From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from list by monty-python.gnu.org with tmda-scanned (Exim 4.30) id 1BEOUE-0001gG-4n for qemu-devel@nongnu.org; Fri, 16 Apr 2004 04:14:10 -0400 Received: from mail by monty-python.gnu.org with spam-scanned (Exim 4.30) id 1BEOMh-0007Py-Mk for qemu-devel@nongnu.org; Fri, 16 Apr 2004 04:06:54 -0400 Received: from [81.228.10.108] (helo=av9-1-sn4.m-sp.skanova.net) by monty-python.gnu.org with esmtp (Exim 4.30) id 1BEOCh-0004AN-Q8 for qemu-devel@nongnu.org; Fri, 16 Apr 2004 03:56:03 -0400 Received: from smtp2-2-sn4.m-sp.skanova.net (smtp2-2-sn4.m-sp.skanova.net [81.228.10.182]) by av9-1-sn4.m-sp.skanova.net (Postfix) with ESMTP id C2B7337E4A for ; Fri, 16 Apr 2004 09:56:02 +0200 (CEST) Received: from putte2k (h151n2fls306o994.telia.com [81.225.243.151]) by smtp2-2-sn4.m-sp.skanova.net (Postfix) with SMTP id 95C5C37E5C for ; Fri, 16 Apr 2004 09:56:02 +0200 (CEST) Message-ID: <000901c42388$509ff080$0401a8c0@putte2k> From: "Mike Nordell" Date: Fri, 16 Apr 2004 09:56:21 +0200 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Subject: [Qemu-devel] Windows 2000 installed without a problem Reply-To: qemu-devel@nongnu.org List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Using "almost" CVS QEMU on Win32, installing NT5sp4 from a CD image, with "only" 64 MB of memory given to the guest, but disabling both NE2k and SB16 by means of setting nb_nics=0 before the "nb_nics1 = nb_nics;" at the end of pc_init() in pc.c, and disabling the calls to AUD_init() and SB16_init(), I've managed to install NT5sp4 (a "slipstreamed" to sp4 image) without too much problems. What I believe is worth noting is that I haven't seen a single gfx "glitch". DTC and failed to install, and some COM component failed to register, but that might be due to lack of a NIC (combined with another missing piece in the emulation, that MS' GUID generation uses in absence of a NIC), but besides that everything else seemingly worked like a charm. I therefore can with some certainty say that either the SB16 or the NE2k code (with the patch from Renzo applied) is a source of bugs - it probably overwrites memory it doesn't own. As the SB16 code has already read memory it didn't own, that'd be my first suspect. Could someone else that's been experiencing spurious guest-crashes, or installation failures, try to do the same - disable NE2k, OSS and SB16 as I outlined, and report failure or success? Both Windows and GNU/Linux hosts would probably be preferred. /Mike