Feature change & code cleanup
Description:
Changed fileno field in H5FD_t struct from 'unsigned long[2]' to just
'unsigned long'.
Changed over code to handle the fileno change.
Chased error API changes
Platforms tested:
h5committested
Bug fix.
Description:
Corrected fencepost error which was incorrectly checking for merging
metadata at the front of the metadata accumulator. This now allows the
metadata cache to cache I/O operations more efficiently.
Platforms tested:
FreeBSD 4.8 (sleipnir) w/C++
Linux 2.4 (burrwhite) w/FORTRAN
Solaris 2.7 (arabica) w/FORTRAN
IRIX64 6.5 (modi4) w/parallel & FORTRAN
(h5committest not run due to my ongoing difficulties with C++ on burrwhite).
Bug fix
Description:
This fixes a bug in the low-level metadata caching code in the library
which could possibly lose metadata during file I/O when a lot of objects are
inserted into a group.
This also fixes a couple of (similar) fencepost bugs in the B-tree
deletion code.
Solution:
For the metadata bug - call the low-level driver's 'write' routine instead
of H5FD_write.
For the B-tree bug - include the correct number of keys.
Platforms tested:
FreeBSD 4.8 (sleipnir) w/C++
Linux 2.4 (burrwhite) w/FORTRAN
Solaris 2.7 (arabica) w/FORTRAN
IRIX64 6.5 (modi4) w/FORTRAN & parallel
(h5committest is still not working for me on burrwhite)
Misc. update:
Update & Bug Fix
Description:
The "free" protocol was missing. Added that to the server side.
When doing a "create" of a file (with no other data structures
created), the freespace in the file wasn't being reclaimed.
Solution:
After adding the free protocol, we put the burden of running through
the FD_free function on the SAP instead of each client.
Platforms tested:
Linux
Misc. update:
Bug Fix & Update
Description:
FPHDF5 was creating files which didn't have the EOA field in the
superblock set correctly. It turns out that the SAP was keeping this
information to itself instead of giving it to the client processes.
Naughty SAP!
Solution:
Have the SAP send this information back to the clients so that they
can update the superblock as necessary. This now creates a file (with
just the root group) that looks correct! Only problem is that there's
extra file space being allocated. Also, at program termination,
there's an infinite loop...
Platforms tested:
H5committests (run by hand on burrwhite, arabica, and modi4)
Misc. update:
Bug fix and Update
Description:
From Quincey's comments on the code I checked in last night:
- In H5F_close call, the "private" processes should call the
H5F_flush with the "CLEAR_ONLY" flag.
- There's no need for a special case for FPHDF5 in the
FD_real_alloc function since FPHDF5 doesn't define an alloc
function.
- The return type of H5Pset_fapl_fphdf5 should be herr_t instead
of hid_t. I don't know how it got that way in the first place.
- The variable names for MPI types and the structure typedefs
should be switched: H5FP_request/H5FP_request_t to
H5FP_request_t/H5FP_request and so on.
- In the H5FP.c module, I was commiting the H5FP_request MPI
datatype but using the wrong offset field...
Platforms tested:
Linux...will test on others, but these are mostly FPHDF5 changes.
Misc. update:
Update
Description:
When doing an allocation, check to see, if it's an FPHDF5
driver, if only the captain should be performing the allocation
Platforms tested:
Linux, Modi4, Sol
Misc. update:
Finish code cleanup
Description:
Wrap up the conversion of H5F_flush's multiple boolean flags into a single
bitfield of flags by pushing the flags down into the H5AC_flush and
H5F_istore_flush routines.
Also, changed the flags from H5_FLUSH_<foo> to H5F_FLUSH_<foo> to be more
consistent with rest of library.
And reverted the changes to H5FDflush and H5FD_flush routines.
Platforms tested:
FreeBSD 4.7 (sleipnir)
Solaris 5.8 (sol)
IRIX64 6.5 (modi4) w/parallel
Misc. update:
Refactoring of Flush Logic
Description:
The Flushing logic passed in multiple flags to indicate what type of
flush it was: closing, invalidate, alloc only. This made extending
the function of the flush logic to handle other flags annoying.
Solution:
I changed it to be just one bitmasked flag to indicate what type of
flushing to do. I also added the CLEAR_ONLY flag, which will be used
in the FPHDF5 stuff.
Platforms tested:
h5committest doesn't work for me (my environment isn't setup on the
other machines I guess). I tested it manually:
Linux parallel & C++
Sol Fortran
Modi4 parallel & Fortran.
