Based on my experience I would say is a problem with the PGI compiler and the HDF5 library. It is a problem with the hdf5 library, not FLASH or the compiler. I think that HDF5-1.8.9 may work with the PGI compiler because someone here submitted a bug fix related to HDF5 not working with the PGI compiler not too long after hdf5-1.8.8 came out.<br>
<br><div class="gmail_quote">On Wed, Dec 19, 2012 at 1:03 PM, Sean Lynch <span dir="ltr"><<a href="mailto:sean.m.lynch@drexel.edu" target="_blank">sean.m.lynch@drexel.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="HOEnZb"><div class="h5">On 12/07/2012 11:53 AM, Sean Lynch wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
I've been having some problems with a Flash4 simulation creating HDF5<br>
output files on TACC Ranger.<br>
<br>
I wasn't sure what was happening so I tried the default Sedov problem<br>
and got the same error:<br>
<br>
WARNING: you have called IO_writePlotfile but no plot_vars are defined.<br>
Put the vars you want in the plotfile in your flash.par (plot_var_1 =<br>
"dens")<br>
*** Wrote plotfile to sedov_hdf5_plt_cnt_0000 ****<br>
Initial plotfile written<br>
<br>
I haven't changed the default flash.par so the plot variable are defined<br>
as:<br>
<br>
plot_var_1 = "dens"<br>
plot_var_2 = "pres"<br>
plot_var_3 = "temp"<br>
<br>
The plot file is written to disk but contains no data for density,<br>
pressure, or temperature.<br>
<br>
I am using the default makefile for <a href="http://ranger.tacc.utexas.edu" target="_blank">ranger.tacc.utexas.edu</a> and have<br>
loaded the hdf5/1.6.5 module.<br>
<br>
Has anyone experienced this behavior?<br>
</blockquote>
<br></div></div>
This seems to be related to the Portland compiler.<br>
<br>
With pgi I get the following warning during compilation (I didn't noticed it before)<br>
<br>
PGF90-W-0164-Overlapping data initializations of haveplotvars (IO_init.F90)<br>
<br>
I did some debugging in IO_init and 'haveplotvars' is always .FALSE. even right after it is declared as .TRUE.. So this seems to be why I was getting no output. It didn't have anything to do with the HDF5 library.<br>
<br>
When I switched the compiler to Intel it compiled fine. It also ran correctly and produced the correct output for the Sedov problem.<br>
<br>
Is this a bug in the pgi compiler or a problem with Flash?<br>
<br>
</blockquote></div><br>