Re: [Lilypond-auto] Issue 3194 in lilypond: crash when too many bars requested

[Thomas Morley]

2013/2/18  <address@hidden>:
>
> Comment #2 on issue 3194 by address@hidden: crash when too many bars requested
> http://code.google.com/p/lilypond/issues/detail?id=3194
>
> Issue 1475 was reported fixed in version 2.13.47.  Can someone check whether
> a) version 2.13.47 will fail with the input
>
>     \repeat unfold 651 { c'1 c  \break }
> (the number may be different depending on available memory, but using 3
> times as much is likely a safe bet given the apparently exponential
> behavior)

I downloaded lilypond-2.13.47.tar.gz from
http://download.linuxaudio.org/lilypond/source/v2.13/
and self-compiled it within LilyDev.

I tested
{
 \repeat unfold xy { c'1 c  \break }
 \bar "|."
}
with different values for xy.

For 2.13.47 I had at least one successful run with xy = 1263, but it
was _not_ reproducable. After testing several other values, it failed
testing the same value (xy = 1263).
??

xy = 2500
returns:
[...]
Preprocessing graphical objects...terminate called after throwing an
instance of 'std::bad_alloc'
  what():  std::bad_alloc
Aborted

xy = 5000
[...]
returns:
Preprocessing graphical objects...terminate called after throwing an
instance of 'std::length_error'
  what():  vector::_M_fill_insert
Aborted

xy = 10000
returns:
[...]
Preprocessing graphical objects...Segmentation fault

With 2.12.3, 2.16.1, 2.17.10
I had one successful run with xy = 1263
But I didn't repeat the test and didn't test other values.

With 2.14.2 I aborted compilation.

Every compilation with every tested versions tooks a very long amount,
up to  ~20 min. 2.14.2 maybe longer.

-Harm