I''m currently having a hard time starting a PV domain on my notebook (HP EliteBook 6930p). This is the output of xm create -c config.cfg: Started domain c08d06e6-cb72-48ab-88a9-24360beff248 (id=3) [ 3.241721] Reserving virtual address space above 0xf5800000 [ 3.241721] Initializing cgroup subsys cpuset [ 3.241721] Initializing cgroup subsys cpu [ 3.241721] Linux version 2.6.30-rc6-andi-tip (root@steve-xen) (gcc version 4.3.2 (Debian 4.3.2-1.1) ) #3 SMP Wed Sep 2 10:33:12 BST 2009 [ 3.241721] KERNEL supported cpus: [ 3.241721] Intel GenuineIntel [ 3.241721] AMD AuthenticAMD [ 3.241721] NSC Geode by NSC [ 3.241721] Cyrix CyrixInstead [ 3.241721] Centaur CentaurHauls [ 3.241721] Transmeta GenuineTMx86 [ 3.241721] Transmeta TransmetaCPU [ 3.241721] UMC UMC UMC UMC [ 3.241721] ACPI in unprivileged domain disabled [ 3.241721] BIOS-provided physical RAM map: [ 3.241721] Xen: 0000000000000000 - 00000000000a0000 (usable) [ 3.241721] Xen: 00000000000a0000 - 0000000000100000 (reserved) [ 3.241721] Xen: 0000000000100000 - 0000000008000000 (usable) [ 3.241721] DMI not present or invalid. [ 3.241721] last_pfn = 0x8000 max_arch_pfn = 0x1000000 [ 3.241721] init_memory_mapping: 0000000000000000-0000000008000000 [ 3.241721] NX (Execute Disable) protection: active [ 3.241721] RAMDISK: 01531000 - 0253b000 [ 3.241721] 0MB HIGHMEM available. [ 3.241721] 128MB LOWMEM available. [ 3.241721] mapped low ram: 0 - 08000000 [ 3.241721] low ram: 0 - 08000000 [ 3.241721] node 0 low ram: 00000000 - 08000000 [ 3.241721] node 0 bootmap 00002000 - 00003000 [ 3.241721] (9 early reservations) ==> bootmem [0000000000 - 0008000000] [ 3.241721] #0 [0000000000 - 0000001000] BIOS data page ==> [0000000000 - 0000001000] [ 3.241721] #1 [000255e000 - 0002574000] XEN PAGETABLES ==> [000255e000 - 0002574000] [ 3.241721] #2 [0000001000 - 0000002000] EX TRAMPOLINE ==> [0000001000 - 0000002000] [ 3.241721] #3 [0000006000 - 0000007000] TRAMPOLINE ==> [0000006000 - 0000007000] [ 3.241721] #4 [0001000000 - 000140b930] TEXT DATA BSS ==> [0001000000 - 000140b930] [ 3.[ 0.004000] Freeing SMP alternatives: 19k freed [ 0.004318] Brought up 1 CPUs [ 0.004649] net_namespace: 1088 bytes [ 0.004659] Booting paravirtualized kernel on Xen [ 0.004666] Xen version: 3.5-unstable (preserve-AD) [ 0.004923] Grant table initialized [ 0.004985] NET: Registered protocol family 16 [ 0.005299] xenbus_probe wake_waiting [ 0.005310] xenbus_probe_init ok [ 0.005599] PCI: setting up Xen PCI frontend stub [ 0.008697] bio: create slab <bio-0> at 0 [ 0.008827] ACPI: Interpreter disabled. [ 0.008898] xenbus_probe_frontend_init bus registered ok [ 0.008898] xenbus_probe_devices device [ 0.060029] xenbus_probe_devices 1/4 vkbd [ 0.060029] xenbus_probe_device_type type vkbd [ 0.060039] xenbus_probe_device_type 1/1 0 [ 0.060092] xenbus_probe_device_type done [ 0.060100] xenbus_probe_devices 2/4 vfb [ 0.060106] xenbus_probe_device_type type vfb [ 0.060119] xenbus_probe_device_type 1/1 0 [ 0.060119] xenbus_probe_device_type done [ 0.060119] xenbus_probe_devices 3/4 vbd [ 0.060119] xenbus_probe_device_type type vbd [ 0.060119] xenbus_probe_device_type 1/1 2049 [ 0.060119] xenbus_probe_device_type done [ 0.060119] xenbus_probe_devices 4/4 console [ 0.060119] xenbus_probe_device_type type console [ 0.060119] xenbus_probe_device_type 1/1 0 [ 0.060119] xenbus_probe_device_type done [ 0.060119] xenbus_probe_devices done [ 0.060119] frontend_probe_and_watch devices probed ok [ 0.060119] frontend_probe_and_watch watch add ok ok [ 0.060119] frontend_probe_and_watch all done [ 0.060119] xen_balloon: Initialising balloon driver. [ 0.060654] Xen: Initializing Xen DMA ops [ 0.064743] pnp: PnP ACPI: disabled [ 0.064752] PnPBIOS: Disabled [ 0.065458] NET: Registered protocol family 2 [ 0.065534] IP route cache hash table entries: 1024 (order: 0, 4096 bytes) [ 0.065741] TCP established hash table entries: 4096 (order: 3, 32768 bytes) [ 0.065762] TCP bind hash table