From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.33) id 1BXQQA-0006Xh-4Y for qemu-devel@nongnu.org; Mon, 07 Jun 2004 16:08:38 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.33) id 1BXQQ8-0006WH-B3 for qemu-devel@nongnu.org; Mon, 07 Jun 2004 16:08:37 -0400 Received: from [199.232.76.173] (helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.33) id 1BXQQ8-0006WE-7V for qemu-devel@nongnu.org; Mon, 07 Jun 2004 16:08:36 -0400 Received: from [81.228.9.107] (helo=av2-1-sn3.vrr.skanova.net) by monty-python.gnu.org with esmtp (Exim 4.34) id 1BXQPT-000339-BY for qemu-devel@nongnu.org; Mon, 07 Jun 2004 16:07:55 -0400 Received: from smtp1-2-sn3.vrr.skanova.net (smtp1-2-sn3.vrr.skanova.net [81.228.9.178]) by av2-1-sn3.vrr.skanova.net (Postfix) with ESMTP id 8B85737E42 for ; Mon, 7 Jun 2004 22:07:54 +0200 (CEST) Received: from putte2k (h151n2fls306o994.telia.com [81.225.243.151]) by smtp1-2-sn3.vrr.skanova.net (Postfix) with SMTP id 6984438012 for ; Mon, 7 Jun 2004 22:07:54 +0200 (CEST) Message-ID: <000401c44ccb$24e25020$0401a8c0@putte2k> From: "Mike Nordell" Date: Mon, 7 Jun 2004 22:07:56 +0200 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Subject: [Qemu-devel] Re: Status of user-mode networking for win32 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 I wrote: > I vote to commit the patch for ne2000, aligning on 16-bit, not 32-bit. > > It made (among other things) "ipconfig /renew" work in Windows 2000 guest, > on Windows 2000 host. Seems I opened my mouth too early. With this patch in place, it seems that every second time I boot the Windows 2000 guest in QEMU I get a somewhat working connection between slirp and the guest OS (I've been able to ICMP ping the host, using it's real IP interface, but ping-ing other hosts on the net does not work(*)). The other times it just won't talk to it as it should (even that I do get slirp debug output). If it works the same or not without the patch I haven't yet tested. Maybe it's better with it, maybe not - I just wanted to retract my absolute support of the patch ASAP until I've been able to reasearch the problem further. (*) I believe the largest part of the problem is that slirp is so far only aware of errno. In Win32, since the C runtime library is just that, a library, while on POSIX I believe it's an integral part of the environment, no e.g. socket errors propagate to errno - and things go south. As an example I can mention a recvfrom() (on a UDP socket, obviously) that fails, and while errno claims 9 (EBADF - "Bad file descriptor"), WSAGetLastError says WSAECONNRESET - which means the previous send resulted in an ICMP port unreachable. /Mike