March 2009 - openSUSE Virtual - openSUSE Mailing Lists

[opensuse-virtual] Updating a OpenSuse 11.0 DomU to OpenSuse 11.1 problem
by Jörg Spilker 18 Apr '09

18 Apr '09

I recently installed OpenSuse 11.0 together with two OpenSuse 11.0 DomU´s. Now i want to update the complete system to OpenSuse 11.1. I had no problem´s to update the Dom0 system to OpenSuse 11.1. However updating the DomU´s seems to be a problem. The first install of the DomU´s was a install into directory with whole LVM´s partitions exported as disks. My layout is as follows: disk=[ 'phy:/dev/domX/daolin_root,xvda,w' , 'phy:/dev/domX/daolin_boot,xvdb,w' , 'phy:/dev/domX/daolin_usr,xvdc,w' , 'phy:/dev/domX/daolin_opt,xvdd,w' , 'phy:/dev/domX/daolin_var,xvde,w' , 'phy:/dev/domX/daolin_src,xvdf,w' , 'phy:/dev/domX/daolin_tmp,xvdg,w' , 'phy:/dev/data/home,xvdh,w' , 'phy:/dev/data/foto,xvdi,w' , 'phy:/dev/data/imap,xvdj,w' , 'phy:/dev/data/news,xvdk,w' , 'phy:/dev/data/squid,xvdl,w' , 'phy:/dev/data/srv,xvdm,w' , 'phy:/dev/hda1,xvdn,w' , 'phy:/dev/backup/backup,xvdo,w' , 'phy:/dev/sr0,xvdy:cdrom,r' , 'phy:/dev/sr1,xvdz:cdrom,r'] for the initial install i did mount the lvm volumens to subdirs of /tmp/install and did a install into this directory. The fstab of the system contains entry like /dev/xvda for mountpoints (i.e a whole disk, not a partition) This was working fine. But i don´t know how i know can update this system to 11.1. Installation into a directory seems to only can handle fresh installs. I created a special startup file for xen with allows a disk install of OpenSuse 11.1. However this doesn´t work because the exported disks are not recognized as expected. When i choose the update option, the installer shows no partitions at all. I have no idea to solve this problem. Do i really have to partition an exported lvm volume first in the DomU before using it as a mountpoint ? Greetings, Jörg -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

2 7

[opensuse-virtual] crash @ pci passthrough of pci/sata device to DomU
by PGNet 31 Mar '09

