configure fixes for HDFFV-9068,9069,9096,9097,9135 from autotools_rework branch:
- Remove Infering parallel compilers (C and Fortran) from configure.ac
- Remove restriction to build shared with parallel
- Cleanup parallel sections in configure.ac
- remove large file support checks
- MPE fixes.
- Do not remove the prefix in the chunk alloc file, otherwise it makes the H5Fcreate fail when prefix is specified.
- fix tabbing in that file too.
tested: jam w/ parallel
fix bug in shapesame test where test fails with more than 10
procs. This is a fix in the test itself and does not have to do with
the library code.
tested cloud machine with parallel.
fix bug in parallel tests where a compact dset is being created
proportional to the process count without regard to the 64KB limit for
compact datasets.
Merge changes that correspond to the 64-bit ID changes (without the actual
switch to 64-bit IDs) to the 1.8 release branch. (Plus a few minor cleanups
and alignments with the trunk that aren't on the branch)
Tested on:
Mac OSX/64 10.9.4 (amazon) w/C++ & FORTRAN
(h5committested on branch already for a week)
- remove configure checks for MPI_File_get_size and Big MPI_File_set_size as they are supported by MPIO implementations today.
- fix bug in t_mpi.c (HDFFV-8856)
tested with h5committest.
Added instead configure variable hardcode_into_lib=no. This removes rpath from lib files on Linux and solaris machines.
Tested with h5committest on jam, koala, ostrich and platypus (cmake), and on emu and kite.
Bring r24864 from trunk to 1.8 branch:
Remove all traces of MPI-POSIX VFD and GPFS detection/code.
Remove remaining traces of stream VFD.
Remove testpar/t_posix_compliant test (it's not actually verifying anything).
Clean up H5D__mpio_opt_possible() further.
Moved environment variable that disables MPI collective operations into
MPI-IO VFD (instead of it being in src/H5S.c).
A few other small code cleanups.
Tested on:
Mac OSX/64 10.9.2 (amazon) w/parallel & serial
(daily tested on trunk)
Bring r24769 & r24772 from trunk to 1.8 branch:
Check in Mohamad's changes to support collective I/O on point selections,
along with some other minor cleanups.
Correct some dynamically sized arrays that fail on Windows.
Tested on:
Mac OSX/64 10.9.2 (amazon) w/parallel
(And has been through the daily test regime on the trunk)
fix bugs in parallel tests exposed in corner cases when running with 1 or 2 processes.
First bug is in testpar/t_mdset.c, where the test reports an error in addition to skipping the test if there are less than three procs. Fix to just skip the test.
Second bug is in testpar/t_dset.c in actual_io_mode tests, where incorrect expected value for IO mode was set if the number of procs running the test is 1.
tested with h5committest.
Bring r23529 from trunk to 1.8 branch:
Stop aliasing property to indicate internal collective metadata operations
with property to perform collective raw data operations from the application.
Tested on:
Mac OSX/64 10.8.3 (amazon) w/parallel
(daily tests on trunk)
HDFFV-8146 - Remove "multi-chunk IO without optimization" sub-feature from MPI I/O optimization for chunked dataset feature
Description:
The “multi-chunk IO without optimization” feature is removed and made the related xfer property (H5FD_MPIO_CHUNK_MULTI_IO) go directly to “multi-chunk-io” feature.
Also update/fix/cleanup testings (chunk collective IO and actual chunk opt mode) accordingly.
Merged from HDF5 trunk r22992.
Tested:
jam-pp (linux32-LE), koala (linux64-LE), ostrich (linuxppc64-BE), fred (mac64-LE), linew (solaris-BE), koala-pp, wallby-pp, cmake (jam-pp), Windows (32-LE cmake)
HDFFV-8143 Provide a routine(s) for telling the user why the library broke collective data access
Description:
Changed to use H5Dopen2() instead of H5Dopen(), which caught by daily test on jam.
This is based on previous commit r22780.
Tested:
jam-pp (linux32-LE)
HDFFV-8143 Provide a routine(s) for telling the user why the library broke collective data access
Description:
Added H5Pget_mpio_no_collective_cause() function that retrive reasons why the collective I/O was broken during Read/Write IO access.
Reasons to break collective I/O:
- SET_INDEPENDENT
- DATATYPE_CONVERSION
- DATA_TRANSFORMS
- MPIPOSIX
- NOT_SIMPLE_OR_SCALAR_DATASPACES (NULL Space)
- POINT_SELECTIONS
- NOT_CONTIGUOUS_OR_CHUNKED_DATASET (Compact or External-Storage)
- FILTERS
Merged from HDF5 trunk r22735, r22741 , 22743, 22744, 22763.
Tested:
jam-pp (linux32-LE), koala-pp (linux64-LE), wallaby-pp
Bring r22758 from trunk to 1.8 branch:
Bring generic improvements from encode/decode property list branch to
the trunk. This includes a better version of the property list comparison
routine, cleaned up compiler warnings, and some cleaned up property list
callbacks. Also, started on changes to clean up parallel test output, so that
it doesn't report successful tests from each process.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug, GCC 4.7.x, FORTRAN, C++, threadsafe and parallel
(h5committested on trunk)
Convert configure.in to configure.ac.
Align TESTS usage with autotools convention.
Update all references from configure.in to configure.ac
Tested: h5committest
Bring r22287 from trunk to 1.8 branch:
Clean up more FUNC_ENTER/FUNC_LEAVE macros and move H5D & H5T code toward
the final design (as exemplified by the H5EA & H5FA code).
Tested on:
Mac OSX/64 10.7.3 (amazon) w/debug & parallel
Bring r22244 from trunk to 1.8 branch:
Correct use of deprecated API routines in test routine.
Tested on:
Linux/PPC 2.6 (ostrich) w/C++ & FORTRAN, w/threadsafe, in debug mode
Linux/64-ia64 2.6 (ember) w/Intel compilers, w/C++ & FORTRAN, parallel,
in production mode
Bring r22219 from trunk to 1.8 branch: (with release manager's blessing)
Merge "file image" changes from feature branch back to trunk.
Tested on:
h5committested
Bring r21442 and r21529 from trunk to 1.8 branch: Add new API routines
for querying if collective raw data I/O was actually performed collectively.
Tested:
Linux/32 2.6 (jam)
(h5committested on trunk)
HDFFV-7639
Purpose:
Remove H5_MPI_SPECIAL_COLLECTIVE_IO_WORKS and
H5_MPI_COMPLEX_DERIVED_DATATYPE_WORKS #defines from source.
Description:
Two advanced parallel functionalities, special collective IO and
complex derived datatypes, are not supported by older
implementations of mpi, and thus our code limits the use of these
features with #ifdefs and has checks in configure to set them (or
not). Unfortunately, configure can't actually run a parallel check
to see if these features are working (nor not) so it resorts to
looking in the config files where they are explicity enabled or
disabled based on versions of mpi, sytems being built on, or for
no documented reason at all (i.e. just set to on or off as some
'default'). Overriding these settings is easy if need be, provided
it is known that it needs to be done to get improved performance,
and oftentimes it is not.
Most new MPI implementations successfully handle the functionality
requested when these #defines are set, and many of the "turn these
features off" cases in the config files are for old (> 5 years)
versions of MPI and retired systems (such as NCSA's tungsten).
Therefore, the decision has been made to remove the support for
these old versions of MPI and systems that cannot handle these
behaviors. The #ifdefs and supporting setup in the config/ files
and configure script has been removed, and the code executed when
these options were not set removed from the source.
In passing, this commit also cleans up some whitespace issues in
both t_mpi.c and H5Dmpio.c. Furthermore, in t_mpi.c, the special
collective IO test was not getting regularly run due to it being
written to work only with four processes (we regularly test with
six, previously with three), and thus it failed when actually run
due to an out of bounds data buffer assignment. It has been
modified to run at any number of processes greater than four, and
the memory problem has been fixed so the test passes.
Tested:
Jam, h5committest, Ember
