| Bugs item #2959788, was opened at 2010-02-26 18:54
Message generated for change (Comment added) made by jpff
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=564599&aid=2959788&group_id=81968
Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: csound
Group: csound5
>Status: Closed
>Resolution: Wont Fix
Priority: 5
Private: No
Submitted By: Rene Djack (rene-djack)
>Assigned to: John ffitch (jpff)
Summary: STKShakers not stable in real time on UBUNTU
Initial Comment:
Below csd give seg fault on Ubuntu 32b when playing in RT with midi.
Problem occurs when playing around midi keys 57 to 61.
(all other STK opcodes are OK in real time)
-fdhnm0 -odac:plughw -b16 -B512 --expression-opt -+rtaudio=alsa -+rtmidi=alsa -Mhw:1,0 --midi-key=4
; MESSAGE : not stable; seg fault with midi-key 57 to 61
sr = 44100
ksmps = 16
nchnls = 1
0dbfs = 2.0
instr 1 ; STK Shakers
print p4
iHz = cpsmidinn(p4)
asig STKShakers iHz, 1.0, 1071, 22, 11, 1, 128, 100, 1, 30
out asig
endin
f 0 300
----------------------------------------------------------------------
>Comment By: John ffitch (jpff)
Date: 2011-01-17 09:53
Message:
I tink I will hav eto mark this as "Will not fix" as it is an STK bug.
I am attempting to contact the stk community to see if they wil fix; there
is an easy change which would fix but a local fix would disappear.
I will leave comment sopen on this after it is closed
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2011-01-10 18:59
Message:
OK found a real problem There re 24 instruments declared but this program
writes to instr 28 and hence overwrites something
798 void Shakers :: noteOn( StkFloat frequency, StkFloat amplitude )
799 {
800 // Yep ... pretty kludgey, but it works!
801 int noteNum = (int) ((12*log(frequency/220.0)/log(2.0)) + 57.01)
% 32;
802 if (instType_ != noteNum) instType_ = this->setupNum(noteNum);
803 shakeEnergy_ += amplitude * MAX_SHAKE * 0.1;
804 if (shakeEnergy_ > MAX_SHAKE) shakeEnergy_ = MAX_SHAKE;
805 if (instType_==10 || instType_==3) ratchetPos_ += 1;
can writ eoff array.
So a Perry Cook error and not csound
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2011-01-10 18:38
Message:
Seems to be a memory problem . Running under gdb I see an illegal memory
acces in clenup in csoundDestroy.
Under valgrind I see much earier an Invalid write of size 8 in
stk::Shakers::setupNum(int) (Shakers.cpp:786)
which suggests unallocated array?
Either way looks lik eth estk interface or the stk code
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-09-01 09:14
Message:
Thanks; that does indeed shed some light. Wll investigate further.
----------------------------------------------------------------------
Comment By: Rene Djack (rene-djack)
Date: 2010-08-30 20:57
Message:
Have tested with gdb, here the result:
gdb csound
(gdb) r "/home/moi/csd/0_Files/Bugs/STK_Shakers_score.csd"
Starting program: /home/moi/csound5-f-debug/csound
"/home/moi/csd/0_Files/Bugs/STK_Shakers_score.csd"
[Thread debugging using libthread_db enabled]
time resolution is 0.416 ns
virtual_keyboard real time MIDI plugin for Csound
0dBFS level = 32768.0
Csound version 5.12 beta (float samples) Aug 30 2010
libsndfile-1.0.21
UnifiedCSD: /home/moi/csd/0_Files/Bugs/STK_Shakers_score.csd
STARTING FILE
Creating options
Creating orchestra
Creating score
orchname: /tmp/csound-UG7iFe.orc
scorename: /tmp/csound-0DEwzU.sco
RAWWAVE_PATH: /home/moi/csound5-f-debug/Opcodes/stk/rawwaves/
orch compiler:
11 (12) lines read
instr 1
Elapsed time at end of orchestra compile: real: 0.002s, CPU: 0.000s
sorting score ...