Update
Description:
Big change to the H5FD.c module:
- Split apart H5FD_alloc and H5FD_free. H5FD_alloc was huge and
H5FD_free had a freeing of the freelist part which I needed to
call from the SAP.
- Added support for FPHDF5. If it's a client, then it sends the
allocation or free request to the SAP. The SAP will call the
same code, but it'll actually do the allocation/freeing in that
case.
Platforms tested:
Linux & Modi4
Bug Fix
Description:
Metadata cache in parallel I/O can cause hangs in applications which
perform independent I/O on chunked datasets, because the metadata cache
can attempt to flush out dirty metadata from only a single process, instead
of collectively from all processes.
Solution:
Pass a dataset transfer property list down from every API function which
could possibly trigger metadata I/O.
Then, split the metadata cache into two sets of entries to allow dirty
metadata to be set aside when a hash table collision occurs during
independent I/O.
Platforms tested:
Tested h5committest {arabica (fortran), eirene (fortran, C++)
modi4 (parallel, fortran)}
FreeBSD 4.7 (sleipnir) serial & parallel
Misc. update:
Updated release_docs/RELEASE
Code cleanup/new feature.
Description:
Split FUNC_LEAVE into API and non-API specific versions. This allows a
solution to compiling this branch with C++, as well as reducing the size
of the binaries produced.
Platforms tested:
FreeBSD 4.7 (sleipnir) w/serial, parallel (including MPE) & thread-safe
Lots of performance improvements & a couple new internal API interfaces.
Description:
Performance Improvements:
- Cached file offset & length sizes in shared file struct, to avoid
constantly looking them up in the FCPL.
- Generic property improvements:
- Added "revision" number to generic property classes to speed
up comparisons.
- Changed method of storing properties from using a hash-table
to the TBBT routines in the library.
- Share the propery names between classes and the lists derived
from them.
- Removed redundant 'def_value' buffer from each property.
- Switching code to use a "copy on write" strategy for
properties in each list, where the properties in each list
are shared with the properties in the class, until a
property's value is changed in a list.
- Fixed error in layout code which was allocating too many buffers.
- Redefined public macros of the form (H5open()/H5check, <variable>)
internally to only be (<variable>), avoiding innumerable useless
calls to H5open() and H5check_version().
- Reuse already zeroed buffers in H5F_contig_fill instead of
constantly re-zeroing them.
- Don't write fill values if writing entire dataset.
- Use gettimeofday() system call instead of time() system when
checking the modification time of a dataset.
- Added reference counted string API and use it for tracking the
names of objects opening in a file (for the ID->name code).
- Removed redundant H5P_get() calls in B-tree routines.
- Redefine H5T datatype macros internally to the library, to avoid
calling H5check redundantly.
- Keep dataspace information for dataset locally instead of reading
from disk each time. Added new module to track open objects
in a file, to allow this (which will be useful eventually for
some FPH5 metadata caching issues).
- Remove H5AC_find macro which was inlining metadata cache lookups,
and call function instead.
- Remove redundant memset() calls from H5G_namei() routine.
- Remove redundant checking of object type when locating objects
in metadata cache and rely on the address only.
- Create default dataset object to use when default dataset creation
property list is used to create datasets, bypassing querying
for all the property list values.
- Use default I/O vector size when performing raw data with the
default dataset transfer property list, instead of querying for
I/O vector size.
- Remove H5P_DEFAULT internally to the library, replacing it with
more specific default property list based on the type of
property list needed.
- Remove redundant memset() calls in object header message (H5O*)
routines.
- Remove redunant memset() calls in data I/O routines.
- Split free-list allocation routines into malloc() and calloc()-
like routines, instead of one combined routine.
- Remove lots of indirection in H5O*() routines.
- Simplify metadata cache entry comparison routine (used when
flushing entire cache out).
- Only enable metadata cache statistics when H5AC_DEBUG is turned
on, instead of always tracking them.
- Simplify address comparison macro (H5F_addr_eq).
- Remove redundant metadata cache entry protections during dataset
creation by protecting the object header once and making all
the modifications necessary for the dataset creation before
unprotecting it.
- Reduce # of "number of element in extent" computations performed
by computing and storing the value during dataspace creation.
- Simplify checking for group location's file information, when file
has not been involving in file-mounting operations.
- Use binary encoding for modification time, instead of ASCII.
- Hoist H5HL_peek calls (to get information in a local heap)
out of loops in many group routine.
- Use static variable for iterators of selections, instead of
dynamically allocation them each time.
