Current master segfaults on an openSuSE Tumbleweed system if a global staff size equal or greater 14.5 is used. It also segfaulted with git master about half a year ago, but I had/have no time to investigate the problem. Maybe someone is interested.
If you'd spent about 2 minutes at least making an effort to simplify this file you'd have noticed that as soon as you remove all the '\break' statements - or rather I removed the 'obreak' command you created and the corresponding strings in your huge file, then it compiles fine. But with the 'break' statement yes it does still segfault on master (2.19.65)
I installed tumbleweed and used the install instructions as per 13.2 OpenSUSE
But really?
We give users who don't have access to create trackers grief for not providing tiny-examples. Is there a particular reason why you didn't provide such an example here?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Maybe something like that should be added to the documentation:
"If you do not see how to transform a correct but segfaulting lilypond source file into a tiny still segfaulting example use a monte carlo search until you succeed or don't report the problem."
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Furthermore, you didn't report a problem, but opened a tracker issue.
If you or someone else has problems tracking a file down to a minimal, even if the reason is running short of time, it woud have been appropriate to ask for help on the user- or probably the bug-list.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Well, in this particular case I have to say that the situation at a segfault makes a minimal example often less necessary for analysis than in other cases since the debugger will provide a lot of information. However, once one needs to work with reverse execution, it's important that the segfault occurs as fast as possible since "reverse execution" is actually tracing with full recording in order to be able to "step backward". So even in the segfault case, an example with minimal running time (rather than minimal source size though both are usually correlated) is helpful.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
If you'd spent about 2 minutes at least making an effort to simplify this file you'd have noticed that as soon as you remove all the '\break' statements - or rather I removed the 'obreak' command you created and the corresponding strings in your huge file, then it compiles fine. But with the 'break' statement yes it does still segfault on master (2.19.65)
I installed tumbleweed and used the install instructions as per 13.2 OpenSUSE
But really?
We give users who don't have access to create trackers grief for not providing tiny-examples. Is there a particular reason why you didn't provide such an example here?
Maybe something like that should be added to the documentation:
"If you do not see how to transform a correct but segfaulting lilypond source file into a tiny still segfaulting example use a monte carlo search until you succeed or don't report the problem."
Why this hostility?
James is simply right.
Furthermore, you didn't report a problem, but opened a tracker issue.
If you or someone else has problems tracking a file down to a minimal, even if the reason is running short of time, it woud have been appropriate to ask for help on the user- or probably the bug-list.
Well, in this particular case I have to say that the situation at a segfault makes a minimal example often less necessary for analysis than in other cases since the debugger will provide a lot of information. However, once one needs to work with reverse execution, it's important that the segfault occurs as fast as possible since "reverse execution" is actually tracing with full recording in order to be able to "step backward". So even in the segfault case, an example with minimal running time (rather than minimal source size though both are usually correlated) is helpful.