Panda3D fails to get LightMutexHolder lock.

Bug #1694898 reported by ponyboy837
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Panda3D
Expired
Undecided
Unassigned

Bug Description

Panda3D seems to fail trying to get a LightMutexHolder lock (https://prnt.sc/feijna) in MouseWatcher when trying to replace a group and returns a memory violation error instead.

The project this was found in is located here (https://github.com/theclashingfritz/libotp)

I don't know much on why this actually happens so feel free to look into it.

Revision history for this message
rdb (rdb) wrote :

The key clue here is in the inspector which shows that "this" is 0. So, it is being called on a NULL pointer, which is why accessing the _lock member fails.

However, looking in pgTop.cxx cull_callback function, it seems there is already an assert to check whether it is NULL before calling replace_group. Are you compiling without assertions, perchance?

Revision history for this message
ponyboy837 (ponyboy837) wrote :

I'm not calling it at all. The graphics engine is what ends up calling it which is a issue. A bit a odd one.

Revision history for this message
ponyboy837 (ponyboy837) wrote :

Also my panda3d is compiled with assertions. (That's the reason why i need to show and hide Mouse Watcher regions to add a region without crashing oddly enough)

Revision history for this message
ponyboy837 (ponyboy837) wrote :

This doesn't look good at all. https://prnt.sc/feqmz2

Revision history for this message
rdb (rdb) wrote :

That (unless the debugger is incorrect) suggests again that something is being called on a nullptr GraphicsEngine.

It could be that something more fundamentally wrong is going on in your build; for example that the compiler settings (_DEBUG or NDEBUG setting, or MSVC version) doesn't match the one one used to produce the Panda build.

Revision history for this message
ponyboy837 (ponyboy837) wrote :

That's very possible. But i don't know what would trigger that kind of issue.

Revision history for this message
ponyboy837 (ponyboy837) wrote :

And that's because the compiling settings should match. (It's a odd life we have)

Revision history for this message
rdb (rdb) wrote :

Could this be a similar issue to https://bugs.launchpad.net/panda3d/+bug/1694353 ; due to improper reference counting?

Revision history for this message
rdb (rdb) wrote :

If you have a reproduction case that can be easily compiled and run by one of the developers, please reopen this bug on GitHub.

Changed in panda3d:
status: New → Incomplete
Revision history for this message
Launchpad Janitor (janitor) wrote :

[Expired for Panda3D because there has been no activity for 60 days.]

Changed in panda3d:
status: Incomplete → Expired
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.