[FLASH-USERS] Laserslab Constant too large or too small for real4 format

[Stylianos.PASSALIDIS at cea.fr]

Hello community,


Compiling  the Laserslab in 3D (both FLASH4.6.1 and FLASH4.6.2) after


./setup -auto LaserSlab -3d +hdf5typeio species=cham,targ +mtmmmt +pm4dev +laser +uhd3t +mgd mgd_meshgroups=6 -parfile=example3d.par


I run into


eos_tabGetSpeciesAnyTableData.F90(111): error #5022: Constant too large or too small for real4 format
  real,parameter :: largeMultiplier = 10.0**77  ! was 10.0**largeExponent
------------------------------------------^
compilation aborted for eos_tabGetSpeciesAnyTableData.F90 (code 1)
make: *** [Makefile:116: eos_tabGetSpeciesAnyTableData.o] Error 1


I tried to explicitly declare them as  real8 or double precision but apparently will not work. So in the  eos_tabGetSpeciesAnyTableData.F90 I change the following:


  real,parameter :: largeExponent = 77.0      ! was 0.25 * log10(HUGE(largeExponent))

  real,parameter :: largeMultiplier = 10.0**77  ! was 10.0**largeExponent


to the old version (removing comment) and it compiles successfully, or so it prints...


But this is quite weird, since I have already compiled Laserslab (FLASH4.6.1) on other machines and I have never encountered such a problem.


Additionally, when running it I encounter, in beginning of the output file, the following:


 read :  skipping the following line due to a syntax error:
 dt_diff_factor = 1.0e+100 # Disable diffusion dt

 read :  skipping the following line due to a syntax error:
 hx_dtFactor    = 1.0e+100


Note that they are the default values in the example3d.par file and I also never had such problem in the past.

Any comments on would be appreciable.

Thank you for your time,

S. Passalidis


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20221028/1ed18f1f/attachment.htm>