From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?ISO-8859-2?Q?=A3ukasz_Ole=B6?= Subject: Re: Re: [vt-d][xen4-rc6] Hangs on startup Date: Mon, 22 Mar 2010 22:56:11 +0100 Message-ID: References: <20100317201446.GA7785@phenom.dumpdata.com> <20100318141909.GA14445@phenom.dumpdata.com> <20100322195010.GB32073@phenom.dumpdata.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-2 Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <20100322195010.GB32073@phenom.dumpdata.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xensource.com Errors-To: xen-devel-bounces@lists.xensource.com To: Konrad Rzeszutek Wilk Cc: xen-devel@lists.xensource.com List-Id: xen-devel@lists.xenproject.org Hi, 2010/3/22 Konrad Rzeszutek Wilk : > What is the version of Xen tools you have? Did you build them from > xen-unstable? Yes, there are from xen-unstable >> "0000:00:1a.1">/sys/bus/pci/devices/0000:00:1a.1/driver/unbind >> root@debian:~# echo "0000:00:1a.1">/sys/bus/pci/drivers/pci-stub/bind > > Where did you get the documentation for those steps? I would suggest you = use > the 'pciback' instead of 'pci-stub' module for this just to see if that > makes the 'pciback' later error go away. Well, I missread http://wiki.xensource.com/xenwiki/VTdHowTo I switched to pciback(I used http://wiki.xensource.com/xenwiki/Assign_hardware_to_DomU_with_PCIBack_as_m= odule) and this error no longer occurs. > Uhh. So you assigned the device to 'pci-stub', and somehow 'pciback' gets > involved. That is not good. PCI back is to be used _only_ with > PV guests - you on the other hand are running an HVM guest. > > Per the earlier statement, I would recommend you use the 'pciback' > instead of 'pci-stub' or just not compile pciback in and see what > happens. Keep in mind: PCI back module is only needed when you want to do= PV PCI > passthrough, which is not what you are doing. > Wiki says something else (http://wiki.xensource.com/xenwiki/VTdHowTo) So pciback or pci-stub for HVM? from wiki: "If using pv-ops dom0, also can use pci-stub to hide devices for assignment" >> [ =A0733.248052] pciback pci-3-0: 22 Couldn't locate PCI device >> (0000:00:1a.0)! perhaps already in-use? > > Can you paste your 'xenstore-ls /local/domain/0' output when this > happening? When using pci-stub: xenstore-ls /local/domain/0 vm =3D "/vm/00000000-0000-0000-0000-000000000000" device =3D "" control =3D "" platform-feature-multiprocessor-suspend =3D "1" error =3D "" backend =3D "" pci =3D "" 1 =3D "" 0 =3D "" error =3D "22 Couldn't locate PCI device (0000:01:01.0)! perhaps already in-use?" memory =3D "" target =3D "1017632" guest =3D "" hvmpv =3D "" data =3D "" cpu =3D "" 1 =3D "" availability =3D "online" 0 =3D "" availability =3D "online" description =3D "" console =3D "" limit =3D "1048576" type =3D "xenconsoled" domid =3D "0" name =3D "Domain-0" device-model =3D "" 1 =3D "" disable_pf =3D "0" state =3D "running" parameter =3D "0x20" backend =3D "" vfb =3D "" 1 =3D "" 0 =3D "" vncunused =3D "1" domain =3D "ExampleHVMDomain" vnc =3D "1" uuid =3D "e4f967e1-9ab9-6170-9392-711b7eee638d" vnclisten =3D "0.0.0.0" frontend =3D "/local/domain/1/device/vfb/0" state =3D "1" online =3D "1" frontend-id =3D "1" location =3D "0.0.0.0:5900" vbd =3D "" 1 =3D "" 768 =3D "" domain =3D "ExampleHVMDomain" frontend =3D "/local/domain/1/device/vbd/768" uuid =3D "f51680ce-7057-ed25-3a4f-aacd78e5f2b5" bootable =3D "1" dev =3D "hda" state =3D "2" params =3D "/media/debian.img" mode =3D "w" online =3D "1" frontend-id =3D "1" type =3D "file" node =3D "/dev/loop1" physical-device =3D "7:1" hotplug-status =3D "connected" 5632 =3D "" domain =3D "ExampleHVMDomain" frontend =3D "/local/domain/1/device/vbd/5632" uuid =3D "44367e07-409f-d676-9047-d84b1f2189e2" bootable =3D "0" dev =3D "hdc" state =3D "2" params =3D "/media/debian.iso" mode =3D "r" online =3D "1" frontend-id =3D "1" type =3D "file" node =3D "/dev/loop2" physical-device =3D "7:2" hotplug-status =3D "connected" pci =3D "" 1 =3D "" 0 =3D "" domain =3D "ExampleHVMDomain" key-0 =3D "01:01.0" vdevfn-0 =3D "20" uuid =3D "01ad7a64-bf43-b782-b4ad-bbcb9060c36a" dev-0 =3D "0000:01:01.0" frontend =3D "/local/domain/1/device/pci/0" state =3D "7" online =3D "1" frontend-id =3D "1" opts-0 =3D "msitranslate=3D1,power_mgmt=3D0" uuid-0 =3D "9db6e99f-31c4-8a7a-96c4-dd88e130abbe" num_devs =3D "1" state-0 =3D "1" console =3D "" 1 =3D "" 0 =3D "" domain =3D "ExampleHVMDomain" protocol =3D "vt100" uuid =3D "ae8d643f-cd4d-2256-2f55-c79736f96311" frontend =3D "/local/domain/1/device/console/0" state =3D "1" location =3D "3" online =3D "1" frontend-id =3D "1" >> (XEN) mm.c:859:d1 Error getting mfn 40004 (pfn 5555555555555555) from >> L1 entry 0000000040004037 for l1e_owner=3D1, pg_owner=3D1 >> ... >> And guest never starts. =A0Perhaps this is related to the firmware which >> is used by qla2xxx? > > So you see this _only_ when you pass in QLA2XXX card, right? When you > don't pass in a PCI card, you don't see this, correct? Exactly. This happens with both pciback and pci-stub. Regards, =A3ukasz Ole=B6