[PATCH] serial: core: Revert port_id use

[PATCH] serial: core: Revert port_id use

[Tony Lindgren]

Guenter reports boot issues with duplicate sysfs entries for multiport
drivers. Let's go back to using port->line for now to fix the regression.

With this change, the serial core port device names are not correct for the
hardware specific 8250 single port drivers, but that's a cosmetic issue for
now.

Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
Reported-by: Guenter Roeck <groeck7@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
---
 drivers/tty/serial/serial_base_bus.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/tty/serial/serial_base_bus.c b/drivers/tty/serial/serial_base_bus.c
--- a/drivers/tty/serial/serial_base_bus.c
+++ b/drivers/tty/serial/serial_base_bus.c
@@ -151,7 +151,7 @@ struct serial_port_device *serial_base_port_add(struct uart_port *port,
 	err = serial_base_device_init(port, &port_dev->dev,
 				      &ctrl_dev->dev, &serial_port_type,
 				      serial_base_port_release,
-				      port->ctrl_id, port->port_id);
+				      port->ctrl_id, port->line);
 	if (err)
 		goto err_put_device;
 
-- 
2.41.0

Re: [PATCH] serial: core: Revert port_id use

[Guenter Roeck]

On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
> Guenter reports boot issues with duplicate sysfs entries for multiport
> drivers. Let's go back to using port->line for now to fix the regression.
> 
> With this change, the serial core port device names are not correct for the
> hardware specific 8250 single port drivers, but that's a cosmetic issue for
> now.
> 
> Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
> Reported-by: Guenter Roeck <groeck7@gmail.com>
> Signed-off-by: Tony Lindgren <tony@atomide.com>

Tested-by: Guenter Roeck <linux@roeck-us.net>

Guenter

Re: [PATCH] serial: core: Revert port_id use

[Tony Lindgren]

Hi,

* Guenter Roeck <linux@roeck-us.net> [230806 13:19]:
> On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
> > Guenter reports boot issues with duplicate sysfs entries for multiport
> > drivers. Let's go back to using port->line for now to fix the regression.
> > 
> > With this change, the serial core port device names are not correct for the
> > hardware specific 8250 single port drivers, but that's a cosmetic issue for
> > now.
> > 
> > Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
> > Reported-by: Guenter Roeck <groeck7@gmail.com>
> > Signed-off-by: Tony Lindgren <tony@atomide.com>
> 
> Tested-by: Guenter Roeck <linux@roeck-us.net>

Thanks for testing.

Guenter, care to also test the patch below on top of this fix and
see if things still behave for you?

I'll send a proper patch assuming things test fine.

Regagrds,

Tony

8< --------------------
diff --git a/drivers/tty/serial/serial_base.h b/drivers/tty/serial/serial_base.h
--- a/drivers/tty/serial/serial_base.h
+++ b/drivers/tty/serial/serial_base.h
@@ -16,6 +16,7 @@ struct device;
 
 struct serial_ctrl_device {
 	struct device dev;
+	struct ida ida;
 };
 
 struct serial_port_device {
diff --git a/drivers/tty/serial/serial_base_bus.c b/drivers/tty/serial/serial_base_bus.c
--- a/drivers/tty/serial/serial_base_bus.c
+++ b/drivers/tty/serial/serial_base_bus.c
@@ -10,6 +10,7 @@
 
 #include <linux/container_of.h>
 #include <linux/device.h>
+#include <linux/idr.h>
 #include <linux/module.h>
 #include <linux/serial_core.h>
 #include <linux/slab.h>
@@ -112,6 +113,8 @@ struct serial_ctrl_device *serial_base_ctrl_add(struct uart_port *port,
 	if (!ctrl_dev)
 		return ERR_PTR(-ENOMEM);
 
