Re: 2.11.30 segmentation fault

[Han-Wen Nienhuys]

2007/8/31, Trevor Bača <address@hidden>:
> Hi,
>
> Crossposted as Google #433.
>
> Running 2.11.30 on a relatively large file (110 measures of six staves
> with 4559 lines of total input) causes the following segmentation
> fault.



>
> GNU LilyPond 2.11.30
> Processing `/Users/trevorbaca/Documents/lilypond/pictures/1768.ly'
> Parsing...
> Interpreting music... [8][16][24]
> Preprocessing graphical
> objects.../Users/trevorbaca/Documents/lascaux/scr/lily: line 4: 14270
> Segmentation fault
>
>
> I've been cutting out parts of the file all morning to produce a
> minimal snippet. But at this point it appears that the seg fault comes
> from the *size* of the inputfile rather than from the *musical
> contents*. (Ie, file compiles absolutely fine to and with measure 109;
> adding measure 110 *in any of the six musical staves* causes the seg
> fault.)

One way of finding this out is to run the thing inside GDB and look at
the stack trace.  Unfortunately, for useful information, you have to
do this in an unstripped binary, which is not included in the
installer.

Usually segfaults are writing to a null pointer, and I remember doing
500 mesaures of 64 staves without problems, so memory overflow seems
unlikely.

-- 
Han-Wen Nienhuys - address@hidden - http://www.xs4all.nl/~hanwen