[Bug 350051] New: -xenpae kernels seem to be unusable when running on more than one VCPUs

newer
[Bug 206838] New: slideshow starts...

bugzilla_noreply＠novell.com

20 Dec 2007 20 Dec '07

16:33

https://bugzilla.novell.com/show_bug.cgi?id=350051 Summary: -xenpae kernels seem to be unusable when running on more than one VCPUs Product: openSUSE 10.3 Version: Final Platform: PC OS/Version: openSUSE 10.3 Status: NEW Severity: Critical Priority: P5 - None Component: Xen AssignedTo: cgriffin@novell.com ReportedBy: nice@titanic.nyme.hu QAContact: qa@suse.de Found By: Other As I mentioned in https://bugzilla.novell.com/show_bug.cgi?id=343181#c19 When I run a Linux system on a -xenpae kernel with more VCPUs either as dom0 or domU, the system will be HIGHLY unstable. I tried it with both the i386's /boot/xen-pae.gz and x86_64's /boot/xen.gz, and the results are the same. Kernel compiling on such a system is equal to suicide for example. First some segmentation fault occurs, then the kernel space's CPU load goes very high, and, at last, the system becomes totally unresponsive. I will attach a typical screenshot (which is combined of more screenshots) of our production mailserver's (SuSE Linux 8.2(!)) tty10 when running paravirtualized under the mentioned corcumstances as a domU. It crashed 4-5 times a day, when running with more VCPUs with a -xenpae kernel. Now it has run for some days without any problem with one VCPU. Running this 32 bit system on a 64bit kernel (either with one or more VCPUs) is OK as well. 64 bit kernels seem to be unaffected by this problem. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

Show replies by date

bugzilla_noreply＠novell.com

20 Dec 20 Dec

16:37

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c1 --- Comment #1 from Tamás Németh 2007-12-20 09:37:34 MST --- Created an attachment (id=188422) --> (https://bugzilla.novell.com/attachment.cgi?id=188422) Crash of domU -xenpae with multiple VCPUS With every crash I (a layman) observed, the "last sysfs file" was /block/xvda1/dev, and the "call trace" always contained something about ext3. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

7 Jan 7 Jan

16:12

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jbeulich@novell.com AssignedTo|cgriffin@novell.com |jbeulich@novell.com Status|NEW |ASSIGNED -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

16:15

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c2 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |NEEDINFO Info Provider| |nice@titanic.nyme.hu --- Comment #2 from Jan Beulich 2008-01-07 09:15:24 MST --- Would mind first clarifying you original statement? SuSE Linux 8.2 can't run para-virtualized (unless you manually swapped the kernel and probably did various other adjustments), so there must be something wrong with your description. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

16:22

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c3 --- Comment #3 from Tamás Németh 2008-01-07 09:22:04 MST --- Of course I swapped the kernel, changed /etc/fstab in order to be able to use xvda drives and so on. I tried 32bit -xen, 32bit -xenpae and 64bit -xen kernels and as I mentioned 64 bit kernels seem to be the best solution so far (Despite https://bugzilla.novell.com/show_bug.cgi?id=349895). -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

16:44

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c4 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |RESOLVED Info Provider|nice@titanic.nyme.hu | Resolution| |INVALID --- Comment #4 from Jan Beulich 2008-01-07 09:44:59 MST --- I'm sorry, but I don't think we can assist you with this entirely unsupported configuration. If you upgraded the DomU to 10.3 ... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

8 Jan 8 Jan

17:45

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c5 Tamás Németh changed: What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |REOPENED Resolution|INVALID | --- Comment #5 from Tamás Németh 2008-01-08 10:45:46 MST --- :) I don't want you to assist me since running the system on a 64 bit kernel seems to be more than a workaround for this problem. Instead, I want to help at improving openSUSE and thus the commercial Novell products. I installed a paravirtualized openSUSE 10.3 on my laptop. So now I have 10.3 both on dom0 and domU. Xeanpae kernels are also present and able to work (I use them right now). This bug seems to be related to https://bugzilla.novell.com/show_bug.cgi?id=343181. Please take a look at comments 19 and 20. What can I help you now? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

9 Jan 9 Jan

10:29

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c6 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |NEEDINFO Info Provider| |nice@titanic.nyme.hu --- Comment #6 from Jan Beulich 2008-01-09 03:29:34 MST --- That bug is about fully virtualized guests, not paravirtualized ones. Also, without providing (detailed) information of what instabilities you observe now that you're using 10.3 only. I can assure you that I regularly build kernels while running a Xen kernel, and it's been a long time since this was anything dangerous. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

