Custom Query (964 matches)

On a Cray T3E with MPI, I'm getting an error writing to history files. Traceback is shown below. I think this has been caused by code changes in the last few weeks. I'm away from work for 10 days from tomorrow, so I'm creating this ticket now and will investigate when I get back (unless someone else does so in the meantime).

It looks to me like a data type mismatch, probably specific to this platform.

: Floating point exception (invalid floating point operation)

 Beginning of Traceback (PE 0):
  Interrupt at address 0x800154e68 in routine 'ncx_put_float_double'.
  Called from line 3756 (address 0x80015eadc) in routine 'ncx_putn_float_double'.
  Called from line 975 (address 0x80014d274) in routine 'ncx_pad_putn_Idouble'.
  Called from line 2123 (address 0x80015170c) in routine 'nc_put_att_double'.
  Called from line 118 (address 0x8001ae430) in routine 'NF_PUT_ATT_REAL'.
  Called from line 268 (address 0x8001bd6e4) in routine 'NF90_PUT_ATT_ONE_FOURBYTEREAL_in_NETCDF'.
  Called from line 222 (address 0x80009fd58) in routine 'DEF_VAR'.
  Called from line 760 (address 0x80008cf88) in routine 'DEF_INFO'.
  Called from line 189 (address 0x80007cd20) in routine 'DEF_HIS'.
  Called from line 100 (address 0x800010508) in routine 'OUTPUT'.
  Called from line 168 (address 0x80000dbf8) in routine 'MAIN3D'.
  Called from line 159 (address 0x800001b64) in routine 'ROMS_RUN_in_OCEAN_CONTROL_MOD'.
  Called from line 109 (address 0x8000015c0) in routine 'OCEAN'.
  Called from line 475 (address 0x800000c94) in routine '$START$'.
 End of Traceback.

In step_floats.F, the following code (lines 278-286)...

!
!  Compute vertical position (grid units) 3D Lagrangian floats.
!
          IF (Ftype(l).eq.flt_Lagran) THEN
            track(izgrd,nfp1,l)=cff1*track(izgrd,nf  ,l)-               &
     &                          cff2*track(izgrd,nfm2,l)+               &
     &                          dt(ng)*(cff3*track(izrhs,nfp1,l)+       &
     &                                  cff4*track(izrhs,nf  ,l)-       &
     &                                  cff3*track(izrhs,nfm1,l))

...uses coefficients cff1, cff2, cff3 and cff4. However cff3 and cff4 have not been initialised, and cff1 and cff2 have values 8/3 and 4/3 (assigned to them at lines 264-265) which don't look right in this expression.

The effect of the uninitialised values is platform-dependent: on a Cray T3E it causes the model to crash.

The time of term diagnostics is not updated when only DIAGNOSTIC_BIO is activated. This type of diagnostics are only available for BIO_FASHAM. The reason for this is that set_diags is not called from main3d for this case. Many thanks to Moritz Lehmann for ​reporting this problem.

Also corrected missing continuation line in wrt_rst.F. Many thanks to Shihnan Chen for ​reporting this bug.