<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">
<div class="moz-cite-prefix"><small><b>On 05/27/2015 05:41 PM,
Carl Ponder wrote:</b></small><br>
</div>
<blockquote cite="mid:5566480D.2020100@nvidia.com" type="cite"><font
color="#003333">Here are some oddities from the "make check"
that I ran:<br>
</font>
<blockquote><font color="#003333"><tt>./nf_test -d .</tt><br>
</font> <font color="#003333"><tt>rank 0: MPI error
(MPI_File_delete) : MPI_ERR_IO: input/output error</tt><br>
</font> <font color="#003333"><tt>rank 0: MPI error
(MPI_File_delete) : MPI_ERR_IO: input/output error</tt><br>
</font> <font color="#003333"><tt>*** TESTING F77 ./nf_test
for CDF-1 ------ pass</tt></font></blockquote>
</blockquote>
<small><b>
On 05/27/2015 06:41 PM, Wei-keng Liao wrote</b></small><br>
<blockquote type="cite">
<blockquote
cite="mid:E70E96AE-89B8-4EB1-8AA8-79B71969539A@eecs.northwestern.edu">
<pre wrap=""><font color="#003333">Those MPI error messages are expected if you are using OpenMPI or older versions of MPICH.
This error happens when trying to delete a non-existing file.
A correct MPI implementation should return MPI_ERR_NO_SUCH_FILE, instead of MPI_ERR_IO.
This has been fixed in the MPI-IO part of the latest MPICH release.
OpenMPI is copying the MPI-IO part of MPICH and has not updated to the latest yet.
You can safely ignore those error messages.</font></pre>
</blockquote>
</blockquote>
</div>
With the 1.7.0 release of PNetCDF, I've started getting there errors
from the "make check":<br>
<blockquote><tt>./nctst ./testfile.nc</tt><br>
<tt>[ivb106:6511] *** An error occurred in
MPI_Type_create_hindexed</tt><br>
<tt>[ivb106:6511] *** reported by process
[46912098861057,46909632806912]</tt><br>
<tt>[ivb106:6511] *** on communicator MPI_COMM_WORLD</tt><br>
<tt>[ivb106:6511] *** MPI_ERR_ARG: invalid argument of some other
kind</tt><br>
<tt>[ivb106:6511] *** MPI_ERRORS_ARE_FATAL (processes in this
communicator will now abort,</tt><br>
<tt>[ivb106:6511] *** and potentially your MPI job)</tt><br>
<tt>*** TESTING C++ ./nctst for APIs with different netCDF
formats ------ make[2]: [testing] Error 13 (ignored)</tt><br>
<br>
<tt>./tst_f90 ./testfile.nc</tt><br>
<tt>[ivb106:7079] *** An error occurred in
MPI_Type_create_hindexed</tt><br>
<tt>[ivb106:7079] *** reported by process
[46912052199425,46909632806912]</tt><br>
<tt>[ivb106:7079] *** on communicator MPI_COMM_WORLD</tt><br>
<tt>[ivb106:7079] *** MPI_ERR_ARG: invalid argument of some other
kind</tt><br>
<tt>[ivb106:7079] *** MPI_ERRORS_ARE_FATAL (processes in this
communicator will now abort,</tt><br>
<tt>[ivb106:7079] *** and potentially your MPI job)</tt><br>
<tt>make[2]: [testing] Error 13 (ignored)</tt><br>
</blockquote>
It looks like the 1.6.1 release passed the same:<br>
<blockquote><tt>./nctst ./testfile.nc</tt><br>
<tt>*** TESTING C++ ./nctst for APIs with different netCDF
formats ------ pass</tt><br>
<br>
<tt>./tst_f90 ./testfile.nc</tt><br>
<tt>*** TESTING F90
./tst_f90 ------ pass</tt><br>
</blockquote>
Can you tell me what the issue is here? I'm using OpenMPI 1.10.2 in
both cases.<br>
The error happens when I use the Intel 15.0, GCC 4.8.5 and PGI 16.4
compilers, so I don't believe the compiler is the problem.<br>
Thanks,<br>
<br>
Carl Ponder<br>
<br>
<br>
<DIV>
<HR>
</DIV>
<DIV>This email message is for the sole use of the intended recipient(s) and may
contain confidential information. Any unauthorized review, use, disclosure
or distribution is prohibited. If you are not the intended recipient,
please contact the sender by reply email and destroy all copies of the original
message. </DIV>
<DIV>
<HR>
</DIV>
</body>
</html>