Problem with the stopping power of polyethylene

It seems that a bug has been introduced in FLUKA. The input attached
was working until a short time ago (with a few trivial changes that
do not concern materials).
The log file says: Floating exception.
The .out and the .err file report a number of error messages such as:
  *** dp/dx:d,imat,ztar,ij,po,eo -0.00108121724 36 1.06943913 1 1.51030313
   1.77802432 ***
  *** x,xoster,ccster,d0ster 0.206735223 0.2 -0.556058849 0. ***

The material to which the messages refer (imat=36) is POLY, defined as
follows:
MATERIAL 0.0 0.0 0.94 36.0 0.0POLY
COMPOUND -0.143711 CARBON -0.856289 HYDROGEN 0.0 POLY
This definition was working in the past. Trying to run again the old
input gives now the same error message.

Looking into the source code, it appears that the message is sent when the
"d" variable is negative (-0.00108121724 in the example shown above).

In the .out file, it can be seen that the Sternheimer parameters of POLY have
problems:
  ***** dp/dx : material number 36 "POLY " *****

    ***** Average excitation energy : 2.1479E+01 eV, weighted Z/A : 9.2137E-01
****
*
    ***** Sternheimer density effect parameters:
****
*
    ***** X0 = 0.2000, X1 = 2.0000, C = -0.5561, A = ******** m = -22.4849
D0 = 0.0000 ****

The value of A ( -0.0626 in a past run without error) has been replaced
by a string of asterisks, and that of m (3.0000 in the past) is now
-22.4849.

Can somebody please find out what has been changed in the code and fix it?

Alberto
and Kelly (Qiu Rui, SLAC RP)