Re: [PATCH] serial_txx9: forcibly init the spinlock for PCI UART used as a console

[Sergei Shtylylov <sshtylyov@ru.mvista.com>]

Hello.

Russell King wrote:

> On Wed, Dec 28, 2005 at 12:34:57AM +0900, Atsushi Nemoto wrote:
> 
>>Thanks for your comment.
>>
>>
>>>>>>>On Tue, 27 Dec 2005 16:38:54 +0300, Sergei Shtylylov <sshtylyov@ru.mvista.com> said:
>>
>>>>The problem is not just only spin_lock_init.  The parameters of
>>>>"console=" option (baudrate, etc.) are not passed for PCI UART.
>>
>>sshtylyov>     They are -- uart_add_one_port() calls console setup
>>sshtylyov> once more when registering PCI UART with serial code.
>>
>>Yes, you are right.  I missed the register_console call in
>>uart_add_one_port().  So your patch will fix the problem.  But I
>>suppose the spinlock should be initialized in serial_core.  How about
>>this?

> I think you're layering work-around on top of work-around on top of
> work-around here.

> I think the first thing you need to resolve is the way you're
> registering your ports.  Firstly, if you're solely PCI-based, there's

    No, this is not the case. The driver serves Toshiba TX39xx/49xx SOC UARTs 
as well as the compatible PCI UART (GOKU-S).

> no need to pre-register all the uart ports at driver initialisation
> time.  Consequently, there's no need to remove them all when you
> remove the module.

    Hm, then the driver would need to keep track of which ports it has 
registered and which it has not...

 > Secondly, the upshot of this is that you only call uart_add_one_port()
 > when you initialise a PCI card.

    Not the case as I've said; uart_add_one_port() should be called on driver 
startup anyway...

> This should result in a cleaner implementation, and the console will
> not be started until you detect the PCI card.

    It will still be started with the console_initcall() in this driver, if 
that code is not also deleted...

WBR, Sergei