Re: Issue 1732 in lilypond: occasional segfault when compiling input/regression/midi/key-initial.ly

[David Kastrup]

address@hidden writes:

> Comment #12 on issue 1732 by address@hidden: occasional
> segfault when compiling input/regression/midi/key-initial.ly
> http://code.google.com/p/lilypond/issues/detail?id=1732
>
> You might want to add the suppressions from the attached file, which
> filters practically all spurious warnings triggered in guile's garbace
> collection...
>
> There are, however, two relevant warnings in that log file, which
> might help pinpoint the problem:
>
> ==7941== Invalid read of size 4
> ==7941==    at 0x812377D:

[...]

> ==7941==  Address 0x3e is not stack'd, malloc'd or (recently) free'd
> ==7941==
> ==7941==
> ==7941== Process terminating with default action of signal 11 (SIGSEGV)
> ==7941==  Access not within mapped region at address 0x3E
> ==7941==    at 0x812377D:

Well, that's what I consider likely a null pointer access (after walking
from 0 to the desired field).  Of course, this is the immediate cause of
the segfault.

However, I have not found the cause of the null pointer, and I have not
seen more information about it in the valgrind dump than I had been able
to find in the coredump already.

-- 
David Kastrup