Pd-list March 2006

pd-list@lists.iem.at

148 participants
291 discussions

Fortran-cross-C usage
by Charles Henry 20 Mar '06

20 Mar '06

Hi, list, I've run into a certain problem with accessing a Fortran subroutine from C. I thouhgt that I could just look up the symbols and make some header for the C function which calls the Fortran subroutine. My Fortran sub is called SolveT, and is found in a module called Toeplitz_Solver. I looked up the symbol in the compiled binary; it's toeplitz_solver_mp_solvet_ The makefile compiles the two things separately: $(F90) -shared -o libsft.a $(FFLAGS) Toeplitz_Solver.o verbose.o timing.o ratintnf.o precision.o $(DFFT) so I compile my Fortran code together, (using the intel fortran compiler) into a library that I'm calling libsft.a Similarly, I put the C code together in another object: gcc $(LINUXINCLUDE) $(LINUXCFLAGS) -o controlfreak~.o -c controlfreak~.c and link them: ld -export-dynamic -shared -o controlfreak~.pd_linux ... controlfreak~.o -lsft -lfftw3f -lc controlfreak~.c has the following header for using this: extern void toeplitz_solver_mp_solvet_(integer * , double *, double *, integer *, double *); I figured since Fortran passes by reference, I would need to define all my variables as pointers. The usage, looks like this, toeplitz_solver_mp_solvet_(&tempn,XT,m,&itref,&factorref); where XT and m are arrays, tempn, itref, and factorref are parameters. m is also the return array for the operation. I've checked it out with valgrind, and the error which pops up, before seg fault, is ==3735== Invalid read of size 8 ==3735== at 0x1C583F6D: toeplitz_solver_mp_solvet_ (in /usr/local/lib/pd/extra/controlfreak~/libsft.a) ==3735== by 0x1B91995F: controlfreak_tilde_perform (in /usr/local/lib/pd/extra/controlfreak~/controlfreak~.pd_linux) ==3735== by 0x80A5859: dsp_tick (d_ugen.c:304) ==3735== Address 0x0 is not stack'd, malloc'd or (recently) free'd ==3735== Process terminating with default action of signal 11 (SIGSEGV) ==3735== Access not within mapped region at address 0x0 ==3735== at 0x1C583F6D: toeplitz_solver_mp_solvet_ (in /usr/local/lib/pd/extra/controlfreak~/libsft.a) ==3735== by 0x1B91995F: controlfreak_tilde_perform (in /usr/local/lib/pd/extra/controlfreak~/controlfreak~.pd_linux) ==3735== by 0x80A5859: dsp_tick (d_ugen.c:304) So, I think I've got the right form for the function prototype. It apparently calls solvet, but I wonder what's going wrong to put that Adress 0x0 in there. Chuck

1 0

possible threading problem
by Charles Henry 20 Mar '06

20 Mar '06

I've been trying valgrind to see what's wrong with my programs in pd. Each time I load pd (not the patches) I get the following error: valgrind --show-below-main=yes --error-limit=no --leak-check=full --show-reachable=yes /usr/local/bin/pd -r 22050 Is this expected? Chuck ==6444== Memcheck, a memory error detector for x86-linux. ==6444== Copyright (C) 2002-2005, and GNU GPL'd, by Julian Seward et al. ==6444== Using valgrind-2.4.0, a program supervision framework for x86-linux. ==6444== Copyright (C) 2000-2005, and GNU GPL'd, by Julian Seward et al. ==6444== For more details, rerun with: -v ==6444== ==6444== Conditional jump or move depends on uninitialised value(s) ==6444== at 0x205BE0: strstr (in /lib/libc-2.3.5.so) ==6444== by 0x3F36E7: __pthread_initialize_minimal (in /lib/libpthread-2.3.5.so) ==6444== by 0x3F3297: (within /lib/libpthread-2.3.5.so) ==6444== by 0x3F2E7F: (within /lib/libpthread-2.3.5.so) ==6444== by 0x1B8F1DCA: call_init (in /lib/ld-2.3.5.so) ==6444== by 0x1B8F1EEC: _dl_init (in /lib/ld-2.3.5.so) ==6444== by 0x1B8E47CE: (within /lib/ld-2.3.5.so) ==6444== ==6444== Conditional jump or move depends on uninitialised value(s) ==6444== at 0x205BE4: strstr (in /lib/libc-2.3.5.so) ==6444== by 0x3F36E7: __pthread_initialize_minimal (in /lib/libpthread-2.3.5.so) ==6444== by 0x3F3297: (within /lib/libpthread-2.3.5.so) ==6444== by 0x3F2E7F: (within /lib/libpthread-2.3.5.so) ==6444== by 0x1B8F1DCA: call_init (in /lib/ld-2.3.5.so) ==6444== by 0x1B8F1EEC: _dl_init (in /lib/ld-2.3.5.so) ==6444== by 0x1B8E47CE: (within /lib/ld-2.3.5.so)

