Updated the H5FS cache code to grab the correct tag and modifiedthe freespace test to use dxpls that have been tagged with
the H5AC__FREESPACE_TAG global tag instead of H5AC_ind_read_dxpl_id.
The library code now expects the owner of the free space manager
to tag it so the owner-less free space managers in the freespace
tag had to be tagged with *something* to avoid cache errors.
Merge pull request #186 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:icc_config_fixes to developCleaned up the Intel compiler flags (Autotools).
* commit '1c5029f21b51977d26b26fd3540065f36d97c683':
Updated the icc flags (C flags only).
Merge pull request #181 in HDFFV/hdf5 from ~BYRN/hdf5_adb:develop to develop* commit '53d3a619f8a076cb1d0ba9cbc207fc07a6fcc5d8':
Change examples to match junit test
Add missing character
Change where test copies libs because of osx1011 SIP
Merge pull request #185 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop_swmr_merge_03 to develop* commit '527f28333cb2ec5b247ec348cf6970e0d114891b':
Another Java oversight (sorry, don't have Java configured on my Mac)
Merge pull request #184 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop_swmr_merge_02 to developMerged minor fixes from revise_chunks.
* commit 'e1026bd354443ae9c4e0cf197feb7191920c9f13':
Correct oversight in Java test and remove direct VFD from SWMR supported drivers.
Merge pull request #183 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop_swmr_merge to developMerged the final library SWMR changes from revise_chunks.
* commit '15e04901a0cea65b2b0fa3ca869341a2b656bfdf':
Bring SWMR support in to the main development branch. (Finally!) More tests and the tool and API wrappers will be coming in over the weekend.
Add some Java methods that were missed earlier.
Minor normalization cleanups.
Merge pull request #182 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:vs2012_round to developMerged fix for missing VS2012 round commands.
* commit '488f19a9cbd2e89e83847a628e59f57f4293fed0':
Fixed a comment.
Added Windows-only versions of the round() functions, which do not exist in VS2012 and earlier.
Updated the cache_tagging test so it correctly analyzes the new fractal heap free space manager behavior.Also converted the debug output of the cache_tagging
test to use the revised H5AC_dump_cache() function.
Added TAG macros to the H5MF package code so H5AC__FREESPACE_TAGgets set on the dxpl correctly. Also restored the TAG macro
to H5FS__cache_hdr_pre_serialize() since otherwise the tag
won't be initialized.
First cut at fixing the tagging issue with the free space managers. * Converted the tag macros to regular ones so the parent's tag is used via the dxpl.* Updated the tag sanity check logic so for free space managers.
* Turned new-style group checks back on in test/evict_on_close.c
Merge pull request #179 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:eoc_valgrind_bugfix to developFix for some evict-on-close issues revealed by Valgrind.
* commit 'd3dbe9d90ee71b231b3264e1d62d3efbae3242ef':
Tentative fix for valgrind issues related to EoC.
Merge pull request #178 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:expanded_eoc_test to developUpdated the evict-on-close test to better exercise groups.
* commit '30bd7247d80f08aee4fc7038df86c43258344de8':
Updated the evict-on-close test to create a large number of subgroups in a containing group. This will ensure that many cache entries will be used, making a better test of EoC functionality.
Merge pull request #177 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop_merge_revise_chunks_10 to develop* commit 'b092dbcdfdfc7477596ae49f816f18e0dadf0fb1':
Bring over another batch (hopefully the last) of non-SWMR "normalization" changes from the revise_chunks branch.
Merge pull request #175 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:new_H5AC_dump_cache to developUpdated H5AC_dump_cache() to emit more information.
* commit '7a8c7c6063e328fdac28aa3f77ead4bdc91860e5':
Updated the cache dump to use the stored type name instead of the (redundant) array of names I previously created.
Updated the format of H5AC/C_dump_cache() to include more information and be easier to read.
Merge pull request #176 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop to develop* commit '2cb2d5a533c2de7503014da3250c3088e1ba089d':
Remove "collective write list" parameter from H5C__flush_single_entry() as it is only used in a couple of places. The collective write list has been moved to be internal to the cache data structure instead.
Eliminate unnecessary data structure for parallel collective metadata cache I/O
Merge pull request #171 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:develop_memfix to developFix for tfile.c valgrind issues.
* commit '191cc19ae0a2feb72d2171a0808fdc71a3b862dd':
Added vlen reclaim functions to a test in tfile.c that leaks memory.
Merge pull request #173 in HDFFV/hdf5 from ~DEROBINS/hdf5_der:gitignore to developUpdated .gitignore to be more manageable and future-proof.
* commit '1aa2373b2e1a651bec507dcb14a89b51815a972c':
Simplified .gitignore by using patterns.
Added src/H5config.h.in~ to .gitignore.
Remove "collective write list" parameter from H5C__flush_single_entry() as itis only used in a couple of places. The collective write list has been moved
to be internal to the cache data structure instead.
Updated the evict-on-close test to create a large number of subgroupsin a containing group. This will ensure that many cache entries will
be used, making a better test of EoC functionality.
Tentative fix for valgrind issues related to EoC.Adds /*out*/ parameters to H5O_close() and H5F_try_close() so
that H5D/G_close() will know when H5O_close() has triggered a
file close and thus the file struct is not reliable.
Also removes the H5F_CLOSING() macro and related which were
formerly used to check if the file was closing.
Merge pull request #172 in HDFFV/hdf5 from ~KOZIOL/hdf5:develop to develop* commit 'ea4d18c3d9c3b6c77030abdf2f0509bc68224559':
Remove unnecessary assert.
Add a flag for H5C__flush_single_entry() to allow the parallel code to request that images for entries be generated, even when not writing the entry from an MPI rank.