The static szip libraries are working as intended, but when shared szip
libraries are present and intended for use they cannot be opened. A check
has been implemented to test if shared szip libraries are functional on
current platform. If they are not, and the user is trying to use them,
szip is now disbaled and the user is informed.
This issue occurs on kagiso, and the resolution was tested on kagiso.
Make H5O_protect/H5O_unprotect use pinned cache entries instead of
protected entries (allowing the object header routines to be semi-reentrant).
Tested on:
FreeBSD 6.2 (duty)
Tested platform:
Kagiso only since it is only a comment block change. If it works in one
machine, it should work in all, I hope. Still need to check the parallel
build on copper.
More cleanups on the object header message handling code, to make it
easier to work with and move forward on the creation order coding.
Various other minor cleanups & warning fixes.
Tested on:
FreeBSD/32 6.2 (duty)
Move datatype messages into new shared message method interface
Clean up various formatting & compiler warnings
Minor optimizations
Tested on:
Linux/32 2.6 (chicago)
Linux/64 2.6 (chicago2)
Move fill value messages into new shared message method calling scheme.
Only update an opened dataset's fill value property if it's different from
the default fill value settings.
Tested on:
Linux/32 2.6 (chicago)
Linux/64 2.6 (chicago2)
Get rid of two different types of fill value struct (merge H5O_fill_t
and H5O_fill_new_t) and clean up & simplify dataset initialization code.
(In preparation for shared object header message method call
refactoring).
Tested on:
FreeBSD/32 6.2 (duty)
Mac OS X/32 10.4.8 (amazon)
Add "set_crt_index" and "get_crt_index" methods for the object header
message class.
Unify fractal heap definitions for shared messages and attributes, under
"object header" fractal heap definitions.
Initial code for adding creation order index to object header messages.
Tested on:
Linux/32 2.6 (chicago)
Linux/64 2.6 (chicago2)
Migrate more object header routines to use the H5O_msg_ prefix and put
them into the src/H5Omessage.c code module.
Tested on:
Mac OS X/32 10.4.8 (amazon)
FreeBSD/32 4.11 (sleipnir)
Linux/32 2.4 (heping)
AIX/32 5.? (copper)
Refactor object header code to separate process of creating an object
header message from the process of writing to an existing one.
Start renaming operations that deal with object header messages to have
"H5O_msg_" prefix...
Tested on:
Mac OS X/32 10.4.8 (amazon)
FreeBSD/32 4.11 (sleipnir)
Linux/32 2.4 (heping)
AIX/32 5.? (copper)
Rename too generic "H5O_FLAG_<foo>" flags that are used as flags for
header messages to "H5O_MSG_FLAG_<foo>", in order to make their use clearer.
Tested on:
FreeBSD/32 4.11 (sleipnir)
Linux/32 2.4 (heping)
Linux/64 2.4 (mir)
Propagate object creation properties up into group, dataset and named
datatype property lists, when those property lists are retrieved for
existing objects in a file.
Also, add H5Tget_create_plist() API routine, to allow named datatype
property lists to be retrieved for named datatypes.
Tested on:
FreeBSD/32 4.11 (sleipnir)
Linux/32 2.4 (heping)
Linux/64 2.4 (mir)
AIX/32 5.? (copper)
Refactor internal code that retrieves the object's type to use the new
H5O_type_t (which only definitions for object types, instead of links as well)
instead of the older H5G_obj_t (which included links, etc.)
Tested on:
Linux/32 2.6 (chicago)
Linux/64 2.6 (chicago2)
This feature is still in progress; Shared Object Header Messages are not
complete as a feature and are not thoroughly tested. There are still
"TODO" comments in the code (comments with the word "JAMES" in them,
so as not to be confused with other TODO comments).
Hopefully this checkin will reduce the liklihood of conflicts as I finish
implementing this feature.
All current tests pass on juniper, copper (parallel), heping, kagiso, and mir.
Refactor generic property list initialization code to put property list
specific routines in property list modules, instead of scattered to the four
winds. Also, introduce property list class initialization objects, to make
adding new property list classes in the library easier.