entries: 4096 (order: 3, 32768 bytes) [ 0.065781] TCP: Hash tables configured (established 4096 bind 4096) [ 0.065789] TCP reno registered [ 0.065856] NET: Registered protocol family 1 [ 0.065915] Trying to unpack rootfs image as initramfs... [ 0.086706] Freeing initrd memory: 16424k freed [ 0.096391] platform rtc_cmos: registered platform RTC device (no PNP device found) [ 0.097145] audit: initializing netlink socket (disabled) [ 0.097179] type=2000 audit(1251890690.402:1): initialized [ 0.097414] HugeTLB registered 2 MB page size, pre-allocated 0 pages [ 0.097520] VFS: Disk quotas dquot_6.5.2 [ 0.097555] Dquot-cache hash table entries: 1024 (order 0, 4096 bytes) [ 0.097645] msgmni has been set to 288 [ 0.097845] alg: No test for stdrng (krng) [ 0.097956] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 253) [ 0.097980] io scheduler noop registered [ 0.097994] io scheduler anticipatory registered [ 0.098013] io scheduler deadline registered [ 0.100079] io scheduler cfq registered (default) [ 0.105018] Console: switching to colour frame buffer device 100x37 [ 0.110575] console [tty0] enabled [ 0.110908] isapnp: ISA Plug & Play support disabled [ 0.111039] Event-channel device installed. [ 0.116334] Serial: 8250/16550 driver, 4 ports, IRQ sharing enabled [ 0.119292] brd: module loaded [ 0.123126] Initialising Xen virtual ethernet driver. [ 0.123445] PNP: No PS/2 controller found. Probing ports directly. [ 0.124306] i8042.c: No controller found. [ 0.124471] mice: PS/2 mouse device common for all mice [ 0.125880] input: Xen Virtual Keyboard as /class/input/input0 [ 0.126018] input: Xen Virtual Pointer as /class/input/input1 [ 0.131184] rtc_cmos rtc_cmos: rtc core: registered rtc_cmos as rtc0 [ 0.131480] cpuidle: using governor ladder [ 0.132622] cpuidle: using governor menu [ 0.133687] No iBFT detected. [ 0.135193] TCP cubic registered [ 0.136494] NET: Registered protocol family 17 [ 0.137584] Using IPI No-Shortcut mode [ 0.138934] registered taskstats version 1 [ 10.236023] XENBUS: Timeout connecting to device: device/vbd/2049 (state 3) [ 10.237188] XENBUS: Device with no driver: device/console/0 [ 10.238358] drivers/rtc/hctosys.c: unable to open rtc device (rtc0) [ 10.239554] Freeing unused kernel memory: 344k freed [ 10.804192] device-mapper: uevent: version 1.0.3 [ 10.807995] device-mapper: ioctl: 4.14.0-ioctl (2008-04-23) initialised: dm-devel@redhat.com After this line, no more output appears and the domain stays in blocked (''b'') state. This is the configuration file: kernel = ''/boot/vmlinuz-2.6.30-rc6-andi-tip'' ramdisk = ''/boot/initrd.img-2.6.30-rc6-andi-tip'' memory = ''128'' vif = [] disk = [''file:/myxendomains/c08d06e6-cb72-48ab-88a9-24360beff248/disk0.img,sda1,w''] root = ''/dev/sda1 ro'' name = ''c08d06e6-cb72-48ab-88a9-24360beff248'' uuid = ''c08d06e6-cb72-48ab-88a9-24360beff248'' extra = "video=xenfb xencons=tty1" vfb = [''type=vnc,vncdisplay=10''] vnc = 1 The kernel comes from git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git, with dom0 support deactivated. All kernel modules have been copied to the domain''s /lib/module. I have successfully tested that domain (Debian lenny installed with debootstrap) on another machine that runs the 2.6.18.8-xen kernel which ships with Xen. But on this notebook, only Xen-unstable with 2.6.30-rc6 works, and I can''t start PV domains. Do you have any hints what the problem could be? _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Wed, Sep 02, 2009 at 11:48:37AM +0100, Andreas Sommer wrote:> I''m currently having a hard time starting a PV domain on my notebook (HP > EliteBook 6930p). This is the output of xm create -c config.cfg: > > > This is the configuration file: > > kernel = ''/boot/vmlinuz-2.6.30-rc6-andi-tip'' > ramdisk = ''/boot/initrd.img-2.6.30-rc6-andi-tip'' > memory = ''128'' > vif = [] > disk = > [''file:/myxendomains/c08d06e6-cb72-48ab-88a9-24360beff248/disk0.img,sda1,w''] > root = ''/dev/sda1 ro'' > name = ''c08d06e6-cb72-48ab-88a9-24360beff248'' > uuid = ''c08d06e6-cb72-48ab-88a9-24360beff248'' > > extra = "video=xenfb xencons=tty1" > vfb = [''type=vnc,vncdisplay=10''] > > vnc = 1 > > > The kernel comes from > git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git, with dom0 > support deactivated. All kernel modules have been copied to the domain''s > /lib/module. > >Why don''t you run vanilla/standard 2.6.30.latest kernel? It has Xen pv_ops domU support included out-of-the-box. xen.git is only needed for dom0 support. (And the latest pv_ops dom0 kernel is 2.6.31-rc6, in rebase/master branch of that git tree). -- Pasi _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Wed, Sep 2, 2009 at 5:48 PM, Andreas Sommer<AndiDog@web.de> wrote:> [ 0.110575] console [tty0] enabledThis one is weird. You should have two consoles, one of them is hvc0> [ 10.807995] device-mapper: ioctl: 4.14.0-ioctl (2008-04-23) initialised: > dm-devel@redhat.com > > After this line, no more output appears and the domain stays in blocked > (''b'') state.> extra = "video=xenfb xencons=tty1" > vfb = [''type=vnc,vncdisplay=10'']Try this : - remove (or mark) those two lines - setup domU''s /etc/inittab to spawn login prompt on hvc0 In my last testing, 2.6.30 somehow won''t work (similar simptom as yours) if I enable Xen frame buffer. If I disable it (comment out vfb line) it works fine on Xen serial console, so perhaps it the same case.> I have successfully tested that domain (Debian lenny installed with > debootstrap) on another machine that runs the 2.6.18.8-xen kernel which > ships with Xen. But on this notebook, only Xen-unstable with 2.6.30-rc6 > works, and I can''t start PV domains.If you know that 2.6.18 kernel works, why not use that for domU kernel? dom0 and domU kernel does not have to be the same version. It doesn''t even have to be the same OS. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha wrote:> On Wed, Sep 2, 2009 at 5:48 PM, Andreas Sommer<AndiDog@web.de> wrote: > > >> [ 0.110575] console [tty0] enabled >> > > This one is weird. You should have two consoles, one of them is hvc0 > >> extra = "video=xenfb xencons=tty1" >> vfb = [''type=vnc,vncdisplay=10''] >> > > Try this : > - remove (or mark) those two lines > - setup domU''s /etc/inittab to spawn login prompt on hvc0 > > In my last testing, 2.6.30 somehow won''t work (similar simptom as > yours) if I enable Xen frame buffer. If I disable it (comment out vfb > line) it works fine on Xen serial console, so perhaps it the same > case. > > >Thanks for the suggestion, I will try that tomorrow.>> I have successfully tested that domain (Debian lenny installed with >> debootstrap) on another machine that runs the 2.6.18.8-xen kernel which >> ships with Xen. But on this notebook, only Xen-unstable with 2.6.30-rc6 >> works, and I can''t start PV domains. >> > > If you know that 2.6.18 kernel works, why not use that for domU > kernel? dom0 and domU kernel does not have to be the same version. It > doesn''t even have to be the same OS. > >In my experience, dom0 and PV domU kernels of different versions don''t /always/ work together perfectly. Anyway, the reason why I''m not using the tested 2.6.18.8-xen kernel is that it does not work - I don''t remember the problem so I''ll try that kernel again tomorrow. _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Fajar A. Nugraha wrote:> >> extra = "video=xenfb xencons=tty1" >> vfb = [''type=vnc,vncdisplay=10''] >> > > Try this : > - remove (or mark) those two lines > - setup domU''s /etc/inittab to spawn login prompt on hvc0 > > In my last testing, 2.6.30 somehow won''t work (similar simptom as > yours) if I enable Xen frame buffer. If I disable it (comment out vfb > line) it works fine on Xen serial console, so perhaps it the same > case. > > > >> I have successfully tested that domain (Debian lenny installed with >> debootstrap) on another machine