Opened 7 years ago

Closed 7 years ago

#10846 closed Bug Report - General (Fixed)

Memory leak in MBE/zeromq logging service

Reported by: tom@… Owned by: beirdo
Priority: minor Milestone: 0.26
Component: MythTV - General Version: Master Head
Severity: medium Keywords: memory leak valgrind
Cc: Ticket locked: no

Description

Since installing latest HEAD branch I am seeing massive memory leaks in the backend process, approx. 1GB per hour is lost while playing back recordings to remote frontends.

MythTV Version : v0.26-pre-663-g2b189a9-dirty MythTV Branch : master Network Protocol : 75 Library API : 0.26.20120617-1 QT Version : 4.6.3 Options compiled in:

linux debug use_hidesyms using_alsa using_oss using_backend using_bindings_perl using_bindings_python using_bindings_php using_dvb using_frontend using_hdhomerun using_ceton using_hdpvr using_iptv using_ivtv using_joystick_menu using_libcrypto using_libdns_sd using_libxml2 using_lirc using_mheg using_opengl_video using_qtwebkit using_qtscript using_qtdbus using_v4l2 using_v4l1 using_x11 using_xrandr using_bindings_perl using_bindings_python using_bindings_php using_mythtranscode using_opengl using_ffmpeg_threads using_live using_mheg using_libxml2

230k gzipped valgrind log located here: http://dr.redpepperracing.com/uploads/vglog.log.gz

All I did was start the backend, and watch a recording on a remote frontned for about 15 minutes. There were no other clients connected, and no recordings were happening at the time.

Is there anything else I can provide to help debug?

Tom

Change History (2)

comment:1 Changed 7 years ago by beirdo

Please do a git pull and try again.

There have been several minor fixes since the version you have.

Also, this is hardly a massive leak. (as I indicated in the email thread). Still needs fixing though.

comment:2 Changed 7 years ago by Jim Stichnoth

Component: MythTV - MythlogserverMythTV - General
Milestone: unknown0.26
Resolution: Fixed
Status: newclosed

Fixed in 0034a85fa007a99992f1aaa3909fa84ce8cfe898.

For reasons not entirely understood, Qt was not taking care of certain deferred delete events, leading to an uncontrolled memory growth that would be conveniently cleaned up only at program exit.

Note: This may also fix similar problems that might have come up in mythfrontend.

Note: See TracTickets for help on using tickets.