11:02

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c7 --- Comment #7 from Tamás Németh 2008-01-09 04:02:36 MST --- Was your machine running the 32bit -xenpae kernel shipped with openSUSE 10.3 (2.6.22.13-0.3-xenpae) on multiple VCPUS while compiling the kernel? I tested this kernel on different processor generations and it proved to be extremely unstable for me when running on multiple VCPUS. Yesterday I compiled this kernel with a configuration being slightly different from the SuSE defaults (no kernel preemption, tick frequency=100Hz), then booted the compiled kernel with two VCPUS. I started to compile a kernel on that system but did not succed. It crashed. I'm afraid, only 64 bit Xen kernels of openSUSE 10.3 are usable on the Intel platform (except that I have to made some workarounds for https://bugzilla.novell.com/show_bug.cgi?id=344877 and https://bugzilla.novell.com/show_bug.cgi?id=349892) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

11:20

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c8 --- Comment #8 from Tamás Németh 2008-01-09 04:20:42 MST ---

...

That bug is about fully virtualized guests, not paravirtualized ones.

I would say that the bug is about the behavior of dom0 and paravirtualized guest in the presence of fully virtualized guests, instead. I don't have too much time, but I will try to provide some detailed informations about the instabilities. What would be useful for you? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

14:25

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c9 --- Comment #9 from Jan Beulich 2008-01-09 07:25:55 MST --- What you're describing sounds more like problems with the originally shipped kernel, which indeed was severly broken. What would help are - reliable (and preferrably not too complex) ways to reproduce your issues - logs (Xen, kernel) showing anomalies - crash data (register+stack traces) - any other technical information you have. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

15:28

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c10 --- Comment #10 from Tamás Németh 2008-01-09 08:28:20 MST --- OK. The first thing I can say is that it's extremely easy to reproduce this issue (at least on the computers near me): -Configure an openSUSE 10.3 system (either dom0 or domU) with a 2.6.22.13-0.3-xenpae kernel (this is the newest one for 10.3 today). -Configure that openSUSE 10.3 (or any other Linux, since the kernel itself is broken) to run on MORE THAN ONE virtual cpu. -Wait a few hours, or if you want a quick "success", start a kernel compiling. -When the kernel compiling suddenly ends, take a look at /dev/tty10, or exatract the needed information in some other way quickly, because the system will become totally unresponsive in a few seconds. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

16:39

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c11 --- Comment #11 from Jan Beulich 2008-01-09 09:39:51 MST ---

...

-When the kernel compiling suddenly ends, take a look at /dev/tty10, or exatract the needed information in some other way quickly, because the system will become totally unresponsive in a few seconds.

Could you get us this information (even if similar to #1, I would still want it from a pure 10.3 setup)? Also, you seem to imply that that's happening with Dom0, too: In that case, I can't see how /block/xvda1/dev would come into the picture unless you also have DomU-s running. Also, are you sure you're not running into the time problem we're aware of and have a fix queued (bug 335121)? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

20:38

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c12 --- Comment #12 from Tamás Németh 2008-01-09 13:38:22 MST ---

...

What you're describing sounds more like problems with the originally shipped kernel, which indeed was severly broken. It happens on the newest 10.3 kernels, too.

...

Also, are you sure you're not running into the time problem we're aware of and have a fix queued (bug 335121)?

Oh, that bug seems to be the same as https://bugzilla.novell.com/show_bug.cgi?id=344877 don't you think so? Unfortunately I'm familiar with that issue but I think, that differs from this one. Seting the clocksource to jiffies was a good workaround for me, but currently I'm testing a 100Hz kernel instead. I will investigate if that bug is identical to this problem.

...

Also, you seem to imply that that's happening with Dom0, too: In that case, I can't see how /block/xvda1/dev would come into the picture unless you also have DomU-s running.

Yes, it definitely happens in in dom0 too, just like bug 335121. I just wrote about xvda because I could observe the tty10 logs of only a domU syses (that infamous SuSE Linux 8.2).

...

Could you get us this information (even if similar to #1, I would still want it from a pure 10.3 setup)?

...

- logs (Xen, kernel) showing anomalies - crash data (register+stack traces) - any other technical information you have.

Where can I get those kernel and Xen logs and crash data from? What do you mean exactly? (dmesg, xm dmesg, tty10, /var/log/messages or /var/log/xen/*) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

10 Jan 10 Jan

08:00

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c13 --- Comment #13 from Jan Beulich 2008-01-10 01:00:50 MST ---

...

Oh, that bug seems to be the same as https://bugzilla.novell.com/show_bug.cgi?id=344877 don't you think so?