Fix daily test failure by using H5Pget_elink_prefix() API routine instead
of looking at the "raw" generic property list information.
Tested on:
Mac OS X/32 10.4.8 (amazon)
FreeBSD/32 4.11 (sleipnir) w/threadsafe
Linux/32 2.4 (heping) w/C++ & FORTRAN
Linux/64 2.4 (mir) w/build-all & 1.6 compat
Move object header creation past several internal possibilities for
errors, making the routine more resistant to errors encountered.
Also, various formatting/compiler warning cleanups.
Tested on:
Linux/32 2.6 (chicago)
Linux/64 2.6 (chicago2)
Clean up some object header code in advance of more major changes in the
near future.
Fix small initialization error in attribute message name buffer copy.
Tweak down the default # of messages for new object headers.
Tested on:
FreeBSD/32 4.11 (sleipnir) w/threadsafe
Linux/32 2.4 (heping) w/FORTRAN & C++
Mac OS/32 10.4.8 (amazon)
Update datasets and the layout, attributes and fill-value
object header messages to use the latest version
of the file format flag.
Tested on:
FreeBSD 4.11 (sleipnir)
Linux/64 2.4 (mir)
Linux/32 2.4 (heping)
Change call from H5O_loc_free() to H5G_name_free() when certain dataset
open failures are generated.
Tested on:
Linux/32 2.6 (chicago)
Too minor to require h5committest
Object header locations (H5O_loc_t's) can now "hold open" a file and
decrement its open object count when they close. This means that
locations (H5O_loc_t's and H5G_loc_t's) should always be freed.
Added more thorough tests to ensure that external files are closed.
1. Provide another option for users to do independent IO with MPI file setview(collectively)
2. With the request of collective IO from users, using Independent IO with MPI file setview if we find collective IO is not good for the applications for IO per chunk(multi-chunk IO) case. Previously we used pure independent IO and that actually performed small IO(IO each row) for this case. The recent performance study suggested the independent IO with file setview can acheieve significantly better performance than collective IO when not many processes participate in the IO.
3. For applications that explicitly choose to do collective IO per chunk case, the library won't do any optimization(gather/broadcast) operations. The library simply passes the collective IO request to MPI-IO.
Tested at copper, kagiso, heping, mir and tungsten(cmpi and mpich)
Kagiso is using LAM, t_mpi test was broken even.
The cchunk10 test failed at heping and mir. I suspected it was an MPICH problem. Will investigate later.
Everything passed at copper.
at tungsten: the old cmpi bug(failed at esetw) is still there. Other tests passed.
Some sequential fheap tests failed at kagiso.
Users can create external links using H5L_create_external(). These links
point to an object in another HDF5 file. Users can alter the behavior of
external links or create new kinds of links by registering callbacks
using the H5L interface.
Added tests, tools support, etc.
Also a number of other, minor changes have been made (some restructuring of
the H5L interface, for instance).
Additional documentation and examples are forthcoming.
Feature
Description:
Revised Link APIs.
Solution:
New link APIs use H5L*
H5*create_expand do not create links to the objects created; this must
be done manually with H5Llink.
Added APIs to link an object given its ID (H5Llink), to copy links (H5Lcopy),
and changed creation APIs (H5Lcreate_hard and H5Lcreate_soft) and query
API (H5Lget_linkinfo instead of H5Gget_objinfo).
All old APIs are still supported in H5Gdeprec.c .
Platforms tested:
sol, mir, copper
Misc. update:
Forgot to update MANIFEST and release docs. Will do after checkin.
Bug #281
Description:
The previous fix for bug #281 was not safe. Files got corrupted in
some cases. Needs a more comprehensive fix for it. Undo the unsafe
fix for now.
Platforms tested:
h5committested.
Code cleanup
Description:
Trim trailing whitespace in Makefile.am and C/C++ source files to make
diffing changes easier.
