On Sun, Nov 29, 2009 at 09:44:13PM -0500, Scot P. Floess wrote:
So, as it turns out, my issue seems to be running a CentOS 5.4 x86_64 host and an i386 F11 VM. I used another machine running CentOS 5.4 i386 host and was able to launch, no changes, the F11 i386 VM.
Is there any reason I should see the failure listed below? I've not yet tried to install F11 x86_64 as a VM on the CentOS 5.4 x86_64 host.
Hmm.. iirc F11 GA kernel has some xen-related bugs (at least NX and xsave). Not sure if your crash is one of those though..
Did you try toggling the NX (No eXecute memory protection) setting in BIOS?
There was workarounds for those in post RHEL 5.4 virttest kernels, so could try those aswell..
Some bugzillas that might be related: https://bugzilla.redhat.com/show_bug.cgi?id=502826 https://bugzilla.redhat.com/show_bug.cgi?id=524719 https://bugzilla.redhat.com/show_bug.cgi?id=525290
You could try updated post-5.4 Xen / dom0 kernel rpms from here:
http://people.redhat.com/clalance/virttest/ or from: http://people.redhat.com/dzickus/el5/
-- Pasi
On Wed, 25 Nov 2009, Scot P. Floess wrote:
I just installed F11... When attempting an "xm create -c" I get the following failure... I do use Cobbler and KOAN and have my setup set to use Xen pv... Do I need to specify something special for the guest kernel? I know for CentOS 5.4 guests - I really don't do anything special with xm create...
Using config file "/etc/xen/fedora-11-vm-i386". Started domain fedora-11-vm-i386 BUG: unable to handle kernel NULL pointer dereference at 00000008 IP: [<c048dc74>] get_page_from_freelist+0x291/0x3a2 *pdpt = 00000000382a7007 Thread overran stack, or stack corrupted Oops: 0000 [#1] SMP last sysfs file: Modules linked in:
Pid: 0, comm: swapper Not tainted (2.6.30.9-96.fc11.i686.PAE #1) EIP: e019:[<c048dc74>] EFLAGS: 00010246 CPU: 0 EIP is at get_page_from_freelist+0x291/0x3a2 EAX: f5458000 EBX: 00000001 ECX: 00000001 EDX: 00000000 ESI: c0907ea0 EDI: 00000000 EBP: c0949ea8 ESP: c0949e60 DS: e021 ES: e021 FS: 00d8 GS: 0000 SS: e021 Process swapper (pid: 0, ti=c0948000 task=c08d2314 task.ti=c0948000) Stack: c161e000 00000001 00000020 c0909fa4 c09b46bc 00000000 c09086c0 00000000 00000001 00000000 000212d0 00000000 00000200 c0907ea0 c0907ea0 000012d0 000000d0 c0909fa0 c0949f04 c048de2e c0909fa0 00000001 00000044 c0389fa8 Call Trace: [<c048de2e>] ? __alloc_pages_internal+0xa9/0x38a [<c043607e>] ? vprintk+0x2da/0x2fc [<c04ab646>] ? alloc_pages_node+0x22/0x28 [<c04ab66c>] ? alloc_slab_page+0x20/0x23 [<c04ab7d0>] ? __slab_alloc+0x11e/0x3ef [<c0577b86>] ? kvasprintf+0x26/0x45 [<c04ac631>] ? __kmalloc+0xbd/0x121 [<c0577b86>] ? kvasprintf+0x26/0x45 [<c0577b86>] ? kvasprintf+0x26/0x45 [<c0577bb6>] ? kasprintf+0x11/0x13 [<c096c6b9>] ? kmem_cache_init+0xbb/0x119 [<c09517ae>] ? start_kernel+0x23c/0x2e6 [<c0951081>] ? i386_start_kernel+0x70/0x77 [<c09540c8>] ? xen_start_kernel+0x434/0x43c Code: 00 89 f1 8b 5d d8 2b 8e ec 0a 00 00 64 a1 24 31 9b c0 03 45 c8 c1 f9 08 69 c9 a3 8b 2e ba 01 5c 88 10 8b 45 e8 ff 15 80 b4 8d c0 <8b> 42 08 89 d1 89 45 e8 8b 02 8b 5a 10 66 85 c0 79 03 8b 4a 0c EIP: [<c048dc74>] get_page_from_freelist+0x291/0x3a2 SS:ESP e021:c0949e60 CR2: 0000000000000008 ---[ end trace 4eaa2a86a8e2da22 ]--- Kernel panic - not syncing: Attempted to kill the idle task! Pid: 0, comm: swapper Tainted: G D 2.6.30.9-96.fc11.i686.PAE #1 Call Trace: [<c0738681>] ? printk+0x14/0x1b [<c07385cb>] panic+0x3e/0xe0 [<c04385af>] do_exit+0x62/0x5e5 [<c0738681>] ? printk+0x14/0x1b [<c073b246>] oops_end+0x99/0xa1 [<c04218b4>] no_context+0x141/0x14b [<c04219dd>] __bad_area_nosemaphore+0x11f/0x127 [<c0574a00>] ? put_dec+0x2e/0x104 [<c0574a00>] ? put_dec+0x2e/0x104 [<c048c4a5
On Wed, 25 Nov 2009, Scot P. Floess wrote:
So, I am going to show my ignorance here... Is there anything special I need to do to use those kernels? I could have sworn when I tried either F 10 or F 11 - I couldn't work it without qemu running. But I could be wrong :)
Maybe I should retry 11...
