[PATCH 1/1] Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.

[PATCH 1/1] Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.

[Konstantin Smola]

kernel/drivers/serial: Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.
Signed-off-by: Konstantin Smola <ksmola51@gmail.com>
---
 kernel/drivers/serial/16550A.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/kernel/drivers/serial/16550A.c b/kernel/drivers/serial/16550A.c
index c9274fb19..325737f65 100644
--- a/kernel/drivers/serial/16550A.c
+++ b/kernel/drivers/serial/16550A.c
@@ -31,6 +31,7 @@ MODULE_VERSION("1.5.2");
 MODULE_LICENSE("GPL");
 
 #define RT_16550_DRIVER_NAME	"xeno_16550A"
+static char mod_name[64]=RT_16550_DRIVER_NAME;
 
 #define MAX_DEVICES		8
 
@@ -1128,6 +1129,8 @@ int __init rt_16550_init(void)
 		name = (char *)(dev + 1);
 		ksformat(name, RTDM_MAX_DEVNAME_LEN, dev->label, i);
 
+		rtdm_printk("%s: %s io=0x%x irq=%d baud_base=%d\n", mod_name, name, (unsigned)rt_16550_base_addr(i),irq[i],baud_base[i]);
+
 		err = rt_16550_init_io(i, name);
 		if (err)
 			goto kfree_out;
-- 
2.27.0

Re: [PATCH 1/1] Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.

[Bezdeka, Florian]

On Sun, 2022-01-23 at 00:18 -0800, Konstantin Smola via Xenomai wrote:
> kernel/drivers/serial: Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.
> Signed-off-by: Konstantin Smola <ksmola51@gmail.com>
> ---
>  kernel/drivers/serial/16550A.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/kernel/drivers/serial/16550A.c b/kernel/drivers/serial/16550A.c
> index c9274fb19..325737f65 100644
> --- a/kernel/drivers/serial/16550A.c
> +++ b/kernel/drivers/serial/16550A.c
> @@ -31,6 +31,7 @@ MODULE_VERSION("1.5.2");
>  MODULE_LICENSE("GPL");
>  
>  #define RT_16550_DRIVER_NAME	"xeno_16550A"
> +static char mod_name[64]=RT_16550_DRIVER_NAME;

How about using RT_16550_DRIVER_NAME directly?

>  
>  #define MAX_DEVICES		8
>  
> @@ -1128,6 +1129,8 @@ int __init rt_16550_init(void)
>  		name = (char *)(dev + 1);
>  		ksformat(name, RTDM_MAX_DEVNAME_LEN, dev->label, i);
>  
> +		rtdm_printk("%s: %s io=0x%x irq=%d baud_base=%d\n", mod_name, name, (unsigned)rt_16550_base_addr(i),irq[i],baud_base[i]);
> +

I guess Jan will comment on that as well, not sure if we really need
such verbose outputs. Technically I wonder if that is the right
location. The initialization might still fail below. So maybe move it
to the place where the device is up and running?

>  		err = rt_16550_init_io(i, name);
>  		if (err)
>  			goto kfree_out;

Re: [PATCH 1/1] Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.

[Jan Kiszka]

On 24.01.22 08:40, Bezdeka, Florian via Xenomai wrote:
> On Sun, 2022-01-23 at 00:18 -0800, Konstantin Smola via Xenomai wrote:
>> kernel/drivers/serial: Added verbosity to 16550A serial driver. dmesg shows the ports, irqs, baudbases.
>> Signed-off-by: Konstantin Smola <ksmola51@gmail.com>
>> ---
>>   kernel/drivers/serial/16550A.c | 3 +++
>>   1 file changed, 3 insertions(+)
>>
>> diff --git a/kernel/drivers/serial/16550A.c b/kernel/drivers/serial/16550A.c
>> index c9274fb19..325737f65 100644
>> --- a/kernel/drivers/serial/16550A.c
>> +++ b/kernel/drivers/serial/16550A.c
>> @@ -31,6 +31,7 @@ MODULE_VERSION("1.5.2");
>>   MODULE_LICENSE("GPL");
>>   
>>   #define RT_16550_DRIVER_NAME	"xeno_16550A"
>> +static char mod_name[64]=RT_16550_DRIVER_NAME;
> 
> How about using RT_16550_DRIVER_NAME directly?
> 
>>   
>>   #define MAX_DEVICES		8
>>   
>> @@ -1128,6 +1129,8 @@ int __init rt_16550_init(void)
>>   		name = (char *)(dev + 1);
>>   		ksformat(name, RTDM_MAX_DEVNAME_LEN, dev->label, i);
>>   
>> +		rtdm_printk("%s: %s io=0x%x irq=%d baud_base=%d\n", mod_name, name, (unsigned)rt_16550_base_addr(i),irq[i],baud_base[i]);
>> +
> 
> I guess Jan will comment on that as well, not sure if we really need
> such verbose outputs.

Yeah, but more for style reasons (overlong line). The output itself is 
OK, comparable to what Linux reports about its TTYs on boot-up too.

> Technically I wonder if that is the right
> location. The initialization might still fail below. So maybe move it
> to the place where the device is up and running?

That is a valid point, specifically as there is no error output otherwise.

Jan

-- 
Siemens AG, Technology
Competence Center Embedded Linux