From 5dd2fda18fbbeab95ba44200e48c8202dae65206 Mon Sep 17 00:00:00 2001 From: Richard Warren Date: Mon, 18 Feb 2019 09:44:42 -0500 Subject: Modify RELEASE.txt in response to PR comments --- release_docs/RELEASE.txt | 32 +++++++++++++++----------------- 1 file changed, 15 insertions(+), 17 deletions(-) diff --git a/release_docs/RELEASE.txt b/release_docs/RELEASE.txt index 1503ac1..31a80f6 100644 --- a/release_docs/RELEASE.txt +++ b/release_docs/RELEASE.txt @@ -250,16 +250,6 @@ New Features (MSB - 2019/01/02, HDFFV-10652) - - Supporting 2GB IO with MPI 2 functionality was implemented and tested. - The proposed support recognizes IO operations > 2GB and if the datatype - is not a derived type, breaks the IO into chunks which can be input or - output with the existing MPI 2 limitations, e.g. size reporting and function - API size/count arguments are restricted to be 32 bit integers. - Derived types in which the datatype size is > 2GB still result in - IO failures. As a result, this BUG remains unresolved. - - (RAW - 2019/02/14, HDFFV-10540) - Fortran Library: ---------------- - Added wrappers for dataset object header minimization calls. @@ -1038,15 +1028,23 @@ Known Problems CPP ptable test fails on VS2017 with Intel compiler, JIRA issue: HDFFV-10628. This test will pass with VS2015 with Intel compiler. - Older MPI-2 libraries such as OpenMPI 2.0.1 and MPICH 2.1.5 were tested + Older MPI libraries such as OpenMPI 2.0.1 and MPICH 2.1.5 were tested while attempting to resolve the Jira issue: HDFFV-10540. The known problems of reading or writing > 2GBs when using MPI-2 was - partially resolved with the MPICH library but OpenMPI testing produces - too many errors. Within a single parallel test (testphdf5), the following - subtests report data verfication errors (cdsetr, eidsetr) before eventually - aborting. More recent versions (v3.1.3 or newer) of OpenMPI appear to have - resolved most of the known testing problems even though occasional test - hangs have been observed. + partially resolved with the MPICH library. The proposed support recognizes + IO operations > 2GB and if the datatype is not a derived type, the library + breaks the IO into chunks which can be input or output with the existing + MPI 2 limitations, i.e. size reporting and function API size/count + arguments are restricted to be 32 bit integers. For derived types larger + than 2GB, MPICH 2.1.5 fails while attempting to read or write data. + OpenMPI in contrast, implements MPI-3 APIs even in the older releases + and thus does not suffer from the 32 bit size limitation described here. + OpenMPI releases prior to v3.1.3 appear to have other datatype issues however, + e.g. within a single parallel test (testphdf5) the subtests (cdsetr, eidsetr) + report data verfication errors before eventually aborting. + The most recent versions of OpenMPI (v3.1.3 or newer) have evidently + resolved these isses and parallel HDF5 testing does not currently report + errors though occasional hangs have been observed. Known problems in previous releases can be found in the HISTORY*.txt files in the HDF5 source. Please report any new problems found to -- cgit v0.12