Re: Segfault 2.15.23 Span_bar_stub_engraver

[Keith OHara]

mike <at> apollinemike.com <mike <at> apollinemike.com> writes:
> Some meh news: I don't get a segfault ...
> Maybe someone else could give it a look?
 
I can't produce the segfault, either.

Jay Anderson <horndude77 <at> gmail.com> writes:
> Removing the "Span_bar_stub_engraver" from my
> PianoStaff fixes the segfault, but what won't work if I do this?

Lyrics inside the piano staff would hit span bars, but you don't have any,
so the engraver should not be doing anything.  
The back-trace, however, seems to say that the engraver is doing something.

> #0  Scheme_hash_table::try_retrieve (this=0x0, k=0x7ffff342f520,
>     v=0x7fffffff8b78) at scm-hash.cc:97
> #1  0x000000000048e331 in Context::where_defined (this=0xd71ff0,
>     sym=0x7ffff342f520, value=0x7fffffff8b78) at context.cc:420
> #2  0x00000000004953a7 in updated_grob_properties (tg=<optimized out>,
>     sym=0x7ffff342f520) at context-property.cc:264
> #3  0x00000000006272fe in Span_bar_stub_engraver::process_acknowledged (
>     this=0xce1d10) at span-bar-stub-engraver.cc:119

When I put a debug print at span-bar-stub-engraver.cc:119
and run your file (with the engraver re-enabled) I never reach that line!

Mike, Maybe it is the empty \new Dynamics {} inside the PianoStaff in Jay's
score?  A print at line 109 shows that LilyPond is considering the empty
Dynamics for inclusion in affected_contexts_.  Maybe we are testing 
de-allocated memory?