In my midi sequencer program I use the following code (in preparation for a
new start):

	def stop(self):
		self.csPerfThread.Stop()
		self.csPerfThread.Join()
		self.csound.Stop()
		self.csound.Reset()
		self.playing = False
		self.started = False

It crashses in libfluidsynth ( I use some soundfonts).
I already tried with time.sleep(x) between the Stop, Join, etc, to no
avail...

Richard



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086.html
Sent from the Csound - General mailing list archive at Nabble.com.

But when does it crash (what function)?
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 10 May 2014, at 15:30, zappfinger  wrote:

> In my midi sequencer program I use the following code (in preparation for a
> new start):
> 
> 	def stop(self):
> 		self.csPerfThread.Stop()
> 		self.csPerfThread.Join()
> 		self.csound.Stop()
> 		self.csound.Reset()
> 		self.playing = False
> 		self.started = False
> 
> It crashses in libfluidsynth ( I use some soundfonts).
> I already tried with time.sleep(x) between the Stop, Join, etc, to no
> avail...
> 
> Richard
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

By the way, there is no need to call csound.Stop(). That is only if you are using csound.Perform() to perform csound (you’re not,
you are using csoundPerformanceThread()).
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 10 May 2014, at 15:30, zappfinger  wrote:

> In my midi sequencer program I use the following code (in preparation for a
> new start):
> 
> 	def stop(self):
> 		self.csPerfThread.Stop()
> 		self.csPerfThread.Join()
> 		self.csound.Stop()
> 		self.csound.Reset()
> 		self.playing = False
> 		self.started = False
> 
> It crashses in libfluidsynth ( I use some soundfonts).
> I already tried with time.sleep(x) between the Stop, Join, etc, to no
> avail...
> 
> Richard
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

It looks like this happens in Reset().
When I comment this out, it does not crash.
I also do not use Stop() now.  



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735099.html
Sent from the Csound - General mailing list archive at Nabble.com.

You might not need to reset as I think it’s done by csoundPerformanceThead.
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 10 May 2014, at 21:12, zappfinger  wrote:

> It looks like this happens in Reset().
> When I comment this out, it does not crash.
> I also do not use Stop() now.  
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735099.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

When I do not reset I get the following:

Csound is already started, call csoundReset()
before starting again 
Score finished in csoundPerformKsmps().

So it seems necessary.





--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735101.html
Sent from the Csound - General mailing list archive at Nabble.com.

Can you prepare a short example of the problem (with csound code etc).
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 10 May 2014, at 21:38, zappfinger  wrote:

> When I do not reset I get the following:
> 
> Csound is already started, call csoundReset()
> before starting again 
> Score finished in csoundPerformKsmps().
> 
> So it seems necessary.
> 
> 
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735101.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

Assuming you have no PySide Qt installed, I prepared a Python/csound version,
stripped of all the Qt GUI stuff, and .qml files I use.
To my surprise this seemed to work at first without crashes, until I
commented out the import pyside (I use nothing of PySide in this code!) line
..
Total weirdness! It also crashes in a different place now, not in fluid..
See the comments in the header.  

3 score tracks are included, for piano, bass and drums. The piece lasts for
30 seconds, but is stopped and restarted a few times in the code. 

Richard
csCrash.py   
track1.sco   
track2.sco   
track10.sco   



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735119.html
Sent from the Csound - General mailing list archive at Nabble.com.

Your test is not crashing here when commenting out PySide (I don’t have PySide anyway). 
It plays and exits cleanly. The only difference is that I do not have your soundfont files and 
I replaced them with the one I have here (but still using 3 separate engines).

This is OSX 10.9 with the latest Csound.

========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 11 May 2014, at 11:18, zappfinger  wrote:

> Assuming you have no PySide Qt installed, I prepared a Python/csound version,
> stripped of all the Qt GUI stuff, and .qml files I use.
> To my surprise this seemed to work at first without crashes, until I
> commented out the import pyside (I use nothing of PySide in this code!) line
> ..
> Total weirdness! It also crashes in a different place now, not in fluid..
> See the comments in the header.  
> 
> 3 score tracks are included, for piano, bass and drums. The piece lasts for
> 30 seconds, but is stopped and restarted a few times in the code. 
> 
> Richard
> csCrash.py   
> track1.sco   
> track2.sco   
> track10.sco   
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735119.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

The piano font I used for track 1 is almost 40 Mb, a bit large, but not
extraordinary.
I replaced it with the Roland GS SF that I use for tracks 2 and 10 (only 3.5
Mb), but it still crashes on me.
I did just install the latest Csound for OSX, 6.03, was still on 6.02, but
that did not help.