- Lookup & insert new entries in one step, avoiding traversing
group's B-tree twice.
- Fixed memory leak in H5Gget_objname_idx() routine (tangential to
performance improvements, but fixed along the way).
- Use free-list for reference counted strings.
- Don't bother copying object names into cached group entries,
since they are re-created when an object is opened.
The benchmark I used to measure these results created several thousand
small (2K) datasets in a file and wrote out the data for them. This is
Elena's "regular.c" benchmark.
These changes resulted in approximately ~4.3x speedup of the
development branch when compared to the previous code in the
development branch and ~1.4x speedup compared to the release
branch.
Additionally, these changes reduce the total memory used (code and
data) by the development branch by ~800KB, bringing the development
branch back into the same ballpark as the release branch.
I'll send out a more detailed description of the benchmark results
as a followup note.
New internal API routines:
Added "reference counted strings" API for tracking strings that get
used by multiple owners without duplicating the strings.
Added "ternary search tree" API for text->object mappings.
Platforms tested:
Tested h5committest {arabica (fortran), eirene (fortran, C++)
modi4 (parallel, fortran)}
Other platforms/configurations tested?
FreeBSD 4.7 (sleipnir) serial & parallel
Solaris 2.6 (baldric) serial
Bug Fix
Description:
Free blocks of space in the file which were merged with a block before them
were not updating the free list information correctly, allowing space in
the file to over-allocated and potentially corrupted.
Solution:
Correct address of free block during merging.
Platforms tested:
Tested h5committest {arabica (fortran), eirene (fortran, C++)
modi4 (parallel, fortran)}
FreeBSD 4.7 (sleipnir)
Purpose:
New API functions
Description:
Added API functions to return pointer to low-level file handle
(H5Fget_vfd_handle and H5FDget_vfd_handle) and related property list
setting functions(H5Pset_family_offset and H5Pset_multi_type).
Platforms tested:
Linux 2.2(eirene), Solaris 2.7(arabica), IRIX64 6.5(modi4)
Code cleanup/New Feature
Description:
Improve the space allocation in the file by re-using freed space more
effectively.
Platforms tested:
FreeBSD 4.6 (sleipnir) w/serial & parallel
Code cleanup
Description:
Combined H5P_isa_class and H5I_object functionality into a new internal
H5P API function: H5P_object_verify, which checks that a property list is
the appropriate class and then returns the property list object associated
with the property list ID.
This reduces the source code by about 200 LOC and trims the library binary
some more.
Platforms tested:
FreeBSD 4.6 (sleipnir)
Code cleanup
Description:
Changed the last HRETURN* statements in the FUNC_ENTER macros into HGOTO*
macros, which reduces the size of the library binary in certain
configurations by another 10%
Platforms tested:
FreeBSD 4.6 (sleipnir) serial & parallel, IRIX64 6.5 (modi4) serial &
parallel
Code cleanup
Description:
Change most (all?) HRETURN_ERROR macros to HGOTO_ERROR macros, along with
HRETURN macros to HGOTO_DONE macros. This unifies the error return path
from functions and reduces the size of the library by up to 10% on some
platforms.
Additionally, I improved a lot of the error cleanup code in many routines.
Platforms tested:
FreeBSD 4.6 (sleipnir) serial & parallel and IRIX64 6.5 (modi4) serial &
parallel.
Code cleanup
Description:
Created a new H5I function which combined the some of the functionality of
H5I_get_type and H5I_object: H5I_object_verify.
Using this new function in the library trims another ~200 lines of code off
the library and makes the resulting binaries smaller and faster also.
Platforms tested:
FreeBSD 4.6 (sleipnir)
New feature.
Description:
Added MPI-posix VFL driver. This driver uses MPI to coordinate actions, but
performs I/O directly with posix sec(2) I/O functions. This driver should
_NOT_ be used if the file accessed is not on a parallel filesystem.
Platforms tested:
FreeBSD 4.6 (sleipnir) w/parallel & IRIX64 6.5 (modi4) w/parallel
Code improvement
Description:
Some small code cleanups and took out the code the was turning off the
metadata cache for parallel I/O (!)
Platforms tested:
IRIX64 6.5 (modi4) w/parallel
Bug fix.
Description:
H5P_get was reading a 'hsize_t' field into a 'size_t' variable, causing
problems on the HP's and Windows.
Solution:
Changed the variable to be an hsize_t
Platforms tested:
HP/UX 11.0 (kelgia)
Bug fix.
Description:
Added extra check to locations where metadata and "small data" blocks were
being returned to the file's list of free space to avoid freeing 0-sized
blocks. This should fix the problems in last night's daily tests.
