[Openmcl-devel] Floating point errors for innocuous function calls

[Liam Healy]

In running a complete suite of tests for GSLL, I'm finding
several inexplicable floating point errors for what should
be legitimate function calls.

    EXPONENTIAL-FUNCTIONS: DIVISION-BY-ZERO detected
performing EXP on (-10.0D0)

GIVENS: DIVISION-BY-ZERO detected
performing SIN on (-39.66)

    HANKEL: FLOATING-POINT-OVERFLOW detected
performing EXP on (-0.18873077993219034D0)

 LINEAR-LEAST-SQUARES: FLOATING-POINT-OVERFLOW detected
performing EXP on (0.1D0)

 MINIMIZATION-ONE: DIVISION-BY-ZERO detected
performing COS on (0.0D0)

plus a couple others of a similar nature.  I suspect these all
have a common origin in CCL.  When I do these calculations
independently, sometimes I get the error but usually I don't

? (cos 0.0d0)
> Error: DIVISION-BY-ZERO detected
>        performing COS on (0.0D0)
> While executing: CCL::%FP-ERROR-FROM-STATUS, in process listener(1).
> Type :POP to abort, :R for a list of available restarts.
> Type :? for other options.

? (cos 0.0d0)
1.0D0


(lisp-implementation-version)
"Version 1.6-dev-r13980M-trunk  (LinuxX8664)"

Liam