From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hollis Blanchard Date: Mon, 03 Mar 2008 05:56:48 +0000 Subject: [kvm-ppc-devel] PCI interrupts Message-Id: <1204523808.5581.11.camel@basalt> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: kvm-ppc@vger.kernel.org Despite the fact that we seem to have UART interrupts going missing, I tried playing with attaching a qemu PCI device (rtl8139) to a KVM 440 guest to see what would happen. First though I had to fix some poorly considered endian-handling qemu code, but after that "ip=dhcp" kernel parameter still doesn't work (ifup succeeds; DHCP fails). Debugging reveals that slirp (qemu's "user networking") does in fact see the DHCP request, and the rtl8139 code does in fact see the reply, but somehow qemu and the guest seem to disagree about which IRQ to expect the interrupt on. Hacking it so that they agree doesn't seem to fix the problem either, so right now I'm digging in to qemu's UIC emulation to see if there's an issue there. Right now I'm not certain if the guest is properly setting polarity (UIC0_PR), but more investigation is needed. Anyways, as an alternative for now, we may be able to get VirtIO working if we add a polling mode. It might be as easy as setting up a timer in virtio_pci.c that calls vp_interrupt() directly... -- Hollis Blanchard IBM Linux Technology Center ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ kvm-ppc-devel mailing list kvm-ppc-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/kvm-ppc-devel