31 Mar '09

this is a suggested "not a bug" redirect from, "Bug 489846 - pci passthrough error non-page-aligned MMIO BAR found" https://bugzilla.novell.com/show_bug.cgi?id=489846 any help debuggung/fixing this is, of course, appreciated! i'm running openSUSE 11.1, uname -ri 2.6.27.21-7-xen x86_64 booting to Xen, @grub title Xen root (hd0,0) kernel /xen.gz dom0_mem=768M vga=gfx-1280x1024x32 console=vga,com1 com1=57600,8n1 module /vmlinuz-xen root=/dev/VG_Dom0/LV_ROOT resume=/dev/VG_Swap/LV_SWAP vga=0x31a console=tty0 console=xvc0,57600 elevator=cfq max_loop=64 pci_devs_to_hide=(04:07.0)(00:14.0)(04:08.0) reassigndev=0004:07.0 module /initrd-xen passing through two devices to the DomU, cat nas.cfg ... pci = [ '04:07.0', '04:08.0' ] ... where, lspci | egrep "04:07.0|04:08.0" 04:07.0 RAID bus controller: Silicon Image, Inc. SiI 3124 PCI-X Serial ATA Controller (rev 02) 04:08.0 FireWire (IEEE 1394): VIA Technologies, Inc. VT6306 Fire II IEEE 1394 OHCI Link Layer Controller (rev c0) note: really only want/need "04:08.0", but "04:07.0" needs to come along for the ride (cref: "Bug 489735 - pci passthrough of sata_sil24 device requires that firewire must be co-assigned to the same guest"" https://bugzilla.novell.com/show_bug.cgi?id=48973 ) crashes @, xm create -c nas.cfg @console, ... sd 0:0:0:0: [sda] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA vol_id[790]: segfault at 0 ip 0000000000000000 sp 00007ffffdb25220 error 14 in vol_id[400000+5000] ------------[ cut here ]------------ kernel BUG at mm/mmap.c:2145! invalid opcode: 0000 [1] SMP last sysfs file: /sys/devices/xen/pci-0/pci0000:00/0000:00:01.0/host1/target1:0:0/1:0:0:0/type CPU 0 Modules linked in: sg sd_mod crc_t10dif sata_sil24 libata ohci1394 scsi_mod forcedeth ieee1394 dock rtc_core rtc_lib joydev linear dm_snapshot dm_mod ext3 jbd mbcache xennet xenblk cdrom Supported: Yes Pid: 790, comm: vol_id Not tainted 2.6.27.21-4-xen #1 RIP: e030:[<ffffffff8028a96b>] [<ffffffff8028a96b>] exit_mmap+0xe6/0xf2 RSP: e02b:ffff8800172ddd78 EFLAGS: 00010202 RAX: 0000000000000000 RBX: ffff8800013a0140 RCX: 0000000000000045 RDX: ffffffffff5fc000 RSI: ffff880016e9d820 RDI: ffff88001780f680 RBP: 0000000000000000 R08: 0000000000000000 R09: ffff880001d92200 R10: 0000000000000010 R11: ffffffff80320c8f R12: ffff880016e9d7c0 R13: ffff880016e9d820 R14: ffff8800172ddf58 R15: ffff8800172ddf08 FS: 00007fe1f5b12780(0000) GS:ffffffff80761080(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process vol_id (pid: 790, threadinfo ffff8800172dc000, task ffff8800168642c0) Stack: 000000000000007d ffff8800013a0140 ffff880016e9d7c0 ffff8800168642c0 ffff880016e9d7c0 ffffffff80230eaa ffff8800168648ac ffffffff80234adf 0000000001241000 ffff8800168642c0 0000000000000001 ffff8800168642c0 Call Trace: [<ffffffff80230eaa>] mmput+0x20/0xbc [<ffffffff80234adf>] exit_mm+0x101/0x10c [<ffffffff80236907>] do_exit+0x208/0x304 [<ffffffff80236aa2>] do_group_exit+0x9f/0xcb [<ffffffff8024028f>] get_signal_to_deliver+0x3b0/0x405 [<ffffffff8020ad3e>] do_signal+0x64/0x191 [<ffffffff8020ae78>] do_notify_resume+0xd/0x2c [<ffffffff8020ba3e>] retint_signal+0x57/0xa9 Code: 7b 18 e8 b5 76 00 00 c7 43 08 00 00 00 00 eb 0b 48 89 ef e8 d4 fe ff ff 48 89 c5 48 85 ed 75 f0 49 83 bc 24 e8 00 00 00 00 74 04 <0f> 0b eb fe 5f 41 58 5b 5d 41 5c c3 41 56 41 be f4 ff ff ff 41 RIP [<ffffffff8028a96b>] exit_mmap+0xe6/0xf2 RSP <ffff8800172ddd78> ---[ end trace de6d924e597e88a3 ]--- Fixing recursive fault but reboot is needed! ---------> xm destroy nas Adding 524280k swap on /dev/mapper/vg0-lv_swap. Priority:-1 extents:1 across:524280k general protection fault: 0000 [2] SMP last sysfs file: /sys/kernel/uevent_seqnum CPU 0 Modules linked in: sg sd_mod crc_t10dif sata_sil24 libata ohci1394 scsi_mod forcedeth ieee1394 dock rtc_core rtc_lib joydev linear dm_snapshot dm_mod ext3 jbd mbcache xennet xenblk cdrom Supported: Yes Pid: 829, comm: vgscan Tainted: G D 2.6.27.21-4-xen #1 RIP: e030:[<ffffffff802ee780>] [<ffffffff802ee780>] sysfs_readdir+0xd9/0x13e RSP: e02b:ffff8800172dfeb8 EFLAGS: 00010282 RAX: 0000000000000000 RBX: ffff8800178fcd20 RCX: ffffffffffffffff RDX: 0000000000000000 RSI: 14ed7f01c5e864b0 RDI: 14ed7f01c5e864b0 RBP: ffff880016e953c0 R08: 0000000000000aaa R09: 000000000000000a R10: 0000000013a97e56 R11: ffffffff803211c1 R12: ffffffff802aabbc R13: ffff8800172dff38 R14: 0000000000000000 R15: ffffffffffffffff FS: 00007fbcf58e0780(0000) GS:ffffffff80761080(0000) knlGS:0000000000000000 CS: e033 DS: 0000 ES: 0000 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process vgscan (pid: 829, threadinfo ffff8800172de000, task ffff880016420300) Stack: 0000000013a97e56 00000000fffffffe ffff880016e953c0 ffff88001749d6e0 ffff88001749d798 ffffffff802aabbc ffff8800172dff38 ffffffff802aae3f 00000000000000ff 00000000fffffff7 ffff880016e953c0 0000000000000000 Call Trace: [<ffffffff802aae3f>] vfs_readdir+0x75/0xa7 [<ffffffff802aaeeb>] sys_getdents64+0x7a/0xbf [<ffffffff8020b3b8>] system_call_fastpath+0x16/0x1b [<00007fbcf40f4f3a>] 0x7fbcf40f4f3a Code: 61 17 00 48 8b 5b 28 eb 04 48 8b 5b 10 48 85 db 74 74 48 8b 43 38 48 39 45 38 77 ed eb 40 48 8b 73 18 44 88 f0 4c 89 f9 48 89 f7 <f2> ae 4c 89 ef 48 f7 d1 48 8d 41 ff 48 8b 4b 38 89 c2 48 89 4d RIP [<ffffffff802ee780>] sysfs_readdir+0xd9/0x13e RSP <ffff8800172dfeb8> ---[ end trace de6d924e597e88a3 ]--- is this a config problem? though i'm told this is not a bug, i _do_ see "kernel BUG at mm/mmap.c:2145!" above. so, general confusion abounds ... thanks. -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 1