+	ida_init(&ctrl_dev->ida);
+
 	err = serial_base_device_init(port, &ctrl_dev->dev,
 				      parent, &serial_ctrl_type,
 				      serial_base_ctrl_release,
@@ -142,16 +145,31 @@ struct serial_port_device *serial_base_port_add(struct uart_port *port,
 						struct serial_ctrl_device *ctrl_dev)
 {
 	struct serial_port_device *port_dev;
+	unsigned int min = 0, max = ~0U;
 	int err;
 
 	port_dev = kzalloc(sizeof(*port_dev), GFP_KERNEL);
 	if (!port_dev)
 		return ERR_PTR(-ENOMEM);
 
+	/* Device driver specified port_id vs automatic assignment? */
+	if (port->port_id) {
+		min = port->port_id;
+		max = port->port_id;
+	}
+
+	err = ida_alloc_range(&ctrl_dev->ida, min, max, GFP_KERNEL);
+	if (err < 0) {
+		kfree(port_dev);
+		return ERR_PTR(err);
+	}
+
+	port->port_id = err;
+
 	err = serial_base_device_init(port, &port_dev->dev,
 				      &ctrl_dev->dev, &serial_port_type,
 				      serial_base_port_release,
-				      port->ctrl_id, port->line);
+				      port->ctrl_id, port->port_id);
 	if (err)
 		goto err_put_device;
 
@@ -165,16 +183,24 @@ struct serial_port_device *serial_base_port_add(struct uart_port *port,
 
 err_put_device:
 	put_device(&port_dev->dev);
+	ida_free(&ctrl_dev->ida, port->port_id);
 
 	return ERR_PTR(err);
 }
 
 void serial_base_port_device_remove(struct serial_port_device *port_dev)
 {
+	struct serial_ctrl_device *ctrl_dev;
+	struct device *parent;
+
 	if (!port_dev)
 		return;
 
+	parent = port_dev->dev.parent;
+	ctrl_dev = to_serial_base_ctrl_device(parent);
+
 	device_del(&port_dev->dev);
+	ida_free(&ctrl_dev->ida, port_dev->port->port_id);
 	put_device(&port_dev->dev);
 }
 
-- 
2.41.0

Re: [PATCH] serial: core: Revert port_id use

[Dhruva Gole]

On Aug 09, 2023 at 08:26:50 +0300, Tony Lindgren wrote:
> Hi,
> 
> * Guenter Roeck <linux@roeck-us.net> [230806 13:19]:
> > On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
> > > Guenter reports boot issues with duplicate sysfs entries for multiport
> > > drivers. Let's go back to using port->line for now to fix the regression.
> > > 
> > > With this change, the serial core port device names are not correct for the
> > > hardware specific 8250 single port drivers, but that's a cosmetic issue for
> > > now.
> > > 
> > > Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
> > > Reported-by: Guenter Roeck <groeck7@gmail.com>
> > > Signed-off-by: Tony Lindgren <tony@atomide.com>
> > 
> > Tested-by: Guenter Roeck <linux@roeck-us.net>
> 
> Thanks for testing.
> 
> Guenter, care to also test the patch below on top of this fix and
> see if things still behave for you?
> 
> I'll send a proper patch assuming things test fine.
> 
> Regagrds,
> 
> Tony
> 
> 8< --------------------
> diff --git a/drivers/tty/serial/serial_base.h b/drivers/tty/serial/serial_base.h
> --- a/drivers/tty/serial/serial_base.h
> +++ b/drivers/tty/serial/serial_base.h
> @@ -16,6 +16,7 @@ struct device;
>  
>  struct serial_ctrl_device {
>  	struct device dev;
> +	struct ida ida;

When you send the proper patch, might want to change variable name to
something other than ida?

[...]
>  
> -- 
> 2.41.0

-- 
Best regards,
Dhruva Gole <d-gole@ti.com>

Re: [PATCH] serial: core: Revert port_id use

[Tony Lindgren]

* Dhruva Gole <d-gole@ti.com> [230809 05:33]:
> On Aug 09, 2023 at 08:26:50 +0300, Tony Lindgren wrote:
> > --- a/drivers/tty/serial/serial_base.h
> > +++ b/drivers/tty/serial/serial_base.h
> > @@ -16,6 +16,7 @@ struct device;
> >  
> >  struct serial_ctrl_device {
> >  	struct device dev;
> > +	struct ida ida;
> 
> When you send the proper patch, might want to change variable name to
> something other than ida?

