From ec2808990dbc15b6652070c1bdf9487cb1b3f38a Mon Sep 17 00:00:00 2001 From: Pepper Date: Fri, 30 Jan 2015 03:55:43 -0500 Subject: fixed leak and sample rate issue --- src/VALOUT | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'src/VALOUT') diff --git a/src/VALOUT b/src/VALOUT index bbb2683..4907c63 100644 --- a/src/VALOUT +++ b/src/VALOUT @@ -3,6 +3,9 @@ ==19216== Using Valgrind-3.10.1 and LibVEX; rerun with -h for copyright info ==19216== Command: ./dssi-render /usr/local/lib/dssi/amsynth_dssi.so -p 1 ==19216== +//so I've got a memory leak here, at least I'm pretty sure that's what this is...basically if a midi file is too long, it segfaults. + +this is the output from valgrind... is this helpful? or do I need to recompile with -g? ==19216== Conditional jump or move depends on uninitialised value(s) ==19216== at 0x6DBEC62: revmodel::update() (in /usr/local/lib/dssi/amsynth_dssi.so) ==19216== by 0x6DBEE88: revmodel::setwet(float) (in /usr/local/lib/dssi/amsynth_dssi.so) @@ -281156,8 +281159,9 @@ dssi-render: Wrote 255 frames to output.wav event table last nframe: 8904056 run_synth(instancehandle, 255, - 1: event_type: NOTE_ONchannel: 0 note: 68 veloci==19216== Argument 'size' of function realloc has a fishy (possibly negative) value: -173348 +so it's invalid size calulcation, need to check those functions ==19216== at 0x4C2C29E: realloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so) -==19216== by 0x4027AE: ladspa_run_sample_callback (ladspa_run_synth.c:15) +==19216== by 0x4027AE: ladspa_run_sample_callback (ladspa_run_synth.c:15) //here? yes: ==19216== by 0x4030FA: get_events (midi_loader.c:228) ==19216== by 0x4047BC: fluid_track_send_events (fluid_midi.c:1212) ==19216== by 0x40326B: load_midi_file (midi_loader.c:297) -- cgit v1.2.3-70-g09d2