linux-mips
[Top] [All Lists]

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

To: sshtylyov@dev.rtsoft.ru
Subject: Re: [PATCH] serial_txx9: forcibly init the spinlock for PCI UART used as a console
From: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Date: Fri, 30 Dec 2005 01:32:42 +0900 (JST)
Cc: rmk@arm.linux.org.uk, linux-mips@linux-mips.org, ralf@linux-mips.org
In-reply-to: <20051228.132544.96684396.nemoto@toshiba-tops.co.jp>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <43B18DD2.3090206@ru.mvista.com> <43B196A9.8010608@dev.rtsoft.ru> <20051228.132544.96684396.nemoto@toshiba-tops.co.jp>
Sender: linux-mips-bounce@linux-mips.org
>>>>> On Wed, 28 Dec 2005 13:25:44 +0900 (JST), Atsushi Nemoto 
>>>>> <anemo@mba.ocn.ne.jp> said:

anemo> * The uart_console() returns 1 even if the console was not
anemo> successfully configured (CON_ENABLED was not set and spinlock
anemo> did not initialized).  So uart_add_one_port() does not
anemo> initialize the spinlock for the console.

This is same for the 8250 driver which also can handle PCI and legacy
ports.

The 8250 driver is working correctly just because
serial8250_isa_init_ports() initializes all spinlocks.  If this way a
right way, spin_lock_init() in uart_add_one_port() seems redundant...

---
Atsushi Nemoto

<Prev in Thread] Current Thread [Next in Thread>