ART Kernel BUG?

BerArt · Unread post by **BerArt** » Mon Feb 04, 2008 6:44 am

I get this messages every few seconds in my ART kernel log:

BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
in_atomic():1, irqs_disabled():0
[<c146856c>] [<c14656af>] [<c1422362>] [<c142236e>] [<c146685c>] [<c1420c43>] [<c1421084>] [<c1468cc5>] [<c1420365>] [<c1420363>] [<c1420c43>] [<c161b8e5>] [<c1460000>] [<c1610000>] =======================
BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
in_atomic():1, irqs_disabled():0
[<c146856c>] [<c146685c>] [<c1468cc5>] [<c1420c43>] [<c1421084>] [<c146a988>] [<c1420c43>] [<c161b8e5>] =======================
BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
in_atomic():1, irqs_disabled():0
[<c146856c>] [<c15ae75e>] [<c143c901>] [<c1420c43>] [<c1421084>] [<c147b9a6>] [<c1420c43>] [<c161b8e5>] =======================
BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
in_atomic():1, irqs_disabled():0
[<c146856c>] [<c1468127>] [<c1420c43>] [<c1421084>] [<c15ae6b5>] [<c1420c43>] [<c161b8e5>] [<c15b0000>] =======================

Does someone now how to solve this please?

Thx!

Unread post by **scott** » Mon Feb 04, 2008 9:34 am

Probably acpi related (power saving mode), you can try turning that off by setting acpi=off in the kernel boot parameters.

BerArt · Unread post by **BerArt** » Mon Feb 04, 2008 10:00 am

I think it is off, or where can I find this parameters?

Unread post by **scott** » Mon Feb 04, 2008 10:21 am

you turn it off in grub.conf, on the kernel line. Add this to the end:

acpi=off

BerArt · Unread post by **BerArt** » Mon Feb 04, 2008 10:37 am

Thx!

I supose I have to reboot?

BerArt · Unread post by **BerArt** » Mon Feb 04, 2008 10:47 am

Just to make sure: (I have to add the RED part)

OLD Line:
kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00
selinux=0

NEW Line:
kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00
selinux=0 rhgb quiet acpi=off

Reboot, problem solved?

BerArt · Unread post by **BerArt** » Tue Feb 05, 2008 9:16 am

Can you please verify before I reboot the server

Thx a lot!

Unread post by **scott** » Tue Feb 05, 2008 9:40 am

yeah, just the acpi=off part. You dont need the other two

BerArt · Unread post by **BerArt** » Tue Feb 05, 2008 9:42 am

Thx!

BerArt · Unread post by **BerArt** » Tue Feb 05, 2008 9:51 am

After the reboot, I recieved the same error:

The line in the grub.conf is now:
kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00 acpi=off

in message log:
Feb 5 14:49:29 plesk2 kernel: BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
Feb 5 14:49:29 plesk2 kernel: in_atomic():1, irqs_disabled():0
Feb 5 14:49:29 plesk2 kernel: [<c146856c>] [<c146ae3d>] [<c1420c43>] [<c1421084>] [<c14430ba>] [<c1420c43>] [<c161b8e5>] =======================

In Kernel log:
BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
in_atomic():1, irqs_disabled():0
[<c146856c>] [<c146ae3d>] [<c1420c43>] [<c1421084>] [<c14430ba>] [<c1420c43>] [<c161b8e5>] =======================

Any other suggestions?

zooming · Unread post by **zooming** » Thu Feb 14, 2008 1:10 pm

Hello,

I'm having the same problem.

I have just installed ASL a week ago in a RHEL5 ES server, with Plesk 8.2.

My server is freezing twice a day.

First apache stops, and if I don't reboot the server, everything stops.

I have several entries like this in the /var/log/messages file:

Feb 14 10:58:01 rhel5 kernel: BUG: sleeping function called from invalid context at include/linux/pagemap.h:165
Feb 14 10:58:01 rhel5 kernel: in_atomic():1, irqs_disabled():0
Feb 14 10:58:01 rhel5 kernel: [<c146856c>] [<c146ae3d>] [<c1420c43>] [<c1421084>] [<c1455ea6>] [<c1420c43>] [<c161b8e5>] =======================

I have tried the noacpi and acpi=off grub parameters, but it didn't work.

If I try to use the default RHEL5 ES kernel, I end up with lots of apache processes consuming all CPU (it looks like ASL doesn't work well without ASL's kernel)

After an apache crash, I can find some error logs in the /var/log/messages file:

Feb 14 10:39:38 rhel5 kernel: grsec: From 189.12.230.249: failed fork with errno -12 by /usr/sbin/httpd[httpd:30666] uid/euid:48/48 gid/egid:48/48, parent /usr/sbin/httpd[httpd:9241] uid/euid:0/0 gid/egid:0/0
Feb 14 10:39:56 rhel5 kernel: grsec: From 189.12.230.249: failed fork with errno -12 by /usr/sbin/httpd[httpd:26633] uid/euid:48/48 gid/egid:48/48, parent /usr/sbin/httpd[httpd:9241] uid/euid:0/0 gid/egid:0/0

This is a critical situation, I'll have to remove ASL soon If I didn't fix this problem. The server Uptime is very slow this month.

Thank you.
Alexandre

BerArt · Unread post by **BerArt** » Thu Feb 14, 2008 1:14 pm

The line in the grub.conf is now:
kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00 acpi=off

(then reboot)

Worked for me, the messages is only comming back one time right after the reboot...

guga-nyc · Unread post by **guga-nyc** » Sat Feb 16, 2008 6:21 pm

If that doesnt work, try it noapic instead.

BerArt · Unread post by **BerArt** » Sun Feb 17, 2008 3:41 am

You mean:

kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00 noapic

instead of

kernel /vmlinuz-2.6.23.1-3.art ro root=/dev/VolGroup00/LogVol00 acpi=off

Is it possible this make a difference?

zooming · Unread post by **zooming** » Sun Feb 17, 2008 9:08 am

Hello,

Neither option worked, the kernel bug messages are still there.

I'm confused, are these two parameters related to the same matter?

http://www.linuxquestions.org/questions ... rs-454675/

Despite of these messages, I'm not sure they are the source of the problem. It seems to be related to apache.

Using the server-status page I can see several pages in the "W" (Sending Reply) state for a long time (four hours).

It looks like these requestes stays there until the server crash.

Thank you.

Alexandre

Atomicorp

ART Kernel BUG?

ART Kernel BUG?

Another option