Colin, it is your call whether the spurious soft lockup warnings should
be fixed or not. Personally, I am fine knowing they are harmless.
Thanks a lot!
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you
Marking as Won't Fix as the error messages are soft lock up messages
are coming from a misbehaving user space app and the watchdog is
detecting this as a *potential* soft lockup issue.
** Changed in: linux (Ubuntu)
Status: Incomplete = Won't Fix
--
kernel 2.6.24 Soft lockup under load on
@Etienne, any change on this bug? Should we mark is ask Won't Fix now?
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs
The bug have been marked Incomplete, but I am not sure what needs to be
provided. A better test case?
AFAIC, I am entirely satisfied with Colin's explanation. I do think the
spurious soft lockup message should be fixed, but it is not mine to
decide.
--
kernel 2.6.24 Soft lockup under load on
** Changed in: linux (Ubuntu)
Sourcepackagename: linux-meta = linux
Importance: Undecided = High
Status: New = Incomplete
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of
Etienne,
In trying to reproduce an OOM and loading issue using the stress test
application,
one needs to consider how the OOM occurs and if can be fixed or helped by
tweaking the virtual memory settings.
Some suggested tweaks to try are as follows:
as root:
echo 2
Let's look at these error messages. The following:
[ 554.116802] BUG: soft lockup - CPU#1 stuck for 11s! [stress:4618]
[ 557.606750] CPU 1:
[ 557.607150] Modules linked in: iptable_filter ip_tables x_tables ipv6 lp
loop af_packet serio_raw intel_agp container evdev button shpchp parport_pc
One other point to make is that the message BUG: soft lockup - CPU#1
stuck for 11s! [stress:4618] is rather alarmist - it makes one think the
CPU is not being used and is stuck. In fact, if one uses the top command
one can see that the CPU is 100% busy running the stress process and not
stuck at
Once we can reproduce this, I will see if upstream commit
126e01bf92dfc5f0ba91e88be02c473e1506d7d9 will help fix this.
Quoting the patch:
softlockup: fix NOHZ wakeup
David Miller reported:
|---
the following commit:
| commit
It appears that this kind of soft lockup kernel message easily
reproduced with the following code on a 2.6.24-xxx hardy server kernel
inside vmware server:
#include stdio.h
#include stdlib.h
#include unistd.h
#include sched.h
void setsched(void)
{
struct sched_param mysched;
Incidentally, how long does it take before the problem occurs? Orders of
minutes, hours, days?
Colin
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed
Generally within a matter of minutes, but sometime it would take longer.
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
ubuntu-bugs
** Changed in: linux-meta (Ubuntu)
Assignee: (unassigned) = Colin King (colin-king)
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
I left the stress test running again on six VM overnight, and only one
got the soft lockup. Attached is the dmesg output. I could probably
provide more of these by re-running the stress test on a bunch of
machine, but I think it might be redundant.
** Attachment added: dmesg-1bis.txt
** Attachment added: dmesg-1.txt
http://launchpadlibrarian.net/21110399/dmesg-1.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
** Attachment added: dmesg-2.txt
http://launchpadlibrarian.net/21110401/dmesg-2.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
--
** Attachment added: cpuinfo-1.txt
http://launchpadlibrarian.net/21110403/cpuinfo-1.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Attachment added: cpuinfo-2.txt
http://launchpadlibrarian.net/21110412/cpuinfo-2.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
** Attachment added: lspci-vvnn-1.txt
http://launchpadlibrarian.net/21110414/lspci-vvnn-1.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
** Attachment added: lspci-vvnn-2.txt
http://launchpadlibrarian.net/21110416/lspci-vvnn-2.txt
--
kernel 2.6.24 Soft lockup under load on VMWare ESX 3.5
https://bugs.launchpad.net/bugs/316187
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to
Notice that I have added the kernel parameter clocksource=acpi_pm, as
discussed in bug #261937 and at:
http://kb.vmware.com/selfservice/microsites/search.do?language=en_UScmd=displayKCexternalId=1007020
However, dmesg on both VM that crashed still show an error such as:
Clocksource tsc
21 matches
Mail list logo