Subject: Re: [PATCH] UIO: Implement a UIO interface for the SMX Cryptengine
From: Ben Nizette <bn@niasdigital.com>
To: Paul Mundt <lethal@linux-sh.org>
Cc: gregkh@suse.de, linux-kernel <linux-kernel@vger.kernel.org>,
       hjk@linutronix.de
In-Reply-To: <20080312100831.GB21854@linux-sh.org>
References: <1205211430.3817.15.camel@moss.renham>
	 <20080312100831.GB21854@linux-sh.org>
Content-Type: text/plain
Organization: Nias Digital
Date: Wed, 12 Mar 2008 21:41:08 +1100
Message-Id: <1205318468.4344.23.camel@moss.renham>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1196
Lines: 31


On Wed, 2008-03-12 at 19:08 +0900, Paul Mundt wrote:
> On Tue, Mar 11, 2008 at 03:57:10PM +1100, Ben Nizette wrote:
> > +
> > +	info->mem[0].size = regs->end - regs->start;
> > +	info->mem[0].memtype = UIO_MEM_PHYS;
> > +
> You have an off-by-1 in the resource size. ioremap and struct resource
> are not equal in their expectations. You probably want to do something
> like:
> 
> 	info->mem[0].size = regs->end - regs->start + 1;
> 	info->mem[0].internal_addr = ioremap(regs->start, info->mem[0].size);
> 	...
> 
You're right, thanks.  I hadn't noticed since we don't actually access
all the way to the end of the region in normal use.

> and make sure that your resource end doesn't overlap with a start address
> of an unrelated resource in your definition. /proc/iomem tends to be
> useful for this.

Yup.  In our case this is fairly easy as the engine just takes up a
whole Chip Select's worth of address space on an AVR32 AT32AP7000.

		--Ben.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/