Heya!
Not sure if anyone of you already read this announcement about my
mutex profiler "mutrace" I wrote a couple of days back:
http://0pointer.de/blog/projects/mutrace.html
Since I wrote that blog story I have added a couple of features to
mutrace that are particularly relevant when developing real-time
applications. I'd like to draw your attention to these, since this
might be of some interest for the audio community.
"mutrace -r" can be used to track down any mutex usage in real-time
threads. As everyone probably knows it's a good idea to abstain from
mutex usage entirely in RT threads. However, actually doing that is
not always realistic, and sometimes the more complex your project gets
the harder it becomes actually making sure that no hidden mutex usage
creeps into your program. Too verify lock-freeness or track down the
culprits you may use "mutrace -r". It will tell you not only if am RT
thread used a mutex, it will also tell you how often it was contended
and how often it changed ownership (i.e. how big the risk of
contention is even if it didn't contend in the specific run.) In
addition it will show you the mutex protocol, so if you are using a
mutex from an RT thread and really cannot do without it, you can at
least make sure it is a PRIO_INHERIT mutex,
And there's also now a second tool called "matrace" in the tarball
that tracks down malloc() usage from RT threads. It's a very simple
tool, which just catches the memory allocations/frees, checks the
scheduling and complains if necessary.
With these two tools two of the biggest donts in RT programming may be
tracked down: use of mutexes and use of malloc.
How does the output look like? As an example I simply ran jackd through
mutrace -r and matrace, for 10s without doing anything else. The
result for mutrace you may find here:
http://0pointer.de/public/mutrace-jackd.txt
(The interesting table is at the end, so scroll down)
Not sure what to make of this, as I don't know the Jack internals too
well. However mutrace reveals that at least one mutex was used from
an RT thread and even contended. Moreover that mutex was not
PRIO_INHERIT.
And here's matrace:
http://0pointer.de/public/matrace-jackd.txt
Seems free() is called a couple of times from an RT thread.
These might, or might not be problems. The tools can just inform you
what is done. If there's something to fix is up to the eye of the
beholder.
Get the tool here:
http://git.0pointer.de/?p=mutrace.git
Have fun,
Lennart
--
Lennart Poettering Red Hat, Inc.
lennart [at] poettering [dot] net
http://0pointer.net/lennart/ GnuPG 0x1A015CC4
Does anyone know anything about this change in alsa:
warning: snd_pcm_sw_params_set_xfer_align is deprecated (declared
at /usr/include/alsa/pcm.h:1115)
I just wanted to check background before fixing the code.
Regards
Victor
The first milestone is reached and result is a tarball that brave souls
may want to download and try. It contains implementation of JACK
multiconfig functionality. JACK server settings can be saved as part of a
studio. Then, loading studio will cause JACK settings stored as part of
the studio to be restored.
Build will produce three operational components:
* ladishd - The daemon, a D-Bus service
* gladish - GTK GUI interface
* ladish_control - Command-line interface
In the tarball you will also find bundled suitable (latest and gratest)
flowcanvas and LADI Tools.
Download:
http://ladish.org/download/ladish-0.1.tar.bz2http://ladish.org/download/ladish-0.1.tar.bz2.sig
Homepage: http://ladish.org/
Roadmap: http://ladish.org/roadmap
---------------------------------------------------------------------
LADI Session Handler or simply ladish is a session management system
for JACK applications on GNU/Linux. Its aim is to allow you to have
many different audio programs running at once, to save their setup,
close them down and then easily reload the setup at some other
time. ladish doesn't deal with any kind of audio or MIDI data itself;
it just runs programs, deals with saving/loading (arbitrary) data and
connects JACK ports together. It can also be used to move entire
sessions between computers, or post sessions on the Internet for
download.
ladish has GUI frontend, gladish, based on lpatchage (LADI Patchage)
and the ladish_control command line app for headless operation. LADI
Tools is set of apps that interface with ladish, JACK server and
a2jmidid
ladish requires D-Bus and JACK compiled with D-Bus support.
LADI Session Handler is rewrite of LASH.
Project goals:
* Save and restore sets of JACK (audio and MIDI) enabled
applications.
* Provide JACK clients with virtual hardware ports, so projects can
be transfered (or backups restored) between computers running
different hardware and backups.
* Don't require session handling library to be used. There is no need
of such library for restoring connections between JACK clients.
* Flow canvas based GUI. Positions of elements on the canvas are
saved/restored.
* Allow clients to use external storage to save its state. This
includes storing internal state to non-filesystem place like memory
of a hardware synth. This also includes storing client internal
state (client project data) in a way that is not directly bound to
ladish project.
* Import/export operations, as opposed to save/load. Save/load
operate in current system and may cause saving data outside of
project itself (external storage). Import/export uses/produces
"tarball" suitable for transferring session data over network to
other computer or storing it in a backup archive.
* Hierarchical or tag-based organization of projects.
* List of JACK applications. Applications are always started through
ladish to have restored runtime environment closer to one existed
before project save.
* Distributed studio - network connected computers. Netjack
configuration is part of the studio and thus is saved/restored.
* Collaborate with the X11 window manager so window properties like
window position, virtual desktop and screen (multimonitor) are
saved/restored.