... done
Elapsed time at end of score sort: real: 0.003s, CPU: 0.000s
Csound version 5.12 beta (float samples) Aug 30 2010
displays suppressed
0dBFS level = 32768.0
orch now loaded
audio buffered in 256 sample-frame blocks
writing 1024-byte blks of floats to /home/moi/Tempo/test.wav (WAV)
SECTION 1:
new alloc for instr 1:
B 0.000 .. 0.300 T 0.300 TT 0.300 M: 5494.4
new alloc for instr 1:
B 0.300 .. 0.700 T 0.700 TT 0.700 M: 6186.5
B 0.700 .. 1.100 T 1.100 TT 1.100 M: 5512.2
B 1.100 .. 1.500 T 1.500 TT 1.500 M: 6164.7
Program received signal SIGSEGV, Segmentation fault.
0x0065a944 in insert (csound=0x804dbc0, insno=1, newevtp=0x8050240)
at Engine/insert.c:126
126 if (ip->actflg && ip->offtim < 0.0 && ip->p1 == newevtp->p[1])
{
(gdb)
(gdb) where
#0 0x0065a944 in insert (csound=0x804dbc0, insno=1, newevtp=0x8050240)
at Engine/insert.c:126
#1 0x00666abc in process_score_event (csound=0x804dbc0, evt=0x8050240,
rtEvt=0) at Engine/musmon.c:689
#2 0x00667440 in sensevents (csound=0x804dbc0) at Engine/musmon.c:874
#3 0x006eebab in csoundPerform (csound=0x804dbc0) at Top/csound.c:1459
#4 0x08049ac1 in main (argc=2, argv=0xbffff444)
at frontends/csound/csound_main.c:136
(gdb)
All other STK opcodes works well.
My system is ubuntu 10.04, csound cvs float and double.
But i have seen this "problem" since about 2 years (file CsoundVST.csd in
csound examples).
First time i use gdb, hope this help.
Cheers,
René
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-08-30 18:49
Message:
I have managed to build STK and it runs with no errors on my Linux system
Suggested that this is an Ubuntu error. From whence came your system?
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-08-30 18:02
Message:
Is thi sin Csound or in the STK system? Can you run under gdb and then
with the where command isolate the function involved? I still do not seem
to be able to build the STK code due to many C++ errors, and my knowledge
of C++ is insufficient to understand it
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-03-05 18:31
Message:
I am having trouble building the STK code, so it may be a while before I
can help
----------------------------------------------------------------------
Comment By: Rene Djack (rene-djack)
Date: 2010-03-05 17:49
Message:
I'have done:
rt-midi play -> midi record -> midi file -> score file --->score
instrument
and isolate a faulty section as below, i have changed the frequency to the
same values.
-Wfd -+rtaudio=null -+rtmidi=null --expression-opt
sr = 44100
ksmps = 16
instr 1 ; STK Shakers
asig STKShakers p4, 1.0, 1071, 22, 11, 1, 128, 100, 1, 30
out asig*100000
endin
i 1 10.0 0.45 261.626
i 1 10.3 0.29 261.626
i 1 10.7 0.48 261.626
i 1 11.1 0.27 261.626
i 1 11.5 0.63 261.626
This give seg fault every time i run it.
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2010-03-04 16:49
Message:
I have tried with score activated (without midi), but i don't success to
have any seg fault.
Seg fault appears if i play "fast" on my midi keyboard.
Below midi note 57 no problem
Higher midi note 61 no problem
Between 57 and 61 seg fault
Note 57 and 59 played at the same time give seg fault, most of the time
but not always!!!
I will record a midi file to test with score.
Thanks
René
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-03-04 16:05
Message:
Does this fail without MIDI? That is, is it the high frequency?
----------------------------------------------------------------------
Comment By: John ffitch (jpff)
Date: 2010-02-28 17:30
Message:
This is Michael's code; referred to him
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=564599&aid=2959788&group_id=81968
------------------------------------------------------------------------------
Protect Your Site and Customers from Malware Attacks
Learn about various malware tactics and how to avoid them. Understand
malware threats, the impact they can have on your business, and how you
can protect your company and customers by using code signing.
http://p.sf.net/sfu/oracle-sfdevnl
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://l |