[uml-devel] run away 2.6.0

[uml-devel] run away 2.6.0

[Patrick Kilian]

Hi,

I found a problem with the 2.6.0-test2 and 2.6.0-test3 kernels. Both 
kernels are clean sources from ftp.kernel.org patched with the 
coresponding patches from user-mode-linux.sourceforge.net. With both 
kernels I did make xconfig ARCH=um and make linux ARCH=um. Startup work 
without a problem and the running uml guest works like a charm. But 
when I issue a "uml# shutdown -n -h now" the cpu load jumps to 100% 
after some seconds and stays up till I kill the uml guest.

Is this a known bug? Is this possibly a error in the .config file? How 
should I proceed to track this bug down?
By the way I' using minit written by Felix von Leitner and not the 
normal System V init.

mfg,
Patrick "Petschge" Kilian

PS: If you need further information like the .config file, just ask for 
it.

-- 
I still think the fastest way to get wholesale IPv6 adoption
is to write a pr0n server that serves thumbnails over IPv[46]
and full-size images over IPv6 only.              -- Ingvar



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] run away 2.6.0

[Jeff Dike]

petschge@gmx.de said:
> How  should I proceed to track this bug down?

Figure out what process is causing the hang and get a stack trace from it.

				Jeff



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] run away 2.6.0

[Patrick "Petschge" Kilian]

Hi,

Jeff Dike wrote:
> Figure out what process is causing the hang and get a stack trace from it.

Ok, it is the shutdown process itself. After issueing a shutdown and after 
the the cpu load hits the roof, I still can login into the uml guest and I 
got the following top screen:

11:35am  up 12298 days,  9:35,  0 users,  load average: 0.48, 0.52, 0.25
14 processes: 12 sleeping, 2 running, 0 zombie, 0 stopped
CPU states:  0.0% user, 100.0% system,  0.0% nice,  0.0% idle
Mem:    29176K av,    7992K used,   21184K free,       0K shrd,     204K 
buff
Swap:       0K av,       0K used,       0K free                    3156K 
cached

  PID USER     PRI  NI  SIZE  RSS SHARE STAT %CPU %MEM   TIME COMMAND
   35 root      24   0  1376  620  1324 R    17.7  2.1   0:06 shutdown
   47 root      22   0  1744  944  1652 R     0.7  3.2   0:00 top
    1 root      15   0    28   28    12 S     0.0  0.0   0:00 init
    2 root      34  19     0    0     0 SWN   0.0  0.0   0:00 ksoftirqd/0
    3 root       5 -10     0    0     0 SW<   0.0  0.0   0:00 events/0
    4 root      15 -10     0    0     0 SW<   0.0  0.0   0:00 kblockd/0
    5 root      25   0     0    0     0 SW    0.0  0.0   0:00 pdflush
    6 root      15   0     0    0     0 SW    0.0  0.0   0:00 pdflush
    7 root      25   0     0    0     0 SW    0.0  0.0   0:00 kswapd0
    8 root      10 -10     0    0     0 SW<   0.0  0.0   0:00 aio/0
   42 root      18   0  2056  952  1752 S     0.0  3.2   0:00 login
   44 root      16   0  2308 1128  2224 S     0.0  3.8   0:00 run
   45 root      21   0  2048 1156  1752 S     0.0  3.9   0:00 login
   46 root      23   0  2356 1320  2224 S     0.0  4.5   0:00 bash

So I restarted the uml guest with the debug swith and did attache gdb to the 
shutdown process and got the following backtrace:

(gdb) bt
#0  0xa01b3fa1 in __kill ()
#1  0xa0037714 in os_usr1_process (pid=3486) at 
arch/um/os-Linux/process.c:90
#2  0xa0024378 in set_user_mode (t=0x0) at 
arch/um/kernel/tt/process_kern.c:446
#3  0xa0025ca7 in sig_handler_common_tt (sig=12, sc_ptr=0xa0b67d28)
    at arch/um/kernel/tt/trap_user.c:43
#4  0xa0021fdf in sig_handler (sig=12, sc=
      {gs = 0, __gsh = 0, fs = 0, __fsh = 0, es = 43, __esh = 0, ds = 43, 
__dsh = 0, edi = 134526247, esi = 0, ebp = 2684353100, esp = 2684353048, 
ebx = 4294967295, edx = 0, ecx = 2684353144, eax = 4294967286, trapno = 14, 
err = 6, eip = 1074537161, cs = 35, __csh = 0, eflags = 2097670, 
esp_at_signal = 2684353048, ss = 43, __ssh = 0, fpstate = 0xa0b67d80, 
oldmask = 503390208, cr2 = 1075040288})
    at arch/um/kernel/trap_user.c:103
#5  <signal handler called>
#6  0x400c22c9 in ?? ()
#7  0x080499d0 in ?? ()
#8  0x08049bdf in ?? ()
#9  0x08049cf6 in ?? ()
#10 0x0804a8f3 in ?? ()
#11 0x40035991 in ?? ()

So I need to get a shutdown with debugging sysmbols compiled in, right?

mfg,
Patrick "Petschge" Kilian

PS.: The uptime looks pretty strange, but I can reproduce this big uptime 
value.

-- 
Erst wenn der letzte FTP Server kostenpflichtig, der letzte
GNU-Sourcecode verkauft, der letzte Algorithmus patentiert, der letzte
Netzknoten kommerzialisiert, die letzte Newsgroup moderiert wird,
werdet Ihr merken, dass man mit Geld allein nicht programmieren kann.



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] run away 2.6.0

[Jeff Dike]

petschge@web.de said:
> Why doesn't gdb show me the function names? 

Because those are all process addresses.  gdb only knows about UML symbols.

				Jeff



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] run away 2.6.0

[Patrick "Petschge" Kilian]

Hi,

Jeff Dike wrote:
>> Why doesn't gdb show me the function names?
> Because those are all process addresses.  gdb only knows about UML
> symbols.
Ok, I admit, I never used gdb for anything else than debugging uml. Please 
could anybody give me a hint or some manuals to read, how I can debug this 
situatuion?

Thanks for your help,
Patrick "Petschge" Kilian

-- 
I still think the fastest way to get wholesale IPv6 adoption
is to write a pr0n server that serves thumbnails over IPv[46]
and full-size images over IPv6 only.              -- Ingvar



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel