juce on amd64 X2

[patrickkidd]

yeah, I still get the broken pipe error after that. Sorry, I'm don't know the alsa api that well.

[jules]

I'm a bit stuck then. Which line causes the broken pipe error?

[patrickkidd]

I'm not sure what happened, but the code suddenly seems to work. We should assume that it was something on my end until then. I still get the broken pipe line, but the audio is very scratchy in the demo and our app. Also, the demo crashed at one point.

I don't have the time at the moment to debug it, but at least I gave you some runtime info for linux on amd64 X2. I'll keep you informed...

[patrickkidd]

the demo from juce 1.39 doesn't run. I'm getting that same backtrace as before:

Code: Select all

0x00000000005624ca in juce::atomicIncrementAndReturn (variable=@0x7fff93e7e2dc)
    at juce_Atomic.h:83
83                  : "cc", "memory");


[jules]

I'm stuck for ideas. Strange that it used to work, because I don't think I changed anything in there, and it's pretty straightforward code.

[patrickkidd]

you know jack support would solve this problem, and would get it working across platforms very easily...

[patrickkidd]

sorry, I meant that jack support would take care of the audio device not showing up. Did you add the patches that you posted to the later versions of juce?

[jules]

Jack is on my to-do-list..

And yes, all the patches I've mentioned on here went into the release.

[kraken]

JACK JACK JACK !!!
JUCE-AUDIO-CUNNINGLY-KICKINGASS

[proppy]

Hi, googled this implementation of atomic operation:
http://lists.helixcommunity.org/piperma ... 02703.html

See following patch for JUCE integration:
http://pastebin.com/826825

And following test case for validation:
http://pastebin.com/826829

The patch also '#ifndef JUCE_64BIT' SSE specific code in juce_LowLevelGraphicsSoftwareRenderer.cpp

with these modifications jucedemo 1.39 run fine on ubuntu amd64.

Hope this help

[jules]

Thanks a lot! That looks pretty sensible, though I can't test it myself. Anyone else tried it?

[patrickkidd]

Code: Select all

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 46991289978016 (LWP 11110)]
0x00000000005624ca in juce::atomicIncrementAndReturn (variable=@0x7fffa9a14e4c) at juce_Atomic.h:83
83             int result;
(gdb) bt 100
#0  0x00000000005624ca in juce::atomicIncrementAndReturn (variable=@0x7fffa9a14e4c) at juce_Atomic.h:83
#1  0x000000000056239e in juce::initialiseJuce_NonGUI () at juce_SystemStats.cpp:64
#2  0x000000000044140e in juce::initialiseJuce_GUI () at juce_Application.cpp:310
#3  0x0000000000441746 in juce::JUCEApplication::main (commandLine=@0x7fffa9a14fe0, app=0x785a30) at juce_Application.cpp:178
#4  0x0000000000441d54 in juce::JUCEApplication::main (argc=1, argv=0x7fffa9a15138, newApp=0x785a30) at juce_Application.cpp:296
#5  0x0000000000407d7f in main (argc=1, argv=0x7fffa9a15138) at ApplicationStartup.cpp:169
(gdb)


[jules]

What a crazy place to crash..

Ok, I really have no idea what's going on here.

There are some commented-out lines in the file that use gcc built-in functions to do the operations instead of asm - I think I commented them out because they're only available in certain gcc versions, but maybe that'd work on yours?

[proppy]

Code: Select all

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 46991289978016 (LWP 11110)]
0x00000000005624ca in juce::atomicIncrementAndReturn (variable=@0x7fffa9a14e4c) at juce_Atomic.h:83
83             int result;
(gdb) bt 100
#0  0x00000000005624ca in juce::atomicIncrementAndReturn (variable=@0x7fffa9a14e4c) at juce_Atomic.h:83
#1  0x000000000056239e in juce::initialiseJuce_NonGUI () at juce_SystemStats.cpp:64
#2  0x000000000044140e in juce::initialiseJuce_GUI () at juce_Application.cpp:310
#3  0x0000000000441746 in juce::JUCEApplication::main (commandLine=@0x7fffa9a14fe0, app=0x785a30) at juce_Application.cpp:178
#4  0x0000000000441d54 in juce::JUCEApplication::main (argc=1, argv=0x7fffa9a15138, newApp=0x785a30) at juce_Application.cpp:296
#5  0x0000000000407d7f in main (argc=1, argv=0x7fffa9a15138) at ApplicationStartup.cpp:169
(gdb)


Can you compile&run the following test:
http://pastebin.com/826829

After having applied my patch on juce source:
http://pastebin.com/826825

Is JUCE_64BIT defined ?

[proppy]

try applying the diff from:

http://farm.aminche.com/juce64.diff

or

http://zeuhl.aminche.com:8080/juce64.diff

pastebin may have added extra ^M which can prevent
the patch from applying correctly.