[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[netCDF #ZBY-408193]: NetCDF building on Blue Gene/Q with IBM XL compiler?

> Russ, thank you. I tried the "-DIBMR2Fortran" before, and got a lot of
> compilation errors with "test_get.F" - then I've found that the F77
> variable wasn't set in my environment. After I set the F77 the
> compilation was OK.
> However, I came across another error - this time during "make check":
> ###  15  FAILURES TESTING nf_rename_att! ###
> *** testing nf_del_att ...
> *** testing nf_set_fill ...
> *** testing nf_set_default_format ...
> Total number of failures:  43568
> STOP 2
> FAIL: nf_test
> Should I open a separate ticket for that?
> Thanks,

It's OK to continue on this ticket, because the title and subject still applies.

I haven't seen the above symptom before.  nf_rename_att is just a thin wrapper
that calls nc_rename_att.  I assume you got no failures testing nc_rename_att,
so it's hard what's going wrong.  Can you follow this in a debugger or
insert some print statements to get more details about what is happening?

We don't have an IBM platform on which to run tests, so we can't reproduce the
problem.  It works fine with gfortran on lots of platforms, as well as 
and ifort.

The only other thing I can suggest is trying the separate Fortran netCDF API,
available in a released version (netcdf-fortran-4.2) or in a beta-test version:


In both cases, you build the Fortran library separately, telling the configure
script where the netCDF C library is installed, according to these instructions:



> --
> Aleksey Yakovlev, DRC Staff SW Developer
> Room 134B, (609) 452-5312
Russ Rew                                         UCAR Unidata Program
address@hidden                      http://www.unidata.ucar.edu

Ticket Details
Ticket ID: ZBY-408193
Department: Support netCDF
Priority: Normal
Status: Closed

NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.