| Project: | Call Session Control Functions |
| Component: | Code |
| Category: | bug |
| Priority: | normal |
| Assigned: | Unassigned |
| Status: | closed |
During my testing I am shutting down the CSCFs by typing a Ctl-C however they don't usually exit by themselves. I have to go to another screen and manually kill them.
I got a core dump from icscf while it's last process was hanging and got the following backtrace from gdb:
#0 0xff1bfeb4 in lwp_yield () from /lib/libc.so.1
#1 0xff304b2c in sched_yield () from /lib/librt.so.1
#2 0x00087194 in qm_sums (qm=0x10ad28) at fastlock.h:323
#3 0x00037b10 in cleanup (show_status=1) at main.c:462
#4 0x00038214 in shutdown_children (sig=1090560,show_status=1) at main.c:561
#5 0x00038b1c in handle_sigs () at main.c:634
#6 0x00039508 in main_loop () at main.c:1110
#7 0x0003aec0 in main (argc=0, argv=0x0) at main.c:1614
... and this one from scscf:
#0 0xff1bfeb4 in lwp_yield () from /lib/libc.so.1
#1 0xff304b2c in sched_yield () from /lib/librt.so.1
#2 0xfede3fb8 in mod_destroy () at ../fastlock.h:323
#3 0x00068dac in destroy_modules () at sr_module.c:393
#4 0x00037a14 in cleanup (show_status=1) at main.c:439
#5 0x00038214 in shutdown_children (sig=1090560,show_status=1) at main.c:561
#6 0x00038b1c in handle_sigs () at main.c:634
#7 0x00039508 in main_loop () at main.c:1110
#8 0x0003aec0 in main (argc=0, argv=0x0) at main.c:1614
-Vance
Comments
Fix found
I have fixed the issue since r327 by removing the lock on process_lock in qm_sums(). However, now the output looks awful. Haven't identified yet the cause of the deadlock (the previous lock_get).