Bug Fix
Description:
Stumbled across a problem when working on the J90 h5dump problem where
I noticed that some of the output for named compound datatypes wasn't being
displayed.
Solution:
Corrected output to display missing information.
Platforms tested:
FreeBSD 4.2 (hawkwind) & Cray J90 (killeen)
Additional test
Description:
Exercise reading background buffer from file when reading only one field
from a compound datatype which has array fields.
Platforms tested:
FreeBSD 4.2 (hawkwind)
Bug Fix
Description:
Compound datatypes weren't reading background data properly when used
with array or vlen fields.
Solution:
Changed to always read background information from file in those situations.
Platforms tested:
FreeBSD 4.2 (hawkwind)
Purpose:
Finalize new API function descriptions
Description:
Edits to recently-added APIs:
H5Pset_sieve_buf_size H5Pset_fapl_log
H5P[sg]et_fapl_mpio H5Pset_fapl_multi
H5P[sg]et_dxpl_multi H5Pset_fapl_split
H5P[sg]et_fapl_gass H5Pset_fapl_dpss
H5P[sg]et_fapl_srb
Added H5P[sg]et_fapl_core.
Addition of new APIs for R1.4 is now complete.
Moved H5Pset_driver and H5Pget_driver_info to TechNotes.
Platforms tested:
IE 5
Fix
Description:
When not building in the same directory, we needed to tell the
checker wherre to look for files.
Solution:
Put $(srcdir) before things which needed it.
Platforms tested:
Linux
Add
Description:
Adding a small amount of testing to the C++ library until "real"
testing can be created. (We use the examples).
Platforms tested:
Linux
Fix
Description:
Needed to change the extention to the file configure expects to
be in the source directory when Binh-Minh changed the extensions
for all the files.
Solution:
Did just that.
Fix
Description:
Windows couldn't handle the .C extension. The filenames had to be
changed to *.cpp instead.
Solution:
Changed in the Makefiles to accomodate said change.
Platforms tested:
Eyeballed.
Purpose:
Support portability
Description:
I forgot that source file extension .C will not work on Windows.
Solution:
Changed all source file from *.C to *.cpp for portability.
Platforms tested:
arabica (sparc-sun-solaris 2.7)
Description:
The prototype of two C API functions, H5Tget_offset and
H5Sget_simple_extent_npoints was changed to fix bug #446,
resulting in the need for updating the two corresponding
C++ API functions.
Solution:
- Changed the return type of AtomType::getOffset from size_t
to int, and the error return value from 0 to -1.
- Changed the return type of DataSpace::getSimpleExtentNpoints
from hsize_t to hssize_t, and the error value from 0 to -1.
Platforms tested:
arabica (sparc-sun-solaris 2.7)
Purpose:
updating all.dsw(adding h4toh5dll and h5toh4dll version).
adding a automatic crude test script of h4toh5 converter on windows.
Description:
Solution:
Platforms tested:
WIn 2000
Purpose:
a bug fix for windows
Description:
ignore checking the permission of file and directory on windows
Solution:
using ifndef WIN32 endif to ignore checking this feature.
Will find how windows deal with file checking stuff later.
Platforms tested:
WIN 2000, and confirmed at eirene(Linux)
Improvement
Description:
Made the Makefiles better.
Solution:
Added standard macros concerning where things are place. Like
libraries and so forth.
Platforms tested:
Linux
Purpose:
Add a crude script for automatic testing h5toh4 converter tool
Description:
the name of the script is called h5toh4testrun and is under ....../hdf5/tools directory.
Solution:
Four parameters should be typed in correctly in order to do the auto test.
1. Debug or Release version you are testing.
2. The top directory path(including the drive name) where hdf5 tools is located.
Such as D:/h5_1.4pre2
3. The full path where hdp is located.
Such as D:/H414/bin
4. If no dll version(currently not tested yet); leave this parameter empty or type DLL.
Platforms tested:
NT 4.0(NT) and 5.0(2000).
Purpose:
turn off "checking file and directory permission right" on windows platform
Description:
Adding #ifndef WIN32 #endif block at the corresponding test-file and test-dir functions.
Solution:
See above.
Platforms tested:
Win NT 4.0,5.0; and confirm the result the same on eirene and arabica
Bug Fix
Description:
The fix we put in to ignore warning 405 was incorrect. The quotes
were confusing to configure.
Solution:
Turns out that there doesn't need to be a space between the -M
flag and the warning number.
Platforms tested:
T3E
Feature
Description:
H5Eprint used to print "thread 0" even in MPI-application.
Added code to print MPI-process rank ID (as in MPI_COMM_WORLD) to
give more information of which process is printing the messages.
Platforms tested:
IRIX64-64, both serial and parallel.
Bug Fix.
Description:
Attributes whose datatype was an object or dataset region reference weren't
dumping correctly.
Solution:
Pass along attribute ID to the code which prints the attributes, so the
reference types can be dereferenced and printed.
Platforms tested:
FreeBSD 4.2 (hawkwind) & IRIX64 6.5 (modi4)
Improvement
Description:
During configure, it tries to find the size of various basic
C data types (int, int8, fast_int8, ...) by running a small
program for each data type. But Pecific blue uses a batch queue
system to launch all jobs and the whole process takes couple
minutes. With over 20 basic types to test, that takes "forever"
for the configure to complete.
Solution:
Let the configure complete, then save all those valid values in
the config.cache file into the specific config file for Pac. Blue.
With the values cached, configure does not need to run those little
problems.
Platform Tested:
Pacific blue with mpicc.