1 0

joystick calibration and two displays in GEM
by list-pd＠robmunro.net 20 Mar '06

20 Mar '06

1). I have a hercules dj console which comes up as a joystick under linux(suse 9.3) it all works fine in pd except that there is a tolerance level for the joystick (i think its part of the joystick driver) below which it returns zero which means that i get dead zones when using the jog wheels on the herc dj console. has anyone played around with joystick config in linux who can give me some pointers? 2). Is there a way to have two gemwin from the same instance of pd? i want to have one display fullscreen on my second desktop and another so i can see it along side my patch on the first desktop. (I am using windows for GEM display as i havent been able to get my ATI mobility raedon 9700 working under suse and communicating beetween computer via OSC). I had a look around on the list and i think some people were saying it was in the process of being made. I am using GEM 0.87. regards, rob munro.

3 2

enabling jack and alsamidi on Pd startup
by Claude Heiland-Allen 20 Mar '06

20 Mar '06

Hi all, Why do I have to start Pd from a script like this to get jack and alsamidi working and connected the way I want? I'm particularly talking about the pd -send "really long sequence of cryptic messages", but jack_connect and aconnect support within Pd would be even more awesome. Can the -jack and -alsamidi flags be made to actually work? Or is -send the future of Pd's command line arguments? It's also annoying that the messages to pd open the midi/audio dialog windows, or is there a way to close them via messages too? Can the ALSA MIDI port names of Pd be made unique for each instance of Pd, for example "Pure Data 0", "Pure Data 1", etc, with the same ID number as the jack instances? I'm having to aconnect Pd using port numbers, because 'aconnect "Pure Data":1 blah:0' selects only the first of the many "Pure Data"s. #!/bin/bash # start pd in the background so we can do stuff after startup pd -send "; pd audio-setapi 5 ; pd audio-dialog 0 0 0 0 0 0 0 0 0 0 0 0 2 0 0 0 48000 50 ; pd midi-setapi 1 ; pd midi-dialog 1 0 0 0 1 0 0 0 1 1 ; pd dsp 1" $* & sleep 5 # hacky delay factor to wait for pd to start and have ports # connect up to 8 pd instances to ecasound via jack jack_connect pure_data_0:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_0:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_1:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_1:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_2:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_2:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_3:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_3:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_4:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_4:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_5:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_5:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_6:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_6:output1 ecasound:record_2 >/dev/null 2>&1 jack_connect pure_data_7:output0 ecasound:record_1 >/dev/null 2>&1 jack_connect pure_data_7:output1 ecasound:record_2 >/dev/null 2>&1 # connect up to 8 pd instances to the hardware midi interface aconnect 129:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 129:0 >/dev/null 2>&1 aconnect 130:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 130:0 >/dev/null 2>&1 aconnect 131:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 131:0 >/dev/null 2>&1 aconnect 132:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 132:0 >/dev/null 2>&1 aconnect 133:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 133:0 >/dev/null 2>&1 aconnect 134:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 134:0 >/dev/null 2>&1 aconnect 135:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 135:0 >/dev/null 2>&1 aconnect 136:1 "EV-10 USB MIDI Interface":0 >/dev/null 2>&1 aconnect "EV-10 USB MIDI Interface":0 136:0 >/dev/null 2>&1 # re-join with the pd process wait $! Claude

3 2

pdoctave makefile
by Gerda Strobl 20 Mar '06

20 Mar '06

Hallo Franz, I'd like to use your pdoctave but in the iem/pdoctave/- folder I can't find a make file or some similar install instructions. Can you please tell me how to install your objects. Thank you. Lg Gerda

1 0

v4l2
by mart ameise 20 Mar '06

20 Mar '06

moinmoin, i am wondering if v4l2 is now supported by PDP? what has to be done to get it running? would be lovely if someone could send me a link to some infos if there are... thanks a lot mart

2 1

Re: [hid] crashes Re: [PD] Couldn't Create sndfiler thread: 1?
by Georg Holzmann 19 Mar '06

19 Mar '06

Hallo! >> I discoverd this behaviour all the time on various computers ... I >> only have >> one working hid binary, compiled 1 year ago, which works everywhere ... >> (but I can try to give you more information about those crashes next >> week >> ...) > > > What platform? Which kernel? Which devices? Which version of [hid]? my platform: debian - kernel 2.6.15 and I also tried it on some other debian systems ... devices: mouse and keyboard version: current cvs (and I also tried it half year ago or so ...) see e.g. also the thread: "[PD] hid segmentation fault" by Gerda Strobl, I got the same backtrace ... did you never got those crashes ? let me know which information would be usefull ... LG Georg -- Bis zu 70% Ihrer Onlinekosten sparen: GMX SmartSurfer! Kostenlos downloaden: http://www.gmx.net/de/go/smartsurfer

2 4

Getting rid of the new help browser
by Frank Barknecht 19 Mar '06

19 Mar '06

Hi, is there some kind of option that allows me to get rid of the new help browser in Pd and get the old file-manager like interface back? I really don't like the look and feel of the new browser, it doesn't allow first-letter keyboard navigation, the size of the viewport is fixed so I cannot see complete filenames etc. and thus I desparatly want the old browser back. I'm on Linux. Thanks and ciao -- Frank Barknecht _ ______footils.org_ __goto10.org__

2 5

Re: [PD] Getting rid of the new help browser
by adam armfield 19 Mar '06

19 Mar '06

i have had problems with the browser on os 10.3.9, it does work but takes a glacial age to appear (i just get the spinning beachball for a minute or so), i've taken to opening up the app with the finder and loading the help files straight from there, it's a lot quicker this is on a 6 year old ibook though ;-) all the best adam ___________________________________________________________ Yahoo! Messenger - NEW crystal clear PC to PC calling worldwide with voicemail http://uk.messenger.yahoo.com

2 1

The automatic connection problem
by Chuckk Hubbard 19 Mar '06

19 Mar '06

When I asked about what messages can be sent to windows, and I read over the documentation given in response, it occurred to me the [mouse( message could possibly eliminate the problem of having to know exactly which number each object was in order to make connections. I threw together this little patch to illustrate. If your only concern was making it not overlap existing onscreen objects, you could place the objects like at -500 -500 or something. I'm not sure how useful this would be, since you would instead have to know the exact position. But if you make the connections at the same time as placing the objects, you can figure it out. I added 1 to the x value and 16 to the y value to get to an outlet. Maybe this will give someone some useful ideas. -Chuckk -- "It is not when truth is dirty, but when it is shallow, that the lover of knowledge is reluctant to step into its waters." -Friedrich Nietzsche, "Thus Spoke Zarathustra"

1 0

← Newer
1
...
9
10
11
12
13
14
15
...
30
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

Pd-list March 2006