Hi,

I'm running a Lustre client on a openvz kernel. After enabling
ve_allow_kthreads I get the warnings below which comes from
__set_special_pid (code pasted below). But what is openvz really warning
about here? Is it an old debug message ? Should I care about it ? 

Thanks
/Jakob



BUG: warning at kernel/exit.c:333/__set_special_pids()

Call Trace:
 [<ffffffff8020af1b>] dump_stack+0x12/0x17
 [<ffffffff8023076e>] __set_special_pids+0xab/0x127
 [<ffffffff8023080c>] set_special_pids+0x22/0x33
 [<ffffffff80232607>] daemonize+0x86/0x2e8
 [<ffffffff88087ae3>] :libcfs:cfs_daemonize+0x13/0x70
 [<ffffffff883114ec>] :lustre:ll_statahead_thread+0x19c/0x17f0
 [<ffffffff8020a5e8>] child_rip+0xa/0x12
DWARF2 unwinder stuck at child_rip+0xa/0x12
Leftover inexact backtrace:
 [<ffffffff88311350>] :lustre:ll_statahead_thread+0x0/0x17f0
 [<ffffffff8020a5de>] child_rip+0x0/0x12




void __set_special_pids(pid_t session, pid_t pgrp)
{
        struct task_struct *curr = current->group_leader;

        WARN_ON(is_virtual_pid(pgrp));
        WARN_ON(is_virtual_pid(session));

        if (curr->signal->session != session) {
                detach_pid(curr, PIDTYPE_SID);
                curr->signal->session = session;
                attach_pid(curr, PIDTYPE_SID, session);
        }
        if (process_group(curr) != pgrp) {
                detach_pid(curr, PIDTYPE_PGID);
                curr->signal->pgrp = pgrp;
                attach_pid(curr, PIDTYPE_PGID, pgrp);
        }
}

AFAIR it can be ignored.
it warn's that some kernel thread with virtual pid (i.e. from inside VE)
changes it's session/pgrp.

BTW, In 2.6.18 (what kernel are you running?) these lines are changed to:
         WARN_ON(__is_virtual_pid(pgrp));
         WARN_ON(__is_virtual_pid(session));

Thanks,
Kirill


Jakob Goldbach wrote:
> Hi,
> 
> I'm running a Lustre client on a openvz kernel. After enabling
> ve_allow_kthreads I get the warnings below which comes from
> __set_special_pid (code pasted below). But what is openvz really warning
> about here? Is it an old debug message ? Should I care about it ? 
> 
> Thanks
> /Jakob
> 
> 
> 
> BUG: warning at kernel/exit.c:333/__set_special_pids()
> 
> Call Trace:
>  [<ffffffff8020af1b>] dump_stack+0x12/0x17
>  [<ffffffff8023076e>] __set_special_pids+0xab/0x127
>  [<ffffffff8023080c>] set_special_pids+0x22/0x33
>  [<ffffffff80232607>] daemonize+0x86/0x2e8
>  [<ffffffff88087ae3>] :libcfs:cfs_daemonize+0x13/0x70
>  [<ffffffff883114ec>] :lustre:ll_statahead_thread+0x19c/0x17f0
>  [<ffffffff8020a5e8>] child_rip+0xa/0x12
> DWARF2 unwinder stuck at child_rip+0xa/0x12
> Leftover inexact backtrace:
>  [<ffffffff88311350>] :lustre:ll_statahead_thread+0x0/0x17f0
>  [<ffffffff8020a5de>] child_rip+0x0/0x12
> 
> 
> 
> 
> void __set_special_pids(pid_t session, pid_t pgrp)
> {
>         struct task_struct *curr = current->group_leader;
> 
>         WARN_ON(is_virtual_pid(pgrp));
>         WARN_ON(is_virtual_pid(session));
> 
>         if (curr->signal->session != session) {
>                 detach_pid(curr, PIDTYPE_SID);
>                 curr->signal->session = session;
>                 attach_pid(curr, PIDTYPE_SID, session);
>         }
>         if (process_group(curr) != pgrp) {
>                 detach_pid(curr, PIDTYPE_PGID);
>                 curr->signal->pgrp = pgrp;
>                 attach_pid(curr, PIDTYPE_PGID, pgrp);
>         }
> }
> 
> 
> 
> 
> 
> 
> 
> 
> 
>

On Wed, 2008-01-09 at 11:55 +0300, Kirill Korotaev wrote:
> AFAIR it can be ignored.

Great. 


> BTW, In 2.6.18 (what kernel are you running?) these lines are changed to:
>          WARN_ON(__is_virtual_pid(pgrp));
>          WARN_ON(__is_virtual_pid(session));
> 

I am using 2.6.18 from Debian. They also distribute
patch-026test015-combined.gz in kernel-patch-openvz.deb 
It contains:

@@ -263,6 +276,9 @@ void __set_special_pids(pid_t session, p
 {
        struct task_struct *curr = current->group_leader;
 
+       WARN_ON(is_virtual_pid(pgrp));
+       WARN_ON(is_virtual_pid(session));
+


/Jakob

Jakob Goldbach wrote:
> On Wed, 2008-01-09 at 11:55 +0300, Kirill Korotaev wrote:
> 
>>AFAIR it can be ignored.
> 
> 
> Great. 
> 
> 
> 
>>BTW, In 2.6.18 (what kernel are you running?) these lines are changed to:
>>         WARN_ON(__is_virtual_pid(pgrp));
>>         WARN_ON(__is_virtual_pid(session));
>>
> 
> 
> I am using 2.6.18 from Debian. They also distribute
> patch-026test015-combined.gz in kernel-patch-openvz.deb 
> It contains:

026test015? Maybe it's a misprint and you mean 028test015?
cause 026xxx series was for 2.6.16 kernels, not 2.6.18.

Regardless of this, it is more than outdated version, so please update
up to latests 028stab051 available.
For debian it can be found here:
http://download.openvz.org/debian-systs/

Thanks,
Kirill

> 026test015? Maybe it's a misprint and you mean 028test015?
> cause 026xxx series was for 2.6.16 kernels, not 2.6.18.
> 

It's not a typo. Apperently Debian ships the 2.6.16 patch to their
2.6.18 kernel

http://packages.debian.org/etch/all/kernel-patch-openvz/filelist

> Regardless of this, it is more than outdated version, so please update
> up to latests 028stab051 available. For debian it can be found here:
> http://download.openvz.org/debian-systs/
> 

Will do. Thanks for hint to our deb repo.


/Jakob