 



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735127.html
Sent from the Csound - General mailing list archive at Nabble.com.

another thing: did you try removing the -M option just to see if the MIDI input has anything to do with this?
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 11 May 2014, at 14:19, zappfinger  wrote:

> The piano font I used for track 1 is almost 40 Mb, a bit large, but not
> extraordinary.
> I replaced it with the Roland GS SF that I use for tracks 2 and 10 (only 3.5
> Mb), but it still crashes on me.
> I did just install the latest Csound for OSX, 6.03, was still on 6.02, but
> that did not help.
> 
> 
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735127.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

Yes, but then I get a 'bus error: 10', shortly after it starts playing again
after the first stop and Python quits:

....
--Csound version 6.03.1 (double samples) May  6 2014
displays suppressed
0dBFS level = 1.0
Created fluidEngine 0x0x7fcafba4f300 with sampling rate = 44100.000000,
chorus on, reverb on, channels 256, voices 4096.
Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
fluidSynth: 0x0x7fcafba4f300  soundFontId: 1.
Created fluidEngine 0x0x7fcaf97310b0 with sampling rate = 44100.000000,
chorus on, reverb on, channels 256, voices 4096.
Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
fluidSynth: 0x0x7fcaf97310b0  soundFontId: 1.
Created fluidEngine 0x0x7fcafbaecb20 with sampling rate = 44100.000000,
chorus on, reverb on, channels 256, voices 4096.
Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
fluidSynth: 0x0x7fcafbaecb20  soundFontId: 1.
orch now loaded
audio buffered in 64 sample-frame blocks
PortAudio V19-devel (built Jun  4 2013 18:37:15)
   0: dac0 (Built-in Output)
   1: dac1 (Soundflower (2ch))
   2: dac2 (Soundflower (64ch))
PortAudio: selected output device 'Built-in Output'
writing 128 sample blks of 64-bit floats to dac 
SECTION 1:
 i  11 time     0.00073:     1.00000
 i  11 time     0.50141:     2.00000
Bus error: 10

BTW, is it not strange that all 3 soundfonts have a soundFontId of 1?

 



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735129.html
Sent from the Csound - General mailing list archive at Nabble.com.

I think the soundFontId is probably OK. As I can’t reproduce any of these crashes, it’s hard for me to suggest much more.
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 11 May 2014, at 15:03, zappfinger  wrote:

> Yes, but then I get a 'bus error: 10', shortly after it starts playing again
> after the first stop and Python quits:
> 
> ....
> --Csound version 6.03.1 (double samples) May  6 2014
> displays suppressed
> 0dBFS level = 1.0
> Created fluidEngine 0x0x7fcafba4f300 with sampling rate = 44100.000000,
> chorus on, reverb on, channels 256, voices 4096.
> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
> fluidSynth: 0x0x7fcafba4f300  soundFontId: 1.
> Created fluidEngine 0x0x7fcaf97310b0 with sampling rate = 44100.000000,
> chorus on, reverb on, channels 256, voices 4096.
> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
> fluidSynth: 0x0x7fcaf97310b0  soundFontId: 1.
> Created fluidEngine 0x0x7fcafbaecb20 with sampling rate = 44100.000000,
> chorus on, reverb on, channels 256, voices 4096.
> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
> fluidSynth: 0x0x7fcafbaecb20  soundFontId: 1.
> orch now loaded
> audio buffered in 64 sample-frame blocks
> PortAudio V19-devel (built Jun  4 2013 18:37:15)
>   0: dac0 (Built-in Output)
>   1: dac1 (Soundflower (2ch))
>   2: dac2 (Soundflower (64ch))
> PortAudio: selected output device 'Built-in Output'
> writing 128 sample blks of 64-bit floats to dac 
> SECTION 1:
> i  11 time     0.00073:     1.00000
> i  11 time     0.50141:     2.00000
> Bus error: 10
> 
> BTW, is it not strange that all 3 soundfonts have a soundFontId of 1?
> 
> 
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735129.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

I tried the python code here.  If it doesn't find proper sf2s, I get a
crash on first note init. When I put in a valid sf2 here to use, it
does not crash, even on resets.  The id's should be fine as the
instruments are registered with three different engines.

Richard: which OSX are you on?  Also, what version of python are you
using?  I'm testing on OSX 10.9.2 and have this when i start python:

Python 2.7.5 (default, Aug 25 2013, 00:04:04)
[GCC 4.2.1 Compatible Apple LLVM 5.0 (clang-500.0.68)] on darwin