Yes, as I implied in that bug's comment #8.

...

Where can I get those kernel and Xen logs and crash data from? What do you mean exactly? (dmesg, xm dmesg, tty10, /var/log/messages or /var/log/xen/*)

dmesg (or equivalent collected via serial) and xm dmesg (likewise using serial if necessary). -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

09:17

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c14 --- Comment #14 from Tamás Németh 2008-01-10 02:17:06 MST --- Hi Jan! Good news! Yesterday my dom0 crashed several times, but always became unresposive so quickly, that I was unable to extract any information. Today, I wanted to crash the 10.3 domU, but had insufficient disk space to install the kernel sources. While yast was installing something on domU, some applications in dom0 freezed in dom0, but the system remained somewhat responsive. So I was able to get the xm dmesg and dmesg info. KDE on tty7 freezed, but I was able to log in on a character terminal. The system time was correct. Dmesg complained about some ext3 thing again. I will the output of dmesg, xm dmesg and /var/log/messages, which is filled with some binary(?) garbage (search for that at "Jan 10 09:35:57"). Some information about my filesystems: milleniumfalcon:/home/tamas/bugreports # mount /dev/sda5 on / type ext3 (rw,acl,user_xattr,data=writeback) proc on /proc type proc (rw) sysfs on /sys type sysfs (rw) debugfs on /sys/kernel/debug type debugfs (rw) udev on /dev type tmpfs (rw) devpts on /dev/pts type devpts (rw,mode=0620,gid=5) /dev/sda1 on /boot type ext2 (rw,acl,user_xattr) /dev/sda8 on /home type ext3 (rw,acl,user_xattr,data=writeback) /dev/sda6 on /usr type ext3 (rw,acl,user_xattr,data=writeback) /dev/sda7 on /var type ext3 (rw,acl,user_xattr,data=writeback) securityfs on /sys/kernel/security type securityfs (rw) milleniumfalcon:/home/tamas/bugreports # tune2fs -l /dev/sda5 tune2fs 1.40.2 (12-Jul-2007) Filesystem volume name: <none> Last mounted on: <not available> Filesystem UUID: 44ffd78c-1df0-41a7-8726-c9a284267286 Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery sparse_super large_file Filesystem flags: signed directory hash Default mount options: (none) Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 263296 Block count: 526120 Reserved block count: 26306 Free blocks: 175840 Free inodes: 222125 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 128 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 15488 Inode blocks per group: 484 Filesystem created: Fri Oct 5 13:30:15 2007 Last mount time: Thu Jan 10 09:50:57 2008 Last write time: Thu Jan 10 09:50:57 2008 Mount count: 116 Maximum mount count: 500 Last checked: Wed Dec 5 07:37:54 2007 Check interval: 5184000 (2 months) Next check after: Sun Feb 3 07:37:54 2008 Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 128 Journal inode: 8 First orphan inode: 78246 Default directory hash: tea Directory Hash Seed: 8eeee4bb-e7af-40c2-943a-8c1c29430534 Journal backup: inode blocks milleniumfalcon:/home/tamas/bugreports # tune2fs -l /dev/sda1 tune2fs 1.40.2 (12-Jul-2007) Filesystem volume name: <none> Last mounted on: <not available> Filesystem UUID: 83959e06-2ac5-4713-b570-0f1ba75fe354 Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: resize_inode dir_index filetype sparse_super Filesystem flags: signed directory hash Default mount options: (none) Filesystem state: not clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 14056 Block count: 56196 Reserved block count: 2809 Free blocks: 9325 Free inodes: 13970 First block: 1 Block size: 1024 Fragment size: 1024 Reserved GDT blocks: 219 Blocks per group: 8192 Fragments per group: 8192 Inodes per group: 2008 Inode blocks per group: 251 Filesystem created: Fri Oct 5 13:30:14 2007 Last mount time: Thu Jan 10 09:51:15 2008 Last write time: Thu Jan 10 09:51:15 2008 Mount count: 1 Maximum mount count: 21 Last checked: Thu Jan 10 09:51:03 2008 Check interval: 15552000 (6 months) Next