[Users] 4gb seg fixup - centos5 HN and VE, with Xen

[Henrik Holmboe]

Hello all,

I have a setup which might not be the intended use for Xen flavor of the
kernel provided by OpenVZ. I am running a Centos 5.2 Xen dom0, which in
turn runs a Centos 5.2 domU. There is no problem at all when I boot that
domU using the Centos provided kernel, but when I boot the latest Xen
flavor of OpenVZ kernel, then I get alot messages similar to the below
on the console:

4gb seg fixup, process klogd (pid 4309), cs:ip 73:001d1aa0
4gb seg fixup, process syslogd (pid 4306), cs:ip 73:00276830
printk: 21 messages suppressed.
4gb seg fixup, process automount (pid 4365), cs:ip 73:00cdd36c
printk: 102 messages suppressed.


Both the Xen dom0 and domU are updated as of today.

Xen domO:
[root at dom0 ~]# uname -r
2.6.18-92.1.18.el5xen

Centos domU (not seeing the problem):
[root at domu ~]# uname -r
2.6.18-92.1.18.el5xen

OpenVZ domU (seeing the console error messages):
[root at domu ~]# uname -r
2.6.18-92.1.13.el5.028stab059.6xen


When I search the mail archive and other resources on the internet they
recommend two possible solutions to this problem. However neither of
them seem to solve it for me.

Suggested solution 1:
[root at domu ~]# mv /lib/tls /lib/tls.disabled
[root at domu ~]# reboot

Suggested solution 2:
[root at domu ~]# echo "hwcap 0 nosegneg" > /etc/ld.so.conf.d/nosegneg.conf
[root at domu ~]# reboot

Notice that both solutions have been tried on the domU and not dom0.
Can't remember if I tried them on dom0 as well, but probably not since
it doesn't seem likely that it could affect the domU.

There is the file
/etc/ld.so.conf.d/kernelcap-2.6.18-92.1.13.el5.028stab059.6.conf that
contains the same as solution 2, so it should already be in effect.


Have anyone of you tested this setup? I don't know where to go from here
so I'm hoping for some input from you guys.

Thanks,
Henrik

-- 
Henrik Holmboe, Stockholm, Sweden
<http://henrik.holmboe.se/contact/>