 0).

Quoted 7 lines
Quoted 7 lines

The driver will load successfully, but won't do anything.

While I prefer user errors to be reported as soon as possible, I don't really mind here, as parallel port drivers are not something worth debating over at this point in time. As a matter of fact, I can't possibly test this change as I no longer have a parallel port on any of my systems.

Signature · 2 lines
With Best Regards,
Andy Shevchenko