[opensuse-virtual] pciback failing to seize _multiple_ devices
by PGNet 27 Mar '09

27 Mar '09

I'm attempting to passthrough _multiple_ PCI cards from openSUSE 11.1 Dom0 -> DomUs. e.g., 00:14.0 Bridge: nVidia Corporation MCP51 Ethernet Controller (rev a3) 04:07.0 RAID bus controller: Silicon Image, Inc. SiI 3124 PCI-X Serial ATA Controller (rev 02) 04:08.0 FireWire (IEEE 1394): VIA Technologies, Inc. VT6306 Fire II IEEE 1394 OHCI Link Layer Controller (rev c0) Following success/example in an old post @ http://archive.netbsd.se/?ml=xen-users&a=2007-12&m=5961315 Subject: Re: [Xen-users] pciback.hide with ISDN Cards Date: 2007-12-20 15:33:53 i've similarly @ /boot/grub/menul.lst ... title Xen (symlink) NORMAL root (hd0,0) kernel /xen.gz module /vmlinuz-xen root=/dev/vg0/ROOT resume=/dev/vg0/SWAP elevator=cfq max_loop=64 physdev_dom0_hide=(00:14.0)(04:07.0)(04:08.0) module /initrd-xen & @ /etc/init.d/boot.local modprobe pciback hide='(00:14.0)(04:07.0)(04:08.0)' echo -n 0000:00:14.0 > /sys/bus/pci/drivers/pciback/permissive echo -n 0000:04:07.0 > /sys/bus/pci/drivers/pciback/permissive echo -n 0000:04:08.0 > /sys/bus/pci/drivers/pciback/permissive &, additionally, @ my DomU .cfg, pci = [ '00:14.0', '04:07.0', '04:08.0' ] after reboot, however, pciback has only "seized" a single device, dmesg | grep pciback pciback 0000:00:14.0: seizing device pciback 0000:00:14.0: PCI INT A -> Link[LMAC] -> GSI 20 (level, low) -> IRQ 20 pciback 0000:00:14.0: PCI INT A disabled pciback 0000:00:14.0: enabling permissive mode configuration space accesses! pciback 0000:00:14.0: permissive mode is potentially unsafe! (note, that in the link's example, multiple devs are successfully seized) @ DomU launch, with pci = [ '00:14.0', '04:07.0', '04:08.0' ] fails to boot, returning, Error: pci: PCI Backend does not own device 0000:04:07.0 See the pciback.hide kernel command-line parameter or bind your slot/device to the PCI backend using sysfs but with just the seized interface pci = [ '00:14.0' ] DomU boots & logs-in ok, & lspci 00:00.0 Bridge: nVidia Corporation MCP51 Ethernet Controller (rev a3) what add'l/corrected config is required for pciback to correctly seize all the devices? or, is this likely a bug? thanks. -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 2

