From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1KQO3n-0001dz-Vs for qemu-devel@nongnu.org; Tue, 05 Aug 2008 11:02:52 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1KQO3i-0001RQ-Cr for qemu-devel@nongnu.org; Tue, 05 Aug 2008 11:02:51 -0400 Received: from [199.232.76.173] (port=38884 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1KQO3i-0001R5-8v for qemu-devel@nongnu.org; Tue, 05 Aug 2008 11:02:46 -0400 Received: from mail.windriver.com ([147.11.1.11]:33810 helo=mail.wrs.com) by monty-python.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1KQO3h-0003mV-Oa for qemu-devel@nongnu.org; Tue, 05 Aug 2008 11:02:46 -0400 Message-ID: <48986B89.8000608@windriver.com> Date: Tue, 05 Aug 2008 10:02:33 -0500 From: Jason Wessel MIME-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH] upgrading emulated UART to 16550A References: <48985284.9060001@eu.citrix.com> In-Reply-To: <48985284.9060001@eu.citrix.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit 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 Cc: Ian Jackson , stefano.stabellini@eu.citrix.com Stefano Stabellini wrote: > This patch upgrades the emulated UART to 16550A, the code comes from > xen-unstable. > The main improvement was introduced with the following patch and > subsequent email thread: > > http://lists.xensource.com/archives/html/xen-devel/2007-12/msg00129.html > > I only made few changes to use qemu_get_clock instead of > clock_gettime, as Anthony suggested a little while ago. > > I didn't spend the time to figure out what was wrong, but I thought I'd try your new port emulation against KGDB which is known to work with real hardware, as well as the existing 16450 qemu uart support. For some reason, the initial interrupt if a character comes into the uart does not seem to get raised, so the kernel routine which handles the traffic never gets invoked. It is possible that the that the kgdb kernel driver has a defect, but this does seem to work on real hardware, so perhaps it is a timing problem or the qemu emulation is not quite accurate with respect to the real hardware. If you like, I can provide you a bzImage and the qemu invocation, such that you could see the problem first hand. Jason.