From mboxrd@z Thu Jan  1 00:00:00 1970
From: David Daney <ddaney@caviumnetworks.com>
Subject: Re: [PATCH 3/5] tty/8250_dw: Add support for OCTEON UARTS.
Date: Wed, 19 Jun 2013 09:45:28 -0700
Message-ID: <51C1E028.2040700@caviumnetworks.com>
References: <1371582775-12141-1-git-send-email-ddaney.cavm@gmail.com> <1371582775-12141-4-git-send-email-ddaney.cavm@gmail.com> <2302882.NVjP8DdXWY@wuerfel>
Mime-Version: 1.0
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
Content-Transfer-Encoding: 7bit
Return-path: <linux-serial-owner@vger.kernel.org>
Received: from mail-by2lp0235.outbound.protection.outlook.com ([207.46.163.235]:1574
	"EHLO na01-by2-obe.outbound.protection.outlook.com"
	rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP
	id S1756707Ab3FSQpd (ORCPT <rfc822;linux-serial@vger.kernel.org>);
	Wed, 19 Jun 2013 12:45:33 -0400
In-Reply-To: <2302882.NVjP8DdXWY@wuerfel>
Sender: linux-serial-owner@vger.kernel.org
List-Id: linux-serial@vger.kernel.org
To: Arnd Bergmann <arnd@arndb.de>
Cc: David Daney <ddaney.cavm@gmail.com>, linux-mips@linux-mips.org, ralf@linux-mips.org, Jamie Iles <jamie@jamieiles.com>, Greg Kroah-Hartman <gregkh@linuxfoundation.org>, Jiri Slaby <jslaby@suse.cz>, linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org, David Daney <david.daney@cavium.com>

On 06/19/2013 03:01 AM, Arnd Bergmann wrote:
> On Tuesday 18 June 2013 12:12:53 David Daney wrote:
>> +static unsigned int dw8250_serial_inq(struct uart_port *p, int offset)
>> +{
>> +       offset <<= p->regshift;
>> +
>> +       return (u8)__raw_readq(p->membase + offset);
>> +}
>> +
>> +static void dw8250_serial_outq(struct uart_port *p, int offset, int value)
>> +{
>> +       struct dw8250_data *d = p->private_data;
>> +
>> +       if (offset == UART_LCR)
>> +               d->last_lcr = value;
>> +
>> +       offset <<= p->regshift;
>> +       __raw_writeq(value, p->membase + offset);
>> +       dw8250_serial_inq(p, UART_LCR);
>> +}
>
> This breaks building on 32 bit architectures as I found on my daily ARM
> builds: __raw_writeq cannot be defined on architectures that don't have
> native 64 bit data access instructions.

I will rework the patch to avoid this problem.


> It's also wrong to use the
> __raw_* variant, which is not guaranteed to be atomic and is not
> endian-safe.

We do runtime probing and only use this function on platforms where it 
is appropriate, so atomicity is not an issue.  As for endianess, I used 
the __raw_ variant precisely because it is correct for both big and 
little endian kernels.

David Daney