On Wed, 25 Nov 2009, Pasi Kärkkäinen wrote:
On Tue, Nov 24, 2009 at 12:24:06PM -0500, Scot P. Floess wrote:
I have a silly question... I did some goggle'ing around and it appeared to me that Fedora 12 should work as a Xen guest (domU). Is this true? I know 10 and 11 I'd have to use KVM....
Fedora 10, Fedora 11, and Fedora 12 all work as Xen PV domU, with the included (pvops) kernels. Earlier Fedora versions work too.
Fedora 12 also works as Xen dom0, if you install the external unofficial xendom0 kernel rpm.
-- Pasi
On Tue, 24 Nov 2009, Pasi Kärkkäinen wrote:
On Mon, Nov 23, 2009 at 05:51:51PM -0500, Charles J Gruener wrote: >I've only got the one virtual disk. /boot is on xvda1 with / and swap >in LVs. The PV is xvda2. >
Double check the F12 /boot type. It needs to be ext3 for now. RHEL5 / CentOS5 doesn't have ext4 pygrub support yet.
Xen packages on F12 host/dom0 do support ext4 pygrub though.
-- Pasi
>Charles > >On Nov 23, 2009, at 4:43 PM, "Andri Möll" andri@dot.ee wrote: > >>If /boot is on another virtual drive, make sure it's the first one in >>the VM's 'disk' listing (/etc/xen/foo). I remember reading that Xen >>gives only the first disk to pygrub as an arg. >>/usr/bin/pygrub is a Python script. Probably printing the 'file' >>var to >>stderr before the line fsimage.open(file, get_fs_offset(file)) shows >>what it's trying to open. >> >> >>Andri >> >> >>On Mon, 2009-11-23 at 15:59 -0500, Charles J Gruener wrote: >>>I specifically call out to create an ext3 filesystem in the >>>kickstart file for /boot. I did neglect to mention that in my >>>original post. >>> >>>Charles >>> >>>On Nov 23, 2009, at 3:56 PM, Andri Möll wrote: >>> >>>>This looks like your host doesn't support the filesystem your >>>>domU's >>>>using. I think Fedora's on ext4 by default. One solution is to >>>>use >>>>ext2/ext3 for the domU's boot media or partition. >>>> >>>> >>>>Andri >>>_______________________________________________ >>>CentOS-virt mailing list >>>CentOS-virt@centos.org >>>http://lists.centos.org/mailman/listinfo/centos-virt >> >>_______________________________________________ >>CentOS-virt mailing list >>CentOS-virt@centos.org >>http://lists.centos.org/mailman/listinfo/centos-virt >_______________________________________________ >CentOS-virt mailing list >CentOS-virt@centos.org >http://lists.centos.org/mailman/listinfo/centos-virt _______________________________________________ CentOS-virt mailing list CentOS-virt@centos.org http://lists.centos.org/mailman/listinfo/centos-virt
Scot P. Floess 27 Lake Royale Louisburg, NC 27549
252-478-8087 (Home) 919-890-8117 (Work)
Chief Architect JPlate http://sourceforge.net/projects/jplate Chief Architect JavaPIM http://sourceforge.net/projects/javapim
Architect Keros http://sourceforge.net/projects/keros
CentOS-virt mailing list CentOS-virt@centos.org http://lists.centos.org/mailman/listinfo/centos-virt
CentOS-virt mailing list CentOS-virt@centos.org http://lists.centos.org/mailman/listinfo/centos-virt
Scot P. Floess 27 Lake Royale Louisburg, NC 27549
252-478-8087 (Home) 919-890-8117 (Work)
Chief Architect JPlate http://sourceforge.net/projects/jplate Chief Architect JavaPIM http://sourceforge.net/projects/javapim
Architect Keros http://sourceforge.net/projects/keros
Scot P. Floess 27 Lake Royale Louisburg, NC 27549
252-478-8087 (Home) 919-890-8117 (Work)
Chief Architect JPlate http://sourceforge.net/projects/jplate Chief Architect JavaPIM http://sourceforge.net/projects/javapim
Architect Keros http://sourceforge.net/projects/keros
Scot P. Floess 27 Lake Royale Louisburg, NC 27549
252-478-8087 (Home) 919-890-8117 (Work)
Chief Architect JPlate http://sourceforge.net/projects/jplate Chief Architect JavaPIM http://sourceforge.net/projects/javapim
Architect Keros http://sourceforge.net/projects/keros
CentOS-virt mailing list CentOS-virt@centos.org http://lists.centos.org/mailman/listinfo/centos-virt