[Cs-dev] csoundCompileCsd()
Date | 2015-06-02 14:21 |
From | Rory Walsh |
Subject | [Cs-dev] csoundCompileCsd() |
Attachments | None None |
I'm having some intermittent crashes with csoundCompileCsd(). What are the normal procedures for using this after a performance has already started? When I simply call compiledCsd() it runs most of the time, but every so often I get a crash(backtrace given below). The reason I started using compileCsd() was so I could avoid calling csoundReset() between updates. But is there something else I should be doing to free up memory before I recompile?
0 0x00007ffff67fc761 mfree /home/rory/sourcecode/cabbageaudio/csound/Engine/memalloc.c 164 1 0x00007ffff6a3dd71 free_instrtxt /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 888 2 0x00007ffff6a3e00b add_to_deadpool /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 937 3 0x00007ffff6a3eb53 insert_instrtxt /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 1151 4 0x00007ffff6a3f266 engineState_merge /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 1296 5 0x00007ffff6a401b1 csoundCompileTree /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 1612 6 0x00007ffff6a406d7 csoundCompileOrc /home/rory/sourcecode/cabbageaudio/csound/Engine/csound_orc_compile.c 1698 7 0x00007ffff695f233 csoundCompileCsd /home/rory/sourcecode/cabbageaudio/csound/Top/main.c 505 8 0x00000000004fcb53 Csound::CompileCsd /usr/local/include/csound/csound.hpp 264 |
Date | 2015-06-02 14:33 |
From | Victor Lazzarini |
Subject | Re: [Cs-dev] csoundCompileCsd() |
You should be free to call this at any point. We did extensive work to recover memory from dead instances and this could be a bug resulting from it. If we had a test code then we could debug. ======================== Dr Victor Lazzarini Dean of Arts, Celtic Studies and Philosophy, Maynooth University, Maynooth, Co Kildare, Ireland Tel: 00 353 7086936 Fax: 00 353 1 7086952 > On 2 Jun 2015, at 14:21, Rory Walsh |
Date | 2015-06-02 15:02 |
From | Rory Walsh |
Subject | Re: [Cs-dev] csoundCompileCsd() |
Attachments | None None |
I'll see if I can throw something together. On 2 June 2015 at 14:33, Victor Lazzarini <Victor.Lazzarini@nuim.ie> wrote: You should be free to call this at any point. We did extensive work to recover memory from dead instances and |
Date | 2015-06-02 15:27 |
From | Rory Walsh |
Subject | Re: [Cs-dev] csoundCompileCsd() |
Attachments | compileCSD.cpp test.csd None None |
As promised. You'll notice that at some point there is a problem with the f0 statement in the score. But taking that out results in a segfault. On 2 June 2015 at 15:02, Rory Walsh <rorywalsh@ear.ie> wrote:
|
Date | 2015-06-02 16:04 |
From | Victor Lazzarini |
Subject | Re: [Cs-dev] csoundCompileCsd() |
Thanks; there is definitely a bug and it is to do with “named” instruments. It appears that if you use numbers only, it is OK. I’ll try to get to the bottom of it. ======================== Dr Victor Lazzarini Dean of Arts, Celtic Studies and Philosophy, Maynooth University, Maynooth, Co Kildare, Ireland Tel: 00 353 7086936 Fax: 00 353 1 7086952 > On 2 Jun 2015, at 15:27, Rory Walsh |
Date | 2015-06-02 16:12 |
From | Rory Walsh |
Subject | Re: [Cs-dev] csoundCompileCsd() |
Attachments | None None |
Thanks Victor. On 2 June 2015 at 16:04, Victor Lazzarini <Victor.Lazzarini@nuim.ie> wrote: Thanks; there is definitely a bug and it is to do with “named” instruments. It appears that if you use |
Date | 2015-06-02 21:42 |
From | Victor Lazzarini |
Subject | Re: [Cs-dev] csoundCompileCsd() |
Ok, I think I’ve stabilised this situation and named instruments are not crashing on recompilation (at least they pass your test). I need to go back and assess the situation a little more, because I want to check that we are recovering as much memory as possible. But for now, the code in GIT should work. ======================== Dr Victor Lazzarini Dean of Arts, Celtic Studies and Philosophy, Maynooth University, Maynooth, Co Kildare, Ireland Tel: 00 353 7086936 Fax: 00 353 1 7086952 > On 2 Jun 2015, at 16:12, Rory Walsh |