I think I found the bug: in csoundapi_tilde.c
lines 84 & 263

static t_int *csoundapi_perform(int *w)

should be

static t_int *csoundapi_perform(t_int *w)

and line

int     i, n, end = x->end, run = x->run;

should be

t_int     i, n, end = x->end, run = x->run;

I am positive this is the bug (it is A bug in
64bit CPUs, so I expect it to be the cause of
your crashes). I can't change in CVS or test it
now, will do when I am back to work.

Chuck, can you test it for me?

thanks

Victor








>
> OK, we'll fix it, don't worry. But it will be when I'm
> back to work. (I need to get myself a 64bit CPU...)
> >
> >
> > The command works okay on my patch, it sits and waits
> > for input.  The example patch does crash Pd.  Could be a
> > 64-bit thing, huh.
> >
> > -Chuckk
> >
> > On 8/1/07, Victor Lazzarini 
> > wrote: >
> > > what about the csoundapi~ example patch?
> > > Your CSD is quite simple and should not crash
> > > anything. Perhaps it's a 64bit issue?
> > >
> > > Is the csound command working OK?
> > >
> > > Victor
> > >
> > > >
> > > >
> > > > I just checked, and taking out f0 didn't prevent the
> > > > crash.  I tried without that table and with it as
> > > > f2. The patch is a [csoundapi~ apitest.csd] object
> > > > with a [set level( message and a [control level $1(
> > > > message going into it, and a dac~ not connected to
> > > anything. >
> > > > The fact that it said "connection reset by peer"
> > > > made me wonder if it was Pd, since it uses a network
> > > > connection between the GUI and the engine.  Now that
> > > > I think of it, I'm using a script where I alter
> > > > their realtime priorities after starting Pd... Just
> > > > tried it only running "pd" from command line, same
> > > > thing.  This time I saw "alsa_pcm" xrun... could it
> > > > be Pd is violating Jack's realtime mandate? Well I
> > > > just tried it with Jack down and the -alsa flag (it
> > > > can't open audio), it still crashes when I start
> > > > processing audio with that patch. There's a Csound
> > > > message first: segmentation fault; and what I think
> > > > is a Pd message, socket receive error: connection
> > > reset by peer. The .csd acts fine from the command
> > > > line. > I am on 64-bit Linux.  Sending csoundapi~
> > > > output to lop~ crashed too, with lop~ not connected
> > > >  to anything. I wonder if it has to do with the
> > > > realtime flag.  I wouldn't think the csoundapi~ with
> > > > that particular .csd would be enough to kill it, but
> > > maybe it is? -Chuckk >
> > > > On 8/1/07, Victor Lazzarini
> > > >  wrote: >
> > > > > I have not yet changed the version on CVS; if it
> > > > > is still crashing, then it's something else I need
> > > > > to find out. If you can isolate what is causing
> > > > > the crash as much as possible it will be quicker
> > > > > for me to debug it. Are you doing anything special
> > > > > in that patch? I don't have PD here so I can't
> > > > open it. >
> > > > > Are you on a 64-bit linux? I am wondering if this
> > > > > is related. What happens if you connect csoundapi~
> > > > > to something else that is not a dac~? I think PD
> > > > > might not call the perform method of an object
> > > > > that is not connected anywhere, that's why it does
> > > > not crash. >
> > > > > Victor
> > > > >
> > > > > Victor
> > > > >
> > > > > >
> > > > > >
> > > > > > Hi Victor.
> > > > > > I tried compiling last night from CVS after
> > > > > > reading this, it still crashes. It tells me it
> > > > > > is csoundapi~ 1.01, using the Csound 5.06.2 API
> > > > > > - is that the right one?  I deleted all but what
> > > > > > I compiled last night to be sure Pd wasn't
> > > > > > pulling it from somewhere else.  I believe I
> > > > > > zeroed everything in SConstruct except the
> > > > > > 64-bit word option and the csoundapi~ option...
> > > > > do I need something else enabled? No rush, I'll
> > > > > > keep trying. > -Chuckk
> > > > > >
> > > > > > On 7/31/07, Victor Lazzarini
> > > > > >  wrote: >
> > > > > > > There is a bug I identified a couple of weeks
> > > > > > > ago with the error messaging system, which
> > > > > > > used to crash PD; I fixed it in CVS and it
> > > > > > > will be there in the next release. I think if
> > > > > > > you do not use the message output (message 0)
> > > > > > > it might not crash.
> > > > > > >
> > > > > > > I think this is probably what is going wrong,