OK I'll change it to port_ida.

Tony

Re: [PATCH] serial: core: Revert port_id use

[Guenter Roeck]

On 8/8/23 22:26, Tony Lindgren wrote:
> Hi,
> 
> * Guenter Roeck <linux@roeck-us.net> [230806 13:19]:
>> On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
>>> Guenter reports boot issues with duplicate sysfs entries for multiport
>>> drivers. Let's go back to using port->line for now to fix the regression.
>>>
>>> With this change, the serial core port device names are not correct for the
>>> hardware specific 8250 single port drivers, but that's a cosmetic issue for
>>> now.
>>>
>>> Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
>>> Reported-by: Guenter Roeck <groeck7@gmail.com>
>>> Signed-off-by: Tony Lindgren <tony@atomide.com>
>>
>> Tested-by: Guenter Roeck <linux@roeck-us.net>
> 
> Thanks for testing.
> 
> Guenter, care to also test the patch below on top of this fix and
> see if things still behave for you?
> 

Queued. We should have test results sometime tomorrow.

Guenter

Re: [PATCH] serial: core: Revert port_id use

[Tony Lindgren]

* Guenter Roeck <linux@roeck-us.net> [230809 05:36]:
> On 8/8/23 22:26, Tony Lindgren wrote:
> > Hi,
> > 
> > * Guenter Roeck <linux@roeck-us.net> [230806 13:19]:
> > > On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
> > > > Guenter reports boot issues with duplicate sysfs entries for multiport
> > > > drivers. Let's go back to using port->line for now to fix the regression.
> > > > 
> > > > With this change, the serial core port device names are not correct for the
> > > > hardware specific 8250 single port drivers, but that's a cosmetic issue for
> > > > now.
> > > > 
> > > > Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
> > > > Reported-by: Guenter Roeck <groeck7@gmail.com>
> > > > Signed-off-by: Tony Lindgren <tony@atomide.com>
> > > 
> > > Tested-by: Guenter Roeck <linux@roeck-us.net>
> > 
> > Thanks for testing.
> > 
> > Guenter, care to also test the patch below on top of this fix and
> > see if things still behave for you?
> > 
> 
> Queued. We should have test results sometime tomorrow.

OK great.

Thanks,

Tony

Re: [PATCH] serial: core: Revert port_id use

[Guenter Roeck]

On 8/8/23 22:26, Tony Lindgren wrote:
> Hi,
> 
> * Guenter Roeck <linux@roeck-us.net> [230806 13:19]:
>> On Sun, Aug 06, 2023 at 09:20:50AM +0300, Tony Lindgren wrote:
>>> Guenter reports boot issues with duplicate sysfs entries for multiport
>>> drivers. Let's go back to using port->line for now to fix the regression.
>>>
>>> With this change, the serial core port device names are not correct for the
>>> hardware specific 8250 single port drivers, but that's a cosmetic issue for
>>> now.
>>>
>>> Fixes: d962de6ae51f ("serial: core: Fix serial core port id to not use port->line")
>>> Reported-by: Guenter Roeck <groeck7@gmail.com>
>>> Signed-off-by: Tony Lindgren <tony@atomide.com>
>>
>> Tested-by: Guenter Roeck <linux@roeck-us.net>
> 
> Thanks for testing.
> 
> Guenter, care to also test the patch below on top of this fix and
> see if things still behave for you?
> 
> I'll send a proper patch assuming things test fine.
> 

Patch below works for me.