--
Nedko Arnaudov <GnuPG KeyID: DE1716B0>
Greetings,
Thank you for all your great feedback on the last two betas!
phasex-0.12.0-beta4 addresses all of the concerns brought up so far:
Velocity: Many of you have noticed that velocity and aftertouch
were completely unsupported. Now, velocity (with aftertouch) can be
mapped to directly to oscillators and LFOs (by setting the source),
used as a modulation source (adjust volume w/ AM or pitch w/ FM), or
mapped to the filter in the filter-lfo section.
GUI: Some of you mentioned that the colors were too dark with too
little contrast, so now there are four GTK theme options: Dark
(original purple background), Light (orange background), System (use
the system GTK theme), and Custom (choose any gtkrc file for your
theme). There have also been issues with getting phasex to fit on
small screens (usually netbooks). The knob images have been trimmed
down vertically (just blank pixels), and the padding between widgets
has been almost completely cut out. The font can be selected in the
preferences. Additionally, a true fullscreen mode has been added.
It is now possible to fit phasex into an 800x600 desktop.
Atom processors: Compiler optimization flags for the Intel Atom
processors have been added to the build system. Run './configure
--enable-arch=atom' to build for the Atom. To force 32- or 64-bit
builds, use 'atom32' or 'atom64'.
The rest is just small tweaks and bugfixes, such as fixing the
segfault on quit issue, fixing all the compiler warnings, minor
updates to the build system, new menu items, etc.
As I don't have access to a netbook right now, please let me know
how it works out with the Atom or other low-power CPU, or on any
machine with a screen smaller than 1024x768. Of course, feedback
from the rest of you is highly welcome, too ;-}
Visit http://sysex.net/phasex/beta for source tarball, Fedora 11
RPMS, or Fedora 8 RPMS.
Cheers,
--ww
Hallo,
This is a serious question for Linux Audio and in general multi media
applications developers.
I am in the program committee of the conference part of a Linux related
even in france (called Salon Linux), and we are discussing about the
2010 conference program.
I may be in the condition to have a Audio or multi media subject added,
but i can do that only if i have serious evidence of a user base or
a developer base in france. Most of the subjects are around enterprise
and industrial development.
The conference is strictly french-speaking, and the access is not
for free (even if the organisation is stricly on a volunteer base).
So, any French developers here, and any evidence of a French user base ?
Maurizio De Cecco
--
Music: http://www.myspace.com/mauriziodececco
Blog: http://maurizio.dececco.name/
Software: http://www.jmax-phoenix.org/
I'm looking for some real time audio analysis software similar to the
melodic range spectrogram in sonic visualizer. Instead of visual output
though, I'm looking for midi output (this can be rough, i'm not going to be
generating audio from it, just trying to check if said notes were played).
I'm not sure if this is possible with clam, I've used it a little, but not
much.
The ideal solution will be multiplatform.
Thanks for any leads,
Nathanael
Hi everyone, I'm wondering what most linux audio developers think about
pyqt vs wxpy for writing audio app guis now that qt is gpl'd.
Specifically I'm interested in tightly controlling the timing of event
loop ( ie making some accurate external clock source like the csound
engine be the time boss ) and making decent faders widgets.
thanks!
iain
Hey guys,
I had a performance the other day with a nicely underpowered PC (500 MHz).
One man show. Everything being run through JACK (guitar + vocals), using
sooperlooper, and jack rack plus some nice TAP effects (for vocals). Set
up for 10ms latencies.
Not long into it I looked at the computer and saw the hard drive light was
very bright. Great. My computer is running updatedb... _now_??? This is
bad. Since I was doing something slightly different... I really couldn't
afford any glitches. So, I was a little nervous.
But I trudged along. No crashes. No xruns. No clicks. No problems.
So, I'm taking a moment to say: Thanks, everybody.
Peace,
Gabriel
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
As I'm new to smart pointers and memory management in C++ I have a short
question on this topic.
I'm now using the following code to traverse the devices and their ports:
snd_seq_client_info_t *clientInfo = 0;
snd_seq_port_info_t *portInfo = 0;
snd_seq_client_info_malloc(&clientInfo);
snd_seq_port_info_malloc(&portInfo);
shared_ptr<snd_seq_client_info_t> clientInfoMemoryHandler(clientInfo,
snd_seq_client_info_free);
shared_ptr<snd_seq_port_info_t> portInfoMemoryHandler(portInfo,
snd_seq_port_info_free);
.
.
.
snd_seq_port_info_set_client( portInfo,
snd_seq_client_info_get_client(clientInfo ) );
snd_seq_port_info_set_port( portInfo, -1 );
while ( snd_seq_query_next_port( seq, portInfo ) >= 0 ) {
string portName = snd_seq_port_info_get_name( portInfo );
unsigned int capability = snd_seq_port_info_get_capability( portInfo );
.
.
.
Well the cleanUp methods are called at block-leaving.
I'm only a bit curious because after giving portInfo and clientInfo to
the shared_ptr for cleanup management I have to use them for all the
alsa queries.
But since I'm not using malloc and free in these queries everything is
ok, isn't it?
Thanks for your help!
Christian
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iEYEARECAAYFAkqiFXwACgkQVC26eJ+o0+3H6wCcCb49zyCeNFpyOksEpKA0uXL7
5NMAoJNXUhpWmMNWGU80RoqoqUp52dSf
=IcCV
-----END PGP SIGNATURE-----
