Re: Samba 4.3 and 4.4 crashes on FreeBSD 11-ALPHA4

On Sun, Jun 19, 2016 at 09:10:08AM +0200, Daniel Engberg wrote:
> Hi,
> 
> Decided to try ALPHA4 and Samba 4.3 and 4.4 doesn't seem very happy.
> 
> Fatal error 'mutex 0x8013ec000 own 0x18cdc is on list 0x8112161a0 0x0' at line 153 in file /usr/src/lib/libthr/thread/thr_mutex.c (errno = 4)
> [2016/06/19 01:14:47.689256,  0] ../lib/util/fault.c:78(fault_report)
>   ===============================================================
> [2016/06/19 01:14:47.689450,  0] ../lib/util/fault.c:79(fault_report)
>   INTERNAL ERROR: Signal 6 in pid 45222 (4.4.3)
>   Please read the Trouble-Shooting section of the Samba HOWTO
> [2016/06/19 01:14:47.689479,  0] ../lib/util/fault.c:81(fault_report)
>   ===============================================================
> [2016/06/19 01:14:47.689498,  0] ../source3/lib/util.c:791(smb_panic_s3)
>   PANIC (pid 45222): internal error
> [2016/06/19 01:14:47.690577,  0] ../source3/lib/util.c:902(log_stack_trace)
>   BACKTRACE: 6 stack frames:
>    #0 0x803c16fe8 <log_stack_trace+0x28> at /usr/local/lib/samba4/libsmbconf.so.0
>    #1 0x803c16ed2 <smb_panic_s3+0x52> at /usr/local/lib/samba4/libsmbconf.so.0
>    #2 0x801696dad <smb_panic+0x13d> at /usr/local/lib/samba4/libsamba-util.so.0
>    #3 0x801696c57 <fault_setup+0x157> at /usr/local/lib/samba4/libsamba-util.so.0
>    #4 0x8014616df <pthread_sigmask+0x4ff> at /lib/libthr.so.3
>    #5 0x801460cbf <pthread_getspecific+0xe9f> at /lib/libthr.so.3
> [2016/06/19 01:14:47.690686,  0] ../source3/lib/dumpcore.c:313(dump_core)
>   unable to change to %N.core
>   refusing to dump core
> 
> Fatal error 'mutex 0x8013eb000 own 0x18d3d is on list 0x812e161a0 0x0' at line 153 in file /usr/src/lib/libthr/thread/thr_mutex.c (errno = 4)
> [2016/06/19 01:26:41.063681,  0] ../lib/util/fault.c:78(fault_report)
>   ===============================================================
> [2016/06/19 01:26:41.063827,  0] ../lib/util/fault.c:79(fault_report)
>   INTERNAL ERROR: Signal 6 in pid 61297 (4.3.9)
>   Please read the Trouble-Shooting section of the Samba HOWTO
> [2016/06/19 01:26:41.063844,  0] ../lib/util/fault.c:81(fault_report)
>   ===============================================================
> [2016/06/19 01:26:41.063855,  0] ../source3/lib/util.c:789(smb_panic_s3)
>   PANIC (pid 61297): internal error
> [2016/06/19 01:26:41.064679,  0] ../source3/lib/util.c:900(log_stack_trace)
>   BACKTRACE: 6 stack frames:
>    #0 0x803846eb8 <log_stack_trace+0x28> at /usr/local/lib/libsmbconf.so.0
>    #1 0x803846da2 <smb_panic_s3+0x52> at /usr/local/lib/libsmbconf.so.0
>    #2 0x80168c12d <smb_panic+0x13d> at /usr/local/lib/libsamba-util.so.0
>    #3 0x80168bfd7 <fault_setup+0x157> at /usr/local/lib/libsamba-util.so.0
>    #4 0x8014606df <pthread_sigmask+0x4ff> at /lib/libthr.so.3
>    #5 0x80145fcbf <pthread_getspecific+0xe9f> at /lib/libthr.so.3
> [2016/06/19 01:26:41.064747,  0] ../source3/lib/dumpcore.c:313(dump_core)
>   unable to change to %N.core
>   refusing to dump core
> 
> This occurs immediately when you start Samba so it's easy to replicate.
> I know that 4.3 worked on r298887 for sure, I'm going to try this in VM too but I'd apperciate if anyone else could give this a go.
> 
> I noticed that there were some changes in libthr which might be a clue....
> 

Please convince samba to allow to dump core on assert (I have no idea how),
then get a backtrace from the core, using ports/devel/gdb.  After that I
may know where to look next.