[Scipy-tickets] [SciPy] #1523: another Arpack crash on OS X python 2.7, 64-bit
SciPy Trac
scipy-tickets@scipy....
Mon Oct 10 04:56:45 CDT 2011
#1523: another Arpack crash on OS X python 2.7, 64-bit
---------------------------------+------------------------------------------
Reporter: rgommers | Owner: wnbell
Type: defect | Status: new
Priority: highest | Milestone: 0.10.0
Component: scipy.sparse.linalg | Version: 0.9.0
Keywords: arpack, segfault |
---------------------------------+------------------------------------------
Comment(by pv):
I doubt this has to do with the *LAMCH routines.
The Z/CLADIV wrapper we have now is in fact quite suspicious. It does a
`CALL` to a Fortran function, so it would be no wonder if it crashes. It
would be safer to replace it with a similar wrapper as for the `*dot*`
functions --- add the corresponding CBLAS-using routine to
`veclib_cabi_c.c`, and call that from `veclib_cabi_f.f`.
***
Maybe we would also would need wrappers for `SDOT`. There seems to be a
relevant comment by "jarno" in the thread you posted above
(http://www.macresearch.org/lapackblas-fortran-106):
The -ff2c flag seems to work, but in practice only with -m32, as SDOT
fails with
-ff2c -m64 (always returns 0). So, with -m64, the only way SDOT works
correctly is
with the never calling convention (no -ff2c), and the only way CDOTU
works correctly
is with the older calling convention (with -ff2c). AFAIK, they are both
in the same
library
--
Ticket URL: <http://projects.scipy.org/scipy/ticket/1523#comment:11>
SciPy <http://www.scipy.org>
SciPy is open-source software for mathematics, science, and engineering.
More information about the Scipy-tickets
mailing list