[LinuxBIOS] Interesting.

[Feng, Libo]

Hi, all,

In our board, we found the trace is sent out: "Issuing SOFT_RESET…" from the function init_cpus in the file init_cpus.c. But when the function runs, the serial port is still far to be initialized, how dose the trace is sent out?

Best Regards

丰立波 Feng Libo @ AMD  Ext: 20906
Mobile Phone: [phone]
Office Phone: [phone]

[Marc Jones]

Feng, Libo wrote:
> Hi, all,
>
> In our board, we found the trace is sent out: "Issuing SOFT_RESET…" from 
> the function init_cpus in the file init_cpus.c. But when the function 
> runs, the serial port is still far to be initialized, how dose the trace 
> is sent out?
>
> Best Regards
>
> 丰立波 Feng Libo @ AMD  Ext: 20906
> Mobile Phone: [phone]
> Office Phone: [phone]

Each mainboard usually has an early uart setup located in auto.c or 
cache_as_ram_auto.c.

For example the Serengeti-Cheetah has the following lines.

w83627hf_enable_serial(SERIAL_DEV, TTYS0_BASE);
uart_init();
console_init();

Marc

[Feng, Libo]

Hi, all,

But the function init_cpus is called before the three functions: w83627hf_enable_serial(SERIAL_DEV, TTYS0_BASE); uart_init(); console_init(); The trace is sent out from the init_cpus.

Best Regards

丰立波 Feng Libo @ AMD  Ext: 20906
Mobile Phone: [phone]
Office Phone: [phone]

-----Original Message-----
From: Marc Jones [mailto:[e-mail]] 
Sent: Friday, November 30, 2007 12:47 AM
To: Feng, Libo
Cc: linuxbios
Subject: Re: [LinuxBIOS] Interesting.

Feng, Libo wrote:
> Hi, all,
>
> In our board, we found the trace is sent out: "Issuing SOFT_RESET…" 
> from the function init_cpus in the file init_cpus.c. But when the 
> function runs, the serial port is still far to be initialized, how 
> dose the trace is sent out?
>
> Best Regards
>
> 丰立波 Feng Libo @ AMD  Ext: 20906
> Mobile Phone: [phone]
> Office Phone: [phone]

Each mainboard usually has an early uart setup located in auto.c or cache_as_ram_auto.c.

For example the Serengeti-Cheetah has the following lines.

w83627hf_enable_serial(SERIAL_DEV, TTYS0_BASE); uart_init(); console_init();

Marc

[yhlu]

On Nov 29, 2007 9:25 PM, Feng, Libo  wrote:
> Hi, all,
>
> But the function init_cpus is called before the three functions: w83627hf_enable_serial(SERIAL_DEV, TTYS0_BASE); uart_init(); console_init(); The trace is sent out from the init_cpus.

that is for next fail reboot.

YH

[Feng, Libo]

Yes, our board reboots for some reason. The function print_debug and other functions must still be located where they had been located last time and all configuration related with serial port is still OK since the reboot, then without the serial port initialization, init_cpus can sent out trace. My understanding is right or not?

Best Regards

??? Feng Libo @ AMD  Ext: 20906
Mobile Phone: [phone]
Office Phone: [phone]

-----Original Message-----
From: yhlu [mailto:[e-mail]] 
Sent: Friday, November 30, 2007 3:46 PM
To: Feng, Libo
Cc: Jones, Marc; linuxbios
Subject: Re: [LinuxBIOS] Interesting.

On Nov 29, 2007 9:25 PM, Feng, Libo  wrote:
> Hi, all,
>
> But the function init_cpus is called before the three functions: w83627hf_enable_serial(SERIAL_DEV, TTYS0_BASE); uart_init(); console_init(); The trace is sent out from the init_cpus.

that is for next fail reboot.

YH