Tested-by: Guenter Roeck <linux@roeck-us.net>

Guenter

> Regagrds,
> 
> Tony
> 
> 8< --------------------
> diff --git a/drivers/tty/serial/serial_base.h b/drivers/tty/serial/serial_base.h
> --- a/drivers/tty/serial/serial_base.h
> +++ b/drivers/tty/serial/serial_base.h
> @@ -16,6 +16,7 @@ struct device;
>   
>   struct serial_ctrl_device {
>   	struct device dev;
> +	struct ida ida;
>   };
>   
>   struct serial_port_device {
> diff --git a/drivers/tty/serial/serial_base_bus.c b/drivers/tty/serial/serial_base_bus.c
> --- a/drivers/tty/serial/serial_base_bus.c
> +++ b/drivers/tty/serial/serial_base_bus.c
> @@ -10,6 +10,7 @@
>   
>   #include <linux/container_of.h>
>   #include <linux/device.h>
> +#include <linux/idr.h>
>   #include <linux/module.h>
>   #include <linux/serial_core.h>
>   #include <linux/slab.h>
> @@ -112,6 +113,8 @@ struct serial_ctrl_device *serial_base_ctrl_add(struct uart_port *port,
>   	if (!ctrl_dev)
>   		return ERR_PTR(-ENOMEM);
>   
> +	ida_init(&ctrl_dev->ida);
> +
>   	err = serial_base_device_init(port, &ctrl_dev->dev,
>   				      parent, &serial_ctrl_type,
>   				      serial_base_ctrl_release,
> @@ -142,16 +145,31 @@ struct serial_port_device *serial_base_port_add(struct uart_port *port,
>   						struct serial_ctrl_device *ctrl_dev)
>   {
>   	struct serial_port_device *port_dev;
> +	unsigned int min = 0, max = ~0U;
>   	int err;
>   
>   	port_dev = kzalloc(sizeof(*port_dev), GFP_KERNEL);
>   	if (!port_dev)
>   		return ERR_PTR(-ENOMEM);
>   
> +	/* Device driver specified port_id vs automatic assignment? */
> +	if (port->port_id) {
> +		min = port->port_id;
> +		max = port->port_id;
> +	}
> +
> +	err = ida_alloc_range(&ctrl_dev->ida, min, max, GFP_KERNEL);
> +	if (err < 0) {
> +		kfree(port_dev);
> +		return ERR_PTR(err);
> +	}
> +
> +	port->port_id = err;
> +
>   	err = serial_base_device_init(port, &port_dev->dev,
>   				      &ctrl_dev->dev, &serial_port_type,
>   				      serial_base_port_release,
> -				      port->ctrl_id, port->line);
> +				      port->ctrl_id, port->port_id);
>   	if (err)
>   		goto err_put_device;
>   
> @@ -165,16 +183,24 @@ struct serial_port_device *serial_base_port_add(struct uart_port *port,
>   
>   err_put_device:
>   	put_device(&port_dev->dev);
> +	ida_free(&ctrl_dev->ida, port->port_id);
>   
>   	return ERR_PTR(err);
>   }
>   
>   void serial_base_port_device_remove(struct serial_port_device *port_dev)
>   {
> +	struct serial_ctrl_device *ctrl_dev;
> +	struct device *parent;
> +
>   	if (!port_dev)
>   		return;
>   
> +	parent = port_dev->dev.parent;
> +	ctrl_dev = to_serial_base_ctrl_device(parent);
> +
>   	device_del(&port_dev->dev);
> +	ida_free(&ctrl_dev->ida, port_dev->port->port_id);
>   	put_device(&port_dev->dev);
>   }
>   

Re: [PATCH] serial: core: Revert port_id use

[Tony Lindgren]

* Guenter Roeck <linux@roeck-us.net> [230809 21:17]:
> On 8/8/23 22:26, Tony Lindgren wrote:
> > Guenter, care to also test the patch below on top of this fix and
> > see if things still behave for you?
> > 
> > I'll send a proper patch assuming things test fine.
> > 
> 
> Patch below works for me.

OK good to hear thanks, patch posted at [0] below.

Regards,

Tony

[0] https://lore.kernel.org/linux-serial/20230810065737.47294-1-tony@atomide.com/T/#u