Date: Mon, 3 Jan 2005 18:02:09 +0000
From: Al Viro <viro@parcelfarce.linux.theplanet.co.uk>
To: Len Brown <len.brown@intel.com>
Cc: Michael Geithe <warpy@gmx.de>,
       Linux Kernel list <linux-kernel@vger.kernel.org>
Subject: Re: __iounmap: bad address c00f0000 (Re: 2.6.10-bk5)
Message-ID: <20050103180209.GW26051@parcelfarce.linux.theplanet.co.uk>
References: <200501030114.55399.warpy@gmx.de> <1104773076.18173.64.camel@d845pe> <1104774583.18174.68.camel@d845pe>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1104774583.18174.68.camel@d845pe>
User-Agent: Mutt/1.4.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1216
Lines: 27

On Mon, Jan 03, 2005 at 12:49:43PM -0500, Len Brown wrote:
> Suggested-by: Al Viro
> Signed-off-by: Len Brown <len.brown@intel.com>
> 
> 
> ===== arch/i386/kernel/dmi_scan.c 1.74 vs edited =====
> --- 1.74/arch/i386/kernel/dmi_scan.c	2004-12-28 14:07:48 -05:00
> +++ edited/arch/i386/kernel/dmi_scan.c	2005-01-03 12:46:33 -05:00
> @@ -126,12 +126,12 @@
>  			dmi_printk((KERN_INFO "DMI table at 0x%08X.\n",
>  				base));
>  			if(dmi_table(base,len, num, decode)==0) {
> -				iounmap(p);
> +				/* too early to call iounmap(p); */

One comment: iounmap() on result of ioremap(ISA address) is a no-op.
The problem in this case is that we are too early in setup sequence
and that confuses iounmap() into not recognizing that fact.  So we
end up trying to find VMA created by ioremap() and (surprise, surprise)
find none.  Which gives the warning in question.  We *still* do nothing,
so everything actually works fine, but we do get confused "WTF is going
on?" in the logs.
-
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/