[opensuse-virtual] paravirtualized USB support (per-device passthrough) in OS 11.1
by PGNet 24 Mar '09

24 Mar '09

i'm running OS 11.1's xen-3.3.1_18494_03-1.8 on 2.6.27.20-5-xen x86_64 from .../Kernel:/SL111_BRANCH/openSUSE_11.1/. pci passthrough of ethernet from Dom0 to guests works fine. i recently added a USB 10/100 Ethernet adapter to my mix. @ Dom0, lsusb | grep -i Belkin Bus 001 Device 004: ID 050d:0121 Belkin Components F5D5050 100Mbps Ethernet dmesg | egrep -i belkin pegasus 2-4:1.0: eth1, Belkin F5D5050 USB Ethernet, 00:05:1b:42:16:f3 it's fully functional. digging to see how to passthrough the single USB dev to a guest, i came across a presentation from the recent Xen/Oracle summit: "Paravirtualized USB Support for Xen: Status Update" www.xen.org/files/xensummit_oracle09/PVUSB.pdf which seems to indicate that support for single-device passthru exists, works, and has decent (~80% of native) performance. iirc, it's currently targeted @ upstream xen sources, and can't be simply/blindly applied to OS's xen sources. anyone know if there _is_ any current effort/progress on per-device USB PV support in OS/SLES? -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

2 1

Re: [opensuse-virtual] Can I delete domU images?
by Nicholas Thirkettle OSB 23 Mar '09

23 Mar '09

This is a great help and saves me from opening Yast2 each time I want to start the guest. Thanks for your help. Br. Nicholas > > You may want to try using the xend commands directly from the command > line. Our experience has been that some comands that do not work well > from the yast2 tool interface will work with xend. For example, some > of our virtual machines will not start correctly if we use yast2, but > if we just use "xm start <machine-name>" from the command line, > everything runs just fine. We are also using opensuse 11.1 as our > development platform. > > Sincerely, > > John > > ---- > What is the formal meaning of the one-line program > #include "/dev/tty" > > J.P. McDermott building 12 > Code 5542 mcdermott(a)itd.nrl.navy.mil > Naval Research Laboratory voice: +1 202.404.8301 > Washington, DC 20375, US fax: +1 202.404.7942 -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 0

Re: [opensuse-virtual] Can I delete domU images?
by Nicholas Thirkettle OSB 23 Mar '09

23 Mar '09

This appears to have worked. Windows XP still reports drive C: as 8GB but when I list the directory it shows the image file is 12GB. My thought is that it will overwrite the zero's as needed. (I'm not sure I know what I'm talking about but I'm learning. Thanks for your help. Br. Nicholas On Monday 23 March 2009 08:22:30 Rob Verduijn wrote: > No garanties, use at your own risk, > make a backup first > > > Maybe this helps : > > This only works for vm's that use an image file that contains a partition > table. > > howto enlarge a image file. > > first dd if=/dev/zero of=dummyfile bs=1024k count=1024 > This creates a 1gb file that is completely empty (or filled with > zeroes) with the name dummyfile > > shut down the machine with the image you wish to enlarge > > append the empty file to your image with the following statement > > cat dummyfile >> path/to/image/you/wish/to/enlarge > > repeat for each gb you wish to add (or make a bigger dummyfile) > > This way there will be additional free space at the end off the image > that you can use. > > Cheers > Rob > -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 0