steven

On Sun, May 11, 2014 at 3:31 PM, Victor Lazzarini
 wrote:
> I think the soundFontId is probably OK. As I can’t reproduce any of these crashes, it’s hard for me to suggest much more.
> ========================
> Dr Victor Lazzarini
> Senior Lecturer
> NUI Maynooth, Ireland
> victor dot lazzarini at nuim dot ie
>
>
>
>
> On 11 May 2014, at 15:03, zappfinger  wrote:
>
>> Yes, but then I get a 'bus error: 10', shortly after it starts playing again
>> after the first stop and Python quits:
>>
>> ....
>> --Csound version 6.03.1 (double samples) May  6 2014
>> displays suppressed
>> 0dBFS level = 1.0
>> Created fluidEngine 0x0x7fcafba4f300 with sampling rate = 44100.000000,
>> chorus on, reverb on, channels 256, voices 4096.
>> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
>> fluidSynth: 0x0x7fcafba4f300  soundFontId: 1.
>> Created fluidEngine 0x0x7fcaf97310b0 with sampling rate = 44100.000000,
>> chorus on, reverb on, channels 256, voices 4096.
>> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
>> fluidSynth: 0x0x7fcaf97310b0  soundFontId: 1.
>> Created fluidEngine 0x0x7fcafbaecb20 with sampling rate = 44100.000000,
>> chorus on, reverb on, channels 256, voices 4096.
>> Loading SoundFont : ./samples/003.3mg GS Roland Sound Set 16 bit Bank.SF2.
>> fluidSynth: 0x0x7fcafbaecb20  soundFontId: 1.
>> orch now loaded
>> audio buffered in 64 sample-frame blocks
>> PortAudio V19-devel (built Jun  4 2013 18:37:15)
>>   0: dac0 (Built-in Output)
>>   1: dac1 (Soundflower (2ch))
>>   2: dac2 (Soundflower (64ch))
>> PortAudio: selected output device 'Built-in Output'
>> writing 128 sample blks of 64-bit floats to dac
>> SECTION 1:
>> i  11 time     0.00073:     1.00000
>> i  11 time     0.50141:     2.00000
>> Bus error: 10
>>
>> BTW, is it not strange that all 3 soundfonts have a soundFontId of 1?
>>
>>
>>
>>
>>
>> --
>> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735129.html
>> Sent from the Csound - General mailing list archive at Nabble.com.
>>
>>
>> Send bugs reports to
>>        https://github.com/csound/csound/issues
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>>
>>
>>
>
>
>
> Send bugs reports to
>         https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>
>

Hi Steven,

I am on the same OS and Python versions you have.

I tried some more things:
deleting the csound object, restarting csound completely by adding these
lines as the first lines in .startthrd():
		csoundInitialize(3)
		self.csound = Csound()

but then it crashes the second time I run or stop it.

Could it be that the fluid opcodes are not removed correctly when csound
exits?

Richard





--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735135.html
Sent from the Csound - General mailing list archive at Nabble.com.

It's entirely possible the fluidopcodes may have a bug.  Do you
experience any problems resetting if you do not use those opcodes?

I have had problems with fluidopcodes crashing API apps in the past
(on Android in particular, if I remember correctly).  They currently
use a global map of fluid engines that isn't well suited for
re-entrancy. (The code should store instances in the CSOUND globalVar
system I think).  I thought though that things were fixed, but perhaps
not.

On Sun, May 11, 2014 at 5:33 PM, zappfinger  wrote:
> Hi Steven,
>
> I am on the same OS and Python versions you have.
>
> I tried some more things:
> deleting the csound object, restarting csound completely by adding these
> lines as the first lines in .startthrd():
>                 csoundInitialize(3)
>                 self.csound = Csound()
>
> but then it crashes the second time I run or stop it.
>
> Could it be that the fluid opcodes are not removed correctly when csound
> exits?
>
> Richard
>
>
>
>
>
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735135.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
> Send bugs reports to
>         https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>
>

By the way, I think csoundInitialize()  is no-op in a second call in a given process, so calling it should not affect anything.
========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 11 May 2014, at 17:33, zappfinger  wrote:

> Hi Steven,
> 
> I am on the same OS and Python versions you have.
> 
> I tried some more things:
> deleting the csound object, restarting csound completely by adding these
> lines as the first lines in .startthrd():
> 		csoundInitialize(3)
> 		self.csound = Csound()
> 
> but then it crashes the second time I run or stop it.
> 
> Could it be that the fluid opcodes are not removed correctly when csound
> exits?
> 
> Richard
> 
> 
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735135.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

There is something strange about my system. I removed the fluid opcodes in my
csd, replaced them by some FM instruments, and it still crashes:

Python quit unexpectedly while using the libfluidsynth.1.dylib plug-in.

How can the plug-in be loaded when I do not use fluid opcodes?

Richard 



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735140.html
Sent from the Csound - General mailing list archive at Nabble.com.

The plugin gets loaded because it’s in the plugin dir. You can move it out of there and it won’t load anymore.

========================
Dr Victor Lazzarini
Senior Lecturer
NUI Maynooth, Ireland
victor dot lazzarini at nuim dot ie




On 11 May 2014, at 22:01, zappfinger  wrote:

> There is something strange about my system. I removed the fluid opcodes in my
> csd, replaced them by some FM instruments, and it still crashes:
> 
> Python quit unexpectedly while using the libfluidsynth.1.dylib plug-in.
> 
> How can the plug-in be loaded when I do not use fluid opcodes?
> 
> Richard 
> 
> 
> 
> --
> View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735140.html
> Sent from the Csound - General mailing list archive at Nabble.com.
> 
> 
> Send bugs reports to
>        https://github.com/csound/csound/issues
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
> 
> 
> 

Ok, I tried it now on a Ubuntu VM and on a Windows machine, all works fine.
Since you and Victor can't reproduce it, it must be something on my Mac.
How can I completely uninstall Csound (including the Python API module) in
order to do a fresh re-install?

Richard



--
View this message in context: http://csound.1045644.n5.nabble.com/crash-in-libfluidsynth-when-stopping-performance-thread-tp5735086p5735220.html
Sent from the Csound - General mailing list archive at Nabble.com.