summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorAlbert Cheng <acheng@hdfgroup.org>2005-12-11 04:15:12 (GMT)
committerAlbert Cheng <acheng@hdfgroup.org>2005-12-11 04:15:12 (GMT)
commit67f39db8a0b332b10a66128da900bb48faca9b40 (patch)
tree632648f3682430fb500be2b3de7ce640376f3e87
parentfe505c2bbf510941cefe397346aa72436e10938d (diff)
downloadhdf5-67f39db8a0b332b10a66128da900bb48faca9b40.zip
hdf5-67f39db8a0b332b10a66128da900bb48faca9b40.tar.gz
hdf5-67f39db8a0b332b10a66128da900bb48faca9b40.tar.bz2
[svn-r11782] Purpose:
bug fix. Description: Mpich2 also has the derived type errors but some mpich2 gave no verison number and some gave 1.0.2. Solution: Added those two cases to catch mpich2 that had these sort of errors. Platforms tested: heping using mpich and mpich2.
-rw-r--r--config/mpich12
1 files changed, 12 insertions, 0 deletions
diff --git a/config/mpich b/config/mpich
index 3e63e56..4c48a2d 100644
--- a/config/mpich
+++ b/config/mpich
@@ -24,6 +24,9 @@
# Notice that this code works only if the mpicc compiler shows its $MPIVERSION
# properly. It is confirmed mpicc does that as far back as v1.2.3.
+# mpich2 do not support it correctly. But mpich2 reports small version number
+# indistiguishable from mpich. Some mpich2 reports blank, some reports 1.0.2.
+
if [ -z "$hdf5_mpi_complex_derived_datatype_works" ]; then
ccversion=`$CC -v 2>/dev/null`
# mpich compiler will give "mpicc for 1.2.x ..."
@@ -34,6 +37,15 @@ if [ -z "$hdf5_mpi_complex_derived_datatype_works" ]; then
1.2.[0-5]*)
hdf5_mpi_complex_derived_datatype_works='no'
;;
+ 1.0.2)
+ # mpich2 is recycling the version number, some report
+ # 1.0.2, some report blank.
+ hdf5_mpi_complex_derived_datatype_works='no'
+ ;;
+ "")
+ # got blank ccversion. Assume it is bad a mpich2.
+ hdf5_mpi_complex_derived_datatype_works='no'
+ ;;
*)
# assume okay
;;