[Pkg-octave-devel] Dynare FTBFS / Octave+gnuplot bug

[Sébastien Villemot <sebastien.villemot@ens.fr>]

Hi everyone,

The new Dynare package randomly FTBFS, as can be seen on kfreebsd-amd64
where it failed the first time and succeeded the second time, and it
failed two times on kfreebsd-i386, while it succeeded on the other
architectures.

The problem is rather intricate and looks like a race condition between
multiple instances of Octave and gnuplot.

I attach two files which isolate the problem on my machine.

The following fails most of the time (but not always, try to run it
several times):

octave foo1.m && octave foo2.m

...while the following succeeds most of the time:

octave foo1.m && sleep 1 && octave foo2.m

The Dynare testsuite runs a similar sequence of Octave instances with
graphics and M-files created on-the-fly, hence the problem reported on
kfreebsd-i386 arch. I had created a poor-man workaround in quilt patch
"testsuite-octave3.2-crash-workaround" of Dynare package, but this is
not enough, at least on some buildds. Increasing the sleep delay would
probably solve the problem, but this is definitely not a satisfactory
solution.

Any ideas? Should I open a bug against Octave3.2 or directly report that
to Octave bug list?

Best,

-- 
Sébastien Villemot

if exist("foo3.m")
  delete("foo3.m")
endif

plot([ 1 2 3 ])

system("echo 1+1 > foo3.m");
foo3;