Re: [opensuse-virtual] Can I delete domU images?
by Nicholas Thirkettle OSB 23 Mar '09

23 Mar '09

Thanks for the help! I forgot to mention that I'm using opensuse 11.1 but I guess that doesn't matter much. I deleted in the VMM and then deleted the file itself. Anyway, it worked. I still have the problem of the small image file. I have tried to enlarge it but only corrupted it. Good thing I made a backup. I'm out of partitions so I may try to put a partition on my external backup drive to test. Thanks again. Br. Nicholas On Monday 23 March 2009 06:54:02 John McDermott (U.S. Navy Employee) wrote: > We have successfully deleted the files in /var/libs/xen/images, but > you have to edit .disk-list by hand. Perhaps the yast2 code needs to > be extended to do this as part of its functions? > > Sincerely, > > John McDermott > > On Mar 21, 2009, at 4:23 PM, Nicholas Thirkettle OSB wrote: > > Hello list, > > > > I tried twice, unsuccessfully, to install operating systems as > > guests in xen. > > The third time I was successful at installing Windows XP. Now I have > > two > > unused images on my drive. I deleted the two that I couldn't use in > > Virtual > > Machine Manager but they are still showing up in /var/libs/xen/ > > images. One of > > the unused is 4 GB and the other is 8 GB. The good windows > > installation has > > already just about used up all of the 8 GB image and I would like to > > know if I > > can delete the two that I don't use and expand the one that I do > > use. I don't > > know if that makes sense. Thank you for any help > > > > Br. Nicholas > > > > -- > > To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org > > For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org > > ---- > What is the formal meaning of the one-line program > #include "/dev/tty" > > J.P. McDermott building 12 > Code 5542 mcdermott(a)itd.nrl.navy.mil > Naval Research Laboratory voice: +1 202.404.8301 > Washington, DC 20375, US fax: +1 202.404.7942 -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 0

[opensuse-virtual] Can I delete domU images?
by Nicholas Thirkettle OSB 21 Mar '09

21 Mar '09

Hello list, I tried twice, unsuccessfully, to install operating systems as guests in xen. The third time I was successful at installing Windows XP. Now I have two unused images on my drive. I deleted the two that I couldn't use in Virtual Machine Manager but they are still showing up in /var/libs/xen/images. One of the unused is 4 GB and the other is 8 GB. The good windows installation has already just about used up all of the 8 GB image and I would like to know if I can delete the two that I don't use and expand the one that I do use. I don't know if that makes sense. Thank you for any help Br. Nicholas -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 0

[opensuse-virtual] sil3124 driver problem _only_ with kernel-xen; kernel-default ok
by PGNet 11 Mar '09

11 Mar '09

i'm running OS11.1 with an external RAID array hanging off a sata_sil24 card. boot to kernel-default is (now, as of bug #463829 resolution) ok, with external drives/array fully functional. boot to kernel-xen, however, fails to boot, dropping me to just a root login @ 'maintenance mode'. the sata_sil24-attached drives fail to be recognized. i've filed a bug, Bug 482220 - IRQ delivery problem (?) for sata_sil24 controller when boot @ kernel-xen https://bugzilla.novell.com/show_bug.cgi?id=482220 that's under Component:Kernel. i've had suggestions both that the issue is, and is not, IRQ related :-/ that's yet to be determined what's for certain is that the problem currently _only_ appears in the presence of kernel-xen. might someone 'here' take a look there, and chime in if/as appropriate? thanks! -- To unsubscribe, e-mail: opensuse-virtual+unsubscribe(a)opensuse.org For additional commands, e-mail: opensuse-virtual+help(a)opensuse.org

1 0