> > > > > > > although I can't check your patch (I am not at
> > > > > > the office). >
> > > > > > > Victor
> > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Hi.
> > > > > > > > I create a simple patch with a dac~ and a
> > > > > > > > csoundapi~ and connect the left
> > > > > > > > outlet/inlet. When I turn on Pd audio
> > > > > > > > processing, Pd disappears and I see some
> > > > > > > > messages (using the -stderr flag). I have
> > > > > > > > both jackd and pd set up for realtime, and
> > > > > > > > at pretty high priorities from the scripts I
> > > > > > > > use to start them.  My audio IRQ is higher
> > > > > > > than jack, which is higher than pd , which is
> > > > > > > > higher than pd-gui.  Pd works fine for other
> > > > > > > > patches. > I compiled Csound5.06 on Linux,
> > > > > > > > once with a bunch of options and once with
> > > > > > > just csoundapi~, replacing the first one. >
> > > > > > > > Does this look like something for the Pd
> > > > > > > > list? Csound is saying segmentation fault...
> > > > > > > >  My csd file is at
> > > > > > > > http://www.badmuthahubbard.com/apitest.csd
> > > > > > > > Below is my Pd output while loading the .csd
> > > > > > > > first and then turning on audio.  Btw, it
> > > > > > > > doesn't crash if I just put the csoundapi~
> > > > > > > and dac~ objects in and turn on audio; only
> > > > > > > > when they are connected. > Thanks.
> > > > > > > > -Chuckk
> > > > > > > >
> > > > > > > > -----------------------------------------
> > > > > > > > csoundapi~ 1.01
> > > > > > > >  A PD csound class using the Csound 5.06.2
> > > > > > > > API (c) V Lazzarini, 2005-2007
> > > > > > > >
> > > > > > > > Localisation of messages is disabled, using
> > > > > > > > default language. time resolution is
> > > > > > > > 1000.000 ns /home/chuckk/apitest.csd
> > > > > > > > /home/chuckk/apitest.csd
> > > > > > > > 0dBFS level = 32768.0
> > > > > > > > Csound version 5.06 beta (float samples) Jul
> > > > > > > > 18 2007 libsndfile-1.0.17
> > > > > > > > UnifiedCSD:  /home/chuckk/apitest.csd
> > > > > > > > STARTING FILE
> > > > > > > > Creating orchestra
> > > > > > > > Creating score
> > > > > > > > orchname:  /tmp/filex9OWtg.orc
> > > > > > > > scorename: /tmp/filejlUP5v.sco
> > > > > > > > orch compiler:
> > > > > > > > 17 lines read
> > > > > > > >
> > > > > > > > instr
> > > > > > > > 11
> > > > > > > >
> > > > > > > > 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 5.06 beta
> > > > > > > > (float samples) Jul 18 2007 displays
> > > > > > > > suppressed 0dBFS level = 32768.0
> > > > > > > > orch now loaded
> > > > > > > > audio buffered in 256 sample-frame blocks
> > > > > > > > SECTION 1:
> > > > > > > > Csound tidy up: Segmentation fault
> > > > > > > > inactive allocs returned to freespace
> > > > > > > > end of score.              overall amps
> > > > > > > >       0.
> > > > > > > >       0.
> > > > > > > >
> > > > > > > >            overall samples out of range
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > 0 errors in performance
> > > > > > > > Elapsed time at end of performance: real:
> > > > > > > > 11.255s, CPU: 0.410s Removing temporary file
> > > > > > > > /tmp/file6b2OHL.srt ... Removing temporary
> > > > > > > > file /tmp/filejlUP5v.sco ... Removing
> > > > > > > > temporary file /tmp/filex9OWtg.orc ...
> > > > > > > > socket receive error: Connection reset by
> > > > > > > > peer (104) subgraph starting at pure_data_0
> > > > > > > timed out (subgraph_wait_fd=7, status = 0,
> > > > > > > state = Triggered) > >
> > > > > > > >
> > > > > > > >
> > > > > > > > --
> > > > > > > > http://www.badmuthahubbard.com
> > > > > > > >
> > > > > > > --
> > > > > > > Send bugs reports to this list.
> > > > > > > To unsubscribe, send email to
> > > > > > csound-unsubscribe@lists.bath.ac.uk >
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > http://www.badmuthahubbard.com
> > > > > >
> > > > > --
> > > > > Send bugs reports to this list.
> > > > > To unsubscribe, send email to
> > > > csound-unsubscribe@lists.bath.ac.uk >
> > > >
> > > >
> > > >
> > > > --
> > > > http://www.badmuthahubbard.com
> > > >
> > > --
> > > Send bugs reports to this list.
> > > To unsubscribe, send email to
> > csound-unsubscribe@lists.bath.ac.uk >
> >
> >
> >
> > --
> > http://www.badmuthahubbard.com
> >
> --
> Send bugs reports to this list.
> To unsubscribe, send email to
> csound-unsubscribe@lists.bath.ac.uk