Re: [PATCH 1/2] x86: Add Isolated Memory Regions for Quark X1000

From: "Bryan O'Donoghue" <pure.logic@nexus-software.ie>
To: "Ong, Boon Leong" <boon.leong.ong@intel.com>,
	Darren Hart <dvhart@infradead.org>
Cc: "tglx@linutronix.de" <tglx@linutronix.de>,
	"mingo@redhat.com" <mingo@redhat.com>,
	"hpa@zytor.com" <hpa@zytor.com>,
	"x86@kernel.org" <x86@kernel.org>,
	"platform-driver-x86@vger.kernel.org" 
	<platform-driver-x86@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 1/2] x86: Add Isolated Memory Regions for Quark X1000
Date: Thu, 08 Jan 2015 12:10:47 +0000	[thread overview]
Message-ID: <54AE73C7.90009@nexus-software.ie> (raw)
In-Reply-To: <AF233D1473C1364ABD51D28909A1B1B7326E0C25@PGSMSX105.gar.corp.intel.com>

On 07/01/15 23:45, Ong, Boon Leong wrote:
>> Since BIOS and grub code both use 0x00000000 as the 'off' address I think it
>> makes sense for the kernel to continue to use that address.
>
> Just add on top of what Daren mentioned in another mail, based on the Quark document,
> the base address can start from zero. Say lo=0, hi=0, and WM & RM may be changed from default value,
> 1st 1KiB will be marked as IMR. It seems to me that there is no good way to test if an IMR is 'occupied' and/or 'enabled'
> since enable-bit is not available. But, what is benefit of testing against lo=0 & hi=0? The logic to calculate size will work under
> lo=0 & hi=0 anway.

Hi Boon Leong.

I think it does make sense to add a check for rmask and wmask in the 
'access all' state when determining if an IMR is enabled on X1000 or not.

>> My own view is that it's not really desirable and easier to debug IMRs
>> generally on a platform if overlaps aren't allowed.
> I do agree on the benefit listed above. Perhaps, you can add explanation here
> to mention the design decision.

Will do.