that runs the 2.6.18.8-xen kernel which >> ships with Xen. But on this notebook, only Xen-unstable with 2.6.30-rc6 >> works, and I can''t start PV domains. >> > > If you know that 2.6.18 kernel works, why not use that for domU > kernel? dom0 and domU kernel does not have to be the same version. It > doesn''t even have to be the same OS. >I tried the following kernels (and removed the framebuffer configuration), with no luck: Output for 2.6.18.8 and 2.6.27.5 (both are from the xenbits.xensource.com GIT repo; output is the same): ... [ 300.030530] RAMDISK driver initialized: 16 RAM disks of 8192K size 1024 blocksize [ 300.031631] Xen virtual console successfully installed as tty1 [ 300.031653] Event-channel device installed. [ 300.037578] blktap_device_init: blktap device major 254 [ 300.037583] blktap_ring_init: blktap ring major: 253 [ 300.057235] netfront: Initialising virtual ethernet driver. [ 300.057804] PNP: No PS/2 controller found. Probing ports directly. [ 300.058654] i8042.c: No controller found. [ 300.058688] mice: PS/2 mouse device common for all mice [ 300.058733] TCP cubic registered [ 300.058738] NET: Registered protocol family 1 [ 300.058743] NET: Registered protocol family 17 [ 300.058747] Using IPI No-Shortcut mode [ 300.058785] PCI IO multiplexer device installed. [ 305.154455] XENBUS: Waiting for devices to initialise: 295s...290s...285s...280s...275s...270s...265s...260s...255s...250s...245s...240s...235s...230s...225s...220s...215s...210s...205s...200s...195s...190s...185s...180s...175s...170s...165s...160s...155s...150s...145s...140s...135s...130s...125s...120s...115s...110s...105s...100s...95s...90s...85s...80s...75s...70s...65s...60s...55s...50s...45s...40s...35s...30s...25s...20s...15s...10s...5s...0s... [ 599.991814] XENBUS: Timeout connecting to device: device/vbd/2049 (local state 3, remote state 1) [ 599.991820] XENBUS: Device not ready: device/vbd/2049 [ 599.991823] XENBUS: Device with no driver: device/console/0 [ 599.991828] drivers/rtc/hctosys.c: unable to open rtc device (rtc0) [ 599.991890] Freeing unused kernel memory: 148k freed Loading, please wait... [ 599.992852] [ 599.992857] *************************************************************** [ 599.992860] *************************************************************** [ 599.992864] ** WARNING: Currently emulating unsupported memory accesses ** [ 599.992867] ** in /lib/tls glibc libraries. The emulation is ** [ 599.992871] ** slow. To ensure full performance you should ** [ 599.992874] ** install a ''xen-friendly'' (nosegneg) version of ** [ 599.992880] ** the library, or disable tls support by executing ** [ 599.992884] ** the following as root: ** [ 599.992887] ** mv /lib/tls /lib/tls.disabled ** [ 599.992892] ** Offending process: init (pid=179) ** [ 599.992896] *************************************************************** [ 599.992899] *************************************************************** [ 599.992902] [ 599.992905] Continuing... [ 604.953911] Begin: Loading essential drivers ... done. Begin: Running /scripts/init-premount ... FATAL: Error inserting fan (/lib/modules/2.6.18.8/kernel/drivers/acpi/fan.ko): No such device FATAL: Error inserting thermal (/lib/modules/2.6.18.8/kernel/drivers/acpi/thermal.ko): No such device done. Begin: Mounting root file system ... Begin: Running /scripts/local-top ... [ 605.203803] device-mapper: ioctl: 4.7.0-ioctl (2006-06-24) initialised: dm-devel@redhat.com done. Begin: Waiting for root file system ... done. Gave up waiting for root device. Common problems: - Boot args (cat /proc/cmdline) - Check rootdelay= (did the system wait long enough?) - Check root= (did the system wait for the right device?) - Missing modules (cat /proc/modules; ls /dev) ALERT! /dev/sda1 does not exist. Dropping to a shell! BusyBox v1.10.2 (Debian 1:1.10.2-2) built-in shell (ash) Enter ''help'' for a list of built-in commands. /bin/sh: can''t access