[Mono-list] Mono process hangs on exit

[Francisco Figueiredo Jr.]

On Wed, May 7, 2008 at 7:21 AM, Dick Porter <dick at ximian.com> wrote:
> On Wed, 2008-05-07 at 00:24 -0300, Francisco Figueiredo Jr. wrote:
>> Hi, all!
>>
>> I'm facing an interesting problem...
>>
>> My mono processes seem to be not finishing correctly. They hang when
>> they were supposed to finish.
>>
>> If I execute nant it hangs in the end instead of returning to bash.
>>
>> Do you have anything like that?
>>
>> Is there something I can do to check what could be happening?
>>
>> I'm using Ubuntu Hardy 8.04 with mono 1.2.6
>
> This sounds like https://bugzilla.novell.com/show_bug.cgi?id=322163
> (especially the ubuntu with mono 1.2.6 part matching the later
> comments.)  If you could add to that bug we will have more data to work
> with.
>


Thanks Dick, for pointer.

I'm having problems with my account right now but this is the
information I could gather:

This is the backtrace:

Thread 3 (Thread 0xb769db90 (LWP 21365)):
#0  0xb7fda410 in __kernel_vsyscall ()
#1  0x4cce6aa5 in pthread_cond_wait@@GLIBC_2.3.2 ()
   from /lib/tls/i686/cmov/libpthread.so.0
#2  0x081088ff in ?? ()
#3  0xb7b971dc in ?? ()
#4  0xb7b971c4 in ?? ()
#5  0x4cceb14f in pthread_getaffinity_np@@GLIBC_2.3.4 ()
   from /lib/tls/i686/cmov/libpthread.so.0
#6  0x0810b3cd in ?? ()
#7  0x00000000 in ?? ()

Thread 2 (Thread 0xb7c25b90 (LWP 21364)):
#0  0xb7fda410 in __kernel_vsyscall ()
#1  0x4ccea196 in nanosleep () from /lib/tls/i686/cmov/libpthread.so.0
#2  0x08105c91 in ?? ()
#3  0xb7c25344 in ?? ()
#4  0x00000000 in ?? ()

Thread 1 (Thread 0xb7fc6940 (LWP 21361)):
#0  0xb7fda410 in __kernel_vsyscall ()
#1  0x4cce7ae7 in pthread_kill () from /lib/tls/i686/cmov/libpthread.so.0
---Type <return> to continue, or q <return> to quit---
#2  0x08131f92 in ?? ()
#3  0xb769db90 in ?? ()
#4  0x0000001e in ?? ()
#5  0x00000001 in ?? ()
#6  0x081255d0 in ?? ()
#7  0x00000000 in ?? ()
#0  0xb7fda410 in __kernel_vsyscall ()


My machine is a  pentium4m with HT, but I disabled HT. I will try to
reenable it again to see if I face this problem too.

This stacktrace is from a nant execution.
I can reproduce it all time.

Interesting enough is that if I just run gmcs, which also uses mono,
it exists ok. Only when I execute any other program it hangs.

I hope it helps.




-- 
Regards,

Francisco Figueiredo Jr.
http://fxjr.blogspot.com
http://www.npgsql.org