Platforms tested:
FreeBSD 4.5 (sleipnir)
Code Improvement
Description:
Split the metadata accumulator code into two parts: one for allowing writes
of the accumulator buffer during reads (when the buffer is dirty and needs
to be flushed to disk in order to hold the new metadata being read in) and
another for only allowing writes of the buffer during writes.
This allows the MPI-I/O VFL driver to use the metadata accumulator (but
only during writes) and benefit from the reduced number of metadata I/O
operations that it brings.
Platforms tested:
IRIX64 6.5 (modi4) w/parallel
Code cleanup/bug fix
Description:
Unused space in the file was being "dropped on the floor" when the metadata
or "small data" block was moved to a new location in the file and there was
unused space in the old block.
Solution:
Put the space left in the allocation block into the free list of space for
the file. This allows it to be [potentially] reused and sometimes allows
the file to be smaller.
Platforms tested:
Solaris 2.7 (arabica) w/FORTRAN and FreeBSD 4.5 (sleipnir) w/C++
New feature.
Description:
Added a "small data" block allocation mechanism to the library, similar to
the mechanism used for allocating metadata currently.
See the RFC for more details:
http://hdf.ncsa.uiuc.edu/RFC/SmallData/SmallData.html
This reduces the number of I/O operations which hit the disk for my test
program from 19 to 15 (i.e. from 393 to 15, overall).
Platforms tested:
Solaris 2.7 (arabica) w/FORTRAN and FreeBSD 4.5 (sleipnir) w/C++
Code improvement
Description:
Now that the metadata accumulator is working correctly, the buffer size
tends to get rather large and stay there.
Solution:
Put a throttle on the accumulator buffer to reduce its size back down when
it isn't needed.
Platforms tested:
Solaris 2.7 (arabica) w/FORTRAN & FreeBSD 4.5 (sleipnir) w/C++
Code improvement
Description:
The metadata aggregation code in the library was not terribly smart about
extending contiguous regions of metadata in the file and would not extend
them as far as possible. This causes space in the file to be wasted, also.
Solution:
Be smarter about extending the space used in the file for metadata by
checking whether new metadata blocks allocated in the file are at the end
of the current metadata aggregation region and append them to the metadata
region if so. This has the nice side benefit of reducing the number of
bytes we waste in the file and reducing the size of the file by a small
amount in some cases.
This reduces the number of I/O operations which hit the disk for my test
program from 53 to 19 (i.e. from 393 to 19, overall).
Platforms tested:
Solaris 2.7 (arabica) w/FORTRAN and FreeBSD 4.5 (sleipnir) w/C++
Code cleanup/bug fix
Description:
The "metadata accumulator" cache in the library (which is designed to catch
small metadata writes/reads and bundle them together into larger I/O
buffers) was incorrectly detecting the important case of metadata pieces
being written sequentially to the file, adjoining but not overlapping.
Additionally, the metadata accumulator was not being used to cache data
read in from disk, only caching writes.
Solution:
Fix accumulator to correctly cache adjoining metadata writes and also to
cache metadata read from disk.
Between these two fixes, the number of I/O requests which resulted in actual
reads/writes to the filesystem dropped from 393 requests to 82 for the
particular test I was using. :-)
Platforms tested:
Solaris 2.7 (arabica) w/FORTRAN & FreeBSD 4.5 (sleipnir) w/C++
Code cleanup
Description:
Broke the FUNC_ENTER macro into several macros, with more specialized
uses (which followup mail will describe). This was designed to move
most/all of the checks which could be done at compile time to that point,
instead of needlessly performing them (over & over :-) at run-time.
This reduces the library's size (and thus staticly linked binaries) and
has a minor speedup effect also.
Platforms tested:
IRIX64 6.5 (modi4) with parallel & FORTRAN enabled, and additional testing
on FreeBSD and Solaris immediately after the checkin.
Code cleanup.
Description:
Took Robb's recent ideas for improving the FUNC_ENTER/FUNC_LEAVE macros
equivalents in the SAF library and adapted them to our library. I added
an additional macro which is equivalent to FUNC_ENTER:
FUNC_ENTER_NOINIT - Has the API tracing code, etc. from FUNC_ENTER but
none of the library or interface initialization code. This is to
be used _only_ for static functions and those which explicitly
cannot have the library or interface initialization code enabled
(like the API termination routines, etc.).
This allowed many more of the functions in the library [but not all yet :-(]
to be wrapped with FUNC_ENTER[_NOINIT]/FUNC_LEAVE pairs.
