muggot / openmcu

openmcu.ru
Other
220 stars 154 forks source link

openmcu-ru critical memory leak #26

Open nvnhcmus opened 6 years ago

nvnhcmus commented 6 years ago

I have a big problem with openmcu, it has a critical leak memory issue.

Is there any solution or tools for me to check and fix the leak memory?

@P/S: I used the current version from github.

tenge-amzn commented 5 years ago

I have a big problem with openmcu, it has a critical leak memory issue.

  • sip registeration with sofia-sip is leaking memory.
  • big memory leak in video call.

Is there any solution or tools for me to check and fix the leak memory?

@P/S: I used the current version from github.

Hi, any findings about the memory leak? I also noticed lots of memory leak issue, but still locating them.

kay27 commented 5 years ago

I do apologize, still can't reproduce it. It might be handful if you post some valgrind data here (any other dumps showing leak details as well). MANY MANY Thanks in advance! P.S. Please take a look if you didn't: https://wiki.videoswitch.ru/en/bugreport

tenge-amzn commented 5 years ago

Hi, I have some valgrind definite loss sample here, they are all related to the message header with sofia lib.

==2662== 31,214 (1,792 direct, 29,422 indirect) bytes in 2 blocks are definitely lost in loss record 2,412 of 2,432 ==2662== at 0x4C2A975: calloc (vg_replace_malloc.c:711) ==2662== by 0x7357236: su_home_new (su_alloc.c:559) ==2662== by 0x72A5CD9: msg_create (msg.c:61) ==2662== by 0x72ACD21: msg_next (msg_parser.c:576) ==2662== by 0x7377E15: tport_parse (tport.c:2928) ==2662== by 0x7377B37: tport_recv_event (tport.c:2874) ==2662== by 0x73777A8: tport_base_wakeup (tport.c:2775) ==2662== by 0x737772D: tport_wakeup (tport.c:2761) ==2662== by 0x7364624: su_epoll_port_wait_events (su_epoll_port.c:510) ==2662== by 0x7360856: su_base_port_step (su_base_port.c:467) ==2662== by 0x735CA36: su_port_step (su_port.h:340) ==2662== by 0x735DBD6: su_root_sleep (su_root.c:886) ==2662== ==2662== 38,282 (824 direct, 37,458 indirect) bytes in 1 blocks are definitely lost in loss record 2,413 of 2,432 ==2662== at 0x4C28BE3: malloc (vg_replace_malloc.c:299) ==2662== by 0x6E4BE67: CRYPTO_malloc (in /opt/mcu/UCBuzzMCUNg-1.0.200345.0/openmcu-ru/lib/libcrypto.so.1.0.0) ==2662== by 0x911134F: SSL_new (in /opt/mcu/UCBuzzMCUNg-1.0.200345.0/openmcu-ru/lib/libssl.so.1.0.0) ==2662== by 0x7385B67: tls_init_secondary (tport_tls.c:493) ==2662== by 0x738322C: tport_tls_init_secondary (tport_type_tls.c:283) ==2662== by 0x737122F: tport_alloc_secondary (tport.c:888) ==2662== by 0x7384076: tport_tls_accept (tport_type_tls.c:579) ==2662== by 0x7377564: tport_wakeup_pri (tport.c:2736) ==2662== by 0x7364624: su_epoll_port_wait_events (su_epoll_port.c:510) ==2662== by 0x7360856: su_base_port_step (su_base_port.c:467) ==2662== by 0x735CA36: su_port_step (su_port.h:340) ==2662== by 0x735DBD6: su_root_sleep (su_root.c:886)