From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jan Kiszka Subject: Re: PCI passthrough on Sony Vaio F11 laptop... Date: Sat, 13 Nov 2010 12:20:10 +0100 Message-ID: <4CDE746A.2090701@web.de> References: <4CDDCE40.9050001@brakkee.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="------------enig5FFEBE987ADC581E196C3AE3" Cc: kvm@vger.kernel.org To: Erik Brakkee Return-path: Received: from fmmailgate02.web.de ([217.72.192.227]:50291 "EHLO fmmailgate02.web.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753669Ab0KMLWl (ORCPT ); Sat, 13 Nov 2010 06:22:41 -0500 In-Reply-To: <4CDDCE40.9050001@brakkee.org> Sender: kvm-owner@vger.kernel.org List-ID: This is an OpenPGP/MIME signed message (RFC 2440 and 3156) --------------enig5FFEBE987ADC581E196C3AE3 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Am 13.11.2010 00:31, Erik Brakkee wrote: > Hi all, >=20 >=20 > I have KVM running successfully on opensuse 11.3 (64bit) with a opensus= e > 11.1 guest. I am using hardware virtualization and want to experiment > with PCI passthrough with the (wired) network card on my laptop because= > I want to repeat similar steps later on a server (with a TV card). The > main specs of the laptop are Intel Core I7 Q720 1.6GHz with 8MB of memo= ry. >=20 > I have followed the instructions on > http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Virtu= alization/chap-Virtualization-PCI_passthrough.html. > Everything is successful until I start the domain. >=20 > In that case I get the following messages in the logs: >=20 > Nov 13 00:23:45 pelican kernel: [ 3302.793043] pci-stub 0000:04:00.0: > claimed by stub > Nov 13 00:23:45 pelican kernel: [ 3302.793197] pci-stub 0000:04:00.0: > claimed by stub > Nov 13 00:23:45 pelican kernel: [ 3302.931578] > kvm_vm_ioctl_deassign_device: device hasn't been assigned before, so > cannot be deassigned > Nov 13 00:23:48 pelican libvirtd: 00:23:48.136: error : > qemuMonitorOpenUnix:269 : monitor socket did not show up.: Connection > refused > Nov 13 00:23:48 pelican libvirtd: 00:23:48.136: error : > qemuConnectMonitor:1192 : Failed to connect monitor for other >=20 > To be absolutely certa CPU0 CPU1 CPU2 =20 > CPU3 CPU4 CPU5 CPU6 CPU7 > 0: 371 0 0 203026 0 =20 > 0 0 0 IO-APIC-edge timer > 1: 12870 0 0 39 173 =20 > 0 0 0 IO-APIC-edge i8042 > 7: 259 0 0 0 0 =20 > 0 0 0 IO-APIC-edge > 8: 0 0 0 4 0 =20 > 0 0 0 IO-APIC-edge rtc0 > 9: 2823 0 0 141 0 =20 > 0 0 0 IO-APIC-fasteoi acpi > 12: 15 0 0 155 0 =20 > 0 0 0 IO-APIC-edge i8042 > 16: 0 298817 0 0 0 =20 > 0 634 0 IO-APIC-fasteoi ehci_hcd:usb1, ohci1394, > hda_intel, nvidia > 17: 0 0 0 0 0 =20 > 0 0 0 IO-APIC-fasteoi mmc0 > 19: 0 0 0 0 0 =20 > 0 0 0 IO-APIC-fasteoi mmc1 > 23: 0 0 99 0 0 =20 > 95116 0 0 IO-APIC-fasteoi ehci_hcd:usb2 > 24: 687633 0 0 0 0 =20 > 0 0 0 HPET_MSI-edge hpet2 > 25: 0 637204 0 0 0 =20 > 0 0 0 HPET_MSI-edge hpet3 > 26: 0 0 199161 0 0 =20 > 0 0 0 HPET_MSI-edge hpet4 > 27: 0 0 0 134606 0 =20 > 0 0 0 HPET_MSI-edge hpet5 > 28: 0 0 0 0 237717 =20 > 0 0 0 HPET_MSI-edge hpet6 > 34: 58784 0 0 7780 0 =20 > 0 0 0 PCI-MSI-edge ahci > 36: 939 0 0 0 138686 =20 > 0 0 0 PCI-MSI-edge iwlagn > 37: 0 0 0 0 0 =20 > 210 0 0 PCI-MSI-edge hda_intel > NMI: 0 0 0 0 0 =20 > 0 0 0 Non-maskable interrupts > LOC: 1535 1438 1347 1257 1168 =20 > 396803 120435 79080 Local timer interrupts > SPU: 0 0 0 0 0 =20 > 0 0 0 Spurious interrupts > PMI: 0 0 0 0 0 =20 > 0 0 0 Performance monitoring interrupts > PND: 0 0 0 0 0 =20 > 0 0 0 Performance pending work > RES: 4164 4332 2050 1413 3244 =20 > 2790 1630 1200 Rescheduling interrupts > CAL: 1394 1395 1279 1195 1431 =20 > 1430 1429 575 Function call interrupts > TLB: 5007 2623 3751 3045 5617 =20 > 3902 4486 4115 TLB shootdowns > TRM: 0 0 0 0 0 =20 > 0 0 0 Thermal event interrupts > THR: 0 0 0 0 0 =20 > 0 0 0 Threshold APIC interrupts > MCE: 0 0 0 0 0 =20 > 0 0 0 Machine check exceptions > MCP: 14 14 14 14 14 =20 > 14 14 14 Machine check polls > ERR: 259 > MIS: 0 > in I also removed the driver for the network card using 'rmmod sky2'. >=20 > I have also specified the intel_iommu=3Don option in the command line (= cat > /proc/cmdline gives: root=3D/dev/system/root resume=3D/dev/system/swap > splash=3Dsilent quiet nomodeset intel_iommu=3Don vga=3D0x314). Also, th= e > output of 'dmesg | grep -i iommu' gives: >=20 > [ 0.000000] Command line: root=3D/dev/system/root > resume=3D/dev/system/swap splash=3Dsilent quiet nomodeset intel_iommu=3D= on > vga=3D0x314 > [ 0.000000] Kernel command line: root=3D/dev/system/root > resume=3D/dev/system/swap splash=3Dsilent quiet nomodeset intel_iommu=3D= on > vga=3D0x314 > [ 0.000000] Intel-IOMMU: enabled >=20 > So it seems that IOMMU is available. >=20 > I read somewhere that there could be a problem in the interrupts and > some users had success by disabling them but I really don't know how to= > debug this. > The output of 'cat /proc/interrupts' is as follows: What IRQ is the sky2 using when assigned to the host? Is it really a shared IRQ (I bet not as it should be using MSI)? Also, check in the libvirt logs what qemu-kvm reports on the console. Jan --------------enig5FFEBE987ADC581E196C3AE3 Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.15 (GNU/Linux) Comment: Using GnuPG with SUSE - http://enigmail.mozdev.org/ iEYEARECAAYFAkzedGoACgkQitSsb3rl5xQePwCgyzGMGAkZBTmNSeTkPZsO/MuV CBEAoJ40db7r0d98k7erhFhdWmZHkqWO =X4z1 -----END PGP SIGNATURE----- --------------enig5FFEBE987ADC581E196C3AE3--