From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:39266)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <kevin@koconnor.net>) id 1bwecI-0000xo-Lb
	for qemu-devel@nongnu.org; Tue, 18 Oct 2016 20:20:23 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <kevin@koconnor.net>) id 1bwecD-0002SN-84
	for qemu-devel@nongnu.org; Tue, 18 Oct 2016 20:20:22 -0400
Received: from mail-qk0-x243.google.com ([2607:f8b0:400d:c09::243]:35840)
	by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16)
	(Exim 4.71) (envelope-from <kevin@koconnor.net>) id 1bwecD-0002RV-2d
	for qemu-devel@nongnu.org; Tue, 18 Oct 2016 20:20:17 -0400
Received: by mail-qk0-x243.google.com with SMTP id z190so725259qkc.3
	for <qemu-devel@nongnu.org>; Tue, 18 Oct 2016 17:20:16 -0700 (PDT)
Date: Tue, 18 Oct 2016 20:20:14 -0400
From: Kevin O'Connor <kevin@koconnor.net>
Message-ID: <20161019002014.GA18055@morn.lan>
References: <1475053640-30483-1-git-send-email-kraxel@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1475053640-30483-1-git-send-email-kraxel@redhat.com>
Subject: Re: [Qemu-devel] [SeaBIOS] [PATCH 0/7] serial console support
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/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: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Gerd Hoffmann <kraxel@redhat.com>
Cc: seabios@seabios.org, qemu-devel@nongnu.org

On Wed, Sep 28, 2016 at 11:07:13AM +0200, Gerd Hoffmann wrote:
>   Hi,
> 
> After a looong break finally the next round
> of the seabios serial console patches.

Hi Gerd,

Sorry for the delay in responding.

I ran some tests on your series and it looks like it causes issues
with some systems that are picky about the vgabios.

It looks like it tickles the skifree bug.  Test case is windows vista
running the ancient dos program skifree.  When we last investigated
this, it looked like anything in the vgabios path that tries to write
to memory above 0xc0000 causes problems.  The SeaBIOS extra stack is
in that range by default.

It originally looked like old versions of x86emu were working.  But
when I apply a trivial debugging patch (see below) I then find I can't
start X.  Test case is an old fedora13 image.  I suspect that random
code movement could cause x86emu to sometimes hit a problematic
instruction and sometimes not.

I don't see these issues with sgabios.

-Kevin


--- a/src/sercon.c
+++ b/src/sercon.c
@@ -483,6 +483,7 @@ sercon_10_splitmode(struct bregs *regs)
         return;
     if (!GET_LOW(sercon_port))
         return;
+    dprintf(1, "here\n");
 
     switch (regs->ah) {
     case 0x01: