Bugs item #2637417, was opened at 2009-02-25 10:48
Message generated for change (Comment added) made by eighthave
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2637417&group_…
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: pd-extended
>Group: v0.41
>Status: Pending
Resolution: None
Priority: 5
Private: No
Submitted By: boonier (boonier)
>Assigned to: Nobody/Anonymous (nobody)
>Summary: When connecting to Jack mode, Pd freezes
Initial Comment:
I've noticed a couple of bug with regard to Jack and the new test build of Pd extended:
1) When connecting to Jack mode, Pd freezes and has to be force quit.
2) Another problem I've noticed is with current release of 0.40.3 extended, is when Pd is not computing audio, cpu usage shoots up around 30% .
This appears to happen when compute audio is first activated, then switched off. Switching on audio and cpu returns to nominal amount. This problem does not affect actual usage.
Specs problems appear on:
os 10.5.6 Intel 2.4 iMac
os 10.4.11 PPC 1 ghz
Jack OSX 0.81
thanks
----------------------------------------------------------------------
>Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-28 11:32
Message:
since there is another bug report about the CPU usage
(https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2791638&group_…),
I renamed this one to be about the first part. Can you give steps to
reproduce the freeze? I haven't seen this using Mac OS X/Intel 10.5.(6|7)
and JackOSX 0.81 using Pd-extended 0.41.4 release candidates.
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-02-25 17:27
Message:
THe latest version - 0.81
----------------------------------------------------------------------
Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-02-25 11:01
Message:
which version of jackosx are you using?
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2637417&group_…
Bugs item #2791638, was opened at 2009-05-14 06:16
Message generated for change (Settings changed) made by eighthave
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2791638&group_…
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: pd-extended
>Group: v0.41
>Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: boonier (boonier)
>Assigned to: IOhannes m zmlnig (zmoelnig)
Summary: Jack OSX cpu load
Initial Comment:
In Jack mode, Blank empty patch
Toggling Compute Audio on and then off results in cpu load jumping up to ~25% from a nominal 1%. Turn Compute Audio back on, and cpu drops back to nominal 1%.
Tested on 2.4ghz iMac , OS 10.5.6. Pd-0.41.4-extended-rc1
----------------------------------------------------------------------
Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-28 11:25
Message:
------------------------------------------------------------------
1. start jack server and attach Pd to it
jackdmp 2.7%
pd 1.7%
- 21.6%, ml_set_interrupts_enabled, mach_kernel
- 7.6%, __nanotime, libSystem.B.dylib
- 6.2%, __memcpy, libSystem.B.dylib
5.6%, lo_unix_scall, mach_kernel
------------------------------------------------------------------
2. then turn on Compute Audio
jackdmp 4.2%
pd 2.1%
- 20.0%, ml_set_interrupts_enabled, mach_kernel
7.1%, lo_unix_scall, mach_kernel
- 6.3%, __nanotime, libSystem.B.dylib
- 4.0%, __memcpy, libSystem.B.dylib
- 4.0%, __gettimeofday, libSystem.B.dylib
- 2.8%, select$DARWIN_EXTSN$NOCANCEL, libSystem.B.dylib
2.2%, lo_alltraps, mach_kernel
2.0%, lo_mach_scall, mach_kernel
------------------------------------------------------------------
3. turn off Compute Audio after its on
jackdmp 4.2%
pd 7.5-10.5%
+ 81.4%, __bzero, libSystem.B.dylib
| + 80.9%, process, pd
| | + 80.9%, Jack::JackClient::Execute(), Jackmp
| | | + 80.9%, Jack::JackPosixThread::ThreadHandler(void*), Jackmp
| | | | + 80.9%, _pthread_start, libSystem.B.dylib
| | | | | 80.9%, thread_start, libSystem.B.dylib
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-05-18 16:04
Message:
Ah, sorry, I misunderstood, I also get similar behavior using Pd version
0.41.4-extended-rc2. Using Jack, I get ~3% CPU when audio is ON and ~8%
CPU when audio is OFF.
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-05-18 05:56
Message:
hi - sorry for delay
hmm i don't understand why when Pd is runing in Jack mode and you switch
audio off, why it would revert to a Portaudio state? I that what you are
saying?
----------------------------------------------------------------------
Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-14 12:24
Message:
For me, Pd without Jack, i.e. just using coreaudio/portaudio, it will use
25% of my CPU just idling. That's a known issue. With Jack, it goes down
to 5-7% at idle. So my guess is that the first time the 25% spike is
caused by the same think.
Is this causing any problems for you?
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2791638&group_…
Bugs item #2791638, was opened at 2009-05-14 06:16
Message generated for change (Comment added) made by eighthave
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2791638&group_…
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: None
Group: v0.42
Status: Pending
Resolution: None
Priority: 5
Private: No
Submitted By: boonier (boonier)
Assigned to: Nobody/Anonymous (nobody)
Summary: Jack OSX cpu load
Initial Comment:
In Jack mode, Blank empty patch
Toggling Compute Audio on and then off results in cpu load jumping up to ~25% from a nominal 1%. Turn Compute Audio back on, and cpu drops back to nominal 1%.
Tested on 2.4ghz iMac , OS 10.5.6. Pd-0.41.4-extended-rc1
----------------------------------------------------------------------
>Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-28 11:25
Message:
------------------------------------------------------------------
1. start jack server and attach Pd to it
jackdmp 2.7%
pd 1.7%
- 21.6%, ml_set_interrupts_enabled, mach_kernel
- 7.6%, __nanotime, libSystem.B.dylib
- 6.2%, __memcpy, libSystem.B.dylib
5.6%, lo_unix_scall, mach_kernel
------------------------------------------------------------------
2. then turn on Compute Audio
jackdmp 4.2%
pd 2.1%
- 20.0%, ml_set_interrupts_enabled, mach_kernel
7.1%, lo_unix_scall, mach_kernel
- 6.3%, __nanotime, libSystem.B.dylib
- 4.0%, __memcpy, libSystem.B.dylib
- 4.0%, __gettimeofday, libSystem.B.dylib
- 2.8%, select$DARWIN_EXTSN$NOCANCEL, libSystem.B.dylib
2.2%, lo_alltraps, mach_kernel
2.0%, lo_mach_scall, mach_kernel
------------------------------------------------------------------
3. turn off Compute Audio after its on
jackdmp 4.2%
pd 7.5-10.5%
+ 81.4%, __bzero, libSystem.B.dylib
| + 80.9%, process, pd
| | + 80.9%, Jack::JackClient::Execute(), Jackmp
| | | + 80.9%, Jack::JackPosixThread::ThreadHandler(void*), Jackmp
| | | | + 80.9%, _pthread_start, libSystem.B.dylib
| | | | | 80.9%, thread_start, libSystem.B.dylib
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-05-18 16:04
Message:
Ah, sorry, I misunderstood, I also get similar behavior using Pd version
0.41.4-extended-rc2. Using Jack, I get ~3% CPU when audio is ON and ~8%
CPU when audio is OFF.
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-05-18 05:56
Message:
hi - sorry for delay
hmm i don't understand why when Pd is runing in Jack mode and you switch
audio off, why it would revert to a Portaudio state? I that what you are
saying?
----------------------------------------------------------------------
Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-14 12:24
Message:
For me, Pd without Jack, i.e. just using coreaudio/portaudio, it will use
25% of my CPU just idling. That's a known issue. With Jack, it goes down
to 5-7% at idle. So my guess is that the first time the 25% spike is
caused by the same think.
Is this causing any problems for you?
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=2791638&group_…
--- On Thu, 5/28/09, Sergei Steshenko <sergstesh(a)yahoo.com> wrote:
> From: Sergei Steshenko <sergstesh(a)yahoo.com>
> Subject: Re: [PD-dev] ffmpeg with gcc 3.4
> To: "pd-dev List" <pd-dev(a)iem.at>, "Hans-Christoph Steiner" <hans(a)at.or.at>
> Date: Thursday, May 28, 2009, 1:44 AM
>
>
> --- On Tue, 5/19/09, Hans-Christoph Steiner <hans(a)at.or.at>
> wrote:
>
> > From: Hans-Christoph Steiner <hans(a)at.or.at>
> > Subject: [PD-dev] ffmpeg with gcc 3.4
> > To: "pd-dev List" <pd-dev(a)iem.at>
> > Date: Tuesday, May 19, 2009, 4:04 PM
> >
> > There is hope, we can still use MSYS's default gcc
> 3.4.5
> > for building ffmpeg, making the MinGW build setup
> much
> > easier. Yes, its in japanese, but we can all
> understand the
> > universal language of "command line".
> >
> > http://feather.cocolog-nifty.com/weblog/2009/04/msys-mingw-gcc3.html
> >
> > Then again, it might be worthwhile to try the gcc 4
> called
> > "Testing"
> >
> > http://sourceforge.net/project/showfiles.php?group_id=2435
> >
> > .hc
> >
> ----------------------------------------------------------------------------
> >
> > 'You people have such restrictive dress for women,’
> she
> > said, hobbling away in three inch heels and panty hose
> to
> > finish out another pink-collar temp pool day. -
> > “Hijab Scene #2", by Mohja Kahf
> >
> >
>
> There are much more recent than the official MinGW 'gcc'
> packages:
>
> http://www.tdragon.net/recentgcc/
>
> There is another also one - cant find it at the moment.
>
>
> Regards,
> Â Sergei.
>
Found it:
http://www.thisiscool.com/gcc_mingw.htm
.
Regards,
Sergei.
--- On Tue, 5/19/09, Hans-Christoph Steiner <hans(a)at.or.at> wrote:
> From: Hans-Christoph Steiner <hans(a)at.or.at>
> Subject: [PD-dev] ffmpeg with gcc 3.4
> To: "pd-dev List" <pd-dev(a)iem.at>
> Date: Tuesday, May 19, 2009, 4:04 PM
>
> There is hope, we can still use MSYS's default gcc 3.4.5
> for building ffmpeg, making the MinGW build setup much
> easier. Yes, its in japanese, but we can all understand the
> universal language of "command line".
>
> http://feather.cocolog-nifty.com/weblog/2009/04/msys-mingw-gcc3.html
>
> Then again, it might be worthwhile to try the gcc 4 called
> "Testing"
>
> http://sourceforge.net/project/showfiles.php?group_id=2435
>
> .hc
> ----------------------------------------------------------------------------
>
> 'You people have such restrictive dress for women,’ she
> said, hobbling away in three inch heels and panty hose to
> finish out another pink-collar temp pool day. -
> “Hijab Scene #2", by Mohja Kahf
>
>
There are much more recent than the official MinGW 'gcc' packages:
http://www.tdragon.net/recentgcc/
There is another also one - cant find it at the moment.
Regards,
Sergei.
--- On Thu, 5/21/09, Hans-Christoph Steiner <hans(a)at.or.at> wrote:
> From: Hans-Christoph Steiner <hans(a)at.or.at>
> Subject: [PD-dev] liboil - lib of CPU optimizations
> To: "pddev List" <pd-dev(a)iem.at>
> Date: Thursday, May 21, 2009, 11:55 AM
>
> Anyone ever looked at liboil? "Liboil is a library of
> simple functions that are optimized for various CPUs."Â
> I just noticed that VLC is using it. It sounds like it
> could be quite useful for Pd.
>
> http://liboil.freedesktop.org/wiki/
>
> .hc
>
>
>
> ----------------------------------------------------------------------------
>
> Programs should be written for people to read, and only
> incidentally for machines to execute.
> - from Structure and Interpretation of Computer Programs
>
For starters, it's part of gtk+ bundle.
Regards,
Sergei.
--- On Tue, 5/26/09, Hans-Christoph Steiner <hans(a)eds.org> wrote:
> From: Hans-Christoph Steiner <hans(a)eds.org>
> Subject: [PD-dev] dlopen() for windows
> To: "pddev List" <pd-dev(a)iem.at>
> Date: Tuesday, May 26, 2009, 12:37 PM
>
> August just brought this potentially handy piece of
> software to my attention:
>
> http://code.google.com/p/dlfcn-win32/
>
> Basically, it allows you to use dlopen() to access the
> native Win32 calls. Or in their words:
> "This library implements a wrapper for dlfcn, as specified
> in POSIX and SUS, around the dynamic link library functions
> found in the Windows API."
>
> We are including it into the MinGW build system, so
> potentially, Pd code could use only dlopen() and work on all
> platforms.
>
> .hc
>
> ----------------------------------------------------------------------------
>
> Access to computers should be unlimited and total. -
> the hacker ethic
>
>
>
'libtool' implements a nice interface for dynamic loading, and the
interface is cross-platform.
Regards,
Sergei.
Bugs item #1982664, was opened at 2008-06-02 17:14
Message generated for change (Comment added) made by eighthave
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1982664&group_…
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: externals
Group: None
>Status: Closed
>Resolution: Fixed
Priority: 7
Private: No
Submitted By: Hans-Christoph Steiner (eighthave)
>Assigned to: Hans-Christoph Steiner (eighthave)
Summary: vbap freezes Pd when it receives a large float value
Initial Comment:
On Mac OS X 10.4.11/Intel running 0.40.3-extended-20080602, sending an 'inf' float value, then banging vbap causes pd to freeze. See attached patch for example
----------------------------------------------------------------------
>Comment By: Hans-Christoph Steiner (eighthave)
Date: 2009-05-27 12:54
Message:
The code was using a while statement with a -= instead of a mod, so huge
values would peg the CPU while it was running thru that loop. Fixed with
this commit:
http://pure-data.svn.sourceforge.net/viewvc/pure-data?view=rev&revision=115…
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1982664&group_…
August just brought this potentially handy piece of software to my
attention:
http://code.google.com/p/dlfcn-win32/
Basically, it allows you to use dlopen() to access the native Win32
calls. Or in their words:
"This library implements a wrapper for dlfcn, as specified in POSIX
and SUS, around the dynamic link library functions found in the
Windows API."
We are including it into the MinGW build system, so potentially, Pd
code could use only dlopen() and work on all platforms.
.hc
----------------------------------------------------------------------------
Access to computers should be unlimited and total. - the hacker ethic
Bugs item #1356455, was opened at 2005-11-14 10:34
Message generated for change (Comment added) made by nobody
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1356455&group_…
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: pd-devel
Group: None
Status: Open
Resolution: None
Priority: 7
Private: No
Submitted By: Jamie Bullock (postlude)
Assigned to: Miller Puckette (millerpuckette)
Summary: [midiout] not working for ALSA
Initial Comment:
If [midiout] is used with ALSA on Linux, no output
is generated by the object.
ALSA version: 1.0.9rc1
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2009-05-24 23:35
Message:
I don't have output with midiout in 0.40.3-extended.
Now is this a bug or there a way around for me?
----------------------------------------------------------------------
Comment By: Nobody/Anonymous (nobody)
Date: 2005-11-15 22:20
Message:
Logged In: NO
thanks!
tested here too and it works.
----------------------------------------------------------------------
Comment By: Jamie Bullock (postlude)
Date: 2005-11-15 10:43
Message:
Logged In: YES
user_id=1361897
Tested [midiout] sysex output with external hardware; works
fine.
----------------------------------------------------------------------
Comment By: IOhannes m zmlnig (zmoelnig)
Date: 2005-11-15 09:20
Message:
Logged In: YES
user_id=564396
as gnter has noticed, there is a memory leak in
sys_alsa_putmidibyte()
attached is a patch against s_midi_alsa.c (in pd-0.39-1-CVS)
which should fix this
----------------------------------------------------------------------
Comment By: IOhannes m zmlnig (zmoelnig)
Date: 2005-11-14 17:33
Message:
Logged In: YES
user_id=564396
the putbytes of the alsa-implementation does not get called.
the attached patch (s_midi.c, x_midi.c; pd=0.39-1-CVS)
should fix this.
the current alsa-implementation of [midiout] uses a kind of
sysex-wrapper to get the raw-data through the alsa-seq
interface; later think whether this is really a good idea,
in the meantime it works.
btw, since i changed the subject to not reflect the version
number of pd, i give it here: confirmed wrong behaviour in
pd-0.39-1 (both vanilla and devel); all pd-versions with
alsamidi until now should be affected
----------------------------------------------------------------------
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=478070&aid=1356455&group_…