reproducible Segmentation Fault using jogwheels on mc6000mk2 in non-vinyl mode

Bug #1584201 reported by blub
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Mixxx
Fix Released
Undecided
Unassigned

Bug Description

I can reproducible let Mixxx crash when using the following setup:

 - Mixxx 2.0.0 on Debian/testing
 - Denon MC6000MK2 attached

doing the following:

 - turning off vinyl mode
 - searching for a CUE point using the jog wheels

After some spinning Mixxx crashes.

I was asked to let it crash a couple of times for more backtraces - so here they are (attached)

Revision history for this message
blub (wc-ubuntu) wrote :
Revision history for this message
blub (wc-ubuntu) wrote :
Revision history for this message
blub (wc-ubuntu) wrote :
Revision history for this message
Sean M. Pappalardo (pegasus-renegadetech) wrote :

Looks like the crash is happening in soundtouch::TDStretch::overlapStereo() which makes me think the controller script may be sending a bad speed value to the Mixxx engine. Are you using a preset for that controller that came with Mixxx? Also, can you try changing the time-shifting library in Preferences to "Rubberband" and see if you can still reproduce the crash? What version of PortAudio is on your system?

Changed in mixxx:
status: New → Incomplete
Revision history for this message
blub (wc-ubuntu) wrote :

The preset is the one coming with Mixxx - no changes made.
In IRC you also requested the version of PortAudio, which is libportaudio2:amd64 19+svn20140130-1.
I will try the Rubberband change in a couple of hours and report back.

Revision history for this message
Uwe Klotz (uklotzde-deactivatedaccount) wrote :

Might be related to the following bug that has recently and hopefully been fixed in 2.1:
https://bugs.launchpad.net/mixxx/+bug/1577042

As an owner of an MC6000mk2 and author of the Mixxx mapping I will try to reproduce this issue.

Changed in mixxx:
assignee: nobody → Uwe Klotz (uklotzde)
Revision history for this message
Uwe Klotz (uklotzde-deactivatedaccount) wrote :

The backtraces for the crashes are the same in both bug reports. But the crash is not related to the controller, because in my case no controller was connected.

Revision history for this message
Uwe Klotz (uklotzde-deactivatedaccount) wrote :

I'm not able to reproduce the crash on the current 2.1.0-alpha-pre. Looks like this issue has been resolved by fixing the memory corruption issue in the SoundTouch engine buffer code recently (see above).

Are you able to compile and test 2.1.0-alpha-pre yourself? The current development version is very stable and performs really well. I use it myself all the time (with tag writing extensions) and recommend it over 2.0.0 ;) The beta release is only blocked by a non-existent Windows build server as I understood from the discussions.

If no new findings occur we should consider this bug as a duplicate and may close it with status "Fix Committed".

Revision history for this message
blub (wc-ubuntu) wrote :

I am and will compile Mixxx from master (I assume? I didn't find any branch / tag related to 2.1) myself, hence I'd recommend leaving this bug report open until I can confirm it being fixed / a duplicate.

Revision history for this message
Uwe Klotz (uklotzde-deactivatedaccount) wrote :

Great, master is the correct starting point! Currently no release branch for 2.1 has been created.

Revision history for this message
raskolnikov (raskolnikov) wrote :

I am suffering the same bug with the M-Audio Xponent wheel. I just tried with git master and the problem is still there. Using Rubberband does workaround, but my computer is slow so I often get the CPU meter on the red with it...

Revision history for this message
Uwe Klotz (uklotzde-deactivatedaccount) wrote :

Did you compile Mixxx from sources? Which OS?

It would be helpful if you attach a backtrace of such a crash. As already mentionend I am not able to reproduce this crash any more. Without further evidence we are not able to spot the corresponding cause somewhere in the code.

Revision history for this message
raskolnikov (raskolnikov) wrote :

Hi Uwe!

Sorry I was too brief. I am using Debian Sid, and built master from source with optimize=native. I am running on a 7-years old laptop with an Intel processor.

Anyways, I can't reproduce it now, so probably it was my fault... Actually I think I had some wrong linker paths and it was still running against some stuff from the distro when I first tried. I will do a longer practice session tomorrow and report back if I have any problems.

Thanks for the quick reply!

Revision history for this message
raskolnikov (raskolnikov) wrote :

I have been practicing the whole week with the new build on master and I can confirm that the bug seems gone.

Also, I played for like 4 hours straight at a club on Friday with this build (but Rubberband instead of SoundTouch) with no problems. Mixxx is really becoming way more stable with time. Thank you all team, you rock!

Changed in mixxx:
status: Incomplete → Fix Committed
Changed in mixxx:
milestone: none → 2.1.0
Changed in mixxx:
status: Fix Committed → Fix Released
Revision history for this message
Swiftb0y (swiftb0y) wrote :

Mixxx now uses GitHub for bug tracking. This bug has been migrated to:
https://github.com/mixxxdj/mixxx/issues/8555

lock status: Metadata changes locked and limited to project staff
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.