On 2018-12-21 4:28 pm, Karlin High wrote:
On 12/21/2018 5:56 PM, Kieren MacMillan wrote:
It's
only 16 pages and LilyPond crashes.
That looks like a [serious] bug to me. I’ve cc'ed the bug list.
Reproduced crash on LilyPond 2.19.80, Windows 7 Pro 64-bit SP1, Intel
Core i5-3450, 24GB RAM.
At crash, lilypond-windows.exe in Task Manager as showing 1,080,744K
of Memory (Private Working Set.) Perhaps this problem is similar to
other large-memory-use crashes reported lately on Windows?
Almost definitely. I ran this in using my setup which is 64-bit Linux
LilyPond running under WSL on 64-bit Windows 10. Peak memory usage
during compilation was around 4.4GiB, so there is no way a 32-bit
process could do this.
It took my computer about two minutes to process all 3000 measures,
which by the way typeset to 25 pages in the end--many of which are
virtually identical save for page and measure numbers.
I tried the following approach to break up the piece into three scores:
%%%%
\score {
{ \time 4/4 \repeat unfold 1000 { c4 d e f } }
}
\score {
{ \set Score.currentBarNumber = #1001 \bar ""
\time 4/4 \repeat unfold 1000 { f1 } }
\layout { indent = 0 }
}
\score {
{ \set Score.currentBarNumber = #2001 \bar ""
\time 4/4 \repeat unfold 1000 { c4 d e f } }
\layout { indent = 0 }
}
%%%%
This only took 42 seconds to complete and I believe it peaked under
2GiB.
-- Aaron Hill
_______________________________________________
bug-lilypond mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/bug-lilypond
