| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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
Tested:
jam (linux32-LE), koala (linux64-LE), ostrich (linuxppc64-BE), tejeda (mac32-LE), linew (solaris-BE)
|
| |
|
|
|
|
| |
Tested: h5committest
|
|
|
|
|
|
| |
autotools TESTS var, rename configure.in to configure.ac, convert test scripts from hard *.sh to configure managed *sh.in files.
Tested: h5committest
|
|
|
|
|
|
|
|
|
|
| |
Bring back some more cleanups from the plist_encode_decode branch, clean up
some formatting and compiler errors, and add a few more property comparison
routines that we've skipped implementing in the past.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug
(Too minor to require h5committest)
|
|
|
|
|
|
| |
libtool 2.4.2
Tested THG machines.
|
|
|
|
|
|
|
|
| |
Propagated Windows and POSIX correctness changes from SEC2 VFD to
the MPI-POSIX VFD. Fixes HDFFV-8058.
Tested on:
jam w/ parallel and fortran enabled
|
|
|
|
|
|
|
|
| |
Minor refactoring to extract managed object length determination.
This makes the tiny, managed, and huge code work alike.
Tested on:
jam (very minor change, no need for full commit test)
|
| |
|
|
|
|
|
|
|
| |
Minor change to consolidate odd bitmasking due to an old bugfix.
Tested on:
jam (very minor change)
|
|
|
|
|
|
|
|
| |
Clean up file driver property handling, along with some misc. warnings.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug, gcc 4.7.1, C++, FORTRAN & threadsafe
(h5committest upcoming)
|
|
|
|
|
|
| |
to get object type instead of H5O_get_info(...TRUE...) saving time in traversing metadata.
(This is part of the checkin to 1.8 branch #22677)
|
|
|
|
| |
H5FD_FLMAP_SINGLE to H5FD_FLMAP_DICHOTOMY.
|
| |
|
|
|
|
|
|
|
|
|
| |
Have free space manager use temporary address space for storing the
section info, until the file is flushed or closed.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug, gcc 4.7.x, C++, FORTRAN & threadsafe
(h5committest forthcoming)
|
|
|
|
|
|
|
|
|
|
| |
Merge "flush me last" & "flush me collectively" feature from the
avoid_truncate branch back to the trunk. (To help enable allowing the
free space section info to reside in temporary address space)
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug, gcc 4.7.x, C++, FORTRAN & threadsafe
(h5committest forthcoming)
|
|
|
|
|
|
|
|
| |
of return
value of FAIL. The function type is a pointer. Fixed it by changing it to NULL.
Tested: In remote machine ADA (aix).
|
|
|
|
|
|
|
|
|
| |
Changes resulting from Klocwork static analysis tool, from Mark Miller
@ LLNL (miller86@llnl.gov).
Tested on:
Mac OS X/64 10.7.4 (amazon) w/debug, C++ & FORTRAN, using gcc 4.7.x
(too minor to require h5committest)
|
| |
|
|
|
|
|
|
|
|
|
| |
Merge some of the changes on the plist_encode_decode branch back to the
trunk.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug
(Too minor to require h5committest)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Switch propert list/class iteration from internal to external form of
iteration, cleaning up and simplifying the code a bit.
Bring other general improvements from plist_encode_decode branch back to
trunk.
Clean up many warnings.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/gcc 4.7, debug and C++ & FORTRAN
(too minor to require h5committest)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Description:
When jumping out from between H5_BEGIN_TAG and H5_END_TAG macros using
HGOTO_ERROR or HGOTO_DONE, the previous metadata tag is not reset on the dxpl.
This could cause problems when, for example, calling H5Ocopy within an
H5Literate callback.
Added new HGOTO_ERROR_TAG and HGOTO_DONE_TAG macros which must be used in place
of the above between H5_BEGIN_TAG and H5_END_TAG.
Tested:
jam, koala, ostrich (h5committest), durandal
|
|
|
|
|
|
|
|
|
| |
Refactor ID class registration to be more like how other "class"s in the
library are registered, and clean up compiler warnings.
Tested on:
Mac OSX/64 10.7.4 (amazon) w/debug
(Too minor to require h5comittest)
|
|
|
|
|
|
| |
fill value. The problem happened in H5P_get_fill_value. I believe it's because the background buffer wasn't big enough. So I corrected it in this checkin.
Tested on jam and koala - simple change.
|
| |
|
|
|
|
|
|
|
|
| |
match what the RM says about the function behavior.
Quincey added it some time ago without realizing it caused test errors (Issue 8072).
Tested on ostrich, MacGoblin - simple change.
|
| |
|
|
|
|
| |
Tested: windows - mingw
|
| |
|
| |
|
|
|
|
|
|
| |
Correct HAVE_WINSOCK name.
Tested: windows
|
|
|
|
|
|
|
| |
(2) fix to extend into the aggregator based on threshold
(3) fix for the "flavor" logging problem
(4) fix for test failures due to the above fixes
(5) bug fix for h5stat when h5stat against a non-existing file
|
|
|
|
|
|
|
|
| |
Correct misspelled debugging #ifdef and remove unused variable.
Tested on:
Mac OS X/64 10.7.4 (amazon) w/serial
(too minor to require h5committest)
|
|
|
|
|
|
|
|
| |
tests and options: DEBUGPKG not addressed.
Still needed: compiler wrapper scripts
Tested: local linux(cmake) and h5committest
|
| |
|
| |
|
|
|
|
|
|
| |
for windows when configured with RelWithDebInfo
Tested: local linux and Windows
|
| |
|
|
|
|
|
| |
remove an un-needed bool
|
|
|
|
|
|
|
| |
DXPL so that we don't have to insert and remove them everytime we
perform collective I/O.
Test with h5committest
|
|
|
|
|
|
|
|
| |
same property in property list multiple times.
fix that bug.
test with h5committest
|
|
|
|
|
|
| |
Minor change to error message text in H5R.c.
Minor change: tested on jam
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Change name of MPE log file produced from "cpilog" (which was the name
used in the example code we were referencing) to "h5log".
Tested on:
None, too trivial, eyeballed only...
|
|
|
|
|
|
|
|
| |
Correct error reported by user (Martin Otte) where we weren't using
realloc'ed buffer in MPI datatype code.
Tested on:
Mac OSX/64 10.7.3 (amazon) w/parallel
|
| |
|
|
|
|
|
|
|
|
| |
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
|
| |
|
| |
|