From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from [140.186.70.92] (port=48285 helo=eggs.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1Q8zOH-0001Fi-Ml
	for qemu-devel@nongnu.org; Sun, 10 Apr 2011 14:29:42 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <jljusten@gmail.com>) id 1Q8zOG-0004vY-KH
	for qemu-devel@nongnu.org; Sun, 10 Apr 2011 14:29:41 -0400
Received: from mail-px0-f174.google.com ([209.85.212.174]:57080)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <jljusten@gmail.com>) id 1Q8zOG-0004vR-Er
	for qemu-devel@nongnu.org; Sun, 10 Apr 2011 14:29:40 -0400
Received: by pxi15 with SMTP id 15so2511111pxi.33
	for <qemu-devel@nongnu.org>; Sun, 10 Apr 2011 11:29:38 -0700 (PDT)
MIME-Version: 1.0
In-Reply-To: <4DA16C6A.7090303@web.de>
References: <1301861786-6637-1-git-send-email-jordan.l.justen@intel.com>
	<4DA16C6A.7090303@web.de>
Date: Sun, 10 Apr 2011 11:29:38 -0700
Message-ID: <BANLkTimT0=HxsH9iRWJ4141u4pPUfcCP=g@mail.gmail.com>
Subject: Re: [Qemu-devel] Re: [PATCH] hw/pflash_cfi02: Fix lazy reset of ROMD
	mode
From: Jordan Justen <jljusten@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Jan Kiszka <jan.kiszka@web.de>
Cc: qemu-devel@nongnu.org

On Sun, Apr 10, 2011 at 01:38, Jan Kiszka <jan.kiszka@web.de> wrote:
> Indeed, that block looks weird to its author today as well. But
> inverting the logic completely defeats the purpose of lazy mode
> switching (will likely file a patch to remove the block).

Looking at the 2nd parameter to the call, and the
pflash_register_memory code, it seems it only makes sense with the
!pfl->rom_mode.

I thought that the goal was to allow multiple write operations before
exiting to rom mode, but that a read will return it to rom mode.  In
that case, this change seemed to fix it.

> We should
> instead switch back using the timer.
>
> So the question is: Did you actually see a problem that the flash was
> stuck in write mode, or did you just stumble over this strange code? In
> the former case, please explain the sequence or provide a trace.

I had enabled the debug trace messages and was probing the flash from
the efi shell on x86-64.  I found that I would always see the debug
messages for reads, after I started writing to the flash.

This change allowed the flash to go back to a rom mode where the debug
prints would then disappear for subsequent reads.

-Jordan