Fix rpmalloc assert for NotePlayHandle

[Veratil]

Did some digging on a common rpmalloc assert raised for me and found this instance of an MM_FREE not being called.

For the quick fix I just added the NotePlayHandle::freeAvailable function, since I didn't go through and see if anything else called NotePlayHandle::free.

[PhysSong]

@Veratil I think this might not fix the assertion failure if you close LMMS after playing a lot of(like 2000 or more) notes simultaneously.

[Veratil]

@Veratil I think this might not fix the assertion failure if you close LMMS after playing a lot of(like 2000 or more) notes simultaneously.

Right, I think that will have to be handled differently. This one at least fixes a blatant one that rpmalloc catches.

[Rossmaxx]

Does this PR still have any effect after #7128?

[TechnoPorg]

Does this PR still have any effect after #7128?

I think it does. I noticed this same memory leak while I was working on 7128 and was going to fix it, but this PR beat me to it. It just needs a rebase to not use the MM_FREE macros.

[Veratil]

Does this PR still have any effect after #7128?

Yes, there's still two MM_ALLOC calls compared to one MM_FREE. I'll rebase this soon.

[Rossmaxx]

I'll rebase this soon.

Wouldn't it be better to close this one and make a new PR off of master instead?

[Veratil]

I'll rebase this soon.

Wouldn't it be better to close this one and make a new PR off of master instead?

No. This PR is fine and there's no reason to close it just to open another doing the same thing. Rebasing is fairly easy and there's a reason it exists.

[Rossmaxx]

Closing in favour of #7345