[svn-r10201] Purpose:

IBM MPI-IO has a bug for MPI derived data type, which is used
to support collective IO. In order to avoid this, we provide
a way to turn off collective IO support for this kind of platform.


Description:
Using a new macro called H5_MPI_COMPLEX_DERIVED_DATATYPE_WORKS
to turn off irregular hyperslab collective IO support at such platforms.

Solution:
Hard code such platforms under hdf5/config.
So far only IBM AIX has been found to have such problems.

Platforms tested:

heping (Linux) and copper(Aix). Testing at tungsten now, however, it
took almost an hour to build parallel HDF5 at tungsten, it is still building. Cannot wait. Have tested this feature at Tungsten a week ago.
Misc. update:

1 files changed, 22 insertions, 0 deletions

diff --git a/configure b/configure
index 64e827f..e5111f4 100755
--- a/configure
+++ b/configure
@@ -47090,6 +47090,28 @@ echo "${ECHO_T}yes" >&6
 echo "${ECHO_T}no" >&6
   fi
 
+echo "$as_me:$LINENO: checking if irregular hyperslab optimization code works inside MPI-IO" >&5
+echo $ECHO_N "checking if irregular hyperslab optimization code works inside MPI-IO... $ECHO_C" >&6
+
+if test "${hdf5_mpi_complex_derived_datatype_works+set}" = set; then
+  echo $ECHO_N "(cached) $ECHO_C" >&6
+else
+  hdf5_mpi_complex_derived_datatype_works=yes
+fi
+
+
+if test ${hdf5_mpi_complex_derived_datatype_works} = "yes"; then
+
+cat >>confdefs.h <<\_ACEOF
+#define MPI_COMPLEX_DERIVED_DATATYPE_WORKS 1
+_ACEOF
+
+  echo "$as_me:$LINENO: result: yes" >&5
+echo "${ECHO_T}yes" >&6
+else
+  echo "$as_me:$LINENO: result: no" >&5
+echo "${ECHO_T}no" >&6
+fi
 fi