check after: Tue Jul 8 10:51:03 2008 Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 128 Default directory hash: tea Directory Hash Seed: f8560e1e-787a-4b6f-8dd5-74f04a01b7f0 milleniumfalcon:/home/tamas/bugreports # tune2fs -l /dev/sda8 tune2fs 1.40.2 (12-Jul-2007) Filesystem volume name: <none> Last mounted on: <not available> Filesystem UUID: 437a38a1-5f94-4eb9-82f5-c177ba4f1bc9 Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery sparse_super large_file Filesystem flags: signed directory hash Default mount options: (none) Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 7847936 Block count: 15687464 Reserved block count: 784373 Free blocks: 4003680 Free inodes: 7659788 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 1020 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 16384 Inode blocks per group: 512 Filesystem created: Fri Oct 5 13:30:32 2007 Last mount time: Thu Jan 10 09:51:15 2008 Last write time: Thu Jan 10 09:51:15 2008 Mount count: 110 Maximum mount count: 500 Last checked: Wed Dec 5 09:45:15 2007 Check interval: 5184000 (2 months) Next check after: Sun Feb 3 09:45:15 2008 Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 128 Journal inode: 8 Default directory hash: tea Directory Hash Seed: b331c34e-cf98-4c58-9ae7-92a12dd5f790 Journal backup: inode blocks milleniumfalcon:/home/tamas/bugreports # tune2fs -l /dev/sda6 tune2fs 1.40.2 (12-Jul-2007) Filesystem volume name: <none> Last mounted on: <not available> Filesystem UUID: b4682c9f-3245-43dc-96d0-a149838b3a9f Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal resize_inode dir_index filetype needs_recovery sparse_super large_file Filesystem flags: signed directory hash Default mount options: (none) Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 656000 Block count: 1311297 Reserved block count: 65564 Free blocks: 161543 Free inodes: 449254 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 320 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 16000 Inode blocks per group: 500 Filesystem created: Fri Oct 5 13:30:19 2007 Last mount time: Thu Jan 10 09:51:15 2008 Last write time: Thu Jan 10 09:51:15 2008 Mount count: 110 Maximum mount count: 500 Last checked: Wed Dec 5 09:55:27 2007 Check interval: 5184000 (2 months) Next check after: Sun Feb 3 09:55:27 2008 Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 128 Journal inode: 8 Default directory hash: tea Directory Hash Seed: c17a9802-ffc0-486a-b362-1e1e1a490ac7 Journal backup: inode blocks milleniumfalcon:/home/tamas/bugreports # tune2fs -l /dev/sda7 tune2fs 1.40.2 (12-Jul-2007) Filesystem volume name: <none> Last mounted on: <not available> Filesystem UUID: 80618a37-629d-46cb-b2b9-2fd2d3e211de Filesystem magic number: 0xEF53 Filesystem revision #: 1 (dynamic) Filesystem features: has_journal ext_attr resize_inode dir_index filetype needs_recovery sparse_super large_file Filesystem flags: signed directory hash Default mount options: (none) Filesystem state: clean Errors behavior: Continue Filesystem OS type: Linux Inode count: 393600 Block count: 787177 Reserved block count: 39358 Free blocks: 642439 Free inodes: 387429 First block: 0 Block size: 4096 Fragment size: 4096 Reserved GDT blocks: 192 Blocks per group: 32768 Fragments per group: 32768 Inodes per group: 15744 Inode blocks per group: 492 Filesystem created: Fri Oct 5 13:30:27 2007 Last mount time: Thu Jan 10 09:51:15 2008 Last write time: Thu Jan 10 09:51:15 2008 Mount count: 110 Maximum mount count: 500 Last checked: Wed Dec 5 09:56:48 2007 Check interval: 5184000 (2 months) Next check after: Sun Feb 3 09:56:48 2008 Reserved blocks uid: 0 (user root) Reserved blocks gid: 0 (group root) First inode: 11 Inode size: 128 Journal inode: 8 First orphan inode: 189300 Default directory hash: tea Directory Hash Seed: b4285e27-a21f-4a41-a0f1-b3491c60af3e Journal backup: inode blocks milleniumfalcon:/home/tamas/bugreports # -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

