[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Segfault when using specific symbol combination

From: David Kastrup
Subject: Re: Segfault when using specific symbol combination
Date: Fri, 06 Mar 2020 16:40:08 +0100
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/28.0.50 (gnu/linux)

Mathias Bockwoldt <address@hidden> writes:

> Dear Lilypond team,
> I found a symbol combination that triggers a segfault in Lilypond
> 2.18.2 on Linux (Linux subsystem for Windows 10). I did not have the
> chance to reproduce on other machines or with other versions.
> I tried looking for the bug in the database and I am 98% sure that it is not 
> there, yet.
> More information about the actual circumstances for the bug is in the 
> attached file.
> Thank you for this great product!
> Mathias
> % This example segfaults:
> % > Parsing...
> % > Interpreting music...
> % > Preprocessing graphical objects...Segmentation fault
> % It does *not* segfault when changing any of these:
> % - remove "--"
> % - remove "~"
> % - use single note instead of two (i.e. dont use <'>)
> % - adding more notes to the end (just segfaults when this "construct" is at 
> the end)
> \version "2.18.2"
> {
>   <g' b'>--~
> }

Doesn't appear to segfault with 2.21.0.  Can corroborate segfault with
the 2.18.2 version in Ubuntu.  I suspect one of several segfault fixes
we had for GCC optimizing (!this) away as being always false, basically
because we called member functions on a null pointer, and the member
functions then checked the condition.

There are other reasons why current compilers do no longer compile
2.18.2, so it would be rather tricky to track this down.

Can you try with 2.20.0?

David Kastrup

reply via email to

[Prev in Thread] Current Thread [Next in Thread]