tty; job control turned off (initramfs) Output for 2.6.31-rc8 (mainline): Using config file "./config.cfg". Started domain c08d06e6-cb72-48ab-88a9-24360beff248 (id=5) Then the CPU goes up to 100% and dom0 hardly reacts to my user input... Output for 2.6.30-rc6-tip (from jeremy repository, but without dom0 support): ... Begin: Mounting root file system ... Begin: Running /scripts/local-top ... [ 10.691178] device-mapper: uevent: version 1.0.3 [ 10.692853] device-mapper: ioctl: 4.14.0-ioctl (2008-04-23) initialised: dm-devel@redhat.com done. Begin: Waiting for root file system ... Then it waits forever for the root file system. Note that it''s in my configuration (root = "/dev/sda1 ro"). So I''m pretty much stuck, no matter which kernel version :(... _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
On Thu, Sep 3, 2009 at 3:29 PM, Andreas Sommer<AndiDog@web.de> wrote:> Fajar A. Nugraha wrote:> extra = "video=xenfb xencons=tty1" > vfb = [''type=vnc,vncdisplay=10''] > > > Try this : > - remove (or mark) those two lines > - setup domU''s /etc/inittab to spawn login prompt on hvc0> I tried the following kernels (and removed the framebuffer configuration), > with no luck: > > Output for 2.6.18.8 and 2.6.27.5 (both are from the xenbits.xensource.com > GIT repo; output is the same):> [ 300.031631] Xen virtual console successfully installed as tty1There''s no mention of hvc0 or xvc0. Do you still have xencons=tty1 line? Try removing it first. I find it easier to debug that way.> [ 599.991814] XENBUS: Timeout connecting to device: device/vbd/2049 (local > state 3, remote state 1) > [ 599.991820] XENBUS: Device not ready: device/vbd/2049That one shows something wrong with domU''s disk. Perhaps you don''t have the block device frontend compiled on domU kernel? Or maybe your dom0 kernel does not have block device backend compiled?> [ 599.992864] ** WARNING: Currently emulating unsupported memory > accesses ** > [ 599.992867] ** in /lib/tls glibc libraries. The emulationThat warning can sometimes be ignored. You can investigate later after domU boots correctly.> Gave up waiting for root device. Common problems: > - Boot args (cat /proc/cmdline) > - Check rootdelay= (did the system wait long enough?) > - Check root= (did the system wait for the right device?) > - Missing modules (cat /proc/modules; ls /dev) > ALERT! /dev/sda1 does not exist. Dropping to a shell!That one is bad :D Make sure : - your domU disk backend is accessible - you have block device frontend driver on domU kernel - you have block device backend driver on dom0 kernel - you use correct name AFAIK later kernels will silently change all reference to hda/sda for domU disk to xvda, so you might want to try changing that.> Output for 2.6.31-rc8 (mainline): > > Using config file "./config.cfg". > Started domain c08d06e6-cb72-48ab-88a9-24360beff248 (id=5) > > Then the CPU goes up to 100% and dom0 hardly reacts to my user input...vanilla kernel is somewhat tricky. In my config I had to compile SCTP support built in to get it to work, otherwise domU can''t start at all. You might want to start with a known good config.> Output for 2.6.30-rc6-tip (from jeremy repository, but without dom0 > support): > > ... > Begin: Mounting root file system ... Begin: Running /scripts/local-top ... > [ 10.691178] device-mapper: uevent: version 1.0.3 > [ 10.692853] device-mapper: ioctl: 4.14.0-ioctl (2008-04-23) initialised: > dm-devel@redhat.com > done. > Begin: Waiting for root file system ... > > Then it waits forever for the root file system. Note that it''s in my > configuration (root = "/dev/sda1 ro").Try changing it to /dev/xvda1> So I''m pretty much stuck, no matter which kernel version :(...You mentioned earlier you already had a working domU on another machine. What happens when you copy it (disk image, kernel, and config) to this machine? It should work. -- Fajar _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users