linux-mips
[Top] [All Lists]

Re: [PATCH] TXx9: Add TX4938 ATA support

To: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Subject: Re: [PATCH] TXx9: Add TX4938 ATA support
From: Sergei Shtylyov <sshtylyov@ru.mvista.com>
Date: Mon, 20 Oct 2008 18:40:56 +0400
Cc: linux-mips@linux-mips.org, linux-ide@vger.kernel.org, bzolnier@gmail.com, ralf@linux-mips.org
In-reply-to: <20081020.231450.51866269.anemo@mba.ocn.ne.jp>
Organization: MontaVista Software Inc.
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <20081017.231036.26097587.anemo@mba.ocn.ne.jp> <48FB7D9E.4030803@ru.mvista.com> <20081020.231450.51866269.anemo@mba.ocn.ne.jp>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Mozilla/5.0 (X11; U; Linux i686; rv:1.7.2) Gecko/20040803
Hello.

Atsushi Nemoto wrote:

+       struct tx4938ide_platform_info pdata = {
+               .ioport_shift = shift,
+               .gbus_clock = tune ? txx9_gbus_clock : 0,

   Any reason not to supply the GBUS clock?

The EBUS channel might be used for both ATA and ISA or other local bus
devices.  In that case, the board setup code should initialize best
timings for all devices and the IDE driver should not overrite it.

Ah, I forgot about the ATA address space being appeandage to ISA address space on TX4939...

   I'm afraid you can't just early return from the set_pio_mode() method...

Do you mean I should use IDE_HFLAG_NO_SET_MODE instead of just
returning from set_pio_mode?

No, that's for the smart RAID controllers that program the transfer modes themselves. In your case, hwif->port_ops->set_pio_mode() must be NULL -- if you're not going to allow the mode programming, that is.

MBR, Sergei

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