From mboxrd@z Thu Jan  1 00:00:00 1970
From: Alex Williamson <alex.williamson@redhat.com>
Subject: Re: PCI device reverse engineering
Date: Fri, 28 Feb 2014 11:48:23 -0700
Message-ID: <1393613303.26901.24.camel@ul30vt.home>
References: <CAJH5N=RfA4ThSGgqY6i=ny9DMVZzOy7=KBUJy2Lef6oApSq=UQ@mail.gmail.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
Cc: kvm@vger.kernel.org
To: ret val <retval386@gmail.com>
Return-path: <kvm-owner@vger.kernel.org>
Received: from mx1.redhat.com ([209.132.183.28]:18639 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1751457AbaB1SsZ (ORCPT <rfc822;kvm@vger.kernel.org>);
	Fri, 28 Feb 2014 13:48:25 -0500
In-Reply-To: <CAJH5N=RfA4ThSGgqY6i=ny9DMVZzOy7=KBUJy2Lef6oApSq=UQ@mail.gmail.com>
Sender: kvm-owner@vger.kernel.org
List-ID: <kvm.vger.kernel.org>

On Fri, 2014-02-28 at 13:39 -0500, ret val wrote:
> I would like to use KVM to reverse engineer/debug a PCI driver. With a
> IOMMU and a pass-through device setup, would it be possible to
> monitor/log all PCI bus traffic to a device? How can I go about
> enabling and viewing this?

You can't monitor bus traffic, you'd still need a PCI analyzer for that,
but you can track all driver and guest OS access to the device.  So you
can at least soft or reverse engineer the driver.  To do that, uncomment
the #define DEBUG_VFIO line near the top of hw/misc/vfio.c and change
all the VFIO_ALLOW_FOO defines from 1 to 0, especially the MMAP one.
Thanks,

Alex