It also reduced the size of the library and executables (by cutting out a
bunch of code which was never executed), I'll e-mail the exact results when
I've finished editing it.
Platforms tested:
IRIX64 6.5 (modi4)
Purpose:
eliminating the compiler warnings in Windows
Solution:
I am eliminating the compiler warnings in Windows.
the last 2 were:
1)
if((oid_list = H5MM_malloc(oid_count*sizeof(hid_t)))==NULL)
HGOTO_ERROR(H5E_RESOURCE, H5E_NOSPACE, NULL,
"memory allocation failed");
D:\disk_w\hdf5\src\H5F.c(2590) : warning C4047: '=' : 'int ' differs in levels
of indirection from 'void *'
on the HGOTO_ERROR call , I replaced the NULL with FAIL
2)
*((float*)d) = *((double*)s);
D:\disk_w\hdf5\src\H5Tconv.c(6426) : warning C4244: '=' : conversion from 'double ' to 'float ', possible loss of data
I added the type cast
*((float*)d) = (float) *((double*)s);
we have now 0 errors 0 warnings on Windows !
Platforms tested:
w2000, linux
New Feature
Description:
The VFL flush function is called immediately before a file is closed.
This can cause duplicate syncronization actions to occur, if the VFL
close function also performs them.
Solution:
Added 'closing' parameter to VFL 'flush' operation. This allows the VFL
flush function to bypass operations that will be duplicated within the VFL
close function.
Additionally, use the 'closing' parameter to bypass calls to MPI_File_sync()
when set. Since MPI_File_close() also syncronizes the file, this avoids
the terrible performance hit taken when calling MPI_File_sync() as the file
is closing.
Platforms tested:
IRIX64 6.5 (modi4)
Bug Fix
Description:
The H5Gget_objinfo() function was not setting the 'fileno' field in the
H5G_stat_t struct passed in.
Solution:
Added a "file serial number" to each file currently open in the library
and put that in the 'fileno' field. If a file is opened twice (with
H5Fopen) and the VFL driver detects that it is the same file (i.e. the
two file structures have the same "shared file info" in the library's
memory structures), they will have the same serial number.
This serial number has two drawbacks:
- If a VFL driver doesn't/can't detect that two calls to H5Fopen with
the same file actually _are_ the same file, each will get a
different serial number
- If the same file is closed and re-opened, the serial number will be
different.
It is be possible to fix the second drawback for many VFL drivers, but it
would be a lot of effort and probably isn't worth it until we've got a
good reason to do it. Dunno if we'll ever be able to fix the first
drawback...
Platforms tested:
FreeBSD 4.5 (sleipnir)
VS: ----------------------------------------------------------------------
Bug Fix
Description:
When file space was returned to the file space free-list for reuse,
occasionally raw data allocations which used space from the free-list
would overlap with the metadata accumulator and get over-written with
the cached information in the accumulator, corrupting the data.
Solution:
Check if the space about to be recycled on the free-list is going to be
used for raw data and also overlaps with the metadata accumulator cache,
avoiding using space that fits those criteria.
This fixes bug #701
Platforms tested:
FreeBSD 4.5 (sleipnir)
Code cleanup
Description:
Windows is generating hundreds of warnings from some of the practices in
the library. Mostly, they are because size_t is 32-bit and hsize_t is
64-bit on Windows and we were carelessly casting the larger values down to
the smaller ones without checking for overflow.
Also, some other small code cleanups,etc.
Solution:
Re-worked some algorithms to eliminate the casts and also added more
overflow checking for assignments and function parameters which needed
casts.
Kent did most of the work, I just went over his changes and fit them into
the the library code a bit better.
Platforms tested:
FreeBSD 4.4 (hawkwind)
Code cleanup
Description:
Get rid of IDs from internal function calls and some small cleanups from
the old-stype => generic property list conversion.
Platforms tested:
FreeBSD 4.4 (hawkwind)
Kludge
Description:
Since we're only about halfway through converting the internal use of
property lists from the "old way" to the generic property lists, we turned
off snapshots to avoid exposing lots of API changes to users, until the
APIs settled down.
Getting the snapshots rolling again seems to have become a priority, so
some changes are going to have to be made now that were going to be
postponed until we were completely finished with the conversion. This
requires that the old API functions be able to deal with both the old
and new property lists smoothly.
Solution:
Kludge together the property list code so that they can transparently handle
dealing with both the old and new property lists
Platforms tested:
FreeBSD 4.4 (hawkwind)
