[SciPy-dev] A stray pointer somwehere...

[Fernando Perez]

Hi all,

I can consistently reproduce a segfault by doing the following at an 
interactive prompt (only newcore is installed, not full newscipy, current svn 
- rev 1429):

import scipy
scipy.test()
scipy.test(10,10)
scipy.test(10,10)
scipy.test(10,10)
scipy.test(10,10)
scipy.test(10,10)

Quit the interpreter -> segfault.

This is on the Itanium2 boxes, icc build.

Note that the segfault only occurs upon quitting python, and it only happens 
if I run the test(10,10) _many times_ in a row.  Just running it once or twice 
doesn't seem to produce the problem.

I know this can be very hard to track, I just leave it here for now to have it 
reported.  I'm working on the icc build still...

Cheers,

f