Platforms tested:
None necessary, whitespace only change
Code checkpoint
Description:
Checkpoint fractal heap improvements, as well as move the free space
manager code that it's using into a separate package.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Linux 2.4/64 (mir) w/C++ & FORTRAN
Linux 2.4/32 (heping)
Solaris 2.9 (shanti)
AIX 5.? (copper) w/FORTRAN & parallel
New APIs to add for collective chunk IO
Description:
Three new APIs
H5Pset_dxpl_mpio_chunk_opt_ratio
H5Pset_dxpl_mpio_chunk_opt_num
H5Pset_dxpl_mpio_chunk_opt
for optional optimization choices from users.
Solution:
Haven't added tests yet, won't affect other parts of the library.
Will add tests after urgent investigations of memory leaking problems from NASA Aura team.
Platforms tested:
heping: both parallel and sequential
shanti
Misc. update:
Bug fix
Description:
Retrieving an object's name could fail (in various ways) under certain
circumstances (mostly having to do with mounted files).
Solution:
Re-write & simplify "get object name" code to fix error in a better way
than adding yet another hack to the previous pile of hacks... :-)
Platforms tested:
FreeBSD 4.11 (sleipnir)
h5committest
New feature
Description:
Add in a combination of Peter's & my code to support copying
variable-length data from one file to another, although currently only
supported with contiguous data storage.
Platforms tested:
FreeBSD 4.11 (sleipnir)
h5committest
New feature
Description:
Check in baseline for compact group revisions, which radically revises the
source code for managing groups and object headers.
WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!!
WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!!
This initiates the "unstable" phase of the 1.7.x branch, leading up
to the 1.8.0 release. Please test this code, but do _NOT_ keep files created
with it - the format will change again before the release and you will not
be able to read your old files!!!
WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!!
WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!! WARNING!!!!
Solution:
There's too many changes to really describe them all, but some of them
include:
- Stop abusing the H5G_entry_t structure and split it into two separate
structures for non-symbol table node use within the library: H5O_loc_t
for object locations in a file and H5G_name_t to store the path to
an opened object. H5G_entry_t is now only used for storing symbol
table entries on disk.
- Retire H5G_namei() in favor of a more general mechanism for traversing
group paths and issuing callbacks on objects located. This gets us out
of the business of hacking H5G_namei() for new features, generally.
- Revised H5O* routines to take a H5O_loc_t instead of H5G_entry_t
- Lots more...
Platforms tested:
h5committested and maybe another dozen configurations.... :-)
New feature
Description:
Add in baseline "object copy" code from Peter [in the form of a new API
routine: H5Gcopy()]. There's still some work to do (like handling variable-
length datatypes and possibly support for references) and it hasn't been tested
on mounted files yet, but the core functionality is there and working
correctly.
I've also got a set of patches to update the 1.6 branch with tweaks to
keep the branches mostly in sync, but Elena will kill me if I import them
before the 1.6.5 release is out... :-)
Platforms tested:
FreeBSD 4.11 (sleipnir)
h5committested
Code cleanup
Description:
Clean up H5Dcreate a little, to prepare for coding standards discussion
today.
Platforms tested:
Just eyeballed, too trivial to require testing.
Code cleanup
Description:
Clean up & standardize a bit in preparation for coding standards
discussion.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Too minor to require h5committest
Bug fix
Description:
Fix "double root open" bug and apply same fix to datasets & named datatypes.
Also fix bug in symbol table entry lookup code that would "transfer" an
object from one file handle to another file handle when the two file handles
were opened on the same actual file.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Too minor to require h5committest
Repair synchronization bug in the metadata cache in PHDF5
Also repair numerous other bugs that surfaced in testing the
bug fix.
Description:
While operations modifying metadata must be collective, we allow
independant reads. This allows metadata caches on different processes
to adjust to different sizes, and to place the entries on their dirty
lists in different orders. Since only process 0 actually writes
metadata to disk (all other processes thought they did, but the writes
were discarded on the theory that they had to be collective), this made
it possible for another process to modify metadata, flush it, and then
read it back in in its original form (pre-modification) form. The
possibilities for file corruption should be obvious.
Solution:
Make the policy that only process 0 can write to file explicit, and
visible to the metadata caches. Thus only process 0 may flush dirty
entries -- all other caches must retain dirty entries until they are
informed by process 0 that the entries are clean.
Synchronization is handled by counting the bytes of dirty cache entries
created, and then synching up between the caches whenever the sum
exceeds an (eventually user specified) limit. Dirty metadata creation
is consistent across all processes because all operations modifying
metadata must be collective.
This change uncovered may bugs which are repaired in this checkin.
It also required modification of H5HL and H5O to allocate file space
on insertion rather than on flush from cache.
Platforms tested:
H5committest, heping(parallel & serial)
Misc. update:
Code cleanup/reorganization
Description:
Merge back some more changes extracted from the "compact group" set.
This bunch cleans up and prepares the H5G_* routines for eventual import of
new features.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Linux 2.4
Mac OS X.4
Code cleanup
Description:
Merge back changes from "compact group" work that improve the
infrastructure of the library and may impact others. In this round of
merging, that includes:
- Move datatype allocation into single internal routine, instead of
duplicated code that was spread out in a dozen or so places.
- Clean up guts of object header routines (H5O_*) to allow for some of
the fancieroperations that need to be performed on groups, along with
some general improvements.
- Added a new error code
- Some minor cleanups in other code....
Platforms tested:
FreeBSD 4.11 (sleipnir)
Linux 2.4
Mac OS X
Bug #281
Description:
A dataset created in serial mode with H5D_ALLOC_TIME_INCR allocation setting
was not extendible, either explicitly by H5Dextend or implicitly by writing
to unallocated chunks. This was because parallel mode expects the allocation
mode be H5D_ALLOC_TIME_INCR only.
Solution:
Modified library to allocate more space when needed or directed if the
file is opened by parallel mode, independent of what the dataset allocation
mode is.
Platforms tested:
Heping pp.
Code cleanup
Description:
Trim trailing whitespace, which is making 'diff'ing the two branches
difficult.
Solution:
Ran this script in each directory:
foreach f (*.[ch] *.cpp)
sed 's/[[:blank:]]*$//' $f > sed.out && mv sed.out $f
end
Platforms tested:
FreeBSD 4.11 (sleipnir)
Too minor to require h5committest
Bug fix
Description:
Rewrite code for mounting files to clean up layers of kludges and implement
a much cleaner and more maintainable design.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Linux 2.4
Bug fix
Description:
Correct memory leak when a dataset is attempted to be opened, but turns
out to be a group or named datatype.
Also, clean up code that was leading to the leak and zero out empty
group entries to help prevent similar errors in the future.
Platforms tested:
FreebSD 4.11 (sleipnir)
Too minor to require h5committest
Bug fix
Description:
Correct error where the dataset sieve buffer was inadvertantly disabled,
resulting in poor raw data I/O performance for certain I/O patterns.
Solution:
Hook dataset sieve buffer size setting back up.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Solaris 2.9 (shanti)
Code cleanup
Description:
Refactor metadata cache to merge "dirtied" flag in with other flags for
H5AC_unprotect and H5C_unprotect.
Platforms tested:
FreeBSD 4.11 (sleipnir)
h5committest
Interim checkin of code changes moving management of the is_dirty flag
into the cache code.
Description:
Prior to this checkin, management of the is_dirty flag was handled
above the level of the metadata cache. This can no longer be allowed,
as it introduces a race condition in the proposed fix for a cache
coherency bug in PHDF5.
Solution:
Move management fo the is_dirty flag to the cache code proper.
Entries are now marked as dirty via a flag on the unprotect call.
Platforms tested:
h5committested
Misc. update:
New feature
Description:
Add group creation & access property lists, dataset access property lists
and named datatype creation & access property lists. Currently have
<foo>_extend() API names, which will need to be changed for the final release.
Platforms tested:
FreeBSD 4.11 (sleipnir)
Linux 2.4 (heping)
