blas error and crash when doing ldivide/rdivide of a scalar by a vector

[John W. Eaton]

On 13-Jan-2008, Jason Riedy wrote:

| I've appended a quick
| workaround inside LAPACK that should generalize to a workaround
| in Octave;

How?  We call DGELSD with lwork == -1 and ask it to tell us how much
work space to use.  Since we typically use ATLAS, this method
obviuosly won't work until packages are fixed.  So working around the
bug means a change to Octave's configure script to detect the bug,
plus switching from the nice query method for finding the workspace
requirements to using an explicit calculation from an obscure
formula.  I'd be happy to consider patches for a workaround for
Octave, but I don't plan to fix this problem myself.

Thanks,

jwe