09:21

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c15 --- Comment #15 from Tamás Németh 2008-01-10 02:21:23 MST --- I've forgotten to mention, that this time both dom0 and domU ran on one CPU. Both of them were started on two VCPUs, but later I reduced the CPU number to one by 'xm vcpu-set DOMAIN 1' in both systems runtime. I also assigned the VCPUs to phisical CPUs by 'xm vcpu-pin'. My notebook has two CPU cores, so dom0 ran on CPU 0, domU ran on CPU 1. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

09:22

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c16 --- Comment #16 from Tamás Németh 2008-01-10 02:22:44 MST --- Created an attachment (id=190026) --> (https://bugzilla.novell.com/attachment.cgi?id=190026) dmesg output from crashed dom0 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

09:23

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c17 --- Comment #17 from Tamás Németh 2008-01-10 02:23:38 MST --- Created an attachment (id=190027) --> (https://bugzilla.novell.com/attachment.cgi?id=190027) xm dmesg output from crashed dom0 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

09:25

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User nice@titanic.nyme.hu added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c18 --- Comment #18 from Tamás Németh 2008-01-10 02:25:02 MST --- Created an attachment (id=190028) --> (https://bugzilla.novell.com/attachment.cgi?id=190028) Garbaged /var/log/messages file from crashed dom0 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

11 Jan 11 Jan

10:44

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #190026|application/octet-stream |text/plain mime type| | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

10:45

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 Jan Beulich changed: What |Removed |Added ---------------------------------------------------------------------------- Attachment #190027|application/octet-stream |text/plain mime type| | -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

11:06

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c19 --- Comment #19 from Jan Beulich 2008-01-11 04:06:28 MST --- A couple of notes after looking at these: - You said Dom0 alone is sufficient to reproduce the problem, so please don't start any DomU (to reduce the number of variables). - Please pass "loglvl=all guest_loglvl=all" as additional parameters (without the quotes) to Xen. - Please exclude at least DRM from the picture (to be absolutely sure, perhaps also the AGP module, despite that having been reviewed for correctness); for simplification purposes, it might be worth excluding other modules, too: pcmcia, ipv6, netfilter, wireless, watchdog, ... -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

31 Jan 31 Jan

10:40

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jim.pye@pyenet.co.nz added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c20 Jim Pye changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |jim.pye@pyenet.co.nz --- Comment #20 from Jim Pye 2008-01-31 03:40:22 MST --- Jan I have been reporting against bug#344877 but this one seems closer to what I am seeing. I have just updated the kernel to version: Linux vmhoster 2.6.22.16-0.1-xenpae #1 SMP 2008/01/23 14:28:52 UTC i686 i686 i386 GNU/Linux from the update site on Dom0 and all DomUs - all running openSUSE 10.3. Even though this version seems more stable when I run a large file copy test in Dom0, that was failing on earlier kernels, it still does not run very well with multiple domUs started. As described above the machine will suddenly go to high util before becoming completely unresponsive. Hard reset (via on/off switch) is needed. Machine is 32bit with 4GB of mem and Dual Core2 CPU (4 CPUs) Standard log files did not report any issues before crash. Where exactly do we need to put the loglvl parameters? Jim -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

13:46

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c21 --- Comment #21 from Jan Beulich 2008-01-31 06:46:06 MST --- The loglvl=all (and likewise guest_loglvl=all) parameter goes on the Xen command line in /boot/grub/menu.lst. If with standard log files you also mean anything Xen prints out over the serial console, then I'd like you to first see whether SysRq-t generates any output at the point of the hang (requires that you can see the system console and that you made sure that SysRq is enabled at all) and second whether sending 'd' over the serial console to Xen provides any output (requires that you switch input to Xen prior to trying). Any output should be attached here. Also, I'd want to see your loaded module list. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

3 Feb 3 Feb

22:06

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jim.pye@pyenet.co.nz added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c22 --- Comment #22 from Jim Pye 2008-02-03 15:06:22 MST --- Created an attachment (id=192963) --> (https://bugzilla.novell.com/attachment.cgi?id=192963) Log files etc. as requested by Jan Within the gzip file the notes.txt file explains the rest of the files included. Please let me know if you require more information. Jim -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

5 Feb 5 Feb

10:13

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User jbeulich@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c23 --- Comment #23 from Jan Beulich 2008-02-05 03:13:48 MST --- Jim, I'm afraid without proper (complete) backtraces and register/stack dumps (or complete SysRq-t output in those cases where the system silently locks up) it's going to be very difficult to get to the root of your problem. Also, I'm not convinced your issue is really the same as the one being discussed here. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

bugzilla_noreply＠novell.com

25 Apr 25 Apr

12:51

New subject: [Bug 350051] -xenpae kernels seem to be unusable when running on more than one VCPUs

https://bugzilla.novell.com/show_bug.cgi?id=350051 User cthiel@novell.com added comment https://bugzilla.novell.com/show_bug.cgi?id=350051#c24 Christoph Thiel changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |RESOLVED Info Provider|nice@titanic.nyme.hu | Resolution| |NORESPONSE --- Comment #24 from Christoph Thiel 2008-04-25 06:51:10 MST --- Closing NOREPSONSE, due to missing information for more than 21 days. Please feel free to reopen and provide the requested information. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

5851

Age (days ago)

5978

Last active (days ago)

List overview

Download

27 comments

1 participants

participants (1)

bugzilla_noreply＠novell.com

[Bug 350051] New: -xenpae kernels seem to be unusable when running on more than one VCPUs

tags

participants (1)